Carbonyl oxygen exchange evidence of imine formation in the ... - NCBI

2 downloads 0 Views 940KB Size Report
... evidence that this rapid exchange is due to an imine form of the enzyme-re- .... Under these condi- tions quantitative conversion of ammonia takes place and.
Proc. Nati. Acad. Sci. USA Vol. 81, pp. 2747-2751, May 1984 Biochemistry

Carbonyl oxygen exchange evidence of imine formation in the glutamate dehydrogenase reaction and identification of the "occult role" of NADPH (isotope exchange kinetics/mechanism of glutamate dehydrogenase/reactions of a-ketoglutarate and ammonia/ nonreductive role of NADPH)

HARVEY F. FISHER

AND

TENKASI S. VISWANATHAN

Laboratory of Molecular Biochemistry, Department of Biochemistry, University of Kansas School of Medicine and the Veterans Administration Medical Center, Kansas City, MO 64128

Communicated by Myron L. Bender, January 20, 1984

Despite a number of experimental findings that are in accord with such a mechanismt direct evidence of the existence of a kinetically competent ketimine intermediate has not yet been obtained. If such an imine were to be formed at any point in the reaction path even for a fleeting instant, exchange of the keto oxygen atom with the solvent must occur since the overall reaction is known to be reversible:

ABSTRACT Although an imine intermediate has long been postulated as participating in the reaction catalyzed by glutamate dehydrogenase (EC 1.4.1.4), direct evidence for a kinetically competent intermediate of this kind has not heretofore been found. We have sought such evidence by studying the exchange of the carbonyl oxygen atom of a-ketoglutarate in a variety of binary, ternary, and quaternary enzyme complexes. We have found that the time course of this exchange is biphasic when the enzyme, a-ketoglutarate, NADPH, and ammonia are all present initially and that the rapid initial phase ends when ammonia is depleted. We present evidence that this rapid exchange is due to an imine form of the enzyme-reduced-coenzyme-substrate-ammonia complex. Formed very rapidly but in very small amounts, this imine can undergo one of two competing fates: (i) hydrolytic reversal to form carbonyl-exchanged a-ketoglutarate with regeneration of ammonia, and (ii) an internal hydride transfer converting the iminoglutarate to glutamate, whereby ammonia is consumed. The agreement of the amplitudes of rapid 180 exchange with predictions based on direct transient-state spectroscopic kinetic studies supports the identity of an enzyme-NADPH-a-iminoglutarate complex as an obligatory intermediate on the enzyme-catalyzed reaction path. The corresponding enzyme-aiminoglutarate binary complex (previously suggested as an intermediate) is formed at a rate that is less than 1/1000th of that of the NADPH-containing complex shown here, and it therefore lacks kinetic competence. This finding points up an important catalytic role for NADPH that does not involve its obvious function as a hydride donor and is distinctly separate from this role. In the case of the glutamate dehydrogenasecatalyzed reaction, this "occult role" clearly involves the induction of ketimine formation on the enzyme surface.

OH 11

NH3 +

-C-co2

-C-co° NH2

H20

NH

PII

-C- C02

0

11 NH3 +--°

OH

-C-co2

2

H20

NH2

(The filled-in O's represent labeled oxygen atoms, while the O's represent the unlabeled species.) To obtain direct evidence for a-ketimine formation in the enzyme-catalyzed reaction, and to determine the molecularity of the complex or complexes involved in its formation, we have measured the rate of exchange of the carbonyl oxygen atom of a-ketoglutarate during the course of the complete reaction and in a variety of conceivably competent complexes. The results of that study are reported here. open

EXPERIMENTAL Materials

Glutamate dehydrogenase (E; EC 1.4.1.4) catalyzes the reductive amination of a-ketoglutarate (K) by NADPH (R). Following the early suggestion of von Euler et al. (1), it has generally been assumed that the mechanism of this reaction involves the initial formation of enzyme-bound a-iminoglutarate from a-ketoglutarate and ammonia (N) followed by the reduction of that imine by NADPH in a subsequent step:

Bovine liver glutamate dehydrogenase obtained as an ammonium sulfate precipitate from Boehringer Mannheim was filtered through Norit A (activated charcoal) as described (5) and dialyzed overnight against 0.1 M potassium phosphate buffer (pH 7.6) before use. The enzyme, NADPH, and NADP concentrations were determined spectrophotometrically, using molar absorption coefficients of = 54,400, E340 = 6300, and 6259 = 17,800 M-1 cm-', respectively. 6280

o

NH3

NH

C -C-- COC

E

E

NADPH

NH2

- COHC_

+ NADP

Abbreviations: E, glutamate dehydrogenase (enzyme); R, NADPH (reduced); 0, NADP (oxidized); K, a-ketoglutaric acid; N, ammonia in all forms; G, L-glutamate. tThe glutamate dehydrogenase catalysis of the reduction of a cyclic imine by NADPH (2), the observation of an isotope effect in the glutamate dehydrogenase reaction when L-[2-2H]glutamate was employed (3), and the trapping of a-iminoglutarate on the enzyme surface by NaBH4 reduction in the absence of NADPH (4) are evidence in support of an imine mechanism for the glutamate dehydrogenase reaction.

[1]

E

H20

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. 2747

Proc. NatL Acad Sci. USA 81

Biochemistry: Fisher and Viswanathan

2748

(1984)

Methods 60

Measurement of the NADPH Reacted. The amount of NADPH (I'1200 AM) oxidized, A[R] in the presence of enzyme (-250 ,uM), a-ketoglutarate (-300 MM), and small amounts of ammonia (2500 The reaction took place in 0.1 M potassium phosphate buffer (pH = 7.5 + 0.1) at 250C. *A plot of the rate constant against the enzyme concentration at fixed concentrations of K and N was linear with a slope of 11.8 M-1

4

0

Proc. Nati Acad Sci USA 81 (1984)

s-1.

most of the enzyme and a-ketoglutarate exist in the form of the ERK complex. We have shown elsewhere that in the presence of ammonia an ERKN complex is formed and that that complex is an obligatory intermediate on the reaction pathway (10). The results described up to this point, then, lead to the conclusion that this ERKN complex itself, its formation, or its dissociation must be responsible for the rapid first-phase exchange phenomenon. Chemical Mechanism of the Rapid Exchange Phenomenon. Two possible mechanisms are consistent with the facts described thus far: (i) an ammonium ion-catalyzed formation of an a-gem-diol; and (ii) a-ketimine formation. The gem-diol mechanism is shown below: 0 11

H20

+ -C -

H

H

0

/

[3]

C\

11

H20

-C-C 2

It should be noted that catalysis by ammonia, ammonium ion, or both is an essential component of this mechanism, since the phenomenon does not occur in its absence. The following points argue strongly against such a mechanism. (i) If both hydration and dehydration steps occur on the enzyme surface, both processes would be expected to be stereospecific, since the analogous addition of ammonia to enzyme-bound a-ketoglutarate produces only L-glutamate. As can be seen in Scheme 3, the dehydration step in such a

0 U) Ir 0-

0

0

I4

0

300

600

900

1200

PRODUCT, 0 OR G, uM FIG. 3. Effect of externally added products, L-glutamate (L) or NADP (0), on the amplitude of the first phase of the 180 exchange reaction described for Fig. 1. For the experiment in which L-glutamate was varied, E = 256 gM, R = 1230 AM, K = 290 ,uM, and NH3 = 35 AM. For the experiment in which NADP was varied E = 259 ALM, R = 1298 ,M, K = 310 /LM, and NH3 = 115 AM.

2750

Biochemistry: Fisher and Viswanathan

Proc. NatL Acad Sci. USA 81

H2180 failed to incorporate heavy oxygen into a-ketoglutarate. This is not surprising, since Hochreiter et al. (4) have reported rate constants of 4.5 x 10-4 M-1 s-1 and 5.45 x lo-3 S-1, respectively, for the formation and dissociation of a-iminoglutarate from K and N at 00C in 0.2 M ammonium acetate solution of pH 9.4; and they have also reported only a 4.5-fold increase in the rate of this iminoglutarate formation when 71 utM enzyme was included in the reaction mixture. The presence of enzyme increased the equilibrium amount of a-iminoglutarate only slightly. These observations are consistent with our finding that the addition of enzyme, even at very high concentrations, leads only to a small increase in the rate of carbonyl oxygen exchange in K in the absence of R. Thus, the enzyme alone neither shifts the equilibrium in favor of a-iminoglutarate nor accelerates its attainment significantly. When NADPH is included in the reaction mixture, Brown et al. (10) have shown that the dissociation constant of ammonia from the quaternary ERKN complex is 0.045 M at 20'C in pH 7.6 buffer, and they have estimated the off rate constant for this step to be >370 s-1. If, as Rife and Cleland (13) have proposed in their mechanism, ammonia adds to the a-ketoglutarate directly in the ERK complex, this implies a >300-fold increase in the equilibrium iminoglutarate concentration and a >10,000-fold increase in the rate of keto imine interconversion brought about by NADPH. Indeed, Rife and Cleland's suggestion that the EK binary complex could be considered to be a dead-end complex for all practical purposes now appears to be a very reasonable viewpoint. Partitioning of the Iminoglutarate (ERKN) Intermediate. Returning to the mechanism postulated in Fig. 4, it may be noted that any given imine intermediate complex must undergo one of two possible fates, a reversible hydrolysis to form ammonia and carbonyl-exchanged a-ketoglutarate or, alternatively, an irreversible internal redox reaction that removes ammonia from the solution. Thus, the ratio of moles of carbonyl oxygen exchanged to moles of ammonia consumed (A[K*I/A[N]) depends on the relative fluxes through these two alternative pathways. In a variety of exchange experiments in which the first-phase amplitudes correspond to less than 20% of complete exchange, we observe values of A[K*]/A[N] between 1.2 and 2.2. Thus, the imine traverses the hydrolysis pathway at least several times more frequently than it undergoes reduction. This is in qualitative agreement with the estimate from transient-state kinetic studies (10) that k2/k3 - 6, assuming that precisely the same reaction

process would remove only the solvent-derived oxygen atom and no exchange would occur. (ii) If the ammonia-catalyzed hydration were to occur on the enzyme surface with subsequent release of the gem-diol into solution, the dehydration would remove one of the two hydroxyl groups randomly and exchange could result. However, we have previously reported (11) the rate of the diol-toketone reaction of a-ketoglutarate, and that rate is too slow to account for the observed rate of exchange by 2 orders of magnitude. It is also useful to note in this connection that adihydroxyglutarate is not a substrate for this enzyme (11). We are left then with the remaining possibility-a mechanism involving ketimine formation. The mechanism shown in Fig. 4 does account for the experimental findings and is consistent with other known features of the enzyme, its complexes, and its catalytic reaction. The portion of the mechanism enclosed by the curved (dashed line) box simply leads to a very rapid exchange of the a-keto oxygen atom of K without any loss of ammonia. The internal redox step of the ER-a-iminoglutarate complex shown in the vertical (dotted line) box, however, does remove ammonia from the solution. This step is for all practical purposes irreversible in the absence of added 0 and G under the range of conditions used in the experiments. Therefore, when the ammonia has been completely removed by the redox step, the rapid exchange portion of the mechanism also stops, since the first step in the curved box can no longer proceed. Thus, the biphasic behavior of the exchange and the dependence of the amplitude of the rapid first phase on the initial ammonia concentration are readily explained. Support for the existence of ER-bound a-iminoglutarate as the central intermediate in this mechanism is provided by our recently reported finding that glutamate dehydrogenase does indeed catalyze the reduction of a cyclic imine by NADPH (2): NH3-(CH2)3 CO-CO2 =

CO°

+

H20

H

lS-COj

[4]

H+-C;

+ NADPH +

CO- + NADP+

H2 n

I

H

(1984)

Comparison of EKN and ERKN Complexes. Frieden (12) has reported that free enzyme incubated with K and N in

H20 /

/

/

//

I7

f

+

71-

OH

NH 11 -C- COi

k -C-CO2 -.. INH ,,NH2

*f*

OH 1

k4 k3

N 2\

NH2

I k55

" /~~~~~~~~~~~~ /

IN

__

/ N~~~~~~~~~

I

LI. ____ _ I/-C-CO 2 + NH3

NH3 * H-C-CO :

I

*

0

III

FIG. 4. Competitive routes for carbonyl oxygen exchange and reduction in the mechanism of the glutamate dehydrogenase reaction. The dashed curved box encloses the exchange route, the dotted box encloses the reductive route. Other symbols are as defined in previous schemes.

Biochemistry: Fisher and Viswanathan steps are being measured in both cases. This agreement is sufficiently good to assure us that the ERKN complex contains bound a-iminoglutarate (or its corresponding carbinolamine) and that this complex is an obligatory intermediate on the reaction path. There is, however, an additional point to which these results lead. It was pointed out by Peter Hemmerich (personal communication) that while discussions of flavin-linked dehydrogenase mechanisms usually revolve around the reactions of the prosthetic group almost exclusively, with almost no mention of the properties of either the substrate or the protein, equivalent discussions of pyridine nucleotide dehydrogenase mechanisms seem to consider the properties of the protein and those of the substrate extensively but rather ignore the coenzyme. The observation is quite on the mark; examination of all of the most detailed mechanisms written for the enzyme, including our own (14) and those of Rife and Cleland (13) and Smith et al. (15), all show NADPH as being involved explicitly only in the obvious hydride-transfer step and then appear to carry it along in the reaction scheme in what might be termed an "occult role." This continued inclusion in subsequent steps without a specified role is an implicit recognition by most investigators of the fact that many of these steps, which do not appear to involve NADPH directly, simply do not go on in its absence. Therefore, the finding reported here that the presence of NADPH at the active site

induces the formation of enzyme-bound a-iminoglutarate, and that this effect does not involve hydride transfer from the coenzyme, would appear to be of some importance. While we still do not know how NADPH exerts this nonreductive effect, we now know that the effect involves the induction of the formation of a-iminoglutarate at the active site of the enzyme, and we also know the point in the reaction sequence at which this effect occurs.

Proc. Nati Acad Sci USA 81 (1984)

2751

We acknowledge helpful discussions with Drs. C. E. Hignite and R. Srinivasan. We also thank Dr. R. Medary for performing the stopped-flow spectrophotometric work reported in this publication. This work was supported in part by grants from the National Science Foundation (PCM 8203880) and the National Institute of General Medical Sciences (GM 15188) and by the Veterans Administration. 1. von Euler, H., Adler, E., Gunther, G. & Das, N. B. (1938) Hoppe-Seyler's Z. Physiol. Chem. 254, 61. 2. Fisher, H. F., Srinivasan, R. & Rougvie, A. E. (1982) J. Biol. Chem. 257, 13208-13210. 3. Fisher, H. F., Bard, J. R. & Prough, R. A. (1970) Biochem. Biophys. Res. Commun. 41, 601-607. 4. Hochreiter, M. C., Patek, D. R. & Schellenberg, K. A. (1972) J. Biol. Chem. 247, 6271-6276. 5. Cross, D. G. & Fisher, H. F. (1970) J. Biol. Chem. 245, 26122621. 6. Cross, D. G. (1972) J. Biol. Chem. 247, 784-789. 7. di Franco, A. (1974) Eur. J. Biochem. 45, 407-424. 8. Proelss, H. F. & Wright, B. W. (1973) Clin. Chem. 19, 11621169. 9. Viswanathan, T. S., Hignite, C. E. & Fisher, H. F. (1982) Anal. Biochem. 123, 295-302. 10. Brown, A., Colen, A. H. & Fisher, H. F. (1978) Biochemistry 17, 2031-2034. 11. Viswanathan, T. S., Johnson, R. E. & Fisher, H. F. (1982) Biochemistry 21, 339-345. 12. Frieden, C. (1959) J. Biol. Chem. 234, 2891-2896. 13. Rife, J. E. & Cleland, W. W. (1980) Biochemistry 19, 23282333. 14. Fisher, H. F. & Colen, A. H. (1978) in Developments in Biochemistry, eds. Singer, T. P. & Ondarza, R. N. (Elsevier/ North-Holland, Amsterdam), Vol. 1, pp. 95-108. 15. Smith, E. L., Ansten, B. M., Blumenthal, K. M. & Nye, J. F. (1975) in The Enzymes, ed. Boyer, P. D. (Academic, New York), Vol. 11, Part A, pp. 293-367.