Succinic Semialdehyde Dehydrogenases of ...

Eur. J. Biochem. 113, 555-561 (1981) FEBS 1981

Succinic Semialdehyde Dehydrogenases of Escherichia coli Their Role in the Degradation of p-Hydroxyphenylacetate and y-Aminobutyrate Mark I. DONNELLY and Ronald A. COOPER Department of Biochemistry, School of Biological Sciences, University of Leicester (Received July 15, 1980)

Two physically and genetically distinct forms of succinic-semialdehyde dehydrogenase have been identified in Escherichia coli B. The two enzymes could be separated by filtration on Sephadex G-150 and their apparent molecular weights were 200000 and 97000. The larger enzyme, which is specific for NADP, is induced by growth on g-aminobutyrate. Its induction is highly coordinated with that of y-aminobutyrate: 2-oxoglutarate transaminase, the enzyme which initiates degradation of y-aminobutyrate. The smaller enzyme, which is induced by growth on p-hydroxyphenylacetate, has been purified to 98 homogeneity by affinity chromatography in conjunction with conventional methods. Under standard assay conditions this enzyme acts preferentially with NAD but reduces NADP at 15 of the rate observed for NAD, primarily because of a difference in K,. Apparent 1.4 pM, respecK,,, values for succinic semialdehyde and NAD are 13.3 1.3 pM and 33.7 tively. The subunit molecular weight was estimated to be 55000, indicating that the native enzyme is dimeric. The NAD-dependent succinic-semialdehyde dehydrogenase is also induced by exposure of cells to exogenous succinic semialdehyde, a treatment which has no effect on the amount of other enzymes of p-hydroxyphenylacetate or y-aminobutyrate metabolism. Apparently the gene for this enzyme functions independently from the genes encoding the other enzymes of p-hydroxyphenylacetate degradation. As a consequence of its induction mechanism, this NAD-dependent dehydrogenase is also present in extracts of E. coli B grown with y-aminobutyrate as sole nitrogen source, in addition to the NADP-specific enzyme involved in y-aminobutyrate metabolism. Presumably the NAD-dependent enzyme is gratuitously induced by succinic semialdehyde formed by transamination of y-aminobutyrate. When grown on p-hydroxyphenylacetate as sole source of carbon, Escherichia coli strains B, C and W synthesize a dehydrogenase which oxidizes succinic semialdehyde, an intermediate of p-hydroxyphenylacetate degradation, to succinate [l](Fig. 1, reaction 4). Extracts prepared from strain B grown on p-hydroxyphenylacetate oxidize succinic semialdehyde four times faster when NAD is used as cosubstrate than when NADP is used. Succinic semialdehyde is also an intermediate in the bacterial degradation of y-aminobutyrate [2] (Fig.1). Studies conducted by Halpern and Enzymes. Succinic-semialdehyde dehydrogenase or succinicsemialdehyde:NAD(P)+ oxidoreductase (EC 1.2.1.16); y-aminoburyrate: 2-oxoglutarate transaminase or 4-aminobutyrate : 2-0x0glutarate aminotransferase (EC 2.6.1.19) ; glutamate dehydrogenase or L-glutamate:NADP' oxidoreductase (deaminating) (EC 1.4.1.4); 5-carboxymethyl-2-hydroxymuconic-semialdehyde dehydrogenase (EC 1 .-.-.-); 5-carboxymethyl-2-hydroxymuconic acid decarboxylase (EC 4.1.1.-).

his colleagues into the utilization of y-aminobutyrate by E. coli K12 describe only NADP-linked succinicsemialdehyde dehydrogenase activity [3- 51. Since Halpern did not use ultracentrifuged extracts, the high NADH-oxidizing activity found in crude extracts of E. coli would preclude accurate evaluation of the enzyme's co-substrate specificity. Therefore, we also undertook a study of y-aminobutyrate degradation in E. coli B. Our interest centered in particular on determining if the enzyme involved in p-hydroxyphenylacetate degradation is the same as that involved in y-aminobutyrate degradation or if different proteins are induced. Bacterial succinic-semialdehyde dehydrogenases, described previously, vary in their cofactor preference [3,6-111. Growth of Pseudomonas putida strain U on p-hydroxyphenylacetate induces dehydrogenase activity which uses both NAD and NADP as cosub-

556

Succinic Semialdehyde Dehydrogenases of E. coli

MATERIALS A N D METHODS

Bacierial Sirains and Culture Merhods OH

OH

p - hydroxyphenyiacetate

, lNAD (1)

NH2

I OH

y -arninobutyrate

(II)

4

INAD (PI

H02CY O z H

Fig. 1. Puthwuys of degradation of p-hydroxyphenylacetic ucid and yminohuz!.ric, uc~id.Both pathways produce succinic semialdehyde, which is oxidized t o succinate by an NAD-dependent or NADPdependent dehydrogenase

strate, whereas only NADP-linked activity is induced in Pseudomontrs T and Acineiohacier [6]. Similarly, species of Pxudomonas grown on compounds related to :j-aminobutyrate produce both NAD-dependent and NADP-dependent activities [7- 101. Of particular significance is the description by Nirenberg and Jakoby [7] 01' two distinct succinic-semialdehyde dehydrogenases found in extracts of a Pseudomonas species grown on ;>-hydroxybutyrate, one specific for NAD, the other for NADP. In the present study we show that E. coli B also possesses two succinic-semialdehyde dehydrogenases. One enzyme, which is specific for NADP, is involved in the degradation of ;I-aminobutyrate. Its synthesis is highly coordinated with that of y-aminobutyrate: 2oxoglutarate Iransaniinase, both in the wild type and in mutants that can use y-aminobutyrate as sole carbon and nitrogen source. The second enzyme is less specific but acts preferentially with NAD. This NAD-linked succinic-semialdehyde dehydrogenase, which functions in the degradation of p-hydroxyphenylacetatc. has been purified and partially characterized. The enzyme appears to be encoded by a discrete regulatory unit and is induced by succinic semialdehyde itself'. Consequently, when E. coli B is grown on 7-aininobutyrate both the NADP-dependent and NAD-dependent dehydrogenases are induced, the latter produced in response to succinic semialdehyde generated during :l-aminobutyrate degradation.

Wild-type Escherichia coli B was from the E. coli Genetic Stock Center, Yale University. This strain grows well on y-aminobutyrdte as sole nitrogen source. Mutants able to grow on y-aminobutyrate a s sole carbon and nitrogen source arose spontaneously from the wild type when incubated on 7-aminobutyrate plates at 30 "C. Cells were cultured at 30 T in 100-ml portions of M63 or (NH4)2SO1-free M63 medium [12] in 250-ml conical flasks supplemented with 1 ml nutrient broth. Carbon and nitrogen sources, separately sterilized, were added to give final concentrations of 15 mM, except p-hydroxyphenylacetate which was used at 5 mM. For physiological studies, cells were harvested during exponential growth. For enzyme purification, cells were cultured at 30 -C with forced aeration in 16 1 M63 medium using p-hydroxyphenylacetate as carbon source. Cells were allowed to grow into stationary phase before harvesting.

Preparation of Exiracu The buffer used for all manipulations of extracts and for enzyme purification was 0.09 M NalHPO?,' KH2P04, pH 7.0, containing 9 2 ) glycerol and 1 inM dithiothreitol. Cells from 100-ml cultures were resuspended in 3 - 4 ml buffer and disrupted by ultrasonic oscillations in an MSE 100-W ultrasonic disintegrator operating at 8 pM peak-to-peak amplitude at 0 .C for two 30-s periods with cooling between for 1 niin. The crude extracts were ultracentrifuged for 90 min at 120000 x g to remove NADH-oxidizing activity. For enzyme purification, 18 g wet weight of cells were suspended in 35 ml buffer and treated similarly except for the use of eight 30-s periods of ultrasonication. Enzyme preparations were maintained at 0 -- 4 -C at all times.

Enzyme Assays Succinic-semialdehyde dehydrogenase activity was assayed by measuring the reduction of NAD or NADP spectrophotometrically at 340 nm. Reaction mixtures contained 0.1 mM succinic semialdehyde and 0.14 mM NAD(P) in 1.0 ml 0.1 M Na2HP04,KH2P04buffer, pH 8.0, at 30':C. The reaction wa5 initiated by addition of enzyme. y-Aminobutyrate : 2-oxoglutarate transaminase was estimated after the method of Zaboura and Halpern [ 5 ] . Reaction mixturcs contained 10 mM NH4C1, 16 mM L-glutainate, 0.1 m M succinic semialdehyde, 0.1 mM NADPH and 1.4 units of L-glutamate dehydrogenase in 1.0 in1 0.1 M NazHP04/KH2P04 buffer, pH 7.0, at 30 ' C . Transaminase activity was estimated from the steady-state rate of NADPH oxidation by glutamate dehydrogenase.

M. I. Donelly and R. A. Cooper

About 1 min was required following addition of enzyme for sufficient 2-oxoglutarate to accumulate to give a linear rate of reaction. 5 - Carboxymethyl- 2 - hydroxymuconic - semialde hyde dehydrogenase and 5-carboxymethyl-2-hydroxymuconate decarboxylase were assayed as described earlier [l]. Protein was determined by the method of Hartree [13] using crystalline bovine serum albumin as a standard. Gel Filtration

Succinic-semialdehyde dehydrogenases were separated by gel filtration through a Sephadex G-150 column (40 x 2.6 cm, standard grade). Extracts containing approximately 1.5 units of total dehydrogenase activity and 4 mg protein in 3 ml were eluted at a flow rate of 24 ml/h under constant pressure. For estimation of the molecular weights [14], the column was calibrated with blue dextran, catalase (bovine), lactate dehydrogenase (rabbit muscle), bovine serum albumin and ovalbumin. under identical conditions. Enzyme Purification The NAD-dependent succinic-semialdehyde dehydrogenase was purified from p-hydroxyphenylacetate-grown cells. The ultracentrifuged extract prepared from 18 g cells was fractionated by addition of ammonium-sulfate-saturated buffer, pH 7.0. Protein precipitating between 40 % and 65 % saturation was resuspended in 17 ml buffer and dialyzed overnight against 2 x 500-ml volumes of buffer. The dialyzed enzyme was applied to a DEAESephacel column (28 x 2.6 cm) equilibrated with buffer. Protein was eluted with an 800-ml linear gradient of 0-0.3 M NaCl at a flow rate of 48 ml/h. Pooled fractions (83 ml) from the DEAE-Sephacel column were applied at a flow rate of 15 ml/h to an AMP-Sepharose column equilibrated with buffer. The column was washed with 30 ml buffer then eluted with a 160-ml linear gradient of 0-0.5 mM NADP. In this procedure only 40 % of the enzymatic activity was bound. However, after regeneration of the column with 8 M urea and re-equilibration with buffer, the activity which had failed to bind initially was rechromatographed and approximately the same total amount of activity was retained. The column was eluted as before and the active fractions from both elutions were pooled. The low capacity of the column was not due to NaCl in the DEAE-Sephacel fractions, since in preliminary tests 0.5 M NaCl failed to elute bound enzyme. Since the purified enzyme was inactivated by ultrafiltration, the pooled AMP-Sepharose fractions (94 ml) were placed in dialysis tubing and water was removed by application of dry Sephadex G-200. The concen-

551

trated enzyme solution was centrifuged to remove precipitated protein and dialyzed exhaustively against buffer to remove NADP. At this stage, the enLyme, purified @-fold, was estimated to be 98% homogeneous. Characterization of ihe Purified Enzyme The purity of the enzyme from AMP-Sepharose chromatography was evaluated by densitometry of a sodium dodecylsulfate/polyacrylamide gel stained for protein using a Pye-Unicam SP8-100 spectrophotometer equipped with a gel scanning accessory. The subunit molecular weight was estimated by sodium dodecylsulfate/polyacrylamide gel electrophoresis [15] with bovine serum albumin, ovalbumin, a-chymotrypsinogen and lysozyme as standards. The enzyme's pH optimum was determined in 0.1 M sodium phosphate, 0.1 M Tris/HCl and 0.1 M sodium glycinate buffers. Apparent K,,, values for succinic semialdehyde and NAD were estimated by computer through an iterative, least-squares fit of the data to the Michaelis-Menten hyperbola [16]. Rates of NAD(P) rcduction were measured spectrophotometrically in 0.1 M Na2HP04/ KH2P04 buffer, pH 8.0, 30'C. NAD and succinic semialdehyde concentrations, when held constant, were 0.6 mM and 0.1 mM respectively. When varied, concentration ranges were: NAD, 14-280 pM; succinic semialdehyde, 3-48 pM; and NADP, 0.44 mM. Induction by Succinic Semialdehyde A culture of wild-type cells, growing exponentially on succinate with (NH4)2S04as nitrogen source, was diluted to an absorbance at 680 nm of 0.1. Filtersterilized succinic semialdehyde and succinate were added to give final concentrations of 4 mM and 12 mM, respectively. As a control, the culture was also diluted into fresh medium containing only succinate as carbon source. After 3 h incubation at 30 "C cell extracts were prepared and assayed for enzymes involved in p-hydroxyphenylacetate and y-aminobutyrate metabolism. Chemicals Succinic semialdehyde was from Sigma (London) Ltd or was prepared by hydrolysis of diethylformylsuccinate, a gift from Dr P. J. Chapman [9]. 5-Carboxymethyl-2-hydroxymuconic semialdehyde and 5carboxymethyl-2-hydroxymuconate were prepared as described by Sparnins et al. [6]. Chromatographic material was purchased from Pharmacia (Great Britain) Ltd. NAD, lactate dehydrogenase and catalase were products of Boehringer (London) Ltd. NADP, NADPH, y-aminobutyrate, L-glutamate and protein molecular weight standards were from Sigma (Lon-

558

Succinic Semialdehyde Dehydrogenases of E. coli

don) Ltd. All other chemicals were of the highest purity commercially available. RESULTS AND DISCUSSION

IdentiJi'cution oj'Two Succinic Semialdehyde Dehydrogenases in E. coli B When wild-type Escherichia coli B is grown on p-hydroxyphenylacetate the ratio of NAD-dependent to NADP-dependent succinic-semialdehyde dehydrogenase activity in extracts is approximately four, but when grown on succinate with y-aminobutyrate as sole source of nitrogen, the ratio is less than two. Heating of the latter extract for 2 min at 50 C destroyed 90(::, of the NAD-dependent activity while reducing NADP-linked activity by only 15 (';,. Clearly more than one succinic-semialdehyde dehydrogenase is produced under these growth conditions. To identify the enzymes involved, extracts were fractionated by gel filtration on Sephadex G-150 and fractions were assayed both for NAD-dependent and NADP-dependent activity. The elution profiles showed that E. coli B can produce at least two succinic-semialdehyde dehydrogenases. The larger enzyme, which is specific for NADP, is induced by growth on y-aminobutyrate. It is present only at low activity in extracts of p-hydroxyphenylacetate-grown cells. NAD-dependent activity is present in high amounts under both growth conditions and coelutes with lower amounts of NADP-dependent activity. The simplest interpretation is that this second peak represents a single enzyme which acts preferentially with N A D as cosubstrate but which can also use NADP. Calibration of the column with proteins of known molecular weight gave apparent molecular weight estimates of 200000 for the NADP-specific enzyme and 97000 for the NAD-preferring enzyme, assuming that all the proteins concerned were of the same shape and solvat ion.

Enzyme Purif i c u n o n und Churacrerizurion Because of the overlapping cofactor specificity of the enzyme\, one cannot ube the rates of NAD and NADP reduction alone to estimate the amounts of

Purification jtzp

Ultracentrifuged extract Ammonium rulphate DCAT-Seph,icef AMP-Seph,iro\e Concentr,rtion dnd dialy\is

the two enzymes in extracts. The exact relative rates of NAD and NADP reduction by the NAD-preferring enzyme must be determined first. For this reason, and to further our understanding of the pathway of p-hydroxyphenylacetate degradation, we purified the NAD-linked succinic-semialdehyde dehydrogenase (Table 1). The final enzyme preparation had a specific activity of 34 units/mg. This corresponds to a 49-fold increase over the activity in the ultracentrifuged extract, prepared from cells grown into stationary phase, and 106-fold over that observed in extracts of exponentially growing cells (Table 2). Analysis of the purified enzyme by sodium dodecylsulfate/polyacrylamide gel electrophoresis showed one major band and a trace contaminant when stained for protein. From a spectrophotometric scan of this gel the enzyme was estimated to be 98 pure. Comparison of the enzyme's mobility in these gels with that of standards of known molecular weight gave an apparent molecular weight for the subunit of 55000. In conjunction with the molecular weight estimate of 97000 by gel filtration, the native enzyme appears to be a dimer of similar or identical subunits. Under standard assay conditions the purified enzyme reduced NADP at 15 /'; of the rate of NAD reduction. Apparent K , values for succinic semialdehyde and N A D were determined at a single cosubstrate concentration. The K , for succinic semialdehyde at 0.6 mM N A D was 13.3 pM (* S.E. = 1.3). The value determined for NAD, at 0.1 mM succinic semialdehyde, was 33.7 pM (k S.E. = 1.4). An accurate estimate of the Km for NADP could not be obtained. Reciprocal plots of the data were non-linear, giving the appearance of substrate inhibition. The linear portion of these plots extrapolated to give K , values of 2- 3 mM, but non-linearity occurred below these concentrations. However, it is clear that the lower activity of the enzyme toward NADP is due primarily to a difference in K,. The purified enzyme is less active in Tris buffer than in phosphate or glycine buffers, and displayed a broad pH optimum with maximum activity at pH 8.2. Activity decreased sharply as the pH was raised above pH 9.2. Its activity is inhibited by NADH and m or p-hydroxybenzaldehyde, but not by succinate I;,,

Volume

Protein

Activity

Specific activity

ml

mg

units

units mg

34 22 5 83 94 10

590 322 45

41 5 349 323 160 74

-

22

0 70 108 7 20 ~

34

Yield

'

0

100 84 78 39 18

M. I. Donnelly and R. A. Cooper

559

Table 2. Specific activities of enzymes in extracts f r o m cells grown under different nutrient conditions Activity due to the NADP-specific and NAD-preferring succinic semialdehyde dehydrogenases were distinguished by correcting the observed rate of NADP reduction for the rate of NADP reduction by the NAD-linked enzyme (see text for details). y-Aminobutyrate: 2-oxoglutarate transaminase activity was also measured Strain

Activity

Growth conditions ~~~

carbon source

succinic semialdehyde dehydrogenase WAD)

nitrogen source

succinic semialdehyde dehadrogenase (NADP)

y-aminobutyrate transaminase

nmol min-' mg - 1 ~

Wild type

y-Aminobutyrate+ mutant

p-hydroxyphenylacetate succinate succinate succinate succinate

NH3 NH3 >-aminobutyrate y-aminobutyrate proline

succinate succinate glucose glucose y-aminobutyrate y-aminobutyrate

NH3 y-aminobutyrate NH3 9-aminobutyratc NH3 y-aminobutyrate

or levulinic acid, the keto homologue of succinic semialdehyde. Physiologicul Role of the Two Enzymes

Knowing the relative activity of the purified succinic semialdehyde dehydrogenase toward NAD and NADP under standard assay conditions ( VNADP/ V N A= ~ 0.15), it is possible to calculate the activity of the individual dehydrogenases when both are present in an extract. The activity of the NADP-specific enzyme can be approximated by subtracting 0.15 times the observed NAD-dependent activity from the observed NADP-dependent activity. When applied to specific activities measured in extracts of cells grown under a variety of conditions, this adjustment produced a clear picture of the physiological role of the two dehydrogenases. This picture was greatly enhanced by including in our study a mutant of E. coli B, which can use y-aminobutyrate as sole source of carbon as well as nitrogen, and by assaying the activity of y-aminobutyrate : 2-oxoglutarate transaminase. This enzyme catalyzes reaction 3 of Fig. 1, which initiates y-aminobutyrate degradation. In E. coli K12 the gene encoding this enzyme exists as part of an operon with genes for a succinic semialdehyde deh ydrogenase and a permease for y-aminobutyrate uptake [17]. In that strain the expression of the two enzymes is highly coordinated [4]. Table 2 lists the specific activities of the two succinic-semialdehyde dehydrogenases and the transaminase. While in most cases all three enzymes are either elevated or at basal levels, there are important exceptions. Most notably, growth onp-hydroxyphenyl-

+ NH3

~~

319 15 180 30 10

34 26 105 21 23

15 30 132 30 34

16 139 6 108 135 148

91 351 89 395 319 408

96 223 3 15 502 232 338

acetate strongly induces the NAD-dependent dehydrogenase while the NADP-specific enzyme and the transaminase remain at uninduced levels. Also, when the mutant strain is grown under non-inducing conditions [succinate with (NH4)2S04] the NADPspecific dehydrogenase and the transaminase are 3-4-fold elevated relative to amounts seen in the wild type, while the activity of the NAD-dependent dehydrogenase is unaltered. It is clear from the mutant that the amount of transaminase activity is more highly correlated with that of the NADP-specific dehydrogenase than with that of the NAD-dependent enzyme. A plot of transaminase activity against the NAD-linked or NADP-linked enzyme illustrates this distinction (Fig. 2). Linear regression analysis shows that the transaminase and NADP-specific dehydrogenase activities are induced in a highly coordinate manner (correlation coefficient = 0.92). Most probably these enzymes exist as a single operon, as described for E. coli K12 [17]. No significant correlation exists between the NAD-dependent succinic-semialdehyde dehydrogenase and the transaminase (correlation coefficient = 0.16). Although not necessary for y-aminobutyrate degradation, this dehydrogenase appears to be induced fortuitously under certain conditions. Induction of the NAD-Dependent Dehydrogenuse by Succinic Semiuldehyde

Further consideration of the data in Table 2 reveals the regulation of the NAD-dependent succinicsemialdehyde dehydrogenase ; the enzyme is induced

Succinic Semialdehyde Dehydrogenases of E. c d i

560 500 .. 400

A

A

'

300 .

-,m E

200

i

Table 3 Enzymutrc ac tivitrey in cells exposed to exogenous uiicrnic semiuldrhvde Extracts from wild-type E colr B, grown on succinate and exposed for 3 h to 4 mM succinic semialdehyde, were assayed for the NADdependent succinic semidldehyde dehydrogendse and for enzymes of i-dminobutyrate degradation [succinic semialdchyde dehydrogenase (NADP) and 1-dmmobutyrate 2-oxoglutdrate trdn~dininase] and of p-hydroxyphenylacetdte degradation [5-carboxymethql-2hydroxymuconic semialdehyde (I) dehydrogen~sednd 5-carboxqmethyl-2-hydroxymucondte (11) decarboxylase, see Fig 11 Control cells were not exposed to succmic aemialdehyde Fully induced activities for these enzymes are 319, 105, 132, 180 and 330 nmol min mg protein respectively Enzyme

Actikity ~

induced nmol

inin

control

' mg ' ~

Succinic semialdehyde dehydrogena\e WAD) Succinic semialdehyde dehydrogenase (NADP) -Aminobutyrdte transnmindse I dehydrogenase 11 decarboxylase Uehydrqenase activity (nrnol rnin-' rng-') Fig. 2. Coordinurr inducrion of' the NAD-dependent o r N A D P dependent succinic semiuldeh.de dehydrogenusr with y-aminohu!vr.ote:2-oxoglutarate trunsuminuse. Extracts were prepared from wild-type cells ( A ) and froin a mutant able to use 7-aminobulyrate as sole source of carbon and nitrogen (A) after growth under different nutrient conditions. The activity of the NAD-linked dehydrogenase ( A ) is not correlated with that of the transaminase (correlation coefficient = 0.16) whereas the activity of the NADPspecific dehydrogenaae (B) i s highly correlated (correlation coefficient = 0.92)

whenever cells metabolize a substrate to succinic semialdehyde. When metabolism of y-aminobutyrate is repressed in the wild type by including (NH4)2S04 in the medium, the enzyme is not formed. The fact that it is induced most strongly by growth on p-hydroxyphenylacetate may reflect the higher intracellular concentration of succinic semialdehyde that is likely to arise in the absence of prior induction of the NADP-specific succinic-semialdehyde dehydrogenase that occurs during growth on y-aminobutyrate. To test the hypothesis that succinic semialdehyde itself induces the NAD-dependent enzyme, wild-type cells growing on succinate with (NH4)2S04 were exposed to succinic semialdehyde then assayed for enzymatic activities. In addition to the enzymes assayed previously. two enzymes from the pathway forp-hydroxyphenylacelate degradation were assayed : 5-carboxymethyl-2-hydroxymuconic-semialdehyde dehydrogenase (reaction 1, Fig. 1) and 5-carboxymethyl2-hydroxyinuconate decarboxylase (reaction 2, Fig. 1). End-products of degradative pathways may be the inducers for some or all of the preceeding enzymes of the pathway [18], so it was important to assess accurately the regulatory function of succinic semi-

285

17

20 22 1

28 30 4 3

7

I

aldehyde. Table 3 compares the activities present in cells exposed to succinic semialdehyde with those found in cells treated identically but not exposed to the inducer. The NAD-dependent succinic-semialdehyde dehydrogenase is the only enzyme induced by succinic semialdehyde, attaining a specific activity approximately equal to that in cells growing exponentially on p-hydroxyphenylacetate (Table 2). The preceeding enzymes ofp-hydroxyphenylacetatemetabolism, the dehydrogenase and decarboxylase, are clearly not induced by succinic semialdehyde (Table 3). The data presented here indicate that the NADdependent succinic-semialdehyde dehydrogenase is regulated independently of the other enzymes for p-hydroxyphenylacetate degradation. Work in progress on the regulation of the earlier reactions of the p-hydroxyphenylacetate pathway in various strains of E. coli support this hypothesis. Although the gene appears to be functionally isolated from the other genes of the p-hydroxyphenylacetate pathway, we have no evidence concerning its physical relationship to these genes. Mutants lacking succinic-semialdehyde dehydrogenase cannot be selected directly from wildtype cells, since the aldolase reaction which produces succinic semialdehyde also produces pyruvate (Fig. 1). The regulation of succinic-semialdehyde dehydrogenase in species of Pseudomonas which degrade p-hydroxyphenylacetate has not been studied. Barbour and Bayly [19], in their study of the regulation of p-hydroxyphenylacetate catabolism in Pseudornonas putida, dealt only with the earlier reactions of the pathway. The report that P. putidu strain U possesses both NAD-dependent and NADP-dependent activity

M. I. Donnelly and R. A. Cooper

when grown on p-hydroxyphenylacetate [6] should be re-examined in the light of the research presented here. As only NADP-linked activity is found in other bacteria when grown on p-hydroxyphenylacetate [6, 201, it is possible that P. puridu U also possesses two distinct dehydrogenases, as occurs in E. coli B. It should be recalled that Nirenberg and Jakoby purified two succinic-semialdehyde dehydrogenases from a species of Pseudomonus grown on y-hydroxybutyrate, one NADP-dependent, the other NAD-dependent [7]. The NAD-dependent enzyme exhibited K, values, pH optimum and relative rate of NADP reduction very similar to those described here for the enzyme from E. coli grown on p-hydroxyphenylacetate. These considerations beg the much more difficult question of whether or not the NAD-linked succinic-semialdehyde dehydrogenase is truly an enzyme of the p-hydroxyphenylacetate pathway or merely has been recruited [21] from some other pathway because of its specificity of induction. Again, knowledge of the physical relationship between the dehydrogenase gene and those for the rest of the pathway’s enzymes will shed light on this problem. Finally, research into the degradation of y-aminobutyrate should be tempered by the knowledge that growth on y-aminobutyrate may fortuitously induce a second succinic-semialdehyde dehydrogenase, which also can act with NADP as cosubstrate. M 1.D was a recipient of a Fulbright-Hays Grant for Graduate Study Abroad

561

REFERENCES 1. Cooper, R. A. & Skinner, M. A. (1980) J . Bacteviol. 143, 302 - 306. 2. Noe, F. F. & Nickerson, W. J. (1958) J. Bucteriol. 75, 674681. 3. Dover, S. & Halpern, Y. S. (1972) J . Bucreriol. IOY, 835-843. 4. Dover, S. & Halpern, Y. S. (1972) J . Bucreriol. 110, 165-170. 5. Zaboura, M. & Halpern, Y. S. (1978) J . Bucreriol. 133, 447451. 6. Sparnins, V. L., Chapman, P. J. & Dagley, S. (1974) J . Bucteriol. 120, 159 - 167. 7. Nirenberg, M . W. & Jakoby, W. B. (1960) J . Bid. Chem. 235, 954 - 960. 8. Jakoby. W. B. & Scott, E. M. (1959) J . Biol. Chem. 234, 937940. 9. Jakoby, W. B. (1962) Meihods Enzymol. 5, 765-778. 10. Padmanabhan, R. & Tchen, T. T. (1969) J . Bucteriol. 100, 398 - 404. 11. Callewaert, D. M., Rosemblatt, M. S., Suzuki, K. & Tchen, T. T. (1973) J . Bid. Chem. 248, 6009-6013. 12. Miller, J. H. (1972) Experiments in Molecular Genetics, p. 431, Cold Spring Harbor Laboratory, New York. 13. Hartree, E. F. (1972) Anal. Biochem. 48, 422-427. 14. Andrews, P. (1964) Biochem. J . 91, 222-233. 15. Weber, K. & Osborn, M. (1969) J . Biol. Chem. 244,4406-4412. 16. Cleland, W. W. (1967) Adv. Enzymol. 29, 1 - 32. 17. Metzer, E., Levitz, R. & Halpern, Y. S. (1979) J . Bucteriol. 137, 1111-1118. 18. Ornston, L. N. & Parke, D. (1976) C u n . Top. Cell. Regul. 12, 209-262. 19. Barbour, M. G. & Bayly, R. C. (1977) Biochem. Biophys. Res. Commun. 76,565 - 571. 20. Spamins, V. L. & Chapman, P. J. (1976) J . Bucteriol. 127, 362 - 366. 21. Dagley, S . (1975) Essuys Biochem. 11, 81-138.

M . I . Donnelly, Department of Biochemistry, University of Minnesota, College of Biological Sciences, 1479 Gortner Avenue, Saint-Paul, Minnesota, USA 55108

R. A . Cooper, Department of Biochemistry, School of Biological Sciences, University of Leicester, University Road, Leicester, Great Britain, LE1 7RH