Comparative studies of human recombinant 74- and 54-kDa L-histidine decarboxylases.

We have expressed and characterized human recombinant 74-kDa (rHDC74) and 54-kDa (rHDC54) L-histidine decarboxylases (HDCs) in Sf9 cells. By immunoblot analysis, rHDC74 and rHDC54 were shown to be localized predominantly in the particulate and soluble fractions, respectively. rHDC74 exhibited histamine-synthesizing activity equivalent to that of rHDC54. The existence of 74- and 54-kDa HDCs was also confirmed in the particulate and supernatant fractions of the cell lysate, respectively, from the human basophilic leukemia cell line KU-812-F. The ratio of HDC activity to immunoreactivity was similar for the two forms of the enzyme. The specific activity of purified rHDC54 (1.12 mumol/mg/min) was comparable to those of HDCs from other mammalian tissues or cells. The purified rHDC54 was eluted as a monomer form from a Superdex-200 column; the molecular mass of the enzyme was approximately 54 kDa on SDS-polyacrylamide gel electrophoresis without 2-mercaptoethanol. The HDC activity of rHDC54 significantly decreased on dialysis against buffer without pyridoxal 5'-phosphate; addition of pyridoxal 5'-phosphate to the dialysate readily increased in the enzyme activity to the original activity. Taken together, these results suggest that human HDC functions as both 74- and 54-kDa forms having equivalent HDC activity, which are localized in the particulate and soluble fractions, respectively, and that the latter form exhibits its activity as a monomer form.

Histamine plays a pivotal role in a wide variety of reactions, such as inflammation, gastric acid secretion, neurotransmission, and cell growth (Kahlson and Rosengren, 1968;Beaven, 1978;Woolley, 1993). L-Histidine decarboxylase (HDC, 1 EC 4.1.1.22) is the only enzyme that catalyzes the formation of a biogenic amine from its precursor, L-histidine (Schayer, 1966). HDCs fall into two classes: those that use pyridoxal 5Ј-phosphate (PLP) as a cofactor and those that have a pyruvoyl coenzyme (Recsei and Snell, 1984). The sequences of the HDCs in the first class from Enterobacter aerogenes, Klebsiella planticola, Morganella morganii, Drosophila melanogaster, mouse, rat, and human show strong homology (Sandmeier et al., 1994). There is spectral or chemical evidence for PLP in a number of these enzymes, but not in the human enzyme. The sequence of the HDC from Lactobacillus 30a, which contains the pyruvoyl coenzyme, is quite different.
Many investigators demonstrated that HDC regulates the level of histamine in tissues or cells, and that modulation of HDC activity under various conditions markedly affects the level of the amine (Schneider et al., 1987;Mamune-Sato et al., 1990;Yatsunami et al., 1993). It is thus of great importance to clarify the mechanism underlying regulation of the HDC activity.
Recently, we reported the structure of the human HDC gene, and proposed the possibility that HDC activity is transcriptionally regulated (Yatsunami and Ohtsu et al., 1994). Accumulated evidence suggests that HDC activity is also regulated at the translational level. HDC cDNAs have been isolated, and characterized, from human basophilic leukemia cells (Yamauchi et al., 1990), human erythroleukemia cells (Zahnow et al., 1991), fetal rat liver (Joseph et al., 1990), and mouse mastocytoma P-815 cells (Yamamoto et al., 1990). The molecular masses of HDC enzymes (74 kDa), calculated from amino acid sequences derived from the respective cDNAs, are larger than those of the purified HDC subunits (53-54 kDa) by about 20 -21 kDa. It is, therefore, conceivable that primary translated products are processed after translation to yield mature enzymes and that such post-translational processing of HDCs changes the activity or nature of the enzymes.
Until recently, however, it had not been clarified whether or not a primary translated 74-kDa HDC has activity, because it has never been isolated from any tissue or cells. Most recently, mouse recombinant 74-kDa HDC exhibiting enzyme activity was expressed in Sf9 cells . However, it was not clearly shown whether or not the enzyme activity was due to the recombinant 74-kDa HDC itself, because the mouse recombinant 74-kDa HDC was unstable and its purification was unsuccessful. In addition, it is totally unknown whether 74-kDa HDCs from other mammalian sources have a molecular nature similar to that of mouse 74-kDa HDC. Although the 54-kDa HDCs purified so far characteristically resemble each other, the nature of 74-kDa HDCs might differ among man, mouse, and rat in some respects, because their C-terminal regions are not so homologous (Mamune-Sato et al., 1992).
To clarify whether or not a 74-kDa HDC has HDC activity, and whether and how 74-and 54-kDa HDCs take part in histamine synthesis, we have expressed cDNAs for human 74and 54-kDa HDCs using a baculovirus system. To our surprise, the human 74-and 54-kDa HDCs exhibited equivalent enzyme activity, and the latter existed as a monomer form.

MATERIALS AND METHODS
Construction of cDNAs for 74-and 54-kDa HDCs-A human HDC cDNA, pTN-2 (Yamauchi et al., 1990;Mamune-Sato et al., 1992), in the Okayama-Berg cDNA cloning/expression vector (Okayama and Berg, 1982) was kindly provided by Dr. Kohei Yamauchi (Tohoku University, * 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. Japan). Recombinant transfer vectors for 74-and 54-kDa HDCs, pVL-HDC74 and pVLHDC54, respectively, were constructed with pTN-2 and the transfer vector, pVL1392, as shown in Fig. 1. Briefly, human recombinant 74-kDa HDC cDNA (pcHDC74) was obtained by deleting a long 5Ј-untranslated sequence in pTN-2 using a specific primer that contains an XbaI site at its 5Ј-end. The sense oligonucleotide used for the construction of pcHDC74 was 5Ј-GGTCTAGAACTAGTGATC-CCAAGGGAGATGATGGA-3Ј, whereas human recombinant 54-kDa HDC cDNA (pcHDC54) was constructed by deleting the 3Ј-end of pcHDC74 using a specific primer, which contains a stop codon just downstream of the Gln-477 codon and a BamHI site at its 3Ј-end. The antisense oligonucleotide used for the construction of pcHDC54 was 5Ј-AGGTGGATCCTTACTGACTCAGGATGAGAG-3Ј. The amplification was performed with a GeneAmp PCR System 9600 (Perkin Elmer), with 25 cycles of denaturation (94°C, 15 s) and annealing plus extension (68°C, 2 min) using Ampli-Taq DNA polymerase (Takara, Kyoto, Japan). pcHDC74 and pcHDC54 were then digested with XbaI and BamHI, and then subcloned into the XbaI-BamHI sites of the pBluescript and pVL1392 vectors. The sequences of the obtained mutants were confirmed with an Automated Laser Fluorometry DNA Sequencer II (Pharmacia LKB, Uppsala, Sweden). Recombinant plasmids pVL-HDC74 and pVLHDC54 (25 g each) were cotransfected into 1.5 ϫ 10 6 Sf9 cells along with 1 g of wild-type AcNPV DNA. The recombinant viruses, referred to as vHDC74 and vHDC54, were plaque-purified as described (Ausubel et al., 1987).
Cell Culture-A human basophilic leukemia cell line, KU-812-F cells, was cultured at 37°C in RPMI 1640 medium supplemented with 10% fetal bovine serum. Cloned Spodoptera frugiperda (Sf9) cells derived from fall armyworm ovaries were maintained as a monolayer culture at 26°C in TNM-FH medium supplemented with 10% fetal bovine serum.
Infection of Sf9 Cells and Extraction of rHDC74 and rHDC54 -Sf9 cells (2 ϫ 10 6 cells/cm 2 ) were incubated at 26°C with either vHDC74 or vHDC54 at a multiplicity of infection of approximately 10. Cell monolayers were harvested by scraping from the dishes 2 or 3 days after infection. Cell pellets were obtained by centrifugation at 800 ϫ g for 5 min, washed twice with phosphate-buffered saline, suspended in 100 mM potassium phosphate (pH 6.8) containing solution A (0.2 mM dithiothreitol, 0.01 mM pyridoxal 5Ј-phosphate, and 2% polyethylene glycol 300), and then disrupted by sonication (20 kHz, 20 s, twice) with a Bioruptor UCD-200TM (Cosmo Bio, Tokyo, Japan). The sonicates were centrifuged at 800 ϫ g for 5 min to remove undisrupted cells.
Subcellular Distribution of HDC in KU-812-F Cells-Cultured KU-812-F cells were washed, suspended in 100 mM potassium phosphate (pH 6.8) containing solution A, and then disrupted by sonication as described above. The sonicates were centrifuged at 800 ϫ g for 5 min to remove undisrupted cells. The resultant supernatants were designated as homogenates. The homogenates were centrifuged at 100,000 ϫ g for 60 min, and then the precipitates obtained were suspended in the original volume of the same buffer.
Production of an Anti-HDC Peptide Antibody-According to the hydrophilic (Hopp and Woods, 1988) and antigenic values (Welling, 1985), a peptide corresponding to amino acids Ser-218 to Lys-232 in the cDNA-derived human HDC sequence was selected. The peptide was synthesized as a multiple antigen peptide (MAP) (Posnett et al., 1988) by the Fmoc (N-(9-fluorenyl)methoxycarbonyl) solid phase method (Merrifield, 1963) with a model 9050 peptide synthesizer (MilliGen/ Biosearch, Framingham, MA). After deprotection with trimethylsilyl bromide/trifluoroacetic acid, the MAP was desalted on a Sephadex G-25 column and then lyophilized.
For production of antiserum, three female Japanese white rabbits were immunized by intradermal injection of MAP (0.1 mg in phosphatebuffered saline) in complete Freund's adjuvant (1:1), and given boosters in incomplete Freund's adjuvant (1:1) at intervals of 2 or 3 weeks. After the fourth booster, each rabbit was bled and antiserum was obtained.
Immunoblotting-The lysate from Sf9 cells, infected with vHDC54 or vHDC74, was subjected to SDS-polyacrylamide gel electrophoresis on a slab gel (4 -20% gradient gel), and the separated proteins were then transferred electrophoretically to a polyvinylidene difluoride membrane in 10 mM CAPS, pH 11, for 2 h at 1 mA/cm 2 . After preincubation with Tris-buffered saline (20 mM Tris-HCl, pH 7.5, containing 150 mM NaCl) containing 5% skim milk, the membrane was incubated with anti-MAP antiserum (1:3000) in Tris-buffered saline containing 5% skim milk for 1 h at room temperature. The detection of the immobilized HDC was performed using an ECL Western blotting kit (Amersham, United Kingdom).
HDC Assay-Histidine decarboxylase activity was assayed by the o-phthalaldehyde method (Shore et al., 1959) as described previously .
Purification of Recombinant 54-kDa HDC-All purification steps were carried out at 0 -4°C.
Sf9 cells (approximately 1.5 ϫ 10 9 cells) transfected with vHDC54 were sonically disrupted as described above. The sonicate was then centrifuged at 100,000 ϫ g for 60 min. The resulting supernatant was used as the crude extract. rHDC54 was purified from the crude extract (559 mg) by ammonium sulfate fractionation and three successive column chromatographies with phenyl-Sepharose HP, Resource Q, and Superdex 200. The purification conditions employed for each step were essentially identical to those used for the purification of mouse HDC , except that 80 mM NaCl was used to elute HDC from the phenyl-Sepharose HP column (mouse HDC was eluted from the column by 5 mM NaCl). The purified enzyme (34 g) was stored at Ϫ80°C.

SDS-Polyacrylamide Gel Electrophoresis and Molecular Mass
Estimation-Polyacrylamide gel electrophoresis in the presence of SDS was performed on slab gels (4 -20% gradient gels) according to Laemmli (Laemmli, 1970). Proteins were stained with a silver stain kit (Daiichi Pure Chemicals Co., Ltd., Tokyo, Japan). As molecular mass standards, an electrophoresis calibration kit (Daiichi Pure Chemicals Co., Ltd, Tokyo, Japan) was used.

Subcellular Distribution of Recombinant 74-and 54-kDa
HDCs-Human recombinant 74-and 54-kDa HDCs, rHDC74 and rHDC54, respectively, were constructed in a baculovirus vector. Gln-477 was given as the C-terminal amino acid of rHDC54, because the C-terminal residue of purified fetal rat liver HDC was reported to be Gln-479, which corresponds to Gln-477 in the human HDC sequence, based on the results of carboxypeptidase digestion of the enzyme (Mizuguchi, 1992). An anti-HDC antibody raised against a synthesized HDC peptide was used to detect rHDC74 and rHDC54. HDC activity expressed by the lysate from the Sf-9 cells was time-and dose-dependent.
A crude extract from Sf9 cells, which had been infected with vHDC74, was fractionated by centrifugation at 100,000 ϫ g for 60 min. An aliquot of each fraction was subjected to HDC assaying and immunoblot analysis. As shown in Fig. 2A, a strong band for rHDC74 was observed for the precipitate fraction. The intensities of the bands of rHDC74 in respective lanes were nicely correlated with the amounts of the respective HDC activities loaded. No other protein bands that were correlated with the amounts of HDC activity were detected. Fig. 2B shows the subcellular distribution of rHDC54. In contrast to rHDC74, an intense band for rHDC54 was also exclusively observed for the 100,000 ϫ g supernatant fraction.
Only 5% of the total HDC activity in the crude extract was recovered in the 100,000 ϫ g precipitate fraction.
When the intensities of the bands were compared between rHDC74 and rHDC54 on a same gel, equivalent amounts of HDC activity of the two species were found to give protein bands with apparently similar intensities (data not shown).
Pyridoxal-5Ј-Phosphate Dependence of Recombinant 54-kDa HDC-In order to determine whether or not the human HDC requires PLP for its activity, rHDC54 in the 100,000 ϫ g supernatant fraction was dialyzed and then assayed for HDC activity. As shown in Table I, dialysis of the enzyme for 16 h against buffer in the absence of PLP lowered the HDC activity to 30% of the original level, whereas the addition of PLP to the dialysate restored the activity to the original level.
Subcellular Distribution of HDC in the Human Basophilic Leukemia Cell Line, KU-812-F-In order to determine whether or not a particulate HDC exists in KU-812-F cells, cultured KU-812-F cells were fractionated by centrifugation as described under "Materials and Methods." Fig. 3A shows the distribution of the activities of HDC and lactate dehydrogenase, a cytosolic marker enzyme. Approximately 71% of the total HDC activity was recovered in the 100,000 ϫ g supernatant fraction, whereas 29% of the total HDC activity was distributed in the particulate fractions. Less than 3% of the total activity of lactate dehydrogenase was detected in the particulate preparation. Fig. 3B shows the result of immunoblot analysis for the particulate and the supernatant fractions. As shown, a strong band with a molecular mass of 54 kDa was observed for the supernatant fraction, whereas for the particulate fraction, a weak but a distinct band corresponding to a molecular mass of 74 kDa was detected. Quantitative analysis by densitometry demonstrates that the ratio of HDC activity to immunoreactivity is similar for the two forms of the enzyme (Fig. 3C).
Purification of Recombinant 54-kDa HDC-It is indispensable to clarify that recombinant HDCs reflect the nature of native mammalian HDCs. To characterize rHDC54, Sf9 cells were infected with vHDC54 on a large scale and then the cells were lysed. A 100,000 ϫ g supernatant fraction was prepared from the cell extract and purified by four purification steps as described under "Materials and Methods." The specific activity of rHDC54 in the 100,000 ϫ g crude extract (1.41 nmol/min/mg) was about 20 times higher than that of the crude HDC from KU-812-F cells. The specific activity of the purified rHDC54 was 1.12 mol/mg/min, and the overall recovery of HDC activity was 4.8%. Fig. 4 shows a typical elution profile on Superdex 200 column chromatography. HDC activity was eluted in fractions where a globular protein with a molecular mass of 54 kDa would be. When the supernatant fraction from cultured KU-812-F cells was applied on a Superdex-200 gel filtration FPLC column, HDC activity was eluted at the same position as the purified rHDC54 was (data not shown).

Characterization of the Purified Recombinant 54-kDa HDC-
Fractions from the Superdex 200 column were subjected to SDS-polyacrylamide gel electrophoresis in the absence of 2-mercaptoethanol and the gel was silver-stained. As shown in Fig. 4 (inset), a single intense band corresponding to a molecular mass of about 54 kDa was exclusively obtained for the active fractions. Immunoblot analysis showed that the anti-HDC antibody only reacted with the single 54-kDa subunit protein (data not shown). DISCUSSION Mammalian HDCs have been characterized and purified from various types of cells and tissues Taguchi et al., 1984;Martin and Bishop, 1986;Grzanna, 1984). These studies, however, focused only on a soluble 53-54-kDa HDC, although the primary translated HDCs are 74 kDa in size. The purpose of this work was to clarify whether and how 74-and 54-kDa HDCs take part in histamine synthesis. Resolution of this question will provide a clue as to the mystery that mammalian HDCs have to exist in two forms.
An anti-HDC peptide antibody, which recognizes only denatured HDC, was raised and used for the detection of human recombinant 74-and 54-kDa HDCs. The peptide sequence, Ser-218 to Lys-232, in the human HDC sequence is highly hydrophilic as well as antigenic and was selected to completely match those of the mouse (Ser-225 to Lys-239) (Yamamoto et al., 1990) and rat (Ser-221 to Lys-235) (Joseph et al., 1990)

Studies of Human Recombinant L-Histidine Decarboxylases
HDCs. Immunoblot analysis using the antibody demonstrated that 87% of the total human rHDC74 expressed in Sf9 cells existed as an insoluble form with histamine-synthesizing activity. Although quantitative determination by immunoblot analysis has limitations, equal amounts of HDC activity of the two species apparently gave bands of equal intensity of rHDC74 and rHDC54, suggesting that the specific activities of the two HDCs are equivalent. It is unlikely that rHDC74 was cleaved to yield rHDC54 during the enzyme assay, and the latter exhibited HDC activity because no processing from rHDC74 to rHDC54 was observed during incubation of the former with histidine for 3.5 h at 37°C (data not shown). When 74-and 54-kDa human HDC cDNAs were subjected to in vitro transcription and translation, synthesis of the respective HDC proteins was observed. In addition, both the in vitro translated proteins exhibited HDC activity (data not shown). Overall, we concluded that human rHDC74 possesses as much histaminesynthesizing ability as rHDC54. Fig. 3 suggests that the insoluble HDC is not an artifact in the present expression study, but that a fairly large amount of an insoluble 74-kDa HDC exists and functions. Approximately 29% of the total HDC activity was found to exist in the insoluble fraction from human leukemia cells, KU-812-F, which is the source of human HDC cDNA. Possible contamination by soluble HDC of these fractions was ruled out because less than 3% of the lactate dehydrogenase activity was detected in either fraction. Immunoblot analysis using an extract of KU-812-F cells supports the existence of 74-kDa HDC in the cells, and the same ratio of activity to immunoreactivity for the 74-kDa and 54-kDa forms. Although the figure contains multiple bands corresponding to higher and lower molecular masses, there is no contradiction between the intensities of the two immunoreactive bands and the respective HDC activities. We have no evidence ruling out the possibility that these bands were derived from aggregation or processing of the respective enzymes, but we tentatively conclude that these bands appeared nonspecifically for the following reasons. First, the activity to immunoreactivity ratio of the two forms is not contradictory. Second, an anti-peptide antibody often gives nonspecific bands when samples are very crude and the amounts of samples, applied to a gel, are very large, which was the case in this experiment.
If, as we have suggested, an insoluble 74-kDa HDC does contribute to histamine synthesis in physiological tissues and cells, how might soluble 54 and particulate 74-kDa HDCs participate in it? This is a difficult question because the molecular nature of the particulate HDC is totally unknown, and we could FIG. 3. Detection of the insoluble and active 74-kDa HDC in KU-812-F cells. A homogenate (800 ϫ g supernatant) from KU-812-F cells was fractionated by centrifugation as described under "Materials and Methods," and then assayed for the activities of HDC and lactate dehydrogenase. Each value is the mean Ϯ S.E. for four determinations (A). The supernatant and precipitate fractions were subjected to immunoblot analysis (B). The intensities of bands corresponding to the 74-and 54-kDa HDCs, indicated by arrows in B, were calculated with an Image analyzer (PDI, Huntington Station, NY) using the software Quantity One®. The result is presented as a percentage of the value obtained for 54-kDa HDC (C). Sup., supernatant; Ppt., precipitate.
FIG. 4. Superdex 200 FPLC column chromatography of the recombinant 54-kDa HDC. The active fractions (114 g) from a Resource Q column were applied to a Superdex 200 FPLC column, and then gel filtration was carried out as described under "Materials and Methods." Aliquots of fractions from the Superdex 200 FPLC column were subjected to SDS-polyacrylamide gel electrophoresis in the absence of 2-mercaptoethanol, and then the gel was silver-stained (inset). q, enzyme activity; OO, A 280 . The molecular mass markers were: blue dextran (2,000 kDa), ferritin (440 kDa), aldolase (158 kDa), bovine serum albumin (66 kDa), ovalbumin (43 kDa), chymotrypsinogen A (25 kDa), and ribonuclease A (13.7 kDa). not determine its cellular location by means of immunohistochemistry using the anti-HDC peptide antibody we raised. We assume that a particulate 74-kDa HDC binds to cellular membranes and promptly responds to signals from outside the cell. In contrast to stored histamine, newly formed histamine is believed to act soon after its synthesis at the site or in the cell where it is formed (Kahlson and Rosengren, 1968). The newly formed histamine, sometimes called induced histamine or nascent histamine, plays an important role in cell growth (Kahlson and Rosengren, 1968;Brandes and LaBella, 1993), platelet aggregation (Saxena et al., 1989), and gastric acid secretion (Kahlson and Rosengren, 1968;Glavin and Brandes, 1988). It is possible that a particulate 74-kDa HDC synthesizes newly formed histamine, whereas a soluble 54-kDa HDC produces stored histamine. In any case, the hydrophobic C-terminal 20-kDa regions of mammalian 74-kDa HDCs probably participate in anchoring of the proteins to cellular membranes. These assumption requires further investigation.
The HDC activity of rHDC54 significantly decreased on dialysis for 16 h against buffer without PLP ( Table I). The HDC activity was not completely abolished on the dialysis, probably because 16 h was not long enough to completely remove PLP from the sample. The addition of PLP to the dialysate readily increased the enzyme activity to the original level. These results strongly suggest that the human HDC is PLP-dependent.
Human rHDC54 was purified by the same procedure as that used for the purification of HDC from mouse mastocytoma cells . In addition, the specific activity of the purified human rHDC54 (1.12 mol/mg/min) was proved to be equivalent to that of the mouse and rat HDCs (800 and 260 nmol/mg/min, respectively), suggesting that human rHDC54 is functionally similar to native HDCs from mammalian tissues or cells. This is reasonable because the homology of the 54-kDa HDC molecules is considerably high among three mammals (Mamune-Sato et al., 1992). However, the activity of 74-kDa HDCs might differ among them because the homology in their C-terminal 20-kDa is not so striking. Assuming that the Cterminal regions of HDC molecules affect their folding and thereby influence the respective HDC activities, the activity of 74-kDa HDC in mouse or rat might not be similar to that of the respective 54-kDa HDC.
Purified human rHDC54 was shown to be eluted as a monomer form with an apparent molecular mass of 54 kDa on gel filtration on a Superdex-200 column. Only a single band of 54-kDa HDC was detected on SDS-polyacrylamide gel electrophoresis of the eluted protein, even in the absence of 2-mercaptoethanol. Until recently, HDC has been purified to homogeneity only from rat liver (Taguchi et al., 1984) and mouse mastocytoma cells . In both cases, active HDCs were reported to exist as a dimer form. A monomer and active HDC was neither detected nor isolated. Human HDC is the first example of an HDC existing as a monomer form. It is unlikely that the active monomer HDC, found in the present study, was an artifact produced during the process of purifica-tion of the recombinant HDC, because HDC activity derived from cultured KU-812-F cells was also eluted at the same position as the purified rHDC54 was on a Superdex-200 gel filtration column. The specific activity of the purified human monomer, rHDC54, was equivalent to that of the dimer HDCs from mouse and rat. However, we cannot rule out the possibility that the C terminus of 54-kDa human HDC is not identical to those of other HDCs, which could contribute to their dimerization. Further studies are needed to clarify this point.