A Molecular Basis for the Enrichment of Bifidobacteria in the Infant Gastrointestinal Tract

Transcription

1 A Molecular Basis for the Enrichment of Bifidobacteria in the Infant Gastrointestinal Tract By DANIEL ANTONIO GARRIDO CORTES B.S. (Universidad de Chile, Santiago) 2005 DISSERTATION Submitted in partial satisfaction of the requirements for the degree of DOCTOR OF PHILOSOPHY in FOOD SCIENCE in the OFFICE OF GRADUATE STUDIES of the UNIVERSITY OF CALIFORNIA DAVIS Approved: David A. Mills, Chair Glenn M. Young Helen E. Raybould Committee in Charge 2012! "!

2 ! ""!

3 Daniel Antonio Garrido March 2012 Food Science and Technology A Molecular Basis for the Enrichment of Bifidobacteria in the Infant Gastrointestinal Tract Abstract Breast milk is a complex fluid that has been shaped by evolution to meet all the nutritional requirements of the newborn during the first months of life. In addition, it is a source of an impressive number of bioactive compounds that help in protection and the development of the infant. Among these, human milk oligosaccharides (HMO) are complex substrates that are selectively fermented by beneficial microorganisms found in the infant intestines. These bacteria, mainly belonging to the Bifidobacterium genus, have been shown to utilize these substrates as the sole carbon source, and the genomic sequence of Bifidobacterium infantis ATCC provided a rationale for this phenotype. In this dissertation, several aspects related to the consumption of HMO were studied at the molecular level. Import of HMO was observed to occur associated to membrane-associated Solute Binding Proteins. Several of these proteins showed specific binding to different molecules found in HMO, and genes encoding them were upregulated during bacterial growth on HMO. Other molecular determinants critical for bacterial consumption of these substrates are glycosyl hydrolases. The properties of fucosidases, hexosaminidases and galactosidases found in B. infantis were analyzed at the! """!

4 enzymatic and molecular level. Specific genes encoding these enzymes were found to be induced during growth on HMO, and important differences in substrate affinities were determined. Finally, we studied the interactions of a large panel of infant-gut isolates of bifidobacteria with milk glycoproteins containing N-linked glycans. Certain isolates were able to release glycans from glycoproteins, and discrete genes encoding endo-nacetylglucosaminidases in these microorganisms explained this phenotype. We further characterized the properties of an endoglycosidase from B. infantis, EndoBI-1, shown to have a remarkable glycolytic activity on different types of N-glycans found in host glycoproteins such as milk lactoferrins and immunoglobulins. Together, these studies represent functional evidence and define the molecular adaptations that specific microorganisms associated to the infant gut have developed in response to components in breast milk. Furthermore, a detailed description of this mutualistic relationship between these microbes and their host highlights a unique model that could be translated to the application of new functional foods with clearer and reproducible benefits on host health.! "#!

5 DEDICATION Ahora vuelas y yo solo consigo cosas el tiempo pasa tan lento a veces y no me acostumbro a este horario Te invitaria a un desayuno sin heroes sin sabor pero con mucho del presente a contarnos historias que se van repitiendo errores que no se evitan pasiones que se viven con tranquilidad dolores que se esconden y alegrias que no aparecen en las fotos Dicen que tu y yo somos uno a veces es tan tarde para algunos arrepentimientos y oraciones incendios y saqueos pero ese tu y yo sabemos es ahora tan comodo en mi corazon y tan dificil de explicar JAGN, November 2010.! "!

6 ACKNOWLEDGMENTS There is always someone missing in these lists but first of all my beloved parents and family, mom dad and my siblings, for the good and the bad we ve come to live, and also for the freedom I ve been given to be who I am supposed to; Andy, thanks for the patience and support along all these years; Dr. David Mills, for giving me the chance, trust and freedom to do research, to create and innovate; all my beloved friends back in Chile, Paula, German, Miguel Musalem, Fabiola Pino, Hernan Bustamante, Cristian Cuello, who have faithfully been with me and supported me regardless of the distance; all the people I ve met here in the states, helping me out to adapt and love this country every day more: Marcos Garcia, Ismael Avechuco, Mohamed Omar, Paulina Montecinos, Pablo Zamora, Felipe Aburto, Fernando Fierro and Alejandro Garcia; and finally my colleagues and friends here in Davis, Dr. David Sela from the beginning you were a great support and I m really thankful to have met you; Riccardo LoCascio and Angela Marcobal I had a great time while you were in the lab and you were a model to follow; Karen Kalanetra thanks for your compliance and support.; and Santiago Ruiz-Moyano, thanks for your patience and support all this time. And finally thanks to everyone that at least for one moment helped me throughout this long but incredible ride called thesis.! "#!

7 TABLE OF CONTENTS CHAPTER I. A MOLECULAR BASIS FOR BIFIDOBACTERIAL ENRICHMENT IN THE INFANT GASTROINTESTINAL TRACT. 1 SUMMARY... 2 INTRODUCTION 3 HUMAN MILK OLIGOSACCHARIDES.. 4 ABUNDANCE AND PHYSIOLOGICAL EFFECTS OF MILK OLIGOSACCHARIDES. 5 MILK OLIGOSACCHARIDES CHARACTERIZATION BY MASS SPECTROMETRY. 9 MIMICKING MILK OLIGOSACCHARIDES ESTABLISHMENT OF THE INFANT MICROBIOTA: IMPACT OF BREAST MILK ON BIFIDOBACTERIA 12 MOLECULAR ADAPTATIONS OF BIFIDOBACTERIA TO UTILIZE HMO.. 21 REFERENCES CHAPTER II. OLIGOSACCHARIDE BINDING PROTEINS FROM BIFIDOBACTERIUM LONGUM SUBSP. INFANTIS REVEAL A PREFERENCE FOR HOST GLYCANS.. 39 SUMMARY.. 40 INTRODUCTION.. 41 MATERIALS AND METHODS. 44 RESULTS. 51! DISCUSSION!68 REFERENCES... 78! "##!

8 ACKNOWLEDGMENTS. 83 SUPPLEMENTARY INFORMATION. 84 CHAPTER III. BIFIDOBACTERIUM LONGUM SUBSP. INFANTIS ATCC 15697!-FUCOSIDASES ARE ACTIVE ON FUCOSYLATED HUMAN MILK OLIGOSACCHARIDES SUMMARY.. 90 INTRODUCTION.. 91 MATERIALS AND METHODS 93 RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES SUPPLEMENTARY INFORMATION CHAPTER IV. "-GALACTOSIDASES IN BIFIDOBACTERIUM LONGUM SUBSP. INFANTIS ATCC ACTIVE ON PLANT AND HUMAN MILK OLIGOSACCHARIDES SUMMARY 138 INTRODUCTION. 139 MATERIALS AND METHODS RESULTS GENETIC LANDSCAPES AND PHYLOGENETIC REPRESENTATION OF B. INFANTIS!- GALACTOSIDASES ! "###

9 GENE EXPRESSION OF!-GALACTOSIDASES IN B. INFANTIS ENZYMATIC CHARACTERIZATIONS RELATIVE AFFINITIES AND SUBSTRATE SPECIFICITY OF B. INFANTIS!-GALACTOSIDASES. 155 DISCUSSION ACKNOWLEDGMENTS REFERENCES SUPPLEMENTARY INFORMATION CHAPTER V. RELEASE AND UTILIZATION OF N-ACETYL-D-GLUCOSAMINE FROM HUMAN MILK OLIGOSACCHARIDES BY BIFIDOBACTERIUM LONGUM SUBSP. INFANTIS. 171 SUMMARY 174 INTRODUCTION 175 MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES. 194 SUPPLEMENTARY INFORMATION CHAPTER VI. ENDO-!-N-ACETYLGLUCOSAMINIDASES FROM INFANT-GUT ASSOCIATED BIFIDOBACTERIA RELEASE COMPLEX N-GLYCANS FROM HUMAN MILK GLYCOPROTEINS 204! "#

10 SUMMARY 205 INTRODUCTION. 207 MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES SUPPLEMENTARY INFORMATION CHAPTER VII. CONCLUSIONS, PERSPECTIVES AND FUTURE DIRECTIONS INTRODUCTION. 253 MOLECULAR DETERMINANTS IN THE ENRICHMENT OF BIFIDOBACTERIA IN INFANTS IMPORT OF HMO AND OTHER PREBIOTICS IN B. INFANTIS GLYCOSYL HYDROLASES ACTIVE ON HMO GLYCOSYL HYDROLASES ACTIVE ON MILK GLYCOPROTEINS FUTURE DIRECTIONS REFERENCES APPENDIX: CURRICULUM VITAE 264! "

11 1 Chapter I Introduction: A Molecular Basis for Bifidobacterial Enrichment in the Infant Gastrointestinal Tract Daniel Garrido 1,3, Daniela Barile 2,3 and David A. Mills 1,3. Departments of Viticulture and Enology 1, Food Science and Technology 2 and Foods for Health Institute 3, University of California Davis, One Shields Ave., Davis, CA Abbreviations: HMO: human milk oligosaccharides; GlcNAc: N-acetylglucosamine; NeuAc: N-acetylneuraminic acid; LNT: lacto-n-tetraose; LNB: lacto-n-biose. Keywords: Human milk, human milk oligosaccharides, Bifidobacterium longum subsp. infantis, glycosyl hydrolase, intestinal microbiota. Sections of this chapter have been published in Advances in Nutrition (Garrido, D., Barile, D., Mills, D.A. (2012) A molecular basis for bifidobacterial enrichment in the infant gastrointestinal tract. Advances in Nutrition, accepted). All three authors contributed in writing this article.

12 2 Summary Bifidobacteria are commonly used as probiotics in dairy foods. Select bifidobacterial species are also early colonizers of the breast-fed infant colon, however the mechanism for this enrichment is unclear. We have previously shown that Bifidobacterium longum subsp. infantis is a prototypical bifidobacterial species that can readily utilize human milk oligosaccharides as a sole carbon source. Mass spectrometry-based glycoprofiling has revealed that numerous B. infantis strains preferentially consume small mass oligosaccharides, abundant in human milks. Genome sequencing revealed that B. infantis possesses a bias towards genes required to utilize mammalian-derived carbohydrates. Many of these genomic features encode enzymes that are active on milk oligosaccharides including a novel 40-kb region dedicated to oligosaccharide utilization. Biochemical and molecular characterization of the encoded glycosidases and transport proteins have further resolved the mechanism by which B. infantis selectively imports and catabolizes milk oligosaccharides. Expression studies indicate that many of these key functions are only induced during growth on milk oligosaccharides and not expressed during growth on other prebiotics. Analysis of numerous B. infantis isolates has confirmed that these genomic features are common among the infantis subspecies and likely constitute a competitive colonization strategy employed by these unique bifidobacteria. By detailed characterization of the molecular mechanisms responsible, these studies provide a conceptual framework for bifidobacterial persistence and host-interaction in the infant gastrointestinal tract mediated, in part, through consumption of human milk oligosaccharides.

13 3 1. Introduction Human milk plays a key role in supporting the survival and development of the offspring, and is a good example of a nutrient that has been shaped by evolution [1]. Strong selective pressures have influenced the composition of this fluid, where factors such as the nutritive and protective needs of the newborn are balanced against the energetic cost of milk production by the mother [2]. The World Health Organization defines breastfeeding as the normal way of providing young infants with the nutrients they need for healthy growth and development [3,4]. Human milk is considered a gold standard for nutrition [5], and is characterized by high amounts of essential nutrients [6], such as lactose, fatty acids and proteins, and also macronutrients such as vitamins and minerals. Human milk is compositionally different from mother to mother, from time to time during lactation and also during different breastfeeding periods [7,8]. Human milk also contains a wide variety of non-essential nutrients, which are not consumed by the infant but display complex and potent bioactive functions [9,10,11]. For example human milk oligosaccharides (HMO) are an array of complex carbohydrates that do not provide a direct nutritional value for the infant. Among other putative roles, these molecules act as prebiotics, serving as growth substrates for specific colonic bacteria in breast-fed infants, mainly belonging to the Bifidobacterium genus. In this review we address recent advances in our understanding of the bifidogenic effect of HMO, and how specific intestinal bifidobacteria have developed strategies for gaining access to the energetic content of HMO.

14 4 2. Human Milk Oligosaccharides HMOs are free soluble carbohydrates that contain from 3 to 15 monosaccharides, linked through a variety of glycosidic bonds. They are synthesized solely in the mammary gland during lactation, and the complex pathways that result in all the structures observed have not been completely elucidated yet. Mammalian milk oligosaccharides are commonly divided in two main groups: neutral oligosaccharides and acidic oligosaccharides. Neutral oligosaccharides are composed of galactose (Gal), N-acetyl-glucosamine (GlcNAc), fucose [12] and a lactose core; acidic oligosaccharides are characterized by these same monomers but also one of more molecules of N-acetylneuraminic acid (NeuAc), commonly known as sialic acid [13]. The non-reducing end in the lactose molecule is linked to backbones of Lacto-N- Biose (Gal!1-3GlcNAc; LNB) and/or lactosamine (Gal!1-4GlcNAc). These linkage differences define isomers in HMO belonging to either type 1 or type 2 chains. Fucose and sialic acid residues are in general attached terminally to these backbones. The activity of different glycosyltransferases in HMOs synthesis, with 13 potential glycosidic bonds, results in more than a thousand possible structural combinations. Oligosaccharides are present in human milk at 5 15 g/l, being the third largest component after lactose and lipids. Their concentration is several-fold compared to other mammalian milks. HMOs composition and abundance progressively changes during lactation, with significant changes among individual mothers [14]. The question of why these indigestible molecules are present at high concentrations in milk has challenged researchers for decades. Initial characterization of human milk oligosaccharides started over fifty years ago [15], yet even now, oligosaccharide analysis of human milk remains

15 5 a challenging task because of the high number of possible structures and their overall complexity. 2.1 Abundance and Physiological Effects of Milk Oligosaccharides HMOs are not affected by transit through the stomach and small intestine, reaching at a high concentration the infant colon [16,17,18]. Apparently certain fractions of oligosaccharides can be found in urine [19]. Kunz et al. [20] used 13C-labeled galactose to demonstrate that oligosaccharides are also partially absorbed intact in the infant s intestine and are then excreted in the urine of breast-fed infants. As a consequence, a full spectrum of physiological effects may be possible not only locally in the gastrointestinal tract but also systemically. Because many milk oligosaccharides contain structural elements that are homologous to glycoconjugates present in the intestinal mucosa and used by pathogens for adherence and invasion, it has been suggested that they act as soluble receptor analogues and inhibit the adhesion of pathogens, thus preventing infection [21]. Fucose and sialic acid containing HMOs are particularly important in pathogen deflection as they are found in the terminal position of both oligosaccharide and glycan chains of many cell surfaces [22]. Therefore, breast milk prevents infant diarrhea and gastrointestinal infections in breast-fed infants caused by several bacterial and viral pathogens [23,24,25,26,27,28,29,30,31,32]. It has also been suggested that HMOs may chemically change the glycome of intestinal epithelial cells, thus altering the ability of certain pathogens to adhere [33]. Certain specific linkages present in milk oligosaccharides depend on the blood group and secretor status of the mother [20,34,35]. The terminal residues of several

16 6 HMOs are homolog to Lewis blood group structures, and are therefore related to the development of immune functions. As a consequence, mothers of different blood groups may offer their infants different levels of protection against specific infectious diseases [21,26]. However, the heterogeneity of HMOs and the presence of more than one pathogen-binding motif in HMOs might compensate for these differences. Figure 1 shows a schematic representation of the most abundant neutral oligosaccharides identified in human milk [36]. The most abundant oligosaccharide produced in the mammary glands is the neutral oligosaccharide lacto-n-tetraose (LNT), which is composed of a LNB molecule linked to the lactose core in!1-3 configuration. LNT and its fucosylated derivatives make up to 70-80% of the total HMOs. Type 2 oligosaccharides are less represented in HMOs, however structures such as lacto-nneohexaose are abundant. The oligosaccharide 2 -fucosyllactose (2 FL) represents only about 2% of the total neutral oligosaccharide content, however it has been described to be responsible of most of the anti-pathogenic activity of human milk [24].

17 Figure 1. Relative abundance of specific neutral isomers in a pool of HMOs. Data was adapted from [36]. More details on the structures can be found in this reference. FL: fucosyllactose; LNnT: lacto-n-neotetraose; LNT: lacto-n-tetraose; LNFP I: lacto-nfucopentaose I; LNnH: lacto-n-neohexaose; MFLNH I: monofucosyl lacto-n-hexaose I; DFLNH a: difucosyl lacto-n-hexaose a; DFLNnO: difucosyl lacto-n-neooctaose; DFLNO: difucosyl lacto-n-octaose; TFLNO: trifucosyl lacto-n-octaose. Red triangles: fucose; yellow circles: galactose; blue circles: glucose; blue squares: N- acetylglucosamine. 7

18 8 The acidic fraction of HMOs has been recently characterized in detail [37]. It represents approximately the 20% of total HMOs, and the most represented oligosaccharide is sialyllactose (SL), present in two isomeric forms,!2-3 and!2-6 [37]. An important role of sialic acid in HMOs is its function in promoting an optimal development of the neural system. The rate of brain growth at the neonatal stage exceeds those of any other organ or body tissue, and by 2 years of age, the brain is already 80% of the adult brain weight [37]. It is possible that sialyloligosaccharides improve both brain ganglioside sialic acid content and learning ability of the infants. Their concentration has been measured over the course of lactation, and a significant variation has been found in different lactation stages as well as among individual mothers. Several researchers found over 1 g/l of sialyl-oligosaccharides in human colostrum and up to the first week of lactation, while the content of sialyloligosaccharides seemed to decrease dramatically in full-term milk, with contents ranging from mg/l [38,39,40,41]. Acidic HMOs have an important role in the deflection of pathogens. For example, 3-sialyllactose inhibits adhesion of clinical isolates of Helicobacter pylori [42]. The ability of HMOs to inhibit enteropathogen attachment decreased when the oligosaccharides were desialylated [27]. Other systemic effects attributed to different fractions of HMOs are for example anti-inflammatory effects in the gut [43,44,45,46]; impact in cellular differentiation and proliferation [47,48], and suppression of allergic responses [49]. These studies rely basically in in vitro models but certainly expand our point of view of the complexity of biological functions executed by HMOs.

19 9 2.2 Milk Oligosaccharides Characterization By Mass Spectrometry Despite recent technological advances, the structural characterization of oligosaccharides from breast milk is still a demanding task because of the elevated number of different molecules that can be found, and for the numerous isomeric forms that can be found for each structure, which in turn can be present at different abundances. Traditionally, oligosaccharide structures were determined using sequential digestions with exoglycosidases, regiospecific chemical degradation and chemical derivatization in combination with multidimensional NMR methods [38]. However, these methods are chemically aggressive and usually cause changes to the molecular structure and do not allow further functional analysis such us prebiotic and antipathogenic activities. Mass spectrometry (MS) has become the method of choice for oligosaccharide analysis. Several features of Fourier transform ion cyclotron resonance (FT-ICR) make it ideal for oligosaccharide analysis without the need of chemical modification [50,51]. The high resolution and mass accuracy readily yields the composition in terms of numbers of the monomeric composition (fucose, hexose, N-acetylhexosamine and sialic acid). The relatively recent implementation of soft ionization techniques, such as electrospray ionization [52] and matrix-assisted laser desorption/ionization (MALDI) in combination with FT-ICR has expanded the utility of MS for the analysis of oligosaccharides. The availability of reliable and consistent tandem-ms processes such as collision-induced dissociation (CID), and infrared multiphoton dissociation (IRMPD) in FT-ICR make it a rapid tool to acquire the key information about glycans. The recent introduction of HPLC-microfluidic Chip-TOF technology that combines the chromatographic separation

20 10 with on-line isomers MS detection provides a new strategy to unambiguously annotate isomeric forms of HMOs in a single run and without the use of chemical derivatization. This technology employs an integrated microfluidic device packed with porous graphitized carbon and an orthogonal time-of-flight mass analyzer [51]. Currently only a few commercial standards are available for milk oligosaccharides; therefore, to-date, the highest number of oligosaccharides identified in both human and bovine milks have been obtained using FT-ICR MS and HPLC- Chip/TOF MS techniques [51]. The HPLC-microfluidic Chip-TOF was recently employed, in combination with exoglycosidases digestion, to annotate and create two complete libraries of the neutral and acidic oligosaccharides found in human milk allowing high-throughput annotation of over 70 structures [36,53]. 2.3 Mimicking Milk Oligosaccharides For several mothers, breastfeeding is not possible or not desired, and therefore there is an increasing need for human milk substitutes. The complexity of HMOs prevents their production at higher scales and commercial applications. Because the cost of even simple synthetic oligosaccharides makes them inaccessible for formula supplementation, other structures have been studied. Fructo-oligosaccharides (FOS) and galacto-oligosaccharides [54] are used in dietary products, basically because of their prebiotic effect, increasing the number of beneficial bifidobacteria both in the infant and adult colonic microbiota. GOS are derived enzymatically from the trans-galactosidation of lactose, while FOS are derived from chicory plants and inulin. FOS and GOS prebiotic activities have been reviewed by Roberfroid [55] and Boehm and Stahl [39]. The wide

21 11 availability of FOS and GOS has enabled numerous in vitro, human and rat studies on their prebiotic effects, and their bifidogenic effect is currently accepted. However other species in the intestinal microbiota, some of them potential pathogens, could utilize these substrates [56,57]. Moreover, it is probable that the deflecting properties of HMOs, based on fucose and sialic acid moieties, are not present in substrates like FOS and GOS. Recent research has demonstrated that in addition to human milk, other mammals milk contain oligosaccharides [58]. For example bovine milk contains several oligosaccharides that are analogous to human milk oligosaccharides, suggesting a similar protective action [59,60,61]. Analysis of oligosaccharides from bovine milk has been hindered by the lack of efficient analytical tools to analyze complex carbohydrate structures present in low concentrations. Our group recently established a highthroughput strategy to annotate the human and bovine milk glycome by using advanced mass spectrometry such as MALDI-FTICR with tandem capabilities (collision-induced dissociation, CID and Infrared multiphoton dissociation, IRMPD) and nanoliquid chromatography mass spectrometry of oligosaccharides employing graphitized carbon chromatography on microchip with a high accuracy mass analyzer [62]. Both human and bovine milk contain significant amounts of sialyloligosaccharides, especially at the early lactation stage. These carbohydrates have gained much attention over the last two decades, as several biological activities have been demonstrated in humans. The presence of fucose in bovine milk oligosaccharides (BMO) has been controversial. Most studies in the literature suggest that free complex oligosaccharides present in human milk are largely absent from bovine milk. Only one small fucose-containing OS was reported more than 20 years ago [63]. Since then, fucose presence in bovine milk has never been

22 12 reported, and that first identification has been questioned by the more recent literature as the authors identified only three structures. It is possible that recent advances in analytical methods allow the detection of new bioactive oligosaccharides in bovine milk. Recently a readily available by-product of cheese processing, cheese whey, has been shown to contain oligosaccharides homologous to HMOs. The idea of using whey to extract oligosaccharides is attractive due to its wide availability and low cost compared to other dairy products. Whey permeate is a by-product obtained after whey ultrafiltration, in order to concentrate whey protein. Whey proteins are retained by the membrane, whereas smaller molecules such as lactose and salts pass through the membrane making up the whey permeate. It was recently shown that whey permeate contain a variety of neutral and acidic oligosaccharides, many of which had identical composition to human milk oligosaccharides [64]. 3. Establishment of the Infant Microbiota: Impact of Breast Milk on Bifidobacteria Immediately after birth, the previously sterile newborn will be exposed to a nonsterile environment, where several microorganisms will begin a colonization process and develop a long-term relationship. Recent studies suggest that the first colonizing bacterial communities in different portions of the body are in an undifferentiated state and are essentially similar regardless of the delivery mode [65]. The establishment of a defined microbiota in the infant colon is a concerted succession of microbes, predominantly bacteria, is also an intricate phenomenon that depends on a diverse number of factors. The first colonizers of the intestinal tract are facultative anaerobic bacteria, such as Escherichia coli, enterococci and streptococci strains, that will predominate in the first

23 13 days of life. These bacteria will consume the oxygen in the intestinal lumen, rendering an anaerobic milieu that is more favorable for strict anaerobes, such as bacteroides and bifidobacteria. The process of colonization is dependent on the mode of delivery. Vaginal delivery allows the direct contact of the neonate with several bacterial sources such as the vaginal and the fecal maternal microbiomes [65,66,67]. Several infant gut isolates have been also found in human milk, which consists then in another reservoir of microorganisms colonizing the infant gut [68,69,70,71]. In the other hand, the hospital environment is the main source of bacteria for caesarian born infants [72], as well as the skin microbiota [65]. A delay in microbial colonization by prominent members of the intestinal microbiota such as Bifidobacterium, Bacteroides and E. coli has been observed [73,74]. Bifidobacterial counts are also lower in caesarian born infants [75,76]. Several studies, using different technologies, indicate that the colonization of the infant distal colon shows a noticeable complexity, with significant intra and inter personal differences [77]. This variability is observed mainly at the taxon level in the infant microbiota, and apparently corresponds to specific life events, diseases and changes in diet. Since the composition of human milk is also highly dynamic [78], these changes can also have an impact in the composition of the gut microbiome. Interestingly, it has been recently suggested that the maturation of this community is less drastic that previously appreciated, with a smooth transition to a microbiota resembling those of adults, especially after the introduction of solids [5].

24 14 Type of feeding, either breast milk or formula milk, is one of the main factors in the development of the infant gut microbiota. Infant formulas are usually based on cow s milk, supplemented with vegetal fatty acids, minerals and vitamins. In spite of recent advances and new ingredients included in formulas [79], the wide array of non-essential nutrients present in breast milk, such as oligosaccharides, antibodies and bioactive proteins, still make these preparations an imperfect supplement for infants. Using varied study designs and techniques, different trials have been conducted to evaluate the composition of the intestinal microbiome in breast and bottle fed infants. This area of research however shows contradictory results. This can be explained by intrinsic genetic, socioeconomic and ethnic differences in newborns, fecal sampling from babies at diverse ages, confounding variables and different trial designs, innovations in infant formulations along the years, and inadequate or non-conclusive microbiological or molecular techniques. A meta-analysis might provide light in this regard. Table 1 summarizes several clinical trials performed in the last 10 years addressing the abundance of bifidobacteria in the feces of breast-fed and formula-fed infants. One of the few aspects that creates consensus is that bifidobacteria is the most abundant genus in the infant gut microbiome, regardless the feeding mode. It is also generally accepted that breast milk guides the development of the infant microbiota preventing pathogen colonization and selecting for bacteria that can consume specific complex oligosaccharides. In general, only a few Bifidobacterium species are recovered from the feces of breast-fed infants: Bifidobacterium longum subsp. longum (B. longum), B. longum subsp. infantis (B. infantis), Bifidobacterium bifidum and Bifidobacterium breve. The bifidobacterial species found in bottle-fed infants differ from study to study

25 15 but usually include the aforementioned and also Bifidobacterium adolescentis and Bifidobacterium pseudocatenulatum, which are also commonly found in adults. Other predominant bacteria in the infant microbiome belong to the genus Bacteroides and Clostridium. They are also abundant in the adult microbiota [80]. Some studies also suggest that the Clostridium perfringens cluster is characteristic of formula fed infants, as well as enterococci and enterobacteria, while in breast milk fed newborns Clostridium difficile strains can be found more frequently, together with staphylococci [81,82]. The initial microbial patterns present in the distal colon have a proven impact on several aspects of the infant health, and moreover, they are thought to have a long-term impact on human health [83]. However, the mechanisms by which some microorganisms are responsible for these health benefits are scarcely understood. In particular only a few studies have addressed the relationship between the predominance of bifidobacteria in the infant intestinal microbiome and their potential health benefits in human health [84]. It is well known that the intestinal microbiota in part affects the development of the immune system. The contact of the intestinal epithelium with bacteria is a requirement for the development of the immune system. Therefore, alterations in the colonization and maturation of the intestinal microbiota have been related to the development of diverse allergies in the infant [85]. Presence of adult-like species of Bifidobacterium, such as B. pseudocatenulatum and B. adolescentis, has been associated to allergies, as well as lower numbers of this genus in the gut microbiota of these infants [70,86]. Apparently also allergic mothers carry lower numbers of this genus in their feces. Other studies have found that instead of Bifidobacterium, members of the

26 16 microbiota such as E. coli and C. difficile are risk markers for these diseases [87]. Also, adult diseases such as intestinal bowel diseases (IBD), and diabetes have been related to alterations in the microbial composition developed in the first years of life [83,88].

27 17 Table 1. Clinical trials assessing Bifidobacterium abundance in the infant microbiota in the last 10 years. Study characteristics Main outcome Detection method for Bifidobacterium Ref. 11 infants, followed No difference in Bifidobacterium TTGE for total bacteria [89] longitudinally during the numbers along the three stages and for bifidobacteria, BF period, at weaning of the study. B. infantis and B. competitive PCR for and post weaning. longum commonly found in quantifying these infants along the trial. bifidobacteria. 6 BF and 6 FF infants, six fecal samples obtained in Dominance of Bifidobacterium in BF at days 12-20; similar Culture methods coupled with 16S rrna PCR [90] the 20 days of life. levels of bifidobacteria and identification; FISH and Bacteroides in FF group. epifluorescent microscopy. 72 infants (58 FF and 14 BF) between 2 and 6 weeks old. Formula was Similar levels of bifidobacteria at the beginning of the study. Higher levels of this genus in FF Culture methods. [91] supplemented with supplemented with 1.5 g/l FOS different concentrations compared to BF infants and FF of FOS. with 3 g/l FOS. 7-8 weeks old infants. BF and OSF groups had FISH and fluorescence [92] Two intervention groups, significantly more bifidobacteria microscopy. infant formula compared to standard formula supplemented with group. FOS/GOS 9:1 (OSF group; n=15) and compared to standard formula (n=19). BF group used as a reference (n=15). 68 infants, 1 mo. 31 BF, B. longum, B. infantis and B. T-RFLP and culture [93] 26 mix-fed, 11 FF. breve were commonly isolated. methods. Similar distribution of Bifidobacterium regardless of feeding type.

28 18 63 healthy BF infants, 19 infants standard FF, 19 Formula groups levels of bifidobacteria are comparable to FISH and epifluorescence microscopy. [94] received a prebiotic formula (GOS/FOS 9:1), breast-fed infants levels. 19 received a probiotic strain. 40 infants, different ages. 7 BF, 15 FF, 18 weaned. Increased bifidobacteria in breast-fed infants qpcr 16S rrna and northern hybridization [95] 32 infants at 6 mo. BF (n = 8), standard formula (n = 8), FF supplemented with prebiotics (n = 8) and probiotics with breast milk (n = 8) groups. 31 mothers were allergic. 3.5 mo infants. BF group (n = 26), compared to a standard FF group (n = 33), FF supplemented with prebiotics (n = 32) and probiotics (n = 25) for 13 weeks. Samplings at three different points infants, 1 mo. 700 BF and 232 FF (four types of formula), 98 were mixed fed. Longitudinal study, 14 infants, 7 BF and 7 FF. 61 mother-infant pairs. 1 mo infants; 50 BF and 11 Similar levels of bifidobacteria in BF, prebiotic and probiotic plus breast milk after treatments, which were significantly higher than those in the FF group. B. longum group found in all infants, followed by B. bifidum and B. breve. BF group had higher levels of bifidobacteria at the beginning of the study. No difference in these levels during the treatment. 98.7% colonization by bifidobacteria. No difference between Bifidobacterium numbers between BF and FF infants. Bifidobacterium is a minor component in the microbiota of these infants 96.7% infants harbored bifidobacteria in their feces. B. FISH-epifluorescence [96] microscopy, species PCR for Bifidobacterium species. Culture methods and [97] FISH-Flow cytometry qpcr [98] Microarray and [77] sequencing of most abundant taxonomic groups in the infant microbiota (SSU rrna) qpcr [70]

29 19 FF. longum group commonly found in breast-milk. Predominance of B. bifidum, B. animalis and B. breve after. 82 healthy infants. Sampling every 4 days, Low detection of bifidobacteria. B. longum and B. breve most Culture methods and PCR detection. [74] between 1 and 3 months. abundant. Focused on exclusively BF infants. 30 healthy children, 8-16 mo. Annual examination, Ocurrence rate of Bifidobacterium was 96.7% in 1 Culture, PCR and sequencing methods. [99] for 6 years. No reference to feeding type. yo infants. B. longum (no subspecies level) and B. bifidum were commonly isolated, B. catenulatum, B. breve, B. animalis, B. adolescentis and B. dentium less found. BF infants (n = 30) compared with different prebiotic formulas. No difference among groups regarding bifidobacterial numbers. qpcr and 16S rrna and FISH for bifidobacteria. [100 ] 2 BF, 3 FF (GOS:lcFOS 9:1) Higher Bifidobacterium numbers in BF than FF at beginning study, later no difference was qpcr 16S rrna for bifidobacteria. Also a mixed-species analysis [101 ] observed. B. infantis, B. longum, for bif using fecal RNA. B. breve and B. bifidum detected in all infants. 110 mothers and their 1 mo infants. Most all of them received breast milk. Regular sampling of 14 infants, 1 to 18 months. 7 BF, 7 FF. Bifidobacteria were found in 90% of infants. B. breve was the most abundant species, along with B. longum, B. infantis, B. bifidum and B. catenulatum. BF infants possessed a less complex microbiota composition. Dominance of bifidobacteria in BF compared to FF. qpcr on fecal DNA for [76] bifidobacteria. FISH and DGGE [102 ]

30 20 Same study as [102] BF bifidobacterial strains DGGE for bifidobacteria, [103 showed higher diversity than FF. and 16S rrna PCR ] B. infantis and B. bifidum were restriction methods. more abundant in all infants. B. breve characterized the BF group. 606 infants, 6 weeks age Bifidobacteria represents the FISH and flow cytometry [104 from 5 different European 40% of total bacteria. ] countries. Different Geographical differences were formulas used, compared observed regarding to breast milk. bifidobacteria. BF infants showed higher levels of Bifidobacterium compared to FF infants. Numbers were also higher in mixed-fed compared to FF infants. 21 infants, between 5 and 16 weeks. 10 BF and 11 FF (supplemented with Similar levels of bifidobacteria in both groups at the beginning of the study and at 2 weeks of Culture methods. [105 ] inulin and FOS). treatment. Higher levels of Sampling at 2 and 4 weeks. bifidobacteria in FF infants after a month. 2 BF, 2 standard FF and 2 B. breve was found in BF Bifidobacterium mixedspecies [76] infants fed with a infants, B. longum in FF and microarray, based prebiotic formula. several Bifidobacterium species were found in the prebiotic group. on [94] Abbreviations: BF, breast-fed; FF, formula-fed; FOS, fructooligosaccharides; GOS, galactooligosaccharides; qpcr: quantitative PCR; TTGE, temporal temperature gradient electrophoresis; DGGE, denature gradient gel electrophoresis; FISH, fluorescent in situ hybridization; T-RFLP, terminal restriction fragment length polymorphism.

31 21 4. Molecular adaptations of bifidobacteria to utilize HMO The inherent complexity of HMO renders it inaccessible to digestive enzymes, therefore they can reach high concentrations in the distal colon. To grow on HMO as a substrate, a microorganism requires specific transporters and/or enzymatic machinery to process these molecules. Recent advances in our understanding of how bifidobacteria utilize HMO correlates also with a general notion of bifidobacteria providing health benefits to the host, given the probiotic or anti-pathogenic activities shown for certain strains. From the early studies of Moro in 1900 [106] and Gyorgy in 1954 [107], a relation between human milk and the selective growth of Bifidobacterium was determined. Growth factors in human milk, or bifidus factor, were thought to be GlcNAc containing oligosaccharides and glycoproteins that stimulated the growth of bifidobacteria specifically in breast-fed infants. Bifidobacterium species are grampositive, strictly anaerobic bacteria, with a fermentative metabolism that produces acetate and lactate as end products [108]. Bifidobacteria are often dominant microorganisms in the breast-fed infant gut microbiota, and they also are significant members of the adult gut microbiome. Ward et al [109], first showed that bifidobacteria can consume HMO as a sole carbon source. The extent of growth observed by B. infantis was higher compared to other species within the genus. Later work demonstrated that determined B. infantis preferentially consumed short chain HMO (degree of polymerization less than seven) but longer chains were used when the total concentration of HMO was reduced [110].

32 22 The genome sequence of B. infantis provided an explanation for this phenotype [111]. Similar to other bifidobacterial genomes, B. infantis possessed a number of genes involved in consumption of complex carbohydrates [52,112]. However, gene clusters dedicated to the metabolism of plant polysaccharides in the closely related B. longum were replaced in B. infantis with gene functions related to HMO consumption. Of particular interest was a 43 kb gene cluster (termed HMO cluster I), which so far has been only found in B. infantis strains. It contained several genes predicted to be involved in the import and metabolism of HMO, such as glycosyl hydrolases and oligosaccharide transport proteins, all within a single locus [111]. The HMO cluster 1, as well as other potentially HMO-associated clusters, was shown to be conserved among different strains of B. infantis [113,114]. Importantly a B. infantis isolate in which the HMO cluster I was partially deleted could only weakly utilize HMO, suggesting a correlation of the presence of the HMO cluster I and vigorous growth on HMO [115]. These observations also confirmed the genetic and functional divergence from B. longum strains, which lacked several of these HMO-related clusters. Several observations suggest that B. infantis imports and degrades HMO intracellularly. For example, glycolytic enzymes potentially active on HMOs encoded within this microorganism lack signal peptide sequences, indicating a cytoplasmic localization. In addition, the B. infantis genome contains several family 1 solute binding proteins (SBPs; pfam01547) with predicted affinity for oligosaccharides, suggesting a link to HMO transport. Recently Garrido et al. [116] determined that 10 of 20 SBPs encoded by B. infantis exhibit a binding preference for prominent mammalian glycans. The affinities of these SBPs covered great part of the spectrum of HMO linkages,

33 23 including type 1 and type 2 HMO (Table 2), also matching the substrates that B. infantis is able to consume in vitro [110]. Genes encoding SBPs that bind type 1 and type 2 chains were specifically expressed during growth on HMO (Table 2), but not on fructooligosaccharides and galactooligosaccharides [116]. The consumption of specific HMO such as LNT (lacto-n-tetraose), LNnT (lacto-n-neotetraose) and certain fucosylated HMO by B. infantis was recently observed, and it was suggested that ABC importers are associated to their import [117]. The enzymatic deconstruction of HMO within B. infantis appears to occur sequentially via an array of glycosyl hydrolases [118], and some of them have been associated to HMO consumption given their gene expression patterns (Table 2). The ability to consume sialylated HMO such as sialyl-lnt is likely mediated by Blon_2348, one of two!-sialidase genes in B. infantis. Only Blon_2348 is up-regulated during growth on HMO, and the encoded enzyme more effectively cleaves both!2-3 and!2-6 linkages found in acidic HMO, compared to Blon_0646 [119]. Fucosylated HMO are highly abundant in breast milk and glycoprofiling of HMO consumption revealed that B. infantis readily utilizes lacto-n-fucopentaoses and lacto-n-difucohexaoses, however only after LNT is consumed first. Two!-fucosidases encoded in the HMO cluster I, Blon_2335 and Blon_2336, are expressed during growth on HMO, and both release fucose from 2 and 3 fucosyl-lactose, Lewis a, Lewis x and fucosylated HMO such as lacto-n-fucopentaose I and III [120]. Other fucosidases in the B. infantis genome, albeit showing high kinetic rates and activity on certain HMO species, did not show induction during bacterial growth on HMO.

34 24 In addition, two!-galactosidases, Blon_2016 and Blon_2334, were recently shown to be active type 1 and type 2 HMO linkages respectively [121]. The genes encoding these enzymes showed similar gene expression levels between glucose and HMO [121]. Finally, two!-hexosaminidases in B. infantis are constitutively expressed during growth on HMO and lactose. Blon_2355 seems to be specific for linear GlcNAc!1-3Gal linkages, while Blon_0732 can additionally release GlcNAc from branched HMO characterized by GlcNAc!1-6Gal [122]. Together, these observations indicate that glycosyl hydrolases in B. infantis are expressed during grown on HMO and can cleave all the different linkages found in these molecules. A summary of the molecular determinants within B. infantis associated to HMO import and deconstruction is presented in Table 2.

35 25 Table 2: Genes in B. infantis up-regulated during bacterial growth on HMO associated to its consumption. HMO transport Blon_2344-Blon_2347 Import of type 2 HMO, such as LNnT or LacNAc containing oligosaccharides. Also bind glycans found in colonic mucins. Blon_2350-Blon_2351 Import of GNB. Blon_2177 Import of LNT and other type 1 HMO. Constitutive expression. Blon_0883 Import of LNB, GNB, and certain fucosylated blood sugar oligosaccharides. Glycosyl hydrolases Blon_2348 Exo!-sialidase, active on!2-3/6 linkages. Blon_2355 "-hexosaminidase. Active on GlcNAc"1-3Gal linkages. Blon_0732 "-hexosaminidase. Active on GlcNAc"1-3/6Gal linkages. Blon_2016 "-galactosidase specific for type 1 HMO (Gal"1-3GlcNAc). Blon_2334 "-galactosidase specific for type 2 HMO (Gal"1-4GlcNAc). Blon_2335!-fucosidase, with preference for Fuc!1-2 linkages found in HMO but also active on Fuc!1-3/4. Blon_2336!-fucosidase, specific for Fuc!1-3/4 linkages found in HMO.

36 26 The utilization of HMO by B. bifidum has also been described in detail [117]. This infant-borne bacterium possesses a wide array of extracellular glycosyl hydrolases, which can cleave linkages found in HMO and mucin oligosaccharides [112,123,124]. A single F1SBP is thought to participate in the import of LNB and GNB (galacto-n-biose), as well as other importers that can import monosaccharides [125]. The induction of several of these enzymes has been also determined during growth in vitro on HMO or pig mucin [112]. HMO consumption by B. bifidum is different from that of B. infantis in that it deploys many extracellular glycosyl hydrolases to digest HMO into components, some of which are consumed while others, such as fucose, are left behind [109]. It remains to be determined if these two different HMO consumption strategies undertaken by these two species enable colonization of differente niches within the infant colon. A comparison between HMO consumption between these two species is presented in Figure 2.

37 27 Figure 2: Possible strategies for HMO consumption in B. bifidum and B. infantis. Yellow circles: galactose; red triangles: fucose; blue squares: GlcNAc, purple diamond; blue circle: glucose. Dashed lines in the HMO figure represent potential linkages.

38 28 B. longum as recently described is routinely found in both infant and adult microbiota [52,113,126,127]. However, unlike B. bifidum and B. infantis, the number of enzymes and transporters involved in the metabolism of HMO in B. longum appears to be limited. A membrane-associated endo-n-acetylgalactosaminidase has been described in certain B. longum strains, suggesting possible mucin oligosaccharide release [128,129]. B. longum, as well as several infant-associated bifidobacteria, possesses a gene cluster dedicated to the metabolism of LNB and GNB, linking type 1 HMO and mucin oligosaccharide consumption [113,125]. If the enrichment of bifidobacteria in the infant colon is the result of co-evolution of specific bifidobacteria and milk components, one might predict that the interface between host-microbe is similarly influenced by milk components. Numerous researchers have demonstrated a beneficial impact of bifidobacterial probiotics on the host in both animal models [130] and human studies [131,132,133]. Recently Fukuda et al. [134] demonstrated that production of acetate, a main end product of bifidobacterial metabolism, is a protective factor modulating intestinal permeability in a mouse model. In that work, production of acetate by certain bifidobacterial strains was linked to specific sugar transporters suggesting that select sugar consumption is a driving factor for protective colonization of the host. If milk glycans evolved as a selective substrate for specific bifidobacterial strains commonly found in infants, it is tempting to speculate that the resultant acetate production by those infant-borne bacteria is one mechanism by which milk-driven enrichment of a bifidobacteria protects the infant. However, other HMO-induced protective interfaces are at play as well. Recently Chichlowski et al. [135] demonstrated

39 29 that growth on HMOs increases intestinal cell binding and enhances protective modulation of tight junction proteins and cytokines. In aggregate these results advance a concept of a unique relationship between milk glycans, enrichment of specific bifidobacteria and protection of the infant host. While a full mechanistic understanding of how specific human milk components promote infant growth, development and protection remains elusive, the application of new approaches in analytical chemistry, glycobiology and genomics have advanced our understanding tremendously. It is now clear that specific structural elements of milk oligosaccharides are crucial for their ability to selectively enrich beneficial bifidobacteria while inhibiting, or acting as poor growth substrates for undesirable and pathogenic bacteria. Moreover the genetic and enzymatic determinants that enable specific bifidobacteria to deconstruct and grow on these unique substrates are increasingly being identified and characterized, providing an emerging mechanistic picture of this enrichment. The consequences of this enrichment for the infant, however, are still relatively unclear. It can be predicted that systems biology tools such as metabolomics, and next generation sequencing of intestinal metagenomes or transcriptomes will help to identify, at a more global level, how breast milk is selective for beneficial microbes, how the target microbes respond to this stimulus, and how they interface with the host. Moreover it is likely that these approaches will help in the design of more specific nutritional formulations, primed to drive enrichment in the infant gut of specific bacterial strains with a proven and understood health benefit.

40 30 REFERENCES 1. German JB, Freeman SL, Lebrilla CB, Mills DA (2008) Human milk oligosaccharides: evolution, structures and bioselectivity as substrates for intestinal bacteria. Nestle Nutr Workshop Ser Pediatr Program 62: ; discussion Cuthbertson WF (1999) Evolution of infant nutrition. Br J Nutr 81: Kramer MS, Kakuma R (2002) Optimal duration of exclusive breastfeeding. Cochrane Database Syst Rev: CD Fewtrell MS, Morgan JB, Duggan C, Gunnlaugsson G, Hibberd PL, et al. (2007) Optimal duration of exclusive breastfeeding: what is the evidence to support current recommendations? Am J Clin Nutr 85: 635S-638S. 5. Walker A (2010) Breast milk as the gold standard for protective nutrients. J Pediatr 156: S Kent JC (2007) How breastfeeding works. J Midwifery Womens Health 52: Neville MC, Allen JC, Archer PC, Casey CE, Seacat J, et al. (1991) Studies in human lactation: milk volume and nutrient composition during weaning and lactogenesis. Am J Clin Nutr 54: Allen JC, Keller RP, Archer P, Neville MC (1991) Studies in human lactation: milk composition and daily secretion rates of macronutrients in the first year of lactation. Am J Clin Nutr 54: Walker WA (2010) Mead Johnson Symposium: functional proteins in human milk: role in infant health and development. J Pediatr 156: S Vidal K, Labeta MO, Schiffrin EJ, Donnet-Hughes A (2001) Soluble CD14 in human breast milk and its role in innate immune responses. Acta Odontol Scand 59: Lonnerdal B (2009) Nutritional roles of lactoferrin. Curr Opin Clin Nutr Metab Care 12: Blum S, Heller ED, Krifucks O, Sela S, Hammer-Muntz O, et al. (2008) Identification of a bovine mastitis Escherichia coli subset. Vet Microbiol 132: Gopal PK, Gill HS (2000) Oligosaccharides and glycoconjugates in bovine milk and colostrum. British Journal of Nutrition 84 Suppl 1: S69-S Chaturvedi P, Warren C, Altaye M, Morrow A, Ruiz-Palacios G, et al. (2001) Fucosylated human milk oligosaccharides vary between individuals and over the course of lactation. Glycobiology 11: Gauhe A, Gyorgy P, Hoover JR, Kuhn R, Rose CS, et al. (1954) Bifidus factor. IV. Preparations obtained from human milk. Arch Biochem Biophys 48: Engfer MB, Stahl B, Finke B, Sawatzki G, Daniel H (2000) Human milk oligosaccharides are resistant to enzymatic hydrolysis in the upper gastrointestinal tract. Am J Clin Nutr 71: Gnoth MJ, Kunz C, Kinne-Saffran E, Rudloff S (2000) Human milk oligosaccharides are minimally digested in vitro. J Nutr 130: Chaturvedi P, Warren CD, Buescher CR, Pickering LK, Newburg DS (2001) Survival of human milk oligosaccharides in the intestine of infants. Adv Exp Med Biol 501:

41 19. Rudloff S, Pohlentz G, Diekmann L, Egge H, Kunz C (1996) Urinary excretion of lactose and oligosaccharides in preterm infants fed human milk or infant formula. Acta Paediatr 85: Kunz C, Rudloff S, Baier W, Klein N, Strobel S (2000) Oligosaccharides in human milk: structural, functional, and metabolic aspects. Annu Rev Nutr 20: Morrow A, Ruiz-Palacios G, Jiang X, Newburg D (2005) Human-milk glycans that inhibit pathogen binding protect breast-feeding infants against infectious diarrhea. Journal of Nutrition 135: Imberty A, Varrot A (2008) Microbial recognition of human cell surface glycoconjugates. Current Opinion in Structural Biology 18: Ruvoen-Clouet N, Mas E, Marionneau S, Guillon P, Lombardo D, et al. (2006) Bilesalt-stimulated lipase and mucins from milk of 'secretor' mothers inhibit the binding of Norwalk virus capsids to their carbohydrate ligands. Biochem J 393: Ruiz-Palacios GM, Cervantes LE, Ramos P, Chavez-Munguia B, Newburg DS (2003) Campylobacter jejuni binds intestinal H(O) antigen (Fuc alpha 1, 2Gal beta 1, 4GlcNAc), and fucosyloligosaccharides of human milk inhibit its binding and infection. J Biol Chem 278: Newburg DS, Ruiz-Palacios GM, Morrow AL (2005) Human milk glycans protect infants against enteric pathogens. Annu Rev Nutr 25: Morrow AL, Rangel JM (2004) Human milk protection against infectious diarrhea: implications for prevention and clinical care. Seminars in Pediatric Infectious Diseases 15: Martin-Sosa S, Martin MJ, Hueso P (2002) The sialylated fraction of milk oligosaccharides is partially responsible for binding to enterotoxigenic and uropathogenic Escherichia coli human strains. Journal of Nutrition 132: Newburg DS, Ruiz-Palacios GM, Altaye M, Chaturvedi P, Meinzen-Derr J, et al. (2004) Innate protection conferred by fucosylated oligosaccharides of human milk against diarrhea in breastfed infants. Glycobiology 14: Hong P, Ninonuevo MR, Lee B, Lebrilla C, Bode L (2009) Human milk oligosaccharides reduce HIV-1-gp120 binding to dendritic cell-specific ICAM3- grabbing non-integrin (DC-SIGN). British Journal of Nutrition 101: Mysore JV, Wigginton T, Simon PM, Zopf D, Heman-Ackah LM, et al. (1999) Treatment of Helicobacter pylori infection in rhesus monkeys using a novel antiadhesion compound. Gastroenterology 117: Coppa GV, Zampini L, Galeazzi T, Facinelli B, Ferrante L, et al. (2006) Human milk oligosaccharides inhibit the adhesion to Caco-2 cells of diarrheal pathogens: Escherichia coli, Vibrio cholerae, and Salmonella fyris. Pediatr Res 59: Hakkarainen J, Toivanen M, Leinonen A, Frangsmyr L, Stromberg N, et al. (2005) Human and bovine milk oligosaccharides inhibit Neisseria meningitidis pili attachment in vitro. Journal of Nutrition 135: Angeloni S, Ridet JL, Kusy N, Gao H, Crevoisier F, et al. (2005) Glycoprofiling with micro-arrays of glycoconjugates and lectins. Glycobiology 15:

42 34. Stahl B, Thurl S, Henker J, Siegel M, Finke B, et al. (2001) Detection of four human milk groups with respect to Lewis-blood-group-dependent oligosaccharides by serologic and chromatographic analysis. Adv Exp Med Biol 501: Thurl S, Munzert M, Henker J, Boehm G, Muller-Werner B, et al. (2010) Variation of human milk oligosaccharides in relation to milk groups and lactational periods. British Journal of Nutrition: 1-11 [Epub ahead of print]. 36. Wu S, Tao N, German JB, Grimm R, Lebrilla CB (2010) Development of an annotated library of neutral human milk oligosaccharides. J Proteome Res 9: Wu S, Grimm R, German JB, Lebrilla CB (2011) Annotation and Structural Analysis of Sialylated Human Milk Oligosaccharides. J Proteome Res. 38. Hardy M (1989) Methods for the analysis of glycoprotein carbohydrates. Methods in Enzymology 179: Boehm G, Stahl B (2007) Oligosaccharides from milk. Journal of Nutrition 137: 847S. 40. Martin-Sosa S, Martin M, Garcia-Pardo L, Hueso P (2003) Sialyloligosaccharides in human and bovine milk and in infant formulas: Variations with the progression of lactation. Journal of Dairy Science 86: Wang B, Brand-Miller J, McVeagh P, Petocz P (2001) Concentration and distribution of sialic acid in human milk and infant formulas. American Journal of Clinical Nutrition 74: Simon PM, Goode PL, Mobasseri A, Zopf D (1997) Inhibition of Helicobacter pylori binding to gastrointestinal epithelial cells by sialic acid-containing oligosaccharides. Infection and immunity 65: Kunz C, Rudloff S (2008) Potential anti-inflammatory and anti-infectious effects of human milk oligosaccharides. Adv Exp Med Biol 606: Terrazas LI, Walsh KL, Piskorska D, McGuire E, Harn DA, Jr. (2001) The schistosome oligosaccharide lacto-n-neotetraose expands Gr1(+) cells that secrete anti-inflammatory cytokines and inhibit proliferation of naive CD4(+) cells: a potential mechanism for immune polarization in helminth infections. J Immunol 167: Schumacher G, Bendas G, Stahl B, Beermann C (2006) Human milk oligosaccharides affect P-selectin binding capacities: in vitro investigation. Nutrition 22: Bode L, Rudloff S, Kunz C, Strobel S, Klein N (2004) Human milk oligosaccharides reduce platelet-neutrophil complex formation leading to a decrease in neutrophil beta 2 integrin expression. J Leukoc Biol 76: Kuntz S, Kunz C, Rudloff S (2009) Oligosaccharides from human milk induce growth arrest via G2/M by influencing growth-related cell cycle genes in intestinal epithelial cells. Br J Nutr 101: Kuntz S, Rudloff S, Kunz C (2008) Oligosaccharides from human milk influence growth-related characteristics of intestinally transformed and non-transformed intestinal cells. Br J Nutr 99: Eiwegger T, Stahl B, Haidl P, Schmitt J, Boehm G, et al. (2010) Prebiotic oligosaccharides: in vitro evidence for gastrointestinal epithelial transfer and immunomodulatory properties. Pediatr Allergy Immunol 21:

43 50. Park Y, Lebrilla C (2005) Application of Fourier transform ion cyclotron resonance mass spectrometry to oligosaccharides. Mass Spectrometry Reviews 24: Niñonuevo M, Lebrilla C (2009) Mass spectrometric methods for analysis of oligosaccharides in human milk. Nutrition Reviews 67: S216-S Schell MA, Karmirantzou M, Snel B, Vilanova D, Berger B, et al. (2002) The genome sequence of Bifidobacterium longum reflects its adaptation to the human gastrointestinal tract. Proc Natl Acad Sci U S A 99: Wu S, Grimm R, German J, Lebrilla C (2010) Annotation and Structural Analysis of Sialylated Human Milk Oligosaccharides. Journal of Proteome Research. 54. Gosling A, Stevens GW, Barber AR, Kentish SE, Gras SL (2010) Recent advances refining galactooligosaccharide production from lactose. Food Chemistry 121: Roberfroid M (2007) Prebiotics: the concept revisited. Journal of Nutrition 137: 830S-837S. 56. Hartemink R, Van Laere KM, Rombouts FM (1997) Growth of enterobacteria on fructo-oligosaccharides. J Appl Microbiol 83: Gourgue-Jeannot C, Kalmokoff ML, Kheradpir E, Kwan J, Lampi BJ, et al. (2006) Dietary fructooligosaccharides alter the cultivable faecal population of rats but do not stimulate the growth of intestinal bifidobacteria. Can J Microbiol 52: Tao N, Wu S, Kim J, An HJ, Hinde K, et al. (2011) Evolutionary Glycomics: Characterization of Milk Oligosaccharides in Primates. J Proteome Res. 59. Tao N, DePeters EJ, Freeman S, German JB, Grimm R, et al. (2008) Bovine milk glycome. Journal of Dairy Science 91: Tao N, DePeters EJ, German JB, Grimm R, Lebrilla CB (2009) Variations in bovine milk oligosaccharides during early and middle lactation stages analyzed by highperformance liquid chromatography-chip/mass spectrometry. Journal of Dairy Science 92: Barile D, Marotta M, Chu C, Mehra R, Grimm R, et al. (2010) Neutral and acidic oligosaccharides in Holstein-Friesian colostrum during the first 3 days of lactation measured by high performance liquid chromatography on a microfluidic chip and time-of-flight mass spectrometry. Journal of Dairy Science 93: Ninonuevo MR, Lebrilla CB (2009) Mass spectrometric methods for analysis of oligosaccharides in human milk. Nutr Rev 67 Suppl 2: S Saito T, Itoh T, Adachi S (1987) Chemical structure of three neutral trisaccharides isolated in free form from bovine colostrum. Carbohydrate research 165: Barile D, Tao N, Lebrilla C, Coisson J, Arlorio M, et al. (2009) Permeate from cheese whey ultrafiltration is a source of milk oligosaccharides. International Dairy Journal 19: Dominguez-Bello MG, Costello EK, Contreras M, Magris M, Hidalgo G, et al. (2010) Delivery mode shapes the acquisition and structure of the initial microbiota across multiple body habitats in newborns. Proc Natl Acad Sci U S A 107: Tannock GW, Fuller R, Smith SL, Hall MA (1990) Plasmid profiling of members of the family Enterobacteriaceae, lactobacilli, and bifidobacteria to study the transmission of bacteria from mother to infant. J Clin Microbiol 28: Matsumiya Y, Kato N, Watanabe K, Kato H (2002) Molecular epidemiological study of vertical transmission of vaginal Lactobacillus species from mothers to newborn 33

44 infants in Japanese, by arbitrarily primed polymerase chain reaction. J Infect Chemother 8: Martin R, Langa S, Reviriego C, Jiminez E, Marin ML, et al. (2003) Human milk is a source of lactic acid bacteria for the infant gut. J Pediatr 143: Solis G, de Los Reyes-Gavilan CG, Fernandez N, Margolles A, Gueimonde M (2010) Establishment and development of lactic acid bacteria and bifidobacteria microbiota in breast-milk and the infant gut. Anaerobe 16: Gronlund MM, Gueimonde M, Laitinen K, Kociubinski G, Gronroos T, et al. (2007) Maternal breast-milk and intestinal bifidobacteria guide the compositional development of the Bifidobacterium microbiota in infants at risk of allergic disease. Clin Exp Allergy 37: Collado MC, Delgado S, Maldonado A, Rodriguez JM (2009) Assessment of the bacterial diversity of breast milk of healthy women by quantitative real-time PCR. Lett Appl Microbiol 48: Martirosian G, Kuipers S, Verbrugh H, van Belkum A, Meisel-Mikolajczyk F (1995) PCR ribotyping and arbitrarily primed PCR for typing strains of Clostridium difficile from a Polish maternity hospital. J Clin Microbiol 33: Adlerberth I, Lindberg E, Aberg N, Hesselmar B, Saalman R, et al. (2006) Reduced enterobacterial and increased staphylococcal colonization of the infantile bowel: an effect of hygienic lifestyle? Pediatr Res 59: Mitsou EK, Kirtzalidou E, Oikonomou I, Liosis G, Kyriacou A (2008) Fecal microflora of Greek healthy neonates. Anaerobe 14: Chen J, Cai W, Feng Y (2007) Development of intestinal bifidobacteria and lactobacilli in breast-fed neonates. Clin Nutr 26: Boesten R, Schuren F, Ben Amor K, Haarman M, Knol J, et al. (2010) Bifidobacterium population analysis in the infant gut by direct mapping of genomic hybridization patterns: potential for monitoring temporal development and effects of dietary regimens. Microbial Biotechnology: no-no. 77. Palmer C, Bik EM, DiGiulio DB, Relman DA, Brown PO (2007) Development of the human infant intestinal microbiota. PLoS Biol 5: e Picciano MF (1998) Human milk: nutritional aspects of a dynamic food. Biol Neonate 74: Koletzko B (2010) Innovations in infant milk feeding: from the past to the future. Nestle Nutr Workshop Ser Pediatr Program 66: Mariat D, Firmesse O, Levenez F, Guimaraes V, Sokol H, et al. (2009) The Firmicutes/Bacteroidetes ratio of the human microbiota changes with age. BMC Microbiol 9: Mountzouris KC, McCartney AL, Gibson GR (2002) Intestinal microflora of human infants and current trends for its nutritional modulation. Br J Nutr 87: Adlerberth I, Wold AE (2009) Establishment of the gut microbiota in Western infants. Acta Paediatr 98: Conroy ME, Shi HN, Walker WA (2009) The long-term health effects of neonatal microbial flora. Curr Opin Allergy Clin Immunol 9: Fujii T, Ohtsuka Y, Lee T, Kudo T, Shoji H, et al. (2006) Bifidobacterium breve enhances transforming growth factor beta1 signaling by regulating Smad7 expression in preterm infants. J Pediatr Gastroenterol Nutr 43:

45 85. Vael C, Desager K (2009) The importance of the development of the intestinal microbiota in infancy. Curr Opin Pediatr 21: Kirjavainen PV, Apostolou E, Arvola T, Salminen SJ, Gibson GR, et al. (2001) Characterizing the composition of intestinal microflora as a prospective treatment target in infant allergic disease. FEMS Immunol Med Microbiol 32: Penders J, Thijs C, van den Brandt PA, Kummeling I, Snijders B, et al. (2007) Gut microbiota composition and development of atopic manifestations in infancy: the KOALA Birth Cohort Study. Gut 56: Le Huerou-Luron I, Blat S, Boudry G (2010) Breast- v. formula-feeding: impacts on the digestive tract and immediate and long-term health effects. Nutr Res Rev 23: Magne F, Hachelaf W, Suau A, Boudraa G, Mangin I, et al. (2006) A longitudinal study of infant faecal microbiota during weaning. FEMS Microbiol Ecol 58: Harmsen HJ, Wildeboer-Veloo AC, Raangs GC, Wagendorp AA, Klijn N, et al. (2000) Analysis of intestinal flora development in breast-fed and formula-fed infants by using molecular identification and detection methods. J Pediatr Gastroenterol Nutr 30: Euler AR, Mitchell DK, Kline R, Pickering LK (2005) Prebiotic effect of fructooligosaccharide supplemented term infant formula at two concentrations compared with unsupplemented formula and human milk. J Pediatr Gastroenterol Nutr 40: Knol J, Scholtens P, Kafka C, Steenbakkers J, Gro S, et al. (2005) Colon microflora in infants fed formula with galacto- and fructo-oligosaccharides: more like breastfed infants. J Pediatr Gastroenterol Nutr 40: Sakata S, Tonooka T, Ishizeki S, Takada M, Sakamoto M, et al. (2005) Cultureindependent analysis of fecal microbiota in infants, with special reference to Bifidobacterium species. FEMS Microbiol Lett 243: Bakker-Zierikzee AM, Alles MS, Knol J, Kok FJ, Tolboom JJ, et al. (2005) Effects of infant formula containing a mixture of galacto- and fructo-oligosaccharides or viable Bifidobacterium animalis on the intestinal microflora during the first 4 months of life. Br J Nutr 94: Hopkins MJ, Macfarlane GT, Furrie E, Fite A, Macfarlane S (2005) Characterisation of intestinal bacteria in infant stools using real-time PCR and northern hybridisation analyses. FEMS Microbiol Ecol 54: Rinne MM, Gueimonde M, Kalliomaki M, Hoppu U, Salminen SJ, et al. (2005) Similar bifidogenic effects of prebiotic-supplemented partially hydrolyzed infant formula and breastfeeding on infant gut microbiota. FEMS Immunol Med Microbiol 43: Brunser O, Figueroa G, Gotteland M, Haschke-Becher E, Magliola C, et al. (2006) Effects of probiotic or prebiotic supplemented milk formulas on fecal microbiota composition of infants. Asia Pac J Clin Nutr 15: Penders J, Thijs C, Vink C, Stelma FF, Snijders B, et al. (2006) Factors influencing the composition of the intestinal microbiota in early infancy. Pediatrics 118:

46 99. Shkoporov AN, Khokhlova EV, Kulagina EV, Smeianov VV, Kafarskaia LI, et al. (2008) Application of several molecular techniques to study numerically predominant Bifidobacterium spp. and Bacteroidales order strains in the feces of healthy children. Biosci Biotechnol Biochem 72: Nakamura N, Gaskins HR, Collier CT, Nava GM, Rai D, et al. (2009) Molecular ecological analysis of fecal bacterial populations from term infants fed formula supplemented with selected blends of prebiotics. Appl Environ Microbiol 75: Klaassens ES, Boesten RJ, Haarman M, Knol J, Schuren FH, et al. (2009) Mixedspecies genomic microarray analysis of fecal samples reveals differential transcriptional responses of bifidobacteria in breast- and formula-fed infants. Appl Environ Microbiol 75: Roger LC, McCartney AL (2010) Longitudinal investigation of the faecal microbiota of healthy full-term infants using fluorescence in situ hybridization and denaturing gradient gel electrophoresis. Microbiology 156: Roger LC, Costabile A, Holland DT, Hoyles L, McCartney AL (2010) Examination of faecal Bifidobacterium populations in breast- and formula-fed infants during the first 18 months of life. Microbiology 156: Fallani M, Young D, Scott J, Norin E, Amarri S, et al. (2010) Intestinal microbiota of 6-week-old infants across Europe: geographic influence beyond delivery mode, breast-feeding, and antibiotics. J Pediatr Gastroenterol Nutr 51: Lugonja NM, Martinov OB, Rasovic MR, Spasic SD, Gojgic G, et al. (2010) A Comparative Investigation of an in vitro and Clinical Test of the Bifidogenic Effect of an Infant Formula. J Clin Biochem Nutr 47: Moro E (1905) Morphologische und biologische Untersuchung ueber die Darmbakterien des Siiuglings. Jahrb f Kinderh 61: Gyorgy P, Norris RF, Rose CS (1954) Bifidus factor. I. A variant of Lactobacillus bifidus requiring a special growth factor. Arch Biochem Biophys 48: Lee JH, O'Sullivan DJ (2010) Genomic insights into bifidobacteria. Microbiol Mol Biol Rev 74: Ward RE, Ninonuevo M, Mills DA, Lebrilla CB, German JB (2007) In vitro fermentability of human milk oligosaccharides by several strains of bifidobacteria. Mol Nutr Food Res 51: LoCascio RG, Ninonuevo MR, Freeman SL, Sela DA, Grimm R, et al. (2007) Glycoprofiling of bifidobacterial consumption of human milk oligosaccharides demonstrates strain specific, preferential consumption of small chain glycans secreted in early human lactation. J Agric Food Chem 55: Sela DA, Chapman J, Adeuya A, Kim JH, Chen F, et al. (2008) The genome sequence of Bifidobacterium longum subsp. infantis reveals adaptations for milk utilization within the infant microbiome. Proc Natl Acad Sci U S A 105: Turroni F, Bottacini F, Foroni E, Mulder I, Kim JH, et al. (2010) Genome analysis of Bifidobacterium bifidum PRL2010 reveals metabolic pathways for hostderived glycan foraging. Proc Natl Acad Sci U S A 107:

47 113. LoCascio RG, Desai P, Sela DA, Weimer B, Mills DA (2010) Broad conservation of milk utilization genes in Bifidobacterium longum subsp. infantis as revealed by comparative genomic hybridization. Appl Environ Microbiol 76: Zivkovic AM, German JB, Lebrilla CB, Mills DA (2011) Human milk glycobiome and its impact on the infant gastrointestinal microbiota. Proc Natl Acad Sci U S A 108 Suppl 1: Locascio RG, Ninonuevo MR, Kronewitter SR, Freeman SL, German JB, et al. (2009) A versatile and scalable strategy for glycoprofiling bifidobacterial consumption of human milk oligosaccharides. Microb Biotechnol 2: Garrido D, Kim JH, German JB, Raybould HE, Mills DA (2011) Oligosaccharide binding proteins from Bifidobacterium longum subsp. infantis reveal a preference for host glycans. PLoS One 6: e Asakuma S, Hatakeyama E, Urashima T, Yoshida E, Katayama T, et al. (2011) Physiology of the consumption of human milk oligosaccharides by infant-gut associated bifidobacteria. J Biol Chem Sela DA, Mills DA (2010) Nursing our microbiota: molecular linkages between bifidobacteria and milk oligosaccharides. Trends Microbiol 18: Sela DA, Li Y, Lerno L, Wu S, Marcobal AM, et al. (2011) An infant-associated bacterial commensal utilizes breast milk sialyloligosaccharides. J Biol Chem 286: Sela DA, Garrido D, Lerno L, Wu S, Tan K, et al. (2011) Bifidobacterium longum subsp. infantis ATCC15697 alpha-fucosidases are active on fucosylated human milk oligosaccharides. Appl Environ Microbiol Yoshida E, Sakurama H, Kiyohara M, Nakajima M, Kitaoka M, et al. (2011) Bifidobacterium longum subsp. infantis uses two different {beta}-galactosidases for selectively degrading type-1 and type-2 human milk oligosaccharides. Glycobiology Garrido D, Ruiz-Moyano S., Mills D.A. (2011) Release and utilization of N-acetyl- D-glucosamine from human milk oligosaccharides by Bifidobacterium longum subsp. infantis. Anaerobe submitted Miwa M, Horimoto T, Kiyohara M, Katayama T, Kitaoka M, et al. (2010) Cooperation of {beta}-galactosidase and {beta}-n-acetylhexosaminidase from bifidobacteria in assimilation of human milk oligosaccharides with type 2 structure. Glycobiology Wada J, Ando T, Kiyohara M, Ashida H, Kitaoka M, et al. (2008) Bifidobacterium bifidum lacto-n-biosidase, a critical enzyme for the degradation of human milk oligosaccharides with a type 1 structure. Appl Environ Microbiol 74: Kitaoka M, Tian J, Nishimoto M (2005) Novel putative galactose operon involving lacto-n-biose phosphorylase in Bifidobacterium longum. Appl Environ Microbiol 71: Parche S, Amon J, Jankovic I, Rezzonico E, Beleut M, et al. (2007) Sugar transport systems of Bifidobacterium longum NCC2705. J Mol Microbiol Biotechnol 12: Hinz SW, Pastink MI, van den Broek LA, Vincken JP, Voragen AG (2005) Bifidobacterium longum endogalactanase liberates galactotriose from type I galactans. Appl Environ Microbiol 71:

48 128. Ruas-Madiedo P, Gueimonde M, Fernandez-Garcia M, de los Reyes-Gavilan CG, Margolles A (2008) Mucin degradation by Bifidobacterium strains isolated from the human intestinal microbiota. Appl Environ Microbiol 74: Ashida H, Maki R, Ozawa H, Tani Y, Kiyohara M, et al. (2008) Characterization of two different endo-alpha-n-acetylgalactosaminidases from probiotic and pathogenic enterobacteria, Bifidobacterium longum and Clostridium perfringens. Glycobiology 18: McCarthy J, O'Mahony L, O'Callaghan L, Sheil B, Vaughan EE, et al. (2003) Double blind, placebo controlled trial of two probiotic strains in interleukin 10 knockout mice and mechanistic link with cytokine balance. Gut 52: Isolauri E, Arvola T, Sutas Y, Moilanen E, Salminen S (2000) Probiotics in the management of atopic eczema. Clin Exp Allergy 30: Saavedra JM, Bauman NA, Oung I, Perman JA, Yolken RH (1994) Feeding of Bifidobacterium bifidum and Streptococcus thermophilus to infants in hospital for prevention of diarrhoea and shedding of rotavirus. Lancet 344: Saavedra JM (2007) Use of probiotics in pediatrics: rationale, mechanisms of action, and practical aspects. Nutr Clin Pract 22: Fukuda S, Toh H, Hase K, Oshima K, Nakanishi Y, et al. (2011) Bifidobacteria can protect from enteropathogenic infection through production of acetate. Nature 469: Chichlowski M, De Lartigue G., German J. B., Raybould H. E., Mills D. A. (2011) Select bifidobacteria cultured on human milk oligosaccharides affect intestinal epithelial function. Journal of Pediatric Gastroenterology and Nutrition In revision. 38

49 39! Chapter II Oligosaccharide Binding Proteins from Bifidobacterium longum subsp. infantis Reveal a Preference for Host Glycans. Daniel Garrido 1,4,5,6, Jae Han Kim 2,4,5,6, J. Bruce German 1,4,5,6, Helen E. Raybould 3,4,5 and David A. Mills 2,4,5,6. Departments of Food Science and Technology 1, Viticulture and Enology 2 and Anatomy, Physiology & Cell Biology 3 ; Foods for Health Institute 4, Functional Glycobiology Program 5 and Robert Mondavi Institute for Wine and Food Sciences 6, University of California Davis, Davis, CA, USA. Sections of this chapter have been published in PLoS ONE (Garrido D, Kim JH, German JB, Raybould HE, Mills DA (2011) Oligosaccharide binding proteins from Bifidobacterium longum subsp. infantis reveal a preference for host glycans. PLoS One 6: e17315.) Conceived and designed the experiments: DG, JHK, DAM, HER. Performed the experiments: DG, JHK. Analyzed the data: DG, JHK, HER, DAM. Contributed reagents/materials/analysis tools: JBG. Wrote the paper: DG, DAM.

50 40! Summary Bifidobacterium longum subsp. infantis (B. infantis) is a common member of the infant intestinal microbiota, and it has been characterized by its foraging capacity for human milk oligosaccharides (HMO). Its genome sequence revealed an overabundance of the Family 1 of solute binding proteins (F1SBPs), part of ABC transporters and associated with the import of oligosaccharides. In this study we have used the Mammalian Glycan Array to determine the specific affinities of these proteins. This was correlated with binding protein expression induced by different prebiotics including HMO. Half of the F1SBPs in B. infantis were determined to bind mammalian oligosaccharides. Their affinities included different blood group structures and mucin oligosaccharides. Related to HMO, other proteins were specific for oligomers of lacto- N-biose (LNB) and polylactosamines with different degrees of fucosylation. Growth on HMO induced the expression of specific binding proteins that import HMO isomers, but also bind blood group and mucin oligosaccharides, suggesting coregulated transport mechanisms. The prebiotic inulin!,-./01.! 23415! 678,9:! +! ;,-.,-<! =5231,->!?,34! 766,-,3:!625!,-31>3,-79!<9:07->@!A2>3!26!341!42>3!<9:07-!B+CDE>!,-!!"#$%&'%($)#.2!-23! 47F1!428292<>!,-!23415!;,6,.2;70315,7@!Finally, some of these proteins were found to be adherent to intestinal epithelial cells in vitro. In conclusion, this study represents further evidence for the particular adaptations of B. infantis to the infant gut environment, and helps to understand the molecular mechanisms involved in this process.

51 41! 1. Introduction The distal portion of the human gastrointestinal tract is colonized by an impressive number of microorganisms, which are collectively referred as the intestinal microbiota. The environment they occupy is considered to be rich in complex dietary and host-derived oligosaccharides, which serve as an important carbon source for many of these microbes. In general, members of the intestinal microbiota devote an important percentage of their genomes to the utilization of complex carbohydrates [1,2]. This specialization requires specific intra or extracellular glycosyl hydrolases to break down the complex oligosaccharides arriving to the colon, and also transport mechanisms to import intact or processed glycans inside the bacterial cell. Bacterial ABC transporters involved in the import of oligosaccharides show an exquisite specificity, conferred by Solute Binding Proteins (SBPs) [3]. Bifidobacterium species represent an important percentage of the total colonic microbiota in infants [4], and they remain at significant levels in adults [5]. In general they are considered beneficial microorganisms. Some strains in this genus are noteworthy as probiotics, and they represent one of the main targets in prebiotic interventions [6]. In particular, bifidobacteria are efficient in the utilization of complex oligosaccharides, such as fructo-oligosaccharides (FOS)! [7], galacto-oligosaccharides (GOS) [8] and inulin [9]. In the last decades several studies have associated breast-feeding with higher concentrations of bifidobacteria in infant feces [10,11]. Formula fed infants seem to have lower numbers of Bifidobacterium, even though it has been argued that improvements in the composition of infant formulas hsve reduced this difference [12].

52 42! Formula fed infants are also characterized by the presence of species commonly found in adults, such as Bifidobacterium adolescentis and Bifidobacterium pseudocatenulatum, while breast-fed infants are colonized more often with Bifidobacterium breve and B. infantis! [13]. Bifidobacterium bifidum and Bifidobacterium longum subsp. longum (B. longum), can be found both in infants and adults [14]. Human milk oligosaccharides (HMO) are abundant in breast milk and they are thought to act as prebiotics selecting for specific bacterial populations in the infant gut. They are characterized by a lactose molecule at the reducing end to which subunits of lacto-n-biose (LNB; type 1 chain; Gal!1-3GlcNAc) or N-acetyl-lactosamine (type 2 chain; Gal!1-4GlcNAc) are attached in tandem [15]. Fucose and sialic acid residues can be located at terminal positions. 200 different HMO structures have been determined, however, four molecular masses can represent up to the 70% of the total molecules, including isomers of lacto-n-tetraose (Gal!1-3GlcNAc!1-3Gal!1-4Glc; LNT), lacto-n-neotetraose (Gal!1-4GlcNAc!1-3Gal!1-4Glc; LNnT), lacto-n-hexaose (LNH), monofucosyl-lacto-n-hexaose and difucosyl lacto-n-hexaose [16]. Interestingly, HMO structures resemble the lacto and neolacto family of glycolipids [17], and polylactosamine chains are characteristic in glycoproteins [18]. Mucin oligosaccharides also share structural similarity with HMO [19]. Recently B. longum, B. infantis, B. breve and B. bifidum were characterized by their capabilities for consuming HMO [20]. Particularly, a gene cluster for metabolism and import of LNB and galacto-n-biose (Gal!1-3GalNAc; GNB), a core substructure in mucin glycans, is found in these microorganisms [21]. Also enzymes involved in the deconstruction of HMO and the metabolism of its constituents have been characterized

53 43! [22,23]. Different strategies exist for HMO consumption. B. longum, B. breve and B. bifidum seem to prefer LNT [24], and potentially rely on extracellular glycosyl hydrolases for reduction of more complex HMO as well as mucin glycans [25], while B. infantis apparently has expanded its metabolic capabilities to use a higher variety of HMO molecules, specializing in the import of complex carbohydrates and intracellular degradation [26]. The genome of this bacterium has been recently sequenced [27], and several features indicated a specialization in carbohydrate utilization. In addition to the LNB/GNB cluster, B. infantis possesses two additional gene clusters containing ABC transporters and glycosyl hydrolases predicted to be involved in HMO utilization. These clusters were also found to contain an important number of Family 1 Solute Binding Proteins (F1SBPs), predicted to import oligosaccharides associated to ABC transporters. Genomic analysis suggests that B. infantis evolved from a plant-derived glycan utilization genotype, to be competitive in the infant colon [27]. Interestingly, all available carbon sources in this environment are oligosaccharides from human origin, including a significant concentration of HMO arriving undigested to the distal colon, as well as intestinal secretions and glycoconjugates from epithelial cells. The F1SBPs in B. infantis might be important to understand the adaptations of this microorganism to the infant gut, as well as the range of oligosaccharides it is able to import into the cell and use as a carbon source. In this work we have characterized B. infantis F1SBPs oligosaccharide affinities, expression during growth on different prebiotics and interactions with epithelial cells in order to better understand the complex nature for oligosaccharide foraging by this important intestinal species.

54 44! 2. Materials and Methods Bacteria and media. B. longum subsp. infantis ATCC was obtained from the American Type Culture Collection (Manassas, VA). Cultures were routinely grown using Mann Rogose Sharp medium with no carbon source, and supplemented with 2% w/v lactose and 0.25 % w/v L-cysteine (Sigma-Aldrich, St. Louis, MO). Bacteria were routinely cultured in a vinyl anaerobic chamber (Coy Laboratory Products, Grass Lake, MI) at 37 C. Chemically competent Escherichia coli BL21 and DH5! were obtained from EMD Chemicals (Gibbstown, NJ), and grown in Luria Broth supplemented with 50 µg/ml ampicilin (Sigma-Aldrich, St. Louis, MO) when necessary at 37 C. F1SBP Cloning and Expression. Genes coding for F1SBPs were described by [27]. Gene features such as lipoprotein attachment site and signal peptide sequence were identified using the Biology workbench 3.2 ( PCR primers were designed to target each gene excluding the signal peptide region. Primers sequences were designed with different 5 restriction sites, and an extra six base pairs for proper enzyme restriction digestion (Table S1). Genomic DNA from B. infantis was purified from 2 ml cultures using the MasterPure Gram Positive DNA Purification Kit (Epicentre Biotechnologies, Madison, WI), following the manufacturer instructions. F1SBP genes were amplified by PCR using 0.5 µm of each forward and reverse primer, 1 ng genomic DNA, 0.2 mm dntp mix (Fermentas, Glen Burnie, MD), and 2 U of PFU Turbo DNA polymerase (Stratagene, Cedar Creek, TX) in a 100 µl final volume. PCR was performed using a PTC200 Thermo Cycler (MJ Research, Ramsey, MN), with an initial denaturation at 95 C for 2 min, and 35 cycles of denaturation at 95 C 1 min, annealing at 58 C for 1 min, extension at 72 C 2 min, and a final extension at 72 C for

55 45! 7 min. PCR products were gel purified and digested with two restriction enzymes (Fermentas, Glen Burnie, MD), which cleaved the sequences shown in bold in Table S1. The expression plasmid pgex-6p-1 (Glutathione-S-Transferase Gene Fusion System, GE Healthcare, Piscataway, NJ), was linearized with the same two enzymes for each PCR product and dephosphorylated using Antarctic Phosphatase (New England Biolabs, Ipswich, MA), and ligated to each F1SBP gene in a 1:3 ratio using 1U of T4 DNA ligase (Promega, Madison, WI), in a 10 µl final volume at 4 C overnight. Ligation reactions were transformed into E. coli BL21 and DH5 chemically competent cells at 42 C for 30 sec and allowed to recover for 1 h at 37 C in SOC media (Invitrogen). Recombinant clones were selected for growth on LB agar with 50 µg/ml ampicilin, and confirmed by plasmid sequencing using primers pgex 5 and pgex 3 (GE Healthcare). Protein expression was carried out using 100 ml LB broth supplemented with 50 µg/ml ampicilin. Cells were grown at 37 in a shaker at 250 rpm (Innova-4000, New Brunswick Scientific, Edison, NJ), until they reached an O.D. of 0.6 when they were supplemented with 1 mm IPTG (Fermentas) for 2 h. In some cases due to low protein solubility, cultures were scaled to 2 L and growth was performed at 28 C. Cells were collected in an Eppendorf 5804 centrifuge (Eppendorf, Hauppauge, NY) 20 min at 4 and at 4000 rpm, and stored at -80 C at least for 18 h. Pellets were resuspended in Bugbuster Protein Extraction Reagent (EMD Chemicals), using 1 ml of the detergent for 20 ml of the original culture, and incubated for 20 min at room temperature. The suspension was centrifuged for 20 min at rpm at 4 C, and the supernatant was applied to 1 ml Bio-Scale Mini Profinity GST cartridges, connected to an EP-1 Econopump (Bio-Rad, Hercules, CA). Protein purification was performed as recommended by

56 46! the manufacturer. Recombinant proteins were evaluated for molecular weight and purity in SDS-PAGE gels. Finally, glutathione was exchanged for PBS buffer using Amicon Ultra-15 Centrifugal Filter Units, with a cut-off of 50 kda (Millipore, Billerica, CA). Glycan Array Analysis. Each GST-tagged F1SBP was screened for glycan interactions by the Core H of the Consortium for Functional Glycomics, using versions 3.1 and up of the Glycan Array. Each protein sample was diluted to 200 µg/ml in binding buffer (20mM Tris-HCL ph 7.4, 150 mm sodium chloride, 2mM calcium chloride, 2mM magnesium chloride, 0.05% Tween 20 and 1% BSA), and assayed to printed glycan slides. Binding and detection were carried out using an anti-gst Alexa 488 conjugate antibody (Molecular Probes, Eugene, OR). Results were expressed as Relative Fluorescent Units (RFU), and the highest and lowest point from each set of six replicates were removed. RNA Extraction. B. infantis ATCC was grown at 37 C under anaerobic conditions on 15 ml of ZMB1 media [69], supplemented in different triplicate experiments with 2% of either lactose, glucose, fructooligosaccharides (FOS, raftilose Synergy 1, Orafti, Malvern, PA), galactooligosaccharides (GOS, Purimune, GTC Nutrition, Golden, CO), inulin (raftiline HP, Orafti, Malvern, PA), or purified HMO [70]. Also, lacto-n-tetraose, lacto-n-neotetraose and galacto-n-biose (V-Labs, Covington, LA) were supplemented at 0.5% to the ZMB media. Optical density was assayed using a PowerWave microplate spectrophotometer (BioTek Instruments, Inc., Winoosky, VT). Cultures were recovered at exponential and stationary phase, as defined by previously obtained growth curves. Cells were immediately pelleted and resuspended in 1 ml of RNA later (Ambion, Austin, TX), and stored overnight at 4 C and then at -

57 47! 80 C until use. Bacterial cells were washed twice with PBS buffer, and pre-lysed with 250 µl of 50 mg/ml lysozyme and 120 µl of mutanolysin at 1000 units/ml (Sigma- Aldrich). Total RNA was extracted using the Ambion RNAqueous kit. Briefly, cells were resuspended in 300 µl of Lysis/Binding solution, and vortexed for 10 sec. After adding 300 µl of 64% ethanol, the lysate was transferred to RNAse-free filter cartridges, and washed with 700 µl of wash solution 1, and twice with 500 µl of wash solution 2/3. RNA was eluted with 50 µl of elution solution twice. RNA was further purified using the Qiagen RNeasy kit. 350 µl of Buffer RLT were added to the RNA obtained, and then 250 µl ethanol were also mixed. The sample was applied to an RNeasy column centrifuged. The columns were washed with 500 µl Buffer RPE twice, and the RNA was eluted with 50 µl RNase-free water done twice. Finally, the RNA was concentrated adding 50 µl LiCl to the RNA, and incubated at 70 C for 48 h, washed with cold 70% ethanol and resuspended in 20 µl DNAse/RNAse free water. RNA concentration was evaluated by absorbance at 260 nm (Nanodrop, Thermo Scientific, Wilmington, DE). Acceptable RNA ratios were A260/A280! 1.8, and A260/A230! 1.5. RNA was checked for integrity in an Agilent Bioanalyzer 2100 with the RNA 6000 Nano Kit (Agilent Technologies, Foster City, CA). Quantitative Real-Time PCR. The relative levels of gene expression for each F1SBP gene under growth with different carbon sources were evaluated by quantitative PCR. The gene Blon_0393, encoding a cysteinyl-trna synthetase, was chosen as an endogenous control [71]. Primers and 5 nuclease probes were designed and synthesized by TibMolBiol (Adelphia, NJ) (Table S1). qpcr was performed in a 7500 Fast Real- Time PCR System using the Taqman One-Step RT-PCR kit (Applied Biosystems, Foster

58 48! City, CA). Every reaction contained primers at 0.5 µm, each probe at 125 nm, 5 ng of RNA and Reverse Transcriptase and reaction buffer as recommended by the manufacturer. Gene expression for each gene under the different conditions representing different carbon sources was evaluated by Relative Standard Curves, both for the endogenous control and each gene on each plate. qpcr reactions were run at 48 C for 30 min, then 95 C for 10 min, and 40 cycles with denaturation at 95 C for 15 sec and annealing and elongation for 1 min at 60 C. Threshold cycle data and relative efficiencies were analyzed using the Q-Gene software ( Gene expression levels for growth on lactose were used as the calibrator. Results were expressed as fold changes in gene expression. Proteomics of B. infantis F1SBPs. B. infantis cultures growing under different substrates and used for RNA extraction were also taken at exponential phase, and normalized to an OD 600 of 1.0 by dilution or concentration. Upon centrifugation, 15 ml of each sample were washed three times with PBS vigorously. Cell pellets were resuspended in 600 µl of lysis buffer containing 100 mm Tris and 8.0 M urea. Cells were mechanically lysed using silica beads and a bead-beater (FastPrep, QBiogene, Carlsbad, CA, USA) for eight cycles of 30 s pulses each with a 30 s interval on ice. Beads and cell debris were removed by centrifugation and the soluble fraction was stored at -80 C for further analysis. Protein concentration was measured with the Protein Assay Kit (BioRad, Hercules, CA, USA). A volume of 200 mg/ml of protein was transferred to a new cap tube and precipitated by ethanol (75% (v/v)) at -20 o C. Upon centrifugation, protein pellets were resuspended in 100 µl of 0.1M Tris/1M urea buffer (ph 8.0). Proteins were then digested with 5 µg of mass spectrometry grade trypsin

59 49! (Promega, Madison, WI, USA) overnight at 37 C. The tryptic peptides were purified using Cap Trap Columns for peptide concentration and a Desalting Cartridge (Michrom, Aurburn, CA, USA) according to the manufacturer s manual. The peptides were eluted in 98% acetonitrile in water and then dried prior to mass spectrometry analysis. The digested protein samples were submitted to the Genome Center Proteomics Core at the University of California, Davis. Protein identification was performed using an Eksigent Nano LC 2-D system coupled to an LTQ ion-trap mass spectrometer (Thermo-Fisher, Waltham, MI, USA) using a Picoview nano-spray source. Peptides were loaded onto a nanotrap (Zorbax 300SB-C18, Agilent Technologies, Santa Clara, CA, USA) at a loading flow rate of 5.0 µl/min. Peptides were then eluted from the trap and separated at a level of nano-scale using a 75 µm x 15 cm New Objectives Picofrit Column packed in-house. The top ten ions in each survey scan were subjected to automatic low-energy CID. Tandem mass spectra were extracted and the charge states deconvoluted by BioWorks version 3.3. All MS/MS samples were analyzed using X! Tandem (GPM-XE manager, ver ). X! Tandem was set up to search against the B. infantis ATCC whole proteome. X! Tandem was searched with a fragment ion mass tolerance of 0.60 Da. Oxidation of methionine was specified as a variable modification in X! Tandem. Cutoff of log(e) for the peptide (log(e)) was set < -2 and the protein with the log(e) <-6 was considered to be present. F1SBP in vitro binding to Caco-2 cells. Recombinant F1SBPs were purified as described previously, and labeled using the Fluoreporter FITC Protein Labeling Kit (Molecular Probes) as recommended by the manufacturer. Caco-2 cells were obtained from the American Type Culture Collection, and routinely cultured in DMEM

60 50! supplemented with 10% Fetal Bovine Serum, 10 mm HEPES and 1% Penicillin- Streptomycin (Invitrogen, Carlsbad, CA) on 75 cm 2 flasks and incubated at 37 with 5% CO 2. Cells were routinely passed to new flasks using 0.25% trypsin (Invitrogen) two weeks after confluence. For flow cytometry experiments, trypsinized Caco-2 cells were fixed in 4% paraformaldehyde (PFA, Sigma-Aldrich), and after two washes with PBS, they were incubated with different concentrations of labeled F1SBPs in PBS buffer or UEA-FITC lectin (Vector laboratories, Burlingame, CA) at 37 C for 1 h. Cells were centrifuged at 3000 x g 4 min, and washed four times with PBS-tween 0.05%. Lectinlabeled cells were analyzed by flow cytometry using a Becton Dickinson FACScan Flow Cytometer events for each sample were recorded using FSC, SSC and an argon laser at 488 nm emission. Fluorescence was measured at 515/540 nm excitation, and analyzed using the CellQuest software. For confocal microscopy, Caco-2 cells were seeded in eight-well chamber slides (Nalgene Nunc International, Naperville, IL), and after two weeks post-confluence they were washed with PBS and fixed with PFA 4% for 20 min. After two washes with PBS, monolayers were incubated with different concentrations of FITC-labeled F1SBPs or UEA-FITC lectin, and incubated at 37 C 1 h. After four washes with PBS 1X, slides were permeabilized with 0.1% Triton X-100 and incubated with 0.1 ng/ml 4',6-diamidino-2-phenylindole (DAPI, Invitrogen, Carlsbad, CA), and finally covered with VectaShield Mounting Medium (Vector laboratories, Burlingame, CA). In some cases Caco-2 monolayers were incubated with F1SBPs for 10 or 30 minutes and then cells were fixed and processed as indicated. Slides were analyzed with an Olympus FV1000 Laser Scanning Confocal microscope (Olympus, Melville, NY).

61 51! 3. Results 3.1. Several B. infantis F1SBPs bind host glycans The genome of B. infantis ATCC has twenty genes containing the Pfam motif Sbp_bac_1, Family 1 of Solute Binding Proteins (F1SBPs), with predicted affinity for oligosaccharides [3]. Their abundance in this genome is as much as twice compared to other bifidobacterial genomes (Table 1). The F1SBPs genes from B. infantis ATCC were PCR amplified, removing the lipoprotein anchor sequences, and then cloned, expressed in Escherichia coli and purified as glutathione-s-transferase (GST) fusion proteins. The oligosaccharide binding properties of the recombinant F1SBPs were analyzed using the Mammalian Glycan Array v3.1![28], provided by the Consortium for Functional Glycomics. This version of the array contains more than 400 different glycans mainly present in mammalian cells. Eleven F1SBPs from B. infantis bound at least one carbohydrate in this array, suggesting a broad nutritional preference for host-derived glycans. Table 2 catalogs the glycans detected by these proteins and their relative affinities. The specificities of some F1SBPs narrowed to relatively few substrates. For example, Blon_0375 bound solely to the trisaccharide globotriose (Gal!1-4Gal"1-4Glc), while Blon_2444 was found to have affinity for maltose. Blon_0343 and Blon_2202 share significant homology [27], and they both recognized two related oligosaccharides: Fuc!1-2Gal, the precursor for the ABO blood group, and Fuc!1-2Gal"1-4Glc, 2 Fucosyl lactose (2FL). These structures occur as epitopes in more complex glycans and many were present in the array, however these proteins only recognized their free forms.

62 52! Table 1: Representation of the Pfam01547 motif in bifidobacterial genomes. Sequenced genome Pfam01547 hits Bifidobacterium adolescentis ATCC Bifidobacterium animalis lactis AD011 6 Bifidobacterium animalis subsp. lactis BB-12 5 Bifidobacterium animalis subsp. lactis Bl-04 6 Bifidobacterium bifidum NCIMB (Draft) 3 Bifidobacterium bifidum PRL Bifidobacterium breve DSM (Draft) 16 Bifidobacterium catenulatum DSM (Draft) 11 Bifidobacterium dentium ATCC Bifidobacterium dentium Bd1 23 Bifidobacterium longum DJO10A 16 Bifidobacterium longum NCC Bifidobacterium longum infantis ATCC Bifidobacterium pseudocatenulatum DSM (Draft) 15

63 53! Other F1SBPs showed a broader and more flexible range of glycans recognized in the array. Blon_0883 and Blon_2177 bound type 1 glycans, characterized by the disaccharide Gal!1-3GlcNAc (LNB), as well as Gal!1-3GalNAc (GNB), one of the core substructures in mucin oligosaccharides. Interestingly, Blon_0883 also strongly bound to Lewis a and the type 1 H-trisaccharide, two main epitopes present in glycolipids and glycoproteins in several cell types including the intestinal epithelium [29,30,31]. However, similarly to Blon_0343 and Blon_2202, Blon_0883 did not recognize these structures as part of more complex oligosaccharides. Blon_2177 also bound other type 1 polymers, such as LNT, LNH, lacto-n-octaose (LNO), all abundant structures in HMO [32], as well as other mammalian carbohydrates such as asialo GM1 [33], the putative receptor for the Clostridium difficile toxin TcdA [34], and sialyl-lnt [35], albeit the latter structures were bound at a lower affinity (Table 2).

64 54! Table 2: Oligosaccharides recognized by B. infantis F1SBPs in the Glycan Array v3.1.! F1SBP and structure detected Common Name and Relevance RFU a Blon_0343 Fuca1-2Galb-Sp8 Fuca1-2Galb1-4Glcb-Sp0 H-disaccharide. Precursor of ABO blood group glycans. 2 Fucosyl lactose. Found in HMO, glycolipids and O-linked glycoproteins.! Blon_0375 Gala1-4Galb1-4Glcb-Sp0 Blon_0883 Galb1-3GlcNAcb-Sp8 Globotriose. Glycosphingolipid in several cellular types.! Lacto-N-Biose. Core of type 1 HMO, present in glycolipids and glycoproteins Galb1-3GalNAcb-Sp8 Galb1-3(Fuca1-4)GlcNAc-Sp0 Galacto-N-Biose. One of the main substructures of mucin glycans, also in other glycoproteins. Lewis a. In glycolipids, glycoproteins, and substructure of lacto-n-fucosyl tetraose in HMO Galb1-3GlcNAcb-Sp0 Lacto-N-Biose Galb1-3(Fuca1-4)GlcNAcb-Sp8 Lewis a Fuca1-2Galb1-3GlcNAcb-Sp0 Type I Blood group H-antigen, also a substructure of LNFPI in HMO. Glycolipids and glycoproteins) Fuca1-2Galb1-3GlcNAcb-Sp8 Type I Blood group H-antigen Galb1-3(Fuca1-4)GlcNAcb-Sp8 Lewis Le a Blon_2177 Galb1-3GalNAcb-Sp8 Galacto-N-Biose 24879!

65 55! Galb1-3GlcNAcb1-3Galb1-4Glcb- Sp10 Galb1-3GlcNAcb1-3Galb1-4GlcNAcb-Sp0 Lacto-N-tetraose. One of the main HMO structures. Also in glycolipids and glycoproteins. Dimeric LNB. Substructure of lacto-n-hexaose in HMO. Also a glycolipid Galb1-3GlcNacb1-3(Galb1-3GlcNacb1-3Galb1-4GlcNacb1-6)Galb1-4Glcb- Sp0 Galb1-3GlcNAcb-Sp8 Galb1-3GlcNAcb1-3(Galb1-3GlcNAcb1-3Galb1-4(Fuca1-3)GlcNAcb1-6)Galb1-4Glc- Sp21 Iso-Lacto-N-Octaose. In HMO and glycolipids Lacto-N-biose 8160 Iso-lacto-N-fucosyl-octaose. HMO and glycolipids.! In 5224 Galb1-3Galb1-4GlcNAcb-Sp8 Galactosyl-lactosamine. Candidate target for TcdA toxin Galb1-3GlcNAcb-Sp0 Lacto-N-Biose 3617 Galb1-3GalNAcb1-4Galb1-4Glcb- Sp8 Asialo ganglioside GM1. Important glycolipid in the intestinal mucose, and target for several bacteria or toxins Galb1-3(Neu5Aca2-6)GlcNAcb1-4Galb1-4Glcb- Sp10 Sialyl lacto-n-tetraose. In HMO and as a glycolipid Blon_2202 Fuca1-2Galb-Sp8 H-disaccharide Fuca1-2Galb1-4Glcb-Sp0 2 Fucosyl lactose.! 11865!

66 56! Blon_2344 Gal!1-4GlcNAc!1-3Gal!1-4GlcNAc!1-3Gal!1-4GlcNAc!-Sp0 Gal!1-4GlcNAc!1-3Gal!1-4GlcNAc!-Sp0 Gal!1-4GlcNAc!1-3(Gal!1-4GlcNAc!1-6)Gal!1-4GlcNAc-Sp0 GalNAc"1-3(Fuc"1-2)Gal!1-4GlcNAc!1-3Gal!1-4GlcNAc!1-3Gal!1-4GlcNAc!-Sp0 Gal!1-4GlcNAc!1-3Gal!1-4Glc!- Sp8 Tri-lactosamine. Substructure of Lacto-N-neooctaose in HMO, also a glycolipid.! Di-lactosamine. Substructure of Lacto-N-neohexaose in HMO, found in human colonic mucin, occurs also as a glycolipid) Branched tri-lactosamine. Substructure of Lacto-n-neooctaose in HMO, also in glycoproteins, glycolipids and human colonic mucin; I antigen. Glycan found in human colonic mucin Lacto-n-neotetraose. One of the main HMO structures. Also in glycolipids and glycoproteins Blon_2347 Galb1-4GlcNAcb1-3Galb1-4GlcNAcb1-3Galb1-4GlcNAcb Sp0 Galb1-4GlcNAcb1-3Galb1-4GlcNAcb Sp0 Galb1-4GlcNAcb1-3Galb1-4Glcb Sp8 Galb1-4GlcNAcb1-3(Galb1-4GlcNAcb1-6)Galb1-4GlcNAc-Sp0 Tri-lactosamine Di-lactosamine Lacto-n-neotetraose Branched tri-lactosamine GalNAca1-3(Fuca1-2)Galb1-4GlcNAcb1-3Galb1- Glycan found in human colonic mucin !

67 57! 4GlcNAcb1-3Galb1-4GlcNAcb-Sp0 Galb1-4GlcNAcb1-3(GlcNAcb1-6)Galb1-4GlcNAc-Sp0 Glycan found in human colonic mucin Galb1-4(Fuca1-3)GlcNAcb1-4Galb1-4(Fuca1-3)GlcNAcb-Sp0 Dimeric Lewis x. Also substructure of difucosyl lacto-nneohexaose in HMO, found in glycolipids. Colonic and liver associated antigen. [15] [65] 7559 Blon_2350 Galb1-3GalNAcb-Sp8 Galacto-N-Biose 1352 Blon_2351 Galb1-3GalNAcb-Sp8 Galacto-N-Biose 1106 Blon_2354 Galb1-3GalNAcb-Sp8 Galacto-N-Biose 983 Blon_2444 Glca1-4Glca-Sp8 Maltose Blon_0043 ND c ND Blon_2015 ND ND Blon_2061 ND ND Blon_2352 ND ND Blon_2357 ND ND Blon_2367 ND ND Blon_2380 ND ND Blon_2414 ND ND Blon_2458 ND ND!!

68 58! B. infantis possesses a unique cluster of ABC transporters and glycosyl hydrolases that was predicted to be important in the HMO+ phenotype of this bacterium. This cluster contains 6 F1SBPs closely related at the sequence level [27]. Blon_2344 and Blon_2347 specifically recognized type 2 glycans (Table 2), binding with different affinities linear and branched polylactosamines, as well as lacto-n-neotetraose (LNnT) and difucosyl lacto-n-hexaose. These structures are characteristic in HMO as well as in glycolipids and glycoproteins [36]. These two F1SBPs also presented affinity for complex neutral oligosaccharides found in human colonic mucins [19] and thus are candidates for interaction with epithelial surfaces and mucins. In the same gene cluster, Blon_2350, Blon_2351 and Blon_2354 specifically detected GNB, supporting their role in binding and/or import of mucin oligosaccharides. However, the low affinity of these latter F1SBPs suggests that they may prefer other glycans not present on the array. 3.2 F1SBP expression during growth on HMO If F1SBPs are involved in binding and transport of complex oligosaccharides to be used as a carbon and energy source in situ, one would predict that their expression would be induced by the presence of such glycans. To examine this, B. infantis cells were grown on lactose, HMO, inulin, fructooligosaccharides (FOS) and galactooligosccharides (GOS) as the only carbon sources, and the relative expression level for each F1SBP was studied using qpcr. The presence of these proteins inside the cells was also determined using a proteomic approach. Growth on HMO as the only carbon source induced the expression of four F1SBP genes in the exponential phase, Blon_0883, Blon_2344, Blon_2347 and Blon_2350!

69 59! ( induced was defined as a minimum two-fold increase in gene expression; Figure 1A). Proteomics also confirmed the expression of these proteins (Table 3). These F1SBPs were not expressed in stationary phase (Figure 1E). Given the affinities of these F1SBPs for HMO structures, the expression results suggest that HMO induces F1SBPs involved in its own import. Moreover, the predominant HMO species consumed by B. infantis are consistent with the glycans recognized by F1SBPs expressed during growth on HMO, such as LNnT, LNnH and mono and difucosyl LNH [24]. Since these four F1SBPs also recognized other oligosaccharides present in mucins or glycoconjugates, the presence of HMO likely induces the import of mucin and epithelial glycans and/or facilitates adhesion of B. infantis to mucins or epithelial surfaces. The relative gene expression for a number of HMO and mucin binding F1SBPs was then determined using LNT and lacto-n-neotetraose (LNnT), representative structures of type 1 and type 2 HMO respectively, and GNB as the only available carbon sources. All four F1SBPs induced by HMO (Blon_0883, Blon_2344, Blon_2347 and Blon_2350) showed an increase in their expression during growth on LNT as well as LNnT (Figure 1F) regardless of their own affinity for type 1 or type 2 chains as determined in the glycan array. This indicates that the induction of proteins involved in HMO import does not depend on the isomeric differences between Gal!1-3GlcNAc and Gal!1-4GlcNAc in HMO. Interestingly, Blon_2177 showed an increase in its relative gene expression level with LNT and LNnT. However, this gene was not induced during growth on the HMO pool (Figures 1A and 1E) nor was it detected by shotgun proteomics (Table 3). This is expected from its HMO affinities and its gene context (Figure 2A).!

70 60! Changes in gene expression during growth on GNB were not as dramatic as those induced by growth on LNT or LNnT, however Blon_2344, Blon_2350 and Blon_2351 did exhibit an increase in gene expression greater than two fold. The binding affinities for these three F1SBPs indicated that they might be involved in the import of intact mucin oligosaccharides as well of GNB.!

71 61! FIGURE 1: Gene expression of B. infantis F1SBPs growing under different prebiotics as the sole carbon source. Numbers in the x-axis represent each F1SBP locus tag (Blon_). A: 2% HMO; B: 2% Inulin; C: 2% FOS; D: 2% GOS. Colored bars indicated F1SBPs with a change in gene expression higher than two fold; E: relative expression levels for some HMO-induced genes in exponential (blue bars) and stationary phase (white bars); F: relative expression levels for some F1SBPs under growth with GNB (purple bars), LNT (green bars) and LNnT (blue bars) as the only carbon source. Levels for growth on lactose (white bars) are shown as a reference.!

72 62! Table 3: Proteomic analysis of F1SBPs in B. infantis cells growing under different substrates.! Locus GOS FOS Inulin HMO tag (Blon_) a log(e) b c P uniq d SpC T % e log(e) P uniq SpC T % log(e) P uniq SpC T % log(e) P uniq SpC T % a Only F1SBPs found in the proteome of B. infantis are shown. (!)*+,-./!01*2-34!01*(5(36327!89!:;!<54=->?!! A!B C43D/!5!4C>(-1!*E!C43DC-!0-023=-F!CF-=!E*1!2G-!3=-423E3A523*4!*E!5!01*2-34?!H467!2G-!0-023=-F!I32G!JK#! I-1-!CF-=?!! =!N0O </!5!4C>(-1!*E!PNQPN!F0-A215!5FF3+4-=!E*1!2G-!3=-423E3A523*4!*E!5!01*2-34?!! -!S/!0-1A-4236-!*E!F-DC-4A-!A*T-15+-?!)3F2-=!01*2-34F!G5T-!JM&S!*E!2G-!01*2-34!01*(5(36327!(7!01*2-34! 01*0G-2!56+*132G>!IG-4!2G-!0-023=-F!I32G!JM&S!*E!01*(5(36327!,(7!0-023=-!01*0G-2!56+*132G>.!I-1-! CF-=?!!!

73 63! A model summarizing the complex oligosaccharide binding preferences of B. infantis growing on HMO is shown in Figure 3A. Cells under these conditions could import intact type 1 chains by Blon_2177, or via previous degradation by an extracellular lacto-n-biosidase to LNB to be imported by Blon_0883. Fucosylated or unfucosylated type 2 chains could also be transported in an intact form into the bacterial cell. Some mucin oligosaccharides could be imported directly into the cell, or after degradation by extracellular glycosyl hydrolases that convert these oligosaccharides into GNB. Some blood group epitopes could also be imported after release from more complex carbohydrates. The import of these molecules thereafter potentially provides the cell with monosaccharides which can be used in central metabolic pathways by this bacterium.!

74 64! FIGURE 2: Genetic landscapes for Blon_2177 (A), Blon_0883 (B) and Blon_0343 (C), and close related homologs available in the Integrated Microbial Genome database![72]. Percentages represent similarity of homolog genes to those in the B. infantis ATCC genome. Locus tags are presented above each gene when available.!

75 65! FIGURE 3: Oligosaccharide binding profile proposed for B. infantis cells growing on HMO (A) and inulin (B). Numbers represent each F1SBP locus tag (Blon_).!

76 66! 3.3 F1SBP expression during growth on other prebiotics Growth on the commercial prebiotic inulin induced the expression of seven F1SBPs (Figure 1B), however only four of these proteins were witnessed in the B. infantis proteome (Table 3). Several observations suggest that Blon_2061 is a candidate F1SBP for the import of fructans in B. infantis. Firstly, this F1SBP did not bind any glycan on the mammalian array. Additionally, the neighboring gene Blon_2056 encodes a!-fructofuranosidase, and a prominent homolog, BAD_1330, is located in a fructooligosaccharide-utilization cluster in B. adolescentis (Figure S1). Interestingly, Blon_2061 was the only F1SBP induced by FOS (Figure 1C). This F1SBP does not have a homolog in B. longum subsp. longum DJO10A, which suggests that oligofructose import occurs through different transport systems in both strains. Surprisingly, inulin increased the expression of B. infantis F1SBPs that recognize mammalian glycans, including two proteins present in the HMO Cluster 1 with affinities for GNB, and type 1 chain binding F1SBPs Blon_2177 and Blon_0883 (Figure 2B and Table 2). These findings suggest a role for inulin inducing a cellular response directed to the intestinal environment similar to HMO, which include expression of F1SBPs involved in the import of complex intestinal glycans. Besides fructans, cells growing on inulin potentially import type 1 chains, GNB and some blood group oligosaccharides, which after digestion to mono or disaccharides by intracellular glycosyl hydrolases could be derived to glycolytic pathways (Figure 3B). Growth on GOS induced two different F1SBPs, Blon_2414 and Blon_2458, which did not bind any of the mammalian glycans on the CFG array (Figure 1D). Blon_2414 and Blon_2458 are located close to a putative!-galactosidase (Blon_2416)!

77 67! and an!-galactosidase (Blon_2460), respectively (Figure S1). Blon_2458 is homolog to BL1521, a B. longum subsp. longum NCC2705 F1SBP that was induced by raffinose [37]. Finally, Blon_2444 was found to be induced by maltose (Figure S2), which is in concordance with its affinity shown by the glycan array and a similar induction of a homolog, BL0141, by maltose in B. longum subsp. longum [37]. 3.4 In vitro interactions between F1SBPs and epithelial cells. Given that several of the F1SBPs exhibited adherence to glycans commonly found on the epithelial cells, we hypothesized that these binding proteins might also interact with intestinal epithelial surfaces. To test this, B. infantis F1SBPs with affinity for mammalian glycans (Table 2) were labeled with FITC and coincubated with differentiated Caco-2 cells, a human colonic adenocarcinoma cell line. Using flow cytometry, a clear shift in the mean fluorescence value was observed when cells were incubated with Blon_2347 and Blon_0375 (Figures 4B-D), as well as with Ulex Europeus Agglutinin lectin (UEA). This lectin is characterized by binding oligosaccharides containing the Fuc!1-2Gal"1-4GlcNAc epitope. No fluorescence was detected when cells were coincubated with other F1SBPs or FITC-streptavidin. These observations were confirmed using confocal microscopy (Figures 4E-I). Blon_2347 and Blon_0375 were found in the cell membrane of fixed Caco-2 cells. When cells were fixed after coincubation with Blon_0375, this particular protein was observed in the cytoplasm of Caco-2 cells. (Figure 4H-I). Since this protein binds globotriose, a glycolipid abundant in some cancer cell lines, it could be hypothesized that its internalization is mediated by this glycolipid in cells expressing it in their cell membranes.!

78 68! 4. Discussion In this work we have functionally characterized a family of proteins in B. infantis predicted to import oligosaccharides as part of ABC importers. Compared to related bifidobacterial genomes these genes seem to be overrepresented in this subspecies (Table 1). Using an array containing oligosaccharides abundant on mammalian cells, we determined that 10 out of 20 F1SBPs B. infantis have affinity for mammalian glycans, including characteristic HMO structures as well as intestinal glycoconjugates. 4.1 Patterns of F1SBP binding to host oligosaccharides. We observed two major trends in glycan binding. Blon_0343, Blon_2202, Blon_0375, Blon_2350, Blon_2351 and Blon_2354 were specific for only one or two oligosaccharides in the glycan array (Table 2). Considering that these substrates were not recognized by the same F1SBPs when they were present at terminal or internal positions in more complex oligosaccharides, it is likely that these proteins require their ligands to be in their free forms. It is possible that the reducing end in these carbohydrates might be critical for F1SBP binding. With the exception of Blon_0375, none of the F1SBPs listed above were able to bind epithelial Caco-2 cells, which are characterized by glycoconjugates containing these epitopes (Table 2 and Table S2). This idea supports the role of these F1SBPs in oligosaccharide import associated to ABC transporters, and suggests a requirement on foreign extracellular glycosyl hydrolases for their release from intestinal glycoconjugates! and subsequent use as a carbon source [38,39,40].!

79 69! FIGURE 4: Binding of F1SBPs to Caco-2 cells in vitro detected by flow cytometry (A-D) and confocal microcopy (E-F). A: FITC-streptavidin; B and E: UEA-FITC 0.5 µg/ml; C: Blon_0375- FITC 1 µg/ml; D and F: Blon_2347-FITC 1 µg/ml; G: Blon_2347-FITC 1 µg/ml and DAPI (blue); H: fixed Caco-2 cells coincubated with 1 µg/ml Blon_0375-FITC; I: Caco-2 cells incubated for 10 min with 1 µg/ml Blon_0375-FITC and stained with DAPI.!

80 70! In contrast, Blon_0883, Blon_2177, Blon_2344 and Blon_2347 exhibited a more flexible capacity to bind either type 1 or type 2 chains on the glycan array. Because of this variable substrate binding capacity it is possible to discern a core glycan epitope interacting with the F1SBP binding pocket. An example is Blon_2177, which binds the disaccharide Gal!1-3HexNAc. Additional LNB subunits can be attached to this disaccharide structure, resulting however in a decrease in the RFU values for this protein. Blon_0883 also prefers Gal!1-3HexNAc, accepting a fucose residue in either end of the disaccharide (Table S2). Blon_2344 and Blon_2347 both bind dilactosamine (Gal!1-4GlcNAc!1-3Gal!1-4GlcNAc!) as a core structure. Modifications to this tetrasaccharide, such as additional lactosamine units, fucose residues or replacing the terminal GlcNAc with glucose did not inhibit binding. Additional studies are needed to determine to what extent the substrates bound by F1SBPs are internalized into the cell and how this interaction occurs at the structural level. 4.2 HMO import in B. infantis ATCC Blon_0883 and Blon_2177, the latter present in the LNB/GNB gene cluster, were determined to bind type 1 glycans. In the glycan array Blon_0883 had affinity for LNB as well as GNB, and it was induced by HMO, LNT and LNnT. Blon_0883 is located in a gene cluster that includes two enzymes predicted to participate in hexosamine metabolism, glucosamine-6-p isomerase and glucosamine deacetylase (Figure 2B). Blon_0883 does not contain a homolog in any other bifidobacterial genome sequenced to date. Homologs are found in other members of the intestinal microbiota such as Eubacterium rectale and Faecalibacterium prausnitzii.!

81 71! The gene clusters in these microorganisms also contain homologs of hexosamine metabolizing enzymes, as well as a LNB phosphorylase, a key enzyme in the metabolism of type 1 chains [21]. Interestingly, B. longum DJO10A and NCC2705 possess a structurally similar gene cluster containing enzymes related to hexosamine metabolism (Figure 2B), however the transporter protein is a homolog to OppA proteins, predicted to import dipeptides (Family 5 of SBPs). As has been described previously for pentose utilization [41], this difference is another example of how B. infantis has adapted subspecies-specific mechanisms for HMO and/or mucin metabolism, but which have diverged for alternative purposes in the subspecies longum. In concordance with this idea, only two of the ten F1SBPs with affinity with host glycans have affinity to proteins in both B. longum sequenced genomes. Blon_2177 is homologous to BL1638 in B. longum NCC2705. Both are located in the LNB/GNB gene cluster that is well conserved among Bifidobacterium species residing in the infant gastrointestinal tract (Figure 2A). Homologs to this F1SBP are present in all B. longum and B. infantis strains [42]. While BL1638 showed similar K d values for GNB and LNB [43], Blon_2177 seems to prefer GNB three times more than LNB, being also able to bind other complex structures such as LNT and LNH. These structural differences may help explain the broader range of HMO structures B. infantis is able to consume. Blon_2177 expression was only detected when cells grew on LNT and LNnT, and similarly to B. longum LMG [44], the F1SBP in this cluster was induced by inulin. Contrary to what might be predicted given its binding patterns, Blon_2177 was not expressed during growth on HMO as indicated by qrt-pcr or proteomics (Table!

82 72! 3). It is possible that the LNT concentration in the pooled HMO preparation is not enough to induce the expression of this protein, or certain HMO species repress its expression. Considering that both Blon_2177 and Blon_0883 bind type 1 chains, Blon_0883 might have a more active role in the import of these structures. Blon_2177 could therefore be critical in the import of non-fucosylated type 1 oligosaccharides, as well as GNB (as a product of mucin glycan degradation). In order to import type 1 HMO via Blon_0883, B. infantis needs to deconstruct these glycans to the disaccharide LNB. Since a lacto-n-biosidase gene has not been found in the B. infantis ATCC genome, it is possible that B. infantis relies on extracellular degradation of type 1 HMO by other intestinal bacteria. The ability to ferment LNB seems to be present among relatively few bifidobacteria [45], and the import of this substrate has been associated so far with a single F1SBP in the LNB/GNB cluster already described. In B. infantis two F1SBPs were found to bind LNB and five bind GNB (Table 2). This suggests a particular abundance of these two core structures in the infant colon, as well as a selective pressure for mechanisms to internalize them. The potential redundancy in these affinities correlates with the absence of certain F1SBPs related to HMO import in some B. infantis strains [42]. LNT is the main HMO isomer preferred by infant-borne bifidobacteria [24]. B. bifidum was recently determined to possess enzymes deconstructing type 2 chains such as LNnT [46], however the role of these glycosyl hydrolases has not been tested in vivo or their expression evaluated at the RNA or protein level. The B. infantis ATCC genome contains a unique gene cluster containing several transport related proteins and glycosyl hydrolases [27]. However as shown here, the F1SBPs in this!

83 73! cluster have overlapping affinities, and bind different complex type 2 chains. They can also interact with intact mucin oligosaccharides as well as the core mucin structure GNB. It is probable that this entire cluster is devoted to the deconstruction of this type of isomers, while the aforementioned clusters Blon_0883 and Blon_2177 participate in type 1 chain metabolism. These findings are in agreement with the preference for the import of complex substrates for intracellular degradation in B. infantis![26], contrasting with B. breve and B. bifidum, which seem to release extracellular glycosyl hydrolases to break down type 1 oligosaccharides to LNB and, potentially, metabolize type 2 oligosaccharides [26,27]. One might predict that this strategy would be applicable for structurally similar mucin and glycolipid oligosaccharides as well Impact of F1SBP expression under growth on HMO or inulin. While the binding behavior clearly predicted the types of oligosaccharides that interact with a given F1SBP and associated transport systems, the expression of these F1SBPs was not strictly governed by growth on the cognate oligosaccharide partners. Indeed, growth of B. infantis on inulin and HMO induced both predicted transporters involved in their own import as well as other, putatively unrelated F1SBPs. As depicted in Figure 3B, F1SBPs induced by inulin, which is a fructose-based oligosaccharide, import GNB into the cell, as well as type 1 chains and blood group glycans. HMO-induced F1SBPs were determined to bind HMO structures, but they can also import mucin oligosaccharides, and blood group structures. These observations suggest an adaptive response to two prebiotic substrates that reach the human colon, and prepare bifidobacteria to encounter other host-derived glycans. If so, an induction of other genes involved in host!

84 74! oligosaccharide metabolism, such as glycosyl hydrolases, could be expected. F1SBPs induced by HMO also showed binding to epithelial cells. If their binding is one of the factors involved in B. infantis adhesion to intestinal cells or mucins, HMO consumption could indirectly help in bacterial attachment because of F1SBPs dual function in the import and binding of intestinal oligosaccharides. Despite their similarities, growth on inulin resulted in induction of a larger number of F1SBP than growth on FOS. This is a surprising result given these fructans differ mostly in their degree of polymerization (DP) (FOS, 2-7; inulin, 10-32). Inulin is also characterized by branched chains. Shorter chain FOS and GOS appeared to induce F1SBPs involved only in their import, in contrast to HMO and inulin which induced additional F1SBPs related to utilization, or binding, of intestinal glycans. These results suggest that different prebiotics may elicit different binding and colonization behaviors in commensal, or probiotic, bacteria in situ. For some F1SBPs which affinities were determined in the glycan array, the particular genes were not induced by any of the substrates used. Blon_0343 and Blon_2202 showed affinity for 2FL and the H-disaccharide, however they were not expressed either at the transcript or protein level under the conditions studied. The localization of a fuconate dehydratase in the vicinity of Blon_0343 supports the idea of their role in the import of certain fucose containing oligosaccharides (Figure 2C). The homolog of Blon_2015 (BL1163; Fig. S1) in B. longum NCC2705 was induced by lactose and GOS, however this was not observed for B. infantis [37]. Further experiments are required to understand the role of some F1SBPs which affinities or the conditions when they are expressed are not clear. For example, Blon_2380 is!

85 75! homologous to BL1330, a protein that was suggested to participate in the import of mannose containing oligosaccharides! [37]. Blon_0043 possesses several homologs in adult-type bifidobacteria, and it is located in the same region that Blon_0047, which is a putative xylan esterase. F1SBPs and their role in host-bacterial interactions. HMO share structural similarity with intestinal glycoconjugates and they prevent the binding of several pathogens to intestinal surfaces [47,48,49]. One competitive advantage for B. infantis may be the ability to both bind and consume HMO and other intestinal glycoconjugates. Thus expression of specific F1SBPs in B. infantis may play a role in acquisition of food substrates as well as assist in adherence to glycoconjugates on epithelial surfaces. SBPs from certain microorganisms have been found to be associated with bacterial adhesion [50,51]. Cultured epithelial cells have been shown to be mostly unresponsive to B. infantis [52]. However, B. infantis is able to downregulate the cytokine induction profile from some gastrointestinal pathogens [53]. B. infantis has also been shown to secrete products that enhance epithelial cell function [54]. Candidates for adhesion and interaction with the epithelium found in other bifidobacteria or lactobacilli such as glycoprotein binding fimbriae or mucus binding proteins were not observed in the genome of B. infantis ATCC [27]. F1SBPs binding specific glycoconjugates in epithelial cells remain as candidates for these functions. Some oligosaccharides detected by F1SBPs are commonly found in epithelial cell membranes or occur as glycoconjugates in mucins. For example, Lewis a is one of the most abundant glycolipids in intestinal epithelial cells![55], and it was one of the ligands bound by Blon_0883. Polylactosamines are also!

86 76! common structures abundant in the complex type of N-linked proteins, as well as in glycolipids, especially in erythrocytes and neutrophiles [56]. Blon_2344 and Blon_2347 possessed an exquisite affinity for these structures and might be important in hostbacterial interactions. Polylactosamines are also one of the binding targets for several pathogens, such as Helicobacter pylori [57], and the heat-labile enterotoxin of E. coli [58]. Therefore, the expression of these particular F1SBPs might prevent in some cases bacterial colonization or toxin intoxication via competitive exclusion. In the glycan array, Blon_0375 was highly specific for globotriose (Gb3), a trisaccharide usually conjugated to ceramide, and it appeared to be internalized by a discrete population of Caco-2 cells. This glycan has been found at low concentrations in human milk![59] and is characteristic of the urothelial epithelium [60] and human fetal meconium [61]. Blon_0375 binding Gb3 might provide a competitive advantage to this bacterium, and suggests that this oligosaccharide must be abundant in the environment where B. infantis resides. This protein could also be important in bacterial adhesion and interaction with intestinal epithelial cells. Gb3 is not abundant in adult epithelial cells, and no detailed description exists on the abundance of this glycolipid in the human intestinal mucosa in the first months of life. However, Gb3 is highly expressed in several cancer cell lines, and in particular during metastasis. Globotriose is also the target for the B-subunit of the Shiga toxin! [62], and it has been successfully studied as an effector against tumors in vivo [63,64]. Bifidobacteria are well-adapted microorganisms that cohabitate the lower section of the gastrointestinal tract, to which several health benefits have been attributed [65,66]. As has been observed with some Bacteroides species, bifidobacteria have been!

87 77! characterized by their ability to utilize complex oligosaccharides, mainly plant-derived but also host-derived such as HMO [67,68]. The existence of discrete molecular mechanisms for the import of these complex structures in bifidobacteria suggests an evolutionary adaptation to the presence of these indigestible dietary oligosaccharides in the human colon. Indeed, the induction of some common F1SBPs between inulin and HMO suggests a common regulatory mechanism. Since genomic analysis suggests B. infantis has displaced some plant-associated sugar catabolism genes with genes for milk sugars [41] it could be argued that the regulation of HMO consumption has simply adapted from an existing inulin-based regulon. Defining the details of the intricate tripartite network between human milk, the developing infant intestinal epithelium and the intestinal microbiota is critical for understanding the biology of the microbial foundation process occurring in the first years of life in the infant gut, as well as for the improvement of infant formulas. In particular, understanding the nutritional preferences for complex oligosaccharides for members of the microbiota will help in the future in the selection of specific desirable probiotic strains, as well as a change towards more specific prebiotic compounds.!

88 78! REFERENCES 1. Schell MA, Karmirantzou M, Snel B, Vilanova D, Berger B, et al. (2002) The genome sequence of Bifidobacterium longum reflects its adaptation to the human gastrointestinal tract. Proc Natl Acad Sci U S A 99: Xu J, Bjursell MK, Himrod J, Deng S, Carmichael LK, et al. (2003) A genomic view of the human-bacteroides thetaiotaomicron symbiosis. Science 299: Tam R, Saier MH, Jr. (1993) Structural, functional, and evolutionary relationships among extracellular solute-binding receptors of bacteria. Microbiol Rev 57: Haarman M, Knol J (2005) Quantitative real-time PCR assays to identify and quantify fecal Bifidobacterium species in infants receiving a prebiotic infant formula. Appl Environ Microbiol 71: Eckburg PB, Bik EM, Bernstein CN, Purdom E, Dethlefsen L, et al. (2005) Diversity of the human intestinal microbial flora. Science 308: Gibson GR, Probert HM, Loo JV, Rastall RA, Roberfroid MB (2004) Dietary modulation of the human colonic microbiota: updating the concept of prebiotics. Nutr Res Rev 17: Bouhnik Y, Raskine L, Simoneau G, Paineau D, Bornet F (2006) The capacity of short-chain fructo-oligosaccharides to stimulate faecal bifidobacteria: a doseresponse relationship study in healthy humans. Nutr J 5: Barboza M, Sela DA, Pirim C, Locascio RG, Freeman SL, et al. (2009) Glycoprofiling bifidobacterial consumption of galacto-oligosaccharides by mass spectrometry reveals strain-specific, preferential consumption of glycans. Appl Environ Microbiol 75: Ramirez-Farias C, Slezak K, Fuller Z, Duncan A, Holtrop G, et al. (2009) Effect of inulin on the human gut microbiota: stimulation of Bifidobacterium adolescentis and Faecalibacterium prausnitzii. Br J Nutr 101: Favier CF, de Vos WM, Akkermans AD (2003) Development of bacterial and bifidobacterial communities in feces of newborn babies. Anaerobe 9: Kleessen B, Bunke H, Tovar K, Noack J, Sawatzki G (1995) Influence of two infant formulas and human milk on the development of the faecal flora in newborn infants. Acta Paediatr 84: Adlerberth I, Wold AE (2009) Establishment of the gut microbiota in Western infants. Acta Paediatr 98: Ventura M, Elli M, Reniero R, Zink R (2001) Molecular microbial analysis of Bifidobacterium isolates from different environments by the species-specific amplified ribosomal DNA restriction analysis (ARDRA). FEMS Microbiol Ecol 36: Ventura M, O'Connell-Motherway M, Leahy S, Moreno-Munoz JA, Fitzgerald GF, et al. (2007) From bacterial genome to functionality; case bifidobacteria. Int J Food Microbiol 120: Kunz C, Rudloff S, Baier W, Klein N, Strobel S (2000) Oligosaccharides in human milk: structural, functional, and metabolic aspects. Annu Rev Nutr 20: !

89 79! 16. Ninonuevo MR, Park Y, Yin H, Zhang J, Ward RE, et al. (2006) A strategy for annotating the human milk glycome. J Agric Food Chem 54: Finne J, Breimer ME, Hansson GC, Karlsson KA, Leffler H, et al. (1989) Novel polyfucosylated N-linked glycopeptides with blood group A, H, X, and Y determinants from human small intestinal epithelial cells. J Biol Chem 264: Wu AM (2003) Carbohydrate structural units in glycoproteins and polysaccharides as important ligands for Gal and GalNAc reactive lectins. J Biomed Sci 10: Podolsky DK (1985) Oligosaccharide structures of isolated human colonic mucin species. J Biol Chem 260: Ward RE, Ninonuevo M, Mills DA, Lebrilla CB, German JB (2007) In vitro fermentability of human milk oligosaccharides by several strains of bifidobacteria. Mol Nutr Food Res 51: Kitaoka M, Tian J, Nishimoto M (2005) Novel putative galactose operon involving lacto-n-biose phosphorylase in Bifidobacterium longum. Appl Environ Microbiol 71: Ashida H, Miyake A, Kiyohara M, Wada J, Yoshida E, et al. (2009) Two distinct alpha-l-fucosidases from Bifidobacterium bifidum are essential for the utilization of fucosylated milk oligosaccharides and glycoconjugates. Glycobiology 19: Nishimoto M, Kitaoka M (2007) Identification of N-acetylhexosamine 1-kinase in the complete lacto-n-biose I/galacto-N-biose metabolic pathway in Bifidobacterium longum. Appl Environ Microbiol 73: LoCascio RG, Ninonuevo MR, Freeman SL, Sela DA, Grimm R, et al. (2007) Glycoprofiling of bifidobacterial consumption of human milk oligosaccharides demonstrates strain specific, preferential consumption of small chain glycans secreted in early human lactation. J Agric Food Chem 55: Turroni F, Bottacini F, Foroni E, Mulder I, Kim JH, et al. (2010) Genome analysis of Bifidobacterium bifidum PRL2010 reveals metabolic pathways for host-derived glycan foraging. Proc Natl Acad Sci U S A 107: Zivkovic AM, German JB, Lebrilla CB, Mills DA (2010) Microbes and Health Sackler Colloquium: Human milk glycobiome and its impact on the infant gastrointestinal microbiota. Proc Natl Acad Sci U S A., in press. 27. Sela DA, Chapman J, Adeuya A, Kim JH, Chen F, et al. (2008) The genome sequence of Bifidobacterium longum subsp. infantis reveals adaptations for milk utilization within the infant microbiome. Proc Natl Acad Sci U S A 105: Blixt O, Head S, Mondala T, Scanlan C, Huflejt ME, et al. (2004) Printed covalent glycan array for ligand profiling of diverse glycan binding proteins. Proc Natl Acad Sci U S A 101: Hakomori SI (2008) Structure and function of glycosphingolipids and sphingolipids: recollections and future trends. Biochim Biophys Acta 1780: Holgersson J, Jovall PA, Breimer ME (1991) Glycosphingolipids of human large intestine: detailed structural characterization with special reference to blood group compounds and bacterial receptor structures. J Biochem 110: !

90 80! 31. Morgan WT, Watkins WM (1969) Genetic and biochemical aspects of human bloodgroup A-, B-, H-, Le-a- and Le-b-specificity. Br Med Bull 25: Kogelberg H, Piskarev VE, Zhang Y, Lawson AM, Chai W (2004) Determination by electrospray mass spectrometry and 1H-NMR spectroscopy of primary structures of variously fucosylated neutral oligosaccharides based on the iso-lacto-n-octaose core. Eur J Biochem 271: Keranen A (1975) Gangliosides of the human gastrointestinal mucosa. Biochim Biophys Acta 409: Krivan HC, Clark GF, Smith DF, Wilkins TD (1986) Cell surface binding site for Clostridium difficile enterotoxin: evidence for a glycoconjugate containing the sequence Gal alpha 1-3Gal beta 1-4GlcNAc. Infect Immun 53: Asakuma S, Akahori M, Kimura K, Watanabe Y, Nakamura T, et al. (2007) Sialyl oligosaccharides of human colostrum: changes in concentration during the first three days of lactation. Biosci Biotechnol Biochem 71: Hakomori S (1999) Antigen structure and genetic basis of histo-blood groups A, B and O: their changes associated with human cancer. Biochim Biophys Acta 1473: Parche S, Amon J, Jankovic I, Rezzonico E, Beleut M, et al. (2007) Sugar transport systems of Bifidobacterium longum NCC2705. J Mol Microbiol Biotechnol 12: Larson G, Watsfeldt P, Falk P, Leffler H, Koprowski H (1987) Fecal excretion of intestinal glycosphingolipids by newborns and young children. FEBS Lett 214: Ruas-Madiedo P, Gueimonde M, Fernandez-Garcia M, de los Reyes-Gavilan CG, Margolles A (2008) Mucin degradation by Bifidobacterium strains isolated from the human intestinal microbiota. Appl Environ Microbiol 74: Wada J, Ando T, Kiyohara M, Ashida H, Kitaoka M, et al. (2008) Bifidobacterium bifidum lacto-n-biosidase, a critical enzyme for the degradation of human milk oligosaccharides with a type 1 structure. Appl Environ Microbiol 74: Sela DA, Mills DA (2010) Nursing our microbiota: molecular linkages between bifidobacteria and milk oligosaccharides. Trends Microbiol 18: Locascio RG, Desai P, Sela DA, Weimer B, Mills DA (2010) Broad conservation of milk utilization genes in Bifidobacterium longum subsp. infantis as revealed by comparative genomic hybridization. Appl Environ Microbiol 76: Suzuki R, Wada J, Katayama T, Fushinobu S, Wakagi T, et al. (2008) Structural and thermodynamic analyses of solute-binding Protein from Bifidobacterium longum specific for core 1 disaccharide and lacto-n-biose I. J Biol Chem 283: Gonzalez R, Klaassens ES, Malinen E, de Vos WM, Vaughan EE (2008) Differential transcriptional response of Bifidobacterium longum to human milk, formula milk, and galactooligosaccharide. Appl Environ Microbiol 74: Xiao JZ, Takahashi S, Nishimoto M, Odamaki T, Yaeshima T, et al. (2010) Distribution of in vitro fermentation ability of lacto-n-biose I, a major building block of human milk oligosaccharides, in bifidobacterial strains. Appl Environ Microbiol 76: !

91 81! 46. Miwa M, Horimoto T, Kiyohara M, Katayama T, Kitaoka M, et al. (2010) Cooperation of {beta}-galactosidase and {beta}-n-acetylhexosaminidase from bifidobacteria in assimilation of human milk oligosaccharides with type 2 structure. Glycobiology 20: Coppa GV, Zampini L, Galeazzi T, Facinelli B, Ferrante L, et al. (2006) Human milk oligosaccharides inhibit the adhesion to Caco-2 cells of diarrheal pathogens: Escherichia coli, Vibrio cholerae, and Salmonella fyris. Pediatr Res 59: Ruiz-Palacios GM, Cervantes LE, Ramos P, Chavez-Munguia B, Newburg DS (2003) Campylobacter jejuni binds intestinal H(O) antigen (Fuc alpha 1, 2Gal beta 1, 4GlcNAc), and fucosyloligosaccharides of human milk inhibit its binding and infection. J Biol Chem 278: Chaturvedi P, Warren CD, Altaye M, Morrow AL, Ruiz-Palacios G, et al. (2001) Fucosylated human milk oligosaccharides vary between individuals and over the course of lactation. Glycobiology 11: Guglielmetti S, Tamagnini I, Mora D, Minuzzo M, Scarafoni A, et al. (2008) Implication of an outer surface lipoprotein in adhesion of Bifidobacterium bifidum to Caco-2 cells. Appl Environ Microbiol 74: Higgins MA, Abbott DW, Boulanger MJ, Boraston AB (2009) Blood group antigen recognition by a solute-binding protein from a serotype 3 strain of Streptococcus pneumoniae. J Mol Biol 388: O'Hara AM, O'Regan P, Fanning A, O'Mahony C, Macsharry J, et al. (2006) Functional modulation of human intestinal epithelial cell responses by Bifidobacterium infantis and Lactobacillus salivarius. Immunology 118: Candela M, Perna F, Carnevali P, Vitali B, Ciati R, et al. (2008) Interaction of probiotic Lactobacillus and Bifidobacterium strains with human intestinal epithelial cells: adhesion properties, competition against enteropathogens and modulation of IL-8 production. Int J Food Microbiol 125: Ewaschuk JB, Diaz H, Meddings L, Diederichs B, Dmytrash A, et al. (2008) Secreted bioactive factors from Bifidobacterium infantis enhance epithelial cell barrier function. Am J Physiol Gastrointest Liver Physiol 295: G Holgersson J, Stromberg N, Breimer ME (1988) Glycolipids of human large intestine: difference in glycolipid expression related to anatomical localization, epithelial/non-epithelial tissue and the ABO, Le and Se phenotypes of the donors. Biochimie 70: Macher BA, Klock JC (1980) Isolation and chemical characterization of neutral glycosphingolipids of human neutrophils. J Biol Chem 255: Miller-Podraza H, Lanne B, Angstrom J, Teneberg S, Milh MA, et al. (2005) Novel binding epitope for Helicobacter pylori found in neolacto carbohydrate chains: structure and cross-binding properties. J Biol Chem 280: Orlandi PA, Critchley DR, Fishman PH (1994) The heat-labile enterotoxin of Escherichia coli binds to polylactosaminoglycan-containing receptors in CaCo-2 human intestinal epithelial cells. Biochemistry 33: Newburg DS, Ashkenazi S, Cleary TG (1992) Human milk contains the Shiga toxin and Shiga-like toxin receptor glycolipid Gb3. J Infect Dis 166: !

92 82! 60. Stromberg N, Marklund BI, Lund B, Ilver D, Hamers A, et al. (1990) Host-specificity of uropathogenic Escherichia coli depends on differences in binding specificity to Gal alpha 1-4Gal-containing isoreceptors. EMBO J 9: Karlsson KA, Larson G (1981) Molecular characterization of cell surface antigens of fetal tissue. Detailed analysis of glycosphingolipids of meconium of a human O Le(a--b+) secretor. J Biol Chem 256: Lindberg AA, Brown JE, Stromberg N, Westling-Ryd M, Schultz JE, et al. (1987) Identification of the carbohydrate receptor for Shiga toxin produced by Shigella dysenteriae type 1. J Biol Chem 262: Falguieres T, Maak M, von Weyhern C, Sarr M, Sastre X, et al. (2008) Human colorectal tumors and metastases express Gb3 and can be targeted by an intestinal pathogen-based delivery tool. Mol Cancer Ther 7: Kovbasnjuk O, Mourtazina R, Baibakov B, Wang T, Elowsky C, et al. (2005) The glycosphingolipid globotriaosylceramide in the metastatic transformation of colon cancer. Proc Natl Acad Sci U S A 102: McCarthy J, O'Mahony L, O'Callaghan L, Sheil B, Vaughan EE, et al. (2003) Double blind, placebo controlled trial of two probiotic strains in interleukin 10 knockout mice and mechanistic link with cytokine balance. Gut 52: O'Mahony L, McCarthy J, Kelly P, Hurley G, Luo F, et al. (2005) Lactobacillus and bifidobacterium in irritable bowel syndrome: symptom responses and relationship to cytokine profiles. Gastroenterology 128: Crociani F, Alessandrini A, Mucci MM, Biavati B (1994) Degradation of complex carbohydrates by Bifidobacterium spp. Int J Food Microbiol 24: Lee JH, O'Sullivan DJ (2010) Genomic insights into bifidobacteria. Microbiol Mol Biol Rev 74: Zhang G, Mills DA, Block DE (2009) Development of chemically defined media supporting high-cell-density growth of lactococci, enterococci, and streptococci. Appl Environ Microbiol 75: Ward RE, Ninonuevo M, Mills DA, Lebrilla CB, German JB (2006) In vitro fermentation of breast milk oligosaccharides by Bifidobacterium infantis and Lactobacillus gasseri. Appl Environ Microbiol 72: Parche S, Beleut M, Rezzonico E, Jacobs D, Arigoni F, et al. (2006) Lactose-overglucose preference in Bifidobacterium longum NCC2705: glcp, encoding a glucose transporter, is subject to lactose repression. J Bacteriol 188: Markowitz VM, Korzeniewski F, Palaniappan K, Szeto E, Werner G, et al. (2006) The integrated microbial genomes (IMG) system. Nucleic Acids Res 34: D !

93 83! ACKNOWLEDGMENTS The authors want to thank to the Consortium of Functional Glycomics (Core H), Dr. David F. Smith and Dr. Jamie Heimburg-Molinaro for their collaboration in this work and support on the Glycan Array experiments, Dr. Carlito Lebrilla for reading the manuscript and providing us with the Glycan Array platform, Dr. David Sela and Dr. Riccardo LoCascio for their contributions and assistance throughout this work, and Dr. Bart Weimer for providing an improved RNA extraction protocol.!

94 84! SUPPLEMENTARY INFORMATION Figure S1: Genetic landscapes for other F1SBP clusters. A: Blon_2015; B: Blon_2061; C: Blon_2414; D: Blon_2468.!

95 85! Figure S2: Fold change in gene expression of Blon_2444 in B. infantis cells grown in maltose (2%) relative to lactose.!

96 86! Table S1: Primers used in this study. Primers used in F1SBP cloning Primer Sequence (5-3 ) BL0043F a ATATCCCGGGCAGCGACACCGCGCAAGATGA BL0043R ATATCTCGAGTCAGTTGGTCTTGACGCCCATGTCTT BL0343F ATATGTCGACGCTCTGGTACTTCGCAGAAAAACAA BL0343R ATATGCGGCCGCTTTCAGTCTGCGTCAGTGGTGACTTTA BL0375F ATATGTCGACGCGGGAAAACCAAGATTTCGTTCT BL0375R ATATGCGGCCGCAGTTACTTCAGCGTCGCC BL0883F ATATGAATTCTCCGATGGCGGCAAGACCACACTCAAATTC BL0883R ATATCTCGAGTTACTGCTTGGCCGCCGCGTTCA BL2015w TGCGAGTGATCGGACAGCGAATC BL2015F ATATCTCGAGGGACGACAACCGTACGGAGATC BL2015R ATATCGGCCGCTACTACTTAATCGTAAAGCCC BL2061F ATATCCCGGGCGAGAACGGCAAGCCAATCGTCAA BL2061R ATATGCGGCCGCCTCTACTTGGTGTACTTGTCGTAC BL2177F ATATGAATTCACGAAGTCCGGCAGCGATGGCGG BL2177R ATATGCGGCCGCCTTCACTCCTTGACGGACAGAC BL2202F ATATGTCGACGCGATGTGACCGCGCAGGACGT BL2202R ATATCTCGAGTCAGTCGGCGTCGGTGGTGACCTTG BL2344w ATGAGAAGAACCGCGATGAGG BL2344F ATATGTCGACGCGAAGGAACCCAATCAGGACAAGA BL2344R ATATGCGGCCGCTCACTTCTTCGTGTAGATGTCGTA BL2347F ATATGAATTCGACGGCAAGCCGATCGTGAGCGTTC BL2347R ATATGCGGCCGCCTTCACTTCTTCGTGTAGATGTCG BL2347w CGCACTGAAGGCGGGCGCCATCAC BL2350F ATATGTCGACATGGCAAGCCGATTGTAAGTGTCTT BL2350R ATATCTCGAGTCACTTCGTGTAGGTGTCGTACCAC BL2350w GCTGTCTGCGTGCGGCGGCGA BL2351F ATATGAATTCGGCAAGCCGATTGTGACGGTTCTGGTC BL2351R ATATGCGGCCGCCTTCACTTCGTGTAGGTGTCGT BL2351w GTGAAGGCGGGTGCCGTTGCGTGC BL2352F ATATGTCGACACGGCAAGCCGATTGTGACGGTTCTGGTCA BL2352R ATATGCGGCCGCCTTCACTTTACGCAAAGGTCGTACCA BL2352w GTGAAGGCGGGTGCCGTTGTGTGT BL2354F ATATGAATTCGGCAAGCCGATTGTGACGGTTCTGGTC BL2354R ATATGCGGCCGCCTTCACTTTACGCAAAGGTCGTA BL2354w CGATGAAGGTGGGTGCTGCGGTA BL2357F ATATGAATTCGATGGTGGACAATCCGACAAGATCGTCTCC BL2357R ATATCTCGAGTCACTGTTCCCAATCCGAATCGAACGTC BL2367F ATATGAATTCGTCCCTACGCCCAAAGAGTCCGACGGTT BL2367R ATATCTCGAGTCACTTCTGGGACAGCGAGTTATTGTAA BL2380F ATATGTCGACACGGTAAGAAAGAAGTCTCCTTCCAGACCT BL2380R ATATCTCGAGTCATGAGTTCAAGTCCTCATTGGCGATC BL2414F ATATCCCGGGCGCCGGCAAGATCCGGCTCA BL2414R ATATGAATTCTCACTGCGCGGCCGCGACCTTG BL2444F ATATGTCGACGAAGCTCGACCTCCGGCGATGACGC BL2444R ATATGCGGCCGCCTTCAGTGCGAGGCCTCGTATT BL2458F ATATGTCGACCCGTTACGCTGGATTTCTTCCAGTTCAAG BL2458R ATATGCGGCCGCCTTCACTCGAAGGTCCTGGCTTGGACC!

97 ! 87! Primers used in qpcr Primer Sequence (5-3 ) Blon_0043F TACGCCTCCTACGCCAAT Blon_0043R CGTACTTCTCGTGGATACCCT Blon_0043TM 6FAM-TTCTTGGTGCCGTTCACCAGTTCCT-BBQ Blon_0343F TTTGAAGGGTAAGATGGTCGTC Blon_0343R TGCTGCCATTGACCCAC Blon_0343TM 6FAM-AAGCCAATGCCGCTGCTTGCTTAT-BBQ Blon_0375F GAAGACGGCCAGTGGATC Blon_0375R AGGAGGTTCGGCATTGTACT Blon_0375TM 6FAM-AGGCTTCGCCATCGCGTCCA-BBQ Blon_0393F TTCACCGAGGCGTACAACA Blon_0393R CGCATCCGTGACCACATAG Blon_0393TM 6FAM-AGAATGCGCTGAATCAGGTCGATCAT-BBQ Blon_0883F ATCGAAGCCGTGTGGATT Blon_0883R CCTCGTTGTAGGCGTCGTA Blon_0883TM 6FAM-ACACCTTCATGTCGGAGGCCAGGT-BBQ Blon_2015F CCCGTCATTTCCGTGTGAT Blon_2015R CCACGTATTCTTTAGGGTCGTAG Blon_2015TM 6FAM-CATCCAGCGTTGCCTCCATGC-BBQ Blon_2061F GGTCCAACTACGCCGAATA Blon_2061R CGTACATGGCGTCGATCA Blon_2061TM 6FAM-CTTGGCGACCAGTACGTCACCCT-PH Blon_2177 F GGTTCCTGAGGTCTTCACCA Blon_2177 R GCCGAGCTTCTCAAATTCA Blon_2177 TM 6FAM-AGTACAAGGACGATTTCGCTTCCGC-PH Blon_2202F ATGAGAACCAGCGCAATAAG Blon_2202R CAGATCGCCGTTCTCATTC Blon_2202TM 6FAM-TCCGGCACGATGACATCGTTGAAC-BBQ Blon_2344F TCAAGAAGCTCGACCCGTTG Blon_2344R TTGGCGTAGAAGCCGTATGT Blon_2344TM 6FAM- ACTACACCTGGCACAGCCCGATGCT-BBQ Blon_2347F AAGCCGATAGGTTCTCCCT Blon_2347R TCGCCTTGGTGTACTTGTCT Blon_2347TM 6FAM-AGCTGGCCAACCTGCTCTACTCCGA-BBQ Blon_2350F GGTCTGTCTGATCGGTTTACG Blon_2350R CTGCGCTGCTCATCATATG Blon_2350TM 6FAM-TCCCGACGAAGTCTCCATTAAGGGC-BBQ Blon_2351F GTTCGGCAGCTTGTCAAGA Blon_2351R AGTGGCAGGAAGTGACTCG Blon_2351TM 6FAM-AACGCCACCTTGGCCTCTGGAGATA-BBQ Blon_2352F GCGCCCTGTAGCTATCGAA Blon_2352R GACGTGAACCTGAATGGATACG Blon_2352TM 6FAM-ACAAGCTGCCAAACGTCAAGGCATT-BBQ Blon_2354F ACTGATAGGCGTCAAGGGAA Blon_2354R GGACAAGATCGCCAATATGC Blon_2354TM 6FAM-AGGACTTGGAGGCCGACTGCGAT-BBQ Blon_2357F GCGGACTTACGAACAGGAAT Blon_2357R GCGGCCTATCTGTATGAGAAG Blon_2357TM 6FAM-ACGTCCAAGGACAGCCAGATTGAGT-BBQ Blon_2367F GTAGTTCCCCACACCGACT Blon_2367R CAACATCGGCATCACCAT Blon_2367TM 6FAM- AATCCACCGCCGATCTGGCC-BBQ

98 88! (Table S1 continued) Primer Sequence (5-3 ) Blon_2380 F CAGGATCAGGCCGTCAA Blon_2380 R TCATTGGCGATCTTGACGA Blon_2380 TM 6FAM-AAGGCCGTGCTCGACGATGACTG--BBQ Blon_2414F GTTCGCTCTTGAGAATGTCCG Blon_2414R ACCAGAAGGTCTACGATCAGG Blon_2414TM 6FAM-CGGCAAGTACTACCTGCACACCAAC-BBQ Blon_2444F GACTTCCTCAAGTGGGTGATC Blon_2444R ACGGCATCATGGTGAAGTT Blon_2444TM 6FAM-AACCTGGGTCTTGCCGGACTCCT BBQ Blon_2458F TCTACGACTTCACCGATGAGC Blon_2458R AGCATGTCGGTGAACTCG Blon_2458TM 6FAM-TCGAGACCTACCTTGCGGAACAGC--BBQ a Sequence in bold represent the restriction site for each restriction enzyme!!

99 89 Chapter III!! Bifidobacterium longum subsp. infantis ATCC 15697!-Fucosidases Are Active on Fucosylated Human Milk Oligosaccharides David A. Sela a,1, Daniel Garrido b,1, Larry Lerno c, Shuai Wu c, Kemin Tan d, Hyun-Ju Eom e, Andrzej Joachimiak d, Carlito B. Lebrilla c and David A. Mills e a Microbiology Graduate Group, b Food Science Graduate Group, c Department of Chemistry, University of California, Davis, Davis, California, USA, d Midwest Center for Structural Genomics and Structural Biology Center, Biosciences, Argonne National Laboratory, Argonne, Illinois, USA, e Department of Viticulture and Enology, University of California, Davis, Davis, California, USA 1 These authors contributed equally to this work. Conceived and designed the experiments: DAS, DG, LL, SW. Performed the experiments: DG, LL, SW, KT, HYE. Analyzed the data: DAS, DG, LL, KT, CBL, DAM. Contributed reagents/materials/analysis tools: AJM, CBL. Wrote the manuscript: DAS, DG.

100 90 Summary Bifidobacterium longum subsp. infantis ATCC utilizes several small-mass neutral human milk oligosaccharides (HMOs), several of which are fucosylated. Whereas previous studies focused on endpoint consumption, a temporal glycan consumption profile revealed a time-dependent effect. Specifically, among preferred HMOs, tetraose was favored early in fermentation, with other oligosaccharides consumed slightly later. In order to utilize fucosylated oligosaccharides, ATCC possesses several fucosidases, implicating GH29 and GH95!-L-fucosidases in a gene cluster dedicated to HMO metabolism. Evaluation of the biochemical kinetics demonstrated that ATCC expresses three fucosidases with a high turnover rate. Moreover, several ATCC fucosidases are active on the linkages inherent to the HMO molecule. Finally, the HMO cluster GH29!-L-fucosidase possesses a crystal structure that is similar to previously characterized fucosidases.

101 91 1. Introduction The genus Bifidobacterium is frequently overrepresented in the breast-fed infant colon relative to its appearance in adults, where these organisms are believed to benefit their host through nutrient supplementation, participating in host energy cycling and binding to preferred host receptor molecules otherwise available to pathogens (12). Selective growth of bifidobacteria has been attributed to utilization of oligosaccharides abundant in human milk (10 to 20 g/liter) that present complex structures resistant to infant digestion (17, 35). Approximately 200 species of human milk oligosaccharides (HMOs) have been characterized that are composed of glucose, galactose, N- acetylglucosamine, and often fucose and/or sialic acid residues via several glycosidic linkages (25). The HMO core is typically elongated from a lactosyl reducing end (Gal!1-4Glc) that is linked via!1-3 (or!1-6 in branched molecules) to serial lacto-n-biose I units (Gal!1-3GlcNAc) or lactosamine (Gal!1-4GlcNAc) with a degree of polymerization of "4. As with other fucosylated glycoconjugates, #1-2/3/4 fucosyl moieties often shield HMOs from digestion unless this linkage at the non-reducing terminus is first cleaved. Similarly, acidic HMOs or milk sialyloligosaccharides (MSOs) obstruct enzymatic degradation with sialyl residues via #2-3/6 linkages. Removal of these termini is postulated to initiate bacterial catabolism of HMOs (2, 8, 30). To this end, it has been recently demonstrated that Bifidobacterium longum subsp. infantis ATCC utilizes milk sialyloligosaccharides via a sialidase encoded within a large gene cluster dedicated to HMO metabolism (30).

102 92 Previous research conducted on bifidobacterial metabolism of fucosylated oligosaccharides identified a Bifidobacterium bifidum!1-2-l-fucosidase that exhibited an atypical inverting mechanism (glycoside hydrolase [GH] family 95), termed AfcA (10, 22). Inverting glycoside hydrolases modify anomeric stereochemistry via a single nucleophilic displacement, mechanistically contrasting with retaining enzymes, which maintain the anomeric configuration through catalysis of a second displacement. A second B. bifidum fucosidase, one that hydrolyzes!1-3/4 linkages, was characterized to be active on the purified HMO lacto-n-fucopentaose II {Gal("1-3)[Fuc(!1-4)]GlcNAc("1-3)Glc("1-4)Glc} and lacto-n-fucopentaose III {Gal("1-4)[Fuc(!1-3)]GlcNAc("1-3)Glc("1-4)Glc}. Both fucosidases are secreted and attach to the B. bifidum extracellular surface. The specific HMOs consumed by B. longum subsp. infantis ATCC have been previously detailed, and several small-mass, fucosylated oligosaccharides are clearly preferred (16, 17). The physiological basis for fucosylated HMO metabolism is evidenced by several potential!-l-fucosidases, including two residing within the HMO gene cluster; however, the B. longum subsp. infantis ATCC genome sequence is decidedly more ambiguous with respect to successive steps in fucose metabolism, as genes of the canonical fuciak pathway have not been detected (29). We describe here our investigation into fucosylated HMO metabolism, including a temporal glycoprofile to monitor consumption preferences through fermentation. Moreover, we have characterized fucosidases expressed by B. longum subsp. infantis ATCC representing an essential activity necessarily employed early in the catabolism of these molecules.

103 93 2. Materials and Methods Bacteria and media. B. longum subsp. infantis ATCC was routinely grown on modified de Mann, Rogosa, and Sharpe medium supplemented with 2% (wt/vol) lactose and 0.25% (wt/vol) L-cysteine (Sigma-Aldrich, St. Louis, MO). Cells were cultured in a vinyl anaerobic chamber (Coy Laboratory Products, Grass Lake, MI) at 37 C and 5% carbon dioxide, 5% hydrogen, and 90% nitrogen. Chemically competent Escherichia coli BL21 Star and Top10 cells were obtained from Invitrogen (Carlsbad, CA), and recombinant clones were grown in Luria broth supplemented with 50 µg/ml of carbenicillin (Teknova, Hollister, CA) when necessary at 37 C. RNA extraction. B. longum subsp. infantis cells were grown on Zhang-Mills-Block 1 (ZMB-1) medium (38), and 2% carbohydrate, such as lactose (Sigma), inulin (raftiline HP; Orafti, Malvern, PA), or purified HMO (gift from J. B. German) was added. Lacto- N-tetraose and lacto-n-neotetraose (V-Labs, Covington, LA) were also added at a final concentration of 0.5%. Growth profiles were monitored by measuring the optical density at 600 nm (OD600) using a PowerWave microplate spectrophotometer (BioTek Instruments, Inc., Winoosky, VT). Each experiment was performed in triplicate. Cells were recovered at exponential phase, pelleted, resuspended in 1 ml of RNAlater (Ambion, Austin, TX), and stored overnight at 4 C and then at!80 C until use. RNA extraction was performed as previously described (7), and RNA was converted to cdna using the High Capacity cdna reverse transcription kit (Applied Biosystems, Foster City, CA). Quantitative real-time PCR. Relative quantification of the transcript levels of B. infantis fucosidase genes was determined using quantitative real-time PCR (qrt-pcr).

104 94 Blon_0393, a gene encoding a cysteinyl-trna synthetase, was used as an endogenous control (27). Primers and 5! nuclease probes were designed and synthesized by TibMolBiol (Adelphia, NJ) (see Table S1 in the supplemental material). qrt-pcr was performed on a 7500 Fast real-time PCR system with TaqMan universal mastermix (Applied Biosystems). Each reaction mixture contained 0.5 µm each primer, with its specific probe at 125 nm, 5 ng of cdna, and reaction buffer as indicated by the manufacturer. qpcrs were run at 95 C for 10 min, 40 cycles of denaturation at 95 C for 10 s, and annealing and elongation for 1 min at 60 C. Relative transcript levels were normalized to lactose as the basal condition. Threshold cycle data and relative efficiencies were analyzed using the Q-Gene software ( Results are expressed as the fold changes in gene expression. Recombinant cloning and protein expression. Gene sequences were analyzed using the Integrated Microbial Genomes (IMG) database (19). For each fucosidase gene, primers were designed (see Table S1 in the supplemental material) with modifications necessary for cloning using the pet101 directional TOPO expression kit (Invitrogen). Genomic DNA from B. longum subsp. infantis was obtained as described in reference 7. Fucosidase genes were amplified by PCR using 0.5 µm each forward and reverse primer, 1 ng genomic DNA, 0.2 mm deoxynucleotriphosphate mix (Fermentas, Glen Burnie, MD), and 2 U of Phusion Hot Start high-fidelity DNA polymerase (Finnzymes, Vantaa, Finland) in a 150-µl final volume. PCR was performed using a PTC200 Thermo Cycler (MJ Research, Ramsey, MN), with initial denaturation at 95 C for 2 min and 35 cycles of denaturation at 95 C 1 min, annealing at 58 C for 1 min, extension at 72 C 2 min, and a

105 95 final extension at 72 C for 7 min. PCR products were gel purified and incorporated into pet101 via the TOPO reaction, and transformation was performed as indicated by the manufacturer. Recombinant BL21 Star clones were confirmed by plasmid sequencing using primers T7prom and T7term. Protein expression was carried out using 100 ml LB broth supplemented with 50 µg/ml carbenicillin, using different optimized growth temperatures and isopropyl-!-d-thiogalactopyranoside (IPTG) concentrations (see Table S1 in the supplemental material). Cells were grown in a shaker at 250 rpm (Innova-4000; New Brunswick Scientific, Edison, NJ) until they reached an OD of 0.6 and were induced with IPTG for 6 h. Cultures were centrifuged at 1,700 " g on an Eppendorf 5804 centrifuge (Hauppauge, NY) for 20 min at 4 C and stored at #80 C. Pellets were reconstituted in Bugbuster protein extraction reagent (EMD Chemicals), using 5 ml of the detergent for 50 ml of culture. Lysozyme (50 µl of a 50-mg/ml stock; Sigma-Aldrich) and DNase I (20 µl of a 10,000-U stock; Roche Applied Sciences) were added, vortexed, and incubated for 20 min at room temperature. The suspension was centrifuged for 20 min at 18,500 " g at 4 C, and the supernatant was applied to 1-ml Bio-Scale Mini Profinity immobilized-metal affinity chromatography cartridges connected to an EP-1 Econo pump (Bio-Rad, Hercules, CA). Protein purification was performed as recommended by the manufacturer. Recombinant proteins were evaluated for molecular mass and purity on 10% SDS-PAGE gels. Finally, imidazole was exchanged for phosphate-buffered saline using Amicon Ultra-15 centrifugal filter units, with a cutoff of 10 kda (Millipore, Billerica, MA). Determination of kinetic parameters. McIlvaine buffer solutions, between ph 4.0 and 8.0, were prepared for optimum ph determinations. Relative activity at different ph

106 96 values was determined by incubating each fucosidase in a 100-µl reaction volume with 2 mg/ml of 2-chloro-4-nitrophenyl-!-L[SCAP]-fucopyranoside (CNP-fucose; Carbosynth, Berkshire, United Kingdom) in 96-well microwell plates at 37 C for 10 min. Reactions were stopped by adding an equal volume of 1 M Na2CO3. Absorbance at 405 nm was determined using a Synergy2 microplate reader (Biotek). The optimum temperature for enzymatic activity was determined in McIlvaine buffer at each enzyme's optimum ph, incubating them with CNP-fucose at 4, 20, 30, 37, 45, 55, and 65 C. Relative activity was determined from A405 reads. For determination of kinetic values, substrate concentrations in the range of 0.1 to 4 mm CNP-fucose were coincubated with constant amounts of each enzyme under optimum conditions as determined. The time course for each reaction was determined previously to fit steady-state assumptions, and rate reactions were determined under initial conditions. The amount of CNP generated in each reaction was estimated using a standard curve. Nonlinear regression was used to determine Km and Vmax values, using the Solver program. Thin-layer chromatography. The activities of!-fucosidases against fucosylated substrates were studied using the optimum conditions determined as described above, in 10-µl reaction mixtures under optimum conditions for different times. 2"-Fucosyl lactose (2"FL), 3"-fucosyl lactose (3"FL), and H-disaccharide (Fuc!1-2Gal) were purchased from V-labs. Aliquots of the reaction mixtures were inactivated at 95 C for 5 min and spotted in thin-layer chromatograph (TLC) glass-silica gel plates (Sigma). A mixture of ethyl acetate, acetic acid, and water in a 2:2:1 ratio was used as solvent. After drying, plates were sprayed with 0.5%!-naphthol and 5% H2SO4 in ethanol. Plates were dried and revealed at 150 C for 10 min.

107 97 Relative substrate preferences for HMO cluster fucosidases. Equimolar concentrations of 2!-fucosyl lactose, 3!-fucosyl lactose, H-disaccharide, and Lewis a and Lewis x (V-labs) were coincubated with 5 µg of either Blon_2335 or Blon_2336 for 10 min at 37 C. Reactions were inactivated by incubation at 95 C for 5 min. The galactose assay kit (Biovision, Mountain View, CA) was used to quantify galactose concentrations present in each sample (in an equimolar ratio to released fucose), after overnight incubation with lactase (2!FL and 3!FL) and a mix of 2 µg of B. longum subsp. infantis enzymes Blon_2016 and Blon_2334 (known to cleave Gal"1-3GlcNAc and Gal"1-4GlcNAc [data not shown]) for Lewis a and Lewis x reactions. Fluorescence was quantified using a standard curve in a Synergy 2 microplate reader. Glycoprofiling. Aliquots of supernatant were spiked with an internal standard consisting of deuterated HMOs as described previously (35, 36). Each sample was desalted by solidphase extraction employing graphitic carbon Top Tip cartridges (Glygen, Columbia, MD) and was evaporated to dryness under vacuum. The dried samples were prepared for liquid chromatography-mass spectrometry (LC-MS) analysis by dissolving the sample into 18 M# water to a final concentration of 20 ppm based on the amount of the internal standard added to each sample. Analytical triplicates were prepared from each sample of supernatant, and each triplicate was analyzed three times by LC-MS analysis. LC-MS analysis was performed with an Agilent HPLC-Chip/TOF accurate mass time-of-flight mass spectrometer equipped with both an Agilent 1200 series capillary flow high-performance LC (HPLC) pump for sample loading and an Agilent 1200 series nanoflow HPLC pump for chromatographic separation. All chromatography was performed on an Agilent glycan chip, a microfluidic platform incorporating an

108 98 enrichment/concentration column, nano-lc column, and nanospray emitter. Both the enrichment and nano-lc column were packed with porous graphitized carbon. A binary solvent system was employed consisting of 3% aqueous acetonitrile plus 0.1% formic acid (solvent A) and 90% aqueous acetonitrile plus 0.1% formic acid (solvent B). Chromatographic separation of the HMOs was achieved using a previously optimized gradient consisting of the following solvent changes: 0 to 2.5 min 0% B, 2.5 to 20 min 16% B, 20 to 30 min 44% B, 30 to 35 min 100% B, 35 to 45 min 100% B, to 65 min 0% B (24). The flow rate was held constant at 300 nl/min during the entirety of the gradient. All analyses were performed in the positive ion mode with the source conditions being adjusted to provide for maximum signal with minimal source fragmentation. Each analysis was processed postacquisition by summing the spectra across the chromatographic range. The m/z values and intensities in these summed spectra were then exported as text files and used to calculate the percent consumption of the fucosylated HMOs. The percent consumption levels of the fucosylated HMOs were calculated based on a previously described method (26, 36). Briefly, the percent consumption of a fucosylated HMO was calculated using the ratio of the intensities of the HMO in the supernatant and the corresponding HMO from a control sample (D/H ratio). The D/H ratio for each fucosylated HMO of interest in both the supernatant and control samples had to be adjusted for overlap of the isotopic envelope of the nondeuterated HMO. This was accomplished by using the following equation, D/H = [(ID/IH)! IH(k)/IH], where ID and IH are the intensities of the monoisotopic peak of deuterated HMO standard and HMO, respectively. The weighting factor k is used to remove the contribution of the

109 99 HMO isotopic envelope from that of the deuterated standard. All weighting factors were calculated using the IonSpec Exact Mass Calculator software (version ). Once the D/H ratio for each supernatant and control sample was determined, the percent consumption was then calculated using the following equation: 1! {[(D/H)sample]/[(D/H)control]} " 100. Fucosidase digestion of purified HMO. Recombinant enzymes were used without further purification. HMO standards LNFP I, LNFP III, and LNDFH I were obtained from Dextra Laboratory (Earley Gate, United Kingdom). In a 0.2-ml PCR tube, Three µl nanopure water were added, followed by 1 µl HMO standard sample and 1 µl enzyme solution (the mole ratio of protein to OS was about 1:100 to 1:200). The reaction mixture was incubated at 37 C in a water bath. The reaction was monitored by matrix-assisted laser desorption ionization Fourier transform infrared spectroscopy-ion cyclotron resonance MS (MALDI FT-ICR MS) at 1 h and 3 h. The HiRes MALDI FT-ICR apparatus (IonSpec, Irvine, CA) has an external MALDI source with a pulsed 355-nm Nd:YAG laser, a hexapole ion guide, an ultrahigh vacuum system maintained by two turbo pumps, one cryopump, and a 7.0-T shielded superconducting magnet. DHB (2,5- dihydroxybenzoic acid) was used as the matrix (8 mg/160 µl in 50% acetonitrile-water [vol/vol]) in the positive ion mode. The HMO solution (0.5 µl) was spotted on a 100- sample stainless steel probe followed by adding 0.25 µl, 0.01 M NaCl solution as a cation dopant, and 0.5 µl matrix solution. The sample was dried in the vacuum chamber before placing it into the ion source. Protein crystallization and X-ray structure determination. For structural characterization of Bifidobacterium longum subsp. infantis ATCC fucosidases, in

110 100 particular Blon_2336, SeMet-labeled Blon_2336 was screened for crystallization conditions using the hanging drop vapor diffusion technique. Diffraction-quality crystals appeared under the condition of 0.8 M succinic acid, ph 7.0, at 16 C. Prior to X-ray data collection, crystals were treated in paratone N oil and were flash-frozen directly in liquid nitrogen. A set of single-wavelength anomalous diffraction (SAD) data was collected near the selenium absorption peak at a temperature of 100 K from a single selenomethionine (SeMet)-labeled crystal. The data were obtained at the 19ID beamline of the Structural Biology Center at the Advanced Photon Source at Argonne National Laboratory, using the program SBCcollect. The intensities were integrated and scaled with the HKL3000 suite (20) (see Table S2 in the supplemental material). Two Blon_2336 molecules (monomers A and B) with a total 20 methionine residues were expected in one asymmetric unit. Twenty heavy atom sites (potentially Se) were located using the program SHELXD (28), and 19 selenium sites were used for phasing with the program MLPHARE (4). After map averaging and density modification (DM) (4), a partial model of 800 residues (84% of two Blon_2336 molecules) with 506 side chains was built via 3 cycles of Arp/warp model building (3). All of the above programs are integrated within the program suite HKL3000 (20). Further model-building efforts to complete the structure were performed manually using the program COOT (5). In and outside the active site of the catalytic domain, there were extra electron densities that appeared similar to a short peptide with a tyrosine residue pointing into the ligand-binding pocket. Only a tyrosine residue was built into the densities for structural refinement purposes. The nature of the short peptide-like ligand remains unknown. The final model was

111 101 refined using the program Phenix.refine (21) (see Table S2). In the final model, 17 residues between A236 and V255 in monomer A and 11 residues between E241 and T254 in monomer B are missing due to the lack of electron densities. Protein structure accession number. The crystal structure was deposited in the Protein Data Bank (PDB:3MO4).

112 Results 3.1 The B. longum subsp. infantis ATCC temporal glycoprofile of preferred neutral HMOs. We have previously demonstrated that B. longum subsp. infantis ATCC prefers small-mass oligosaccharides secreted early in lactation (17). Several of these HMOs are fucosylated and represent the most abundant species within the aggregate pool. Recently, we monitored oligosaccharide consumption during a fermentation to resolve substrate preferences while controlling for spurious degradation, which revealed that bacteria are capable of utilizing sialylated milk oligosaccharides (30). Using a similar approach, FT-ICR MS was employed to ascertain at what point during fermentation B. longum subsp. infantis ATCC utilizes neutral compositions from a mixed population of purified HMOs (i.e., <200 molecular species). Six low-molecular-mass neutral HMOs were examined, with fermentation culture supernatants sampled at various cell physiological stages corresponding to increasing optical densities (OD600) of approximately 0.2, 0.3, 0.6, 0.75, and 1.0 (Fig. 1). The six HMO species corresponded to those that are consumed by B. longum subsp. infantis efficiently and have been termed preferred (i.e., lacto-n-tetraose [LNT], lacto-n-hexaose [LNH], F-LNH, and DF-LNH), with two additional compositions that are not utilized well and are referred to as nonpreferred (i.e., fucosylated lacto-n-octaose [F-LNO] and DF-LNO). While fucosylated LNT (F-LNT) possesses a comparable mass and is readily consumed, it was excluded due to an extremely low abundance within our HMO pool (<0.7%) (17). We defined consumption conservatively as!35% elimination from the culture supernatant.

113 103 Figure 1: Temporal glycoprofile of abundant neutral HMO consumption by B. longum subsp. infantis ATCC HMO composition results are shown for a representative isomer signifying a characteristic oligosaccharide composition. All samples were analyzed at least in duplicate. LNT, lacto-n-tetraose-like composition; LNH, lacto-nhexaose-like composition; F-LNH, fucosylated lacto-n-hexaose-like composition; DF- LNH, difucosylated lacto-n-hexaose-like composition; F-LNO, fucosylated lacto-noctaose-like composition; DF-LNO, difucosylated lacto-n-hexaose-like composition.

114 104 Interestingly, the LNT-like composition (Gal!1-3/4GlcNAc!1-3Gal!1-4Glc) was consumed early in fermentation and to a greater degree than other neutral oligosaccharides. LNT fermentation terminated at the commencement of stationary phase, with near-complete elimination from the culture supernatant (98.97 ± 0.13% [mean ± standard error of the mean]). With respect to oligosaccharides of increasing degrees of polymerization (DPs), the next two preferred neutral oligosaccharides, LNH (3 Gal:2 GlcNAc:1 Glc) and F- LNH (3 Gal:2 GlcNAc:1 Glc:1 Fuc), were consumed efficiently and approached extinction at ± 0.10% and ± 0.09% elimination, respectively. LNH and F- LNH utilization levels, however, appeared to occur at a later stage than LNT, achieving >60% elimination only at stationary phase. Of interest, LNH was not preferred over F- LNH, despite similar defucosylation prior to metabolism. This suggests a differential transport efficiency between these two oligosaccharides, which differ by one fucosyl residue. DF-LNH differs from F-LNH by an additional fucosyl moiety, and with the initiation of stationary phase is consumed to a lesser extent than the other three preferred oligosaccharides ( 76%). Structural constraints presented by the difucosylated molecule likely contribute to this potentially deleterious effect on transport efficiency. The clearest differences in utilization were observed, as expected, in the nonpreferred lacto-n-octaose series (F-LNO, 4 Gal:3 GlcNAc:1 Glc:1 Fuc) and (DF- LNO, 4 Gal; 3 GlcNAc:2 Fuc:1 Glc), which may represent the DP limit to which B. longum subsp. infantis ATCC efficiently transports HMOs. Translocation of intact molecules is considered critical for B. longum subsp. infantis milk oligosaccharide

115 105 metabolism, as their HMO-active glycoside hydrolases are localized intracellularly (1, 2, 29, 30). Of the preferred oligosaccharides, LNT is favored, as it disappeared earlier from the recovered fermentate than LNH, F-LNH, and DF-LNH, although the large intersample variation in early fermentation (e.g., LNH utilization) precludes a definitive conclusion. This error may be the result of the difficulty in obtaining biological replicates identical in the amount of arbitrarily defined phases. Similar variability has been observed in previous glycoprofiling studies of bifidobacterial consumption of HMOs (16, 18). 3.2 Identification of!-fucosidases in the ATCC genome. The ATCC predicted proteome was scanned for evidence of fucosidase sequence signatures. An examination of GH29 (pfam01120)!-fucosidases revealed the presence of a GH29!- fucosidase (Blon_2336) that localized to a cluster of genes linked with HMO utilization. This!1-3/4 fucosidase exhibited 33% amino acid identity to the fucosidase domain of the B. bifidum NCIMB AfcB fucosidase (BbifN4_ ). Interestingly, the Blon_2336 sequence diverged less from the B. bifidum AfcB than from other ATCC fucosidases (Fig. 2). Clearly, this does not match the organismal phylogeny and suggests different functions for these fucosidases and/or lateral gene transfer. Moreover, Blon_2336 was predicted to be expressed intracellularly, as it does not possess the N- terminal signal sequence, C-terminal transmembrane region, or the LPXTG motif displayed by the larger, membrane-anchored B. bifidum AfcB (478 versus 1,468 amino acids [aa]). Both AfcB and Blon_2336 include a carbohydrate-binding module 32

116 106 (CBM32) domain (PF00754), which has been shown to ligate lactosyl/galactosyl residues for the Micromonospora viridifaciens sialidase (23), lactosamine (Gal!1-3/4GlcNAc) in association with a Clostridium perfringens!-hexosaminidase (6) and found in the LNTcleaving lacto-n-biosidase (EC ) characterized for B. bifidum (34). This strongly suggests that the CBM32 domain binds HMO epitopes and may be common to HMO-active glycoside hydrolases, consistent with Blon_2336 integration within the HMO cluster.

117 107 Figure 2: Phylogenetic relationships of fucosidases encoded by select bacteria. Branch lengths are in the same units (number of amino acid substitutions per site) as those of the evolutionary distances used to construct the tree. The evolutionary history was inferred by the maximum likelihood method, followed by 100 bootstrapped replicates. The organism and loci are listed for those fucosidases found in glycoside hydrolase family 29 (A) and glycoside hydrolase family 95 (B).

118 108 Directly adjacent to Blon_2336 is an encoded GH95!-fucosidase (Blon_2335) predicted to cleave!1-2 linkages by an atypical inverting mechanism. This gene has a much shorter sequence than the homologous B. bifidum afca (BbifN4_ ), with the two enzymes predicted to be comprised of 782 and 1,959 amino acid residues, respectively. These two enzymes exhibited 26% identity along their alignable region, which corresponded solely to the B. bifidum AfcA GH95 catalytic domain. In contrast, an N-terminal signal peptide, C-terminal anchor, or an immunoglobulin-like domain was not identified in the ATCC AfcA. These three elements, in concert, allocate the B. bifidum AfcA to its extracellular surface. Two additional GH29!-L-fucosidase-encoding open reading frames, Blon_0248 and Blon_0426, appear elsewhere on the ATCC chromosome and are highly homologous at the nucleotide level, sharing 98% identity along their length (1,350 bp), with divergence occurring at the 3" end. These paralogs are 95% identical along the length of their deduced 449-amino-acid sequence, moderately diverging at the C terminus external to their GH29!-L-fucosidase domain (amino acids 4 to 360) and hinting at similar catalytic activities. As previously reported (29), Blon_0426 and an associated permease appear to have arisen from a recent duplication event that may have displaced elements of arabinose metabolism in ATCC This is one of several features by which B. longum subsp. infantis metabolism was redirected from plant oligosaccharides toward mammalian sugars (31). Both Blon_0248 and Blon_0426 are predicted to be intracellular, as is the Bifidobacterium dentium ATCC homolog (BIFDEN_00927), although its encoded protein only shares 28% identity with the ATCC

119 Blon_0248 protein (33). The phylogenetic relationship between currently sequenced bifidobacteria GH29!-L-fucosidases is presented in Fig. 2. Finally, there is an additional gene that has been annotated as a "-galactosidase (COG1874) that also possesses a pfam01120!-fucosidase domain at the 5# terminus (Blon_0346). This gene has a "-galactosidase trimerization domain (pfam08532), which suggests this enzyme may be active as a trimer polypeptide. As indicated in Fig. 2, this enzyme is quite divergent from other!-fucosidases, including those found in ATCC !-Fucosidase expression profiles. In order to link genetic structure with an encoded enzyme's substrate, and to determine the extent to which a given substrate induces expression, the expression profiles of all five ATCC 15697!-L-fucosidases were examined (Fig. 3). With the exception of Blon_0248, all fucosidases were induced $2- fold, relative to lactose, when ATCC was grown on the polyfructan inulin, which is devoid of fucosyl residues. A similar pattern has been observed in HMO-interacting solute-binding proteins induced in bifidobacteria subsisting on inulin (7). Interestingly, our complex mix of HMOs only moderately induced ( 2-fold) Blon_2335, while it repressed Blon_0248 and Blon_0426. This appears to indicate a carbon catabolite repression by one or more of the molecules present in the HMO mix. This repression may be due to the abundance of lacto-n-tetraose, because when cells were grown on this substrate, Blon_0248 and Blon_0426 were similarly repressed; however, the other three fucosidases were induced, while ATCC consumed this single-isomer HMO. LNT's structural isomer, lacto-n-neotetraose (LNnT), which differs only by a terminal "1-4

120 110 galactosyl linkage, displayed a different expression pattern. Accordingly, LNnT induces the same three enzymes, with Blon_2335 and Blon_2336 induced to a lesser extent. However, repression of Blon_0248 and Blon_0426 is not evident, as they appeared to be transcribed to a similar degree as with cells grown on lactose.

121 111 Figure 3: ATCC fucosidase gene expression during carbohydrate fermentation. Gene expression was calculated relative to levels when grown on lactose as the sole carbon source. Averages from three independent experiments are shown, and bars represent standard errors of the means.

122 112 Biochemical characterization of ATCC 15697!-fucosidases. In order to evaluate the activity of B. longum subsp. infantis ATCC 15697!-fucosidases, these enzymes were cloned and expressed in a heterologous host. Table S3 in the supplemental maerial indicates optimal expression conditions in addition to ph and temperature optima. All five fucosidases were most active at 37 C and possessed ph optima between 6 and 7.5 (Table S3). The Km values of the fucosidases were <1.0, with the exception of Blon_2335, which had a much lower affinity for CNP-fucose (Km = ± mm) than the other enzymes (Table 1). Blon_0346, Blon_0426, and Blon_2335 have relatively higher turnover rates (>3.5 s"1) than the other two enzymes, which in turn lead to high efficiency ratios (Kcat/Km) for Blon_0346 and Blon_0426.

123 113 Table 1: Biochemical attributes of ATCC 15697!-fucosidases.! Values are means ± standard errors of the means Vmax (nmol/s) Km (mm) Kcat (s -1 ) Kcat/km (M-1 s-1) Blon_ ± ± ± ± Blon_ ± ± ± ± Blon_ ± ± ± ± Blon_ ± ± ± ± Blon_ ± ± ± ± 51.34

124 !-Fucosidase activity on purified HMOs and linkage specificities. A mass spectrometry-based approach was employed to determine the extent to which ATCC fucosidases are active on HMO linkages (see Fig. S1 in the supplemental material). Purified lacto-n-fucopentaose I (LNFP I) and LNFP III were incubated with the fucosidases to assay!1-3 and!1-2 linkages, respectively. Blon_0346 did not exhibit any activity on purified HMOs. Blon_0248 displayed a defucosylation of the!1-3 LNFP III after 1 h. Its paralog, Blon_0426, incompletely defucosylated LNFP III during the course of a 3-hour incubation with the enzyme. Blon_2335, a GH95 enzyme predicted to cleave!1-2 linkages, does indeed hydrolyze this bond in LNFP I after 1 h. It also partially cleaves!1-3 linkages. Finally, the other HMO cluster Blon_2336 cleaves!1-3 linkages inherent to LNFP III and partially degrades!1-2 linkages. Thin-layer chromatography was employed to confirm the MS data on small-chain oligosaccharides (DP, "3) (Table 2). Blon_2335 had activity on all three substrates tested, including 2-fucosyllactose, 3-fucosyllactose, and H-2 disaccharide (Fuc!1-2Gal). Consistent with the prediction that Blon_2335 is an!1-2 fucosidase, it exhibited stronger cleavage of the two!1-2 substrates. The other HMO cluster enzyme, Blon_2336, cleaved 3-fucosyllactose, consistent with its role as an!1-3/4 fucosidase. None of the other three enzymes displayed activities on these substrates. The one exception was Blon_0346, which had minor activity toward the H-2 disaccharide. The relative activities on these substrates (see Table S4 in the supplemental material) indicate that Blon_2335 has a preference for!1-2 linkages and Blon_2336 is more active on!1-3/4 linkages.

125 115 Table 2: TLC assay results for!-fucosidase linkage specificity. TLC signals were scored qualitatively: ", no signal; +, weak; ++, moderate; +++, strong. Linkage specificity to:!-fucosidase 2#FL 3#FL Fuca1-2Gal Blon_0248 " " " Blon_0426 " " " Blon_0346 " " + Blon_ Blon_2336 " +++ "

126 Structural characterization of Blon_2336. To gain insight into the threedimensional architecture of the HMO cluster GH29 fucosidase, Blon_2336 was crystallized and its structure was solved. Blon_2336 is monomeric, and each molecule consists of a catalytic domain of a (!/")8-barrel fold and a!-sandwich domain, which was earlier predicted to be a CBM32 domain (Fig. 4) (13).

127 117 Figure 4: Ribbon drawing of the overall structure of Blon_2336. The N-terminal catalytic domain is shown by the various colors, from the blue at the N terminus to red at the C terminus. The!-strands and "-helices that form the (!/")8 barrel are labeled in black. The C-terminal carbohydrate-binding domain is shown in lime green, with 9!-strands labeled in magenta. A tyrosine from a possible short peptide at the active site of the catalytic domain (see description in the text) is drawn in stick format. A dashed curved line indicates a missing part in the final structural model (see the text). The N and C termini of Blon_2336 are also labeled.

128 118 The catalytic domain of Blon_2336 can be structurally aligned well with the other four!-l-fucosidases, including one from Thermotoga maritime (TM0306) (32) and three from Bacteroides thetaiotaomicron (BT2192 [PDB:3EYP], BT2970 [PDB:3GZA], and BT3798 [14]). The root mean square deviation (RMSD) values of the pairwise structural alignments (secondary structural matching [SSM] fitting, C! only) between Blon_2336 and other four were 1.46 to 1.75 Å, with 244 to 293 aa of 333 aa (Blon_2336) aligned. This indicates that the overall structures of these!-l-fucosidases are quite conserved, although the primary sequence identities between Blon_2336 and others are 23 to 41%. The structural alignments of Blon_2336 with either TM0306 or BT3798 provide valuable information on the active site and potential catalytic residues of Blon_2336, for which extensive studies of their complexes with ligands and inhibitors have been reported (14, 32). Figure 5 shows a structural alignment of Blon_2336 with TM0306 in complex with fucose (32). Although the sequence identity of the catalytic domains of Blon_2336 and TM0306 is only about 26%, the residues that form their active sites are highly conserved. The residue D224 in TM0306 is the catalytic nucleophile. Its corresponding residue in Blon_2336 is D172. The residue E266 in TM0306 is the catalytic acid/base, which is located after the "6-strand. The region between the "6 and "7 strands is one of the most diversified regions in the!-l-fucosidases in terms of length and structural conformation. However, residue E217 of Blon_2336, shortly after the end of the "6 strand, is the most likely catalytic acid/base of Blon_2336 (Fig. 5). The equivalent amino acids of E217 in BT2192 and BT2970 are E243 and E240, respectively. Even in their apo-forms, these two residues in the BT2192 and BT2970 structures point to their active

129 119 sites. Upon ligand binding, the E217 of Blon_2336 can flip around to point its side chain toward the active site, as does the catalytic acid/base found in TM0306 and BT3798 complex structures. The conformation of E217 of Blon_2336 observed in this structure may represent one of the possible conformations in its apo-forms. The structural flexibility of part of a catalytic site has been widely observed and assists in the recognition and recruitment of substrate. The function of residue R254 of TM0306, or of its equivalent, R262 of BT3798, is unknown. In the active site, this is the only residue that appears to have no counterpart in Blon_2336 or in BT2192 or BT2970. The C-terminal domain, as predicted, is a!-sandwich domain. A Dali structural homology search ( using the C-terminal domain resulted in more than 700 hits, including different chains of the same molecules. All of the top 45 hits except 2 were CBM32 domains, with similar scores (Z, 14.4 to 16.7; RMSD, 1.5 to 2.4 Å; sequence identity, 12 to 25%). They included BT2192 and BT2970, which in fact were not outstanding in the top hits list. Interestingly, the C-terminal domain of Blon_2336 has no significant structural homology with that of BT3798 and TM0306.

130 120 Figure 5: Structural alignment of the active sites of Blon_2336 and TM0306 in complex with fucose. Blon_2336 is shown in cyan, and TM0306 is shown in green-yellow. All active site residues and the fucose from the TM0306/fucose complex are drawn in stick format. The residues from Blon_2336 and TM0306 are labeled in blue and orange, respectively. Equivalent residues from two molecules are paired together. Some visible strands from the (!/")8 barrel are labeled in black. The tyrosine shown in Fig. 4 is approximately at the position of the fucose in this figure, and it is not shown for the sake of clarity.

131 Discussion That LNT is consumed early, and to a greater degree, is significant, as LNT is invariably the most abundant oligosaccharide in breast milk and is consumed by a diversity of bifidobacteria, whereas larger HMOs are efficiently utilized by B. longum subsp. infantis and select infant-associated phylotypes (16, 17). LNT also represents the smallest HMO molecule, with a DP of 4, and as such it is a core structure incorporated within larger HMO species. It is possible that LNT is favored over higher-dp oligosaccharides earlier in fermentation, as it is linear and does not present a second!- linkage (1-6) to its terminal lactosyl, as do HMOs with DPs of "6, such as LNH and F- LNH. It is tempting to speculate that this!1-6 N-acetylglucosaminyl linkage inherent to branched HMOs may hinder translocation, albeit with minimal consequence to final consumption. B. longum subsp. infantis VIII-240 was previously characterized (15) as having a relatively strong #1-2 fucolytic activity, a trait now known to be encoded by Blon_2335. This gene was first identified as afca in B. bifidum (2, 11, 22). In addition to B. bifidum and B. longum subsp. infantis, the Bifidobacterium breve DSM20213 and Bifidobacterium pseudocatenulatum DSM20438 genomes include afca homologs, BIFBRE_01822 and BIFPSEUDO_02151, respectively, that are highly identical (both 77%) to the ATCC afca. It is remarkable that these four infant-type bifidobacteria possess afca homologs and that other bifidobacteria that usually colonize adults have not been found to possess one. Further investigation is needed to determine if afca is a key gene linked to infant gut colonization.

132 122 Of the four afca+ bifidobacteria, only B. bifidum secretes its AfcA, advancing a competitive strategy predicated on the import of extracellular hydrolysis products of HMOs (2, 34). In contrast, all five ATCC fucosidases are likely found in the cytosol, further distinguishing B. longum subsp. infantis from B. bifidum by the need to efficiently transport oligosaccharides (29,,31). Accordingly, several ABC transporters appear in the ATCC HMO cluster with solute-binding proteins known to target milk oligosaccharide substrates (7, 29). While the biology underlying specific fucosylated oligosaccharide transport has yet to be elucidated, the ATCC genome includes several transporters that may facilitate environmental scavenging when soluble fucose is encountered. Likely candidates include two deduced proteins assigned to the fucose permease COG0738 (Blon_2307 and Blon_0962) and two that are genetically linked to fucosidases (Blon_0247 and Blon_0426), with various degrees of homology to the few previously characterized fucose permeases. In addition to determining the location of the genes of fucosylated oligosaccharide transport, how B. longum subsp. infantis metabolizes fucose remains enigmatic. An examination of the ATCC genome did not reveal the genes of the canonical fucose utilization pathway. An alternative explanation is that fucose is metabolized via a cryptic oxidative pathway. There is nascent evidence for the existence of an oxidative pathway, based on identification of previously characterized oxidative pathway genes (37). There are two gene clusters within the B. longum subsp. infantis ATCC genome that are putatively involved with oxidative catabolism of fucose. One is localized to the HMO cluster (Blon_2337-Blon_2340) and is immediately adjacent to the fucosidases encoded by Blon_2335 and Blon_2336. Of the six enzymes involved in the oxidative pathway,

133 123 there is evidence for the existence of four of them in the ATCC genome (37). It is possible that some of these enzymes are multifunctional or have yet to be identified. Another possibility is that ATCC does not metabolize fucose but rather cleaves this moiety from oligosaccharides to afford access to an utilizable portion of the molecule. Of interest, one ATCC 15967!-fucosidase (Blon_0426) displayed considerably higher enzymatic efficiency toward CNP-fucose than its paralog (Blon_0248). This hints that these!-fucosidases are active on disparate substrates despite their sequence similarity. It is unknown what biological function, if any, is bolstered by Blon_0426's higher efficiency, detectable on CNP-fucose. However, based on MS analyses, Blon_0248 did exhibit greater activity on purified HMOs, indicating that this enzyme likely processes longer-chain fucosylated glycans. Curiously, it was inactive on smallerchain oligosaccharides as tested by TLC. This contrasts with the HMO cluster enzymes (Blon_2335 and Blon_2336) that clearly are active on both these small-chain linkages and large-chain linkages as assayed by TLC and MS, respectively. In addition, the crystal structure of Blon_2336 further demonstrates the structural conservation of the catalytic domain, in particular the active site of!-l-fucosidase, as well as the structural uniqueness of the GH29 family glycosyl hydrolases. That the active site component residues, including a catalytic nucleophile, are highly conserved can be hardly appreciated simply by sequence alignment due to generally low amino acid identities between these fucosidases. The slight positional variation of the catalytic acid/base and one of the most sequentially diversified regions after the catalytic residue,

134 124 as described earlier, may be one structural determinant of their substrate specificities and catalytic efficiencies. The Blon_2336 structure, together with other four!-l-fucosidase structures, also supports the implication that the catalytic acid/base residue of!-l-fucosidase of mammals, including humans, is an aspartate residue, based on an analysis of a large number of samples of!-l-fucosidase genes (9). The key residue substitution may lead to different substrate specificities and potentially different catalytic mechanisms. The physiologic or even pathological relationship between the!- L-fucosidases of humans and the!- L-fucosidases of enteric bacteria such as bifidobacteria remains an interesting area for further investigation. In conclusion, B. longum subsp. infantis ATCC utilizes several small-mass neutral oligosaccharides, some of which are fucosylated. Whereas previous MS glycoprofiles of neutral HMO consumption examined an end point in early stationary phase, this study revealed a temporal-dependent utilization profile. Specifically, among preferred HMOs, LNT is favored early, with LNH and F-LNH utilized fully though slightly later in the fermentation. DF-LNH is utilized at the same point as LNH and F- LNH, although it is consumed to a lesser degree. In order to utilize fucosylated HMOs, the ATCC chromosome encodes five fucosidase genes, four of which are GH29!- L-fucosidases. Interestingly, the HMO cluster fucosidase genes provide the greatest evidence for involvement in HMO metabolism, as they are active on a wider assortment of HMO linkages and their expression is induced by this purified substrate.

135 125 Acknowledgments This work was supported by grants from the University of California Discovery Grant Program, the California Dairy Research Foundation, USDA NRI-CSREES award , and National Institutes of Health NICHD awards R01HD059127, R01HD065122, and R01HD D.A.S. was supported by a predoctoral training grant (NIH-NIGMS T32-GM08799). This work was also supported by National Institutes of Health grant GM and by the U.S. Department of Energy (DOE), Office of Biological and Environmental Research under contract DE-AC02-06CH The contents of this report are solely the responsibility of the authors and do not necessarily represent the official views of the NIGMS, NIH, or DOE.

136 126 REFERENCES 1. Asakuma S, et al Physiology of the consumption of human milk oligosaccharides by infant-gut associated bifidobacteria. J. Biol. Chem. 286: Ashida H, et al Two distinct alpha-l-fucosidases from Bifidobacterium bifidum are essential for the utilization of fucosylated milk oligosaccharides and glycoconjugates. Glycobiology 19: Cohen SX, et al Towards complete validated models in the next generation of ARP/wARP. Acta Crystallogr. D Biol. Crystallogr. 60: Collaborative Computation Project Number The CCP4 suite: programs for protein crystallography. Acta Crystallogr. D Biol. Crystallogr. 50: Emsley P, Cowtan K COOT: model-building tools for molecular graphics. Acta Crystallogr. D Biol. Crystallogr. 60: Ficko-Blean E, Boraston AB The interaction of a carbohydrate-binding module from a Clostridium perfringens N-acetyl-beta-hexosaminidase with its carbohydrate receptor. J. Biol. Chem. 281: Garrido D, Kim JH, German JB, Raybould HE, Mills DA Oligosaccharide binding proteins from Bifidobacterium longum subsp. infantis reveal a preference for host glycans. PLoS One 6:e Gyorgy P, Jeanloz RW, von Nicolai H, Zilliken F Undialyzable growth factors for Lactobacillus bifidus var. pennsylvanicus. Protective effect of sialic acid bound to glycoproteins and oligosaccharides against bacterial degradation. Eur. J. Biochem. 43: Intra J, Perotti ME, Pavesi G, Horner D Comparative and phylogenetic analysis of alpha-l-fucosidase genes. Gene 392: Katayama T, Fujita K, Yamamoto K Novel bifidobacterial glycosidases acting on sugar chains of mucin glycoproteins. J. Biosci. Bioeng. 99: Katayama T, et al Molecular cloning and characterization of Bifidobacterium bifidum 1,2-alpha-L-fucosidase (AfcA), a novel inverting glycosidase (glycoside hydrolase family 95). J. Bacteriol. 186: Kleerebezem M, Vaughan EE Probiotic and gut lactobacilli and bifidobacteria: molecular approaches to study diversity and activity. Annu. Rev. Microbiol. 63: Krissinel E, Henrick K Inference of macromolecular assemblies from crystalline state. J. Mol. Biol. 372: Lammerts van Bueren A, et al Analysis of the reaction coordinate of alpha-lfucosidases: a combined structural and quantum mechanical approach. J. Am. Chem. Soc. 132: Larson G, Falk P, Hoskins LC Degradation of human intestinal glycosphingolipids by extracellular glycosidases from mucin-degrading bacteria of the human fecal flora. J. Biol. Chem. 263: LoCascio RG, et al A versatile and scalable strategy for glycoprofiling bifidobacterial consumption of human milk oligosaccharides. Microb. Biotechnol. 2:

137 LoCascio RG, et al Glycoprofiling of bifidobacterial consumption of human milk oligosaccharides demonstrates strain specific, preferential consumption of small chain glycans secreted in early human lactation. J. Agric. Food Chem. 55: Marcobal A, et al Consumption of human milk oligosaccharides by gut-related microbes. J. Agric. Food Chem. 58: Markowitz VM, et al The integrated microbial genomes (IMG) system. Nucleic Acids Res. 34:D344 D Minor W, Cymborowski M, Otwinowski Z, Chruszcz M HKL-3000: the integration of data reduction and structure solution. From diffraction images to an initial model in minutes. Acta Crystallogr. D Biol. Crystallogr. 62: Murshudov GN, Vagin AA, Dodson EJ Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallogr. D Biol. Crystallogr. 53: Nagae M, et al Structural basis of the catalytic reaction mechanism of novel 1,2-alpha-L-fucosidase from Bifidobacterium bifidum. J. Biol. Chem. 282: Newstead SL, Watson JN, Bennet AJ, Taylor G Galactose recognition by the carbohydrate-binding module of a bacterial sialidase. Acta Crystallogr. D Biol. Crystallogr. 61: Ninonuevo M, et al Nanoliquid chromatography-mass spectrometry of oligosaccharides employing graphitized carbon chromatography on microchip with a high-accuracy mass analyzer. Electrophoresis 26: Ninonuevo MR, et al A strategy for annotating the human milk glycome. J. Agric. Food Chem. 54: Ninonuevo MR, et al Methods for the quantitation of human milk oligosaccharides in bacterial fermentation by mass spectrometry. Anal. Biochem. 361: Parche S, et al Lactose-over-glucose preference in Bifidobacterium longum NCC2705: glcp, encoding a glucose transporter, is subject to lactose repression. J. Bacteriol. 188: Schneider TR, Sheldrick GM Substructure solution with SHELXD. Acta Crystallogr. D Biol. Crystallogr. 58: Sela DA, et al The genome sequence of Bifidobacterium longum subsp. infantis reveals adaptations for milk utilization within the infant microbiome. Proc. Natl. Acad. Sci. U. S. A. 105: Sela DA, et al An infant-associated bacterial commensal utilizes breast milk sialyloligosaccharides. J. Biol. Chem. 286: Sela DA, Mills DA Nursing our microbiota: molecular linkages between bifidobacteria and milk oligosaccharides. Trends Microbiol. 18: Sulzenbacher G, et al Crystal structure of Thermotoga maritima alpha-lfucosidase. Insights into the catalytic mechanism and the molecular basis for fucosidosis. J. Biol. Chem. 279: Ventura M, et al The Bifidobacterium dentium Bd1 genome sequence reflects its genetic adaptation to the human oral cavity. PLoS Genet. 5:e Wada J, et al Bifidobacterium bifidum lacto-n-biosidase, a critical enzyme for the degradation of human milk oligosaccharides with a type 1 structure. Appl. Environ. Microbiol. 74:

138 Ward RE, Ninonuevo M, Mills DA, Lebrilla CB, German JB In vitro fermentation of breast milk oligosaccharides by Bifidobacterium infantis and Lactobacillus gasseri. Appl. Environ. Microbiol. 72: Xie Y, Liu J, Zhang J, Hedrick JL, Lebrilla CB Method for the comparative glycomic analyses of O-linked, mucin-type oligosaccharides. Anal. Chem. 76: Yew WS, et al Evolution of enzymatic activities in the enolase superfamily: L- fuconate dehydratase from Xanthomonas campestris. Biochemistry 45: Zhang G, Mills DA, Block DE Development of chemically defined media supporting high-cell-density growth of lactococci, enterococci, and streptococci. Appl. Environ. Microbiol. 75:

139 129 SUPPLEMENTARY INFORMATION Table S1: Primers and probes used in this study qpcr Sequence (5-3 ) Blon0248 R Blon0248 F Blon0248 TM Blon0426 F Blon0426 R Blon0426 TM Blon2335 F Blon2335 R Blon2335 TM Blon2336 F Blon2336 R Blon2336 TM 0346F-q 0346R-q AGCGACGGACGAAGTGC ATTCCGCCGTGACGCTT 6FAM-CCACTTCCGAGGCCGGAACAT--BBQ GATGTGGATTCCGCAACC AAACGCACGACCTGAGG 6FAM-TGGCAACGTTGTCGTGGAGGC BBQ CCTGTTCAACCAGGATGAGTC CCGTCCACGACGAAGTAG 6FAM-TCTGGCTTCCATCTGGCCGATCAT--BBQ ATCACGCTCACCCTCCC ACATCGTCGAAGCGGAGT 6FAM-AGCCCACGACGATCAACGCCA--BBQ ACCATCACGGCTACCTGTTC ACGGGCATCCGTTTATTACA Gene cloning Sequence (5-3 ) WF a 0248WR 0426WR BL0248cF BL0248cR BL0426cF BL0426cR BL0346F BL0346R BL2335F BL2335R BL2336F BL2336R ACCAACAACCAGCAACCAAT ATCGAATACGGCACCTTCAG GACCGCCTTCATGGATAAGA CACCATGGTGTTGTTCATGGCCAATC GCGACGGACGAAGTGCACGAC CACCATGGTGTTGTTCATGGCCAATC GTGTCGAGCAAAACGCACGAC CACCATGATGTCCACCGCAACTGGT TTTCAGCGACAACTCGATGA CACCATGAAACTCACTTTCGATGGGA CCTGCGGACAAGCCCCTTGAACG CACCATGAACAATCCTGCAGATGCG GATGCGCACGGCAGCCGCGCGGC

140 130 Table S2: Primers and probes used in this study 1 Not including cloning artifact; 2 Including Bijvoet pairs; 3 (Last resolution bin, A); 4 The residue V206 is located in a sharp turn. The residue E30 is located at the end of a helix and forms a hydrogen bond to Y443 and salt bridges to R351 and R444 respectively. The two residues in each monomer are well-defined in electron density maps.

141 131 Table S3:!-L-fucosidase characteristics Optimum expression conditions Optimum ph enzyme Optimum enzyme Temperature Blon_ C 0.5 mm IPTG 6h C Blon_ C 0.5 mm IPTG 6h C Blon_ C 0.5 mm IPTG 6h C Blon_ C 0.5 mm IPTG 6h C Blon_ C 0.5 mm IPTG 6h C

142 Table S4: Substrate preferences of HMO cluster I fucosidases 132

143 133 Figure S1: Digestion of fucosylated HMO with recombinant fucosidases. HMO standards were obtained from a commercial source. Incubation with Blon_0248 (A), Blon_0426(B), Blon_2335 (C), Blon_2336 (D) are shown.

144 134

145 135

146 136

147 137 Chapter IV!-galactosidases in Bifidobacterium longum subsp. infantis ATCC Active on Plant and Human Milk Oligosaccharides Daniel Garrido 1,4,5,6, Santiago Ruiz-Moyano 2,4,5,6, Rogelio Jimenez-Espinoza 3,6, Hyun-Ju Eom 2,4,5,6, David E. Block 2,3,7 and David A. Mills 2,4,5,6. Departments of 1 Food Science and Technology, 2 Viticulture and Enology, and 3 Chemical Engineering, University of California Davis, CA USA; 4 Foods for Health Institute, 5 Functional Glycobiology Program, and 6 Robert Mondavi Institute for Wine and Food Sciences, University of California Davis. Abbreviations: HMO, human milk oligosaccharides; GOS, galactooligosaccharides; FOS, fructooligosaccharides; TLC, thin layer chromatography; GlcNAc: N-acetylglucosamine; LNB: lacto-n-biose; LacNAc: N-acetyl-lactosamine; LNT: lacto-ntetraose; LNnT: lacto-n-neotetraose. Conceived and designed the experiments: DG DM. Performed the experiments: DG SRM RJM HYE. Analyzed the data: DG DEB DAM. Contributed reagents/materials/analysis tools: DEB. Wrote the manuscript: DG.

148 138 Summary!-galactosidases are critical enzymes that participate in the assimilation of lactose and the bacterial utilization of complex oligosaccharides containing galactose. Bifidobacteria contain several homologs of these enzymes, however only a few have been characterized, and studies have focused on their transglycosylation properties. These enzymes are critical to understand the physiology of the genus and predict their capacities to utilize galactose-containing carbohydrates. In this work we have characterized the properties of five!-galactosidases found in Bifidobacterium longum subsp. infantis, a common isolate of the infant microbiota. They belong to Glycosyl Hydrolase families 2 or 42, and with the exception of Blon_0268, have several homologs in other bifidobacteria. Blon_0268 and Blon_2416 showed activity against galactooligosaccharides and 4 and 6 galactobiose, and were induced when B. infantis grew on GOS. Blon_2016 and Blon_2334 were active on other galactosyl linkages, including Lacto-N-biose and N- acetyl-lactosamine, which represent the core structures of human milk oligosaccharides. These two enzymes also showed the highest kinetic efficiency given their k cat /K m ratio. Blon_2334 showed lactase activity, and Blon_2016 was induced when B. infantis grew on lacto-n-tetraose or lacto-n-neotetraose, two characteristic isomers found in human milk oligosaccharides. These results provide a complete picture of the ability of this infant gut isolate to utilize lactose, and other substrates commonly available in the colon such as human milk and galactooligosaccharides.

149 Introduction The Bifidobacterium genus is characterized by gram-positive strictly anaerobic rods, which are common inhabitants of the intestinal tract of humans [1]. Their abundance is noticeable in the infant gut microbiota, especially in breast-fed infants [2,3], where bifidobacteria can represent up to the 90% of the total bacteria in this environment [4]. Bifidobacteria still represent a significant proportion of the total species found in the adult microbiota [5], however different species can be found in both environments. A health-promoting status has been attributed to some strains in the genus [6]. Some strains are considered as probiotics [7,8], and an increase in bifidobacterial numbers is desirable after consumption of prebiotics [9,10]. One of the most characteristic aspects of Bifidobacterium species is their capability to decompose and use complex oligosaccharides as a carbon and energy source [11,12]. In adults, diet provides the intestinal microbiota with a great variety of oligo- and polysaccharides, which are usually resistant to enzymatic degradation in the intestinal lumen, and therefore reach distal portions of the intestine. Different Bifidobacterium species are capable of metabolizing complex plant oligosaccharides such as cellodextrins and amyloses [13], raffinose [14], arabinooligosaccharides [15,16], xylooligosaccharides [17], FOS and inulin [18], galactans and GOS [19,20,21] among others. GOS are synthetically produced by microbial!-galactosidases, which under specific conditions carry on transglycosylation reactions starting from lactose. These enzymes are widespread in bifidobacteria, and some of them have remarkable yields in GOS synthesis [22,23,24,25,26]. GOS have a degree of polymerization between 1 and 5 and are composed of galactose oligomers in!1-3/4/6 linkages, with a terminal glucose

150 140 residue [27]. These substrates have been extensively studied for their prebiotic status, promoting the growth of beneficial microorganisms such as bifidobacteria, therefore providing different health benefits [9]. In breast-fed infants, the main carbon sources available for the intestinal microbiota are human milk oligosaccharides (HMO) [28]. Only a few bacterial species have been shown to use these substrates [29,30,31], and the mechanisms involved in HMO consumption in B. bifidum and B. longum subsp. infantis (B. infantis) are beginning to be understood [32,33,34,35,36,37]. HMO are characterized by repeats of lacto-n-biose (Gal!1-3GlcNAc), or N-acetyl lactosamine (Gal!1-4GlcNAc), to which fucose and/or sialic acid residues are added at terminal positions. Given the presence of several copies of genes encoding!-galactosidases in different bifidobacteria, it can be predicted that they display different physiological functions or act on different substrates. Besides being competitive in HMO utilization, B. infantis shows a vigorous growth on simple substrates such as lactose, or more complex such as GOS. It is then expected that!-galactosidases in B. infantis display different specificities and their induction be observed only under specific physiological conditions. In order to contribute to our understanding about the adaptations of B. infantis ATCC to use complex oligosaccharides, in this work we have characterized the biochemical properties of its!-galactosidases, their activities on a variety of plant and host oligosaccharides, as well as some of the conditions by which they are induced.

151 Materials and Methods Microorganisms and media. Mann-Rogose-Sharp (MRS) broth supplemented with 0.25% w/v L-cysteine (Sigma-Aldrich) was used for routine growth of B. infantis ATCC under anaerobic conditions (Coy Laboratory Products, Grass Lake, MI) at 37 C. Chemically competent Escherichia coli BL21 Star and Top10 cells were obtained from Invitrogen (Carlsbad, CA), and transformants were cultured at 37 C in Luria Broth with 50 µg/ml Carbenicillin (Teknova, Hollister CA) when necessary. Phylogenetic comparisons. Genes encoding for!-galactosidases in the genome of B. infantis were retrieved by the Integrated Microbial Genome database, IMG; [38]. Homologs to Blon_0268, Blon_2016, Blon_2123, Blon_2334 and Blon_2416 were also identified using IMG, and search was manually curated with other relevant homologous sequences. These were compared by Multiple Sequence Alignment using ClustalW. Phylogenetic representations were obtained using the Maximum Likelihood algorithm using MEGA 5. Gene Expression Analysis. For RNA extraction, B. infantis cells were grown on chemically defined ZMB-1 [39], to which 2% w/v of carbon sources such as glucose, lactose (Sigma), FOS (raftilose Synergy 1, Orafti), inulin (raftiline HP, Orafti), GOS (Purimune, GTC Nutrition) and purified HMO [29] were added. Lacto-N-tetraose (LNT) and lacto-n-neotetraose (LNnT) (V-Labs) were added at 0.5% w/v. Growth was monitored using a PowerWave microplate spectrophotometer (BioTek Instruments, Inc.) under anaerobic conditions at 37 C. Cells at exponential phase were pelleted at x g for 2 min, resuspended in 1 ml of RNA later (Ambion), stored at 4 C overnight and then at -80 C until later use. RNA extraction was performed using the RNAqueous

152 142 Ambion kit (Ambion) and cdna was obtained from 10 µg of RNA using the High Capacity cdna Reverse Transcription kit (Applied Biosystems). For the relative quantification of!-galactosidase encoded genes, the Fast Sybr Green Master Mix (Applied Biosystems) was used, using the gene Blon_0393 as the endogenous control [40], and reaction conditions as recommended by manufacturer. Primer efficiency was determined and normalized in each plate using standard curves for each primer set. The Primer3 software was used for primers design (Table S1), and the q-gene software was used for relative quantification analysis.!-galactosidase gene cloning. B. infantis ATCC genomic DNA was obtained from overnight cultures on MRS, using the MasterPure Gram Positive DNA Purification Kit (Epicentre Biotechnologies), following the manufacturer instructions. Primers used for PCR cloning are shown in table S1. PCR reactions contained 0.2 mm dntps (Fermentas), 1 ng DNA, 0.5 µm of each primer, and 2 U of Phusion DNA Polymerase (Finnzymes) in a 150 µl final volume. PCR was performed in a Verity 96 well thermal cycler (Applied Biosystems), using the following program: initial denaturation at 98 C for 2 min; 35 cycles of denaturation at 98 C 30 s, annealing at 58 C for 30 s, and extension at 72 C 3 min; and a final extension at 72 C for 7 min. PCR products were gel purified (Qiaquick Gel Extraction Kit, Qiagen) and cloned into pet101 using the Champion pet101 Directional TOPO Expression Kit (Invitrogen), following manufacturer instructions. Plasmids were transformed into BL21 star E. coli cells as well as Top10 cells for plasmid storage, and transformants were confirmed for the correct insert sequence by plasmid sequencing using primers T7prom and T7term (Invitrogen).

153 143 Recombinant protein expression and purification. E. coli BL21 Star cells were grown in 100 ml LB broth with 50 µg/ml carbenicillin in a shaker at 250 rpm (Innova-4000, New Brunswick Scientific, Edison, NJ) at 37 C until cultures reached an O.D. of 0.6. Recombinant proteins were induced for 6 hours with the following optimized conditions: 0.5 mm IPTG (USB) at 24 C for Blon_0268 and Blon_2416; 0.5 mm IPTG at 28 C for Blon_2016; 1 mm IPTG at 24 C (Blon_2123) and 0.5 mm IPTG at 28 C for Blon_2334. Cultures were centrifuged in 50 ml falcon tubes at 4000 rpm in an Eppendorf 5804 centrifuge (Eppendorf) for 20 min at 4 C, and pellets were kept at -80 C until use. Cells were resuspended in Bugbuster Protein Extraction Reagent (EMD Chemicals), using 5 ml of the buffer for every 100 ml of culture. Lysozyme (Sigma Aldrich, 50 µl of 50 mg/ml stock), and DNAse I (Roche Applied Sciences; 20 µl of U stock) were added to help in bacterial lysis. The suspensions were vortexed and incubated for 10 min at room temperature, and centrifuged for 20 min at rpm at 4 C. Supernatants were recovered and applied to 1 ml Bio-Scale Mini Profinity IMAC cartridges, connected to an EP-1 Econo-pump (Bio-Rad). Protein purification was performed as recommended by the manufacturer, but proteins were eluted using an imidazole gradient between 20 and 250 mm in washing buffer. Recombinant!-galactosidases were checked for purity and correct molecular weight using 10% SDS-PAGE gels (Bio-Rad). Finally the elution was exchanged for PBS using Amicon Ultra-15 Centrifugal Filter Units, with a cut-off of 50 kda (Millipore). Protein concentrations were determined using the Bio-Rad protein assay, with a standard curve using Bovine Serum Albumin (Sigma). Determination of kinetic parameters. Enzymatic assays were carried out using orthonitrophenyl-!-galactoside (ONPG; Sigma) at a concentration of 2 mg/ml and 1-10 µg of

154 144 each recombinant enzyme. Optimum ph for each enzyme was determined using McIlvaine buffers, with values from 4.0 to 8.0. Reactions were performed in triplicate in 96 microwell plates, and contained 80 µl of each buffer, 15 µl of substrate, and 5 µl of enzyme. Reactions were incubated for 10 min at 37 C, and stopped adding equal volumes of 1M Na 2 CO 3. Absorbance at 420 nm was determined using a Synergy2 microplate reader (Biotek). For determination of optimum temperatures, enzymatic assays were performed at optimum ph and at 4 C, 30 C, 37 C, 45 C, 55 C and 65 C. Relative activity was determined from OD 420 values. Kinetic constants were obtained using substrate concentrations in the range of 0.1 to 4 mm of ONPG and µg of each enzyme. McIlvaine buffer at optimum ph were used, and reaction times were preestablished to be within the initial rate of reaction. Amounts of o-nitrophenol produced in each reaction were calculated from a standard curve and A 420 values. Nonlinear regression was used to determine K m and V max, fitting experimental values to the Michaelis Menten equation, using the tool Solver on Microsoft Excel.!-galactosidase specificity determination. Recombinant enzymes were coincubated in phosphate buffer and 2 µg of the following substrates (V-labs unless indicated): D-lactose (Sigma), 3 Fucosyl lactose, Gal!1-4Gal (4 galactobiose), Gal!1-6Gal (6 galactobiose), Gal!1-3GlcNAc (LNB), Gal!1-4GlcNAc (LacNAc) and 10 µg of commercial galactooligosaccharides (Purimmune GOS). Reactions were carried out in 10 µl for different times at 37 C, and inactivated at 95 C for 5 min. Aliquots of the reactions were spotted in Sigma-Aldrich TLC plates. A mixture of n-propanol, acetic acid and water in a 2:1:1 ratio was used as solvent. Plates were dryed and sprayed with 0.5% "- napthol and 5% H 2 SO 4 in ethanol, and revealed at 150 C for 10 minutes. Specificity was

155 145 also tested with p-nitrophenyl conjugated substrates on colorimetric reactions (purchased from Sigma unless mentioned): 4-nitrophenyl-!-galactopyranoside (Acros Organics, Pittsburgh, PA), 4-nitrophenyl "-D-glucopyranoside, 4-nitrophenyl-"-D-fucopyranoside, 4-nitrophenyl N-acetyl-!-D-glucosaminide and 4-Nitrophenyl N-acetyl-"-Dgalactosaminide. Activity on LNB-pnp (p-nitrophenyl 2-Acetamido-2-deoxy-3-O-("-Dgalactopyranosyl)-"- D-glucopyranoside) and LacNAc (p-nitrophenyl 2-Acetamido-2- deoxy-4-o-("-d-galactopyranosyl)-"- D-glucopyranoside; Toronto Research Chemicals), were evaluated incubating each "-galactosidase with 10 µg of each substrate and 1 µg of Blon_0732, a recombinant "-hexosaminidase from B. infantis (data not shown), in a 100 µl volume. Reactions proceeded for 10 min and OD 405 was measured as described above. Evaluation of relative affinities of!-galactosidases. Equimolar concentrations of lactose (Sigma), 4 galactobiose (V-labs), 6 galactobiose (V-labs), 3 galactosyl lactose (Carbosynth, UK) and lacto-n-tetraose (V-labs) were coincubated with the same amount (1-20 µg) of each of the five recombinant "-galactosidases for 10 min at their optimum temperatures and ph in McIlvaine buffer. Reactions were inactivated by incubation at 95 C for 5 min. The Galactose Assay Kit (Biovision, Mountain View CA) was used to quantify galactose concentrations present in each sample, following the manufacturer instructions. Fluorescence was quantified using a standard curve in a Synergy 2 microplate reader.

156 Results 3.1 Genetic landscapes and phylogenetic representation of B. infantis!- galactosidases. The genome of B. infantis ATCC contains five genes predicted to encode for!-galactosidases, EC (Figure 1). A sixth gene, Blon_0438, is annotated as a!-galactosidase, but apparently it is a truncated enzyme with only 205 aminoacids. Blon_2460 is predicted to be an "-galactosidase and it was not considered in this study. Some of these genes are in the proximity of carbohydrate transporters (Figure 1), suggesting a coregulated transcription. For example Blon_0268 and Blon_2334 are located next to sugar permeases, and Blon_2416 is in a gene cluster containing an ABC transporter with affinity for oligosaccharides and a Glycosyl Hydrolase (GH) family 43, predicted to be a!-xylosidase. Blon_2334 is thought to be part of the HMO cluster I [32], which contains several genes predicted to be important in metabolism of HMO in this bacterium. These enzymes belong to either GH2 or GH42, as defined by the Carbohydrate-Active Enzymes database ( [41]). These families are formed mainly by bacterial!-galactosidases, they have a retaining mechanism and several show transglycosylation activities. A phylogenetic comparison of their aminoacidic sequences with homologs found in the IMG database clearly separates both GH families (Figure 2). This dendrogram shows that the!-galactosidases in B. infantis do not share considerable homology among them, and close homologous sequences are found throughout the genus Bifidobacterium. Only Blon_0268 seems to be unique to this strain (or subspecies), since

157 147 more related sequences are found in other gram-positive species, some of them found in the intestinal microbiota. Close homologs of Blon_2016 and Blon_2416 in B. infantis HL96 have been previously characterized [24,26], but these studies have focused on their transglycosylation properties. 3.2 Gene expression of!-galactosidases in B. infantis. A relative quantification of the expression levels for each!-galactosidase gene was performed growing B. infantis at exponential phase using different carbohydrates or oligosaccharides as the only medium carbon source. Since a lactose preference over glucose exists in B. longum [40], and probably in B. infantis, results were normalized to gene expression levels of cells growing on lactose (Figure 3). We considered significant a level of induction more than two fold, as previously determined to be reliable compared to proteomic data [34]. Two genes, Blon_2123 and Blon_2416, are more than four fold repressed when cells grow on glucose, probably indicating that they are expressed in cells using lactose as a carbon source. On the other side, Blon_0268 was induced eight fold on glucose relative to lactose. This gene was also induced on inulin, which is composed of fructose monomers. Shorter FOS however did not induce the expression of any!-galactosidase. Growth on commercial GOS had the greatest impact, leading to the induction of Blon_0268, Blon_2334 and Blon_2416.

158 148 FIGURE 1: Genetic landscapes for!-galactosidase genes in B. infantis ATCC Locus tags are provided above each arrow. F1SBP: Family 1 Solute binding protein.

159 149 FIGURE 2: Phylogenetic relationships among B. infantis!-galactosidases and close related sequences. Asterisks indicate enzymes present in strain ATCC

160 150 FIGURE 3: Relative quantification of the gene expression of!-galactosidases after growth on several substrates. Results are expressed in a log scale, and error bars represent three biological replicates. Dashed lines indicate a two fold in gene expression.

161 151 When a pool of HMO was used as the only carbon source, apparently none of the enzymes showed a clear induction. Blon_2123 and Blon_2416 were actually repressed as defined previously. It is probably that some of these enzymes have a constitutive expression, and/or some individual HMO exert a catabolic repression on some genes, probably related to a hierarchy in the consumption of certain isomers [34]. This idea is supported by the fact that under growth on lacto-n-tetraose (LNT; Gal!1-3GlcNAc!1-3Gal!1-4Glc), and on lacto-n-neotetraose (LNnT; Gal!1-4GlcNAc!1-3Gal!1-4Glc) Blon_2016 is induced more than two fold, and Blon_2334 gene levels were higher when cells grew on these substrates compared to HMO. We also analyzed the expression levels for some adjacent genes next to Blon_0268 and Blon_2334 (Figure 4), encoding for transporters of the major facilitator superfamily. Even though the carbohydrate affinity of these transporters (Blon_0267, Blon_2331 and Blon_2332) is not known, their induction by GOS, as well as glucose (Blon_0268), provides a notion of the affinity of these importers and their neighbor!- galactosidases. 3.3 Enzymatic characterizations. In order to study some of the physicochemical properties of these enzymes, they were cloned and expressed in E. coli, and purified with a his-tag. These proteins showed a high activity when coincubated with ONPG, and this substrate was used for evaluating different kinetic parameters. Optimum ph values were relatively acidic for Blon_0268 and Blon_2016, and more neutral for the other three enzymes (Table 1). As observed in other!-galactosidases, the optimum temperature varied between 45 C and 55 C for all enzymes, excepting Blon_2334 which showed the

162 152 highest activity at 37 C. Using these conditions, kinetic parameters were determined using ONPG. All enzymes fitted the Michaelis Menten model, and as suggested by their lower K m values, Blon_0268, Blon_2016 and Blon_2334 had more affinity for ONPG compared to the other!-galactosidases, Turnover rates were the highest for Blon_2016, and together with Blon_2334, these enzymes showed the greatest kinetic efficiency given by a k cat /K m ratio over 1x10 6 s -1 M -1. A search on the BRENDA collection [42], indicates that these values are among the highest when compared to other!-galactosidases in other organisms. These values were also 1000 times higher than the kinetic efficiencies observed for Blon_0268 and Blon_2416.

163 153 FIGURE 3: Relative quantification of the gene expression of putative sugar permeases associated to!-galactosidases in B. infantis on the substrates indicated in the x-axis. Dashed lines indicate a two fold in gene expression.

164 154 Table 1: Enzyme kinetic parameters and optimums for B. infantis!-galactosidases. Optimum ph Optimum Temperature K m (mm) k cat (s -1 ) k cat /K m (s -1 M -1 ) Blon_ C - 55 C , Blon_ C , ,943, Blon_ C - 55 C , Blon_ C ,472, Blon_ C ,585.03

165 Relative affinities and substrate specificity of B. infantis!-galactosidases. We also studied the activity of these enzymes on common galactosyl linkages. 4- and 6- galactobiose (Gal!1-4Gal and Gal!1-6Gal), as well as galactosyl lactose (Gal!1-3Gal!1-4Glc) are common products in transglycosylation reactions [27]. Plant polysaccharides also contain these linkages as building blocks. As observed using thin layer chromatography (TLC) plates (Figure 5A, lanes 2-4; 9-11 and Figure 5C, lane 3) and comparing the relative affinities for these substrates (Table 2), Blon_0268 was able release galactose from these three substrates, however with a preference for 6 galactobiose. Blon_2416 showed instead a preference for 4 galactobiose over galactosyl lactose (Figure 5C, lanes 6-7 and 13-14; Figure 5D lane 7). Blon_2123 in the other hand was not particularly active on any of these substrates, only showing a partial preference for 6 galactobiose.

166 156 Table 2: Relative affinities of B. infantis!-galactosidases for different substrates. Results are expressed as a percentage normalized to galactose released from ONPG. Blon_0268 Blon_2016 Blon_2123 Blon_2334 Blon_2416 ONPG Lactose galactobiose galactobiose Gal-lac LNT

167 157 FIGURE 5: Determination of the substrate specificities of B. infantis!-galactosidases on different galactosyl linkages by TLC indicated in the bottom of the figure. (A) Coincubations of Blon_0268 and Blon_2334 for 5, 20 and 60 min. Lanes 1 and 8: standards; lanes 2-4: Blon_0268 with 4-galactobiose; lanes 5-7: Blon_2334 with 4-galactobiose, min; lanes 9-11: Blon_0268 with 6-galactobiose; lanes 12-14: Blon_2334 with 6-galactobiose. Lane 15: Lactose and galactose; lane 16: glucose; lanes 17-19: Blon_0268 with lactose; lanes 20-22: Blon_2334 with lactose. Lane 23: Commercial GOS; lane 24: lactose and galactose; lanes 25-27: Blon_0268 with GOS; lanes 28-30: Blon_2334 with GOS. (B) Coincubations of Blon_2016, Blon_2123 and Blon_2416 with 4 or 6 galactobiose for 20 and 60. Lanes 1 and 8: galactose and 4 or 6 galactobiose standards; lanes 2-3: Blon_2016 on 4-galactobiose; lanes 4-5: Blon_2123 on 4- galactobiose; lanes 6-7: Blon_2416 on 4-galactobiose; lanes 9-10: Blon_2016 on 6-galactobiose; lanes 11-12: Blon_2123 on 6-galactobiose; lanes 13-14: Blon_2416 on 6-galactobiose. (C) Glycolytic activity on 3 galactosyl lactose. Enzymes were incubated with this substrate for 60. Lane 1: 3 galactosyl lactose; lane 2: galactose and lactose; lane 3: Blon_0268; lane 4: Blon_2016; lane 5: Blon_2123; lane 6: Blon_2334; lane 7: Blon_2416. (D) Coincubations of Blon_2016, Blon_2123 and Blon_2416 with commercial GOS for 20 and 60. Lane 1: GOS; lane 2: lactose and galactose; lane 3: glucose; lanes 4-5: Blon_2016; lanes 6-7: Blon_2123; lanes 8-9: Blon_2416.

168 158 As described by Yoshida et al [43], Blon_2334 has a higher lactase activity compared to other enzymes, and Blon_0268 and Blon_2016 that showed only a minor activity on this substrate (Figure 5A, lanes and Table 2). Interestingly, four of the five!-galactosidases including Blon_2334 displayed a significant hydrolytic activity on commercial GOS, releasing galactose and glucose from this complex substrate (Figure 5A, lanes and Figure 5D, lanes 4-9). None of these enzymes showed "-galactosidase, "-fucosidase,!-glucosidase,!- N-Acetyl-glucosaminidase or!-n-acetyl-galactosaminidase activity. Activity on 3 Fucosyl lactose (Gal!1-4Glc"1-3Fuc) was not appreciated for any of these enzymes (data not shown). Blon_2016 is an enzyme that has been recently found to be highly active on type 1 HMO such as lacto-n-tetraose (LNT; Gal!1-4GlcNAc!1-3Gal!1-4Glc), and it had the highest enzymatic efficiency given by the k cat /K m ratio on ONPG. The results presented in this study indicate that this enzyme shows a broader spectrum of specificities, being able to cleave 4 and 6 galactobiose (Figure 5C, lanes 2-3 and 8-9), 3 galactosyl lactose (Figure 5C, lane 4), and GOS (Figure 5D, lanes 4-5), displaying however higher affinity for substrates such as LNT (Table 2). We finally tested the activity of these enzymes on lacto-n-biose (LNB; Gal!1-3GlcNAc) and N-Acetyl-lactosamine (LacNAc; Gal!1-4GlcNAc), core structures of type 1 and type 2 HMO [44]. After coincubating these substrates with B. infantis!- galactosidases, only Blon_2016 and Blon_2334 were active on these linkages (Figure 6). When these substrates, conjugated with p-nitrophenol (LNB-pnp and LacNAc-pnp,) were coincubated with a!-hexosaminidase in B. infantis (Blon_0732, data not shown) and

169 159 each!-galactosidase, release of p-nitrophenol was only appreciated form Blon_2016 and Blon_2334 reactions on both substrates. These results suggest that these enzymes are active on HMO degradation.

170 160 FIGURE 5:!-galactosidases activity on HMO linkages. Lane 1: LNB control; lane 2: galactose; lanes 3 and 4: Blon_2016 on LNB, min; lanes 5 and 6: Blon_2334 on LNB, min; lane 7: LacNAc; lane 8: GlcNAc; lane 9: galactose; lanes 10 and 11: Blon_2016 on LacNAc, min; lanes 12-13: Blon_2334, min; lane 14: galactose.

171 161 FIGURE 6:!-galactosidases acting on HMO linkages. A: Coincubation of B. infantis!- galactosidases with LNB and LacNAc, analyzed on TLC plates. B: Release of p- nitrophenol from LNB-pnp and LacNAc-pnp, after coincubating these substrates with each!-galactosidase and a!-hexosaminidase. Results are presented as relative enzymatic activity, were performed in duplicates and also included activity on ONPG and GlcNAcpnp (results not shown).

172 Discussion In this work we addressed different functional aspects of B. infantis!-galactosidases, such as their phylogenetic relationship with other bifidobacterial enzymes, their induction or repression under specific growth substrates, as well as their specificities and kinetic properties. These results provide clues into the physiology of this infant isolate and how it has evolved to use complex carbohydrates. 4.1 Lactose metabolism.!-galactosidases in B. infantis showed high activity using ONPG as a substrate, however only one of them was active on lactose. Apparently Blon_2334 is a key enzyme in the metabolism of this substrate and the derivation of glucose and galactose into central metabolic pathways. Lactose is the preferential carbon source in close related B. longum NCC2705 [40]. Blon_2334 is homolog to BL0978 in B. infantis, which also has lactase activity. Given the importance of lactose for this microorganism, is probable that Blon_2334 has a constitutive expression, and Blon_2332 and Blon_2331, putative sugar permeases, participate in lactose uptake. It is also interesting that this!-galactosidase is significantly efficient given by its high k cat /K m ratio, which probably is key to the vigorous growth of B. infantis on this disaccharide. 4.2 GOS and plant oligosaccharide metabolism. Synthesis of GOS is achieved using bacterial!-galactosidases. GOS resemble the structure of complex oligosaccharides found in plants that are part of the adult diet. Given that intestinal enzymes do not break down these substrates, these can be used by intestinal bacteria as a carbon source. Therefore, it is expected that molecular responses to GOS in bifidobacteria are similar to those triggered by certain plant oligosaccharides. When B. infantis grew on GOS several

173 163!-galactosidases were induced (Figure 3). Among these, Blon_2416 showed glycolytic activity on GOS, and a Family 1 Solute Binding Protein next to it (Blon_2414; Figure 1), was also induced when B. infantis grew on commercial GOS [34]. Given the specificities of this enzyme is probable that it participates in the metabolism of oligosaccharides containing Gal!1-4Gal or Gal!1-6Gal linkages. Commercial GOS synthesized using Bacillus circulans!-galactosidase render these structures [45]. In addition to the potential involvement of the Blon_2416 cluster in plant oligosaccharide metabolism, Blon_0268 and the genes nearby might constitute another cluster related to this type of metabolism. This!-galactosidase was induced when B. infantis grew on GOS, cleaved 4- and 6- galactobiose, and also released galactose from GOS. Blon_2334 was induced several fold during growth on GOS, releasing also glucose and galactose from this substrate. This information indicates that this enzyme has also a broad spectrum of activities, given that it also efficiently cleaves lacto-n-neotetraose, a type 2 HMO (Yoshida, et al., 2011). D-galactose is a monosaccharide found in several di- and oligosaccharides of human, bacterial and plant origin. Bifidobacteria metabolizes this substrate via the Leloir pathway (Holden et al., 2003), where!-galactose is isomerized to "-galactose by an aldose epimerase (Blon_0897 in B. infantis), and phosphorylated by a galactokinase (Blon_2062). "-1-P galactose enters the Leloir pathway, and at least two sets of genes exist in B. infantis, one encoded by the LNB/GNB cluster (Nishimoto & Kitaoka, 2007), and other by genes Blon_2062 and Blon_2063, which final product is Glc-6-P. Several of these genes are upregulated during growth in HMO, lactose and GOS (Sela, et al., 2008), indicating that galactose released from complex oligosaccharides is utilized in central metabolic pathways in B. infantis.

174 HMO metabolism. The evidence presented in this study suggests the presence of genetic systems potentially competitive for plant oligosaccharide utilization. However, after finding certain gene replacements thought to accommodate functions related to host carbohydrate metabolism [32], B. infantis is predicted to preferentially obtain carbon and energy sources from host-derived oligosaccharides. Two!-galactosidases, Blon_2016 and Blon_2334, showed catalytic activity on LNB and LacNAc, structural backbones of HMO. These enzymes also released galactose from a pool of HMO (data not shown). Interestingly, these two enzymes also showed high enzymatic efficiencies (k cat /K m ), over 1x10 6 s -1 M -1. These values are among the highest reported for any!-galactosidase, as described in the BRENDA database, using ONPG or 4-nitrophenyl-!-galactoside. These results suggest that in addition to genetic and genomic adaptations, improvements in enzymatic catalytic activity is another factor that reflects the microbial adaptation to a particular environment or substrate, as exemplified by B. infantis and the utilization of HMO. Biochemically, Blon_2016 apparently shows a loose specificity, since it was able to break down all the substrates studied except lactose. However, its potential role in HMO consumption is further emphasized by its gene induction under B. infantis growth on LNT or LNnT, two representative HMO isomers. While Blon_2334 presented the same expression level in lactose and HMO, in a previous study this enzyme was detected in B. infantis cells growing on HMO [32]. The role of this enzyme in HMO metabolism was also suggested by its location in the beginning of the HMO cluster I already described. These enzymes possess several homologs in close related bifidobacteria, however it is not clear that they retain the same activities.

175 165 In conclusion, this study presents molecular evidence for the consumption of GOS, a commonly used prebiotic, by a prominent member of the infant gut microbiota, B. infantis. Enzymes releasing galactose from complex HMO have also been identified. Discrete gene products that were up-regulated during growth on these substrates, and their ability to hydrolyze GOS, HMO and core linkages found in these oligosaccharides was characterized. This study provides evidence for the molecular adaptations of members of the intestinal microbiota to GOS and HMO and helps to understand the prebiotic properties of these oligosaccharides.

176 166 Acknowledgements The authors would like to thank Dr. David Sela for his collaboration and critical help on this project. Grants that supported this project were the University of California Discovery Grant Program, the California Dairy Research Foundation, United States Department of Agriculture National Research Initiative-Cooperative State Research, Education, and Extension Service Award and National Institutes of Health-National Institute of Child Health & Human Development award 1R01HD Daniel Garrido currently holds a Fulbright-Conicyt scholarship.

177 167 REFERENCES 1. Turroni F, van Sinderen D, Ventura M (2010) Genomics and ecological overview of the genus Bifidobacterium. Int J Food Microbiol. 2. Gronlund MM, Gueimonde M, Laitinen K, Kociubinski G, Gronroos T, et al. (2007) Maternal breast-milk and intestinal bifidobacteria guide the compositional development of the Bifidobacterium microbiota in infants at risk of allergic disease. Clin Exp Allergy 37: Solis G, de Los Reyes-Gavilan CG, Fernandez N, Margolles A, Gueimonde M (2010) Establishment and development of lactic acid bacteria and bifidobacteria microbiota in breast-milk and the infant gut. Anaerobe 16: Boesten R, Schuren F, Ben Amor K, Haarman M, Knol J, et al. (2010) Bifidobacterium population analysis in the infant gut by direct mapping of genomic hybridization patterns: potential for monitoring temporal development and effects of dietary regimens. Microbial Biotechnology: no-no. 5. Eckburg PB, Bik EM, Bernstein CN, Purdom E, Dethlefsen L, et al. (2005) Diversity of the human intestinal microbial flora. Science 308: Lee JH, O'Sullivan DJ (2010) Genomic insights into bifidobacteria. Microbiol Mol Biol Rev 74: Qiao H, Duffy LC, Griffiths E, Dryja D, Leavens A, et al. (2002) Immune responses in rhesus rotavirus-challenged BALB/c mice treated with bifidobacteria and prebiotic supplements. Pediatr Res 51: McCarthy J, O'Mahony L, O'Callaghan L, Sheil B, Vaughan EE, et al. (2003) Double blind, placebo controlled trial of two probiotic strains in interleukin 10 knockout mice and mechanistic link with cytokine balance. Gut 52: Gibson GR, Probert HM, Loo JV, Rastall RA, Roberfroid MB (2004) Dietary modulation of the human colonic microbiota: updating the concept of prebiotics. Nutr Res Rev 17: Roberfroid M (2007) Prebiotics: the concept revisited. J Nutr 137: 830S-837S. 11. Sela DA, Mills DA (2010) Nursing our microbiota: molecular linkages between bifidobacteria and milk oligosaccharides. Trends Microbiol 18: Pokusaeva K, Fitzgerald GF, van Sinderen D (2011) Carbohydrate metabolism in Bifidobacteria. Genes Nutr. 13. Pokusaeva K, O'Connell-Motherway M, Zomer A, Macsharry J, Fitzgerald GF, et al. (2011) Cellodextrin utilization by bifidobacterium breve UCC2003. Appl Environ Microbiol 77: Dinoto A, Marques TM, Sakamoto K, Fukiya S, Watanabe J, et al. (2006) Population dynamics of Bifidobacterium species in human feces during raffinose administration monitored by fluorescence in situ hybridization-flow cytometry. Appl Environ Microbiol 72: Van Laere KM, Beldman G, Voragen AG (1997) A new arabinofuranohydrolase from Bifidobacterium adolescentis able to remove arabinosyl residues from double-substituted xylose units in arabinoxylan. Appl Microbiol Biotechnol 47: Lagaert S, Pollet A, Delcour JA, Lavigne R, Courtin CM, et al. (2010) Substrate specificity of three recombinant alpha-l-arabinofuranosidases from

178 168 Bifidobacterium adolescentis and their divergent action on arabinoxylan and arabinoxylan oligosaccharides. Biochem Biophys Res Commun 402: Gilad O, Jacobsen S, Stuer-Lauridsen B, Pedersen MB, Garrigues C, et al. (2010) Combined transcriptome and proteome analysis of Bifidobacterium animalis subsp. lactis BB-12 grown on xylo-oligosaccharides and a model of their utilization. Appl Environ Microbiol 76: Omori T, Ueno K, Muramatsu K, Kikuchi M, Onodera S, et al. (2010) Characterization of recombinant beta-fructofuranosidase from Bifidobacterium adolescentis G1. Chem Cent J 4: Hinz SW, Pastink MI, van den Broek LA, Vincken JP, Voragen AG (2005) Bifidobacterium longum endogalactanase liberates galactotriose from type I galactans. Appl Environ Microbiol 71: O'Connell Motherway M, Fitzgerald GF, van Sinderen D (2011) Metabolism of a plant derived galactose-containing polysaccharide by Bifidobacterium breve UCC2003. Microb Biotechnol 4: Barboza M, Sela DA, Pirim C, Locascio RG, Freeman SL, et al. (2009) Glycoprofiling bifidobacterial consumption of galacto-oligosaccharides by mass spectrometry reveals strain-specific, preferential consumption of glycans. Appl Environ Microbiol 75: Van Laere KM, Abee T, Schols HA, Beldman G, Voragen AG (2000) Characterization of a novel beta-galactosidase from Bifidobacterium adolescentis DSM active towards transgalactooligosaccharides. Appl Environ Microbiol 66: Rabiu BA, Jay AJ, Gibson GR, Rastall RA (2001) Synthesis and fermentation properties of novel galacto-oligosaccharides by beta-galactosidases from Bifidobacterium species. Appl Environ Microbiol 67: Hung MN, Lee BH (2002) Purification and characterization of a recombinant betagalactosidase with transgalactosylation activity from Bifidobacterium infantis HL96. Appl Microbiol Biotechnol 58: Hinz SW, van den Brock LA, Beldman G, Vincken JP, Voragen AG (2004) betagalactosidase from Bifidobacterium adolescentis DSM20083 prefers beta(1,4)- galactosides over lactose. Appl Microbiol Biotechnol 66: Hung MN, Xia Z, Hu NT, Lee BH (2001) Molecular and biochemical analysis of two beta-galactosidases from Bifidobacterium infantis HL96. Appl Environ Microbiol 67: Gosling A, Stevens GW, Barber AR, Kentish SE, Gras SL (2010) Recent advances refining galactooligosaccharide production from lactose. Food Chemistry 121: Zivkovic AM, German JB, Lebrilla CB, Mills DA (2010) Microbes and Health Sackler Colloquium: Human milk glycobiome and its impact on the infant gastrointestinal microbiota. Proc Natl Acad Sci U S A. 29. Ward RE, Ninonuevo M, Mills DA, Lebrilla CB, German JB (2006) In vitro fermentation of breast milk oligosaccharides by Bifidobacterium infantis and Lactobacillus gasseri. Appl Environ Microbiol 72: LoCascio RG, Ninonuevo MR, Freeman SL, Sela DA, Grimm R, et al. (2007) Glycoprofiling of bifidobacterial consumption of human milk oligosaccharides

179 169 demonstrates strain specific, preferential consumption of small chain glycans secreted in early human lactation. J Agric Food Chem 55: Xiao JZ, Takahashi S, Nishimoto M, Odamaki T, Yaeshima T, et al. (2010) Distribution of in vitro fermentation ability of lacto-n-biose I, a major building block of human milk oligosaccharides, in bifidobacterial strains. Appl Environ Microbiol 76: Sela DA, Chapman J, Adeuya A, Kim JH, Chen F, et al. (2008) The genome sequence of Bifidobacterium longum subsp. infantis reveals adaptations for milk utilization within the infant microbiome. Proc Natl Acad Sci U S A 105: LoCascio RG, Desai P, Sela DA, Weimer B, Mills DA (2010) Broad conservation of milk utilization genes in Bifidobacterium longum subsp. infantis as revealed by comparative genomic hybridization. Appl Environ Microbiol 76: Garrido D, Kim JH, German JB, Raybould HE, Mills DA (2011) Oligosaccharide binding proteins from Bifidobacterium longum subsp. infantis reveal a preference for host glycans. PLoS One 6: e Kiyohara M, Tachizawa A, Nishimoto M, Kitaoka M, Ashida H, et al. (2009) Prebiotic effect of lacto-n-biose I on bifidobacterial growth. Biosci Biotechnol Biochem 73: Miwa M, Horimoto T, Kiyohara M, Katayama T, Kitaoka M, et al. (2010) Cooperation of {beta}-galactosidase and {beta}-n-acetylhexosaminidase from bifidobacteria in assimilation of human milk oligosaccharides with type 2 structure. Glycobiology. 37. Sela DA, Li Y, Lerno L, Wu S, Marcobal AM, et al. (2011) An infant-associated bacterial commensal utilizes breast milk sialyloligosaccharides. J Biol Chem 286: Markowitz VM, Korzeniewski F, Palaniappan K, Szeto E, Werner G, et al. (2006) The integrated microbial genomes (IMG) system. Nucleic Acids Res 34: D Zhang G, Mills DA, Block DE (2009) Development of chemically defined media supporting high-cell-density growth of lactococci, enterococci, and streptococci. Appl Environ Microbiol 75: Parche S, Beleut M, Rezzonico E, Jacobs D, Arigoni F, et al. (2006) Lactose-overglucose preference in Bifidobacterium longum NCC2705: glcp, encoding a glucose transporter, is subject to lactose repression. J Bacteriol 188: Cantarel BL, Coutinho PM, Rancurel C, Bernard T, Lombard V, et al. (2009) The Carbohydrate-Active EnZymes database (CAZy): an expert resource for Glycogenomics. Nucleic Acids Res 37: D Scheer M, Grote A, Chang A, Schomburg I, Munaretto C, et al. (2011) BRENDA, the enzyme information system in Nucleic Acids Res 39: D Yoshida E, Sakurama H, Kiyohara M, Nakajima M, Kitaoka M, et al. (2011) Bifidobacterium longum subsp. infantis uses two different {beta}-galactosidases for selectively degrading type-1 and type-2 human milk oligosaccharides. Glycobiology. 44. Kunz C, Rudloff S, Baier W, Klein N, Strobel S (2000) Oligosaccharides in human milk: structural, functional, and metabolic aspects. Annu Rev Nutr 20:

180 Usui T, Kubota S, Ohi H (1993) A convenient synthesis of beta-d-galactosyl disaccharide derivatives using the beta-d-galactosidase from Bacillus circulans. Carbohydr Res 244:

181 171 Table S1: Primers used in this study 1. Gene cloning of B. infantis!-galactosidases. Primer Sequence (5 3 ) Blon_0268F Blon_0268R Blon_2016F Blon_2016R Blon_2123F Blon_2123R Blon_2334F Blon_2334R Blon_2416FN Blon_2416RN CACCATGGAGCGAATCCAATACCC CTGCACGTAGCCGTAGGTGTG CACCATGGAACATAGAGCGTTCAAG CAGCTTGACGACGAGTACGCCGTT CACCTGCCGCGGTGCGCACCACCGT CGGGTTCGGGCGTTTCATCACGA CACCATGACAGACGTCACACATGTC GATCAGCTCGAGATCGACGTCGAG CACCATGACCGACACCATGGCACACACC TGCCGCGGTGCGCACCACCGTCAC 2. qpcr of B. infantis genes (Locus tag indicated in primer name) Primer Sequence (5 3 ) 268qF 268qR 2016qF 2016qR 2123qF 2123qR 2334qF 2334qR 2416qF 2416qR 0267qF 0267qR TCTGCTACACGCAGACCTCC ACACCTGCACGTAGCCGTAG AGCACTACAAGGACAACCCCTAT TACTTCTTCTCGCACCACTTCTG CATGAAATCAACCATTGGCAT ACGATGGTCTTGTATGAACTCCA TGTTGCCGTACTCCAGCTTC GGGATATGGTACTGGGGGTG TCACCAGAACCTCATGGAC ATCAGATCGTGCTTGAACGC CTGTGGGGCACCTGCTACGC CGCGAGCATCACCGGCATGA

182 qF 2331qR 2332qF 2332qR CCGCCCAGATGATCGCCGTC ACGGAGGAGACCAGACCGCC ATCCCGGCGTTGTCCTCCGA CGGTCCAACCCGACTGGTGC

183 173! Chapter V!! Release and Utilization of N-acetyl-D-glucosamine from Human Milk Oligosaccharides by Bifidobacterium longum subsp. infantis. Daniel Garrido 1,3,4,5, Santiago Ruiz-Moyano 2,3,4,5 and David A. Mills 2,3,4,5. Departments of 1 Food Science and Technology and 2 Viticulture and Enology, University of California Davis, CA USA; 3 Foods for Health Institute, 4 Functional Glycobiology Program, and 5 Robert Mondavi Institute for Wine and Food Sciences, University of California Davis. 1 Shields Ave, Davis CA Abbreviations: HMO: human milk oligosaccharides; LNT: lacto-n-tetraose; LNnT: lacto- N-neotetraose; LNB: lacto-n-biose; GlcNAc: N-acetyl-D-glucosamine; GNB: galacto-nbiose; LNH: lacto-n-hexaose. Conceived and designed the experiments: DG, DAM. Performed the experiments: DG, SRM. Analyzed the data: DG, DAM. Wrote the manuscript: DG. Sections of this manuscript have been submitted for publication to Anaerobe.!

184 174! Summary Human milk contains high amounts of complex oligosaccharides, which can be utilized especially by Bifidobacterium species in the infant gut as a carbon and energy source. N- acetyl-d-glucosamine is a building block of these oligosaccharides, and molecular details on the release and utilization of this monosaccharide are not fully understood. In this work we have studied some of the enzymatic properties of three!-hexosaminidases encoded by the genome of the intestinal isolate Bifidobacterium longum subsp. infantis ATCC and the gene expression of the corresponding genes during bacterial growth on human milk oligosaccharides. These enzymes belong to the Glycosyl Hydrolase family 20, with several homologs in bifidobacteria. Their optimum ph was 5.0 and had an optimum temperature of 37 C. The three enzymes were active on the GlcNAc!1-3 linkage found in lacto-n-tetraose, the most abundant human milk oligosaccharide. Blon_0459 and Blon_0732, but not Blon_2355, cleaved branched GlcNAc!1-6 linkages found in lacto-n-hexaose, another oligosaccharide abundant in breast milk. B. infantis!- hexosaminidases were induced during early growth in vitro on HMO, and also during growth on lacto-n-tetraose or lacto-n-neotetraose. The up-regulation by HMO isomers of enzymes that convert this monosaccharide into UDP-N-acetyl-glucosamine suggested that this activated sugar is used in peptidoglycan biosynthesis. These results emphasize the complexity of HMO consumption by this infant gut isolate, and provide new clues into this process.!

185 175! 1. Introduction Human milk represents the first food being introduced to breast-fed infants. It meets all the nutritional requirements of the newborn, and its composition has been refined through evolution to ensure the wellbeing of the infant [1]. An increasing number of positive health outcomes has been linked to breast milk intake in early infancy [2,3]. Breastfeeding has a profound impact on the establishment of the intestinal microbiota. Human milk oligosaccharides (HMO) are present in high concentrations in breast milk [4], and certain bacterial populations in the infant gut microbiota are able to use HMO as a carbon source, especially those belonging to the Bifidobacterium genus [5,6]. Evidence has also assigned HMO a protective role against pathogen colonization [7]. HMO are composed by five monosaccharides: glucose (Glc), galactose (Gal), fucose [8], N-acetyl-neuraminic acid (or sialic acid, NeuAc) and N-acetyl-glucosamine (GlcNAc). Several combinations of glycosidic bonds add structural complexity to HMO [4,9], which are not degradable by intestinal enzymes. Newer analytical techniques have recently expanded the number of known HMO species [10,11], revealing a constellation of structures that are not represented in other mammals [12]. N-acetyl-glucosamine (GlcNAc) forms part of the building blocks found in HMO, such as lacto-n-biose (Gal!1-3GlcNAc) and N-acetyl lactosamine (Gal!1-4GlcNAc), as well as different core structures found in glycoproteins and glycolipids abundant in the intestinal lumen as secretions [13]. GlcNAc is also a structural component of bacterial peptidoglycan, for which metabolic pathways have been well described, based on the de novo synthesis of UDP-GlcNAc [14,15].!

186 176! Bifidobacterium longum subsp. infantis (B. infantis) is among the few species consistently found to colonize the gastrointestinal tract in breast-fed infants [16,17]. The ability of this bacterium to vigorously utilize HMO as the only carbon source in vitro [18] has been explained by a unique genetic divergence from other bifidobacteria to metabolize these substrates [19]. Some of these adaptations include the replacement of certain gene clusters for other genes associated to the utilization of fucose and sialic acid [20]; an abundance of ABC transporters and associated Family 1 Solute Binding Proteins (F1SBPs) with affinity for different HMO structures [21]; and finally a higher number of predicted glycosyl hydrolases that target linkages present in HMO. These studies suggest that consumption of HMO by B. infantis includes the import of intact oligosaccharides [21,22] followed by degradation of HMO via the action of intracellular!1-2/3/4 fucosidases [23],!2-3/6-sialidades [24], and "1-3/4 galactosidases [25]. "- hexosaminidases are another set of enzymes thought to play an important role in the consumption of HMO by B. infantis. In this work we examined some of the properties of the three "-hexosaminidases present in B. infantis, as well as the different patterns of gene expression on GlcNAc metabolic pathways induced by different oligosaccharides including HMO.!

187 177! 2. Materials and methods Bacteria and media. For routine experiments, B. longum subsp. infantis ATCC was grown on de Mann-Rogose-Sharp broth, supplemented with 0.05 % w/v L-cysteine (Sigma-Aldrich, St. Louis, MO). Cells were anaerobically grown in a vinyl chamber (Coy Laboratory Products, Grass Lake, MI) at 37 C, 5% carbon dioxide, 5% hydrogen, and 90% nitrogen. Competent Escherichia coli BL21 Star and Top10 cells were from Invitrogen (Carlsbad, CA). RNA extraction and quantitative PCR. B. infantis cells were grown on chemically defined Zhang-Mills-Block 1 media (ZMB-1, [26]), supplemented with 2% w/v of carbon sources such as lactose or glucose (Sigma); purified HMO [27]; and lacto-n-tetraose or lacto-n-neotetraose (V-labs, Covington LA). Optical density at 600 nm was continuously assayed using a PowerWave microplate spectrophotometer (BioTek Instruments, Inc., Winoosky, VT), under anaerobic conditions. Each experiment was done at least in duplicate. Cells at exponential phase growing on the different substrates were centrifuged at x g for 2 min, and resuspended in 1 ml of RNA later (Ambion, Austin, TX), stored overnight at 4 C and then at -80 C until use. Cells growing on HMO were recovered at different OD values representing early exponential phase, mid-exponential phase I, mid-exponential phase II, late exponential phase and stationary phase (Figure S2). B. infantis cells growing on LNT or LNnT were also recovered at early and midexponential phase. RNA extraction and cdna conversion were performed as previously described [21]. RNA quality was checked in a Bioanalyzer 2100 using the RNA 6000 Nano Kit (Agilent Technologies, Foster City, CA). Relative quantification for each gene listed in Table S1 was obtained as previously described [21], using the Fast Sybr Green!

188 178! Master Mix (Applied Biosystems, Foster City, CA), and reaction conditions as recommended by manufacturer. Primers for qpcr were designed using the Primer3 software (Table S1). Bioinformatic analyses. The Integrated Microbial Genomes (IMG;! [28]) database was used: to examine the genetic landscapes of genes predicted to participate in GlcNAc metabolism in the genome of B. infantis ATCC 15697; to infer metabolic pathways found in the genome of B. infantis ATCC Molecular cloning and protein purification.!-hexosaminidase genes (Blon_0459, Blon_0732 and Blon_2355) were PCR-amplified from chromosomal DNA of B. infantis ATCC using the primers in Table S1. PCR reactions contained 0.5 µm of each primer, 1 ng DNA, 0.2 mm dntps (Fermentas, Glen Burnie, MD), and 2 U of Phusion DNA Polymerase (Finnzymes, Vantaa, Finland) in a 150 µl final volume. PCR was performed in a PTC200 Thermo Cycler (MJ Research, Ramsey, MN), using the following program: initial denaturation at 98 C for 30 s, 35 cycles of denaturation at 98 C 10 s, annealing at 58 C for 30 s, extension at 72 C 2 min, and a final extension at 72 C for 7 min. PCR products were gel purified (Qiaquick Gel Extraction Kit, Qiagen) and cloned into pet101 using the Champion pet101 Directional TOPO Expression Kit (Invitrogen, Carlsbad, CA). The proper cloning of these fragments was confirmed by plasmid sequencing using primers T7prom and T7term (Invitrogen). BL21 Star clones containing the recombinant genes were grown in 100 ml LB broth with 50 µg/ml carbenicillin at 37 C in a shaker at 250 rpm (Innova-4000, New Brunswick Scientific, Edison, NJ) until cultures reached and O.D. of 0.6, and induced for 6 hours with 0.5 mm IPTG at 28 C (Blon_0459 and Blon_0732), or with 1 mm IPTG at 37 C (Blon_2355).!

189 179! Bacterial cells were pelleted at 2000 x g in an Eppendorf 5804 centrifuge (Eppendorf, Hauppauge, NY) for 20 min at 4 and stored at -80 C. Cell pellets were resuspended in Bugbuster Protein Extraction Reagent (EMD Chemicals,!Gibbstown, NJ), using 5 ml of the detergent for every 50 ml of culture. To obtain complete lysis, lysozyme (Sigma Aldrich, 50 µl of 50 mg/ml stock), and 200 U DNase I (Roche Applied Sciences, IN) were added. The suspension was vortexed and incubated for 10 min at room temperature, and centrifuged for 20 min at x g at 4 C. The supernatant was applied to Bio- Scale Mini Profinity IMAC cartridges, connected to an EP-1 Econo-pump (Bio-Rad, Hercules, CA), and proteins were purified as recommended by the manufacturer. The purity and molecular weight of each recombinant protein was evaluated using 10% SDS- PAGE gels (Bio-Rad). Buffer was exchanged for PBS using Amicon Ultra-15 Centrifugal Filter Units, with a cut-off of 10 kda (Millipore, Billerica, CA). Determination of kinetic parameters. Relative enzymatic activity was determined using 2 mg/ml of 4-nitrophenyl-N-Acetyl-!-D-Glucosaminide (GlcNAc-pnp; Sigma) and 10 µg of each recombinant enzyme. For optimum ph determinations, McIlvaine buffers [29], with ph from 4.0 to 8.0 were prepared, and reactions contained 80 µl of each buffer, 10 µl of substrate, and 10 µg of enzyme in 100 µl. Reactions were performed in triplicate in 96 microwell plates and incubated for 10 min at 37 C. After stopping each reaction by adding equal volumes of 1M Na 2 CO 3, absorbance at 405 nm was read (Synergy2 microplate reader, Biotek, Winoosky, VT). Optimum temperatures were determined at each enzyme optimum ph, and reactions were carried out 4 C, 20 C, 30 C, 37 C, 45 C, 55 C and 65 C. Relative activity was determined from A 405 reads. Kinetic constants were obtained using substrate concentrations in the range of 0.1 and 4 mm of GlcNAc-!

190 180! pnp and 100 µg of each enzyme in 50 mm of Na 2 HPO 4 under optimum conditions and times within the initial rate of reaction. Amounts of p-nitrophenol produced were estimated from a standard curve and A 405 values. Non-linear regression was used to determine K m and V max, using the tool Solver in Microsoft Excel. Enzymatic specificity using 4-nitrophenyl- conjugated substrates was also determined using 10 µg of enzyme, under optimal conditions for ten minutes, with 10 µl of each of these substrates at 2 mg/ml in a 100 µl volume reaction (purchased from Sigma unless mentioned): 4- nitrophenyl-!-galactopyranoside (Acros Organics, Pittsburgh, PA), 4-nitrophenyl "-Dglucopyranoside, 4-nitrophenyl-!-D-fucopyranoside, 4-nitrophenyl N-acetyl-!-Dglucosaminide and 4-Nitrophenyl N-acetyl-"-D-galactosaminide. Thin Layer Chromatography. Chitobiose (GlcNAc"1-4GlcNAc; V-labs) at a concentration of 1 mg/ml was coincubated with 10 µg of each "-hexosaminidase in 50 mm Na 2 HPO 4 at ph 5.0. Reactions were carried out for 60 min at 37 C, and products were inactivated at 95 C for 5 minutes. Lacto-N-tetraose and lacto-n-hexaose at 1 mg/ml (0.5 µg final) were also coincubated with these enzymes, but with and without the addition of 10 units of "1-3 or "1-4 galactosidase, or both (New Englands Biolabs), for 1 h at 37 C and ph 5.5. Reactions were heat inactivated 95 C for 5 minutes and spotted in a TLC glass Silica-gel plates (Sigma). A mixture of ethyl acetate, acetic acid and water in a 2:1:1 ratio was used as solvent. After drying, plates were sprayed with 0.5%!-napthol and 5% H 2 SO 4 in ethanol. Plates were dried and revealed at 150 C for 10 minutes.!

191 181! 3. Results 3.1 Distribution of!-hexosaminidase genes in B. infantis. A search on the genome of B. infantis ATCC indicated three genes encoding!-hexosaminidases, each containing the pfam motif and belonging to the glycosyl hydrolase family 20 (GH20). The encoded enzymes lack of signal peptides therefore they are predicted to be intracellular, and sequence identity is in average 50% among them. While Blon_0459 and Blon_0732 are located separately in the genome with no evident functional association to other neighboring genes, Blon_2355 is part of the HMO cluster I, a unique locus found in B. infantis species that contains several F1SBPs, ABC permeases and glycosyl hydrolases related to the import and degradation of HMO [19][30]. Interestingly, the protein sequences of B. infantis!-hexosaminidases do not significantly resemble those of B. bifidum, another infant gut isolate with known ability to consume HMO [31]. Phylogenetically related genes encoding!-hexosaminidases are presented in Figure S Enzymatic properties and substrate specificity of B. infantis!-hexosaminidases. In order to understand the possible role of these enzymes on HMO degradation, we cloned, expressed and purified these proteins from E. coli using an N-terminal his-tag. A partial enzymatic characterization using 4-Nitrophenyl N-acetyl-!-D-glucosaminide (GlcNAc-pnp) indicated that these three enzymes had a ph optimum of 5.0 and reached maximum activity at 37 C (Table 1). While K m and k cat values for GlcNAc-pnp were higher for Blon_0732, Blon_2355 showed the higher enzymatic efficiency compared to the other!-hexosaminidases as determined by the k cat /K m ratio (Table 1). Interestingly,!

192 182! none of these enzymes showed appreciable!-n-acetyl-d-galactosaminidase activity, and no detectable activity was observed against Glc-!-pnp, Gal-!-pnp, GlcNAc-"-pnp, GalNAc-"-pnp or Fuc-"-pnp substrates (data not shown).!

193 183! Table 1:!-hexosaminidases optimum reaction conditions and kinetics Optimum ph Optimum Temperature K m (mm) K cat (s -1 ) k cat /K m (s -1 M -1 ) Blon_ C 0.33 ± ± x10 5 ± 6.49x10 3 Blon_ C 1.64 ± ± x10 4 ± 5.43x10 2 Blon_ C 0.48 ± ± x10 6 ± 3.73x10 4!

194 184! All three enzymes degraded chitobiose (GlcNAc!1-4GlcNAc; Figure S1), as determined using TLC. Lacto-N-tetraose (LNT) and lacto-n-hexaose (LNH) are abundant isomers found in human milk, and they were used to study some of the linkage preferences of these enzymes (Figure 1). Incubation of LNT with a commercial!1-3 galactosidase generated galactose and lacto-n-triose. Coincubation of this!- galactosidase and Blon_0459, Blon_0732 or Blon_2355 generated a spot with the same migration as lactose, indicating that these three enzymes are able to cleave the GlcNAc!1-3Gal linkage (Figure 1, lanes 9-12). Unlike LNT, LNH is a branched oligosaccharide. Using a!1-3 galactosidase, LNH was reduced to a pentamer, exposing a GlcNAc!1-3Gal linkage that was successfully cleaved by Blon_0459 and Blon_0732, and to a lesser extent by Blon_2355 (Figure 1, lanes 17-19). Using a!1-4 galactosidase, the pentasaccharide generated contained now a GlcNAc!1-6Gal linkage, which was cleaved by Blon_0732, partially by Blon_0459, but not at all by Blon_2355 (Figure 1, lanes 20-22). Finally, co-incubation of LNH with the two specific!-galactosidases generated a tetramer, which after addition of Blon_0459 or Blon_0732 was reduced to galactose, GlcNAc and glucose (Figure 1, lanes 23-25). This catalytic activity was not observed for Blon_2355.!

195 185! Figure 1: Thin layer chromatography of co-incubations of B. infantis!-hexosaminidases with LNT or LNH after treatment with!-galactosidases. Structures are illustrated below the figure. Lanes 1-8 and 26-28: standards (as indicated in the figure); lane 9: LNT with specific!1-3 galactosidase; lane 10-12: LNT with a!1-3 galactosidase and Blon_0459, Blon_0732 or Blon_2355. Lane 13: LNH; lanes 14-16: LNH with either a!1-3, a!1-4 or both specific galactosidases; lanes 17-19: LNH with a!1-3 galactosidase and Blon_0459, Blon_0732 and Blon_2355 respectively; lanes 20-22: LNH with a!1-4 galactosidase and either Blon_0459, Blon_0732 and Blon_2355; lanes 23-25: LNH with both!1-3 and!1-4 galactosidase, as well as either Blon_0459, Blon_0732 and Blon_2355.!

196 186! 3.3 Gene expression of!-hexosaminidases in B. infantis under growth on different carbohydrates. The gene expression of the three encoded!-hexosaminidases in B. infantis was first evaluated at different time points of growth using HMO as the sole carbon source (Figure S2). These levels were compared to those of B. infantis growing exponentially on lactose. All three genes were induced over two-fold at early exponential phase (Figure 2A), suggesting their participation in HMO metabolism. Interestingly, a gradual decrease in their expression was observed at mid and late exponential phase, as well as during stationary phase where these genes were apparently repressed. A similar trend was observed for B. infantis cells grown on LNT or LNnT (Figure 2B). The three encoded!-hexosaminidases were up-regulated during early growth on these substrates, but their expression drastically decreased during mid-exponential growth. Only Blon_2355 was significantly expressed at mid exponential growth on LNnT, but not on LNT. 3.4 Gene expression of GlcNAc-related metabolic pathways. B. infantis possess several gene clusters that potentially participate in the metabolism of GlcNAc and other monosaccharides besides the HMO cluster I (Figure S3). Several genes in the LNB/GNB cluster (genes Blon_2171 to Blon_2177 in B. infantis) were shown to be up-regulated to a different extent during growth on LNT or LNnT at mid exponential phase (Figure 3A). These genes include Blon_2173, encoding an N-acetylhexosamine-1-kinase, Blon_2174, a lacto-n-biose phosphorylase, and Blon_2172, galactose-1-phosphate uridyltransferase. The expression levels for Blon_2171 were not significantly altered but this gene shows a constitutive expression [19]. Their enzymatic products constitute a Leloir-like pathway that generates Glc-1-P and UDP-GlcNAc using LNB or GNB as substrates [32,33].!

197 187! Figure 2: Relative quantification of the gene expression levels of!-hexosaminidase genes at different time points of growth on HMO (A), LNT or LNnT (B). Expression is relative to their levels during growth on lactose. Error bars represent two biological replicates.!

198 188! Blon_0882, a gene encoding a GlcNAc-6-phosphate deacetylase (naga), and Blon_0881, which gene product is a glucosamine-6-phosphate isomerase (nagb), were induced during exponential growth on HMO, LNT and LNnT (Figure 3B). These genes are located next to an ABC importer that was previously shown to bind specific monoand disaccharides containing GlcNAc [21]. Finally another set of genes in the genome of this bacterium represents a phosphotransferase (PTS) system with predicted affinity for GlcNAc. These genes, Blon_2470 (IIA subunit) and Blon_2471 (IIBC subunit), as well as other components in the PTS system, were repressed between 5-10 fold during logarithmic growth using HMO, LNT or LNnT as the sole carbon source (Figure 3C). Only Blon_0177 and Blon_0178, encoding a phosphocarrier protein and a phosphoenolpyruvate protein phosphotransferase respectively, were induced by glucose, probably associated to another PTS system that imports glucose in B. infantis (Blon_2183).!

199 189! Figure 3: Fold change in gene expression for genes predicted to participate in the metabolism of GlcNAc in B. infantis after exponential growth on the substrates listed in the x-axis. Results are normalized to levels on lactose (dashed line), and error bars represent three biological replicates. A: Genes in the LNB/GNB pathway; B: GlcNAc-6-P deacetylase and GlcN-6-P isomerase; C: (results are presented in log scale) a PTS system putative for GlcNAc import and metabolism.!

200 190! 4. Discussion Certain infant-borne Bifidobacterium species are known to consume HMO, one of the most abundant components of human milk [20,22]. B. infantis is characterized by the preferential import of intact HMO with different degrees of polymerization [18]. The lack of lacto-n-biosidase activity in the strain ATCC suggests that glycosyl hydrolases act sequentially on different HMO isomers. An!-sialidase encoded by the HMO cluster I released sialic acid from HMO [24]. Recently!-fucosidases with activity on different fucosylated HMO [23] and "-galactosidases acting on "1-3/4 linkages of HMO have been characterized [25]. In this work we have studied the properties and expression of three "- hexosaminidases encoded in the genome of B. infantis ATCC The corresponding genes share only 26-28% identity with BbhI, a "-hexosaminidase found in B. bifidum JCM 1254 and active on LNnT and partially on LNH [31]. Interestingly, "- hexosaminidases in B. infantis showed different substrates affinities. Blon_2355 activity was apparently limited to linear oligosaccharides such as LNT, while Blon_0732 and to a lesser extent Blon_0459 cleaved both linear GlcNAc"1-3Gal in LNT and branched GlcNAc"1-6Gal as found in LNH. The higher enzymatic efficiency of Blon_2355, given by its k cat /k m ratio (Table 1) might also indicate that this enzyme is highly specialized in the release of GlcNAc from linear oligosaccharides. Considering their enzymatic activities and their induction during growth on HMO, the results in this study suggests that Blon_0459, Blon_0732 and Blon_2355 participate in the release of GlcNAc from HMO. Interestingly, their up-regulation was restricted to the early exponential phase, and conversely their expression apparently is down-regulated!

201 191! during mid and late logarithmic growth as well as stationary phase. This result was not expected, since other important genes associated to HMO consumption were readily induced during mid and late exponential phase [21,23]. This tight regulation of the expression of these genes might still allow B. infantis!-hexosaminidases to be active during subsequent phases of growth. Bacterial transcriptional responses to HMO are therefore complex and dymami, and this is likely to be related to the heterologous nature of HMO, based on chemically diverse isomers found at different concentrations. Genomic analysis indicated three different genetic clusters potentially involved in the metabolism of GlcNAc by B. infantis (Figure S3). Genes present in the LNB/GNB cluster were shown to be up-regulated during B. infantis growth on LNT and LNnT (Figure 3A). This cluster includes an ABC importer with affinity for LNB and GNB, a lacto-n-biose phosphorylase (LnbP) and two other enzymes that constitute an alternative Leloir pathway, GalT and GalE [32]. It is possible that an N-acetylhexosamine-1-kinase in this cluster (Blon_2173; NahK), participates in GlcNAc metabolism, generating GlcNAc-1-P from GlcNAc (Figure S4A). However, downstream pathways metabolizing GlcNAc to glycolysis require GlcNAc-6-P [34,35]. The genome of B. infantis does not contain any evident function that can interconvert these substrates, except genes that participate in the de novo biosynthesis of UDP-GlcNAc (Figure S4B), catalyzing the inverse reaction. An alternative is the conversion of GlcNAc-1-P into UDP-GlcNAc by Blon_2172, an encoded galactose-1-phosphate uridyltransferase with affinity for GlcNAc [33], also induced by LNT and LNnT. It is possible that under these conditions, the UDP-GlcNAc formed is used directly in peptidoglycan synthesis, linking HMO consumption and GlcNAc release with cell wall biogenesis. This idea of GlcNAc being used as a!

202 192! peptidoglycan building block is also supported by the down-regulation of a PTS system specific for GlcNAc import (Figure 3C), which generates GlcNAc-6-P from GlcNAc. Mutational and biochemical experiments are required to support this hypothesis. Growth on HMO and some of their isomers also led to the up-regulation of GlcNAc- 6-P deacetylase (Blon_0882; nagb) and Blon_0881 (glucosamine-6-p isomerase). If GlcNAc released from HMO is converted into GlcNAc-1-P or UDP-GlcNAc, it is possible that the induction of Blon_0882 and Blon_0881 is more associated with metabolism of sialic acid. This pathway includes the formation of GlcNAc-6-P (Figure S4C). Other genes in this pathway were also shown to be up-regulated during growth on HMO compared to glucose and lactose (Figure S5). Based on these results we suggest that under these conditions sialic acid present in HMO represents a carbon and energy source for B. infantis, while GlcNAc might be used as a building block in peptidoglycan synthesis. In summary, this study presents evidence for molecular mechanisms found in the infant gut isolate B. infantis, associated to the release of GlcNAc from complex HMO. We showed that encoded!-hexosaminidases are active on HMO, and are up-regulated only during early growth on HMO and some of its isomers. This work also suggests that GlcNAc could be directly used as a substrate for peptidoglycan synthesis, while other monosaccharides found in HMO could be used in glycolysis. This results complement previous work presented by our group and others, unraveling the adaptations this microorganism has evolved to respond and utilize human milk components such as HMO.!

203 193! Acknowledgements Daniel Garrido was funded in part through a Fulbright-Conicyt Chile scholarship, and Santiago Ruiz-Moyano was supported by the Ministry of Education and Science of Spain and University of Extremadura, Spain. This work was supported by grants from the University of California Discovery Grant Program, the California Dairy Research Foundation and National Institutes of Health NICHD Awards R01HD059127, R01HD065122, and R01HD The authors would like to thank Dr. David Sela for his important help on the elaboration of this work.!!

204 194! REFERENCES 1. German JB, Freeman SL, Lebrilla CB, Mills DA (2008) Human milk oligosaccharides: evolution, structures and bioselectivity as substrates for intestinal bacteria. Nestle Nutr Workshop Ser Pediatr Program 62: ; discussion Le Huerou-Luron I, Blat S, Boudry G (2010) Breast- v. formula-feeding: impacts on the digestive tract and immediate and long-term health effects. Nutr Res Rev 23: Adlerberth I, Wold AE (2009) Establishment of the gut microbiota in Western infants. Acta Paediatr 98: Kunz C, Rudloff S, Baier W, Klein N, Strobel S (2000) Oligosaccharides in human milk: structural, functional, and metabolic aspects. Annu Rev Nutr 20: Marcobal A, Barboza M, Froehlich JW, Block DE, German JB, et al. (2010) Consumption of human milk oligosaccharides by gut-related microbes. J Agric Food Chem 58: Kiyohara M, Tachizawa A, Nishimoto M, Kitaoka M, Ashida H, et al. (2009) Prebiotic effect of lacto-n-biose I on bifidobacterial growth. Biosci Biotechnol Biochem 73: Newburg DS, Ruiz-Palacios GM, Morrow AL (2005) Human milk glycans protect infants against enteric pathogens. Annu Rev Nutr 25: Blum S, Heller ED, Krifucks O, Sela S, Hammer-Muntz O, et al. (2008) Identification of a bovine mastitis Escherichia coli subset. Vet Microbiol 132: Ninonuevo MR, Park Y, Yin H, Zhang J, Ward RE, et al. (2006) A strategy for annotating the human milk glycome. J Agric Food Chem 54: Wu S, Tao N, German JB, Grimm R, Lebrilla CB (2010) Development of an annotated library of neutral human milk oligosaccharides. J Proteome Res 9: Wu S, Grimm R, German JB, Lebrilla CB (2011) Annotation and structural analysis of sialylated human milk oligosaccharides. J Proteome Res 10: Tao N, Wu S, Kim J, An HJ, Hinde K, et al. (2011) Evolutionary glycomics: characterization of milk oligosaccharides in primates. J Proteome Res 10: Varki A (2009) Essentials of glycobiology. Cold Spring Harbor, N.Y.: Cold Spring Harbor Laboratory Press. xxix, 784 p. p. 14. Foley S, Stolarczyk E, Mouni F, Brassart C, Vidal O, et al. (2008) Characterisation of glutamine fructose-6-phosphate amidotransferase (EC ) and N- acetylglucosamine metabolism in Bifidobacterium. Arch Microbiol 189: Ghosh S, Rao KH, Sengupta M, Bhattacharya SK, Datta A (2011) Two gene clusters co-ordinate for a functional N-acetylglucosamine catabolic pathway in Vibrio cholerae. Mol Microbiol 80: Favier CF, de Vos WM, Akkermans AD (2003) Development of bacterial and bifidobacterial communities in feces of newborn babies. Anaerobe 9: Roger LC, McCartney AL (2010) Longitudinal investigation of the faecal microbiota of healthy full-term infants using fluorescence in situ hybridization and denaturing gradient gel electrophoresis. Microbiology 156: !

205 195! 18. LoCascio RG, Ninonuevo MR, Freeman SL, Sela DA, Grimm R, et al. (2007) Glycoprofiling of bifidobacterial consumption of human milk oligosaccharides demonstrates strain specific, preferential consumption of small chain glycans secreted in early human lactation. J Agric Food Chem 55: Sela DA, Chapman J, Adeuya A, Kim JH, Chen F, et al. (2008) The genome sequence of Bifidobacterium longum subsp. infantis reveals adaptations for milk utilization within the infant microbiome. Proc Natl Acad Sci U S A 105: Sela DA, Mills DA (2010) Nursing our microbiota: molecular linkages between bifidobacteria and milk oligosaccharides. Trends Microbiol 18: Garrido D, Kim JH, German JB, Raybould HE, Mills DA (2011) Oligosaccharide binding proteins from Bifidobacterium longum subsp. infantis reveal a preference for host glycans. PLoS One 6: e Asakuma S, Hatakeyama E, Urashima T, Yoshida E, Katayama T, et al. (2011) Physiology of consumption of human milk oligosaccharides by infant gutassociated bifidobacteria. J Biol Chem 286: Sela DA, Garrido D, Lerno L, Wu S, Tan K, et al. (2011) Bifidobacterium longum subsp. infantis ATCC15697 alpha-fucosidases are active on fucosylated human milk oligosaccharides. Appl Environ Microbiol. 24. Sela DA, Li Y, Lerno L, Wu S, Marcobal AM, et al. (2011) An infant-associated bacterial commensal utilizes breast milk sialyloligosaccharides. J Biol Chem 286: Yoshida E, Sakurama H, Kiyohara M, Nakajima M, Kitaoka M, et al. (2011) Bifidobacterium longum subsp. infantis uses two different {beta}-galactosidases for selectively degrading type-1 and type-2 human milk oligosaccharides. Glycobiology. 26. Zhang G, Mills DA, Block DE (2009) Development of chemically defined media supporting high-cell-density growth of lactococci, enterococci, and streptococci. Appl Environ Microbiol 75: Ward RE, Ninonuevo M, Mills DA, Lebrilla CB, German JB (2006) In vitro fermentation of breast milk oligosaccharides by Bifidobacterium infantis and Lactobacillus gasseri. Appl Environ Microbiol 72: Markowitz VM, Korzeniewski F, Palaniappan K, Szeto E, Werner G, et al. (2006) The integrated microbial genomes (IMG) system. Nucleic Acids Res 34: D McIlvaine TC (1921) A buffer solution for colorimetric comparaison. Journal of Biological Chemistry 49: LoCascio RG, Desai P, Sela DA, Weimer B, Mills DA (2010) Broad conservation of milk utilization genes in Bifidobacterium longum subsp. infantis as revealed by comparative genomic hybridization. Appl Environ Microbiol 76: Miwa M, Horimoto T, Kiyohara M, Katayama T, Kitaoka M, et al. (2010) Cooperation of beta-galactosidase and beta-n-acetylhexosaminidase from bifidobacteria in assimilation of human milk oligosaccharides with type 2 structure. Glycobiology 20: !

206 196! 32. Kitaoka M, Tian J, Nishimoto M (2005) Novel putative galactose operon involving lacto-n-biose phosphorylase in Bifidobacterium longum. Appl Environ Microbiol 71: Nishimoto M, Kitaoka M (2007) Identification of N-acetylhexosamine 1-kinase in the complete lacto-n-biose I/galacto-N-biose metabolic pathway in Bifidobacterium longum. Appl Environ Microbiol 73: Vogler AP, Lengeler JW (1989) Analysis of the nag regulon from Escherichia coli K12 and Klebsiella pneumoniae and of its regulation. Mol Gen Genet 219: Vincent F, Davies GJ, Brannigan JA (2005) Structure and kinetics of a monomeric glucosamine 6-phosphate deaminase: missing link of the NagB superfamily? J Biol Chem 280: !!!!!!!!!!!!!!!!!!

207 197! SUPPLEMENTARY INFORMATION Table S1: Primers used in this study 1. Gene cloning of Blon_0459, Blon_0732 and Blon_2355. Primer Sequence (5 3 ) 0459FT CACCATGAGCGATCAAGCAACCCTG 0459RT GGATTGGAAGCGCAGATGGTC 0732FT CACCGTGCCCACTTCCGAACATAAGGCC 0732RT CAGAGCGCCCCTACGCAGAATGTC 2355FT CACCATGGTGCAGGAACCAACATTGGAA 2355RT AATGGAACGCAGCACGTCCGCCGC 2. qpcr of B. infantis gene expression (gene ID # indicated in primer name) Primer Sequence (5 3 ) 0177qF TCCGGTCGGCATTCACGCAC 0177qR GGCAACGGTCTCGGCGTTGT 0178qF TGGTCTGCGCACGCTGAAGG 0178qR GGCACCTCGGCCATCACACC 0393qF TTCACCGAGGCGTACAACA 0393qR CGCATCCGTGACCACATAG 0459qF GGCGACTTCGGCCACGTCAA 0459qR AGTTCACCGCTGGCGTCGTG 0732qF ACGCTGGACCGCACATTGGG 0732qR AACGCCAGCAGTTCCTCGCC 0881qF GGCCACGTCGGCTTCAACGA 0881qR GAACGCCAGCAGCACGAGGT 0882qF TCGTTTCCCGCGTGACCACG 0882qR CCACGTAGCCGGGGGTCAGA 2171qF CCACCAACCCGTACGGCACC 2171qR ACGGGGTGAGATTCGCCGGA 2172qF GCGGTGAGGAGGACCGTTGC 2172qR TCGTCGAGATCGGCGGGGTT 2173qF TGGACATCGTGCGCACCACC 2173qR AATCACCGAACGCCCTGCCG 2174qF TCGGCGAACCCGTGCTCAAC 2174qR CGTTCCAGCAGTCGGGCGTT 2355qF ACGCGCCGCGCAATAGGAAT 2355qR GGACGTGACTCGTGGCCGTG 2470qF CACGATGCTGGTGAGTGC 2470qR CCGGAACCGGTAAGATCC 2471qF ACAACCGTTTCAGCAAGACC 2471qR GAGCAGACGGTTGAAGAAGG 2401qF CGCATCCCGACGCCATTGGT 2401qR GTTCGGCGATAGCGCGTCCA 2402qF CGACTGCGGCTGCATCGAAGA 2402qR TGGCCGCATCGGACGGAGTA 2349qF GCGGCTCCCATCTCACGAGC 2349qR ATTGCGGCCGGGGTCAATGG!

208 198! SUPPLEMENTARY FIGURES Figure S1: Chitinase activity of B. infantis!-hexosaminidases. Lane 1: Chitobiose (GlcNAc!1-4GlcNAc); lane 2: GlcNAc; lanes 3-5: co-incubations of chitobiose with Blon_0459, Blon_0732 or Blon_2355.!

209 199! Figure S2: Growth of B. infantis in HMO, LNT and LNnT. Arrows indicate points were cells were recovered and gene expression was evaluated.!

210 200! Figure S3: Representation of gene clusters potentially involved in the metabolism of GlcNAc in B. infantis, including the GlcNAc-6-P deacetylase/isomerase cluster (A); LNB/GNB cluster (B), and a PTS specific for GlcNAc (C).!

211 201! Figure S4: Proposed metabolic pathways for the fate of GlcNAc in B. infantis, as obtained from the Kyoto Encyclopedia of Genes and Genomes (KEGG) and the gene expression data obtained in this study. A: Formation of UDP-GlcNAc from GlcNAc by the LNB/GNB cluster; B: predicted de novo synthesis of UDP-GlcNAc; C: sialic acid metabolic pathway, including genes Blon_0882 and Blon_0881.!

212 202! Figure S5: Relative quantification of the gene expression levels of genes predicted to be involved in metabolism of sialic acid, after bacterial growth on substrates shown in the x- axis. Gene information is provided in figure S4C. Values were normalized to lactose control (dashed line). Error bars represent three biological replicates.!

213 203! Figure S6: Phylogenetic relations between!-hexosaminidases found in B. infantis and other bifidobacteria. Asterisks indicate enzymes present in strain ATCC !!

214 204! Chapter VI Endo-!-N-acetylglucosaminidases from Infant-gut Associated Bifidobacteria Release Complex N-glycans from Human Milk Glycoproteins. Daniel Garrido 1,4, Charles Nwosu 2,4, Santiago Ruiz-Moyano 1,4, Danielle Aldredge 2,4, J. Bruce German 2,3,4, Carlito B. Lebrilla 2,4 and David A. Mills 1,4. Departments of Viticulture and Enology 1, Chemistry 2, Food Science and Technology 3 and Foods for Health Institute 4, University of California Davis. 1 Shields Ave, Davis CA Abbreviations: GlcNAc: N-acetylglucosamine; Man: mannose; Fuc: fucose; GH: glycosyl hydrolase; hlf: human lactoferrin; blf: bovine lactoferrin; IgA: immunoglobulin A; IgG: immunoglobulin G; RNAseB: ribonuclease B. Conceived and designed the experiments: DG, CW, DAM. Performed the experiments: DG, CW, SRM, DA. Analyzed the data: DG, CW, DA, CBL, DAM. Contributed reagents/materials/analysis tools: JBG, CBL. Wrote the paper: DG, DAM. Sections of this manuscript have been submitted for publication to Molecular and Cell Proteomics.

215 205! Summary Breastfeeding is one of the main factors guiding the composition of the infant gut microbiota in the first months of life. This process is shaped in part by the high amounts of human milk oligosaccharides that serve as a carbon source for saccharolytic bacteria such as Bifidobacterium species. Infant-borne bifidobacteria have developed various molecular strategies for utilizing these oligosaccharides as a carbon source. Here we hypothesized that these species can also interact with structurally similar N-glycans found in host glycoproteins as those found in human milk and the intestinal epithelium. Endo-!-N-acetylglucosaminidases were identified in certain isolates of Bifidobacterium longum subsp. longum, Bifidobacterium longum subsp. infantis and Bifidobacterium breve, and their presence correlated with the ability of these strains to deglycosylate bovine ribonuclease B. An endoglycosidase from B. infantis ATCC 15697, EndoBI-1, was shown to be active in complex and high mannose N-linked glycans. Its activity was not affected by core fucosylation, antenna number or terminal fucosylation or extensive sialylation, releasing several N-glycans from human lactoferrin and immunoglobulins A and G. Extensive N-deglycosylation of whole breast milk was also observed after coincubation with this enzyme. An active site mutant of EndoBI-1 bound to N- glycosylated proteins, and specifically recognized Man 3 GlcNAc 2 ("1-6Fuc), the core structure of human N-glycans. EndoBI-1 is constituvely expressed in B. infantis, and incubation of the bacterium with human or bovine lactoferrin led to the induction of genes associated to import and consumption of human milk oligosaccharides, suggesting linked regulatory mechanisms among these glycans. This work reveals an unprecedented interaction of bifidobacteria with host N-glycans, and describes the properties of EndoBI-

216 206! 1, a novel endoglycosidase active on diverse host glycoproteins with varying N- glycosylation types and decorations.

217 207! 1. Introduction Breast milk is an intriguing and complex fluid that supports the growth, development and protection of the newborn. Its composition includes essential nutrients such as lactose, fatty acids and proteins [1], as well as a myriad of bioactive compounds critical for the protection and correct development of the infant in the first months of life [2]. Breastfeeding is one of the main factors in the establishment of the intestinal microbiota in infants [3]. The presence of certain species of Bifidobacterium is commonly observed in breast-fed infants [4], and the dominance of these microorganisms is thought to be associated with beneficial health effects [5,6]. This enrichment has been explained by the ability of bifidobacteria to degrade and utilize human milk oligosaccharides (HMO) as a carbon source [7,8,9]. Proteins represent an important fraction of breast milk. A great variability exists among different proteins types and concentrations across different mothers and stages of lactation [10]. Milk proteins are readily utilized by the infant [11], and are also critical in the protection of the newborn. For example, human lactoferrin (hlf) is one of the most abundant proteins in human milk, and hlf or its derived peptides display broad antimicrobial and anti-inflammatory effects, among several other biological activities [12]. Virtually all secreted proteins in eukaryotes, including those in human milk, are glycosylated [13]. While milk caseins are O-linked glycosylated, lactoferrin and immunoglobulins contain N-linked glycans [13,14]. Asparagine-linked glycosylation is the most common post-translational modification of eukaryotic proteins [15]. In general, the role of N-linked glycosylation in folding, secretion and resistance to proteolysis is

218 208! understood for several proteins [16,17], and several examples have exemplified the crucial role of N-glycans in protein function, such as bacterial recognition [18], intracellular signaling [19] and antigen binding and presentation [20]. Interestingly, certain microorganisms, mostly pathogens, have acquired the ability to release N-glycans from glycoproteins. This trait is associated with the use of these oligosaccharides as a carbon source [21], or altering the biological function of certain glycoproteins such as immunoglobulins [22]. Bacterial Endo-!-N-acetylglucosaminidases (EC ; endoglycosidases) are widespread enzymes that cleave the N-N -diacetyl chitobiose moiety characteristic of the pentasaccharide Man 3 GlcNAc 2 found in all N- glycans [23]. These enzymes belong to glycosyl hydrolase families GH18 or GH85. Prominent examples are EndoH from Streptomyces plicatus [24], EndoE from Enterococcus faecalis [25] and EndoS from Streptococcus pyogenes [26], while EndoD from Streptococcus pneumoniae [27] is a member of GH85. Their substrate specificities are usually limited to either high mannose or complex N-glycans and some require additional exoglycosidases for glycan release. The niche that infant-gut associated bifidobacteria colonize is characterized by high amounts of milk oligosaccharides, as well as proteins or peptides arriving from breast milk or from the developing infant gut. While some bifidobacteria apparently can use mucin O-linked oligosaccharides as a carbon source [28,29], if these microorganisms can interact with N-glycosylated proteins has yet to be addressed. In this work we explored the ability of gut isolates of bifidobacteria to release N-glycans from host glycoproteins, and we also describe some of the properties of an endoglycosidase that released high mannose and complex N-glycans from host glycoproteins.

219 209! 2. Materials and Methods. Bacteria and media. Bifidobacterium strains used in this study (Table S1) were obtained from the Japanese Collection of Microorganisms (Riken Biosource Center Japan), the American Type Culture Collection (Manassas, VA), and the University of California Davis Viticulture and Enology Culture Collection (Davis, CA). For routine experiments, bifidobacteria were grown on de Mann-Rogose-Sharp (MRS) broth supplemented with 0.05 % w/v L-cysteine (Sigma-Aldrich, St. Louis, MO). Chemically defined Zhang- Mills-Block-1 (ZMB-1) media [30] was used for evaluation of bacterial growth on glycoproteins or transcriptional analyses. Cells were anaerobically grown in a vinyl chamber (Coy Laboratory Products, Grass Lake, MI) at 37 C for 24 h, in an atmosphere consisting of 5% carbon dioxide, 5% hydrogen, and 90% nitrogen. Chemicals. Cyanogen bromide (CNBr) activated sepharose 4B (S4B) beads, ribonuclease B from bovine pancreas (RNAseB), immunoglobulin G from human serum [31], immunoglobulin A from human colostrum (IgA), lactoferrin from human milk (hlf), lactoferrin from bovine milk (blf) and 2,5- dihydroxylbenzoic acid (DHB) were all obtained from Sigma Aldrich (St. Louis, MO). Graphitized carbon cartridges were purchased from Grace Davison Discovery Sciences (Deerfield, IL). All chemicals used were either of analytical grade or better. Claristar yeast mannoprotein was a gift from DSM Food Specialties (Parsippany, NJ). Incubations and growth of bifidobacteria on glycoproteins. Bifidobacterial isolates were grown on 2 ml of MRS with no carbon source (mmrs), supplemented with 2% lactose to mid-late exponential phase. Two hundred µl of culture were centrifuged for 1 min at x g, and resuspended in 200 µl of mmrs supplemented with 5 mg/ml of

220 210! RNAseB. Incubations were run for 18 hours, and supernatants were recovered after centrifugation 1 min at x g. A 1:10 dilution of each supernatant was denatured in glycoprotein denaturing buffer (0.5% SDS and 40 mm DTT) and analyzed on 4-15% precast SDS-PAGE gels (Bio-Rad, Carlsbad CA). Growth of specific bacteria was also analyzed on 96 well plates containing 200 µl of ZMB-1 media supplemented with 10 mg/ml of hlf or blf, or 50 mg/ml of Claristar yeast mannoprotein. Cultures were inoculated at 2% and grown for 72 h in a PowerWave microplate reader (BioTek Instruments, Inc., Winoosky, VT), under anaerobic conditions at 37 C. Growth was monitored using the Gen software (BioTek). Cultures were grown in triplicate, and controls with no glycoprotein and no bacteria were run and subtracted from OD 600 values. Endoglycosidase sequence determination. Protein coding sequences belonging to GH18 found in the genomes in B. infantis ATCC (Blon_2468), B. infantis 157F (BLIF_1310) and Enterococcus faecalis OG1RF (EndoE!) were aligned using MUSCLE. Conserved regions were selected and converted to DNA to design degenerate primers (Table S2). A similar approach was used with sequences encoding GH85 enzymes, found in the published genome sequences of B. longum DJO10A (BLD_0197), B. longum NCC2703 (BL1335) and B. breve UCC2003. Genomic DNA was prepared from overnight cultures on MRS for each strain used in this study using the DNeasy Blood & Tissue Kit (Qiagen, Valencia CA). Fifty µl PCR reactions contained 1 U of Phusion DNA polymerase (Finnzymes, Vantaa, Finland), 1 ng of DNA, 0.2 mm of dntps and 2.5 µm of each degenerate primer (Table S2), and were run in a PTC200 Thermo Cycler (MJ Research, Ramsey, MN). The PCR program included an initial denaturation at 98 C for 30 s, 30 cycles of denaturation at 98 C 10 s, annealing at 55 C for 30 s, extension at 72

221 211! C 1 min, and a final extension at 72 C for 7 min. PCR products were purified using the Qiaquick PCR product purification kit (Qiagen), and sequenced at the UC Davis DNA sequencing facility. Sequences encoding GH18 enzymes were analyzed using BioEdit 7.1.3, and later expanded and fully determined using the DNA Walking SpeedUp Premix Kit (Seegene, Rockville MD), and the TSP142 primers listed in Table S2. GH85- encoding gene sequences were directly determined using primers GH85cF and GH85cR. Bioinformatic analyses. The Integrated Microbial Genomes [32] database was used to find GH18 and GH85 protein sequences in Bifidobacterium genomes and to determine genetic landscapes for GH18-type and GH85-type genes found in the genomes of B. infantis ATCC 15697, B. infantis 157F and B. longum DJO10A. Multiple sequence alignments were performed using MUSCLE, using the Maximum Likelihood algorithm in MEGA v 5.0. Gene cloning and expression. Genomic DNA from B. infantis ATCC 15697, B. infantis SC142 and B. longum DJO10A was amplified with the cloning primers indicated in Table S2, targeting GH18 or GH85 sequences. Signal peptides and transmembrane domains were not amplified to facilitate protein expression and purification from E. coli. Gene amplification by PCR, cloning, protein expression and purification were performed as [33]. Induction was performed with 0.5 mm IPTG at 28 C (EndoBI-1, EndoBI-2 and EndoBI-1mut), or with 1 mm IPTG at 37 C (EndoBB). Proteins were concentrated using Amicon Ultra 30 kda 4 ml columns, and buffer was exchanged for saline sodium citrate 1X using Bio-Gel P-30 in SSC Buffer columns (Bio-Rad). Glycoprotein digestion by bifidobacterial endoglycosidases. Optimal enzymatic conditions for endoglycosidases EndoBI-1, EndoBI-2 and EndoBB were determined by

222 212! incubation with RNAseB. Reactions were performed in a 10 µl volume and included 4 µg of RNAseB, 1 µg of each enzyme and 4 µl of 0.2 M Na 2 HPO 4 with ph values between 5.0 and 7.0 at 37 C. Reactions were run for 1 h, stopped with 1 M Na 2 CO 3, treated with the denaturing buffer described above and loaded into 4-15% precast polyacrylamide SDS gels. Optimal temperature for each reaction was determined at each optimal ph, and reactions were performed at 4, 30, 37, 45, 55 and 65 C for 1 h. Heat resistance was evaluated by incubating each glycosidase at 95 C for 1, 5 and 30 min, and enzyme reactions were then carried out under optimal conditions. Digestions of hlf and blf (Sigma) were performed under optimal conditions using 4 µg of each glycoprotein and incubated for 18 h with 1 µg of each endoglycosidase, or 1 µl of glycerol-free peptide:nglycosidase F (PNGaseF 500U/µl; New England Biolabs, Ipswich, MA). Finally 20 µl of a fresh breast milk sample (kindly provided by a lactation study directed by Jennifer Smilowitz at UC Davis) was incubated for 18 h at 37 C with 10 µg of EndoBI-1, 10 µg of EndoBI-1 D184N or 1 µl of PNGaseF in 20 mm Na 2 HPO 4 ph 5.0. Lactoferrin and human milk digestions were evaluated in 7.5% precast SDS-PAGE gels under denaturing conditions. All experiments were performed at least in duplicates. EndoBI-1 immobilization to sepharose beads. To eliminate unwanted interferences and contamination from the enzyme solution or glycoprotein, EndoBI-1 was immobilized to sepharose beads activated with CNBr. This also allowed for multiple usage of the enzyme on different samples over a few weeks. CNBr-activated sepharose beads of µm diameter were covalently coupled to EndoBI-1 via the well-established coupling chemistry [34]. The actual immobilization of EndoBI-1 to the sepharose beads was achieved using a slightly modified version of the protocol reported earlier [35,36]. In this

223 213! study 150 mg of the lyophilized S4B beads were coupled to 300!g of EndoBI-1 prior to the glycoprotein digestion. Glycan release by EndoBI-1. Model glycoproteins used in this study were RNAseB, blf, hlf, IgA and IgG. Each glycoprotein was individually digested with beadimmobilized EndoBI-1 while thoroughly rinsing the beads after each digestion to eliminate cross-contamination. Glycoproteins were prepared in 0.2 M Na 2 HPO 4 ph 5.0 at 1 mg/ml in a final volume of 300!L in 1.5 ml tubes. The digestion mixture including the beads was incubated at 37 C overnight with gentle agitation. The resultant digest mixture was then carefully drawn out following centrifugation. Purification of the resultant glycans was then achieved via solid phase extraction [37] using C18 and graphitized carbon cartridges as earlier described by our group [38]. In this study, a clean mixture of the resultant glycans was then eluted with 9 ml of 0.05% trifluoroacetic acid in 40% acetonitrile (ACN) in water (v/v) followed by vacuum drying using a speed vac prior to MS analysis. Instrumentation. Glycans purified by SPE were completely dried in a speed vac, reconstituted in 50!L of deionized water and were ready for MS analysis without further treatment. Glycan stock solutions (0.75!L) were individually spotted on a stainless steel MALDI target with each spot mixed with an equal volume of DHB matrix solution made up of 0.05 mg/ml DHB in 50% ACN: 50% water. The glycan-dhb spots were then allowed to dry prior to the actual MS analyses. In this study, an IonSpec HiRes MALDI FT-ICR mass spectrometer (Lake Forest, CA) equipped with an external ion source based on a third harmonic Nd:YAG laser (355 nm) and a 7.0 Tesla actively shielded superconducting magnet, served as the platform for all the experiments described herein.

224 214! Glycans were analyzed with the MALDI FT-ICR MS instrument in both the positive- and negative ion modes. Once obtained, the mass spectra were externally calibrated using a series of maltooligosaccharides as previously described [39]. Site directed mutagenesis. A plasmid containing the EndoBI-1 sequence (omitting signal sequence or transmembrane domains) was resynthesized with mutagenic primers AmpR and 2468mutF (Table S2) using the Change-IT multiple mutation site directed mutagenesis kit (USB Corporation, Santa Clara CA) and following manufacturer instructions. Mutated plasmids were cloned into Top10 competent cells (Invitrogen), and after verifying the proper mutation were transformed into BL21 competent cells. EndoBI- 1 D184N was purified as described in the previous section, with induction carried on with 0.5 mm IPTG at 28 C for 6 h. Glycan array analysis. Purified EndoBI-1 D184N (100 µg/ml, 200 µl), was analyzed for glycan binding by the Consortium for Functional Glycomics using the Mammalian Printed Array v5.0. Protocols are available at Detection was performed using an Anti-His-FITC antibody (Invitrogen). B. infantis gene expression. B. infantis cells were grown on ZMB-1 media with 2% lactose as describe above. Six ml of an exponential culture (OD ) were centrifuged for 1 min at x g, and immediately resuspended in 5 ml of prewarmed ZMB-1 supplemented with either human lactoferrin or bovine lactoferrin (5 mg/ml). Cultures were rapidly returned to anaerobic conditions, and 1 ml of each culture was taken anaerobically every hour. One ml of the original culture grown on lactose (t=0), and hourly time points of incubations with blf or hlf (t=1-3 h), were centrifuged at x g for 1 min, and the pellet was resuspended in 1 ml of RNAlater (Ambion,

225 215! Austin, TX). The experiment was done in duplicate. Cell suspensions were stored overnight at 4 C and then at -80 C until use. RNA extraction, quality check and cdna conversion were performed as previously described [40]. Relative quantification for genes listed in table S2 was performed in a 7500 Fast Real-Time PCR System (Applied Biosystems, Foster City, CA) and using the Fast Sybr Green Master Mix (Applied Biosystems). Reaction conditions were as recommended by the manufacturer using 0.5 µm of each primer. Primers for qpcr were designed using the NCBI primer design tool, checking for specificity across the B. infantis ATCC genome (Table S2). Fluorescence assays. Binding of EndoBI-1 D184N to different glycoproteins was determined after overnight coating in microtiter 96 well plates of 20 µmoles of RNAseB, blf, hlf or BSA in PBS buffer at room temperature. The experiment was performed in triplicates. Wells were washed with PBS three times, and blocked after incubation with BSA 3% at RT for 1h. Twenty µmol of EndoBI-1 D184N or BSA were added to the wells and incubated for 2 h at 37 C in PBS buffer adjusted to ph 5.0. Wells were washed three times with PBS-Tween %, and incubated for 1 h with a 1:500 dilution of FITC-Anti-His (C-term) antibody (Invitrogen). After 4 washes with PBS- Tween, fluorescence was monitored in a Synergy2 Microplate reader (Biotek), at 485/530 nm emission/excitation. In another set of experiments, breast milk samples incubated overnight with EndoBI-1, EndoBI-1 D184N or PNGaseF as performed above were coated overnight in a microtiter 96-well plate at room temperature. After washing three times with PBS buffer, wells were incubated with a 1:500 dilution of 5 mg/ml of fluorescein labeled Concavalin A (Vector labs, Burlingame CA) for 1 h at 37 C. Wells were washed four times with PBS-Tween %, and fluorescence was read as

226 216! described above. This experiment was repeated twice. Statistical analysis of the data was carried out by one-way analysis of variance, and the means were separated by Tukey s honest significant differences test using the SPSS software package version (SPSS Inc., Chicago, Illinois, USA).

227 217! 3. Results 3.1 Infant gut isolates of bifidobacteria display endo-n-acetylglucosaminidase activity. Bovine ribonuclease B (RNAseB) is a 17 kda model glycoprotein that contains one glycosylation site, composed of high mannose N-linked glycans. Cleavage by endoglycosidases results in a molecule of 14 kda. Overnight incubations of a panel of 76 bifidobacterial isolates (Table S1) with RNAseB suggested that endoglycosidase activity is present in only certain isolates. Incubation of B. infantis ATCC with 5 mg/ml of RNAseB led to a gradual deglycosylation of this glycoprotein over time (Figure 1A). In general other B. infantis strains degraded RNAseB weakly (Figure 1C). Only a few B. longum subsp. longum (B. longum) isolates and none of the B. bifidum strains examined displayed this phenotype (Figure 1B). Interestingly, certain isolates of B. breve such as KA179 and JCM7019, completely deglycosylated RNAseB (Figure 1D). 3.2 Distribution of endo-n-acetylglucosaminidase gene sequences in bifidobacteria. Based on known endo-n-acetylglucosaminidase sequences found in bifidobacteria, degenerate primers were used to search for GH18 or GH85 sequences in the strains used in this study. Certain isolates were found to contain a gene encoding either a GH18 or GH85 enzyme. All strains containing one of these sequences cleaved RNAseB in vitro, and conversely strains lacking these genes did not show endoglycosidase activity (Table S1). This suggested a correlation between the presence of a GH18 or GH85 enzyme and the observed RNAseB deglycosylation phenotype.

228 218! Figure 1: Endoglycosidase activity found in Bifidobacterium isolates. A: Deglycosylation of RNAseB by B. infantis ATCC over time. Overnight incubation with RNAseB was evaluated for other isolates of B. longum (B), B. infantis (C) or B. breve (D). E: Phylogenetic representation of endoglycosidase sequences found in bifidobacterial isolates.

229 219! A phylogenetic tree (Figure 1E) classified these protein sequences in three types. One group was exclusively found in B. infantis strains including the type strain ATCC (termed GH18a), and distantly related to EndoE. Another set of sequences found in strains of B. infantis, B. breve and B. longum shared 60% aminoacid identity with GH18a, and it was termed GH18b. Sequences belonging to GH85 were almost exclusively found in fecal B. breve isolates. Multiple alignments revealed a high degree of conservation of the proposed active site for each glycosidase family (Figure S1; [25,41]). The genomic landscape for these genes also supported their association with carbohydrate metabolism. Blon_2468 in B. infantis ATCC is located in a gene cluster that in addition contains a phosphotransferase (PTS) system specific for N- acetylglucosamine (Figure S2). BLIF_1310 in B. infantis 157F (GH18b), and BLD_0197 in B. longum DJO10A (GH85) are located near ABC transporters predicted to import oligosaccharides and two or three putative!-mannosidases (Figure S2). 3.3 Enzymatic properties of bifidobacterial endo-n-acetylglucosaminidases. Based on the sequence alignments obtained (Figure 1E), a representative gene of each glycosyl hydrolase type was cloned, expressed and purified in E. coli. The recombinant endo-"-nacetylglucosaminidases from B. infantis ATCC (EndoBI-1; Blon_2468), B. infantis SC142 (EndoBI-2), and B. longum DJO10A (EndoBB; BLD_0197), all exhibited maximum glycolytic activity at ph 5.0 and their optimal temperatures ranged from 37 C to 45 C (Figure S3). Interestingly, the enzymatic activity of EndoBI-1 and EndoBI-2 was not significantly impaired by incubation at 95 C for 1 or 5 minutes, implying that

230 220! they are heat resistant enzymes (Figure 2A). This property was not observed for EndoBB. Other properties of these enzymes are listed in Table 1. Human lactoferrin (hlf) contains core fucosylated complex N-glycans, predominantly at two glycosylation sites [42]. Bovine lactoferrin (blf) represents a minor fraction of bovine milk, and it contains mainly high mannose N-linked glycans at five glycosylation sites [36]. Overnight incubations of blf and hlf with the three recombinant Bifidobacterium endoglycosidases indicated that EndoBI-1 and EndoBI-2 were able to cleave blf and also hlf, as observed by discrete changes in MW on SDS-PAGE gels (Figures 2B and 2C). EndoBB did not display glycolytic activity against blf or hlf (Figure 2C).

231 221! Figure 2: Properties of the recombinant endoglycosidases in bifidobacteria. A: Heat tolerance of EndoBI-1, EndoBI-2 or EndoBB evaluated in SDS-PAGE gels, as evaluated by RNAseB deglycosylation with each enzyme incubated at 95 C for the times indicated. B: Coincubations of blf and hlf with EndoBI-1 (1), EndoBI-2 (2), EndoBB (3) or PNGaseF (4). Control (C) non-digested reactions were included in both experiments.

232 222! Table 1: General properties of the endo-n-acetylglucosaminidases described in this study EndoBI-1 EndoBI-2 EndoBB Family GH18 GH18 GH85 Calculated MW 47 kda 47 kda 98 kda (recombinant protein) Transmembrane domains Optimum ph Optimum temperature C C C Heat resistance Yes Yes No

233 223! 3.4 EndoBI-1 cleaves the chitobiose core of high-mannose and complex N-glycans. In order to better determine the enzymatic properties of EndoBI-1, the enzyme was immobilized using sepharose beads and incubated with several glycoproteins with varying glycosylation types. Glycans released were analyzed using MALDI FT-ICR mass spectrometry. Similarly to related endoglycosidases, EndoBI-1 acted on the chitobiose core of N-glycans, probably leaving a GlcNAc residue (and an!1-6 fucose in core fucosylated glycans) attached to the glycosylated asparagine residue. Activity on blf was detected and as expected predominantly oligomannose N-glycans were observed, with a minor amount of complex/hybrid glycans (Figure 3A; [43]). N-glycans released from RNAseB contained between 5 and 9 mannose residues (Figure S4). EndoBI-1 was also shown to deglycosylate host glycoproteins such as hlf, IgA and IgG (Figures 3B, 3C and Figure S4). These proteins are characterized by complex core fucosylated glycans, and are essentially resistant to several commercial endoglycosidases under native conditions. Glycans released from hlf and IgA were bi and triantennary, and contained up to two sialic acid residues and up to three fucoses attached to the lactosamine chains. The profile of the glycans released by EndoBI-1 was similar to PNGaseF [42,44,45]. IgG deglycosylation by EndoBI-1 revealed biantennary complex N-glycans with lesser fucosylation and sialylation compared to IgA, and similar to those observed after PNGaseF cleavage (Figure S4, [46,47]).

234 224! Figure 3: MALDI-FT-ICR MS analysis of glycans released after coincubation of EndoBI-1 with A: blf; B: hlf (positive mode); C: IgA (positive mode). Green circles: mannose; blue squares: GlcNAc; yellow circles: galactose; red triangles: fucose; pink diamonds: sialic acid.

235 225! 3.5 A mutant of EndoBI-1 binds specifically the core of N-linked glycans. The conserved active site of GH18 enzymes includes the motif D-X-E where both Asn and Glu are crucial for activity [48]. Asp184 in EndoBI-1 was mutated by site-directed mutagenesis to Asn184 (EndoBI-1 D184N). The mutated enzyme specifically bound to the core of N-glycans, Man 3 GlcNAc 2, across 600 glycans in a mammalian glycan array (Figure 4A). Interestingly, EndoBI-1 D184N also showed significant binding to the!1-6 fucosylated pentasaccharide, characteristic of human N-linked glycoproteins. When equimolar amounts of RNAseB, blf and hlf were coated to microwell plates, EndoBI-1 D184N showed a significant binding to these proteins compared to a non-glycosylated control (Figure 4B). 3.6 EndoBI-1 has glycosidase activity on human milk glycoproteins. Breast milk is a complex fluid, characterized by diverse amounts of N-linked, O-linked and nonglycosylated proteins. Overnight incubation of a fresh human milk sample with EndoBI-1 or PNGaseF produced a shift in the molecular weight of mainly one protein, probably lactoferrin as deduced from its MW (Figure 5A). No change was observed when a breast milk sample was incubated with EndoBI-1 D184N. In a parallel experiment, the total amount of N-linked glycans, estimated as the amount of!-mannose detected by the lectin Concavalin A conjugated to FITC (ConA-FITC), was determined in digested milk samples. EndoBI-1 and PNGaseF, but not EndoBI D184N, significantly decreased the amount of!-mannose in breast milk (Figure 5B), suggesting an extensive removal of N- linked glycans.

236 226! Figure 4: Properties of EndoBI-1 D184N. A: Glycan array analysis of the mutant enzyme binding to mammalian glycans (x-axis). Bars represent SD of sextuplicates. Green circles: mannose; blue squares: GlcNAc. B: Binding of EndoBI-1 or EndoBI-D184N to coated glycoproteins, as detected by a FITC-Anti His antibody. Error bars represent SD from triplicate experiments. Asterisks represent samples with p < 0.05 compared to BSA.

237 227! 3.7 Growth of bifidobacteria on N-glycosylated proteins. The ability of selected bifidobacteria to use hlf or blf as the sole carbon source was tested in vitro. Under the conditions assayed, apparently bifidobacteria did not use these substrates and no growth was appreciated. A more vigorous and significant growth was appreciated when 5% of yeast mannoprotein, heavily N-glycosylated cell wall proteins purified from Saccharomyces cerevisiae, were used as the sole carbon source (Figure S5). A higher OD 600 was obtained under these conditions for B. breve SC139 and B. breve KA179, strains that showed high endoglycosidase activity against RNAseB (Figure 1).

238 228! Figure 5: EndoBI-1 activity in breast milk. A: SDS-PAGE gel of overnight incubation of human milk (lane 1, control) with EndoBI-1 (lane 2), EndoBI-1 D184N (lane 3) or PNGaseF (lane 4). Protein identities were deduced from [13]. B: Amount of N-glycosylation (proportional to!- mannose) in samples from A. Error bars represent SD from triplicate experiments. Asterisks represent samples with p < 0.05 compared to control.

239 229! 3.8 Impact of hlf and blf on B. infantis gene expression. The molecular response of B. infantis ATCC to blf and hlf was later tested coincubating the microorganism with 5 mg/ml of blf and hlf in a resting cell assay. This revealed an increased expression of Blon_2468 (EndoBI-1) by comparison to cells grown on glucose, and the level of expression of this gene was similar to that from cells grown on lactose (Figure S6). Coincubations with blf or hlf resulted in higher expression of other genes adjacent to Blon_2468 including Blon_2470 and Blon_2471, encoding part of a PTS system specific for GlcNAc, (Figure 6A). A similar trend was observed for Blon_0177 and Blon_0178, genes also associated to PTS systems in B. infantis. Other genes induced by these glycoproteins were Blon_0881 and to a lesser extent Blon_0882, key enzymes that participate in metabolism of GlcNAc and sialic acid. Putative genes in B. infantis associated to mannose metabolism (Blon_2380, solute binding protein for mannooligosaccharides, and Blon_0868 and Blon_0869,!-mannosidases) were not affected by the presence of blf or hlf. Conversely, several genes associated to the import and consumption of human milk oligosaccharides in B. infantis were significantly induced by hlf, and to a lesser extent blf (Figure 6B). In general the highest induction was observed after 1 hour of incubation. These genes included Blon_2344, Blon_2347, Blon_0883 and Blon_2177, solute-binding proteins that bind different classes of HMO associated to ABC transporters, as well as Blon_2335 and Blon_2336, two key fucosidases in the B. infantis genome [33,40].

240 230! Figure 6: Fold changes in gene expression for B. infantis ATCC genes during time coincubation with blf or hlf, as indicated in the figure legend. Locus tags are described in the text. Error bars represent SD from three biological replicates. A: Genes associated to GlcNAc metabolism and located close to EndoBI-1; B: Genes previously described to be associated or induced by HMO.

241 231! Discussion Bifidobacteria are common members of the infant and adult gut microbiota. Their presence in this environment is generally regarded as beneficial to the host, however the mechanisms involved are far from understood. Human milk provides the newborn a range of bioactive glycans either as free oligosaccharides or as conjugates bound to proteins or lipids. The potential prebiotic role of free HMOs to enrich infant-borne bifidobacteria has been recently defined at the molecular level [49,50]. However, little work has examined the prebiotic impact of N-linked or O-linked human milk glycoproteins. In bifidobacteria, endoglycosidase activity on N-linked glycoproteins was tested across a large panel of bifidobacteria isolated from infant feces and found only in certain fecal isolates of B. infantis (40%), B. longum (21%) and B. breve (36%), with the latter species showing the highest glycolytic activity on RNAseB. Genes encoding endo-nacetylglucosaminidases belonging to GH18a, GH18b or GH85 were found only in isolates with endoglycosidase activity, suggesting that the presence of these genes explain the observed phenotype. In this study we determined that Bifidobacterium strains do not extensively use host glycoproteins as the sole carbon source compared to certain pathogens under the conditions assayed. However, several strains possessing an endoglycosidase showed a modest growth on pure yeast mannoproteins (Figure S5), which consist of high-mannose heavily N-glycosylated cell wall proteins from yeast. This suggested that bifidobacteria can utilize N-glycans as the sole carbon source under certain conditions, and also indicated that this substrate could be used as a selective agent for bifidobacteria displaying endoglycosidase activity. The ability to release N-glycans from

242 232! host proteins has been mainly associated to GH18 enzymes found in bacterial pathogens such as E. faecalis [51], S. pyogenes [52] and Capnocytophaga canimorsus [21]. These bacteria can extensively grow on different glycoproteins as a carbon source, and as in the case of EndoS from S. pyogenes, IgG-specific deglycosylation severely impairs its recognition by immune effectors, increasing bacterial survival in blood [26]. It is possible that bifidobacterial endoglycosidases could also modulate of the activity of host glycoproteins. An increasing amount of evidence suggests a crucial role for N-glycans in the function of several host proteins [20,53,54]. For example, recognition of Gram-positive bacteria by IgA is dependent on its glycosylation [18], and intracellular signaling and NF-kB activation of the toll-like receptor 3 [19] is modulated by N-glycans. C-type lectins, galectins and sialic-acid-binding Ig-like lectins are immune and cell response mediators that specifically recognize different epitopes in N-glycans [23]. While lactoferrin N-linked glycosylation is variable during lactation [13], studies about the impact of glycosylation of this protein with respect to its resistance to proteolysis and iron binding [55,56,57,58] are contradictory. If certain Bifidobacterium isolates have the ability to remove N-glycans from lactoferrin, destabilization of the protein could favor to the production of antimicrobial peptides such as lactoferricin B or lactoferrampin. Interestingly, some studies have suggested that lactoferrin has a bifidogenic effect [59,60]. While in the present work specific enzymes in bifidobacteria have been determined to deglycosylate human lactoferrin, more studies are needed to address closer the impact of lactoferrin or its derived peptides on these microorganisms. GH18 and GH85 endoglycosidases specifically cleave the N-N - diacetylchitobiose core of N-linked glycans. Here we studied some of the enzymatic

243 233! properties of EndoBI-1 and EndoBI-2, representatives of two clades of GH18 sequences found in bifidobacteria (Figure 1E). While their aminoacid sequences were only 60% identical and possessed different gene contexts (Figure S2), their active sites were conserved, and both acted on blf and hlf, containing oligomannose and complex N- glycans respectively. Further description of the N-glycans released from EndoBI-1 by mass spectrometry indicated that it was also active on IgA and IgG. The enzyme apparently did not recognize O-linked glycans or HMO (data not shown). In general the specificity of most known endoglycosidases is limited to high mannose glycans (for example EndoH [24]). EndoS from S. pyogenes acts solely on IgG [26], and the affinities of EndoE! from E. faecalis for other proteins than RNAseB have not been further studied [25]. Endoglycosidases F1, F2 and F3 from Elizabethkingia miricola show a preference for either high mannose or complex oligosaccharides [61]. Comparatively, several features suggested that EndoBI-1 might prove as a novel tool for diverse applications. Found in a beneficial microbe, heat inactivation for 95 C for 5 minutes did not severely impacted its activity. The glycolytic activity found in a varied range of glycoproteins showed that this enzyme cleaved the chitobiose core of high mannose N-glycans (RNAseB and blf), and core!1-6 fucosylated, bi or triantennary complex N-glycans with up to two sialic acid and up to three fucose residues decorating the lactosamine chains (Figure 3 and S4). More precise kinetic studies are required for determining the impact of these modifications on enzyme activity. These analyses were done using native glycoproteins, and a much greater deglycosylation rate after denaturation is expected. Finally, we evaluated the activity of EndoBI-1 in human milk, a complex matrix of lipids, oligosaccharides and proteins with disparate glycosylation types. The endoglycosidase

244 234! was successful in removing a significant proportion of the total amount of N-glycans from human milk (Figure 5). EndoBI-1 D184N has a mutation in the active site, and while lacking of catalytic activity, it was expected to retain its ability to bind N-glycan. Specific binding of the mutant to Man 3 GlcNAc 2 (!1-6Fuc) in a glycan array supported the affinity and specificity of the enzyme for N-glycans, especially those of host-origin, many of which are core fucosylated. We also partially characterized EndoBB (BLD_0197) from B. longum DJO10A, representative of GH85 sequences found in infant gut bifidobacteria. The activity of this enzyme apparently was much more limited, cleaving RNAseB but not blf or hlf. Bifidobacterial GH85 enzymes are distantly related to EndoD from S. pneumoniae [27]. EndoD acts on complex core fucosylated N-glycans, but only when lactosamine chains have been trimmed by exoglycosidases [27]. It is possible that B. breve GH85 endoglycosidases collaborates with additional glycosyl hydrolases. The presence of!- mannosidases and an ABC importer for oligosaccharides adjacent to these genes suggests that their function is related. It is also possible that these clusters are also active on mannose-based oligosaccharides from plant origin. Strains of B. infantis have been studied by their remarkable ability to use HMO as the only carbon source [62]. Genes induced by HMO in B. infantis, such as solute binding proteins and!-fucosidases [33,40], were also up-regulated by hlf and blf (Figure 6). These results suggested that, while not extensively using these glycoproteins as a carbon source (at least under the conditions tested in this study), B. infantis was still able to respond to these substrates in a similar fashion as to HMO.

245 235! In conclusion, in this work we described the interaction of infant-gut associated bifidobacteria with N-linked glycans found in host glycoproteins such as those found in breast milk, and we determined and characterized the discrete molecular determinants associated with this interaction. Finally, we have characterized the enzymatic properties of EndoBI-1 from B. infantis, which showed a remarkable activity on a wide range of host N-linked glycans.

246 236! Acknowledgments The authors would like to thank Dr. Jennifer Smilowitz and Lan Guan for providing reagents for this study, Dr. Alberto Martin Gonzalez for his support in statistical analysis and finally Dr. David F. Smith and Dr. Jamie Heimburg-Molinaro from the Consortium for Functional Glycomics Core H for their glycan array analysis. Daniel Garrido was sponsored in part through a Fulbright-Conicyt Chile scholarship, and Santiago Ruiz- Moyano was supported by the Ministry of Education and Science of Spain and University of Extremadura, Spain. This work was supported by the University of California Discovery Grant Program, the California Dairy Research Foundation and National Institute of Health Award R01HD

247 237! REFERENCES 1. Neville MC (2009) Classic studies of mammary development and milk secretion: J Mammary Gland Biol Neoplasia 14: German JB, Dillard CJ, Ward RE (2002) Bioactive components in milk. Curr Opin Clin Nutr Metab Care 5: Adlerberth I, Wold AE (2009) Establishment of the gut microbiota in Western infants. Acta Paediatr 98: Roger LC, McCartney AL (2010) Longitudinal investigation of the faecal microbiota of healthy full-term infants using fluorescence in situ hybridization and denaturing gradient gel electrophoresis. Microbiology 156: Le Huerou-Luron I, Blat S, Boudry G (2010) Breast- v. formula-feeding: impacts on the digestive tract and immediate and long-term health effects. Nutr Res Rev 23: Conroy ME, Shi HN, Walker WA (2009) The long-term health effects of neonatal microbial flora. Curr Opin Allergy Clin Immunol 9: Ward RE, Ninonuevo M, Mills DA, Lebrilla CB, German JB (2007) In vitro fermentability of human milk oligosaccharides by several strains of bifidobacteria. Mol Nutr Food Res 51: Kunz C, Rudloff S, Baier W, Klein N, Strobel S (2000) Oligosaccharides in human milk: structural, functional, and metabolic aspects. Annu Rev Nutr 20: Sela DA, Chapman J, Adeuya A, Kim JH, Chen F, et al. (2008) The genome sequence of Bifidobacterium longum subsp. infantis reveals adaptations for milk utilization within the infant microbiome. Proc Natl Acad Sci U S A 105: Mitoulas LR, Kent JC, Cox DB, Owens RA, Sherriff JL, et al. (2002) Variation in fat, lactose and protein in human milk over 24 h and throughout the first year of lactation. Br J Nutr 88: Prentice A, Ewing G, Roberts SB, Lucas A, MacCarthy A, et al. (1987) The nutritional role of breastmilk IgA and lactoferrin. Acta Paediatr Scand 76: Gonzalez-Chavez SA, Arevalo-Gallegos S, Rascon-Cruz Q (2009) Lactoferrin: structure, function and applications. Int J Antimicrob Agents 33: 301 e Froehlich JW, Dodds ED, Barboza M, McJimpsey EL, Seipert RR, et al. (2010) Glycoprotein expression in human milk during lactation. J Agric Food Chem 58: Picariello G, Ferranti P, Mamone G, Roepstorff P, Addeo F (2008) Identification of N-linked glycoproteins in human milk by hydrophilic interaction liquid chromatography and mass spectrometry. Proteomics 8: Apweiler R, Hermjakob H, Sharon N (1999) On the frequency of protein glycosylation, as deduced from analysis of the SWISS-PROT database. Biochim Biophys Acta 1473: Weber AN, Morse MA, Gay NJ (2004) Four N-linked glycosylation sites in human toll-like receptor 2 cooperate to direct efficient biosynthesis and secretion. J Biol Chem 279: Roth J, Zuber C, Park S, Jang I, Lee Y, et al. (2010) Protein N-glycosylation, protein folding, and protein quality control. Mol Cells 30: Mathias A, Corthesy B (2011) Recognition of gram-positive intestinal bacteria by hybridoma- and colostrum-derived secretory immunoglobulin A is mediated by carbohydrates. J Biol Chem 286: Sun J, Duffy KE, Ranjith-Kumar CT, Xiong J, Lamb RJ, et al. (2006) Structural and functional analyses of the human Toll-like receptor 3. Role of glycosylation. J Biol Chem 281: Ryan SO, Bonomo JA, Zhao F, Cobb BA (2011) MHCII glycosylation modulates Bacteroides fragilis carbohydrate antigen presentation. J Exp Med 208: Renzi F, Manfredi P, Mally M, Moes S, Jeno P, et al. (2011) The N-glycan glycoprotein deglycosylation complex (Gpd) from Capnocytophaga canimorsus deglycosylates human IgG. PLoS Pathog 7: e Collin M, Shannon O, Bjorck L (2008) IgG glycan hydrolysis by a bacterial enzyme as a therapy against autoimmune conditions. Proc Natl Acad Sci U S A 105: Varki A (2009) Essentials of glycobiology. Cold Spring Harbor, N.Y.: Cold Spring Harbor Laboratory Press. xxix, 784 p. p. 24. Trimble RB, Maley F (1977) The use of endo-beta-n-acetylglucosaminidase H in characterizing the structure and function of glycoproteins. Biochem Biophys Res Commun 78:

248 238! 25. Collin M, Fischetti VA (2004) A novel secreted endoglycosidase from Enterococcus faecalis with activity on human immunoglobulin G and ribonuclease B. J Biol Chem 279: Allhorn M, Olin AI, Nimmerjahn F, Collin M (2008) Human IgG/Fc gamma R interactions are modulated by streptococcal IgG glycan hydrolysis. PLoS One 3: e Muramatsu H, Tachikui H, Ushida H, Song X, Qiu Y, et al. (2001) Molecular cloning and expression of endo-beta-n-acetylglucosaminidase D, which acts on the core structure of complex type asparagine-linked oligosaccharides. J Biochem 129: Ruas-Madiedo P, Gueimonde M, Fernandez-Garcia M, de los Reyes-Gavilan CG, Margolles A (2008) Mucin degradation by Bifidobacterium strains isolated from the human intestinal microbiota. Appl Environ Microbiol 74: Fujita K, Oura F, Nagamine N, Katayama T, Hiratake J, et al. (2005) Identification and molecular cloning of a novel glycoside hydrolase family of core 1 type O-glycan-specific endo-alpha-nacetylgalactosaminidase from Bifidobacterium longum. J Biol Chem 280: Zhang G, Mills DA, Block DE (2009) Development of chemically defined media supporting high-celldensity growth of lactococci, enterococci, and streptococci. Appl Environ Microbiol 75: Higgins MA, Abbott DW, Boulanger MJ, Boraston AB (2009) Blood group antigen recognition by a solute-binding protein from a serotype 3 strain of Streptococcus pneumoniae. J Mol Biol 388: Markowitz VM, Korzeniewski F, Palaniappan K, Szeto E, Werner G, et al. (2006) The integrated microbial genomes (IMG) system. Nucleic Acids Res 34: D Sela DA, Garrido D, Lerno L, Wu S, Tan K, et al. (2012) Bifidobacterium longum subsp. infantis ATCC alpha-fucosidases Are Active on Fucosylated Human Milk Oligosaccharides. Appl Environ Microbiol 78: Deguchi K, Ito H, Baba T, Hirabayashi A, Nakagawa H, et al. (2007) Structural analysis of O- glycopeptides employing negative- and positive-ion multi-stage mass spectra obtained by collision-induced and electron-capture dissociations in linear ion trap time-of-flight mass spectrometry. Rapid Commun Mass Spectrom 21: Nwosu CC, Strum JS, An HJ, Lebrilla CB (2010) Enhanced detection and identification of glycopeptides in negative ion mode mass spectrometry. Anal Chem 82: Nwosu CC, Seipert RR, Strum JS, Hua SS, An HJ, et al. (2011) Simultaneous and extensive sitespecific N- and O-glycosylation analysis in protein mixtures. J Proteome Res 10: Rentz DM, Locascio JJ, Becker JA, Moran EK, Eng E, et al. (2010) Cognition, reserve, and amyloid deposition in normal aging. Ann Neurol 67: An HJ, Peavy TR, Hedrick JL, Lebrilla CB (2003) Determination of N-glycosylation sites and site heterogeneity in glycoproteins. Anal Chem 75: Clowers BH, Dodds ED, Seipert RR, Lebrilla CB (2008) Dual polarity accurate mass calibration for electrospray ionization and matrix-assisted laser desorption/ionization mass spectrometry using maltooligosaccharides. Anal Biochem 381: Garrido D, Kim JH, German JB, Raybould HE, Mills DA (2011) Oligosaccharide binding proteins from Bifidobacterium longum subsp. infantis reveal a preference for host glycans. PLoS One 6: e Yamamoto S, Muramatsu H, Muramatsu T (2005) Mutational studies on endo- -Nacetylglucosaminidase D which hydrolyzes core portion of asparagine-linked complex type oligosaccharides. Glycoconjugate Journal 22: Yu T, Guo C, Wang J, Hao P, Sui S, et al. (2011) Comprehensive characterization of the site-specific N-glycosylation of wild-type and recombinant human lactoferrin expressed in the milk of transgenic cloned cattle. Glycobiology 21: Hua S, Nwosu CC, Strum JS, Seipert RR, An HJ, et al. (2011) Site-specific protein glycosylation analysis with glycan isomer differentiation. Anal Bioanal Chem. 44. Barboza M, Pinzon J, Wickramasinghe S, Froehlich JW, Moeller I, et al. (2012) Glycosylation of human milk lactoferrin exhibits dynamic changes during early lactationenhancing its role in pathogenic bacteria-host interactions. Mol Cell Proteomics. 45. Pierce-Cretel A, Pamblanco M, Strecker G, Montreuil J, Spik G, et al. (1982) Primary structure of the N-glycosidically linked sialoglycans of secretory immunoglobulins A from human milk. Eur J Biochem 125:

249 239! 46. Raju TS, Briggs JB, Borge SM, Jones AJ (2000) Species-specific variation in glycosylation of IgG: evidence for the species-specific sialylation and branch-specific galactosylation and importance for engineering recombinant glycoprotein therapeutics. Glycobiology 10: Wuhrer M, Stam JC, van de Geijn FE, Koeleman CA, Verrips CT, et al. (2007) Glycosylation profiling of immunoglobulin G (IgG) subclasses from human serum. Proteomics 7: Synstad B, Gaseidnes S, Van Aalten DM, Vriend G, Nielsen JE, et al. (2004) Mutational and computational analysis of the role of conserved residues in the active site of a family 18 chitinase. Eur J Biochem 271: Zivkovic AM, German JB, Lebrilla CB, Mills DA (2011) Human milk glycobiome and its impact on the infant gastrointestinal microbiota. Proc Natl Acad Sci U S A 108 Suppl 1: Garrido D, Barile, D., Mills, D.A. (2012) A molecular basis for bifidobacterial enrichment in the infant gastrointestinal tract. Advances in Nutrition accepted. 51. Roberts G, Homer KA, Tarelli E, Philpott-Howard J, Devriese LA, et al. (2001) Distribution of endobeta-n-acetylglucosaminidase amongst enterococci. J Med Microbiol 50: Collin M, Olsen A (2001) EndoS, a novel secreted protein from Streptococcus pyogenes with endoglycosidase activity on human IgG. EMBO J 20: Almogren A, Bonner A, Perkins SJ, Kerr MA (2009) Functional and structural characterisation of human colostrum free secretory component. Mol Immunol 46: Vagin O, Tokhtaeva E, Yakubov I, Shevchenko E, Sachs G (2008) Inverse correlation between the extent of N-glycan branching and intercellular adhesion in epithelia. Contribution of the Na,K- ATPase beta1 subunit. J Biol Chem 283: Davidson LA, Lonnerdal B (1988) Specific binding of lactoferrin to brush-border membrane: ontogeny and effect of glycan chain. Am J Physiol 254: G van Berkel PH, Geerts ME, van Veen HA, Kooiman PM, Pieper FR, et al. (1995) Glycosylated and unglycosylated human lactoferrins both bind iron and show identical affinities towards human lysozyme and bacterial lipopolysaccharide, but differ in their susceptibilities towards tryptic proteolysis. Biochem J 312 ( Pt 1): van Veen HA, Geerts ME, van Berkel PH, Nuijens JH (2004) The role of N-linked glycosylation in the protection of human and bovine lactoferrin against tryptic proteolysis. Eur J Biochem 271: Brines RD, Brock JH (1983) The effect of trypsin and chymotrypsin on the in vitro antimicrobial and iron-binding properties of lactoferrin in human milk and bovine colostrum. Unusual resistance of human apolactoferrin to proteolytic digestion. Biochim Biophys Acta 759: Rahman MM, Kim WS, Ito T, Kumura H, Shimazaki K (2009) Growth promotion and cell binding ability of bovine lactoferrin to Bifidobacterium longum. Anaerobe 15: Petschow BW, Talbott RD, Batema RP (1999) Ability of lactoferrin to promote the growth of Bifidobacterium spp. in vitro is independent of receptor binding capacity and iron saturation level. J Med Microbiol 48: Trimble RB, Tarentino AL (1991) Identification of distinct endoglycosidase (endo) activities in Flavobacterium meningosepticum: endo F1, endo F2, and endo F3. Endo F1 and endo H hydrolyze only high mannose and hybrid glycans. J Biol Chem 266: Locascio RG, Ninonuevo MR, Kronewitter SR, Freeman SL, German JB, et al. (2009) A versatile and scalable strategy for glycoprofiling bifidobacterial consumption of human milk oligosaccharides. Microb Biotechnol 2: Roy D, Ward P, Champagne G (1996) Differentiation of bifidobacteria by use of pulsed-field gel electrophoresis and polymerase chain reaction. Int J Food Microbiol 29:

250 240! SUPPLEMENTARY INFORMATION Table S1: Bacterial strains used in this study. Code Identification Additional strain information Source GH gene type Endoglycosid ase activity ATCC B. longum subsp. infantis JCM1222; DSM20088 Intestine of infant GH18a Yes ATCC B. longum subsp. infantis JCM1210; DSM20223 Intestine of infant - No ATCC B. longum subsp. infantis JCM1260; DSM20218 Infant feces GH18a Yes ATCC B. longum subsp. infantis JCM1272; DSM20090 Intestine of infant GH18a Yes JCM 7007 B. longum subsp. infantis LMG18901 Infant feces GH18a Yes JCM 7009 B. longum subsp. infantis LMG18902 Infant feces GH18a Yes JCM 7011 B. longum subsp. infantis Infant feces GH18a Yes JCM B. longum subsp. infantis Isolates Infant feces GH18a Yes SC30 B. longum subsp. infantis Isolates Infant feces - No SC97 B. longum subsp. infantis Isolates Infant feces - No SC117 B. longum subsp. infantis Isolates Infant feces - No SC142 B. longum subsp. infantis Isolates Infant feces GH18b Yes SC143 B. longum subsp. infantis Isolates Infant feces GH18b Yes SC145 B. longum subsp. infantis Isolates Infant feces - No SC268 B. longum subsp. infantis Isolates Infant feces - No SC417 B. longum subsp. infantis Isolates Infant feces - No SC523 B. longum subsp. infantis Isolates Infant feces - No SC569 B. longum subsp. infantis Isolates Infant feces - No SC600 B. longum subsp. infantis Isolates Infant feces - No SC605 B. longum subsp. infantis Isolates Infant feces - No SC638 B. longum subsp. infantis Isolates Infant feces - No SC638 B. longum subsp. infantis Isolates Infant feces - No DJO10A B. longum subsp. longum Isolates Infant feces GH85 Yes SC91 B. longum subsp. longum Isolates Infant feces - No SC116 B. longum subsp. longum Isolates Infant feces GH18b Yes SC156 B. longum subsp. longum Isolates Infant feces - No SC215 B. longum subsp. longum Isolates Infant feces - No SC249 B. longum subsp. longum Isolates Infant feces - No

251 241! SC280 B. longum subsp. longum Isolates Infant feces - No SC513 B. longum subsp. longum Isolates Infant feces - No SC536 B. longum subsp. longum Isolates Infant feces - No SC558 B. longum subsp. longum Isolates Infant feces - No SC592 B. longum subsp. longum Isolates Infant feces - No SC596 B. longum subsp. longum Isolates Infant feces - No SC618 B. longum subsp. longum Isolates Infant feces - No SC630 B. longum subsp. longum Isolates Infant feces GH18b Yes SC633 B. longum subsp. longum Isolates Infant feces - No SC657 B. longum subsp. longum Isolates Infant feces - No SC662 B. longum subsp. longum Isolates Infant feces - No SC700 B. longum subsp. longum Isolates Infant feces - No SC706 B. longum subsp. longum Isolates Infant feces GH18b Yes ATCC B. breve JCM1273; DSM20091 Intestine of infant GH85 Yes ATCC B. breve JCM1192; DSM20213 Intestine of infant - No ATCC B. breve JCM7016 Intestine of infant - No JCM 7017 B. breve Human feces - No JCM 7019 B. breve Infant feces GH85 Yes JCM 7020 B. breve Infant feces GH85 Yes S-17c B. breve [63] Infant feces - No S-46 B. breve [63] Infant feces - No SC81 B. breve Isolates Infant feces - No SC95 B. breve Isolates Infant feces GH85 Yes SC139 B. breve Isolates Infant feces GH85 Yes SC154 B. breve Isolates Infant feces - No SC500 B. breve Isolates Infant feces - No SC506 B. breve Isolates Infant feces GH85 Yes SC507 B. breve Isolates Infant feces - No SC522 B. breve Isolates Infant feces - No SC559 B. breve Isolates Infant feces Yes SC567 B. breve Isolates Infant feces - No SC568 B. breve Isolates Infant feces GH85 Yes SC573 B. breve Isolates Infant feces - No SC580 B. breve Isolates Infant feces - No

252 242! KA179 B. breve Isolates Infant feces GH85 Yes JCM1254 B. bifidum DSM20082 Intestine of adult - No ATCC B. bifidum JCM1255; DSM20456 Infant feces - No ATCC B. bifidum JCM1209; DSM20082 Infant feces - No JCM 7002 B. bifidum Human feces - No JCM 7003 B. bifidum Human feces - No JCM 7004 B. bifidum Intestine of infant - No ATCC B. bifidum JCM1255 Infant feces - No KA75 B. bifidum Starter culture Probioplus - No SC112 B. bifidum Isolates Infant feces - No SC126 B. bifidum Isolates Infant feces - No SC555 B. bifidum Isolates Infant feces - No SC572 B. bifidum Isolates Infant feces - No SC583 B. bifidum Isolates Infant feces - No : As determined with degenerate primers. (-) represents PCR product not detected. : As determined after coincubation with RNAseB and SDS-PAGE. Yes: activity found; No: activity not detected.

253 243! Table S2: Primers used in this study Primer name Primer sequence (5-3 ) a) Degenerate primers GH85degF GH85degr GH18degF GH18degR TAYTGGCARTAYGTNGAY CCAYTTYTCRTCRTCYTC CTNGAYATHGAYATGGAR NGANCCRTAYTGYTGRTA b) DNA walking TSP142-5F1 TSP142-5F2 TSP142-5F3 TSP142-5R1 TSP142-5R2 TSP142-5R3 CAACCGAGGTCATGTACGTT CGTAATCGCTCTTGAGCTTGTC ACTGGGAACGTAGCTGAACA AACGTACATGACCTCGGTTG CACGATGTTCCTTTACGACACC GACACCAATGGCAGCTACACTG c) Cloning of bifidobacterial endoglycosidases 2468F R11 142cF 142cR GH85cF GH85cR CACCATGAATGCGGACGCCGTTTCTCCGAC GCCGGTCGCACTCAGTTGCTTCGG CACCATGGTTGCGAACGCCCAGGAGGGGGA CGCCGCGTTTCTGGCCGTGGTCA CACCATGACCAAGTACACGATCACACCGGAG GAGGATCGCGCCGTCTGCGCCACGTACC d) Site directed mutagenesis 2468mutP (PO4)-GATATCGACATGCAGGCGCACCCGAAT

254 244! e) qpcr Blon0393qF TTCACCGAGGCGTACAACA Blon0393qR CGCATCCGTGACCACATAG Blon2468qF ACAGAGCCACCCCTGCGATG Blon2468qR GCCGGTTCCGACGCCAGATT Blon2470qF CACGATGCTGGTGAGTGC Blon2470qR CCGGAACCGGTAAGATCC Blon2471qF ACAACCGTTTCAGCAAGACC Blon2471qR GAGCAGACGGTTGAAGAAGG Blon2472qF ATGATCGCCGTCACGATATT Blon2472qR GAACATCAGCAGGGAGAAGC Blon0177qF TCCGGTCGGCATTCACGCAC Blon0177qR GGCAACGGTCTCGGCGTTGT Blon0178qF TGGTCTGCGCACGCTGAAGG Blon0178qR GGCACCTCGGCCATCACACC Blon0881qF GGCCACGTCGGCTTCAACGA Blon0881qR GAACGCCAGCAGCACGAGGT Blon0882qF TCGTTTCCCGCGTGACCACG Blon0882qR CCACGTAGCCGGGGGTCAGA Blon0883qF ATCGAAGCCGTGTGGATT Blon0883qR CCTCGTTGTAGGCGTCGTA Blon0868qF ACAGCTCGCGGTGGAGTCCT Blon0868qR TCCAGCGGCTTGCCTTTCGG Blon0869qF GCAGCAGCGTGTCAAACCGC Blon0869qR GCCGGGAACGCGGAAAGGTT Blon2335qF CCTGTTCAACCAGGATGAGTC Blon2335qR CCGTCCACGACGAAGTAG

255 245! Blon2336qF Blon2336qR Blon2177qF Blon2177qR Blon2344qF Blon2344qR Blon2347qF Blon2347qR ATCACGCTCACCCTCCC ACATCGTCGAAGCGGAGT GGTTCCTGAGGTCTTCACCA GCCGAGCTTCTCAAATTCA TCAAGAAGCTCGACCCGTTG TTGGCGTAGAAGCCGTATGT AAGCCGATAGGTTCTCCCT TCGCCTTGGTGTACTTGTCT

256 246! Figure S1: Multiple alignment of Bifidobacterium endoglycosidase GH18 or GH85 sequences found in this study. Numbers indicate aminoacid positions. Underlined are active site sequences as described in (25, 40).

257 247! Figure S2: Representation of gene landscapes for endoglycosidases found in organisms listed in the figure. The IMG database was used to obtain gene coordinates. GH: glycosyl hydrolase; SBP: solute-binding protein; PTS: phosphotransferase system.

258 248! Figure S3: Optimum ph (A) and temperature (B) for EndoBI-1, EndoBI-2 and EndoBB as determined by 1 h incubations with RNAseB and evaluated in 4-15% SDS-PAGE gels.

259 249! Figure S4: MALDI-FT-ICR MS analysis of glycans released after coincubation of EndoBI-1 with A: RNAseB; B: IgG; C: hlf in negative mode; D: IgA in negative mode.

260 250! Figure S5: Growth of bifidobacterial isolates on 5% of yeast mannoprotein. Lines are representative of three replicates, and blanks containing the same media but no bacteria were withdrawn from the curves.

261 251! Figure S6: Fold changes in gene expression of EndoBI-1 during time coincubation with blf or hlf, as indicated in the figure legend. Locus tags are described in the text. Error bars represent SD from three biological replicates.