Megasphaera elsdenii Grown on Glucose or Lactate

Transcription

1 APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Jan. 1993, p /93/1255-5$2./ Copyright 1993, American Society for Microbiology Vol. 59, No. 1 Presence of Lactate Dehydrogenase and Lactate Racemase in Megasphaera elsdenii Grown on Glucose or Lactate TSUNEO HINO* AND SHINJI KURODA Department ofagriculture, Meiji University, Higashimita, Tama-ku, Kawasaki 214, Japan Received 3 July 1992/Accepted 25 October 1992 Activity of D-lactate dehydrogenase (D-LDH) was shown not only in cell extracts from Megasphaera elsdenii grown on DL-lactate, but also in cell extracts from glucose-grown cells, although glucose-grown cells contained approximately half as much D-LDH as DL-lactate-grown cells. This indicates that the D-LDH of M. elsdenii is a constitutive enzyme. However, lactate racemase (LR) activity was present in DL-lactate-grown cells, but was not detected in glucose-grown cells, suggesting that LR is induced by lactate. Acetate, propionate, and butyrate were produced similarly from both D- and L-lactate, indicating that LR can be induced by both D- and L-lactate. These results suggest that the primary reason for the inability ofm. elsdenii to produce propionate from glucose is that cells fermenting glucose do not synthesize LR, which is induced by lactate. The production of propionate, a glucogenic substance, in the rumen is particularly important for ruminant physiology and nutrition. Megasphaera elsdenii is one of the main propionate producers in the rumen, especially when animals are fed high-grain diets (12). This bacterium produces propionate from lactate which is produced by other bacteria (1, 12, 2). It has been reported that M. elsdenii is the most important organism that contributes to the utilization of lactate in the rumen (8), possibly playing an important role in the prevention of rumen acidosis (2). M. elsdenii is very unique in that it produces propionate from lactate but does not form propionate from glucose (15). This is in contrast to most of other bacteria producing propionate from glucose, e.g., Selenomonas ruminantium, Succinimonas amylolytica, and Propionibacterium acnes (12, 16). Veillonella alcalescens forms propionate from lactate, but does not ferment sugars (12, 18). Our interest was centered on why M. elsdenii does not produce propionate from glucose. The fermentation pathways of glucose and lactate in M. elsdenii are shown in Fig. 1, based on the knowledge reported so far (2, 3, 5, 22). First, it seems reasonable to assume that the reaction from pyruvate to D-lactate, a branching point, is a key to this problem. M. elsdenii has NAD-independent D-lactate dehydrogenase (id-ldh) catalizing the reaction from D-lactate to pyruvate (6). This type of LDH is known to be present in a variety of bacteria and usually converts lactate to pyruvate with no reverse reaction reported in intact cells (9). On the contrary, NAD-dependent LDH (d-ldh) forms lactate from pyruvate, and there is no evidence that it functions in the opposite direction in vivo (9). Theoretically, however, the LDH reaction itself must be reversible because it is an oxidation-reduction reaction. In fact, the conversion of pyruvate to lactate was demonstrated in a cell extract from M. elsdenii (13). The direction basically depends on the ratio of [pyruvate] [electron donor] to [lactate] [electron acceptor], where electron donor and acceptor mean reduced and oxidized forms of electron carrier, respectively. Even though the equilibrium position of the id-ldh reaction in * Corresponding author. 255 M. elsdenii strongly favors pyruvate formation, the reverse reaction should occur if the ratio of pyruvate to D-lactate becomes high enough under certain conditions. Such a situation may be brought about if D-lactate is readily metabolized to propionate (Fig. 1). We made an assumption that the metabolism from D-lactate to propionate, or the id-ldh reaction itself, is suppressed when M. elsdenii is fermenting glucose. As the first approach to this problem, we focused on id-ldh and lactate racemase (LR), which catalyzes the reaction after the id-ldh reaction. MATERIALS AND METHODS Organism and culture conditions. M. elsdenii NIAH-112, donated by H. Minato (National Institute of Animal Health, Tukuba, Japan), was grown as described previously (11). The growth medium contained (grams per liter): KH2PO4,.9; K2HPO4,.9; (NH4)2S4, 1.8; NaCl, 1.8; MgSO4. 7H2,.83; CaCl2. 2H2,.24; Trypticase (BBL Microbiology Systems, Cockeysville, Md.), 1.; yeast extract (Difco), 1.; cysteine-hcl,.6; and an energy substrate. Cells were cultured in 12-ml serum vials (11), and culture was terminated at the end of exponential growth. When we collected bacterial cells, 15 mm glucose or 9 mm DL-lactate was added, and 1-liter culture bottles were used. Culture was terminated at the end of log phase before the growth rate decelerated. Final cell numbers (A6w) were almost equal between the cultures with 15 mm glucose and 9 mm DLlactate. The cultures were cooled in an ice bath, and the cells were collected by centrifugation (1, x g, 15 min) at 4 C. The cell pellet was stored under N2 at -8 C and subjected to experiments within a few days. Extraction of cells and preparation of crude enzyme solution. Frozen cells (.5 to 1 g [dry weight]) were thawed out overnight in a refrigerator (4 C) and then suspended in ice-cold 1 mm potassium phosphate buffer (KP1, ph 6.) containing 1 mm dithiothreitol. All the treatments described below were performed at or 4 C. The cells were disrupted by ultrasonication (Ultrasonifier model US-15; Nippon Seiki Ltd.) at maximum power output for 1 min (actual working time, 3% duty cycle). Microscopic examination

2 256 HINO AND KURODA IGlucose IL_Lactatel L. c.lacty-coa a jatp? LR I Pyruvate.. FD Lactate Acrylyl-Co A I D-LDH 4 Acetyl-CoA Acrylatel Propionyl-CoA KATP+\ Acetate Butyrate Propionate FIG. 1. Possible metabolic pathways from glucose, lactate, and acrylate to main VFA in M. elsdenii. CoA, coenzyme A. I TABLE 1. Production of main VFA from glucose, D-lactate, L-lactate, and acrylate by M. elsdenii Substrate (mm) APPL. ENVIRON. MICROBIOL. VFA production (mm) Acetate Propionate Butyrate Glucose (5) Glucose (5) + acrylate (2) D-Lactate (3) L-Lactate (3) L-Lactate (3) + acrylate (2) indicated that more than 95% of the cells were broken. After cell debris was removed by centrifugation (15, x g, 15 min), 2.7 M (NH4)2SO4 (final concentration) was added to the supernatant and left for 2 h. The precipitate from centrifugation (1, x g, 15 min) was dissolved in an appropriate volume of 1 mm KPj (ph 6.) without dithiothreitol and used as a crude enzyme solution. Assay for id-ldh. Activity of id-ldh was determined by observing the A42 changed by the reduction of ferricyanide by D-lactate (6, 17). The assay mixture contained 1 mm K3Fe(CN)6,.2 mg of bovine serum albumin per ml, 125 mm sodium D-lactate (Sigma), and.1 ml of crude enzyme solution per ml in 1 mm KP1 (ph 7.). The reaction was initiated by adding substrate (D-lactate) after the absorbance had become constant. Assays were performed at 23 to 25 C by continuously monitoring the absorbance. Assay for LR. LR activity was estimated by the following two methods. The first method is to link the LR reaction to the id-ldh reaction (Fig. 2A), i.e., L-lactate is first racemized to D-lactate and the conversion of the D-lactate formed to pyruvate is coupled with the reduction of dichlorophenol indophenol (DCIP), a much more sensitive electron-accepting indicator than ferricyanide (6). This reaction is observed by measuring the A6. If the cell extract contains both enzymes, DCIP should be reduced. In this method, 125 mm sodium L-lactate (Sigma) and 6,uM DCIP were used instead of D-lactate and K3Fe(CN)6, respectively. Other reagents and assay conditions were the same as described above. If no or little id-ldh is included in the sample, however, it is impossible to estimate LR activity by the above method. For this reason, the second method was developed as follows. D-Lactate formed from L-lactate was determined by using a commercial preparation of dd-ldh in which id- LDH, if present, was inhibited by Cu2+ as described below (Fig. 2B). The overall reaction was observed by measuring the reduction of NAD at 34 nm. In this case, the assay mixture contained 4,uM NAD, 125 mm L-lactate, 2 U of dd-ldh (EC , from Leuconostoc mesenteroides; Oriental Yeast Ltd., Tokyo, Japan) per ml, and.1 ml of (A) LR i D- LDH L-Lactate- --D-Lactate - Pyruvate DCI P(ox) DCI P(red) (B) id-ldh LR, 'Cu2~' *2"'A L- Lactate -.D-Lactate Pyruvate FIG. 2. NAD dd-ldh Reactions used for the estimation of LR activity. crude enzyme solution per ml in 1 mm KP1 (ph 7.2). The reaction was initiated by adding L-lactate. Separation of id-ldh and LR by DEAE-cellulose column chromatography. id-ldh and LR were separated by the methods of Brockman and Wood (6) and Olson and Massey (17) with slight modifications. The crude enzyme solution (equivalent to ca. 2 g [dry weight] of cells) was applied to a 45-ml DEAE-cellulose column (1.8 by 3 cm) equilibrated with 1 mm KPi (ph 6., without dithiothreitol). The enzymes were then eluted with increasing concentrations of KP1 in a stepwise manner, i.e., 1 ml each of 1, 3, and 6 mm KPi (ph 6., with 1 mm dithiothreitol). The eluate was collected in 1-ml fractions, and id-ldh activity was determined by DCIP reduction after the addition of 6,uM ZnCl2. LR activity was measured by the NAD reduction method described above. Proteins for molecular weight markers were purchased from Sigma Chemical Co.: carbonic anhydrase (29,); chicken egg albumin (45,); bovine serum albumin, monomer (66,); dimer (132,); and alcohol dehydrogenase from yeasts (15,). Determination of VFA and protein. Volatile fatty acids (VFA) were determined by gas chromatography (1). Protein content in the crude enzyme solution was assayed by using Coomassie brilliant blue (4). RESULTS VFA production from glucose, D-lactate, L-lactate, and acrylate. In earlier studies, the extract of M. elsdenii grown on lactate converted acrylate to acetate, propionate, and butyrate (13, 14). In our study, however, M. elsdenii did not grow on acrylate, indicating that at least the strain used in this study has an extremely low capacity to generate ATP from acrylate, i.e., to oxidize acrylate to acetate and butyrate (Fig. 1). On the other hand, when acrylate was given to M. elsdenii with glucose, acrylate was metabolized exclusively to propionate (Table 1). The cells fermenting L-lactate metabolized acrylate to mainly propionate with much smaller amounts of acetate and butyrate (Table 1). These results suggest that the cells fermenting glucose are devoid of the enzyme(s) needed to produce propionate, in the pathway between pyruvate and acrylyl-coenzyme A (Fig. 1). It appears to be most likely that such cells lack id-ldh and/or LR. Activity of id-ldh in crude enzyme solution from cells grown on glucose or DL-lactate. As shown in Fig. 3, both the cells grown on glucose and those grown on DL-lactate had id-ldh activity, the initial velocity of the former sample being nearly half of that of the latter. Since approximately equal amounts of cells were used to prepare extracts and the protein contents of the two extracts were almost equal, it is evident that the lactate-grown cells contained nearly twice as much id-ldh as the glucose-grown cells. Figure 4 shows

3 VOL. 59, 1993 LACTATE DEHYDROGENASE AND RACEMASE IN M. ELSDENII 257.5\ 5 1 FIG. 3. id-ldh activity in the crude enzyme solution from M. elsdenii grown on glucose (curve b) or DL-lactate (curve c), determined by ferricyanide reduction. Curve a, control (without D-lactate). The arrow indicates the addition of D-lactate. the effects of metal ions on id-ldh activity. Addition of Zn2+ slightly stimulated the reaction, which is consistent with the fact that Zn2+ is a cofactor of this enzyme (6). The complete inhibition of activity by EDTA may be due to the removal of Zn2+ from the enzyme. When Zn2+ was replaced by Mn2+ or Cu2' added in excess, enzyme activity was completely lost, whereas Co2' and Mg2" partially inactivated the enzyme. These results confirm that the reduction of ferricyanide is catalyzed by id-ldh (6). The effects of metal ions were exactly the same between the glucose- and lactate-grown cells, suggesting that M. elsdenii synthesizes identical id-ldh irrespective of the fermentation substrate. Further confirmation was made by examining the substrate specificity and response to inhibitors (Table 2), with results essentially similar to those reported previously (6, 9). Activity of LR in cells grown on glucose or DL-lactate. Figure 5 shows LR activity measured by linking to the id-ldh reaction. The extract from lactate-grown cells reduced DCIP when L-lactate was added as a substrate, clearly demonstrating the presence of LR (Fig. SB). The velocity of the reaction was lower with L-lactate than with D-lactate, indicating that the LR reaction is rate limiting. On the other hand, no LR activity was detected in glucose-grown cells (Fig. SA). As described above, id-ldh of M. elsdenii was completely inactivated by.1 mm CUSO4, but the dd-ldh was not pre_paration inactivated by this concentration of Cu +. Figure 6 shows LR activity measured by coupling to the dd-ldh reaction in the presence of.1 mm CuS4. TABLE 2. Substrate specificity of id-ldh from M. elsdenii and effect of some inhibitors on id-ldh activity" Substrate Inhibitor Relative activity of id-ldhb (125 mm) (mm) DL-Lactatec Glucosec D-Lactate 1 41 DL-Malate 2 DL-2-HBA DL-3-HBA 5 2 D-Lactate MIA (1) D-Lactate Oxalate (2) 96 4 D-Lactate Oxamate (2) 94 4 D-Lactate Glyoxylate (2) 4 3 aabbreviations: HBA, hydroxybutyric acid; MIA, monoiodoacetic acid. b Based on AA42 minute-' milligram of proteinm. c Cells grown on DL-lactate or glucose. Results essentially similar to the above-mentioned results were obtained, confirming that lactate-grown cells contained LR but glucose-grown cells did not possess LR to any detectable degree. Separation of id-ldh and LR by DEAE-cellulose column chromatography. Figure 7 shows the elution pattern of id-ldh and LR, in which the id-ldh peak is seen in both the glucose- and lactate-grown cells. Since the two samples were prepared from an equal amount of cells and protein content was nearly equal between the two samples, the glucose-grown cells must have contained about half as much id-ldh as the lactate-grown cells. On the other hand, the LR peak emerged when the sample from lactate-grown cells was applied, but with glucose-grown cells, no peak was detected corresponding to the LR peak. Measurement of the activity with the partially purified enzymes excludes to some extent the possibility that the glucose-grown cells contained a substance(s) inhibiting enzyme activity. From the elution pattern, it could be deduced that the molecular weights of id-ldh and LR are ca. 1, and 5,, respectively. The molecular weight of id-ldh from M. elsdenii has been reported to be 15, (17), but no report is available on LR. These results confirm that cells grown on glucose do not possess LR, although they contain id-ldh. DISCUSSION It was revealed that LR is not synthesized by M. elsdenii growing on glucose, providing a primary answer to the 1..- a FIG. 4. Effect of metal ions on id-ldh activity in the crude enzyme solution from cells grown on DL-lactate (A) or glucose (B). Curves: a, 1 mm EDTA; b, control (without addition of a metal ion); c,.1 mm ZnSO4; d,.1 mm CuSO4; e,.1 mm MnSO4; f,.1 mm MgSO4; g,.1 mm CoS4. O.5 a -B _& 5 FIG. 5. LR activity in the crude enzyme solution from cells grown on glucose (A) or DL-lactate (B), measured by DCIP reduction. Substrate: curve a, L-lactate; curve b, D-lactate. 1

4 258 HINO AND KURODA APPL. ENvIRON. MICROBIOL. It - b3 FIG. 6. LR activity in the crude enzyme solution from cells grown on glucose (curve c) or DL-lactate (curve b), measured by NAD reduction by dd-ldh in the presence of.1 mm CuS4. Curve a, positive control (2 mm D-lactate). question why propionate is not formed from glucose. It appears that LR is induced by its substrate, lactate. Since VFA production was the same whether D- or L-lactate was the substrate, i.e., propionate was produced from D-lactate, and acetate and butyrate were formed from L-lactate (Table 1), LR must be induced by both D- and L-lactate. Fermentation of D-lactate was similar to that of L-lactate, indicating that the LR reaction is readily reversible (Table 1). On the other hand, id-ldh was synthesized in large amounts, even when the organism was fermenting glucose. This is rather strange, because the enzyme does not actually function under this condition. Possibly, id-ldh of M. elsdenii is a constitutive enzyme, and pyruvate may stimulate the synthesis of the enzyme. This is in contrast with many other bacteria in which LDH is induced by lactate (9). Even though id-ldh is present, virtually no D-lactate must be formed from pyruvate unless LR is simultaneously present, since the equilibrium for the id-ldh reaction appears to be far to the formation of pyruvate. Thus, as D-lactate is not formed from glucose, LR should not be synthesized in the cells growing on glucose. When acrylate was added in thke presence of glucose, presumably L-lactate was not accumulated to the level at which LR is induced: substantialu' all the acrylate was metabolized to propionate (Table 1), possibly due to lack of LR. Even when acrylate was provided with L-lactate, most of the acrylate was metabolized to propionate, indicating that backward reac- -j E >.E ( :) to ' cc J >1 E u -1."e a, E It _.5 - W ix Fraction number FIG. 7. Elution pattern of id-ldh and LR from DEAE-cellulose column. (A) Glucose-grown cells; (B) DL-lactate-grown cells. The eluate was collected in fractions after the void volume of buffer solution had been discarded. Activities of id-ldh and LR were measured by the reduction of DCIP and NAD, respectively. tions from acrylyl-coenzyme A toward L-lactate hardly proceed. There may be other reasons why the id-ldh reaction does not proceed from pyruvate to D-lactate in vivo, but evidently the lack of LR synthesis in the cells fermenting glucose provides the most decisive explanation for the fact that propionate is not produced from glucose. In the rumen, M. elsdenii appears to occupy an ecological niche as a lactate utilizer (12, 2) and to be valuable as an organism that produces propionate from lactate. Since M. elsdenii is not subject to catabolite repression by glucose or maltose, its contribution to lactate catabolism is considered to increase after the feeding of readily fermentable carbohydrate which represses lactate utilization by Selenomonas species and other lactate utilizers (7, 19, 21). REFERENCES 1. Baldwin, R. L Pathways of carbohydrate metabolism, p In R. W. Dougherty (ed.), Physiology of digestion in the ruminant. Butterworths, London. 2. Baldwin, R. L., and L. P. Milligan Electron transport in Peptostreptococcus elsdenii. Biochim. Biophys. Acta 92: Baldwin, R. L., W. A. Wood, and R. S. Emery Lactate metabolism by Peptostreptococcus elsdenii: evidence for lactyl coenzyme A dehydrase. Biochim. Biophys. Acta 97: Bradford, M. M A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72: Brockman, H. L., and W. A. Wood Electron-transferring flavoprotein of Peptostreptococcus elsdenii that functions in the reduction of acrylyl-coenzyme A. J. Bacteriol. 124: Brockman, H. L., and W. A. Wood D-Lactate dehydrogenase of Peptostreptococcus elsdenii. J. Bacteriol. 124: Counotte, G. H. M., A. Lankhorst, and R. A. Prins Role of DL-lactate as an intermediate in rumen metabolism of dairy cows. J. Anim. Sci. 56: Counotte, G. H. M., R. A. Prins, R. H. A. M. Janssen, and M. J. A. DeBie Role of Megasphaera elsdenii in the fermentation of DL-[2-'3C]lactate in the rumen of cattle. Appi. Environ. Microbiol. 42: Garvie, E. I Bacterial lactate dehydrogenases. Microbiol. Rev. 44: Hino, T Action of monensin on rumen protozoa. Jpn. J. Zootech. Sci. 53: Hino, T., K. Miyazaki, and S. Kuroda Role of extracellular acetate in the fermentation of glucose by a ruminal bacterium, Megasphaera elsdenii. J. Gen. Appl. Microbiol. 37: Hungate, R. E The rumen and Its microbes, p Academic Press, Inc., New York. 13. Ladd, J. N., and D. J. Walker The fermentation of lactate and acrylate by the rumen micro-organism LC. Biochem. J. 71: Lewis, D., and S. R. Elsden The fermentation of L-threonine, L-serine, L-cysteine and acrylic acid by a gram-negative coccus. Biochem. J. 6: Marounek, M., K. Fliegrova, and S. Bartos Metabolism and some characteristics of ruminal strains of Megasphaera elsdenii. Appl. Environ. Microbiol. 55: Moore, W. F. C., and E. P. Cato Validity of Propionibacterium acnes (Gilchrist) Douglas and Gunter comb. nov. J. Bacteriol. 85: Olson, S. T., and V. Massey Purification and properties of the flavoenzyme D-lactate dehydrogenase from Megasphaera elsdenii. Biochemistry 18: Rogosa, M The genus Veillonella. I. General cultural, ecological, and biochemical considerations. J. Bacteriol. 87: Russell, J. B., and R. L. Baldwin Substrate preferences in

5 VOL. 59, 1993 LACTATE DEHYDROGENASE AND RACEMASE IN M. ELSDENII 259 rumen bacteria: evidence of catabolite regulatory mechanisms. Appl. Environ. Microbiol. 36: Russell, J. B., M. A. Cotta, and D. B. Dombrowski Rumen bacterial competition in continuous culture: Streptococcus bovis versus Megasphaera elsdenii. Appl. Environ. Microbiol. 41: Stewart, C. S., and M. P. Bryant The rumen bacteria, p In P. N. Hobson (ed.), The rumen microbial ecosystem. Elsevier Applied Science, London. 22. Tung, K. K., and W. A. Wood Purification, new assay, and properties of coenzyme A transferase from Peptostreptococcus elsdenii. J. Bacteriol. 124: