NAD-linked, GSH- and factor-independent aldehyde dehydrogenase of the methylotrophic bacterium, Hyphomicrobium X

ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS Vol. 271, No. 1, May 15, pp. 240-245,1989

NAD-Linked,

GSH- and Factor-Independent Aldehyde Dehydrogenase the Methylotrophic Bacterium, Hyphomicrobium X PAULUS

A. POELS

AND

JOHANNIS

of

A. DUINE’

Department qf’ Microbiology and Enzymology, Delft University of Technology, Julinduan 67, ;?6.# BC Delft, The Netherlads Received September 27,1988, and in revised form January

18,1989

Cell-free extracts of Hyphomicrobium X showed NAD-dependent aldehyde dehydrogenase activity, provided that NAD addition preceded that of aldehyde. Activity was lost rather rapidly, especially during purification attempts, but this could be partially masked by including a time-dependent restoration step with thiol compounds in the protocol. The nature of the assay buffer appeared to be critical and stimulation occurred on incorporation of K’ ions in the mixture. An even higher specific activity could be achieved by 1,4-dithiothreitol (DTT) treatment of the preparation, followed by removal of DTT, and assaying in the absence of thiol compounds under anaerobic conditions. Exposure of such a preparation to O2 led to a significant decrease in activity within a couple of hours. Immediate inactivation occurred on addition of HzOs, but this could be prevented completely by prior addition of NAD. Since GSH does not participate in the reaction and no stimulating factor was detected, the role of thiol compounds is most probably confined to restoration or prevention of damage to an OZ-sensitive, necessary thiol group. Since the same features were found for cell-free extract as for the partially purified enzyme, only one enzyme type seems to be present. Although the enzyme is a general aldehyde dehydrogenase, the kinetic parameters and the specific activity of the cell-free extract for formaldehyde indicate that it may play a role in formaldehyde dissimilation by Hyph.omicrobium X. The NAD-linked, GSH- and factor-independent aldehyde dehydrogenase described here appears to be different in several respects from the formaldehyde dehydrogenase of Pseudomonas putida (EC 1.2.1.46) (despite showing similar behavior toward coenzymes and factors) but resembles the aldehyde dehydrogenase from baker’s yeast (EC 1.2.1.5). lc’1989 Academic press. I~C

Microbial dissimilation of methane, methanol, or methylated compounds frequently proceeds via a metabolic pathway in which a conversion of formaldehyde into formate is involved. Although several formaldehyde oxidoreductases have been found in cell-free extracts (see Ref. (1) for a compilation), the in vivo significance of these enzymes is not always clear. For instance, Hyphomicrobium X contains a dyelinked aldehyde dehydrogenase (2) but the enzyme has a low specific activity for form-

1 To whom correspondence

should be addressed

0003-9861/89 $3.00 Copyright All rights

0 1989 by Academic Press, Inc. of reproduction in any form reserved.

aldehyde in cell-free extracts, is synthesized constitutively, and its kinetic parameters seem unfavorable for keeping formaldehyde at a low level in the cell (3). Similarly, although methanol dehydrogenase (EC 1.1.99.8) oxidizes formaldehyde very efficiently in the in vitro assay, several arguments can be put forward to reject the idea that this enzyme plays a substantial role in formaldehyde dissimilation (3, 4). On the other hand, although an NAD(P)dependent formaldehyde dehydrogenase has not been found in this bacterium (2), formaldehyde dissimilation to formate must occur in this type of methylotroph. 240

NAD-LINKED

ALDEHYDE

DEHYDROGENASE

To escape from this dilemma, a route comprising tetrahydrofolate-linked enzymes has been proposed (5). Many different types of NAD(P)-linked (form)aldehyde dehydrogenases have been described (6). Variability exists in substrate specificity, coenzyme specificity, stimulation or inhibition of the activity by certain salts, and the dependency on glutathione (reduced) GSH” or other (co)factors. Recently a novel type of formaldehyde dehydrogenase (NAD linked, GSH independent), involved in degradation of methylated compounds, was discovered (7). The activity of this enzyme depends on the addition of a factor with unknown structure to the assay. Since the existence of such an enzyme might have been overlooked in methylotrophic bacteria, an investigation was started on Hyphomicrobilnm X. It appeared, however, that such an enzyme was absent and instead of this, another type of NAD-dependent aldehyde dehydrogenase was found by accident, as reported here. The similarity of this enzyme with other (form)aldehyde dehydrogenases and its significance for formaldehyde dissimilation by Hyph,omicrobium X, are discussed. MATERIALS

AND

METHODS

Materials. NAD, @-mercaptoethanol, GSH, L-(+)cysteine, and phenazine methosulfate were from Merck. Coomassie brilliant blue G-250 was from Serva. 1,4-Dithiothreitol (DTT), formaldehyde dehydrogenase (NAD linked, GSH dependent) (EC 1.2.1.1) from Candida boidinii, and formaldehyde dehydrogenase (NAD linked, GSH independent) (EC 1.2.1.46) from Pseudomonas putida were from Sigma. Formate dehydrogenase (NAD linked) (EC 1.2.1.2) from yeast was from Boehringer. Cell-~free extruct. Hyphomicrobium X was cultured on a mineral medium, supplemented with 0.3% (v/ v) methanol, 0.3% (v/v) ethanol, 0.5R (w/v) sodium formate, or 0.5% (w/v) sodium acetate and harvested as described (8). Frozen cells were mixed with an buffer, pH 9.0, conequal volume of 0.1 M Tris-HCl taining 0.5% (v/v) Triton X-100 and the mixture was passed twice through a French pressure cell at 110

a Abbreviations GSH, glutathione no)ethanesulfonic

used: DTT, 1,4-dithiothreitol; (reduced); Ches, 2-(cyclohexylamiacid.

FROM

Hyphomicrobium

X

241

MPa. The suspension (its viscosity was lowered by adding deoxgribonuclease) was centrifuged at 48,000~ for 30 min at 4°C (all subsequent manipulations were performed at 4”C), giving the cell-free extract. Purtial r~zynre pwi~cchm. After adding (NH,),SO1 to 45%’ (w/v) saturation to the cell-free extract prepared from methanol-grown cells (10 g wet u-t), the precipitate was collected by centrifugation (48,OOOy,10 min). The precipitate was dissolved in and dialyzed against (400 volumes, 16 h) 20 mM potassium phosphate, pH 7.0. The dialysis residue was centrifuged (48,OOOg, 10 min) and the supernatant applied to a DEAE-Sepharose fast flow (Pharmacia) column (10 X 1 cm) equilibrated with 20 mM potassium phosphate, pH 7.0. The column was eluted with 0.1 M potassium phosphate, pH 7.0. Active fractions were pooled and dialyzed against (50 vol, 16 h) 20 mM potassium phosphate, pH 7.0. The dialysis residue was concentrated by pressure filtration. IU’T treatnrr~~t. Enzy-me preparations were made anaerobic by degassing on a vacuum line and flushing with N, for 20 min. Subsequently, incubation occurred with DTT (10 ITIM) for 1 h and the mixture was dialyzed under anaerobiosis against (500 vol, 16 h) 20 mM potassium phosphate, pH 7.0. Enzyme assays and kinetics. The enzyme was assayed at room temperature by measuring NADH, formation at 340 nm. Two different methods were used: method A for comparative purposes during purification and to show the effect of preincubation with DTT; method B for kinetic experiments where thiohemiketal formation from the substrate and DTT should be avoided and to show the effect of O2 on enzyme subjected to the DTT treatment. Method A: To 600 kl 0.1 M sodium pyrophosphate buffer, pH 9.0, containing 0.17 M KCI, 100 ~10.1 M DTT, 100 ~1 20 mM NAD, and 100 ~1 enzyme solution were added in that order. The mixture was incubated for 12 min and the reaction was started by adding 100 ~1 120 rIIM formaldehyde. Method B: In this assay, all stock solutions were made anaerobic by degassing and flushing with NB, while the manipulations to fill the cuvette were carried out in an anaerobic cabinet. The cuvette was stoppered with Suba seals and transferred to the spectrophotometer and the reaction started by adding the anaerobic formaldehyde solution via a syringe provided with a hypodermic needle. To 600 ~1 0.1 M sodium pyrophosphate buffer, pH 9.0, 100 ~1 20 rIIM NAD, 100 ~1 enzyme solution, 100 ~1 HaO, and 100 ~1 12 mM formaldehyde were added in that order. Calculations were made using a molar absorption coefficient for NADH of 6220 M ’ cm-l at 340 nm (9). For kinetic studies in assay system B, substrate concentrations were varied while that of NAD was kept constant. The apparent kinetic parameters were determined with the direct linear plot method (10). To check the stoichiometry of the reaction, NADH, pro-

242

POELS

AND

duction was compared to product formation (formate and benzoate) using formaldehyde and benzaldehyde as a substrate. Ar~alytical methods. Protein determinations were performed according to Bradford (11) with bovine serum albumin as a standard. Formaldehyde and acetaldehyde were determined according to Avigad (12); formic acid was determined with formate dehydrogenase (13); benzoic acid was determined with HPLC on a reversed-phase column (C,, Radial PAK) (Waters) with 25% (v/v) methanol, containing 0.4% (v/v) H3P04, as an eluant, at a flow rate of 1.5 ml/min. The effluent was monitored with a Hewlett-Packard 1040A photodiode array detector. The system was calibrated with a known amount of benzoic acid. RESULTS

Properties

AND

DISCUSSION

of the Enzyme

Cell-free extract from methanol-grown cells showed NAD-linked formaldehyde dehydrogenase activity, provided that NAD addition preceded that of formaldehyde. The nature of the buffer appeared to be important: most buffers tested at pH 9.0 (Ches, NH,Cl, and Tris) showed far less activity than sodium pyrophosphate buffer while sodium borate buffer showed no activity at all. A pH optimum of 10 was found for assay A. The enzyme lost activity rapidly if assayed with assay A in the absence of DTT. This suggests a damaging effect of O2 on the enzyme, in accordance with the observation that if cell disruption occurred under anaerobic conditions, much higher activities were found. All these peculiarities may be the reason this enzyme has been overlooked so far in cell-free extracts of Hyphomicrobium X (2). The results for cells cultured on other carbon sources were variable: ethanol- and acetate-grown cells showed activities similar to those of methanol-grown cells whereas formate-grown cells showed no activity. Also during purification steps, very severe losses of activity occurred if measured with method A in the absence of DTT. Even when measured with the complete method A, the yield is low (Table I). A better recovery was attained when the final preparation underwent DTT treatment and was assayed with method B. This is most probably due to restoration of damage to an 02-sensitive thiol group in the en-

DUINE

zyme as the reactivation with DTT proceeded in a time-dependent manner (Table II, Fig. 1). Although Fig. 1 suggests a maximal level of restoration, activities measured with method A were always lower than those with method B. It should be realized, however, that DTT and formaldehyde will form thiohemiketal adducts so that the genuine substrate concentration for method A is in fact unknown. Furthermore, the presence of the thiol compound may influence the reaction also in other respects, as was apparent from the strong activating effect of KC1 in assay method A, while this effect was absent in method B. The stimulation is probably related to the K+ ion since no stimulation was observed with MgC12. Other features, given in Table II, are also in accordance with an essential thiol group in the enzyme. This group is effectively shielded by binding with NAD, as shown by the observations that HzOz or substrate addition before NAD addition blocked the enzyme almost completely but not if NAD was added first. Restoration of damage induced by O2 could be partially restored by DTT or P-mercaptoethanol but less effectively with cysteine or GSH. From the substrate specificity presented in Table III, it appears that the enzyme is a general aldehyde dehydrogenase. Corresponding alcohols, even at high concentrations, were not substrates. Severe substrate inhibition was observed, as shown in Fig. 2, for formaldehyde. Therefore, when testing substrate specificity with method B, substrate concentrations outside the range of inhibition were used. Since this method is reliable with respect to real substrate concentrations, propionaldehyde appears to be the best substrate for this enzyme. Keeping the NAD concentration constant and varying the substrate concentration in assay B, it was found that K:, and V’ for formaldehyde were 0.55 and 234, and for propionaldehyde they were 0.06 mM and 141 nmol mini’ (mg protein))‘, respectively. On varying the NAD concentration, keeping the formaldehyde concentration constant at 1.2 mM, the K& and V’ found for NAD were 0.35 mM and 144 nmol mini’ (mg protein))*, respec-

NAD-LINKED

ALDEHYDE

DEHYDROGENASE TABLE

FROM

Hyphomicrobium

X

243

I

PARTIAL PURIFICATION OF THE ENZYME

Preparation Cell-free extract (NH,)$O, precipitate DEAE-Sepharose eluate

Activity (pm01 min ‘)

Protein (md

Specific activity (nmol mm’ rng-i)

Recovery (So)

Purification factor

111 84 29 (75)

2307 1260 344 (311)

48 67 84 (241)

100 76 26 (68)

1.0 1.4 1.8 (5.0)

Note. Preparations were assayed with method A. For the DEAE-Sepharose eluate preparation, in parentheses refer to assay method B and DTT treatment. The recovery and purification factor were calculated using the values of the cell-free extract determined in the same way.

tively. NADP could not replace NAD. The reverse reaction (using NADHJformate, NADH/acetate, or acetyl-CoA) was not detected, either at pH 7 or at pH 9. No dismutase activity was present since stoichiometric product formation occurred with formaldehyde (1883 nmol formate; 1868 nmol NADH2) and benzaldehyde (1630 nmol benzoic acid; 1636 nmol NADH,). Since substrate specificity and specific features were identical for the cell-free extract and the partially purified enzyme, Hyphomicrobium X seems to have one sole NAD-linked aldehyde dehydrogenase (the dye-linked aldehyde dehydrogenase remained in the supernatant after ammonium sulfate precipitation). Comparison with Other NAD-Linked (Form)aldehyde Dehydrogenases The well-known NAD-linked formaldehyde dehydrogenase (EC 1.2.1.1) specifically requires GSH in the reaction. So from the results of assays of cell-free extracts with and without the incorporation of GSH, the presence of this enzyme has been claimed in several methylotrophic bacteria (1, 6). Although GSH also has a positive effect on the activity of cell-free extracts of Hyphomicrobium X, it is clear that the enzyme from this organism is quite different since the thiol compounds required in method A do not participate in the reaction but prevent or restore damage caused by OZ to the enzyme. Two types of NAD-linked, GSH-independent formaldehyde dehydrogenase

the values in this case

have been found to be involved in C1 metabolism. The first type requires an as yet unidentified factor for activity, as was reported for the enzymes from Rhodococcus erythropolis (7) and Methylococcus capsulatus, strain Bath (14) (the factors required differ in many properties in the two cases). For that reason alone, it is obvious that the enzyme from Hyphomicrobium X is quite different. The second type (EC 1.2.1.46) has been detected in Pseudomonas strains (15, 16) and does not require an extra cofactor. The enzyme is involved in the dissimilation pathway of methylated compounds. At first sight, it seems similar to the Hyphomicrobium enzyme since they show the same behavior with respect to cofactor requirements. On close examination, however, important differences appear to exist. From the literature (15-1’7) and/or from this work (results not shown), it appears that the P. putida enzyme has the following aberrant properties: it does not convert higher aldehydes than acetaldehyde; higher alcohols are substrates (at high concentration); the enzyme is stable and does not require thiol compounds in the assay; the order of substrate addition is not critical; inhibition with HeOz proceeds rather slowly while formaldehyde is a competitive inhibitor for this process. In view of these differences and the broader substrate specificity, it seems appropriate to indicate the enzyme from Hyphomicrobium X as an NAD-linked, GSH- and factor-independent aldehyde dehydrogenase. Several types of NAD-linked, GSH- and factor-independent aldehyde dehydroge-

244

POELS TABLE

AND

DUINE

II

SOME PROPERTIES OF THE ENZYME: THE EFFECT OF DIFFERENT THIOLS Activity Method A Condition Assayed without preincubation Without DTT @-Mercaptoethanol (20 mM) instead of DTT L-Cysteine (20 mM) instead of DTT GSH (20 mM) instead of DTT Without KC1 Conditions for inhibition Formaldehyde before NAD addition Hz02 (1 mM), applied before NAD addition H,O, (1 mM), applied after NAD addition HgClz (10 mM)

(% ) Method B

13 6 0

100

4

8

8

10 Time

12

1,

(min.)

FIG. 1. The effect of preincubation with DTT. Partially purified enzyme was incubated for different times with DTT and assayed with method A (activities are expressed as a percentage of the highest activity which could be attained).

67 17 37

0

2

0 12 100 0

Note. Effects were measured by assaying partially purified enzyme with modified methods A and B or by treating the enzyme, as indicated. For the measurements performed with method B, the enzyme preparation underwent the DTT treatment. Activities are indicated as a percentage of the values obtained under the usual conditions (84 and 241 nmol formaldehyde oxidized min’ (mg protein))’ with methods A and B, respectively).

its substrate specificity is not restricted to formaldehyde, a role in C1 metabolism cannot be excluded. Cell-free extracts prepared under anaerobic conditions showed an activity of 0.1 pmol formaldehyde oxidized mini’ (mg protein)-’ with method A as well as with B. This value is low but not insignificant and could even be higher since the assay was not optimized. In addition, levels of formate dehydrogenase in serine-type methylotrophs are in a similar

TABLE

nases have been found in nonmethylotrophic organisms. One of these (EC 1.2.1.5) has been described for baker’s yeast (18). This enzyme resembles that of Hyphomicrobium in many properties: it becomes specifically activated with K+ ions; the substrate specificity is similar, showing a preference for propionaldehyde; rapid loss of activity occurs; and the presence of thiol compounds is required in the assay. Further investigations on the physicochemical properties are necessary to define the precise extent of resemblance. Physiologicul Signi&cance Although the enzyme is synthesized during growth on several carbon sources and

III

SUBSTRATE SPECIFICITY Activity Substrate Formaldehyde Acetaldehyde Propionaldehyde Butyraldehyde Octanal Benzaldehyde

(%)

Method A

Method B

100 55 50 28 13 66

54 68 100 68 32 49

Note. Measurements were performed with the substrates having concentrations as that of formaldehyde in methods A and B. Activities are expressed as a percentage of the values obtained with the best substrate.

NAD-LINKED

0

1.0

ALDEHYDE

2.0 (Formaldehyde)-’

DEHYDROGENASE

4.0

3.0 (mM)-’

FIN. 2. Initial reaction rates with varying formaldehyde concentrations. Activities of partially purified enzyme were determined with method B, using varying concentrations of formaldehyde. The data are presented in the form of a Lineweaver-Burk plot.

range (0.1 and 0.4 pmol formate oxidized min-’ (mg protein)-’ (19,20). Since also the K,, value for formaldehyde is favorable (compared to that of the dye-linked aldehyde dehydrogenase (2)), the enzyme could be involved in cellular formaldehyde dissimilation. Anyhow, in contrast with what has been reported (2), levels of NAD-dependent formaldehyde dehydrogenase activity are not zero so that the occurrence of the enzyme should be taken seriously in physiological considerations on Hyphomicrobium X. REFERENCES 1. ZATMAN, L. J. (1981) i?l Microbial Growth on C, Compounds (Dalton, H., Ed.), pp. 42-54, Heyden, London.

FROM

Hyphomicrobium

X

245

2. MARISON, I. W., AND ATTWOOD, M. M. (1980) J. Gen. Microbid. 117,305-313. 3. ATT~OOD, M. M., AND QUAYLE, J. R. (1984) irr. Microbial Growth on C, Compounds (Crawford, R. L. and Hanson, R. S., Eds.), pp. 315-323, American Society for Microbiology, Washington, DC. 4. GROENEVELD, A., DIJKSTRA, M., ANI) DUINE, J. A. (1984) FEMS Microbiol. L&t. 25,311-314. 5. MARISON, I. W., AND ATTWOOD, M. M. (1982) J. Gen. Microbid. 128,1441-1446. 6. ANTHONY, C. (1982) The Biochemistry of Methylotrophs, pp. 187-193, Academic Press, London. 7. EGGEI,ING, L., ANT) SAHM, H. (1985) Eur. J. Biothem. 150,129-134. 8. DUINE, J. A., FRANK, J., AND WESTERLING, J. (1978) Biochim Biophys. Acta 524, 277-287. 9. DAWSON, M. C., ELLIOTT, D. C., ELLIOTT, W. H., AND JONES, K. M (1986) Data for Biochemical Research, pp. 122-123, Clarendon Press, Oxford. 10. EISENTHAL, R., ANI) CORNISH-BOWDEN, A. (1974) Biochem. J. 139, ‘715-720. 11. BRADFORD, M. M. (1976) And Biochem. 72, 248254. 12. AVI(:AD, G. (1983) Ad. Bioch~~ 134,499-505. 13. HOPNER, T., AND KNAPPE, J. (1974) in Methods in Enzymatic Analysis, (Bergmeyer, H. U., Ed.), 2nd cd., Vol. 3, pp. 1551-1554, Academic Press, London. 14. STIRLING:, D. I., AND DALTON, H. (1978) -1 Gm. Microbiol. 107, 19-29. 15. HOHNI~OSEK, W., Ossw~r,n, B., AND LINGENS, F. (1980) Hoppe S&rk 2. Physinl. Chem. 361, 1763-1766. 16. ANDO, M., YOSHI~~OTO, T., O~~JSI~I, S., RIKITAKE, K., SHIBATA, S., AND Tx:R~J, D. (1979) J. Biothem. 85,1165-1172. 17. OG~TSHI,S., ANDO, M., AND TXJR~, D. (1986)A~~ri~. Bid. Chem.. 50,2503-2507. 18. BLACK, S. (1951) Arch. Biochem. Biophys. 34, 8691. 19. JOHNSON, P. A., AND QIJAYLE, J. R. (1964) Riothem. J. 93,281-290. 20. ROITSCH, T., ANI) STOI.P, H. (1985) Arch. Microbid. 143,233-236.