Eur. J. Biochem. 74, 533-537 (1977)

Are the Aerobic and Anaerobic Phosphofructokinases of Escherichia coli Different? Jorge BABUL, John P. ROBINSON, and Dan G. FRAENKEL Department of Microbiology and Molecular Genetics, Harvard Medical School (Received September 14/December 21, 1976)

Phosphofructokinase has been purified from Escherichia coli strain K-12 grown in a glucoselimited chemostat, both aerobically and anaerobically. The enzymes migrated together in polyacrylamide gel electrophoresis, had the same subunit size in denaturing (dodecylsulfate) gels ( M , approx. 34000) and the same kinetic characteristics as described earlier for E. coli phosphofructokinase [e.g. Blangy e f al. (1968) J . Mol. Biol. 31, 13-35]: a sigmoid curve of velocity vs. fructose 6-phosphate concentration, activation by ADP, and inhibition by phosphoenolpyruvate. Findings [e.g. Doelle (1975) Eur. J . Biochem. 50, 335-3421 of quite different enzymes in aerobic and anaerobic cells were not confirmed.

According to several reports phosphofructokinase from Escherichia coli strain K-12 differs structurally and kinetically depending on whether the cells have been grown aerobically or anaerobically [l - 31. In this paper we show, to the contrary, that the enzymes from the two growth conditions are the same. Reichelt and Doelle [4] found, using a glucoselimited chemostat, that phosphofructokinase (specific activity of crude extracts) was constant over a range of input p o , values for which dissolved oxygen remained in the culture, but that at lower input PO, values the specific activity was higher (at a P O , of zero the factor was 2.6). Over the same range, cell yield and phosphofructokinase varied inversely. Slightly purified extracts [ l ] showed the kinetic properties of phosphofructokinase from the chemostat cells to differ according to the growth conditions. For example, the curve of velocity vs. fructose 6-P concentration was markedly more sigmoidal for the activity from anaerobic cells. In polyacrylamide gel electrophoresis the activities from the aerobic and anaerobic cultures had different RFvalues [2]. Further fractionation gave the following results [3]. Two enzymes were found in the aerobic cells, type I, 70% of the total, ‘ATP insensitive’, and type I1 ‘ATP sensitive’, both types were dimers of subunits of M , 73000. In anaerobic cells an ATP-sensitive enzyme was found; it was a tetramer of subunits of M , 90000. We have been studying the biochemical genetics of E. _. coli phosphofructokinase for some time (see ~~

Ahhreviarion. Fructose-6-P, fructose 6-phosphate. Enn-yrnes. Phosphofructokinase: ATP : ~-fruclose-6-phosphate 1-phosphotransferase (EC 2.7.1.1 1).

[5,6] and references therein). The wild-type strain contains a main activity (approx. 90%) now called phosphofructokinase-I. This is the enzyme which was so extensively characterized by Blangy et al. [7-91. It shows a sigmoid curve of activity vs. fructose-6-P, is activated by nucleoside diphosphates, and is inhibited by phosphoenolpyruvate [7,8]. Phosphofructokinase-I is a tetramer of subunits of M , 36000 [9], probably specified by the gene pfkA [6]. Another activity, phosphofructokinase-11, is physically separable from phosphofructokinase-I and not allosteric. Phosphofructokinase-I1 was originally described in a strain carrying pjkA+ and a suppressor mutation, pjkB1; in such a strain the two enzymes are found in similar amounts [lo]. But a low level of apparent phosphofructokinase-I1 was also seen [lo, 111 in wild type strains (pfkA+ pjkB+). Kotlarz et al. [ l l ] measured the relative amounts of the two enzymes in a wild-type strain grown in batch culture in several carbon sources, aerobically and anaerobically. The low amount (2-1073 of phosphofructokinase-I1 hardly varied, and the variation in total activity among the different cultures was accounted for by phosphofructokinase-I, the allosteric enzyme (e.g., twice as much in anaerobic as in aerobic glucose cultures). Those experiments used differential assays of the two enzymes in crude extracts; it was also shown that antiserum to phosphofructokinase-I did not bind phosphofructokinase-11. Earlier, with publication of [l], we purified phosphofructokinase from aerobic and anaerobic cultures, but found that both in kinetics and course of purification the predominant activity from each culture was typical

534

Different Aerobic and Anaerobic Phosphofructokinases in E . coli?

phosphofructokinase-I (unpublished). With the appearance of the detailed report describing the different aerobic and anaerobic enzymes [3], we have reinvestigated the matter, this time using chemostat cultures (as in [l]; our earlier work, as well as [3], used batch cultures). We find again that with respect to kinetics and subunit size the pure enzymes from aerobic and anaerobic cultures do not differ: they are both phosphofructokinase-1.

Table 1. Purification ofphosphofiuctokinase Yield refers to total activity of the crude extract, a value which 4 of presumed phosphofructokinase-I1 includes about 10 0

MATERIALS AND METHODS The strain of E. coli K-12 was DF1651 [6]: Fedd galK his pps pyrD str tyrA, and wild type in phosphofructokinase genes @@At pfkB+). It was grown in a chemostat of 400 ml culture volume, using minimal medium 63 (see [6]) containing 1 pg/ml thiamineHCl, 25 pg/ml histidine and tyrosine, 50 pg/ml uracil, and 0.8 mg/ml glucose. Temperature was kept at 38 "C with an internal heating element and a thermostat. Flow rate was 0.19/h. Sparging at 6 l/h was with 95 % N2/5 0 4 COZ,for anaerobic conditions, and air for aerobic conditions. A New Brunswick Scientific Co. oxygen electrode showed 0% 0 2 in the anaerobic culture and the same concentration as air for the aerobic culture. (These conditions duplicated, as closely as possible, those of [4].) After 6 volumes had passed through the chamber (30 h), the effluent was collected at 4 "C, harvested twice daily by centrifugation and stored at -20 "C. Cells from the anaerobic cultures were stored under Nz/C02. The specific activities (units/mg protein) in crude extracts of these cells were 0.22 for aerobic cells and 0.49 for anaerobic cells (average of 4 chemostat runs each). 880/0 and 92 %, respectively, of the activity in these crude extracts was precipitated by antibody to phosphofructokinase-I [ a The enzyme was usually purified (Table 1) from about 5 g wet weight of cells. The cells were thawed and suspended (12% w/v) in 10 mM Tris-HC1 (pH 7.6 at 25 "C) containing 5 mM MgC12, 14 mM 2-mercaptoethanol, and 0.1 mM EDTA (buffer A). Then they were disrupted in a Biosonik I11 (Bronwill Scientific Co.) sonicator and centrifuged 2 h at 87000 x g. The supernatant fluid was dialyzed overnight against 4 1 of buffer A, applied to a 1.5 x 23-cm column of cibacron-blue F-3 GA (Ciba) coupled to Sepharose [12] equilibrated with the same buffer, and the column was successively washed with 3 bed volumes of buffer A, 5 bed volumes of buffer A supplemented with 0.1 M Tris-HC1 (ph 7.6), and 5 bed volumes of buffer A with 0.1 M Tris-HC1, 1 mM N A D + , and 0.1 vol. saturated (4 " C )ammonium sulfate. Phosphofructokinase activity was eluted using a 440-ml linear gradient of ATP, 0-2.5 mM, in the lastmentioned buffer (with NAD+ omitted). The enzyme

Fraction

Total activity

Yield

units Aerobic cells Extract Cibacron-blue-Sepharose AMP-Sepharose

206 148 91

100 12 44

Anaerobic cells Extract Blue-dextran AMP-Sepharose

301 243 182

100 81 61

appeared after 80 - 90 ml elution. Fractions containing 3 or more units were pooled (Table 1, cibacron-blueSepharose fractions) and concentrated using an Amicon ultrafiltration cell with a PM 10 membrane. The concentrated enzyme was dialyzed overnight against 4 1 of buffer A containing also 0.3 M KCl and applied to a 1.5 x 23-cm column of AMP-Sepharose (agarose-hexane-adenosine 5'-monophosphate, AGAMP type 2, P-L Biochemicals) equilibrated with the same buffer. After a wash with 6 bed volumes of the same buffer, a 440-ml linear gradient of ATP (02.5 mM) was applied, the enzyme eluting after about 100 ml. Fractions with more than 3 units were pooled (Table 1, AMP-Sepharose fractions). Native 7 % polyacrylamide gels used the method of Davis [13]: The enzymes were first dialyzed vs. 1 mM Tris, 7.7 mM glycine, 0.1 mM ATP and portions mixed with glycerol and bromthymol blue before application to gels. Electrophoresis was at 2 mA/gel at 5 "C with 0.1 mM ATP in the running buffer. 10 % dodecylsulfate gels were according to Laemmli [14]: the enzymes were first dialyzed vs. 0.062 M Tris-HC1 pH 6.8, 2 % dodecylsulfate; to portions, 2-mercaptoethanol was added to 5 %, the mixtures were placed in a boiling bath for 1.5 min and, after addition of glycerol and bromphenol blue, applied to the gels. Electrophoresis was at 2 mA/gel in the presence of 0.1 dodecylsulfate. Protein was stained with Coomassie blue R250. Enzyme kinetics used, per 1 ml, 0.1 M Tris-HC1, 0.2 mM NADH, sodium fructose-6-P as indicated, 1 mM sodium ATP, 10 mM MgC12, 2 mM NH4C1, a dialyzed mixture of auxiliary enzymes (Boehringer) containing 20 pg fructose- 1,6-bisphosphate aldolase, 3 pg triosephosphate isomerase, and 30 pg cc-glycerophosphate dehydrogenase, 1 mM sodium creatine phosphate, and 25 pg creatine phosphokinase (Sigma) ; modifications are given in the legend to Fig.3. A340 was followed at 25 "C with a Gilford recording spectrophotometer. Enzyme units are pmol/min.

'x

J. Babul, J . P. Robinson, and D. G. Fraenkel

RESULTS AND DISCUSSION

E. coli phosphofructokinase has been previously purified to apparent homogeneity by a variety of methods [6-&lo, 11,15,16], themorerecent ofwhich include a blue-dye affinity column [6,10,11,16]. An AMP affinity column has also been useful in purification of some phosphofructokinases [17,18] and the present preparation used the two affinity columns in sequence (Materials and Methods). The crude extracts did contain an approximately 10% level of activity not precipitated by antiserum to phosphofructokinase-I ;the pure enzymes (AMP-Sepharose fractions, Table 1) were completely sensitive to this antiserum. If, as seems likely, the non-precipitated activity in the orude extract is phosphofructokinase-I1 it would not have bound to the first column (to be reported) and indeed there was a little activity in that column wash. N o other minor phosphofructokinase activities were found. The pure enzymes migrated together on native polyacrylamide gels (Fig. 1) and had the same size subunit (Fig.2), M , about 34000 as known for phosphofructokinase-I. Kinetics are shown in Fig. 3 for the same two preparations and for two others. The three usual characteristics of phosphofructokinase-I are evident : sigmoidicity with fructose-6-P,

Fig. 1. Nurive polyacrylamide gel electrophoresis. Gel 1 , 3 units enzyme from aerobic cultures; gel 2, 4.1 units enzyme from anaerobic cultures; gel 3, 2.1 units of each enzyme (aerobic and anaerobic). Migration from top t o bottom. Protein was stained with Coomassie blue

535

activation by ADP, and inhibition by phosphoenolpyruvate. The slight differences between the four separate preparations are not significant. Considering the relatively high overall yields of pure enzymes, their similarities, and the absence of other active fractions, we conclude that in these cells, aerobic or anaerobic, there is no substantial level of phosphofructokinase other than phosphofructokinase-I. Several possible explanations may be considered for the differences between these results and those of the Queensland laboratory. Strain Differences. The Queensland group used E. coli ‘strain K-12 from the department culture collection’ [4], otherwise unspecified. The present experiments used the K-12 strain DF1651. However, we have also used another E. coli K-12 strain (the prototroph HfrC, also called K-10) grown aerobically and anaerobically in glucose-limited chemostats, and partially purified their phosphofructokinase with a cibacron-blue affinity column ;again, no marked kinetic differences were found (data not shown). Strain K-10 is the starting strain for our genetic work, and is the strain used by Kotlarz et al. in their assessment of the amounts of phosphofructokinases-I and I1 in a wildtype strain [ l l ] . Yet another E. coli K-12 strain (Hfr 3000) was used in the earlier enzymological work [7-91, i.e., for the original definition of phos-

Fig. 2. Polyacrylamide gel electrophoresis of’ dissociated enzyme. Gel 1,2.4 units enzyme fromaerobic cultures; gel 2,2.4 unitsenzyme from anaerobic cultures; gel 3, 2.4 units of each enzyme (aerobic and anaerobic); gel 4, 2 pg ovalbumin (subunit M , 43000) and 5 pg rabbit muscle glyceraldehyde-3-phosphate dehydrogenase (subunit M , 36000). Migration from top to bottom

536

Different Aerobic and Anaerobic Phosphofructokinases in E. coli? If

A

B

[Fructose-6-P] (rnM)

Fig. 3 . Phosphqfiuctokinase activity. ( A ) Aerobic; ( B ) anaerobic. The two preparations shown in Fig. 1 and 2 were used (A, 0) as well as two separate preparations from diKerent chemostat runs (A,0). In each panel the middle curve shows velocity in the standard assay without added effectors, as a function of fructose-6-P concentration; the right hand curve is the same assay in the presence of0.5 mM phosphoenolpyruvate (P-Prv); and the left hand curve is the assay in the presence of 0.5 mM A D P (in which case creatine phosphokinase was omitted). Each assay used about 4 m U enzyme

phofructokinase-I. We feel confident in assigning to the three hona fide K-12 strains the same major type of phosphbfructokinase, phosphofructokinase-I. Purity and Yield. Typical activities of phosphofructokinase in crude extracts of the wild-type strain range from 0.1-0.7 unit/mg [ l l ] , while the pure enzyme has a specific activity of about 200 units/mg [8]. Doelle’s preparations were partially purified, starting with specific activities in crude extracts of 0.48 unit/mg (aerobic cells) and 0.52 (anaerobic). His ‘aerobic type I’ enzyme had a specific activity of 31 in 6 % yield, and ‘aerobic type 11’ had a specific activity of 2 in 357; yield; the ‘anaerobic enzyme’ had a specific activity of 23 in 9 % yield (57% of the enzyme in anaerobic cells was discarded as not precipitating between 40 and 50% ammonium sulphate). If any one of those fractions was phosphofructokinase I, it would have been impure or denatured. It is unclear how subunit sizes in dodecylsulfate gels ( M , 73000 aerobic, 90000 anaerobic) were obtained. Considering the low yields it is also possible that originally minor fractions (e.g., phosphofructokinase-11) were purified. Possible Artefacts. (a) Phosphofructokinase-I is easy to desensitize [lo]. (b) It needs NH: or K + for maximal activity [lo], a requirement not always recognized [3,7,8]. One of these ions in sufficient amount may be supplied by the other reagents in an uncontrolled way. (c) Since ADP activates, the assay (at least at low fructose-6-P concentration) needs an ATP-generating system [7]; this was not used [1,3]. With impure enzymes it is even more difficult to assess the effect of adenine nucleotides [19]. ATP inhibition has been considered by others [7,19,20]. Could Phosphofructokinase-I be the End Result of the Alteration during Purification of the Several Activities Described hy Doelle. This seems unlikely,

for the purification steps are not ‘harsh’, and, admitting the difficulties of kinetic studies with crude fractions, our experience has been that the kinetic properties do not obviously change during the purification. Thus we conclude that the principal phosphofructokinase of aerobic and anaerobic E. coli is the same, phosphofructokinase-I, and the transitions between aerobic and anaerobic conditions may involve a change in rate of synthesis of this enzyme, as originally suggested [4]. It is still possible that more detailed studies will reveal significant differences between the enzymes from aerobic and anaerobic cells, or from other growth conditions; and the question of the form of the protein in the cells is of much interest. Reeves and Sols [20] used whole cells of E. coli strain K-12 (Hfr 3300), made permeable to substrates, and found that although the kinetic characteristics were generally the familiar ones of phosphofructokinase-I there were some differences, such as a Hill coefficient of 2 instead of 4. This work was supported by grants from the National Science Foundation (BMS72-01882 A04) and the National Institutes of Health (5 R01 GM 21, 098-02).

REFERENCES 1. Thomas, A. D., Doelle, H. W., Westwood, A. W. & Gordon, G . L. (1972) J . Bacteriol. 112, 1099- 1105. 2. Doelle, H. W. (1974) FEBS Lett. 49, 220-222. 3. Doelle, H. W. (1975) Eur. J . Biocl7eni. 50, 335-342. 4. Reichelt, J. L. & Doelle, H. W. (1971) Anronir Van Lecuwc.~hoek J. Microhiol. Serol. 37, 497 - 506. 5. Fraenkel, D. G. & Vinopal, R. T. (1973) An/7u. Rev. Microhiol. 27, 69- 100. 6. Vinopal, R. T., Clifton, D. & Fraenkel, D. G. (1975) J . B w teriol. 122, 1162- 1171. 7. Blangy, D., Buc, H.& Monod, J. (1968)J. Mol. B i d . 31.13-35. 8. Blangy, D. (1971) Biochamic ( P ~ i r i s S3, ) 135-144. 9. Blangy, D. (1968) FEBS Lcw. 2, 109- 111.

537

J . Babul, J . P. Robinson, and D. G. Fraenkel 10. Fraenkel, D . G., Kotlarz, D. & Buc, H. (1973) J . B i d Chem. 248, 4865 - 4866. 11. Kotlarz, D., Garreau, H. & Buc, H. (1970) Biochim. Biophys. Acta, 381, 257-268. 12. Ryan, L. D. & Vestling, C . S. (1974) Arch. Bioc,hern. Biophys. 160, 279 - 284. 13. Davis, B. J. (1964) Ann. New York Acad. Sci. 121, 404-427. 14. Laemmli, U. K . (1970) Nature (Lond.) 227, 680-685. 15. Griffin, C. G., Houck, B. N. & Brand, L. (1967) Biochcm. Biop1ij.s. Res. Commun. 27, 287- 293.

16. Thompson, S. T., Cass, K. H. & Stellwagen, E. (1975) P t w . Nut1 Acad. Sci. U.S.A. 72, 669-672. 17. Comer, M . G., Craven, D. V., Harvey, M. J., Atkinson, A. & Dean, P. D. G . (1975) Eur. J . Biorhem. 55, 201 -209. 18. Hengartner, H. & Harris, J. 1. (1975) FEBS Lctt. 55, 282-285. 19. Atkinson, D. E. (1966) Annu. Rev. Biochem. 35, 85- 124. 20. Reeves, R. E. & Sols, A. (1973) Biochem. B i o p h j ~Res. . Cononuri. 50. 459 466. ~

J . Babul, Departamento de Quimica, Facultad de Ciencias, Universidad de Chile, Casilla 653, Santiago, Chile J. P. Robinson, Department of Plant Biology and Microbiology, Queen Mary College. University of London, Great Britain, E l 4NS

D. G. Frdenkel*, Department of Microbiology and Molecular Genetics, Harvdrd University Medical School, 25 Shattuck Street, Boston, Massachusetts, U.S.A. 021 15 .-

* T o whom correspondence should be addressed.

Are the aerobic and anaerobic phosphofructokinases of Escherichia coli different?

Eur. J. Biochem. 74, 533-537 (1977) Are the Aerobic and Anaerobic Phosphofructokinases of Escherichia coli Different? Jorge BABUL, John P. ROBINSON,...
NAN Sizes 0 Downloads 0 Views