Vol. 125, No. 1 Printed in U.S.A.

JOURNAL OF BACTERIOLOGY, Jan. 1976, p. 14-18 Copyright 0 1976 American Society for Microbiology

Microcalorimetric Study of the Anaerobic Growth of Escherichia coli: Growth Thermograms in a Synthetic Medium A. BELAICH

AND

J. P. BELAICH*

Laboratoire de Chimie Bacterienne, C.N.R.S., 13274 Marseille Cedex 2, France Received for publication 29 July 1975

A microcalorimetric technique was used for studying the growth of Escherichia coli during anaerobiosis. The growth thermograms obtained are complex and the shape of curves is dependent on the hydrogen lyase activity of the cells. Fermentation balances are given for different culture conditions, and simple growth thermograms are obtained when the hydrogen lyase activity is inhibited. A great deal of work has been done on microbial growth using calorimetry (2-5, 13, 17). When the organism studied has a simple degradative pathway for carbohydrate, for example Zymomonas mobilis growing anaerobically on glucose (5) or Saccharomyces cerevisiae growing on hexose (13), the energy-limited growth thermograms were very simple and their analysis was easy. This is not the case for organisms having complex degradative pathways. The purpose of this paper is to study the complex hexose fermentation of Escherichia coli using the calorimetric technique. MATERIALS AND METHODS Bacterial strains. The two K-12 strains of E. coli used (322, 376 P4x) were obtained from J. Puig (Laboratoire de Chimie Bacterienne de Marseille). The strain 376 P4x is a pleiotropic mutant that has lost, in addition to the nitrate reductase activity, the formate hydrogen lyase system. Medium. Bacteria were grown in a minimal medium containing (per liter: 1.00 g of NH4Cl, 0.237 g of MgSO4 7H20, 0.002 g of anhydrous CaCl2, 0.005 g of FeSO4 7H20, and 0.5 mg of thiamin. The medium was buffered by a mixture of KH2PO4 and K2HPO4 to give a final phosphate concentration of 0.1 M and a pH which can be adjusted between 6.5 and 7.8. Methionine was added at a final concentration of 50 mg per liter for experiments with strain 376 P4x. The carbohydrates were separately sterilized in distilled water and pipetted into the medium at a suitable concentration. All the experiments were carried out anaerobically under pure argon or nitrogen at 30 C. Cultivation techniques. Twelve liters of medium were incubated in a Chemap F0020 fermentor (Chemapec Zurich). The pH was maintained constant to within 0.1 pH unit by automatic addition of potassium hydroxide. Nitrogen was continuously flushed through the medium. Samples were withdrawn at different times of growth and at different times of the stationary phase for the assay of the fermentation products. 14

Microcalorimetric technique. The microcalorimetric experiments were performed with a Tian Calvet differential isothermic calorimeter, the properties of which are described in a monograph of Calvet and Prat (8). The experiments were done as previously described (2). Analytical methods. Optical densities were measured with a Jean et Constant (Prolabo, Paris) spectrophotometer at 475 nm. A curve giving the correlation between the optical density and the dry weight concentration of cells was established for each energetic substrate used. Glucose was estimated by the Glucostat method (Worthington Biochemicals Corp.). Ethanol was determined by the microdiffusion procedure of Conway (9). Carboxylic acids were separated using the chromatographic technique described by Bove and Raveux (7). The organic acids were extracted with ether at pH 2, separated by partition chromatography, and titrated with 0.01 N sodium hydroxide solution. Lactic acid was also determined by the Barker and Summerson method (1). Hydrogen lyase activity was measured by the estimation of the rate of H2 formation in a Warburg apparatus at 30 C using the technique described by Gest and Peck (10). The cell suspensions were prepared from cultures incubated at 30 C for 10 or 40 h. The cells were washed with degassed water and resuspended in phosphate buffer (0.1 M; pH 6.5) under nitrogen atmosphere. All the solutions used for the determination of hydrogen lyase activity contained chloramphenicol (100 jig/ml) to prevent any induction of protein synthesis. RESULTS

Figure 1 shows a thermogram of E. coli K-12 322 growing at pH 6.5 on synthetic medium in which gLucose was the limiting growth factor. This thermogram is different from those obtained with Z. mobilis (2) and S. cerevisiae (4) growing under the same physiological conditions. The difference is due to two thermic events that occur during the growth of E. coli, whereas only one thermic peak was observed

ANAEROBIC GROWTH OF E. COLI

VOL. 125, 1976

during the growth of the other two organisms. Surprisingly, the E. coli growth thermogram carried out at pH 7.8 (with the same initial glucose concentration) shows only one thermic event (Fig. 2). Three hypothesis can be advanced to explain the difference between the thermograms obtained at pH 6.5 and at pH 7.8: (i) at pH 6.5 the concentration of energy substrate (i.e., glucose) is not really the limiting factor for growth but it is limiting at pH 7.8; (ii) at pH 6.5 the growth of E. coli is diauxic; *and (iii) at pH 6.5 two different types of fermentation occur with and without growth. Hypothesis (i) can be discarded since (Fig. 3) the plots giving the final bacterial density versus glucose fermented are identical for growth at both pH values. These experiments show that the glucose growth yields are the same at pH 6.5 and 7.8. Moreover, the fact that the final optical densities obtained for the glucose concentrations used in the two experiments (thermograms of Fig. 1 and 3) are on the linear part of Fig. 2 shows that glucose was the limiting factor of growth in both cases. Hypothesis (ii) can also be discarded. Indeed

15

increase of biomass occurs between points C and D of the thermograms of the Fig. 1. That means that no detectable growth occurs from any products of glucose metabolism. The remaining hypothesis (iii) seems to us more probable. To test this, fermentation balances were measured at pH 6.5 during the exponential growth phase, at the point of glucose depletion and 40 h after the beginning of the stationary phase. The latter two corresponding, respectively, to point C and point D of Fig. 1. Table 1 shows the results obtained. It can be seen that formic acid is accumulated in the medium during all of the exponential growth phase and until point C. At this moment, and only at this moment, the quantity of formic acid starts to decrease and 40 h after point C practically all the formic acid

no

caLhr-1

ime hr7

FIG. 1. Thermogram of E. coli K-12 322 growing anaerobically in glucose-limited media at pH 6.5. Point C: glucose exhaustion.

df4(col.hr-l )

Glucose conc. mg/ml

FIG. 3. Final bacterial density mented strain E. coli K-12 322.

versus

glucose-fer-

TABLE 1. Metabolic products formed in the fermentationa of glucose by E. coli strain 322 Products

(mmol/100 mmol of glucose

Exponential growthb

Glucose exhaustionb

85 112 72 12 Traces

87 110 73 11 Traces

40 h after

glucose exhaustion

fermented)

time

FIG. 2. Thermogram of E. coli K-12 322 growing anaerobically in glucose-limited media at pH 7.8.

Ethanol Formic acid Acetic acid Succinic acid Lactic acid

apH 6.5 at 30 C. b Percentage recovered was 80%.

84 5 71 11 Traces

16

BELAICH AND BELAICH

J. BACTERIOL.

has disappeared. At pH 7.8 the results are quite different (Table 2). In this case all the formic acid evolved during growth is still in the medium after 40 h. From these determinations it can be concluded that hydrogen lyase activity is lacking during the growth phase at pH 6.5 and present only after the depletion of glucose. Therefore, it seems as though the second heat peak of Fig. 1 is related to the cleavage of formic acid. Figure 4 shows a growth thermogram of a hydrogen lyase-less mutant of E. coli (strain 376 P4x) at pH 6.5, i.e., under the same physiological conditions as those used for the experiment of Fig. 1. In this case the thermogram is one of classical energy limited growth, that is, no second heat peak appears after the depletion of glucose. Also, as reported in Table 3, the formic acid is accumulated during growth and does not disappear even 40 h after the end of the growth. This experiment is consistent with hypothesis (iii). The glucose fermentation balances reported in Tables 1, 2, and 3 show significant differences only for formic acid. All the other products determined, that is, ethanol, acetic acid, and succinic acid, are in the same proportions for TABLE 2. Metabolic products formed in the fermentationa ofglucose by E. coli strain 322 40 h

Products

after (mmol/100 Exponential Glucose exhausgrowth" glucose mmol of glucose exhaus.tion fermente) fermented)to tion

Ethanol

86 110 71 12

Formic acid Acetic acid Succinic acid Lactic acid

86 110 70 12

85 107 68 12

Traces

apH 7.8 at 30 C. Percentage recovered was 79%.

the wild type and hydrogen lyase-less mutant strains, whatever the pH. Surprisingly, no lactic acid was found. These results are in good agreement with those reported by Stockes (18), but not with those of Blackwood et al. (6) who found large quantities of lactic acid. Moreover, the percentage of glucose carbon recovered as fermentation products before the cleavage of formic acid was only 80%; the carbon coming from glucose and incorporated in the cell material was estimated according to the method of Luria (12). The fact that no hydrogen lyase activity occurs during the logarithmic growth phase raises the problem of the regulation of this enzymatic activity. Do the components of the hydrogen lyase system exist during the log phase? The answer to this question is given by Table 4 where it can be seen that resting cells prepared from a logarithmically growing culture have the same hydrogen lyase activity as resting cells prepared from a stationary phase culture. TABLE 3. Metabolic products formed in the fermentationa of glucose by E. coli strain 376 P4x Products (mmol/100 mmol of glucose fermented)

growth"

Ethanol Formic acid Acetic acid Succinic acid Lactic acid

87 110 69 13 Traces

Glucose exhaustionc

40 h after glucose exhaustion

86 115 71 10

86 103 73 11

a pH

6.5 at 30 C. b Percentage recovered was 80%. c Percentage recovered was 78%.

TABLE 4. The rate of hydrogen evolutiona

QH, evolved (,l/h per mg dO

dt

[dry weight] of cells)

cal hr1 Cells

Q3

With HCOOH

With HCCOH and glucose

Q2-

Harvested during exponential growth

25.0 29.6

29.3

0.1

Harvested 30 h after glucose exhaustion

33.3 33.3

26.8 24.7

Z121

~~510

~

15

tieh

FIG. 4. Thermogram of E. coli 376 p4x growing anaerobically in glucose-limited media at pH 6.5.

a Assays (30 C) were done in a Warburg apparatus with 0.3 ml of cell suspension in 0.1 M potassium phosphate buffer (pH 6.5); KOH in center well. Final fluid volume = 1.5 ml.

ANAEROBIC GROWTH OF E. COLI

VOL. 125, 1976

17

would seem that the second heat evolution is due to the cleavage of formic acid. Unfortunately the enthalpy of formic acid cleavage in the buffer used in the calorimetric experiments is too small to explain the second heat peak. So it is possible to estimate the heat corresponding to the cleavage of the formic acid and the formation of carbonate in the medium from the data of Table 5. The four reactions which occur DISCUSSION during the entire process at pH 6.5 are The E. coli anaerobic growth thermograms HCO,- + H,PO,- _ H,CO, + HPO4- AH1 (1) reported here show biphasic curves at pH 6.5 but not at pH 7.8. From the glucose balance AH2 (2) -_H2 + CO, H2CO2 data obtained with the wild-type strain and CO, + H,O H,COS AH3 (3) with the hydrogen lyase-less mutant strain, it

Moreover, glucose does not inhibit the hydrogen lyase activity (Fig. 5 and 6). From these results it can be concluded that all the components of the hydrogen lyase system are present during the log phase but without activity. The inhibitor of the hydrogen lyase activity is not simply the energy substrate but is probably related to the phenomenon of growth.

H.P04- AH4 (4) The quantity of CO2 converted to HCO,- can be calculated from the determination of the ratio K [CO,2/[H,,CO,,] 950, and the constant K' = [HI[HCO- ]/[H2CO ] 2.4 x 10-4 measured by Roughton (13). The concentration of H2CO3 H,CO, +

HP042-

=

HCO, -

+

=

=

is:

10-PH. [HCO,] (5) A- [HCO,] with A [CO 2] + [HCO3,,. From the data of Roughton and from equation (5) it can be calculated that, at pH 6.5, 0.44 mol of HCO,- is formed per mol of CO0 evolved during the reaction (2). time The heat quantity evolved per mole of formic FIG. 5. Hydrogenelyase activity of the cells. Cells acid cleaved by the hydrogenlyase is AH = AH are harvested during exponential growth. The rate of + AH2 + 0.44 AH, + 0.44 AH, with AH1 = H, evolved is measured in a Warburg apparatus under 0.714 kcal + 0.22 kcal = 0.93 kcal, AH, = nitrogen atmosphere. (a and b) Sodium formate; (c) +101.71 kcal 98.69 kcal = +3.02 kcal, AH, = sodium formate, glucose added in point A; (d) sodium +2.6 kcal (Roughton [13]), AH, = -0.714 kcal formate and glucose. + 1.2 kcal = +0.486 kcal (AH,, AH2, AH, are calculated from the data of Table 5). AH = 0.93 K' K

=

-

+ 3.02 + 0.44 (2.6 + 0.486), AH

=

+ 5.30 kcal.

TABLE 5. Quantities used for the calculation of the heat of cleavage of formic acid in H, and CO,a

Determinants

FIG. 6. Hydrogenelyase activity of the cells. Cells harvested 30 h after glucose exhaustion. The H, evolved is measured in a Warburg apparatus under nitrogen atmosphere. (a and b) Sodium formate; (c and d) sodium formate and glucose. are

kcal/mol

-101.71b Heat of formation of HCO,H aq. -98.69c Heat of formation of CO. aq. - 68.38c Heat of formation of water -0.22d Heat of ionization of HCO.H +0.714d Heat of ionization of H,PO4- 1.2e Heat of ionization of H,CO, a In the experimental conditions, all the CO, evolved remained dissolved in the medium (19). aq., Aqueous. bReference 14. cReference 15. dReference 11. ' Reference 16.

18

BELAICH AND BELAICH

Since a gross mean of 1.1 mol of formic acid is formed from 1 mol of glucose, the heat corresponding to the cleavage of the formic acid is equal to about +5.83 kcal and does not explain at all the heat evolved during the second heat peak (about -20 kcal per mol of glucose fermented). An unknown reaction related to the hydrogen lyase system and occurring during the second thermogenic peak will have to be involved. As yet we have been unable to identify this reaction. Nevertheless the simple thermograms of energy-limited growth obtained both at pH 7.8 and 6.5 with hydrogen lyase-less mutant permit the determination of the affinity of E. coli for various energy source according the colorimetric method of Belaich et al. (5).

1.

2.

3.

4. 5.

6.

LITERATURE CITED Barker, S. B., and W. H. Summerson. 1940. The colorimetric determination of lactic acid in biological material. J. Biol. Chem. 138:535-554. Belaich, J. P. 1963. Thermogenese et croissance de Pseudomonas lindneri en glucose limitant. C. R. Soc. Biol. 157:316-322. Belaich, J. P., M. Murgier, A. Belaich, and P. Simonpietri. 1971. Application de la microcalorimetrie a la mesure de l'affinite de cellules microbiennes pour leur substrat energetique. In First European Biophysics Congress. Baden, Austria. Belaich, J. P., and J. C Senez. 1965. Les developpements recents de la microcalorimetrie et de la thermogenese. C.N.R.S. 156:381-394. Belaich, J. P., J. C. Senez, and M. Murgier. 1968. Microcalorimetric study of glucose permeation in microbial cells. J. Bacteriol. 95:17,50-1757. Blackwood, A. C., A. C. Neish, and G. A. Ledingham. 1956. Dissimilation of glucose at controlled pH values

J. BACTERIOL. by pigmented and nonpigmented strains of Escherichia coli. J. Bacteriol. 72:497-499. 7. Bove, J., and R. Raveux. 1957. La separation et la determination des acides carboxyliques de Cl a C. par chromatographie de partage sur colonne de silice. Bull. Soc. Chim. France 3:376-382. 8. Calvet, E., and H. Prat. 1956. Microcalorimetrie, p. 60-62. Masson (ed). Paris. 9. Conway, E. J. 19f50. Microdiffusion analysis and volumetric error. Crosley Lockwood and Son, London. 10. Gest, H., and H. D. Peck, Jr. 1959. A study of the hydrogen lyase reaction with systems derived from normal and anaerogenic coli-aerogenes bacteria. J. Bacteriol. 70:326-334. 11 Izatt, R. M., and J. J. Christensen. 1970. Heats of proton ionization, pK, and related thermodynamic quantities, p. J58-J173. In Herbert A. Sober (ed.), Handbook of biochemistry, 2nd ed. The Chemical Rubber Co., Cleveland. 12. Luria, S. E. 1960. The bacterial protoplasm: composition and organisation, p. 1-34. In I. C. Gunsalus and R. Y. Stanier (ed.), The bacteria, vol. I. Academic Press Inc., New York. 13. Murgier, M., and J. P. Belaich. 1971. Microcalorimetric determination of the affinity of Saccharomyces cerevisiae for some carbohydrate growth substrate. J. Bacteriol. 105:573-579. 14. National Bureau of Standards. 1968. Technical note 270-3. National Bureau of Standards, Washington, D.C. 15. Rossini, X., et al. 1952. Selected values of chemical thermodynamic properties. U.S. Government Printing Office, Washington, D.C. 16. Roughton, F. J. W. 1941. The kinetics and rapid thermochemistry of carbonic acid. J. Am. Chem. Soc. 63:2930-2934. 17. Senez, J. C., and J. P. Belaich. 1963. Mecanismes de regulation des activities cellulaires chez les microorganismes. Colloq. Int. C.N.R.S. 124:357-369. 18. Stokes, J. L. 1949. Fermentation of glucose by suspension of Escherichia coli. J. Bacteriol. 57:147-158. 19. Umbreit, W. W., R. H. Burris, and J. F. Stauiffer. 1957. Manometric techniques, p. 18-27. Burgess Publishing

Co., Minneapolis.