The Role of Tyr-169 of Trimethylamine Dehydrogenase in Substrate Oxidation and Magnetic Interaction between FMN Cofactor and the 4Fe/4S Center*

Tyr-169 in trimethylamine dehydrogenase is one component of a triad also comprising residues His-172 and Asp-267. Its role in catalysis and in mediating the magnetic interaction between FMN cofactor and the 4Fe/4S center have been investigated by stopped-flow and EPR spectroscopy of a Tyr-169 to Phe (Y169F) mutant of the enzyme. Tyr-169 is shown to play an important role in catalysis (mutation to phenylalanine reduces the limiting rate constant for bleaching of the active site flavin by about 100-fold) but does not serve as a general base in the course of catalysis. In addition, we are able to resolve two kinetically influential ionizations involved in both the reaction of free enzyme with free substrate (as reflected ink lim/K d ), and in the breakdown of the E ox·S complex (as reflected in k lim). In EPR studies of the Y169F mutant, it is found that the ability of the Y169F enzyme to form the spin-interacting state between flavin semiquinone and reduced 4Fe/4S center characteristic of wild-type enzyme is significantly compromised. The present results are consistent with Tyr-169 representing the ionizable group of pK a ∼9.5, previously identified in pH-jump studies of electron transfer, whose deprotonation must occur for the spin-interacting state to be established.

Trimethylamine dehydrogenase (TMADH, EC 1.5.99.7), 1 an iron-sulfur containing flavoprotein from the bacterium Methylophilus methylotrophus (sp. W 3 A 1 ), catalyzes the oxidative demethylation of trimethylamine to dimethylamine and formaldehyde. The enzyme is a homodimer, and each subunit contains an unusual covalently linked 6-S-cysteinyl FMN cofactor and a bacterial ferredoxin-type 4Fe/4S center, as well as 1 equivalent of tightly bound ADP of unknown function (1)(2)(3)(4)(5)(6). The physiological electron acceptor of TMADH is an electrontransferring flavoprotein, a 62-kDa heterodimer containing 1 equivalent each of FAD (7) and AMP (8). Electron-transferring flavoprotein is thought to oxidize reduced TMADH in two suc-cessive one-electron steps, cycling between the quinone and (anionic) semiquinone oxidation states. The availability of a high resolution structure for TMADH (6) and the cloned and overexpressed gene for the enzyme (10,11) has made it possible to examine many aspects of the reaction mechanism by conventional site-directed mutagenesis. These have included studies of the role of (i) the 6-S-cysteinyl FMN in catalysis (11)(12)(13), (ii) cation-bonding in substrate recognition (14), and (iii) residues on the surface of TMADH involved in electron transfer to electron-transferring flavoprotein (15).
The reaction of TMADH with trimethylamine exhibits three sequential kinetic phases (16 -18): a fast phase that represents bleaching of the 6-S-cysteinyl FMN, an intermediate phase that reflects intramolecular electron transfer from dihydroflavin to the 4Fe/4S center to generate the flavin semiquinone and reduced 4Fe/4S center, and a slow phase that involves formation of an unusual spin-interacting state of the enzyme in which the unpaired magnetic moments of the reduced 4Fe/4S center and flavin semiquinone are strongly ferromagnetically coupled (18 -21).
In the crystal structure of TMADH, Tyr-169 lies in van der Waals contact with the pyrimidine ring of the flavin cofactor, and is hydrogen-bonded to His-172 (which is also in van der Waals contact with the flavin). In order to ascertain the catalytic significance of Tyr-169 in TMADH, we have isolated a Y169F mutant enzyme and analyzed its kinetic behavior. We find that mutation of this residue to phenylalanine reduces the limiting rate constant for flavin reduction by a factor of approximately 100, but does not function as an active site base. The mutation also significantly reduces the ability of the flavin semiquinone to interact magnetically with the reduced 4Fe/4S center of the two-electron reduced enzyme. This is due to a substantial decrease in the equilibrium amount of enzyme possessing flavin semiquinone and reduced 4Fe/4S center, presumably by perturbing the semiquinone/hydroquinone halfpotential of the active site flavin, and a decrease in the magnetic interaction between the centers in mutant enzymes possessing this electron distribution. The result suggests that Tyr-169 in all likelihood represents the ionizable group of pK a ϳ 9.5, previously identified in pH-jump studies of electron transfer (22), whose deprotonation must occur for the spininteraction state to be established.

EXPERIMENTAL PROCEDURES
Chemicals and Enzymes-Complex bacteriological media were from Unipath and all media were prepared as described by Sambrook et al. (23). Trimethylamine, 2,6-dichlorophenolindophenol, phenazine methosulfate, tetramethylammonium chloride (TMAC), and all buffers were from Sigma. Sodium dithionite was obtained from Virginia Chemicals. Perdeuterated trimethylamine HCl (99.7% D) was from CK Gas Products Ltd. All other chemicals were of analytical grade where possible. Wild-type TMADH was purified from M. methylotrophus as described by Steenkamp and Mallinson (1) and modified by Wilson et al. (24). The concentration of wild-type enzyme was determined using an extinction coefficient of 27.3 mM Ϫ1 cm Ϫ1 at 443 nm in 50 mM potassium phosphate buffer, pH 7.0. Recombinant Y169F TMADH was generated and isolated as described elsewhere (14) and expressed using the plasmid pSV2tmdveg (11). With this expression system, the enzyme as isolated possesses its full complement of 4Fe/4S center and ADP, but a significant portion of the enzyme lacks the flavin (25,26). Using spectrophotometric methods reported elsewhere (11), the fraction of flavinylated enzyme in the Y169F preparations used in the present study was estimated as approximately 50%; the mutant enzyme was found to be stoichiometrically assembled with the 4Fe/4S center and ADP. The concentration of the Y169F mutant of TMADH was determined using an effective extinction coefficient (20.0 mM Ϫ1 cm Ϫ1 at 443 nm) for oxidized enzyme, calculated from the extent of the spectral change elicited by excess substrate (only the flavinylated enzyme is reducible by substrate). Enzyme solutions of the desired pH were obtained by adding microliter volumes of a concentrated enzyme stock to buffer at the desired pH.
The 4Fe/4S center of TMADH was selectively inactivated by treatment with ferricenium hexafluorophosphate at pH 10, essentially as described by Huang et al. (27). The protein (30 M) was incubated with ferricenium hexafluorophosphate (3 mM) contained in 50 mM potassium borate buffer, pH 10, at room temperature for 6 h. Excess oxidant was removed by size-exclusion chromatography using Sephadex G-25 equilibrated in 20 mM potassium phosphate buffer, pH 7.0. The flavin site of TMADH treated in this way remains reducible by trimethylamine, but the 4Fe/4S enter becomes redox inert and the enzyme is unable to pass electrons on to electron-transferring flavoprotein (27).
UV/Visible and EPR Spectroscopy-UV/visible spectra were recorded using a Hewlett-Packard 8452A single-beam diode array spectrophotometer. EPR spectra were obtained using a Brü ker ER 300 EPR spectrometer equipped with a ER035M gaussmeter and a Hewlett-Packard 5352B microwave frequency counter. Instrument settings were 9.45 GHz microwave frequency, 1.00 mW microwave power, 10 G field modulation, and 100 kHz modulation amplitude. Temperature was maintained at 15 K using a Brü ker ER 4112HV liquid helium cryostat with an Oxford Instruments ITC4 temperature controller. EPR samples were prepared as follows: 300 l of enzyme solutions at the desired pH were placed in quartz EPR tubes using a long-needle Hamilton syringe, then treated with either 20 -30 l of a concentrated substrate solution prepared in the same buffer (sufficient to give at least 5-fold stoichiometric excess over the enzyme) or with a comparable amount of TMAC followed by reduction with sodium dithionite. For the substrate-reduced samples, a final enzyme concentration of 100 M was used for the wild-type while 200 M was used for Y169F to compensate for the incomplete flavinylation in this mutant protein to facilitate direct comparisons in the data. UV/visible spectra of each sample were recorded before and after each addition using a special spectrophotometer cell holder, which accommodates EPR tubes. Samples were then thoroughly mixed and slowly frozen by hand in liquid nitrogen. EPR spectra were recorded at both half-and high-field (in separate sweeps) for each sample.
Kinetic Measurements-Steady-state kinetic measurements were performed with a 1-cm light path in a final volume of 1 ml. The desired concentrations of trimethylamine, phenazine methosulfate, and 2,6dichlorophenolindophenol were obtained by making microliter additions from stock solutions to the assay mixture. Assays were performed in 100 mM sodium pyrophosphate buffer, pH 8.5. Reaction was initiated by the addition of substrate, and the decrease in absorbance at 600 nm due to reduction of 2,6-dichlorophenolindophenol (⑀ ϭ 21, 900 M Ϫ1 cm Ϫ1 ) was measured using a Hewlett-Packard 8452A diode array spectrophotometer. All data were collected at 30°C. Data were fitted to the appropriate rate equation using the fitting program Grafit (28).
Rapid kinetic experiments were performed using an Applied Photophysics SX.17MV stopped-flow spectrophotometer. Time-dependent reduction of TMADH by trimethylamine at pH 6.5 and 7.0 was performed using a photodiode array detector. Spectral deconvolution was perfomed by global analysis and numerical integration methods using PROKIN software (Applied Photophysics). For single wavelength studies, data collected at 443 nm were analyzed using nonlinear least FIG. 1. Steady-state kinetic analyses of the reaction of wildtype and Y169F TMADH with trimethylamine. Open circles, data for wild-type TMADH; closed circles, data for Y169F TMADH. Data for the wild-type were fitted using the equation described by Falzon and Davidson (30); Y169F data were fitted using the standard Michaelis-Menten equation. squares regression on an Archimedes 410-1 microcomputer using Spectrakinetics software (Applied Photophysics). Experiments were performed by mixing TMADH in buffer of the desired pH, with an equal volume of trimethylamine at the desired concentration in the same buffer. The concentration of substrate was always at least 10-fold greater than that of TMADH, thereby ensuring pseudo first-order conditions. For each substrate concentration used, at least four replicate measurements were collected and averaged. Substrate-reduced TMADH is quite stable to reoxidation in aerobic environments (half-life about 50 min, Ref. 24), and consequently these stopped-flow experiments were carried out under aerobic conditions.
The absorbance change at 443 nm for Y169F TMADH at values of pH Ͼ 7.0 was essentially monophasic, with a single rate constant obtained from fits of the data to Equation 1, where C is a constant related to the initial absorbance and b is an offset value to account for a non-zero baseline. Below pH 7, however, kinetic transients for Y169F TMADH were biphasic and best fit as the sum of two exponentials using Equation 2, where k obs1 and k obs2 are the observed rate constants for the faster and slower phases, respectively, and C 1 and C 2 are related to the initial absorbance; again, b is an offset. The observed rate constants were found to exhibit hyperbolic dependence on substrate concentration and the reaction sequence was modeled as shown in the general scheme, Data were then fitted to obtain related K d and k lim values using k obs ϭ (29).

RESULTS
Steady-state Kinetic Analyses-As reported previously, the steady-state kinetics of wild-type TMADH exhibit excess substrate inhibition (18,30,31). By contrast, Y169F TMADH exhibits well behaved steady-state behavior: no evidence for substrate inhibition was seen even at very high substrate concentrations (up to 45 mM; Fig. 1). The kinetic parameters K m TMA and k cat for the mutant protein are 63 Ϯ 3 M and 2.6 Ϯ 0.03 s Ϫ1 at pH 8.5 and 30°C, which are approximately 5-fold higher and 6-fold lower, respectively, than those determined for the wild-type enzyme (13.7 Ϯ 1.7 M and 15.6 Ϯ 2.4 s Ϫ1 , respectively; Ref. 14). The results suggest that Tyr-169 plays only a relatively small role in the overall catalytic efficiency of the enzyme in the steady-state.
The Reaction of Y169F TMADH with TMA-Reduction of the flavin in Y169F TMADH by substrate was examined by stopped-flow spectroscopy at 443 nm over the pH range 6.0 to 11.0. Unlike wild-type enzyme, in which flavin reduction is essentially monophasic over this pH range (18), flavin reduction in Y169F is biphasic below pH 7.0 ( Fig. 2). At pH 7.0, it is monophasic at high substrate concentrations but biphasic at [TMA] Ͻ 20 mM, suggesting that ionic strength may influence the kinetic behavior. To explore this possibility, the reaction was repeated in lower ionic strength buffer (20 mM phosphate buffer instead of 100 mM, pH 7.0) and biphasic kinetic transients are seen even at high substrate concentrations. However, the limiting rate constant for flavin reduction (faster phase of the biphasic reaction or the single kinetic phase in monophasic reaction) is essentially unchanged. The effect of ionic strength is therefore restricted to controlling the range of substrate concentration over which biphasic kinetic behavior is seen (probably by influencing a kinetically relevant ionization; see below) rather than influencing the limiting rate of flavin reduction. An analysis of the amplitudes for each of the two phases seen at pH 7.0 and below indicates that the slower kinetic phase becomes increasingly prominent as pH decreases. The two kinetic phases also become more clearly resolved, principally due to a decrease in the rate constant for the slower phase. The pH dependence of the amplitudes for the two phases indicates the presence of a kinetically influential ionization of apparent pK a 6.2 Ϯ 0.2 (Fig. 2, inset). While there will be small differences in ionic strength across the pH range used for the determination of this value, these are not expected to compromise the analysis significantly.
Spectral Changes Associated with the Reaction of Y169F TMADH with Substrate-The reductive half-reaction of wildtype TMADH with TMA is triphasic (18). However, at the end of this half-reaction, the distribution of the two electrons derived from substrate in the enzyme is affected by pH (22,32): at high pH (pH 7.5 and above), formation of flavin semiquinone and reduced 4Fe/4S center is favored; at low pH (e.g. pH 6.5), dihydroflavin and oxidized 4Fe/4S center is preferred, reducing the overall kinetics to nearly monophasic behavior. Similar pH effects on electron distribution were also seen for the Y169F TMADH. To simplify the analysis of the absorbance change associated with flavin reduction in Y169F TMADH, the experiment was performed at pH 6.5, thereby effectively eliminating the two slower phases associated with intramolecular electron transfer observed at higher pH values.
Upon completion of the faster phase of flavin reduction the spectrum resembles that of a mixture of oxidized enzyme and enzyme in the dihydroflavin form (Fig. 3). Following completion of the slower phase, the spectrum is that of the dihydroflavin form. The spectral form seen at the end of the faster phase rules out a sequential two-step reduction process involving a flavin semiquinone intermediate since this would give rise to a characteristic flavin semiquinone spectrum. 2 At 1000 s after initial mixing of enzyme with substrate, the observed spectrum indicates that electron transfer to the 4Fe/4S center is indeed far from complete (data not shown). To further confirm that internal electron transfer to the 4Fe/4S center is not implicated in the biphasic behavior seen here, the 4Fe/4S center in the enzyme was selectively inactivated by ferricenium-PF 6 (see "Experimental Procedures"), which is known to render the 4Fe/4S center redox inert. The reductive half-reaction of the modified enzyme was studied at pH 6.5, 7.0, and 10.0. Again, the reaction is biphasic at pH 6.5 and 7.0 and monophasic at pH 10.0, as seen in the untreated Y169F TMADH, while ferricenium-PF6 inactivated wild-type enzyme exhibits monophasic behavior at all pH values examined (data not shown).
Substrate Concentration Dependence and Kinetic Isotope Effects for Substrate Oxidation-The substrate concentration dependence of flavin reduction with Y169F TMADH has been investigated at pH 7.0 (using 20 mM phosphate buffer so that the biphasic behavior could be resolved throughout the entire substrate concentration range) and pH 6.5 (using 100 mM buffer). 3 At pH 7.0, both phases for flavin reduction in Y169F TMADH exhibit hyperbolic dependence on [TMA]. The limiting rate constant for the faster phase (43 s Ϫ1 Ϯ 1.7) was about 21-fold less than that seen with wild-type enzyme (903 s Ϫ1 Ϯ 50) ( Table I). The dissociation constants calculated for the two phases seen with the mutant protein are 35 Ϯ 3.4 and 38 Ϯ 4 mM, respectively, considerably larger than that seen with wildtype TMADH (6 mM, Table I). At pH 7.0, both phases were found to be sensitive to a kinetic isotope effect of approximately 7, as seen with wild-type TMADH, when perdeuterated TMA was used (Table I) indicating that the observed kinetics involve C-H bond breakage. At pH 6.5, the faster phase of the reaction exhibits hyperbolic dependence on [TMA], with k lim and K d of 9 Ϯ 0.8 s Ϫ1 and 55 Ϯ 10 mM, respectively. The slower phase, however, is essentially independent of [TMA] over the concentration range studied (except in the very low substrate concentration regime), with an observed rate constant of 0.03 s Ϫ1 . There is also an associated loss of primary kinetic isotope effect on the reaction (Table I), indicating that cleavage of the C-H bond is no longer rate-limiting. Given the additional observation that the slow phase of the reaction seen at pH 7.0 is significantly slower than k cat , we have not pursued the nature of this slow phase further in the present work.
Kinetically Influential Ionizations in Y169F-The reaction of 2 The transient formation of an anionic semiquinone would normally be observed readily at 365 nm. However, single wavelength studies performed at 365 nm indicated that transient formation of an anionic semiquinone did not occur en route to formation of the dihydroflavin. 3 Given that trimethylamine is predominantly protonated at the pH values employed here and the wide substrate concentration range used in the experiments (0 to 120 mM), control experiments have been performed in the absence and presence of 0.2 M potassium chloride to study the effect of ionic strength on kinetics. Additionally, in separate experiments and at selected pH values (6.5, 7.0, and 7.5), ionic strength was kept constant by balancing the substrate and potassium chloride concentrations over the entire substrate concentration range studied. In all cases, the values for the limiting rate constants for both phases and enzyme-substrate dissociation constants were found to be identical (within experimental error). The data therefore demonstrate that ionic strength influences only the relative spectral change associated with each of the two kinetic phases (presumably by perturbing the apparent pK a of about 6.2) and does not affect the observed rate constants for each phase or the corresponding dissociation constant for the E⅐S complex. Y169F TMADH with trimethylamine has been investigated as a function of pH at 5°C. The pH dependence of k lim seen with Y169F enzyme reveals two reasonably well resolved pK a values (pK a 6.7 Ϯ 0.2 and 9.5 Ϯ 0.3; Fig. 4A), whereas only one ionization is observed for the wild-type enzyme (18). A plot of k lim /K d versus pH gives a bell-shaped curve (Fig. 4B), as seen in wild-type enzyme (18), with two pK a values of 9.7 Ϯ 0.1 and 11.0 Ϯ 0.1 attributable to the ionization of free enzyme and free substrate (pK a of TMA is 9.81), respectively. As in the case of wild-type enzyme, substrate is found to bind preferentially in the cationic form (14,18). Comparison of the pH profiles for Y169F and wild-type enzymes indicates that the pK a values that control flavin reduction are slightly perturbed on mutating Tyr-169 to Phe. However, all kinetically influential ionizations seen in the wild-type enzyme remain in the Y169F mutant enzyme, indicating that Tyr-169 in the wild-type enzyme either does not ionize over the pH range investigated, or that its ionization is not kinetically influential for flavin reduction.
UV/Visible Spectra of Y169F TMADH-The UV/visible absorption spectra for the oxidized and substrate-reduced forms of wild-type and Y169F TMADH, along with the corresponding [oxidized] minus [substrate-reduced] difference spectra are shown in Fig. 5. Oxidized wild-type protein exhibits an A 444 nm/A 382 nm absorbance ratio of about 1.3, whereas that for the Y169F TMADH gives a ratio around 1.03 due to incomplete flavinylation when expressed in Escherichia coli. Interestingly, the absorption change elicited by reduction with excess substrate for this mutant protein is different from that seen with wild-type protein under the same conditions (which we have previously shown to be identical in native and recombinant wild-type enzyme; Ref. 11). 4 In particular, the difference max-imum at 365 nm (reflecting accumulation of the flavin semiquinone form) is absent in the mutant. The observed spectral change seen with the Y169F mutant is in fact quite reminiscent of that generated by reduction of wild-type protein to the twoelectron reduced level using sodium dithionite at pH 8.0, where the enzyme principally contains flavin semiquinone and reduced 4Fe/4S center but their magnetic moments do not interact (22). This interpretation is further supported by the EPR spectroscopic studies discussed below. The implication is that the distribution of reducing equivalents between the 4Fe/4S and flavin centers in the two proteins are different; a larger portion of Y169F TMADH exists as flavin hydroquinone and oxidized 4Fe/4S center rather than flavin semiquinone and reduced 4Fe/4S center, especially at pH 7.0.
Full reduction of wild-type TMADH, which requires 3 reducing equivalents, is observed when titrated with sodium dithionite, however, the enzyme takes up only two electrons when reduced with excess substrate or reduced by sodium dithionite in the presence of TMAC (a substrate analog and inhibitor of 4 It is to be emphasized that while the incomplete flavinylation of the recombinant Y169F mutant prevents a direct spectral comparison of its absorbance spectrum with that of the fully flavinylated native protein, a direct comparison can be made between the substrate-induced difference spectra seen with the two forms of the enzyme, as only that portion of the mutant protein possessing flavin can become reduced by substrate. All discussion and conclusions here are based on these substrate-induced difference spectra and not on the absolute spectra themselves. Use of substrate as reductant ensures that the entirety of the spectral change seen with the mutant arises from enzyme that possesses the full complement of redox-active cofactors, as the deflavo form of the enzyme cannot be reduced by substrate. In determining the concentration of the Y169F protein, we use an effective extinction coefficient that gives the concentration of the fully functional, flavincontaining portion of the enzyme, not simply the total concentration of polypeptide. In both Figs. 5 and 6, spectra of native and Y169F enzyme are presented that have been normalized on a per-flavin basis so that a direct comparison can be made in extinction coefficient and EPR intensity. TMADH) (19,31): binding of the substrate analog perturbs the reduction potential of the flavin semiquinone/hydroquinone couple such that full reduction of the enzyme does not occur. When Y169F TMADH is reduced with sodium dithionite in the presence of TMAC at pH 7.0, however, the final difference absorption spectrum resembles that for three-electron reduction of wild-type enzyme (data not shown). This indicates that full reduction has occurred, consistent with the EPR studies described below. The results indicate that the oxidation-reduction properties of the mutant protein are perturbed and that the ability of the Y169F mutant to form the spin-interacting state is compromised.
EPR Spectra of Y169F TMADH-EPR spectra of wild-type and Y169F TMADH reacted with excess substrate are shown in Fig. 6, A-D. Under these conditions, the wild-type protein contains flavin semiquinone and a reduced iron-sulfur center whose magnetic moments interact strongly with each other and give rise to a spin-interaction state with characteristic g ϳ 2 complex high-field EPR signal and an intense g ϳ 4 half-field signal (Fig. 6, A and B). For the mutant protein, on the other hand, the signal centered at g ϳ 2 is primarily a combination of the axial signal of flavin semiquinone and the rhombic signal of reduced 4Fe/4S center (Fig. 6D). The complex EPR signal associated with the spin-interacting state seen in the wild-type protein is not observed. In addition, the intensity of the g ϳ 4 signal that is diagnostic of the spin-interacting state is greatly reduced in the mutant compared with wild-type protein (Fig.  6C). (The g ϳ 4.3 signal seen in the spectrum is due to trace amounts of adventitious iron in the sample.) The experiments have also been performed at pH 10.0 and similar results are observed (data not shown). This indicates that the mutant protein contains a substantial amount of flavin semiquinone and reduced iron-sulfur center under the present experimental conditions, but the magnetic moments of the two unpaired spins (which given the manner in which the samples were prepared must exist in the same enzyme molecule) do not interact with each other strongly as in the wild-type protein. Fig. 6, E and F, shows the EPR spectra observed when the mutant protein is reduced by sodium dithionite in the presence of TMAC. The rhombic EPR signal of the reduced 4Fe/4S center is observed at high-field (Fig. 6F) and no half-field signal associated with the spin-interaction state is seen (Fig. 6E), which is consistent with the UV/visible results that the mutant protein is fully reduced to a three-electron reduction level by sodium dithionite even in the presence of TMAC. DISCUSSION Tyr-169 is one of three amino acids comprising a novel Tyr-His-Asp triad in the active site of TMADH. Our data for the Y169F mutant clearly indicate that C-H bond cleavage and FMN reduction occur in the mutant enzyme, albeit at a limiting rate that is approximately 100-fold slower than is seen with FIG. 6. EPR spectra of wild-type and Y169F TMADH at pH 7.0. A, halffield EPR spectrum of substrate-reduced wild-type TMADH. B, high-field EPR spectrum of substrate-reduced wild-type TMADH. C, half-field EPR spectrum of substrate-reduced Y169F TMADH. D, high-field EPR spectrum of substrate-reduced Y169F TMADH. E, half-field EPR spectrum of Y169F TMADH reduced with sodium dithionite in the presence of TMAC. F, high-field EPR spectrum of Y169F TMADH reduced with sodium dithionite in the presence of TMAC.
wild-type enzyme. Tyr-169 is in van der Waals contact with the flavin isoalloxazine ring, and local adjustments in active site structure (both physical and electronic) as a result of the mutation are likely to be responsible, at least in part, for the slower rates observed in the flavin reduction of Y169F TMADH. The pH-dependence profiles for Y169F are similar to those for wild-type enzyme, indicating that Tyr-169 is not the group whose ionization facilitates substrate oxidation and/or substrate binding in wild-type enzyme. However, oxidation of substrate is controlled by an additional ionization with pK a of 6.2 in Y169F TMADH, not observed in the wild-type enzyme.
At pH values below this pK a , enzyme reduction occurs as two kinetically resolvable steps, only the faster of which appears to be catalytically significant. The identity of the amino acid residue responsible for the additional ionization (pK a value 6.2) that controls the expression of the biphasic reductive transients in Y169F at low pH remains to be determined, however, it cannot be His-172 (which H-bonds to Tyr-169 in wild-type enzyme), since recent studies of a mutant H172Q TMADH suggest that ionization of this residue occurs around pH 8. 5 Similarly, this work indicates that none of these residues is likely to be involved in abstraction of a proton from substrate to form a carbanion intermediate (18). Indeed the base catalysis accounts for only a quite modest portion of the enzyme-catalyzed rate acceleration for substrate oxidation (17,18). The lack of an obvious base to support a carbanion mechanism provides indirect support for homolytic C-H bond cleavage analogous to the mechanism that has been proposed for the mechanism of monoamine oxidase (as discussed in Ref. 18).
The present work clearly demonstrates that Tyr-169 also plays an important role in mediating the spin-interaction between the flavin semiquinone and reduced 4Fe/4S center in TMADH. Previous work has shown that formation of the spininteracting state of TMADH is governed by a basic residue located at or near the active site, with pK a value around 9.5 (22), and our results implicate Tyr-169 as this basic residue. Although Tyr-169 lies opposite the flavin ring from the ironsulfur in TMADH, the importance of this residue in forming the spin-interaction between the two centers can be rationalized in the context of the x-ray crystal structure of TMADH (6). Tyr-169 is located near the C(2) ϭ O group of the flavin isoalloxazine ring and its van der Waal's surface is in contact with that of the flavin ring. When a negative charge is developed on the hydroxyl group of Tyr-169 side chain, due to electrostatic repulsion the unpaired electron density on the flavin isoalloxazine ring is reasonably expected to be forced to redistribute away from this residue toward the 4Fe/4S center, effectively reducing the spin-spin distance. This may also induce a larger dipole moment on the flavin isoalloxazine ring, which could be important in promoting the formation of the spin-interacting state.
The present work also indicates that the reduction potential of the semiquinone/hydroquinone flavin couple is perturbed in the Y169F enzyme, as reflected in the shift in oxidation-reduc-tion equilibrium inferred from the UV/visible spectra. This is supported by the steady-state kinetic study demonstrating that substrate inhibition is absent in the Y169F TMADH, and the spectroscopic studies showing that full reduction of Y169F is achieved with dithionite even in the presence of TMAC. As described in Ref. 18, substrate inhibition in wild-type TMADH is accounted for by perturbation of the semiquinone/hydroquinone flavin couple upon substrate binding to partially reduced enzyme. Potentiometric studies on Y169F will soon help to further illustrate this point.