Histidine Ammonia-lyase THE USE OF 4-FLUOROHISTIDINE IN IDENTIFICATION OF THE RATE-DETERMINING STEP

The (Y, 6 eliminations of NH3 from L-histidine and 4-fluoro-L-histidine by histidine ammonia-lyase appear to occur by similar mechanisms, although a large difference in V,,, for the two reactions was observed. Both reactions were shown to be reversible with an equilibrium constant of 4 to 5. The presteady state kinetics of the deamination of 4-fluoro-L-histidine indicates that the rate-determining step precedes the dissociation of ammonia from the enzyme. The isotope effect of 1.4 to 2.0 observed with 4-fluoro-DL-[fl-2Hz]histidine or DL-@-2Hz]histidine indicates that the C-H bond breakage is at least partially rate-determining for the deamination of both substrates.


SUMMARY
The (Y, 6 eliminations of NH3 from L-histidine and 4-fluoro-L-histidine by histidine ammonia-lyase appear to occur by similar mechanisms, although a large difference in V,,, for the two reactions was observed. Both reactions were shown to be reversible with an equilibrium constant of 4 to 5. The presteady state kinetics of the deamination of 4-fluoro-L-histidine indicates that the rate-determining step precedes the dissociation of ammonia from the enzyme. The isotope effect of 1.4 to 2.0 observed with 4-fluoro-DL-[fl-2Hz]histidine or DL-@-2Hz]histidine indicates that the C-H bond breakage is at least partially rate-determining for the deamination of both substrates.   of both a hydrogen and a urocanate exchange reaction with histidine had previously been taken as evidence that the slow step in the formation of urocanate is the dissociation of NH3 from the enzyme (6, 9). Since this step is common to both substrates, the difference in V,,,,, slown above indicates that dissociation of ammonia from 1 the enzyme cannot be a common rate-limiting step; alternatively, the slow step in both cases may precede the loss of NHa, or the slow step may be different for the two substrates. The former hypothesis was tested by spectrophotometric measurement of presteady state kinetics, as shown in Fig. 3. At the top of the figure, a rapid kinetic experiment with 4-fluorohistidine is reported. No initial burst of activity is detected; if the rate-limiting step which controls the steady state rate followed the formation of 4-fluorourocanate, a burst of activity (29) should result in an increased absorbance of 0.08 to 0.3 at zero time. At all times the observed rate is linear with time and not significantly different from the steady state rate, as summarized in Table I. Further- to less than 0.3 nmol of fluorourocanate formed per nmol of enzyme tetramer at 2 pH values and in the presence of metal ions. In the case of fluorohistidine, therefore, the rate-limiting step under these conditions precedes the dissociation of amrnonia from the enzyme.
The results of the experiments performed with L-histidine are not as conclusive.
A representative experiment is shown at the bottom of Fig. 3. At the earliest time measurable after mixing (25 ms at this electronic band pass setting), the absorbance change is occurring at the steady state rate (Table I), indicating the absence of any process with a rate constant greater than 100 s-i. However, extrapolation to zero time consistently shows an initial absorbance which could correspond to the very fast release of as much as 2 nmol of urocanic acid per nmol of enzyme tetramer. The size of this initial absorbance is not very reproducible; at lower pH and in the presence of EDTA, both of which decrease the steady state rate, the zero time absorbance can be decreased to as little as 0.5 nmol of urocanic acid per nmol of enzyme tetramer. In the minimum time period between COW secutive experiments (15 to 20 s) the absorbance of the solution will reach a high value (Fig. 3); therefore, we believe that the initial absorbance is an artifact, resulting from incomplete removal of product from the observation chamber, prior to the mixing of enzyme and substrate solutions for the next experiment. This objection is not true for 4-fluorohistidine experiments, because of the slow reaction rate. Isotope E$ects-The absence of a demonstrable initial burst of activity in the presteady state kinetic experiments described above indicates that the rate-determining step in the deamination of 4-fluorohistidine could correspond to the breaking of the C-N bond, the C-H bond, or to a concerted mechanism. In the latter pathways, substitution of the hydrogen in the fi position by a deuterium should decrease the rate of the over-all reaction. Lineweavcr-Uurk plots of the velocity of the reaction as a function of concentration of 4-fluoro-nn-histidinc and of 4fluoro-un-[P-*Hz]histidine are shown in Fig. 4 (bottom). The Michaelis constants derived from these experiments are summarized in Table II. Isotopic replacement of the fl-hydrogens does not affect the K, values, but does decrease li,,, both in the presence of metal ions and of EDTA. Isotope effects of 1.7 and 2.0 were calculated. As showrl in Fig. 4 (top), isotope effects of 1.4 and 1.5 are observed with nn-histidine and its p deuterated analog. These findings suggest that the mechanism of the reaction is similar with both substrates and that breaking of the COvalent C-H bond determines at least partially the reaction rate.
Since an exchange of 3H20 with histidine was observed previously for the deamination of histidine (6), the rate of deamination of the deuterated substrate in *HZ0 was compared with that of nn-histidine in HZ0 as a function of pH and pD, in order to eliminate the possibility that some exchange of deuterium in the p position will decrease the isotope effect under the conditions described above. No significant change in the isotope effect was observed under these conditions (data not shown).
Exchange Reactiolzs-Peterkofsky has shown that solvent tritium and [i4C]urocanic acid can be incorporated into reisolated  hi&line by exchange reactions (6). The rate of the exchange reactions below pH 8 was found to exceed that of net product formation. A study of the initial rates of urocanate and 3Hz0 cxchauge was undcrtakcu to determine if proton retnoval prccedes, follows, or is simultaneous with C-N bond cleavage.
At p1-I 7.7, urocanatc cschangc is readily detectable (Fig. 5); under these csontlitions, both exchange and product formation are inhibited by high urocanate concentrations. The rate of fluorourocanate cschange is much lower (Fig. 5, right). 1Jntler conditions in which the rate of product formation is one-tenth that observed with hjstidiue, the rate of 4-fluorourocanate exchauge is 250 tirncs smaller thau [*Y,']urocanate exchange. Furthermore, the optimum concentratiou of 4-fluorourocanate for the exchange reaction (not reached) is higher than that of urocanate (Fig. 5).4 The rate of the tritium exchange reaction is shown in Fig. 6 for both L-hi&dine and 4-fluoro-L-histidine. As in the urocanic acid exchange, the rate of tritium exchange is much smaller with 4-fluorohistidine than with histidine. For both exchanges, the reduced rates with the fluoro analog reflect the considerably lower value of V,,,. It can also be noted that the rate of tritium exchange is uot greatly affected by urocanate concetltration. At low pH, the rate of tritium eschange is linear, even at early times. When % exchange and ['%]urocanate exchange into histidirle were tneasured simultaneously, the rate of 3H exchange exceeds slightly that of [YZ]urocanate exchange (Table III). and 60 and 120 min (8.5 and 11 mM 4-fluorourocanate).
Increased absorption due to 4-fluorourocanate formation was measured at 286 nm. 4-Fluorourocanate formed per min per mg, A; 4-fluoro[3H]urocanate incorporated into 4-fluorohistidine per min per mg, l reversible formatiou of an amino enzyme intermediate and a slow irreversible dissociation of the amino enzyme, accounting for the apparent irreversibility of the over-all reaction (6). However, it was subsequently shown that after prolonged incubation, the reaction is reversible with an equilibrium constant of 3 to 5 (8). To clarify this situation, the inhibition of urocanate formation by NH4HC03 in the presence of EDTAS was studied. The results could all be expressed as linear double reciprocal plots, compatible with Tc'Hs being a competitive inhibitor with respect to histidine. However, the nature of the inhibition could not be proved unequivocably, due to the high apparent Ki of NH4HC03 (0.2 M). We were able to confirm the reversibility of the reaction with urocanic acid as well as with 4-fluorourocanic acid within short incubation times by use of high concentrations of enzyme. The time course of such reversibility experiments is summarized in Table IV. The equilibrium constants derived from these experiments are similar for both substrates and are comparable to those published earlier for urocanic acid (8).
The dependence of the rate of the reverse reactiou on the con-5 In view of the ability of NHaHCO, to bind the metal activators (37) and thereby inhibit the enzyme noncompetitively, the reaction was carried out in the presence of EDTA.  2.40 a The incubation mixture was 0.05 M potassium phosphate buffer (pH 7.2), 0.1 mM Cd %+, 1 rnM mercaptoethylamine, 20 mM L-histidine, and 4 mM [Wlurocanic acid and contained 0.016 mg of enzyme and 1 mCi of 3H,0 in a final volume of 0.2 ml. The reaction was carried out as described in Fig. 6 and under "Materials and Methods." b Urocanic acid incorporation and tritium incorporation into histidine were measured after separation by thin layer chromatography. The specific activity of urocanic acid after 25 min of incubation was corrected for the increase in concentration and the decrease in radioactive material, and the specific activity used to calculate the micromoles per ml at 25 min was the average of the specific activities at zero time and at 25 min. centration of urocanic acid and of 4-fluorourocanic acid was also measured. In contrast to the forward reaction, where the K, of histidine is higher than that of 4-fluorohistidine, the estimated K, for 4-fluorourocanate (5 mM) is higher than that for urocanate (3 md. vm,, values for the back reaction with these substrates (0.002 and 0.007 uriits/mg, respectively) are much smaller than those of the forward reaction. Furthermore, the rate of formation of histidine is not significantly affected bJ-IGDTA, but the formation of 4-fluorohistidine is reducctl to undetectable levels in the presence of the metal-binding agent. Since both rcvcrse reactions were not performed in the presence of saturating corn centrations of NH3, these values of V max arc not true constants. (specific activity 50,000 cpm/Mmol) as indicated, was incubated at room temperature in the presence of 2 ~1 of toluene/0.16 ml with and without enzyme. The enzyme concentrations were 0.2 mg/ml (I and II), 0.4 mg/ml (III), and 1 mg/ml (IV). Histidinc or 4-fluorohistidine were determined as indicated under "Materials and Methods," 15.'to 75-J aliquots for each time point.
b The concentration of NHdHC03 was 0.26 M; histidine content was determined with an amino acid analyzer on a 3OOJ aliquot.
As argued below, their difference should be far greater than observed.

Substitution
in various positions of the imidazole ring of Lhistidine yields analogs with substrate properties markedly different from those of the natural substrate for hi&dine ammonialyase. For example, 2-fluorohistidinc is a substrate, with K, = 0.02 M, L-NT-methylhistidine is a weak competitive inhibitor, Ki = 0.1 nz, and L-N'T-meth.lhistidilie is ncithcr a substrate nor an inhibitor.
In contrast, 4-fluorohistidiue is a competitive irlhibitor with a Ki lower than the K, of histidine (Fig. 1). The 4-fluoro compound is alTo a substrate for the cnzymc; since its rate of deamination is 30. to 100.fold smaller than that of histidine, this analog provides a useful tool for study of the mechanism of the enzymic reaction.
It is conceivable that the low rate of deamination in the case of the 4-fluoro compound might result from an inhomogeneity in the enzyme preparation, only a small fraction of the population being active with the analog. However, it seems unlikely that the I<, of the analog would give such good agreement with its Ki with regard to histidine, if heterogeneity were the case. Such agreement is to be expected, of course, if it is both substrate and inhibitor for the same enzyme. Other results also argue against this notion.
The absence of a detectable initial burst of activity in the rapid kinetic experiments (Table I) indicates that the rate-determining step, in the case of 4-fluorohistidine, camlot follow the formation of 4-fluorourocanic acid (29) and probably involves C-H or C-N bond cleavage, or both, in a concerted mechanism. The stopped flow data with histidine are less conclusive because the fast rection rate makes the measurement of zero time absorbance less accurate. However, since the deamination of both substrates is stimulated by metal ions (11-15) and 1',,,, values are reduced by EDTA to a similar extent for each, even at high pH, formation of product (the cx ,fl elimination reaction) is most likely the common rate-limiting step. The observation that both substrates show small but reproducible P-deuterium isotope effects, 1.4 to 2.0, suggests that cleavage of a C-H bond determines at least partially the reaction rate in each case (30). However, the small size of the effects, particularly for histidine, indicates that secondary isotope effects should also be considered (31). Thus, in a carbonium ion mechanism, the rate of C-N bond cleavage could be reduced by the two fi-2H atoms through hyperconjugation; C-H bond cleavage may also be reduced by the nonleaving 2H. An equilibrium effect, due to the presence of this atom in the urocanate produced, could also be postulated.
Secondary isotope effects arising from such causes are usually even srnaller than those reported here, particularly for 4-fluorohistidine (31). In the deamination of a mixture of L-[U-14C]histidine and (P-S)-L-[3Hl]histidine, a small but reproducible isotope discrimination (1.1 to 1.3) was observed. This would be expected to be even larger in the absence of the tritium exchange reaction. A similar isotope discrimination was reported in the deamination of (&!+L-[3Hl]phenylalanine and of (/KS-I-[3Hl]tyrosine by phenylalanine ammonia-lyase. Hanson et al. suggested a concerted elimination mechanism for both substrates (5, 32, 33). Such results, together with those obtained with 4-nitro-L-histidine7 are more compatible with a primary isotope effect. A final caveat, in the interpretation of all isotope effects in enzymic reactions, is that loss of a substrate proton from some group on the enzyme may be kinetically important; in such a case, the size of the isotope effect might be expected to be independent of the substrate used.
If the reaction is considered to follow an "ordered uni-bi" mechanism (in the terminology of Cleland (5, 24)), it is possible to estimate V,,,, for the reverse reactions, using the Haldane relationship K,, = vZ;,;;Ki,h-~a 2 h,s where K,, is the equilibrium constant, 8, and 112 are the maximum velocities for the forward and back reactions, and the other K values are the Michaelis constants for each substrate. From the data in Fig. 2 and Table IV, the values obtained are 0.04 and 0.0025 units/mg for urocanate and 4-fluorourocanate, respectively. The experimental results are in proper order and indicate that formation of an amino enzyme intermediate, while relatively slow, is probably not the rate-limiting step of the reverse reaction for the two substrates. Whereas EDTA markedly inhibits both forward reactions, it inhibits the reverse reaction with fluorourocanate much more than with urocanate. This result is puzzling and is being explored further.
In the rate-lirniting LY ,p elimination, the slow step may be rupture of the C-H or C-N bond, or concerted rupture of both. i2s shown in Table III, tritium exchange into histidine is somewhat faster than urocanic exchange. If isotope discrimination were taken into account, the difference could be even greater. This result, together with the facts that tritium exchange can be observed under initial rate conditions which do not support significant urocanate exchange (Figs. 5 and 6), and that the rate of tritium exchange is not significantly dependent on the urocanate concentration , suggests that C-H bond cleavage may precede C-N bond cleavage slightly (9). Separation of the steps may not be sufficient to invoke a carbanion mechanism, as postulated by Bright (35) for the deamination of P-methylaspartate. In the latter case, C-N bond cleavage is considered rate-limiting, and a large tritium exchange is inhibited by the elimination product, mesaconate. Alternatively, if the rate of urocanate dissociation is not significantly faster than the rate of the preceding elimination reaction, a difference in the rates of the urocanate and tritium exchanges could also be observed in a concerted mechanism.
It is not clear why 4-fluorohistidine is such a poor substrate for the enzyme. Although the fluorine atom has a marked effect in depressing the basicity of the imidazole ring, it has little effect 011 that of the primary amino group.3 The absence of a significant isotope effect or1 the K, values of both substrates suggests that both may be essentially dissociation constants. If so, the analog seems to bind to the enzyme more strongly than does histidinc. Similarities in isotope effects and in the inhibitory effect of EDTA suggest that the mechanism of the reaction is the same in both cases. Perhaps the differences in V,,, result from the inability of the analog to force the enzyme into an active form, as required by the "induced-fit" theory (36).