On the Mechanism of 5-Enolpyruvylshikimate S-Phosphate Synthetase*

SUMMARY The enzymic synthesis of 5-enolpyruvylshikimate 3-phos-phate (ES-3-P) from enolpyruvate phosphate (labeled with IsO in the C-O-P oxygen) and shikimate 3-phosphate occurred with C-O cleavage of the pyruvate ester. ES-3-P formed in a Da0 reaction medium showed incorporation of approximately 1.3 atoms of deuterium, and consisted of 39% dz molecules, 53 % dl, and 8% do. Enolpyruvate phosphate isolated from the same reaction mixture had 0.72 atom of deuterium with a similar dz:dl ratio. In both compounds, the nuclear magnetic resonances of the vinyl methylene protons were diminished equally in intensity. ES-3-P formed in a tritiated medium contained 0.42 atom of 3H in the vinyl methylene hydrogens. A reversible addition-elimination mechanism is proposed in which protonation of carbon 3 of enolpyruvate is associated with a nucleophilic attack on carbon 2 by the 5-hydroxyl group of shikimate 3-phosphate. A methyl group of unrestricted rotation is formed in the resulting postulated intermediate. Elimination of orthophosphate yields ES-3-P, and elimination of shikimate j-phosphate

13. SPRISSON** From the Department of Biochemistry, Columbia New Yodi 10032 SUMMARY The enzymic synthesis of 5-enolpyruvylshikimate 3-phosphate (ES-3-P) from enolpyruvate phosphate (labeled with IsO in the C-O-P oxygen) and shikimate 3-phosphate occurred with C-O cleavage of the pyruvate ester.
ES-3-P formed in a Da0 reaction medium showed incorporation of approximately 1.3 atoms of deuterium, and consisted of 39% dz molecules, 53 % dl, and 8% do. Enolpyruvate phosphate isolated from the same reaction mixture had 0.72 atom of deuterium with a similar dz:dl ratio. In both compounds, the nuclear magnetic resonances of the vinyl methylene protons were diminished equally in intensity. ES-3-P formed in a tritiated medium contained 0.42 atom of 3H in the vinyl methylene hydrogens.
A reversible addition-elimination mechanism is proposed in which protonation of carbon 3 of enolpyruvate phosphate is associated with a nucleophilic attack on carbon 2 by the 5hydroxyl group of shikimate 3-phosphate.
A methyl group of unrestricted rotation is formed in the resulting postulated intermediate.
Elimination of orthophosphate then yields ES-3-P, and elimination of shikimate j-phosphate yields enolpyruvate phosphate. A procedure was developed for preparing enolpyruvate phosphate with '*O in the C-O-P oxygen. The isolation of shikimate 3-phosphate was improved. ES-3-P synthetase from Salmonella was purified 80-fold by a simple procedure.
ES-3-P" synthetase represents a rare type of reaction in \yllicll the enolpyruvyl moiety of enolpyruvyl phosphat,e is transferred apl)arently unchanged to a recipient molecule. The E-3-1' synthetase reaction was previously considered (1) a reversible addit,ion-elimination reaction (Scheme I). Protonation of carbon 3 of enolpyruvate-P, facilitated through electron donat,ion by the ester oxygen, was assumed to be associated with a nucleophilic attack on carbon 2 of enolpyrurnte~l' by the 3-hydrosyl group of shikimate-3-P.
Elimination of l'i then yielded ES-3-l'. This mechanism is supported by studies of the rcnct'ion in l&O and in 3H20, and with [lsO]enolpyruvatc-l', which are described in the present report. EXPERI>IEN'AL PROCIWURF, ['*O]Enolpy~uvate-P--,\ solution of 3.0 g of freshly recrystallized bromopyruvic acid (2) in 5 g of H&80 (20.1 atom r0 excess) eras allowed to stand in the dark for 24 houl,s at room temperature. The water was removed by lyophilization and was reutilized in an identical manner to incorporate I80 into t,wo additional 3.0-g lots of bromopyruric acid. The combined products were recrystallized three times from purified, dry chloroform t,o yield 7.6 g of white crystals. Behavior on melting: first tranait,ion 52-55"; m.p. of glass, 73.5-76" (2). '*O concentration (3) 10.30 f 0.02 atom 7. excess (average of t,hree independent analyses). The IsO concentration in the carboxyl group was dctermincd by decarboxylating the bromopyruvic acid in diphenylaminediphenylmethane (4) 5). &uantit,ative yields of (IO2 R-cre obtained at reflux temperat,ure (270') m 5 min, but considernl)lc reaction occurred also at, 100". The following I80 values IT-ere obtained in COZ evolved under the indicated conditions: (n) at 100" for 25 min, 8.79 atom y0 excess; (b) discarding the CO2 after the above t.reatmcnt, and refluxing for 5 min, 8 yield) by-the method of Clark and Kirby (6,7).3 The I80 concentration of this product was 5.25 f 0.09 atom y0 excess (average of three analyses), or the same atoms of I80 escess per mole, within experimental error, as in the bromopyruvic acid (6 X 5.25/100 = 0.315; 3 X 10.3/100 = 0.309). Although decarboxylation of enolpyruvate-1' in refluxing diphenylamine-diphenylmethane gave essentially quantitative yields of OS, its 180 concentration (6.4 to 6.9 atom '% excess) was variable and lower than expected from the value found in the decarboxylation of the bromopyruvic acid. C'*O2 obtained by refluxing in quinoline showed somewhat higher results (7.87 atom y0 excess). These low isotope values were caused by exchange with phosphate oxygen& probably as a result of internal anhydride formation (8), since exchange was also shown to occur between inorganic phosphate and carboxyl oxygens under these vigorous conditions. nL,-Alanine (8 mg) was heated for 6 min under reflux with 13 mg of KH#*04 (29.1 atom $ZO excess) in a mixture of 1 g each of diphenylamine and diphenylmethane. The evolved COz contained 7.18 atom y0 excess l*O.
Preparation of Xhiki,nate-S-P---Several minor changes were introduced into the procedure of Weiss and Mingioli (9). Aerobatter aerogenes A-170-40 was grown as previously described (9) wit,11 rapid shaking (500 ml in a 2-liter flask) for 90 hours at 37", and cells were removed by centrifugation.
The culture fluid (12 liters) was assayed for shikimate-3-P by determining shikimate (loj on appropriate samples before and after hydrolysis with potato acid phosphatase. Found were 1020 mg of shikimate-3-P and 50 mg of shikimate per liter. In a typical hydrolysis, a solution (1 ml) of 0.5 pmole of shikimate-3-P, 500 pmoles of acetate buffer, pH 5.0, and 0.4 mg of potato acid phosphatase, was incubat,ed at 37" for 3 hours, heated at 100" for 3 min, and clarified by centrifugation.
A 0.2.ml sample of the supernatant solution was analyzed for shikimate (10). The results were in good agreement with values obtained by bioassay (1).
Granulated charcoal (400 g of Nuchar C190, 30 mesh) was heated in boiling water for 30 min and allowed to settle overnight.
Floating particles were removed by decantation, and the procedure was repeated.
A chromatographic column (50 x 7.5 cm) fitted with a coarse sintered glass plate and a layer of glass wool was filled with water, a slurry of the charcoal was poured into the column while water was withdrawn at the bottom. The charcoal (43 x 7.5 cm) was covered immediately with a 2-to 3.cm layer of Berkshire sand.
The culture filtrate was brought to pH 3 with HCl, treated 3 We are grateful to Professor V. hl. Clark for placing this method at our disposal prior to publication. with 0.5 ml of toluene, and loaded on t)he column at a rate of 300 to 400 ml per hour.
The column was then washed with 2 liters each of the following solvents adjusted to pH 2.5 with HCl: wat,er, 1.5, 5, and 10% ethanol, and finally with 500 ml of 5y0 unacidified ethanol.
Shikimate-3-P was eluted with 2500 ml of 5% ethanol, 3 liters of 10% ethanol, and 2 liters of 25% ethanol. The fractions were combined and concentrated in a vacuum. The residual colorless syrup was diluted with water to 150 ml and contained 7 g of shikimate-3-P.
The solution was adjusted to pH 7 with NH,OH and treated with 90 ml of 0.5 M barium acetate. Absolute ethanol (950 ml) was added to give 80% (v/v) ethanol, and the solution was allowed to stand overnight at 0". The precipitate was collected by centrifugation, washed twice with 200 ml of 75% ethanol and once with 200 ml of absolute ethanol, and dried over P205 in a vacuum.
Yield was 12.8 g of a white powder.
Continued elution of the column with 10 liters of 257, ethanol gave a slightly yellow syrup (assaying 2.0 g of shikimate-3-l') from which were isolat.ed 2.9 g of slightly tan barium salt.
Both preparations of shikimat,e-3-l' gave one spot in descending chromatography on paper with n-butyl alcohol-acetic acidwater (100:6:2, v/v) (11) and with n-propyl alcohol-15 M NHIOH-water (6:3:1, v/v) (la), an acid molybdate spray (12) being used to detect phosphate esters. Chromatography on Dowex l-Cl-at pH 9 with a lithium chloride gradient gave one symmetrical peak (1). A minor phosphorylated impurity was detected in paper electrophoresis for 1 hour at pH 5.5 and 900 volts (0.1 in. acetate buffer; length of paper, 50 cm). This minor component may have resulted from isomerization of shikimate 3-phosphate to 4-phosphate (9). Hydrolysis of both barium salts with alkaline phosphatase gave rise t'o 1 .O mole of shikimate and 1.0 mole of Pi per equivalent weight of 500 (9).
Assay of ES-S-P Xynthetase Activity-Disappearance of enolpyruvate-P was measured as previously described (13). Active fractions obtained by chromatography of SaZmonelZa extracts on DEAE-cellulose were free of phosphatase and could be studied by Pi release (14).
Formation of [31f]RS-S-P in Tritiated Reaction ilfixture-ES-3-P was synthesized, on the preparative scale previously described, with a 0.4 to 0.6 saturated (NH&SO4 fraction of E. coli K-12 58-278 (1). The reaction mixture (500 ml) contained approximately 400 mCi of 31-IZ0, and was incubated at 37" for 80 min. The yield of barium salt of ES-3-P was 144 mg, which was 90 y0 pure by 2,4-dinitrophenylhydrazine assay and 87 y0 pure by bioassay (1). A sample (48 mg) was dissolved in 17 ml of water, and BaES-3-P was reprecipitated with 34 ml of absolute ethanol to yield 34 mg of product.
Bioassay indicated a purity of 95%; radioactivity, measured as described below, was unchanged. Samples of the barium salt were combusted in O2 in sealed Vycor tubes, the resulting water was added t'o scintillation mixture, Issue of October 25, 1971 Bondinell, Vnelc, Knowles, Xprecher, ant1 Xprinson 6193 and counted under standard conditions (16). Alternatively, samples were combusted in a Packard "Oxidizer" and the resulting water was counted similarly.
;1 sample of the original incubation mixture (1.5 ml) was freeze-dried, the condensed water was diluted loo-fold (w/w), and 0.0500 g was assayed by scintillation counting under standard conditions.
A total activity of 399 f 8 mCi (average of four independent analyses) was obtained for the 3Hz0 present in the reaction mixture, or 3.20 x lOlo dpm per mole.
The solution was evaporated to dryness in a vacuum, and the residue was treated several times with water and evaporated to dryness. The barium salt of shikimate-3-P was isolated. It had no radioactivity.
A similar reaction mixture was incubated as described above except that shikimate-3-l' was omit,ted. Enolpyruvate-P was isolated by chromatography (I), precipitated as barium salt in the presence of a 6-fold excess of barium acetate, and converted to the crystalline tricyclohexylammonium salt. The latter was pure by enzymic assay and had no radioactivity (yield, 50% based on initial enolpyruvate-P).
Parijbtion of l&!-P Synthetase from Salmonella typhi-Irlzeium-Wild type cells of Salmonella LT-2 were grown overnight at 37" in 5 ml of enriched medium (0.2% casein hydrolysate and 0.270 yeast extract), and 1 ml was transferred into 50 ml of the same medium.
These cultures were shaken at 37" for 8 hours, and IO-ml aliquots were inoculated into 500 ml of minimal medium (17) in 2-liter flasks. The cells were grown with rapid shaking at 37" for 6 hours, harvested by centrifugation in the cold, and washed once with cold 0.05 LI Tris succinate buffer, pIT 6.8 (this buffer, pH 6.8, was used throughout, and all operations were at 2-4").
The yield of wet cells was 38 g from 7 liters of medium.
Chilled suspensions of cells (1 g/4 ml of 0.01 JI buffer) were disrupted by subjecting 50.ml portions to oscillation in a h1.S.E. loo-watt ultrasonic disintegrator for three periods of 23 min each, alternating with 30-s rest periods, and were centrifuged at 2" for 45 min at 45,000 x g. The resulting cstracts contained approximately 25 mg of protein (18) per ml, and had a specific activity of 2.2 units (micromoles of ES-3-P per mg per hour).
They were either stored at -15" or chromntographed as described below.
Whatman DEAE-cellulose (DE-52, 150 g) was stirred vigorously three times in 1 liter-portions of 0.01 M buffer and fine particles were removed as completely as possible.
It was then stirred overnight in 500 ml of 0.2 M buffer, fine particles were again removed, and the remainder was washed thoroughly with 0.01 JI buffer and used to prepare a column, 2.5 x 45 cm. The crude ext,ract (140 ml) was applied to the top of the column under slight pressure, and chromatographed with a linear gradient of 1000 ml of 0.01 M buffer in the mixing chamber and 1000 ml of 0.2 ht buffer in the reservoir.
Fractions of 10 ml were collected at a flow rate of 0.5 ml per min. Protein concentration was estimated from the ratio of absorption at A260:A280. A sharp peak of ES-3-P synthetase activity4 appeared in tubes 59 to 67 (approximately 2.0 mg of protein per tube) which was divided into three fractions comprising tubes 61 through 63 4 Phosphatase activity followed immediately after this peak, and was separated owing to its low levels in cells grown as described above.
Fraction I was divided into l-ml portions which were mixed with 1 ml of glycerol and stored at -15". Activity remained unchanged for over 1 year.
Isolation of [2H]ES-S-P from Reactions Conducted in D,O-Monocyclohexylammonium enolpyruvate-P (6, 7) was deionized, neutralized to pH 7.0 with KOH, and its purity checked with pyruvate kinase coupled to lactate dehydrogenase (19). Potassium shikimate-3-P was prepared similarly and analyzed for esterified phosphate (14). Solutions containing, respectively, 0.5 mmole of enolpyruvate-I', 0.5 mmole of shikimate-3-P, 25 mmoles of Tris chloride (pH 7.1) and 5 mmoles of KF, were brought to dryness by rotary evaporation in a vacuum, and exchanged twice with DzO (99.8 mole '%) by dissolving the residues in this solvent and removing it by evaporation.
The above components were dissolved in DSO, mixed, made up to 450 to 490 ml, and warmed to 37".
In one experiment the enzyme preparation was a fraction obtained by chromatography of a Salmonella extract as described above, and was dialyzed against saturated (NH&S04. The precipitate was separated by centrifugation, and dissolved in 50 ml of 0.01 M Tris buffer in DzO (pD 6.8). The protein concentration was 0.27 mg per ml with a specific activity of 44, thus affording approsimately 600 units. In a second experiment, 4 ml of Fraction I (stored in 50% glycerol as described above, approximately 500 units) was diluted to 20 ml with 0.05 hr Tris succinate buffer in DzO, pH 7.0, and concentrated to 1 to 2 ml by ultrafiltration.
This procedure was repeated three times and the concentrated solution was diMed with 8 ml of DzO buffer.
The enzyme solution was added to give a final volume of 500 ml, and the course of the reaction at 37" was followed by assaying enolpyruvate-P disappearance. Equilibrium was reached at 500/, conversion in 105 min, and the reaction was st'opped after 135 min by adding 1.0 N KOH to pH 8.0 and heat,ing in a boiling water bath for 5 min. DzO was removec poration in a vacuum, and the residue was dissolved ml of water. Shikimate-3-P, enolpyruvate-P, and ES-3 separated by chromatography on Dowex l-X8 (Cl-) and isolated as barium salts (1).
Conversion of Barium Salts to Methyl Esters and Ethers-The barium salts of shikimate-3-P, enolpyruvate-P, and ES-3-P were converted to the acids with Dowex 50-H+, neutralized with KOH, and treated with twice the calculated amount of AgNO, solution.
Ethanol (2 volumes) was added, the mixture was stored at 4" overnight, and the silver salt was separated by centrifugation, washed with 67 y0 ethanol, and dried in a vacuum. A suspension of the powdered silver salt in methyl iodide was treated with an excess of AgzO, refluxed with stirring for 3 hours, cooled, diluted with ether, and filtered (20). The clear fi1trat.e was evaporated to dryness, and t'he residue was dissolved in chloroform and applied to a 20.cm2 preparative thin layer chromatography plate (silica gel, 2 mm thick) which was then developed with acetone-chloroform (1:4 v/v). Product bands were located by staining the edges with iodine vapor, scraped off the plate, and eluted with chloroform.
Mass Spectra--Mass spectra of the methyl derivatives were obtained on a CEC 21.1lOC spectrometer q-it11 perfluorokerosene Vol. 246,No. "0 as a reference. The concentrat'ion of deuterated species was responding to 4 x 3.18 = 12.7 for 180 in tile C-O-P oxygen. determined by enlarging the molecular ion region, and calculat-This value was very close to the 13.2 atom y0 excess 180 calcuing intensities of ;\I+, M+ + 1, and M+ + 2 from the average lated for the ester linkage of the enolpyruvate-P. The syntheof several scans (21). Correct#ions were made for natural sis of ES+l', therefore, took place with C-O cleavage of the abundances from the spectra of standards. enolpyruvate-P. Nucleur Magnetic Resonance Spectra-Solutions of t,he barium salts of enolpyruvate-I' and ES-3-l' were deionized, neutralized with KOH to pH 7.5, evaporated to dryness, and exchanged several times wit,11 DzO. NXR spectra were recorded at an apl"'oximate concentration of 0.25 M on a Ovarian A-60 spectrometer. Chemical shifts were determined in parts per million downfield from an internal standard of sodium 2,2-dimethyl2silapentane-5sulfonate.
Spectra of the methyl derivatives were obtained in CDC13 with tetramethylsilane as internal standard. The extent of 2H incorporation was determined by integrating NMR spectra of labeled and unlabeled salts of ES-S-P and trimethyl est#ers of enolpyruvate-P (in which the O-methyl proton resonances served as internal standards).
A similar incubation of shikimate-3-P and enolpyrucate-1' was conducted in the presence of Y130. The ES-3-P was isolated, purified as previously described (l), and reprecipitated to constant activity.
The product had an activity of 6.62 X 10" dpm per mole, corresponding to 21y0 of t'he specific activity of the medium (i.e. 21y0 of 2 atoms of FI, or 0.42 atom).
In a similar experiment. % was not incorporated into enolpyruvat'e-1' when incubated Tvith enzyme in the absence of shikimate-3-P.

RESULTS
The first step in the synthesis of [180]enolpyruvate-r~atc-P was labeling of the carbonyl oxygen of bromopyruvate by exchange with HJ*O.
Analysis of the CO2 obtained by thermal decarboxylation of the bromopyruvatc in diphenylamine-diphenylmethane showed considerable incorporation of 180 into the carboxyl group (8.84 atom 7. excess). The I*0 in the carbonyl oxygen, calculated by difference between the total I80 and that of the carboxyl group, was 13.2 atom '$& excess.
A further investigation was undertaken of the W-3-1 synt'hctase reaction in D,O.
A much more purified enzyme preparation. was used in this work, and reactants as well as ES-3-I' \vei'e isolated and studied by N;LIR and mass spectrometrg.
The results of the second experiment described under "Experimental Procedure" are shown in Table I. Essemially ident,ical results were obtained in the first cspcrime~lt.
The atoms of I80 excess per mole in both bromopyruvate and enolpyruvate-P were identical, i.e. 0.309 and 0.315, respectively. Although the I80 of the COZ from bromopyruvate decarboxylation was constant (see "Experimental Procedure"), the isotope concentration in CO2 from enolpyruvate-P was variable and lower than expcct'ed, presumably owing to exchange with phosphate oxygen atoms.
(A similar labeling of COz occurred when alaninc was decnrboxylated in the presence of KH2P1804). The 180 of the C-O-P oxygen was, therefore, assumed to be the same as that of the carbonyl oxygen of bromopyruvate, i.e. 13.2 atom y0 excess.
The 60 MHz NMR spectrum of ES-3-l', Fig. 1, showed 2 vinyl methylene protons as doublets at' 6 4.71 (I-1,) and 5.21 (He): J ye?n = 2.5 Hz. Assignment of these resonances, I-IA frans to the carboxyl and HB cis to the carbosyl, was based on the observation that a /3 proton cis t,o the carboxyl group of an a,/% unsaturated carboxylic acid absorbed downfield relative to the corresponding tram /I proton (22). The vinyl met'hylene prot'ons of chorismic acid were assigned in a similar manner (23). The ring proton resonances, from 60 and 100 MHz spectra, were analyzed and assigned in a straightforward manner (Table II), and closely resembled those i n the spect,rum of shikimic acid (24), except that H-3 in ES-3-P was also coupled with l)hoxphorns, J3,p = 8.5 Hz.
The [l*O]enolpyruvate-P was incubated with shikimate-3-P and a bacterial extract containing ES-3-P synthetase.
The Pi formed in the reaction contained 3.18 atom y0 excess 180, cor- The spectrum of enolpyruvate-P showed the vinyl methylene protons at 6 5.18 (I-1,) and 5.35 (II,), each as two overlapping doublets owing to geminal coupling (Jsn) and coupling with phosphorus.
This assignment has recently been cst,ablishetl unequivocally by Cohn et al. in a KMR study of 1 -[W]enolpyru vatc-P (25). It was not possible to evaluate separately each coupling constant, but from t,he separation between the outer peaks of the appareut triplets at 6 5.18 and 5.35, it' was deter mined that JAB + JAp = 2.4 Hz, and JAn + Jnp = 2.8 Hz, respectively.
Similarly, th e vinyl prot,ons of trimethyl enolpyrurate-P appeared at 6 5.63 (lZI,) and 5.97 (Hn), each as t,\vo overlapping doublets, and it was determined that JBs + JAP = 4.85 Hz, and JAB + Jnr = 4.65 Hz, respectively. It may be seen that the relative magnitudes of Jap and JBP of enolpyru vate-P have been reversed in the trimethyl ester (cf. Reference 25). The protons of the carboxylic and phosphate ester methyl groups appeared, respectively, as a singlet at 6 3.88, and a douhlet at 6 3.92 (JH,p = 11.5 Hz). Labeled species of ES-3-P and trimethyl enolpyruvate-I' showed diminished intensities only in resonances assigned to vinyl methylene protons.
Deut.erium was equally distributed between HA and Hn in both compounds.
In 60 MHz spectra Hn of [2H]ES-3-P was quite distinct', but I-IA was obscured by other resonances.
However, in a 100 MIIz spect,rum at, 60" HA and Hn were well resolved singlets of equal intensity. small amount of unlabeled molecules. Integration of the 60 MHz spectra showed that 1.1 atoms of deuterium was incorporated into ES-3-P,5 and 0.75 atom into trimethyl enolpyruvate-I'.
The amount of deuterium labeling observed by NMR was corroborated by mass spectrometry (Table I). All the methyl derivatives gave intense molecular ions suitable for determination of deuterium concentration by comparison of spectra obt'ained with the labeled and unlabeled compounds. By this method ES-3-P had 1.3 atoms deuterium, and enolpyruvate-P had 0.72 atom deuterium.
The proportion of labeled species was higher in ES-3-P than in enolpyruvate-P, but the ratio of singly labeled to doubly labeled molecules was essentially the same in both compounds.
Shikimate-3-P was unlabeled, as would be expected. 5 Assuming HA = HB, the value obtained for HB was multiplied by 2.
The mechanism of ES-3-P formation has previously been considered (1) as a reversible addition-elimination reaction (Scheme 1). It was assumed that protonation of carbon 3 of enolpyruvate-P, resulting from electron donation by the ester oxygen, was associated with a nucleophilic attack by the 5-hydroxyl group of shikimate-3-P on carbon 2. Elimination of Pi from the postulated intermediate could yield ES&P, and in reverse direction, elimination of shikimate-3-P would yield enolpyruvate-P.
This mechanism predicts C-O cleavage of the ester bond and exchange of hydrogen from the medium with vinyl methylene protons of ES-3-P and enolpyruvate-P.
Both predictions have been realized in the experiments described above. The Pi released in the reaction contained all of the I80 present in the ester oxygen of the enolpyruvate-P.
When the reaction was carried out in 3Hz0, 21% of 2 atoms of tritium (0.42 atom) was found in ES-3-P.
It is reasonable to assume that the tritium was introduced on the enolpyruvyl side chain of ES-3-P, since there are no known enzymic reactions that labilize the carbon-bound hydrogen atoms of shikimate-3-P. Furthermore, acid hydrolysis of [3H]ES-3-P gave shikimate-3-l' devoid of radioactivity.
The tritium must have been introduced during ES-3-P formation, since enolpyruvate-P did not become labeled when incubated with enzyme extract in the absence of shikimate-3-P. Although low labeling in ES-3-P could have resulted from isotope effects, the relatively small incorporation of tritium from the medium made it difficult to interpret the result in relation to mechanism of the reaction.
ES-3-P and enolpyruvate-P were therefore isolated from an enzymic synthesis in DsO. As shown in Table I atom of deuterium was introduced into ES-3-P from a nearly proton-free DzO medium.
Mass spectrometry of pentamethyl ES-3-P showed 1.3 atoms of deuterium distributed in 8% do species, 53% dl, and 39% dz. The further observation that the vinyl proton magnetic resonances were equally diminished in intensity indicates that the two possible dl species A similar distribution of deuterium was found in enolpyruvate-P, although only 50% of the molecules were labeled compared to 92y0 in ES-3-P.
These results indicate that a methyl group with unrestricted rotation was formed at carbon 3 of enolpyruvate-P (Scheme 1). It is not clear why isotopic equilibrium was not reached between the vinyl methylene hydrogens of ES-3-P and those of enolpyruvate-P under the experimental conditions described above. In the absence of initial rate data on deuterium incorporation into product and starting material it is not possible to calculate isotope effects for the forward and reverse steps of the equilibrium. However, the difference in extent of labeling between ES-3-P and enolpyruvate-P suggests that the postulated intermediate breaks down faster to ES-3-P than to enolpyruvate-P.
Isotope effects in protonation of vinyl carbon atoms as well as elimination reactions may account for the 3-fold greater incorporation of deuterium than of tritium into ES-3-P under similar experimental conditions. Addition-elimination mechanisms are prevalent in reactions of enol ethers. The acid-catalyzed synthesis of enol ethers from enolic compounds and CH3180H was found to occur with complete transfer of 180 to the products (26). The results were interpreted as suggesting an addition-elimination mechanism for acid-catalyzed synthesis and hydrolysis of enol ethers. A similar conclusion was reached in a study of enol ether hydrolysis (27).
The only other reaction known in which the enolpyruvyl moiety of enolpyruvyl-P is transferred apparently unchanged, occurs in the formation of UDP-N-acetylenolpyruvylglucosamine, an intermediate in the synthesis of UDP-A-acetylmuramic acid (28,29). The mechanism of the latter reaction is probably similar to that of ES-3-P synthetase.