Tissue and host tropism of influenza viruses: Importance of quantitative analysis

Transcription

1 Science in China Series C: Life Sciences 2009 SCIENCE IN CHINA PRESS Review life.scichina.com Tissue and host tropism of influenza viruses: Importance of quantitative analysis ZHANG Hong 1,2 1 Z-BioMed, Inc., Rockville, MD 20855, USA; 2 Department of Respiratory Medicine, Affiliated Hospital of Zunyi Medical College, Zunyi , China It is generally accepted that human influenza viruses preferentially bind to cell-surface glycoproteins/ glycolipids containing sialic acids in α2,6-linkage; while avian and equine influenza viruses preferentially bind to those containing sialic acids in α2,3-linkage. Even though this generalized view is accurate for H3 subtype isolates, it may not be accurate and absolute for all subtypes of influenza A viruses and, therefore, needs to be reevaluated carefully and realistically. Some of the studies published in major scientific journals on the subject of tissue tropism of influenza viruses are inconsistent and caused confusion in the scientific community. One of the reasons for the inconsistency is that most studies were quantitative descriptions of sialic acid receptor distributions based on lectin or influenza virus immunohistochemistry results with limited numbers of stained cells. In addition, recent studies indicate that α2,3- and α2,6-linked sialic acids are not the sole receptors determining tissue and host tropism of influenza viruses. In fact, determinants for tissue and host tropism of human, avian and animal influenza viruses are more complex than what has been generally accepted. Other factors, such as glycan topology, concentration of invading viruses, local density of receptors, lipid raft microdomains, coreceptors or sialic acid-independent receptors, may also be important. To more efficiently control the global spread of pandemic influenza such as the current circulating influenza A H1N1, it is crucial to clarify the determinants for tissue and host tropism of influenza viruses through quantitative analysis of experimental results. In this review, I will comment on some conflicting issues related to tissue and host tropism of influenza viruses, discuss the importance of quantitative analysis of lectin and influenza virus immunohistochemistry results and point out directions for future studies in this area, which should lead to a better understanding of tissue and host tropism of influenza viruses. tissue and host tropism, influenza viruses, sialic acid receptors, quantitative analysis The influenza viruses belong to the Orthomyxoviridae family and are enveloped animal viruses containing a segmented single-stranded RNA genome [1,2]. They are classified into three types, A, B and C, based on their immunologically distinct nucleoprotein and matrix protein antigens. Type A influenza viruses are further grouped into antigenic subtypes according to their specific surface glycoproteins hemagglutinin (HA) and neuraminidase (NA). Influenza A occurs in humans, other mammals, and birds and sixteen distinct HA subtypes and nine NA subtypes are currently recognized. Influenza A attracts the most attention because it is associated with high morbidity and mortality in people of all ages, caused millions of deaths in the previous three influenza pandemics, is the cause of the current influenza H1N1 pandemic and has the potential to cause more pandemics in the future [3,4]. A new human influenza virus could initiate as an avian or animal influenza virus that adapts to humans through accumulation of mutations or as a hybrid influ- Received October 1, 2009; accepted November 11, 2009 doi: /s x Corresponding author ( hzhang@zbiomed.com ) Citation: ZHANG Hong. Tissue and host tropism of influenza viruses: Importance of quantitative analysis. Sci China Ser C-Life Sci, 2009, 52(12): , doi: /s x

2 enza virus containing a combination of genes derived from an avian, an animal and a human influenza virus. Therefore, influenza pandemics arise when a new influenza virus emerges from human, avian or animal influenza viruses or from reassortants of them, infects humans, and spreads efficiently among people. The first transmission of the highly pathogenic H5N1 avian influenza A virus directly from chickens to humans in Hong Kong was reported in 1997 [5]. As of 31 August 2009, a total of 440 laboratory-confirmed human cases of avian influenza A (H5N1) virus infections from 15 countries have been reported to WHO and 262 of them (59.5%) have been fatal ( In China, 25 of the 38 confirmed cases (65.8%) have been fatal. Among the 45 confirmed cases (12 deaths) reported to WHO in 2009, 7 were from China and 4 of them were fatal. These events demonstrate that avian influenza viruses can infect human directly and China is an important country for the surveillance, prevention and control of potential influenza pandemics. Continued circulation of H5N1 and other avian influenza viruses in Asia shows that a future pandemic from avian influenza is a real threat. The current circulating influenza A H1N1 viruses were probably originated from the combination of six gene segments (PB2, PB1, PA, HA, NP, and NS1) from swine influenza H1N2 viruses circulating in the USA from 1999 to 2001 (nucleotide sequence similarities >96%) and two gene segments (NA and M1) from swine influenza viruses circulating in Europe from 1985 to 1999 (nucleotide sequence similarities >93%) [6]. The gene segments of these influenza A H1N1 viruses could also be traced back to human or avian origins if those nucleotide sequences with similarities lower than 93% were considered. Since the first report of the swine influenza H1N1 in April 2009 [7], there have been over cases and at least 6770 deaths (1.29%) reported to the WHO from all regions of the world (www. who. int). The WHO Director-General raised the level of influenza pandemic alert from phase 5 to phase 6 on June 11, 2009 following the advice from the Emergency Committee which concluded that the criteria for a pandemic had been met. Currently, pandemic H1N1 influenza virus continues to be the predominant circulating virus of influenza in the world and all pandemic H1N influenza viruses sequenced so far have been antigenically and genetically very similar to the first sequenced new influenza virus (A/California/7/2009). At this stage, the influenza H1N1 pandemic can still be characterized as being moderate in terms of severity and mortality rate. The epidemiology, molecular biology, pathology, and pathogenic factors of influenza viruses including H5N1 in humans have been comprehensively reviewed recently [8 14]. This article will briefly review the generalized view of receptors for influenza viruses, comment on some conflicting issues related to tissue and host tropism of influenza viruses, discuss the importance of quantitative analysis of lectin and influenza virus immunohistochemistry results and point out directions for future studies in this area. 1 Generalized view of receptors for influenza viruses The generalized view is that avian influenza viruses replicate in the gastrointestinal tract of the host and preferentially bind to cell-surface glycoproteins or glycolipids containing sialic acids in α2,3-linkage (SAα2,3); while human influenza viruses replicate in the host s respiratory tract and bind to respiratory epithelial cells via sialic acids attached to glycoproteins or glycolipids containing sialic acids in α2,6-linkage (SAα2,6). Hundreds of articles including many reviews related to influenza virus receptors have been published during last 20 years. Many review articles and several articles published in major scientific journals such as Science, Nature and Nature Medicine made generalized statements about influenza virus receptors and cited two original articles published in 1983 [15,16]. Here are some examples: Human and avian influenza A viruses differ in their recognition of host cell receptors: the former preferentially recognize receptors with saccharides terminating in sialic acid-α2,6-galactose (SAα2,6Gal), whereas the latter prefer those ending in SAα2,3Gal [17] ; Avian influenza viruses bind to cell-surface glycoproteins or glycolipids containing terminal sialyl-galactosyl residues linked by 2-3-linkages [Neu5Ac(α2-3)Gal], whereas human viruses, including the earliest available isolates from 1957 and 1968 pandemics, bind to receptors that contain terminal 2-6-linked sialyl-galactosyl moieties [Neu5Ac(α2-6)Gal] [18] ; HA binds to receptors containing glycans with terminal sialic acids, where their precise linkage determines species preference. A switch in receptor specificity from sialic acids connected to galactose in α2-3 linkages (avian) to α2-6 linkages (human) is a major obstacle for influenza A viruses to 1102 ZHANG H. Sci China Ser C-Life Sci Dec vol. 52 no

3 cross the species barrier and to adapt to a new host [19] ; Human viruses of the H1, H2 and H3 subtypes that are known to have caused pandemics in 1918, 1957, and 1968, respectively, recognize α2,6-linked sialic acid [20] ; and Human influenza strains preferentially bind to sialic acid residues linked to galactose by the α2,6 linkage, while avian and equine influenza strains recognize sialic acid linked to galactose by α2,3 linkage. [21] In one of their original articles published in 1983, Rogers et al. [15] modified human type B erythrocytes to contain sialyloligosaccharides of defined sequence with different sialyltransferases purified from porcine submaxillary glands and rat liver, and examined differential adsorption of influenza virus strains isolated from a variety of hosts. Hemagglutination results of human and animal strains (Table 1 of the article by Rogers et al. [15] ) indicated that receptor determinants of human (SAα2,6Gal) and avian (SAα2,3Gal) influenza virus isolates were accurate only for H3 subtype influenza A viruses [15 ]. In contrast, subtype H2 human strain (A/Japan/305/57) showed the same HA titers (1024) for both SAα2,6Gal and SAα2,3Gal receptors; subtype H1 human strains A/PR/8/34 and A/FM/1/47 preferentially bound to avian specific receptor SAα2, 3Gal (HA titers of 256 and 512) instead of human specific receptor SAα2,6Gal (HA titers of 128 and 256); and subtype H1 avian strain (A/duck/Alberta/513/78) bound to both human and avian receptors (HA titers of 512 for SAα2,6Gal and 1024 for SAα2,3Gal). Therefore, their results with H1 and H2 subtype influenza viruses were contradictory to those with H3 subtype influenza viruses and did not support the generalized view of influenza virus receptor specificities for either human (SAα2,6Gal) or avian (SAα2,3Gal) receptors. Rogers et al. were probably aware of their contradictory results of H1 and H2 subtypes at the time and cautiously concluded in the abstract of the article Receptor specificity appeared, to some extent, to be dependent on the species from which the virus was isolated. In particular, human isolates of the H3 serotype all agglutinated cells containing the SAα2,6Gal linkage, but not cells bearing the SAα2,3Galβ1,3GalNAc sequence. In contrast, antigenically similar (H3) isolates from avian and equine species preferentially bound erythrocytes containing the SAα2,3Gal linkage [15]. It is unclear how Receptor specificity appeared, to some extent, to be dependent on the species (true for H3 subtype isolates only) could have evolved to become generalized statements such as human influenza viruses preferentially recognize receptors with SAα2,6Gal, whereas avian influenza viruses preferentially recognize receptors with SAα2,3Gal (implying isolates of all HA subtypes). In another article of Rogers et al., the sequencing bands showing mutations of CTG-leucine (human specific) to CAG-glutamine (avian specific) or ATG-methionine (specific for both human and avian receptors) at amino acid 226 of HA1 were not clear enough to make definitive conclusions [16]. In conclusion, the generalized view that avian and human influenza viruses preferentially bind to cell-surface glycoproteins or glycolipids containing sialic acids in α2,3-linkage and α2,6-linkage respectively may not be accurate and absolute for all subtypes of influenza A viruses, and therefore, would need to be reevaluated carefully and realistically with quantitative analysis of more experimental results. 2 Distribution of sialic acid receptors and cell/tissue tropism of influenza viruses Lectins such as Sambucus nigra agglutinin (SNA) and Maackia amurensis agglutinin (MAA) have been used to measure the distribution of SAα2,3Gal or SAα2,6 terminated glycoprotein or glycolipids in tissue sections by immunohistochemistry methods. Labeled influenza viruses have also been used to determine the cell and tissue tropism of human, avian and animal influenza viruses. To illustrate that some of the studies published in major scientific journals are inconsistent and caused confusion on the subject of cell and tissue distribution of SAα2,3 and SAα2,6 receptors as well as tropism of influenza viruses, those related studies have been compared and summarized in the Table 1. Matrosovich et al. demonstrated in 2004 that human and avian influenza viruses target different cell types in cultures of human airway epithelium: human influenza viruses preferentially infected non-ciliated cells, while avian influenza viruses mainly infected ciliated cells [18]. They found that tropism for a specific type of airway epithelial cells depends on the virus host species, rather than on virus type, subtype, or strain and this pattern correlated with the predominant localization of human influenza receptors (α2-6-linked sialic acids) on non-ciliated cells and avian influenza receptors (α2-3-linked ZHANG H. Sci China Ser C-Life Sci Dec vol. 52 no

4 Table 1 Summary of recently published studies on the subject of tissue distribution of SAα-2,3 and SAα-2,6 receptors and in6fection of human tissues with influenza viruses Published Studies PNAS 2004 [18] Nature 2006 [22] Science 2006 [23] Respir Res 2007 [24] FASEB J 2008 [25] Distribution of SAα-2,3 and SAα-2,6 Receptors In Tissues ΜΑΑ Binding for SAα-2,3Gal receptor (avian) Nasal/ pharynx Trachea/ bronchus Lung aveolus Bound to some ciliated primary epithelial cells Bound to some ciliated primary epithelial cells Occasionally detected on epithelial cells in nasal mucosa Found on non-ciliated cuboidal bronchiolar cells at the junction between bronchiole and alveolus Found on cells lining the alveolar wall Strong MMA1 binding but no MMA2 binding in the epithelium MAA1 bound to bronchial cells; Strong MMA binding but no MMA2 binding to ciliated and nonciliated cells MAA1 bound to alveolar macrophages but not pneumocytes; Detected on a small proportion of non-ciliated and a few ciliated epithelial cells Detected on non-ciliated and a few ciliated epithelial cells and endothelial cells (EC) Detected on type II cells and endothelial cells MAA2 bound to pneumocytes Other tissues Detected on brain neuron and EC; placenta EC; liver Kupffer cells; kidney glomerular EC; heart EC; intestines neurons and EC; and T cells in spleen SΝΑ Binding for SAα-2,6Gal receptor (human) Nasal/ pharynx Trachea/ bronchus Strong staining on non-ciliated primary epithelial cells Strong staining on nonciliated primary epithelial cells Dominant on epithelial cells in nasal mucosa Found on trachea and bronchus epithelia cells Strong binding of SNA in the epithelium Strong binding of SNA in ciliated and non-ciliated cells Detected on ciliated and non-ciliated epithelial cells and EC Detected on ciliated and nonciliated epithelial cells and EC Lung/aveolus Found on alveolar cells Minimal SNA binding Detected on pneumocytes and EC Other tissues Detected on brain EC; placenta Hofbauer cells and EC; liver bile duct epithelial cells and EC; kidney glomerular EC and distal tubule epithelial cells; heart EC; intestines EC; and B cells in spleen (to be continued on the next page)

5 (Continued) Tissues Infected by Influenza Viruses Human viruses Preferentially infected nonciliated cells Human A/Kawasaki/173/01 (H1N1) bound extensively to epithelial cells in the bronchi and, to a lesser degree, to alveolar cells; H5N1 virus (A/Hong Kong/213/03) infected both alveolar cells and bronchial epithelial cells Human A/Netherlands/213/03 (H3N2) bound to epithelial cells in trachea Avian viruses Mainly infected ciliated cells Avian A/duck/Mongolia/301/01 (H3N2) and A/duck/Vietnam/5001/05 (H5N1) viruses didn t infect bronchial epithelial cells; but they bound extensively to alveolar cells; A/duck/Vietnam/5001/05 did infect alveolar cells H5N1 virus (A/Vietnam/1194/04) attached predominantly to type II pneumocytes, alveolar macrophages, and nonciliated cuboidal epithelial cells in terminal bronchioles ; Attachment became progressively rarer toward the trachea Conclusions 1. The predominant localization of human receptors (SAα2,6) localized predominantly on non-ciliated epithelial cells; 2. Avian receptors (SAα2,3) localized predominantly on ciliated epithelial cells; 3. Ciliated cells most likely serve as primary target cells in cases where avian viruses cause human disease. 1. Avian and human flu viruses seem to target different regions of a patient s respiratory tract ; 2. Although not quantitative, these results indicate that there could be functional significance in the preferential expression of SAα2,3Gal and SAα2,6Gal molecules on human airway cells ; 3. Findings indicate that H5N1 viruses can replicate efficiently only in cells in the lower region of the respiratory tract. 1. The findings fit with the limited pathology data for H5N1 virus infection in humans ; 2. The findings contrast with the idea that avian influenza viruses generally have little affinity for human respiratory tissues ; 3. The predilection of H5N1 virus for type II pneumocytes and alveolar macrophages may contribute to the severity of the pulmonary lesion. 1. MAA binding pattern may be dependent on the supplier and different isoforms show a different tissue distribution; 2. Identified a heterogeneous distribution of SNA and MAA binding in bronchial epithelium with no clear distinction between ciliated and non-ciliated cells; 3. SNA binding was widespread in the upper than the lower respiratory tract. 1. HuIV-Rs were abundantly present in the respiratory tract and lungs. They were also detected on Hofbauer cells, glomerular cells, splenic B cells, and in the liver ; 2. Endothelial cells of all organs examined expressed both human and avian receptors; 3. The distribution pattern of AIV-Rs is partially inconsistent with the pattern of infected cells as detected in previous studies, which suggests there may be other receptors or mechanisms involved in H5N1 infection. SNA, a lectin from the elderberry plant Sambucus nigra agglutinin; MAA, Maackia amurensis agglutinin; EC, endothelial cells.

6 sialic acids) on ciliated cells. Based on quantitative analysis of immunostaining of influenza virus-infected cells, Matrosovich et al. concluded that ciliated cells most likely serve as primary target cells in those rare cases where avian viruses cause human disease [18]. In contrast, Shinya et al. reported in their 2006 article that α2-3-linked sialic acids (avian influenza receptors) were found on non-ciliated cuboidal bronchiolar cells and α2-6-linked sialic acids (human influenza receptors) were dominant on epithelial cells in nasal mucosa, with α2-3-linked sialic acids being occasionally detected [22]. However, a substantial number of cells lining the alveolar wall also expressed α2-3-linked sialic acids. Shinya et al. observed that human-derived viruses (such as A/Kawasaki/173/01) bound extensively to epithelial cells in the bronchi and, to a lesser degree, to alveolar cells (3 stained cells shown in Figure 2b of the article); by contrast, avian viruses (such as A/duck/ Mongolia/301/ 01) bound extensively to alveolar cells (8 stained cells shown in Figure 2d) but less widely to bronchial epithelial cells based on qualitative analysis of their immunohistochemical results. They concluded that A/Hong Kong/213/03 virus infected bronchial epithelial cells by showing about 9 stained cells (Supplementary Figure 4c) and alveolar cells by showing about 12 stained cells (Supplementary Figure 4d). Shinya et al. suggested, Although not quantitative, these results indicate that there could be functional significance in the preferential expression of SAα2,3Gal and SAα2,6Gal molecules on human airway cells and Our findings indicate that H5N1 viruses can replicate efficiently only in cells in the lower region of the respiratory tract. [22] In their article published in 2006, van Riel et al. [23] identified the cell types in the lower respiratory tract of humans and four animal species (mouse, ferret, macaque, and cat) to which H5N1 virus (A/Vietnam/1194/04) attached by showing 1 to 3 stained cells in each of the tissue sections using immunohistochemistry technique. They found that H5N1 virus attached mainly to type II pneumocytes, alveolar macrophages, and non-ciliated cuboidal epithelial cells in terminal bronchioles; and attachment became progressively rarer toward the traches. They concluded, These findings fit with the limited pathology data for H5N1 virus infection in humans, which show diffuse alveolar damage and the presence of H5N1 virus antigen in type II pneumocytes. However, they contrast with the idea that avian influenza viruses generally have little affinity for human respiratory tissues. They also suggested The predilection of H5N1 virus for type II pneumocytes and alveolar macrophages may contribute to the severity of the pulmonary lesion. [23] Using Maackia amurensis lectin II (MAL-II, MAA2) and Sambubus nigra agglutinin (SNA), Yao et al. investigated the distribution of α2,3-linked and α2,6-linked sialic acids in various organs including upper and lower respiratory tracts of two H5N1 cases and 14 control cases [24]. They observed that avian influenza receptors (α2,3-linked sialic acids) were detected on different cells including type II pneumocytes, a limited number of epithelial cells of the upper respiratory tract, the bronchi, bronchioli, and trachea, splenic T cells, and neurons in the brain and intestines. They suggested, the relative lack of AIV-Rs in the upper airway may be one of the factors preventing efficient human-to-human transmission of H5N1 influenza [24]. Yao et al. found that human influenza receptors (α2,6-linked sialic acids) were abundantly present in the respiratory tract and lungs and were detected on Hofbauer cells, glomerular cells, splenic B cells, and in the liver. Yao et al. also found that endothelial cells of all organs examined expressed both avian and human receptor types which may account for the multiple organ involvement in H5N1 influenza [24]. In their 2007 article, Nicholls et al. compared the lectin binding properties of MAA from different suppliers and found that there were differences in tissue distribution of the α2,3 linked SA when 2 different isomers of MAA (MAA1 for SAα2,3Galb1,4GlcNac and MAA2 for SAα2,3Galb1,3GlcNac) lectin were used. MAA1 had widespread binding throughout the upper and lower respiratory tract. By comparison, MAA2 binding was mainly restricted to the alveolar epithelial cells of the lung [25]. They suggested in the article We also urge attention to the exact isoform of MAA present in lectins supplied by different manufacturers. These results imply a need for a re-evaluation of the findings reported in previous studies on the tissue distribution of SA receptor types. They also mentioned, Since MAA1 bound to non-saα2,3 glycans it has not been used as extensively by some researchers as MAA2. For instance, Shinya et al. have recently demonstrated SAα2,3Gal expression only in the lung but not the bronchus or 1106 ZHANG H. Sci China Ser C-Life Sci Dec vol. 52 no

7 upper respiratory tract [7]. Since they only used MAA2 lectin binding (Y. Kawaoka, personal communication) our MAA2 results are in accord with theirs. But our finding of MAA1 binding in the upper and lower respiratory tract does have implications for the possible distribution of receptors for avian influenza viruses including the currently circulating H5N1 viruses. It is clear that results by Shinya et al. SAα2,6Gal is dominant on epithelial cells in nasal mucosa, with SAα2,3Gal being occasionally detected [22] and their conclusion Our findings may provide a rational explanation for why H5N1 viruses at present rarely infect and spread between humans although they can replicate efficiently in the lungs [22] were incomplete because they only used MAA2. As summarized in Korteweg and Gu s review article, Conflicting results have been reported as to the cell type expressing avian influenza virus receptors in the trachea and bronchi. [8]. Tollis et al. also stated the determinants involved in host range restriction have not been completely clarified The H5N1 viruses isolated from human in Hong Kong were shown to retain the receptor binding properties of avian viruses, thus suggesting that gene products other than HA may contribute to host range restriction [12]. The summarized results of recently published studies on the subject of cell and tissue distribution of SAα2,3 and SAα2,6 receptors and tissue tropism of influenza viruses show the inconsistency of these studies, because some of the studies were qualitative descriptions of receptor distributions based on immunohistochemistry results with limited numbers of stained cells and without common standard controls (Table 1). To make published results repeatable and comparable among different laboratories, immunohistochemistry results should be analyzed quantitatively. 3 Alpha 2-3 and alpha 2-6 linked sialic acids are not the sole receptors determining cell and host tropism for influenza viruses More than 750 articles related to influenza virus receptors have been published during last 10 years and attention has primarily been given to the role of alpha 2-3 and alpha 2-6 linked sialic acids as receptors of human, avian and animal influenza viruses. However, conflicting results have been reported in major scientific journals on the subject of the cell and tissue distribution of influenza virus receptors (Table 1) [18,22 26]. As Yao et al. concluded after investigating the expression of avian (AIV-Rs) and human influenza receptors (HuIV-Rs) in various organs (upper and lower respiratory tracts, brain, kidney and intestines, etc.) of two H5N1 cases and control cases, the distribution pattern of AIV-Rs is partially inconsistent with the pattern of infected cells as detected in previous studies, which suggests there may be other receptors or mechanisms involved in H5N1 infection and The absence of AIV-Rs in several cell types, including ciliated epithelial cells of the trachea, intestinal epithelial cells, cytotrophoblasts, and Hofbauer cells appears to be in contrast with the detected H5N1-infection of these cells. This suggests that other receptors or coreceptors might be involved in virustarget-cell interaction in H5N1 avian influenza [25]. They also mentioned that It is remarkable, however, that despite the widespread and abundant expression of AIV-Rs in the lungs, only a limited number of pneumocytes have been found to be infected in previous studies, implying that the presence of AIV-Rs is not the only factor affecting the capability of AIV to infect cells. [25] Recent studies published by other groups also indicate that α2,3- and α2,6-linked sialic acids are not the sole receptors determining tissue tropism of influenza viruses [26 32]. The recent article by Nicholls et al. demonstrated that ex vivo cultures of human nasopharyngeal, adenoid and tonsillar tissues can be infected with H5N1 viruses in spite of an apparent lack of sialic acid α2-3 virus receptors [26]. Previous studies by other groups using MDCK cells [27] and human airway tracheobronchial epithelial cells [28] have shown that removal of sialic acid from cultured cells by broad-spectrum sialidases does not abolish infection by human influenza viruses. Using human A549 and Hela cells with high levels of α2-6 sialic acid and CHO cells that have only α2-3 sialic acid, Kumari et al. found that absence of α2-6 sialic acid does not protect a cell from influenza infection and the presence of high levels of α2-6 SA on a cell surface does not guarantee productive replication of a virus with α2-6 receptor specificity [29]. Binding studies of influenza viruses to sialic acids show that reassortant viruses with A/NWS/33 HA bind to α2,8-linked sialic acid [30]. In addition, Glaser et al. [31] have demonstrated that human influenza viruses can infect knock-out mice lacking a major α2,6 sialyltransferase and grow to ZHANG H. Sci China Ser C-Life Sci Dec vol. 52 no

8 similar titers in the lung and trachea as compared to wild-type mice. Results from these studies indicate that α2-3 and α2-6 linked sialic acids are not the sole receptors determining cell and host tropism of influenza viruses and suggest that the absence of these receptors does not necessarily mean cells are protected from infection by influenza viruses automatically. Therefore, it is over simplified to assume that removal of sialic acid receptors from human respiratory tracts would eliminate infections by H5N1, H1N1 and other influenza viruses [33]. Other factors, such as glycan topology [34], local density of receptors, concentration of invading viruses, lipid raft microdomains [35,36], coreceptors or sialic acid-independent receptors, may also be important. In light of the findings discussed above, more attention should be given to other cell surface components in the human respiratory tract important for interactions with avian, swine and human influenza viruses, which will lead to a better understanding of cell and host tropism of influenza viruses and provide new targets for more effective prevention and treatment of seasonal and pandemic influenza. 4 Quantitative analysis of cell and tissue tropism of influenza viruses Three of the articles published in major scientific journals were selected as examples to illustrate why it is important to quantitatively analyze immunohistochemical staining results. In the article published in 2007, Nicholls et al. designed a study to determine the tissue tropism of seven influenza A viruses (one H1N1, one H3N2 and five H5N1) in the upper and lower respiratory tracts [26]. These seven influenza viruses included A/Hong Kong/54/98 (H1N1), A/Hong Kong/1174/99 (H3N2), A/Vietnam/3046/04 (H5N1), A/Hong Kong/483/95 (H5N1), A/Hong Kong/213/03 (H5N1), A/Vietnam/1203/04 (H5N1) and A/Chicken/Indonesia/ BL/03 (H5N1). Nicholls et al. qualitatively reported the cellular location of influenza A nucleoprotein in different tissues and in primary nasopharyngeal cells infected with influenza viruses H1N1, H3N2 and one of the five H5N1 viruses, A/Vietnam/3046/04. These results would be more convincing if the authors had performed a quantitative analysis of the data by scanning more fields under a microscope to count more stained epithelial cells. It would be interesting as well for the scientific community to know whether the other four H5N1 viruses, especially A/Chicken/Indonesia/ BL/03, have the same cellular tropism as that of A/Vietnam/3046/04 [26]. In the same article, the thermal inactivation curve shown in Figure 1e indicated that it took 8 hours to reduce the virus (A/Vietnam/3046/04) titer from 10 4 to % tissue culture infectious doses per ml (TCID 50 / ml). However, as shown in Figure 1h, it took 18 hours to reduce the virus titer from 10 5 to 10 4 TCID 50 /ml for the same influenza virus. If the thermal inactivation curves in Figures 1e and h were switched, then the results of Figure 1e might have not supported the conclusion that productive viral infection in the tissue fragments from the nasopharynx was demonstrated by an increase in the viral yield in culture supernatants (Figure 1e). To show productive viral infection of influenza viruses in primary cultured cells, Gu et al. carried out three independent experiments with each experiment in duplicate [37], therefore, their data could be averaged and expressed as a mean±sd which demonstrated that virus titer was increased continuously from 24 to 48 and 72 hours after infection. In contrast, Nicholls et al. showed decreased virus titers from 20 to 24 hours after infection in Figure 1e and from 24 to 48 hours after infection in Figure 1h using the same virus at the same input dose (10 10 TCID 50 /ml) [26]. It would be helpful if the authors explained why the thermal inactivation curves for the same virus are different between the two experiments and why virus titers did not continuously increase from 20 to 24 hours in Figure 1e and from 24 to 48 hours in Figure 1h after infection. One of the possible explanations is that each experiment was carried out only once and the results may not be reproducible. Furthermore, the authors mislabeled the influenza virus A/Hong Kong/54/98 (H1N1) used in the study as 54/98 (H5N1) in Figure 2d. Nicholls et al. also reported results of lectin histochemistry on the ex vivo biopsies and archival tissues (Supplementary Figure 2), although they realized lectin histochemistry should not be regarded as definitive [26]. There are some concerns regarding their lectin histochemistry studies: (i) Published data are not quantitative; and (ii) the definition of strong positive staining is not given. As described by Matrosovich et al. in their 2004 article [18] and Gu et al. in their 2007 article [37], which quantitatively described immunostaining studies of influenza virus-infected cells, it is feasible to analyze the lectin histochemistry results quantitatively. Matrosovich et al. counted cells by observing en face with 1108 ZHANG H. Sci China Ser C-Life Sci Dec vol. 52 no

9 oil-immersion at 1000 magnification, analyzed fields for each sample, averaged the results and calculated the percentages of specific cells with respect to the total amount of cells. However, Nicholls et al. choose to qualitatively describe their lectin histochemistry results as strong or poor staining instead of quantitatively counting lectin stained/unstained cells and comparing the percentages of lectin stained cells between tissues from the upper and lower respiratory tracts. Without quantitative analysis, it is possible that one group could report the presence of specific receptors by showing 3 or more lectin or virus bound epithelial cells in one tissue section and another group could report the absence of the same receptors by showing 2 or less lectin or virus bound epithelial cells in another tissue section. The above examples have been used to show the importance of quantitative descriptions of sialic acid receptor distributions based immunohistochemistry results. In order to make published results comparable among different laboratories and avoid potential confusion in the field of tissue and host tropism of influenza viruses, every laboratory should analyze lectin and virus immunohistochemistry results quantitatively before reporting in any scientific journals. 5 Conclusion remarks Determinants of tissue and host tropism of human, avian and animal influenza viruses are more complex than what has been generally accepted. Even though the generalized view is accurate for H3 subtype isolates, it may not be accurate and absolute for all subtypes of influenza A viruses based on the analysis of the original results published in 1983 and, therefore, needs to be reevaluated carefully and realistically. Recent published studies indicate that α2-3 and α2-6 linked sialic acids are not the sole receptors determining tissue and host tropism of human and avian influenza viruses [26 32] Other factors, such as glycan topology [34], lipid raft microdomains [35,36], local density of sialic acid receptors, concentration of invading viruses, coreceptors and sialic acid-independent receptors, may also be important in determining tissue and host tropism of human, avian and animal influenza viruses. In light of the findings discussed above, more attention should be given to other cell surface components in the human respiratory tract important for interactions with influenza viruses. Some of the studies published in major scientific journals are inconsistent and caused confusion on the subject of tissue tropism of influenza viruses. To be more effective in controlling the global spread of pandemic influenza such as the current circulating influenza A H1N1, it is crucial to clarify the determinants for tissue and host tropism of influenza viruses through quantitative analysis of experimental results, which should lead to a better understanding of tissue and host tropism of influenza viruses and provide new targets for effective prevention and treatment of seasonal and pandemic influenza. 1 Stephenson I, Nicholson K G. Influenza: vaccination and treatment. Eur Respir J, 2001, 17: [DOI] 2 Nicholson K G, Wood J M, Zambon M. Influenza, Lancet, 2003, 362: [DOI] 3 Laver G, Garman E. Pandemic influenza: Its origin and control. Microbes Infect, 2002, 4: [DOI] 4 Sorrell E M, Ramirez-Nieto G C, Gomez-Osorio I G, et al. Genesis of pandemic influenza. Cytogenet Genome Res, 2007, 117: [DOI] 5 Centers for Disease Control and Prevention (CDC). Isolation of avian influenza A (H5N1) viruses from humans-hong Kong, May- December MMWR Morb Mortal Wkly Rep, 1997, 46: Zhang H, Chen L. Possible origin of current influenza A H1N1 viruses. Lancet Infect Dis, 2009, 9: [DOI] 7 Centers for Disease Control and Prevention (CDC). Swine influenza A (H1N1) infection in two children-southern California, March-April MMWR Morb Mortal Wkly Rep, 2009, 58: Korteweg C, Gu J. Pathology, molecular biology, and pathogenesis of avian influenza A (H5N1) infection in humans. Am J Pathol, 2008, 172: [DOI] 9 Maines T R, Szretter K J, Perrone L, et al. Pathogenesis of emerging avian influenza viruses in mammals and the host innate immune response. Immunol Rev, 2008, 225: 68 84[DOI] 10 Taubenberger J K, Morens D M. The pathology of influenza virus infections. Annu Rev Pathol, 2008, 3: [DOI] 11 Wong S S, Yuen K Y. Avian influenza virus infections in humans. Chest, 2006, 129: [DOI] 12 Tollis M, Di Trani L. Recent developments in avian influenza research: epidemiology and immunoprophylaxis. Vet J, 2002, 164: [DOI] 13 Zambon M C. The pathogenesis of influenza in humans. Rev Med Virol, 2001, 11: [DOI] 14 Subbarao K, Shaw M W. Molecular aspects of avian influenza (H5N1) viruses isolated from humans. Rev Med Virol, 2000, 10: [DOI] 15 Rogers G N, Paulson J C. Receptor determinants of human and animal influenza virus isolates: Differences in receptor specificity of the H3 hemagglutinin based on species of origin. Virology, 1983, 127: ZHANG H. Sci China Ser C-Life Sci Dec vol. 52 no

10 [DOI] 16 Rogers G N, Paulson J C, Daniels R S, et al. Single amino acid substitutions in influenza haemagglutinin change receptor binding specificity. Nature, 1983, 304: 76 78[DOI] 17 Yamada S, Suzuki Y, Suzuki T, et al. Haemagglutinin mutations responsible for the binding of H5N1 influenza A viruses to human-type receptors. Nature, 2006, 444: [DOI] 18 Matrosovich M N, Matrosovich T Y, Gray T, et al. Human and avian influenza viruses target different cell types in cultures of human airway epithelium. Proc Natl Acad Sci USA, 2004, 101: [DOI] 19 Steven J, Blixt O, Tumpey T M, et al. Structure and receptor specificity of the hemagglutinin from an H5N1 influenza virus. Science, 2006, 312: [DOI] 20 Gamblin S J, Haire L F, Russell R J, et al. The structure and receptor binding properties of the 1918 influenza hemagglutinin. Science, 2004, 303: [DOI] 21 de Jong M D, Hien T T. Avian influenza A (H5N1). J Clin Virol, 2006, 35: 2 13[DOI] 22 Shinya K, Ebina M, Yamada S, et al. Avian flu: influenza virus receptors in the human airway. Nature, 2006, 440: [DOI] 23 van Riel D, Munster V J, de Wit E, et al. H5N1 virus attachment to lower respiratory tract. Science, 2006, 312:399[DOI] 24 Nicholls J M, Bourner A J, Chen H, et al. Sialic acid receptor detection in the human respiratory tract: evidence for widespread distribution of potential binding sites for human and avian influenza viruses. Respir Res, 2007, 8: 73[DOI] 25 Yao L, Korteweg C, Hsueh W, et al. Avian influenza receptor expression in H5N1-infected and noninfected human tissues. FASEB J, 2008, 22: [DOI] 26 Nicholls J M, Chan M C, Chan W Y, et al. Tropism of avian influenza A (H5N1) in the upper and lower respiratory tract. Nat Med, 2007, 13: [DOI] 27 Stray S J, Cummings R D, Air G M. Influenza virus infection of desialylated cells. Glycobiology, 2000, 10: [DOI] 28 Thompson C I, Barclay W S, Zambon M C, et al. Infection of human airway epithelium by human and avian strains of influenza a virus. J Virol, 2006, 80: [DOI] 29 Kumari K, Gulati S, Smith D F, et al. Receptor binding specificity of recent human H3N2 influenza viruses. Virol J, 2007, 4: 42[DOI] 30 Wu W, Air G M. Binding of influenza viruses to sialic acids: Reassortant viruses with A/NWS/33 hemagglutinin bind to alpha2,8-linked sialic acid. Virology, 2004, 325: [DOI] 31 Glaser L, Conenello G, Paulson J, et al. Effective replication of human influenza viruses in mice lacking a major alpha2,6 sialyltransferase. Virus Res, 2007, 126: 9 18[DOI] 32 Chu V C, Whittaker G R. Influenza virus entry and infection require host cell N-linked glycoprotein. Proc Natl Acad Sci USA, 2004, 101: [DOI] 33 Zhang H. Concerns of using fusion protein as an experimental drug to combat seasonal and pandemic influenza. J Antimicrob Chemother, 2008, 62: [DOI] 34 Chandrasekaran A, Srinivasan A, Raman R, et al. Glycan topology determines human adaptation of avian H5N1 virus hemagglutinin. Nat Biotechnol, 2008, 26: [DOI] 35 Scheiffele P, Roth M G, Simons K. Interaction of influenza virus haemagglutinin with sphingolipid-cholesterol membrane domains via its transmembrane domain, EMBO J, 1997, 16: [DOI] 36 Takeda M, Leser G P, Russell C J, et al. Influenza virus hemagglutinin concentrates in lipid raft microdomains for efficient viral fusion. Proc Natl Acad Sci USA, 2003, 100: [DOI] 37 Gu J, Xie Z, Gao Z, et al. H5N1 infection of the respiratory tract and beyond: A molecular pathology study. Lancet, 2007, 370: [DOI] 1110 ZHANG H. Sci China Ser C-Life Sci Dec vol. 52 no