Pyruvoyl-dependent Histidine Decarboxylases MECHANISM OF CLEAVAGE OF THE PROENZYME FROM LACTOBACZLLUS BUCHNERP

of

That this mechanism, initially demonstrated for a mutant proenzyme with a slowed activation rate, also pertains for activation of wild-type proHisDCase under growth conditions * This study was supported in part by Grants AM 19898 and AI 13940 from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
3 Present address, The Rockefeller University, New York, NY
was demonstrated by supplying [hydr~xy- '~O]serine to the growing cultures of Lactobacillus 30a and showing that while control serine residues from the a chain of the subsequently isolated HisDCase contained only in their hydroxyl groups, the carboxyl-terminal serine residue from the j3 chain contained equal amounts of lSO in both the hydroxyl and carboxyl groups (2).
Pyruvoyl-dependent HisDCases from three other bacterial species are now known (1). Although proHisDCase has been detected in only one of these, Lactobacillus buchneri (l), the finding (3) that HisDCases from all three organisms have the same sequence at the j 3 chain COOH terminus and the same sequence at the a chain NH2 terminus as that shown for the Lactobacillus 30a enzyme in Equation 1 implies that all four enzymes arise from inactive proenzymes that share a common activation sequence and that activation proceeds by the same mechanism in each case. To test this implication, we grew L. buchneri with [hydroxyl-"O]serine under the same conditions used previously for Lactobacillus 30a and found to our surprise that none of the serine residues from the isolated HisDCase contained "0. However, as described below, when pyridoxamine was omitted from the growth medium to minimize serine degradation and synthesis, the same transfer of "0 observed in Lactobacillus 30a was observed in L. buchneri.

EXPERIMENTAL PROCEDURES
Fractionation and Immobilization of Rabbit Antibodies to Histidine Decarboxylase-Rabbit antiserum (100 ml) to histidine decarboxylase from Lactobacillus 30a was prepared as described previously (4). The precipitate that formed upon addition of sodium sulfate at 25 "C to 18% concentration was collected by centrifugation, dissolved in 50 ml of 0.15 M NaC1, and reprecipitated by addition of sodium sulfate (final concentration, 16%). The precipitate was collected by centrifugation, dissolved in 50 ml of 17 mM potassium phosphate, pH 6.3 (Buffer A), and dialyzed against this buffer for 24 h at 0°C. The solution was then clarified by centrifugation and applied to a column of DE-52 (30 ml) equilibrated with buffer A (5). Protein-containing fractions obtained by elution of the column with Buffer A were pooled, and the precipitate that formed upon addition of ammonium sulfate at 0 ' C to 50% of saturation was collected by centrifugation, dissolved in 15 ml of 0.15 M NaCl, and dialyzed against 0.15 M NaCl, 1 mM sodium borate, pH 8.4 (Buffer B) for 24 h at 0 "C. The yield of immunoglobulin, determined spectrophotometrically at 280 nm using . E: % = 15, was 490 mg. Agarose (30 ml, Sepharose 4B) suspended in 120 ml of 1 M sodium carbonate was activated with 5 g of cyanogen bromide in 5 ml of acetonitrile for 2 min at 0 "C (4), washed with 400 ulin in 40 ml of buffer B. After 14 h at 4 "C less than 1% of the ml of ice-cold water, and gently stirred with 450 mg of immunoglobprotein remained in solution. The gel was washed with 1 liter each of Buffer B and 0.1 M ammonium acetate, pH 4.8, and stored in Buffer B.
Purification of HisDCase from L. buchneri-Cultures (3 liters) were grown to stationary phase on crude medium (6) or on the defined medium of Guirard and Snell (7) modified to contain L-histidine (5 g/liter) and ~-[hydroxyl-'~O]serine (15 mg/liter). In some cases the defined medium was further modified by omission of pyridoxamine and addition of D-glUtamiC acid (30 mg/liter). The cells were har-vested by centrifugation, washed with cold water, and acetone dried. Acetone-dried cells (2 g) were suspended in 20 ml of 0.2 M ammonium acetate, pH 4.8, and subjected to sonic oscillation for a total of 9 min using an Ultrasonic model W-375 sonicator. Insoluble material was removed by c e n t~f u~t i o n . T h e p H of t h e s u~r n a t a n t solution was adjusted to neutrality by addition of 5 N NaOH, and the solution was stirred gently with immunoglobulin-agarose (15 ml suspended in 30 ml of Buffer B) at 25 "C. After 30 min no HisDCase activity remained in the supernatant solution. The gel was washed with 500 ml each of Buffer F3,O.l M ammonium acetate, pH 4.8, and Buffer B. The washed gel was then incubated with 15 mi of 0.2 M glycine-HCI, pH 2.5, for 5 min to elute the bound enzyme. The supernatant was collected, neutralized by addition of 1.5 ml of 1 M potassium phosphate, pH 7.6, reduced in volume by ultrafiltration using an Amicon YM-10 membrane, dialyzed against water for 24 h, and l y o p~~z e d .
Purity of preparations was confirmed by sodium dodecyl sulfate/polyacrylamide gel electrophoresis (8). The yield of purified enzyme was around 5 mg. Other M e t~-S y n t h e s i s of ~-[ h y d~x y~-l~]~r i n e and procedures used for trimethylsilylation and gas chromatographyfmass spectrometry have been described (2). Histidine decarboxylase activity was determined manometrically (9).

~~n t i~~a t~n of the COOH-ter~ina~ Amino Acid of the #3
Chain-Digestion of the j 3 chain of histidine decarboxylase from L. buchneri with carboxypeptidase resulted in the rapid release of nearly 1 serine residue/mol of @ chain (Fig. 1). Alanine, threonine, and met~onine were released at slower rates. When carboxypeptidase digestion was carried out in '80-enriched water, no "0 label was present in the serine carboxyl poup while the other amino acids released contained the theoretical amount of "0 in their carboxyl groups ( Table  I). These results confirm that serine is the COOH-terminal residue of the /3 chain and that the serine residues released by carboxypeptidase digestion were derived solely from the COOH terminus of the chain. They agree fully with the known carboxyl-te~inal sequence (-~-T-T-A-S) for the #3 chain of this protein (3).
Analysis of Histidine Decarboxyhe from Cells Grown in the Presence of t-[hydroxyl-lsO]Seriw-"-hen histidine decarboxylase was isolated from L. bwhneri grown on defined medium containing 33 atom % ~-[hydroxyl-'*O]senne, essentially no label was present in the serine residues of either the cy or the /? chain ( Fig. 2A). When the defined medium was modified by the addition of all D-and L-amino acids required for cell growth and omission of pyridoxamine to minimize  2B). Serine residues from the (Y chain contained 31 atom % "0 in the hydroxyl group and essentially no l8O in the carboxyl group. The COOH-terminal serine of the #3 chain contained 33 atom % "0 in the hydroxyl group, and in addition 31% of the carboxyl groups of this residue contained 1 atom of lSO (Fig. 2C and Table 11). No lSO was present in the other amino acid residues of the enzyme. This result was also observed with HisDCase from Lactobacillus 30a and demonstrates that the hydroxyl oxygen of a serine residue is incorporated into the newly formed carboxyl group at the COOH terminus of the /3 chain during proenzyme activation in both organisms.

DISCUSSION
Lactobacillus 30a is a nutritionally demanding organism; it requires both vitamin B-6 and each of the amino acids, including serine, for growth (7). its incorporation of added ['%]serine into HisDCase, as reported earlier (2), is, there- buchneri HisDCase. A, serine residues released from the @ chain by digestion with carboxypeptidase Y in H2'0. Unlabeled standard and serine residues from the (x and j3 chains isolated from cells grown on defined medium containing both ~-[hydroxyl-"O]serine and vitamin B-6 (pyridoxamine) showed this same spectrum. B, serine residues from the a! chain or C, the COOH-term^^ serine residue from the @ chain, both from HisDCase of cells grown on defined medium containing ~-[hydroxyZ-'%O] serine but modified by omission of vitamin B-6. The spectral peak at m/z 204 represents Me& serine (unlabeled) that has lost the fragment containing the carboxyl group of serine; that at 218 is the fragment that lacks the -CHzOSiMes group (see references in Ref. group. fore, the expected result. The failure of L. buchneri to behave similarly under the same conditions demonstrates that the labeled serine was either degraded, diluted out by biosynthesized serine, or both. This circumstance would normally preclude use of ["O]serine for examination of the activation mechanism for the proHisDCase of this organism. However, many amino acids ( i n c l u~n g serine) become nutrition~ly essential for lactic acid bacteria that normally synthesize them when vitamin B-6 is absent or supplied at very low levels (10)(11)(12). These observations and the fact that the known metabolic pathways leading to and from serine all include pyridoxal-P-dependent enzymes suggested that L. buchneri, dependent HisDCases known share a common amino acid sequence a t the presumed activation site, indicate that this mechanism for proenzyme cleavage and formation of the essential pyruvoyl residue is general for this group of enzymes. Whether it also accounts for formation of the essential pyruvoyl group of other enzymes that contain this residue is not yet known.