Cloning and Tissue Distribution of a Novel Serine Protease esp-1 from Human Eosinophils

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS ARTICLE NO.

252, 307–312 (1998)

RC989645

Cloning and Tissue Distribution of a Novel Serine Protease esp-1 from Human Eosinophils Masahiro Inoue,* Naotomo Kanbe,† Motohiro Kurosawa,‡ and Hiroshi Kido* *Division of Enzyme Chemistry, Institute for Enzyme Research, University of Tokushima, 3 Kuramoto-cho, Tokushima 770-8503, Japan; †Department of Dermatology, Gunma University School of Medicine, 3-39-22 Showa-machi, Maebashi, Gunma 371-8511, Japan; and ‡Department of Geriatric Medicine, Hirosaki University School of Medicine, 5 Zaifu-cho, Hirosaki, Aomori 036-8562, Japan

Received September 29, 1998

We have cloned a novel serine protease designated as esp-1 from human eosinophils. The amino acid sequence deduced from the cDNA showed that ESP-1 comprises a signal peptide of 18 amino acids, a propeptide of 23 amino acids, an active form sequence of 273 amino acids starting from an Ile-Val-Gly-Gly-Glu motif, the catalytic triad of serine proteases that has been characterized as the essential amino acid residues for the proteolytic activity, and a hydrophobic amino acid stretch in the carboxyl terminus, suggesting this enzyme is a novel membrane-type serine protease. The tissue distributions of esp-1 expression revealed that this protease is not only expressed in human eosinophils, but also widely expressed in mononuclear cells and various tissues other than skeletal muscle and kidney and is most abundant in testis and prostate, and moderately so in lung, spleen and pancreas. © 1998 Academic Press

A large number of differentiated blood cells express characteristic serine proteases, and each protease in these cells plays a specific role in patho-physiological functions in acute and chronic inflammation. Although granzymes from cytotoxic T cells and natural killer (NK) cells (2), cathepsins, elastase, collagenase and gelatinase from neutrophils (3), and tryptase and chymase from mast cells (4) have been extensively studied, the proteases from eosinophils and basophils have not been characterized, because of the limitation of the numbers of these cells in the blood. In order to characterize a novel serine protease(s) in human eosinophils and to deduce its function in the progression of an allergic disease, we attempted the cloning of a serine protease by reverse transcription-polymerase chain reaction (RT-PCR) from purified eosinophils of atopic patients with eosinophilia. In this paper, we first report a novel membrane-bound serine protease deduced

from the amino acid sequence in human eosinophils, determine the distribution of the gene in various organs, and discuss the role of the enzyme. MATERIALS AND METHODS Eosinophil isolation. Peripheral blood was obtained from patients with eosinophilia with informed consent. The isolation of eosinophils and neutrophils was carried out according to the methods previously described (5). Cloning of a novel serine protease sequence from eosinophils. Total RNA was isolated from human eoshinophils by the guanidiumthiocyanate-chloroform method (6). One mg of total RNA was subjected to cDNA synthesis using SuperscriptII (Life Technologies) in the presence of an Oligo-dT-NotI primer: 59-AACTGGAAGAATTCGCGGCCGCAGGAATTTTTTTTTTTTTTTTTTV-39. After the cDNA synthesis, PCR was carried out using a combination of degenerate oligonucleotides that encode the conserved amino-acid sequence within the active sites of the serine proteases, His and Ser (7). After PCR, the products were separated on a 2% agarose/TAE gel. The 400-600 bp bands were excised, and ligated to TA-Vector including pGEMT-easy (Promega) and PCR2.1 (Invitrogen). White colonies were randomly picked up and their sequences were verified. To obtain the full length cDNA, the modified 59and 39RACE described below were carried out. One mg of total RNA was subjected to cDNA synthesis using SuperscriptII in presence of the Oligo-dT-NotI primer and Cap-switch oligoII (CSII):59-AAGCAGTGGTATCAACGCAGAGTACGCGGG-39. After the cDNA synthesis reaction, excess primers were removed with a PCR-purification kit (Quiagen). The resultant cDNA was used for the following PCR. Modified 59Race: PCR was carried out using esp-1 68: 59-CCAAACTGGACCATCCACC-39 and a PCR primer: 59-AAGCA-GTGGTATCAACGCAGAGT-39 (Clontech). Heminested PCR was carried out using esp-1 49: 59-CGGAGGGATCACTAAGGTCAC-39and the PCR primer (Clontech). The PCR products were applied to a 2 % agarose/ TAE gel. The 300 bp band was excised and subcloned into PCR2.1, and the sequences of two independent clones were determined. 39RACE: First PCR was performed using a NotI primer: 59CTGGAAGAATTCGCGGCCGCAGG-39and esp-1 424: 59-GGAGACAT-GGTTTGTGCTGGC-39. Heminested PCR was performed using esp-1 456: 59-CGGGAAGGATGCCTGCTTCG-39and the NotI primer. The PCR products were applied to a 2 % agarose/TAE gel. The 400 bp band was excised, and ligated to PCR2.1 (Invitrogen), and then the sequences of the two independent clones were verified. Amplification of full length cDNA of esp-1. PCR was carried out using the following parameters: One min denaturation at 95°C,

307

0006-291X/98 $25.00 Copyright © 1998 by Academic Press All rights of reproduction in any form reserved.

Vol. 252, No. 2, 1998

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

followed by 35 cycles of 15 sec denaturation at 95°C, 30 sec annealing at 62°C, and 2.5 min extension at 72°C, with final extension of 7 min at 72°C. The primers, esp-1 -9: 59-GAGGAG-GCCATGGGCGCGC39and esp-1 1058: 59-CCTGCAAGGCATCAACTGG-AATGTG-39, were used for the amplification of the cDNA. The 1100 bp band on a 1% agarose/TAE gel was excised. The purified DNA fragment was subjected to subcloning into PCR2.1. Five independent clones were picked up and all their sequences were examined. Esp-1 expression in eosinophils but not in neutrophils. The purity of CD16-negative granulocytes (eosinophils) was checked by RT-PCR using the primers of FcgRIII, oligonucleotides, 465: 59-TCATTTGTCTTGAGGGTC-39and 466: 59-GTCTCTTTCTGCTTGGTG39, described elsewhere (8). Expression of the esp-1 gene was also determined by RT-PCR using the primers, esp-1 -9 and esp-1 68, under the following conditions: One min denaturation at 95°C, followed by 35 cycles of 15 sec denaturation at 95°C, 30 sec annealing at 62°C, and 1.5 min extension at 72°C, with final extension of 7 min. A b-actin primer set purchased from Stratagene was used for RTPCR according to the manufacturer’s protocol except for an annealing temperature of 62°C instead of 55°C. Separated mononuclear cells, granulocytes, and CD16-positive granulocytes (neutrophils) were used as controls for all RT-PCR. Each panel contained 100 bp ladder marker (Life Technologies) in the left lane, with reversetranscriptase negative RT-PCR on the right. The amount of cDNA used for PCR was approximately 50ng. The PCR products were run on 2% agarose/TAE gels. The gels were stained with an ethidium bromide solution, followed by transfer of the products to Hybond N1 membranes. The Hybond N1 membranes except those with b-actin products were hybridized with 32P-labeled oligonucleotide probes: 59-ACAAACATTTGAAGC-TCA-39for FcgRIII and esp-1 49 for esp-1. Tissue distribution of the esp-1 gene. The human multiple tissue cDNA, MTC Panels I and II, was purchased from Clontech and used for measurement of the expression level of esp-1. PCR was performed according to the manufacturer’s protocol. The pair of primers used in these reactions was esp-1 778S: 59-AGCTG-GGGAGTGGGCTGTGGTC-39and esp-1 949AS: 59-ATGGGCTCAGGTAG-GCTCAGAC-39. PCR was carried out under the following conditions: One min denaturation at 95°C, followed by 32 cycles of 15 sec denaturation at 95°C, 30 sec annealing and extension at 68°C, with final extension of 7 min. The PCR products were run on 2% agarose/TAE gels and visualized by ethidium bromide staining, transferred to a Hybond N1 membrane (Amersham Pharmacia Biotech) and then hybridized with a 32P-labeled oligonucleotide probe: 59-AAGCTGATGGCCCAGAGTGG-3. Amplification of the glyceroaldehyde-3phosphate dehydrogenase gene was performed according to the manufacturer’s protocol. Northern blot analysis of the esp-1 gene. Total RNA and mRNA were isolated from Hela S3 cells and used for Northern blot analysis. mRNA (0.5 mg) was separated on a 0.7% formaldehyde-/MOPS agarose gel and then transferred to a Hybond N1 membrane. The blot was hybridized with 32P-labeled full length esp-1 cDNA. The size of the esp-1 transcript was estimated in comparison with the sizes of 28S and 18S RNA, as markers. Expression of the ESP-1 protein in HEK293 cells. The esp-1 gene in PCR2.1 was amplified by PCR, and then subcloned into PME18S-FL by blunt-end ligation. The pair of primers used for the amplification was esp-1 -9 and esp-1 970: 59-GGGCTCAGGTAGGCTCAGACCG-39. The sequence of the esp-1 gene cloned in PME18SFL(PME-esp1) was confirmed. Two mg of PME-esp-1 was transfected into HEK293 cells using FUGENE6 (Boehringer Mannheim). Two days after the transfection, cells were lysed with 100ml of SDS-gel loading buffer containing 2-mercaptoethanol, directly added to a 6-well culture dish, and then sonicated extensively. Fifteen ml aliquots of cellular lysates derived from PME-esp-1 transfected and mock-transfected cells were separated by 10-20% gradient SDSPAGE and then transferred to a PVDF membrane (Millipore). The blot was incubated with affinity-purified specific rabbit antiserum

against the ESP-1 peptide, followed by a horseradish peroxidaselabeled anti-rabbit IgG antibodies. The resultant blot was visualized with ECL (Amersham Pharmacia Biotech) according to the manufacturer’s protocol.

RESULTS AND DISCUSSION Identification of a Novel Serine Protease in Human Eosinophils We searched for a novel serine protease(s) by PCR using the degenerate oligonucleotides in the cDNA of purified eosinophils. For this purpose, we selected the consensus sequences of various serine proteases around the active-site residues, His and Ser, for a set of primers, and succeeded in finding a novel serine protease (7, 9). This novel serine protease, designated as esp-1, consists of 1082 bp with a Kozak sequence (10), 59-ggccatgg-39, and a poly A additional sequence, as shown in FIG. 1. A blast sequence homology search showed that the putative amino acid sequence of ESP-1 is most similar to that reported for human prostasin (11). The overall identity is 40.85%, suggesting that this ESP-1 is not an isoform of the known protease. The putative amino acid sequence of ESP-1 starts at nucleotide 10, as judged with the Kozak’s rule (10). The signal peptide sequence of ESP-1 is composed of 18 amino acids, as determined with a Nakai server, that predicts the structure and intracellular localization of a protein. The N-terminal sequence of the pro-form of ESP-1 starts at Arg19 and ends with Val314. The activated form of ESP-1 starts at the Ile-Val-Gly-Gly-Glu sequence at residues 42-46, indicated by the filled arrow, as judged from the homology of the N-terminal residues of various activated serine proteases (12). The active-site residues that comprise catalytic triad of serine proteases, such as His, Asp and Ser, are residues 82, 137 and 238, respectively, in ESP-1, as judged from the homology with other serine proteases (9). Potential N-glycosylation sites are located at Asn167 and Asn273. In addition to the structural similarity of ESP-1 to other serine proteases, this ESP-1 protein possesses a unique stretch of carboxyl-terminal hydrophobic amino acids at residues 299-314. These data suggest that the ESP-1 protein is a membrane-type serine protease. It will be a challenge to reveal whether this protease stays in the ER, Golgi or plasma membrane. esp-1 Gene Expression in Eosinophils but Not in Neutrophils Random sequencing of the PCR products that encode both the His and Ser residues revealed sequences identical to those of caldecrine (13) and mast cell tryptase (14) in addition to the novel sequence of esp-1. To rule out the possibility of contamination by neutrophils, that contain various kinds of proteases (3), we highly

308

Vol. 252, No. 2, 1998

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

fractionated granuloctes by MACS using CD16 monoclonal antibodies (Miltenyi Biotec GmbH). The fractionated cells were subjected to RT-PCR analysis to check the messages for FcgRIII (CD16), esp-1 and b-actin. As a result, amplification of CD16- positive granulocytes, that mainly consist of neutrophils, showed an amplified 142 bp band of CD16, while CD16negative granulocytes, that mainly consist of eosinophils, did not (FIG. 2a). The data indicate that there is little contamination by neutrophils of the CD16negative granulocytes. In the mononuclear cells, the expression of FcgRIII turned out to be positive (Fig. 2a). This might be due to the presence of NK cells and macrophages among the mononuclear cells (8). On the other hand, RT-PCR analysis of esp-1 gene expression revealed an amplified 193 bp band of esp-1 for CD16negative granulocytes and mononuclear cells, but not CD 16- positive granulocytes (FIG. 2b, top). The small differences in the levels of amplified products of FcgRIII and esp-1 between fractionated CD16negative cells and total granulocytes might be due to too many cycles of PCR or the incubation with CD16 monoclonal antibodies on ice for 1 hr in the process of separation of neutrophils and eosinophils, that decreases the expression levels of esp-1 and FcgRIII but not that of b-actin. Each blot was hybridized with a 32 P-labeled probe, that ensures the identification of the corresponding gene (FIG. 2a and b, bottom). The amounts of cDNA used in these PCRs were determined by PCR using b-actin primers, and they turned out to be identical (FIG. 2c). The overall results indicated that the esp-1 gene is expressed in eosinophils and mononuclear cells, but not neutrophils. Although the physiological role of ESP-1 has not been clarified, it may play a pivotal role in the extravasculization of eosinophils through destruction of the basement membrane, or the activation of other proteases such as metalloprotease and/or gelatinase. To prove these hypotheses, we aim to obtain functional recombinant proteins and monoclonal antibodies to ESP-1. Expression of the esp-1 Gene in Various Human Tissues

FIG. 1. The nucleotide and deduced amino acid sequences of the esp-1 gene. A putative signal sequence is underlined. Asterisks(*) indicate potential sites of the active-site residues of a serine protease. The closed arrow shows an Ile-Val-Gly-Gly-Glu sequence, the possible N-terminal sequence of an active serine protease. A putative polyadenylation signal is dashed-underlined. Sharps (#) indicate potential glycosylation sites.

MTC-panels I and II were used to study the expression levels of esp-1 in various human tissues in comparison to the results of Northern blot analysis, for two major reasons: Rare messages can be detected, and semi-quantitation of rare messages is possible. The results showed that this gene was expressed not only in eosinophils but also was highly expressed in testis and prostate, moderately so in lung, pancreas and spleen, weakly so in thymus, colon and peripheral blood leukocytes (PBL) (FIG. 3, top). Esp-1 expression was detected in neither kidney nor skeletal muscle (FIG. 3, top). Hybridization using a 32P-labeled oligonucleo-

309

Vol. 252, No. 2, 1998

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

FIG. 2. Expression of esp-1 mRNA in eosinophils. Fractionated mononuclear cells and granulocytes, that mainly consist of eosinophils: CD16(2) and neutrophils: CD16(1) were subjected to RT-PCR using primers of FcgRIII (CD16) (a), esp-1 (b), and b-actin (c). Top: ethidium bromide staining of the gel. The 100bp markers are on the left and the RT-negative control in the right lane. Arrows indicate the RT-PCR products. Bottom: hybridaization with a 32P-labeled oligonucleotide probe.

tides probe showed the amplified products were of the esp-1 gene (FIG. 3, middle). The amounts of cDNA used for PCR were constant, since the expression levels of the glyceroaldehyde-3-phosphate dehydrogenese message were almost identical among the various human tissues tested (FIG. 3, bottom). We have not identified the cells containing ESP-1 in these high expression tissues yet. Therefore we could not completely rule out the possibility that the ESP-1 expressed in these tissues was derived from mononuclear cells or eosinophils. In general, normal testis and prostate, however,

are not tissues which contain large quantities of mononuclear cells and/or eosinophils. Since esp-1 was highly expressed in the male reproductive system, one can think that this ESP-1 is related to the fertilization process (15). Determination of the Size of the esp-1 Transcript The size of the esp-1 message was determined by Northern blot analysis of mRNA extracted from Hela S3 cells. We have cloned an identical gene to esp-1 from

310

Vol. 252, No. 2, 1998

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

FIG. 3. Expression of esp-1 mRNA in various human tissues. Top: PCR products of esp-1. The 100bp markers are on the left. Middle: Southern blot hybridaization of the top panel. Bottom: PCR products of glyceroaldehyde-3-phosphate dehydrogenase.

FIG. 5. Western blot analysis of the ESP-1 protein in esp-1 transfected HEK293 cells. The arrows on the left indicate the marker proteins (APPRO).

Hela S3 cDNA libraries (unpublished data). As deduced from the sizes of 28S and 18S RNA, the full length of esp-1 mRNA is approximately 1.2-1.4 kb (FIG. 4).

Transfection of the esp-1 Gene to HEK293 Cells

FIG. 4. Northern blot analysis: Determination of the size of esp-1 mRNA. The arrows on the left indicate the positions of 28S and 18S RNA. The arrow on the right indicates the esp-1 mRNA.

The esp-1 gene under the control of the promoter, SRa, was introduced to HEK293 cells (Human Embroynic Kidney cells). Mock-transfected HEK293 cells were used as a negative control. Western blot analysis of the cellular lysates showed that only the esp-1 gene transfected cells gave rise to a band corresponding to an approximate molecular mass of 35 kDa (FIG. 5). We further examined the myc-tagged esp-1 gene expressed in Cos7 cells, that gave rise to a protein band with an approximately identical molecular mass on Western blotting with anti-myc Ab (9E10) (data not shown). Although these data indicate that esp-1 is correctly transcribed and translated to a protein with a molecular mass of 35kDa, the molecular masses deduced from the amino-acid compositions of the pro-form and active-form are 33.1 and 30.6 kDa, respectively. This may be due to the glycosylation of this protease that contains two potential glycosylation sites at residues 167 and 273. However, the functional recombinant enzyme is necessary to reveal whether this 35 kDa protein is the active-form or the pro-form of ESP-1.

311

Vol. 252, No. 2, 1998

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

ACKNOWLEDGMENTS We wish to thank to Mr. Mitsuhiro Miyake (Tokushima University) for his technical help and Dr. Kazuo Maruyama (Tokyo Medical and Dental University School of Medicine) for supplying the plasmid vector, pME18S-FL.

REFERENCES 1. Pham, C. T., and Ley, T. J. (1998) Semin. Immunol. 9, 127–133. 2. Smyth, M. J., O’Connor, M. D., and Trapani, J. A. (1996) J. Leukoc. Biol. 60, 555–562. 3. Borregaard, N., and Cowland B. J. (1997) Blood 89, 3503–3521. 4. Kaliner, A. M., and Metcalfe, D. D. (1993) Lung Biology in Health and Disease 62, The Mast Cell in Health and Disease, Dekker, New York. 5. Hansel, T. T., De Vries, I. J., Iff, T., Rihs, S., Wandzilak, M., Betz, S., Blaser, K., and Walker, C. (1991) J. Immunol. Methods 145, 105–110. 6. Chomczynski, P., and Sacchi, N. (1987) Anal. Biochem. 162, 156 –159.

7. Yamashiro, K., Tsuruoka, N., Kodama, S., Tsujimoto, M., Yamamura, Y., Tanaka, T., Nakazato, H., and Yamaguchi, N. (1997) Biochim. Biophys. Acta 1350, 11–14. 8. Ravetch, V. J., and Perussia, B. (1989) J. Exp. Med. 170, 481– 497. 9. Kohno, N., Yamagata, K., Yamada, S., Kashiwabara, S., Sakai, Y., and Baba, T. (1998) Biochem. Biophys. Res. Commun. 245, 658 – 665. 10. Kozak, M. (1987) Nucleic Acids Res. 20, 8125– 8132. 11. Yu, J. X., Chao, L., and Chao, J. (1994) J. Biol. Chem. 269, 18843–18848. 12. Davis, E. M., Fujikawa, K., Kurauchi, K., and Kisiel, W. (1979) Adv. Enzymol. 48, 277–318. 13. Tomomura, A., Akiyama, M., Itoh, H., Yoshino, I., Tomomura, M., Nishii, Y., Noikura, T., and Saheki, T. (1996) FEBS Lett. 386, 26 –28. 14. Miller, J. S., Westin, E. H., and Schwartz, L. B. (1989) J. Clin. Invest. 84, 1188 –1195. 15. Yanagimachi, R. (1994) in The Physiology of Reproduction (Knobil, E., and Neill, J., Eds.), pp. 189 –317, Raven Press, NY.

312