Molar Growth Yields of Certain Lactic Acid Bacteria

Transcription

1 JOURNAL OF BACTERIOLOGY, JUIY 1968, P Copyright 1968 American Society for Microbiology Vol. 96, No. I Printed in U.S.A. Molar Growth Yields of Certain Lactic Acid Bacteria as Influenced by Autolysis HASSAN H. MOUSTAFA AND E. B. COLLINS Department offood Science and Technology, University of California, Davis, California Received for publication 20 April 1968 Molar growth yields determined from batch cultures of Streptococcus diacetilactis and S. faecalis were appreciably greater at the peaks of maximal growth than after continued incubation and considerable autolysis. The higher molar growth yields were about equal to those determined in a continuous culture. Autolysis during logarithmic growth was minimal. The average Y value for adenosine triphosphate (ATP), determined by using limiting concentrations of glucose, galactose, lactose, and maltose for growing S. diacetilactis and limiting concentrations of glucose for growing S. lactis, S. cremoris, and S. faecalis, was This is close to the Y (arginine) value of 17.8 determined with S. faecalis, but 62% greater than the generally accepted value of Data are presented indicating that the oftenused Y (ATP) value of 10.5 is erroneously low. The dry-weight yield of bacterial cells per mole of adenosine triphosphate [Y (ATP)] generated from the catabolism of an energy source is assumed to be constant. A generally accepted value of 10.5 (2, 5) has been used as a biological constant in the elucidation of unknown metabolic pathways (10, 11, 24, 28, 31). Accordingly, an organism that ferments glucose by the hexose diphosphate pathway should yield 21 g (dry weight) of cells per mole of glucose; i.e., Y (glucose) should be 21. Nevertheless, Bauchop and Elsden (2) found that one organism, Propionibacterium pentosaceum, gave yields approximately twice those from Streptococcus faecalis or Saccharomyces cerevisiae. From the average value of 10.5 they had found for Y (ATP), the high yield from P. pentosaceum could not be accounted for by known anaerobic pathways. Gunsalus and Shuster (10) theorized that microorganisms should be able to synthesize 33.3 g (dry weight) of cells with 1 mole of ATP. Beck and Shugart (3) recently determined the yields of two strains of S. faecalis var. liquefaciens from glucose and arginine in a complex medium. From Bauchop and Elsden's results, showing that Y (ATP) equals 10.5, they concluded that the organisms produced one extra mole of ATP per mole of glucose. Forrest and Walker (9) found Y (glucose) for S. faecalis to be 32 with concentrations of glucose less than 6 umoles per ml, and 21 with higher concentrations. They concluded that fermentation of the lower concentrations of 117 glucose by S. faecalis was not of the homolactic type. Hobson (14) found Y (glucose) to be 62 for Selenomonas ruminantium var. lactyliticas growing anaerobically in continuous culture at optimal growth rate. He ruled out the possibility that the high yield was due to the fermentation of lactate and considered 62 to be the correct Y (glucose) value for the organism. He concluded that the organism is able to obtain energy from presently unknown fermentation pathways, or that the generally accepted value of 10.5 for Y (ATP) is not universal. Hungate (16) found the Y (glucose) value for Ruminococcus albus growing in a continuous culture to be 55. Hobson and Summers (15) found a Y (fructose) value of 60 and Y (glycerol) values of 15.8 to 20.2 for an organism called Bacterium 5S in a continuous culture. These same investigators found Y (maltose) values for Bacteroides amylophilus in batch cultures to vary between 110 and 77 for initial respective maltose concentrations of 2.04 to 11.7,umoles/ml. From data obtained with a continuous culture, they calculated the Y (maltose) value to be 160. The difference was attributed to a greater maintenance requirement for the batch cultures. They concluded from their results that Y (ATP) values are about 20 for B. amylophilus growing on maltose, 20 for S. ruminantium growing on glucose, 15 for Bacterium 5S growing on fructose, and 10 for Bacterium 5S growing on glycerol (14, 15).

2 118 MOUSTAFA AND COLLINS J. BACTERIOL. We have studied a rapid decline (autolysis) that occurs in cultures of several lactic acid bacteria immediately after the populations reach maximal concentration (21). This paper reports the influence of this lysis of cells on the molar growth yields of certain lactic acid bacteria. The yields reported herein are higher than the yields found by Bauchop and Elsden (2), but are lower than the theoretical limit of Gunsalus and Shuster (10). MATERIALS AND METHODS Organisms. Used in this study were Streptococcus diacetilactis DRCI and 18-16, S. lactis C2, S. cremoris K2E8, and S. faecalis NCTC6783. The cultures were propagated routinely in skim milk fortified with 0.75% nonfat milk solids. Inocula for experiments were prepared by growing the cultures for 8 to 10 hr in appropriate media, separating the cells by centrifugation at 2,000 X g for 20 min, and resuspending them in 0.01 M phosphate buffer, ph 7.0. The incubation temperature was 37 C for S. faecalis and 32 C for the other species. Media. The partially defined casein-hydrolysate medium (BEO medium) reported by Bauchop and Elsden (2), adjusted to ph 6.6 or 6.8, was used for most of the experiments with S. faecalis. The acidhydrolyzed Oxoid casein (Consolidated Laboratories, Inc., Chicago Heights, Ill.) included in the medium was treated according to the procedure they described (2). In some experiments, 10% of acid-hydrolyzed "vitamin-free" casein (Nutritional Biochemical Corp., Cleveland, Ohio) replaced the Oxoid casein hydrolysate (BEV medium). A complex medium and a partially defined casein-hydrolysate medium (HCM medium) were used in studying S. diacetilactis, S. lactis, and S. cremoris. The complex medium contained 1% peptone, 1.5% yeast extract, 0.1% acidhydrolyzed "vitamin-free" casein, 0.2% K2HPO4, 0.025% MgSO4, and 0.2% sodium acetate, adjusted to ph 6.5. The partially defined casein-hydrolysate medium (HCM medium) was that reported by Harvey and Collins (13) as modified by Moustafa and Collins (21), adjusted to ph 7.1, unless indicated otherwise. In some experiments with S. diacetilactis, the casein hydrolysate of this partially defined medium was replaced by the following amino acids (100,ug of each per ml): threonine, methionine, leucine, valine, arginine, aspartic acid, glutamic acid, lysine, isoleucine, cysteine, histidine, alanine, and tryptophan. (The five amino acids in italics were essential for growth of the organism; the others were necessary for growth comparable to that obtained in the medium with casein hydrolysate.) Each medium was filtered through a type AA Millipore filter (pore size, 0.8,; Millipore Corp., Bedford, Mass.) to remove slight cloudiness. Large test tubes containing 150 to 200 ml of media to be used for batch cultures were sterilized at 121 C for 15 min. Larger quantities of media to be used for continuous cultures (6 to 8 liters in 12-liter reservoirs) were sterilized for 30 min. The compounds serving as energy sources were sterilized by filtration through microporous porcelain filters (no. VFA-54-03; Selas Flotronics, Spring House, Pa.) and added aseptically to the media. Cultivation methodr. Batch cultures were grown in large test tubes (50 X 290 mm) fitted with sampling assemblies and partly immersed in a constant-temperature water bath. Each culture was mixed with a Teflon-coated magnetic stirrer. Strictly anaerobic conditions, when used, were obtained with the procedure of Bauchop and Elsden (2), except that the amount of alkaline pyrogallol was doubled. Samples were obtained aseptically with sterile syringes through rubber septa in the sampling assemblies. The oxygenfree air withdrawn from the assemblies (to force culture into them for sampling) was subsequently replaced with sterile oxygen-free nitrogen. Continuous-culture studies were carried out with a modification of the apparatus of Rosenberger and Elsden (27). The growth vessel was a test tube (50 or 93 X 290 mm, depending on the desired dilution rate) fitted with a rubber stoppper that had inlets for air or nitrogen and fresh medium, and an outlet for overflow. Air was filtered through two glass tubes (25 X 240 mm) filled with sterile glass wool; nitrogen (99.997% purity) was passed through a train of alkaline pyrogallol (32), distilled water, and two tubes of sterile glass wool. Medium was fed from a 12-liter reservoir through Tygon flexible plastic tubing [½2 inch (0.079 cm) inner diameter and %32 inch (0.395 cm) outer diameter] into a syringe-needle tip mounted in the rubber stopper of the growth vessel. For experiments in which anaerobic conditions were required, the air entering the 12-liter reservoir to replace the medium was first passed through alkaline pyrogallol and sterile cotton. Flow rate of the medium was controlled by a model T8 Sigmamotor pump with a Revco Zeromax Speedchanger through which the Tygon tube passed (Sigmamotor, Inc., Middleport, N.Y.). Error in the flow rate was found to be within 2%o. The effluent culture fluid was removed, and the volume in the growth vessel was kept constant by application of slight vacuum or nitrogen pressure. Samples were taken from the overflow. For experiments requiring chemical analysis of the effluent, a type AA Millipore filter (pore size, 0.8,u) was connected directly to the outlet to remove bacterial cells from the effluent. The slight vacuum or nitrogen pressure forced the effluent through the filter. Continuous cultures were started as batch cultures; feeding was begun when the optical density reached a maximum. Contamination was not detected during several weeks of operation, but experiments were usually terminated after about 7 days to avoid bacterial growth on the walls of the growth vessel, as observed by Larsen and Dimmick (18). Antifoam. Dow Corning Antifoam A (Dow Corning Corp., Midland, Mich.), a silicone defoamer, was used to prevent foaming in anaerobic continuous cultures of S. faecalis in BEO medium, which contained Oxoid casein. Approximately 0.1 g/liter was added to the medium before autoclaving. Growth and yield measurements. Growth of cultures was followed by periodically measuring optical density (OD) at 600 m,m with a Beckman spectrophotometer,

3 VOL. 96, 1968 EFFECT OF AUTOLYSIS ON MOLAR GROWTH 119 model DB, with 1-cm cuvettes. Bacterial cell mass was determined from OD measurements with the corresponding growth medium as the internal standard, or gravimetrically after drying washed cells to constant weight at 105 C. Gravimetrically determined dry weights were found to be essentially the same as those estimated from standard curves relating OD to dry weight, and OD of the media did not change detectably during growth of the bacteria, i.e., the changes (between 0 and unit) were beyond the sensitivity of the spectrophotometer. Logarithmically growing cells were used for determining standard curves relating OD to dry weight. The organisms were grown in appropriate media, and cells were separated by centrifugation at 4 C for 30 min in a refrigerated centrifuge at 8,000 X g. The cells were suspended and washed three times in volumes of distilled water equal to that of the original growth medium. Different amounts of the washed cells were suspended in distilled water, OD determinations were made (with distilled water as the internal standard), and a known volume (50 to 60 ml) of each dilution was placed in a preweighed dish. These samples were dried to constant weight at 105 C. OD values were plotted against dry weights (ug/ml), and regression analysis was used to determine the best fit lines. OD units corresponding to 1 mg (dry weight)/ml for S. diacetilactis DRCI and S. diacetilactis 18-16, S. faecalis 6783 grown in BEV and BEO media, S. cremoris K2E8, and S. lactis C2 were 3.75, 3.46, 2.61, 2.91, 4.36, and 3.16, respectively. The relationships between OD and dry weight were linear up to an OD of 0.6 to 0.8. Chemical methods. Arginine was determined colorimetrically by the method of Rosenberg et al. (26). Neither citrulline nor ornithine, nor a mixture of all E 0 (0 0 20L of the amino acids used except arginine, interfered with the method. Citrulline was determined by the Archibald method as modified by Oginsky (23). Neither ornithine nor any of the amino acids used interfered with this method. Glucose was determined colorimetrically with a glucostat (Worthington Biochemical Corp., Freehold, N.J.) by the procedure of Chayken (4). Reducing carbohydrates were determined colorimetrically as described by Nelson (22). Lactic acid was determined by the colorimetric method of Barker and Summerson (1). RESULTS Influence of autolysis on molar growth yields. Molar growth yields for S. diacetilactis, with limiting concentrations of glucose, were found erratic and dependent on the time of harvesting the cells. Changes in OD during growth and subsequent lysis in the partially defined media with limiting concentrations of glucose are shown for S. diacetilactis DRCI and S. faecalis in Fig. 1. A Y (glucose) value calculated for S. diacetilactis after incubation for 6 hr was 47% less than a value calculated at the peak of maximal growth. A value calculated for S. faecalis after 15 hr of incubation was 33% less than that calculated at the peak. Similar results were obtained with S. diacetilactis growing in the complex medium, i.e., the Y (glucose) calculated after incubation for 38 hr was 44% less than a value calculated at the peak of maximal growth. Subsequently, molar growth yields with batch cultures were determined at the peaks of maximal 0.10 Im O5 hmolssau 3smoles/ml Streptococcus diocetilactis DRCI Strestococcus faecalis / HOURS Changes in optical density (OD) ofstreptococcus diacetilactis DRCI and S. faecalis 6783 during growth FIG. 1. and subsequent lysis. The organisms were grown and held at 32 C and 37 C, respectively, in the partially defined media containing the indicated amounts of glucose. S. faecalis was grown anaerobically at ph 6.8. Symbols: (0) BEO medium; (0) BEV medium (S. faecalis) and HCM medium (S. diacetilactis).

4 120 MOUSTAFA AND COLLINS J. BACTERIOL. growth. There was a linear relationship between dry weight of cells and concentration of the energy source in all experiments, up through a concentration of 5,moles/ml for the media containing "vitamin-free" casein (HCM and BEV media) and 10 umoles/ml for the BEO medium which contained Oxoid casein (Fig. 2). There was slight growth of S. diacetilactis DRCI in the complex medium and of S. faecalis in the BEO medium (without an additional energy source), and the straight-line curves relating dry weight to glucose concentration passed slightly above the origin. However, in HCM and BEV media, there was no growth of S. diacetilactis or S. faecalis without an additional energy source (Fig. 2). Molar growth yields for S. diacetilactis DRCI. Table 1 shows yields for S. diacetilactis DRCI in the HCM medium with limiting concentrations of four energy sources. Yields determined with batch cultures were not appreciably different from yields determined in a continuous culture. However, assuming conversion of the sugars to lactic acid via the hexose diphosphate pathway, the results indicated values for Y (ATP) that are 50 to 70% higher than the generally accepted value of 10.5 (2). The high yields could be the result of one or more of the following possibilities. (i) The organism was capable of utilizing for energy some compound(s) in the medium besides the intended energy source, e.g., arginine (2). (ii) The organism accomplished some energy-yielding side-reaction(s) that provided 1 extra mole of ATP per mole of hexose utilized, i.e., 3 moles of ATP per mole of glucose or galactose and 6 moles of ATP per mole of lactose or maltose. (iii) There is no Y (ATP) value that is constant for all organisms (14, 15). (iv) The value of 10.5 for Y (ATP) is erroneously low. Absence of energy sources in the basal medium. The possibility that S. diacetilactis DRCI could utilize arginine or some other constituent of the basal medium as an extra source of energy was investigated by using a continuous culture. Since S. faecalis requires a small amount of fermentable carbohydrate (2 to 3,umoles of glucose/ml) for utilization of arginine as a source of energy in synthetic media (2, 7, 8), we included a small amount of glucose (2,umoles/ml) in the HCM medium (ph 6.5). When the steady-state population had been reached, the medium was replaced by a new reservoir of the same medium plus arginine (9.0,umoles/ml). The OD of the overflow did not increase, which indicates that energy was not obtained from the arginine; but the ph increased to 6.8, indicating that this strain of S. diacetilactis can hydrolyze arginine to citrulline. These results were substantiated by finding that in batch cultures the organism did not produce ammonia from citrulline but did produce ammonia from arginine. The casein hydrolysate of the HCM medium was replaced with a mixture of amino acids. With this modified medium, Y (glucose) was the same as it had been with the medium containing casein hydrolysate. The organism was found to have an absolute requirement for arginine. The modified SUBSTRATE (jumoies/mi) FIG. 2. Yields of Streptococcus diacetilactis DRCI and S. faecalis 6783 with the indicated amounts of the energy sources. S. diacetilactis was grown at 32 C in quantities of HCM medium at ph 7.1. S. faecalis was grown anaerobically at 37 C in quantities ofbeo (a) and BEV (@) media at ph 6.8. Dry weight was determined from standard curves relating OD to dry weight. Symbols: (circles) glucose; (o), galactose; (V) maltose.

5 VOL. 96, 1968 EFFECT OF AUTOLYSIS ON MOLAR GROWTH 121 medium contained 0.48,umoles of arginine per ml, but this small amount could not account for the high cell yields. Further, variations of 1 to 5,umoles/ml in the concentrations of glucose, arginine, or both, did not appreciably alter the value obtained for Y (glucose). The fact that yields were constantly high over the wide variations in the concentration of glucose also eliminated the possibility that some other constituent of the chemically defined medium served as a source of energy. Energy not obtained from side reactions. The possibility that S. diacetilactis DRCI could gain 3 moles of ATP per mole of glucose or galactose metabolized was investigated and eliminated. The use of strictly anaerobic conditions in a continuous culture and in batch cultures did not decrease the yield of cells (dry weight) per mole of glucose. The lactic acid yield was greater than 80% of the theoretical amount. Cell density in a continuous culture in air did not increase upon addition to the medium of pyruvate (6.8,moles/ ml), lipoate (3.2 mg/100 ml), or citrate (6.8,umoles/ml). Further, additional species of lactic acid bacteria that we tested gave cell yields that were not appreciably different from those for S. diacetilactis (Table 2; Fig. 2), i.e., if yields were determined at the peaks of maximal growth. Autolysis started immediately thereafter for each of these species, including S. faecalis (Fig. 1). TABLE 1. Molar growth yields for Streptococcus diacetilactis DRCJa Energy source Batch cultures Glucose... Galactose... Maltose... Maltose... Continuous cultures (D = 0.46 hr-') Glucose... Lactose... Concnb (AM) Y (substrate) (g, dry wt/ mole) and and ATP/ mole of substratec (moles) a The organism was grown at 32 C in HCM medium with limiting concentrations of energy sources. Dry weight was determined from a standard curve relating OD to dry weight (batch cultures) or gravimetrically (continuous cultures). b Chemical analysis showed that all of each of the energy sources had disappeared when growth stopped. c Assumed TABLE 2. Cell yields of lactic acid bacteria growing on glucose in partially defined caseinhydrolysate mediaa Y (glucose) Y (ATP)b Organism MNedium (g, dry wt/ (g, dry wt/ mole) mole) Streptococcus diacetilactis HCM S. cremoris K2E8... HCM S. lactis C2... HCM S. faecalis 6783c... BEV S. faecalis 6783C... BEO Batch cultures. Dry weight was determined from standard curves relating OD to dry weight. b Assumed: 2 moles of ATP per mole of glucose. c Strictly anaerobic conditions. Y (arginine) for S..faecalis It was necessary to redetermine Y (arginine) for S. faecalis. An experiment with BEO medium indicated that growth yields from glucose in continuous culture were the same for variations of the dilution rate from to 0.59 hr-' (Fig. 3). However, growth yields from glucose plus arginine varied with the dilution rate, with a maximum at D = 0.25 hr-' (Fig. 3). S. faecalis stopped growing when arginine was substituted in place of glucose plus arginine (Fig. 4), indicating that a small amount of glucose is necessary for the utilization of arginine as a source of energy. Our experiment for determining Y (arginine) was similar to the experiment of Bauchop and Elsden (2), modified as follows. We used four levels of arginine (2.87, 6.05, 9.10, and 12.26,umoles/ml) and determined residual energy sources at each steady state. Our dilution rate was 0.24 to 0.25 hr-1, instead of hr-'. The results (Fig. 5; Table 3) indicate that increases in cell density beyond the first steady state resulted from utilization of the added arginine. The increase in cell density from to 285,ug/ml was not proportional to the increase in arginine, indicating that some nutrient other than arginine became the growth-limiting factor, and that data obtained with concentrations of arginine higher than 9.1,moles/ml should not be used in calculating Y (arginine). The average of apparent Y (arginine) values corresponding to concentrations of arginine up to 9.1,umoles/ml (Table 3) is This value is erroneous, however, since it was calculated without correcting for the amounts of arginine not utilized or converted only to citrulline. The average of corrected Y (arginine) values for concentrations of arginine up to 9.1,umoles/ ml is The same experiment was repeated with BEV medium. Similar results were obtained,

6 122 MOUSTAFA AND COLLINS J. BACTERIOL. except that the nonlimiting concentration of arginine was lower (slightly above 4.83,moles/ ml). The apparent Y (arginine) value for concentrations up to 4.83,umoles/ml was 15.5, and the corrected Y (arginine) value was DISCUSSION Autolysis after the occurrence of maximal bacterial populations can have pronounced influences on the estimations and subsequent interpretations of molar growth yields determined from batch cultures. Lysis has been observed after the occurrence of maximal population with various bacteria grown on limiting energy sources (12, 20, 21, 29). We have studied and explained this autolysis with regard to S. diacetilactis, and have observed that it does not occur with cells grown on galactose, lactose, maltose, or at 17 C on glucose (21). Our results indicated that if chloramphenicol was added to logarithmically growing cells to stop them from growing prior to the depletion of glucose, the cells would be susceptible to lysis (21). Thus, it seemed plausible that some lysis might occur simultaneously with growth during the logarithmic phase and influence cell yields. However, a comparison of the yields determined in this study (Table 1) reveals that lysis of cells during the logarithmic growth phase is minimal. For S. diacetilactis growing logarithmically in a continuous culture, the Y (ATP) calculated from growth on lactose, on which lysis did not occur (21), was less than 1% greater than the Y (ATP) calculated from Y (glucose). I >' 200 -J C) I- CD 3d D - VO DILUTION RATE, D (hr-l) FIG. 3. Influence of dilution rate, D (hr-'), on Y (glucose) and Y (arginine) with Streptococcus faecalis 6783 in continuous culture. The organism was grown anaerobically at 37 C in BEO medium at ph 6.6. Dry weight was determined gravimetrically. Symbols: (0), 2.3,moles ofglucose per ml; (0), 2.3,moles of glucose plus 9.0,umoles of arginine per ml. 0-6 I ~~~(3) 200 // -J -j o o (2) HOURS FIG. 4. Influence of glucose on the utilization of arginine as a source ofenergy by Streptococcus faecalis The organism was grown anaerobically in continuous culture at 37 C in BEO medium, ph 6.6, containing 2.3,umoles of glucose per ml. At point 2, the medium beingfed was replaced with original medium plus 9.0,umoles of arginine per ml. At point 3, the medium was replaced with medium containing no glucose and 9.0,umoles of arginine per ml. Dry weight was determined from a standard curve relating OD to dry weight. 400 E 3oo 5200 _ (5) ~ HOURS FIG. 5. Growth of Streptococcus faecalis 6783 in continuous culture for determining Y (arginine). The organism was grown anaerobically at 37 C in BEO medium, ph 6.6, containing 2.3 pumoles of glucose per ml. At points 2, 3, 4, 5, and 6, the medium beingfed was replaced with quantities of the original medium plus 2.87, 6.05, 9.10, 12.26, and 0,umoles ofarginine per ml, respectively. Dry weight wasdeterminedgravimetrically. Apparently, opportunity for autolysis was greater in batch cultures than in continuous cultures, since in batch cultures the cell yield from galactose, on which lysis did not occur (21), was 8% greater than that from glucose. Despite this indication that lysis was somewhat greater in the batch cultures (though yields were determined at

7 VOL. 96, 1968 EFFECT OF AUTOLYSIS ON MOLAR GROWTH 123 TABLE 3. Data obtained in the determination of Y (arginine) for Streptococcus faecalis 6783a Substrate concn (mm) Dry wt of cells Y (arginine) (JAg/mi) (pg/,4mole) Steady A Steady Determination state state Argi- A Argi- A Argi- Glucose Argi- concn of concn of Ani-e rgi-e nine Steady A- Cr nine (arginine nnrginin added added uti- state A Cell parent rected +citrul- (acirgiinl dddade lized cell yield ralectaled line)b line) yield vle vle 1st Steady state Before growth After growth nd Steady state Before growth After growth rd Steady state Before growth After growth th Steady state Before growth After growth th Steady state Before growth After growth a Growth conditions are given in Fig. 5. Amounts of citrulline were determined by chemical analysis. c Apparent Y (arginine) = A cell yield/a arginine added. d Corrected Y (arginine) = A cell yield/a arginine utilized. maximal OD), the molar growth yield for glucose was 5 % greater in batch cultures than in a continuous culture, possibly attributable to faster growth and lower maintenance in the batch cultures. Results indicated that S. diacetilactis DRCI did not utilize arginine or other constituents of HCM medium as energy sources and that there were no energy-yielding side reactions in its metabolism of glucose. These results and data on the fermentation balance in the metabolism of glucose by this organism (unpublished data) suggest that the fermentation of hexoses was homolactic, i.e., S. diacetilactis gained only 2 moles of ATP per mole of glucose or other fermentable hexose. Lactose and maltose (2.0 to 5.0,umoles/ml) resulted in molar growth yields that were about double those obtained with hexoses, suggesting that the glycosidic bond energy is lost in the metabolism of these disaccharides by S. diacetilactis. It is assumed that this organism hydrolyzes lactose by means of a hydrolytic enzyme, f-d-galactosidase, similar to E. coli (19) and S. lactis (6), and that each mole of lactose yielded only 4 moles of ATP. S. diacetilactis apparently utilized for energy only half of the maltose molecule at low concentrations (0.5 to 1.5,umoles/ml). This is probably explained by our findings (21) which indicate that S. diacetilactis phosphorylyzes maltose with the formation of glucose and :-glucose-l-phosphate. The utilization of :-glucose-i-phosphate by S. diacetilactis at higher concentrations of maltose (2.0 to 5.0,umoles/ml) possibly involved hydrolysis of the ester to glucose and inorganic phosphate (at low ph and reduced inorganic phosphate content of the medium), since f- glucose-l-phosphate is not the usual substrate for phosphoglucomutase. Y (arginine) must equal Y (ATP), since arginine dihydrolase yields 1 mole of ATP per mole of arginine (17, 30). The average value for Y (ATP) which we determined for S. faecalis growing on arginine is This is not much different from the Y (ATP) values we determined for the bacteria growing on sugars (Tables 1 and 2), but about 70% greater than the value of 10.5 reported for the same strain of S. faecalis by Bauchop and Elsden (2). They substantiated an average value of 10.5 for Y (ATP) (range = 8.3 to 12.6, determined from batch cultures of three organisms with various substrates) with data indicating that Y (arginine) for S. faecalis in continuous culture was Our results indicate that growth yields can be influenced considerably if autolysis is permitted in batch cultures before the determinations of dry weight, or if concentrations of glucose added to BEO medium are greater than 10,moles/ml, or if both conditions exist. Also, we found that S. faecalis does not utilize arginine efficiently in continuous culture at

8 124 MOUSTAFA AND COLLINS J. BACTERIOL. the dilution rate (0.116 hr-) used by Bauchop and Elsden (2). The organism used arginine most efficiently at a dilution rate of about 0.25 hr-' (Fig. 3), and at this dilution rate there was unused arginine in the supernatant fluid at each steady state (Table 3). The apparent Y (arginine) value calculated from our data at a dilution rate of hr-1 without correction for unused arginine is The quantities of arginine and citrulline that accumulated in the medium during the fifth steady state (Fig. 5), when growth apparently was limited by some other nutrient, were smaller than expected (Table 3). Possibly, excess arginine and citrulline were hydrolyzed only to ornithine and carbamylphosphate, both of which accumulated. These compounds were not detectable with the methods used for determining arginine and citrulline. Other possibilities are that carbamylphosphate was formed and hydrolyzed to CO2 and NH3 through an uncoupled reaction (activation of a shunt pathway bypassing the energyyielding steps), or that ATP produced in the usual manner was itself hydrolyzed by an adenosine triphosphatase. These results are similar to findings of Rosenberger and Elsden (27), who found S. faecalis 6783 to continue producing lactic acid from glucose after growth was limited in a tryptophan-limiting medium. These findings emphasize the importance of establishing linearity between concentrations of the energy source and cell yields in determining molar growth yields. Results relating dry weight to glucose concentration determined with batch cultures in BEO medium gave a straight line that intercepts the Y axis above the origin (Fig. 2). This positive value is too great to be attributed to the arginine present in the basal medium, and suggests that the basal medium contained some unknown compound that served as a source of energy. In our study, the average Y (ATP) value, determined by using limiting concentrations of carbohydrates as sources of energy for homolactic acid bacteria, is 17.0 (range = 15.3 to 20.0). This value is close to the average Y (arginine) value of 17.8 determined with S. faecalis, but only about 53% of the theoretical value (33.3) calculated by Gunsalus and Shuster (10). There probably is a universally correct value for Y (ATP), but there are problems involved in making an accurate experimental determination of it. Bacteria growing with a limiting energy source direct a considerable amount of the energy they derive from catabolism of the substrate to maintenance, osmotic work, and other functions not concerned with synthesis of cell material. The amount of energy required for maintenance varies with growth rate and from one organism to another. (The apparent high yield of 20 g of dry weight per mole of ATP for S. lactis might indicate low maintenance for this organism.) Additionally, some of the energy source is incorporated into cell material. The homolactic acid bacteria used in our experiments have complex nutritional requirements and ferment glucose by the hexose diphosphate pathway; they use glucose almost exclusively as a source of energy. According to Harvey and Collins (13), S. diacetilactis that was growing with a limiting concentration of glucose incorporated only 3.3% of the metabolized glucose into cell material. Bauchop and Elsden (2) found S. faecalis to assimilate into cell material less than 1% of the glucose it metabolized. By plotting 1/observed yield versus 1/specific dilution rate, according to the method reported by Pirt (25), we determined that the Y (glucose) value for S. diacetilactis DRCI, corrected for maintenance, is This value, when corrected for the glucose incorporated into cell material (3.3%), becomes 43.4, corresponding to a corrected Y (ATP) value of This value is about 25% greater than the values determined experimentally in this study, but only about two-thirds of the theoretical value calculated by Gunsalus and Shuster (10). ACKNOWLEDGMENT This investigation was supported by Public Health Service research grant UI from the National Center for Urban and Industrial Health. LrrERATURE CITED 1. Barker, J. B., and W. H. Summerson The colorimetric determination of lactic acid in biological material. J. Biol. Chem. 138: Bauchop, T., and S. R. Elsden The growth of microorganisms in relation to their energy supply. J. Gen. Microbiol. 23: Beck, R. W., and L. R. Shugart Molar growth yields in Streptococcus faecalis var. liquefaciens. J. Bacteriol. 92: Chayken, S Biochemistry laboratory techniques, p John Wiley & Sons, Inc., New York. 5. Chen, S. L Energy requirement for microbial growth. Nature 202: Citti, J. E., W. E. Sandine, and P. R. Elliker ,3-Galactosidase of Streptococcus lactis. J. Bacteriol. 89: Deibel, R. H Utilization of arginine as an energy source for the growth of Streptococcus faecalis. J. Bacteriol. 87: Elsden, S. R., and J. L. Peel Metabolism of carbohydrates and related compounds. Ann. Rev. Microbiol. 12: Forrest, W. W., and D. J. Walker Synthesis of reserve materials for endogenous metabolism in Streptococcus faecalis. J. Bacteriol. 89:

9 VOL. 96, 1968 EFFECT OF AUTOLYSIS ON MOLAR GROWTH Gunsalus, I. C., and C. W. Shuster Energy yielding metabolism in bacteria, p In I. C. Gunsalus and R. Y. Stanier (ed.), The bacteria, vol. 2. Academic Press, Inc., New York. 11. Hadjipetrou, L. P., J. P. Gerrits, F. A. Teulings, and A. H. Stouthamer Relation between energy production and growth of Aerobacter aerogenes. J. Gen. Microbiol. 36: Hadjipetrou, L. P., and A. H. Stouthamer Autolysis of Bacillus subtilis by glucose depletion. Antonie Van Leeuwenhoek J. Microbiol. Serol. 29: Harvey, R. J., and E. B. Collins Roles of citrate and acetoin in the metabolism of Streptococcus diacetilactis. J. Bacteriol. 86: Hobson, P. N Continuous culture of some anaerobic and facultatively anaerobic rumen bacteria. J. Gen. Microbiol. 38: Hobson, P. N., and R. Summers The continuous culture of anaerobic bacteria. J. Gen. Microbiol. 47: Hungate, R. E Polysaccharide storage and growth efficiency in Ruminococcus albus. J. Bacteriol. 86: Knivett, V. A Phosphorylation coupled with anaerobic breakdown of citrulline. Biochem. J. 56: Larsen, D. H., and R. L. Dimmick Attachment and growth of bacteria on surfaces of continuous-culture vessels. J. Bacteriol. 88: Lederberg, J The beta-d-galactosidase of Escherichia coli, strain K-12. J. Bacteriol. 60: Monod, J Sur un phenomene de lyse lie a l'inanition carbonie. Ann. Inst. Pasteur 68: Moustafa, H. H., and E. B. Collins Role of galactose or glucose-l-phosphate in preventing the lysis of Streptococcus diacefilactis. J. Bacteriol. 95: Nelson, N A photometric adaptation of the Somogyi method for the determination of glucose. J. Biol. Chem. 153: Oginsky, E. L Isolation and determination of arginine and citrulline. II. Citrulline, p In S. P. Colowick and N. 0. Kaplan (ed.), Methods in enzymology, vol. 3. Academic Press, Inc., New York. 24. Oxenbrugh, M. S., and A. M. Snoswell Use of molar growth yields for the evaluation of energy-producing pathways in Lactobacillus plantarum. J. Bacteriol. 89: Pirt, S. J The maintenance energy of bacteria in growing cultures. Proc. Roy. Soc. (London) Ser. B. 163: Rosenberg, H., A. H. Ennor, and J. F. Morrison The estimation of arginine. Biochem. J. 63: Rosenberger, R. F., and S. R. Elsden The yields of Streptococcus faecalis grown in continuous culture. J. Gen. Microbiol. 22: Senez, J. C Some considerations on the energetics of bacterial growth. Bacteriol. Rev. 26: Shockman, G. D., M. J. Conover, J. J. Kolb, P. M. Phillips, L. S. Riley, and G. Toennies Lysis of Streptococcus faecalis. J. Bacteriol. 81: Slade, H. D., C. C. Doughty, and W. C. Slamp The synthesis of high-energy phosphate in the citrulline ureidase reaction by soluble enzymes of Pseudomonas. Arch. Biochem. Biophys. 48: Stouthamer, A. H Energy production in Gluconobacter liquefaciens. Biochim. Biophys. Acta 56: Umbreit, W. W., R. H. Burris, and J. F. Stauffer Manometric techniques, p. 69. Burgess Publishing Co., Minneapolis.