A Recombinant Rat Regenerating Protein Is Mitogenic to Pancreatic Derived Cells

Journal of Surgical Research 89, 60 – 65 (2000) doi:10.1006/jsre.1999.5800, available online at http://www.idealibrary.com on

A Recombinant Rat Regenerating Protein Is Mitogenic to Pancreatic Derived Cells 1 Joshua L. Levine, M.D., Ketul J. Patel, M.D., Qing-hu Zheng, M.S., Alan R. Shuldiner, M.D.,* and Michael E. Zenilman, M.D., FACS Department of Surgery, Albert Einstein College of Medicine, Montefiore Medical Center, Bronx, New York 10461; and *Department of Medicine, University of Maryland, Baltimore, Maryland Submitted for publication October 18, 1999

regeneration [1, 2]. We and others have recently shown that the reg I (and reg III) proteins are mitogenic to pancreatic derived cells [3, 4]. Although reg I is expressed at high levels in normal pancreas, it has been difficult to isolate pure endogenous protein in levels sufficient for in vitro and in vivo studies. Therefore, studies directed toward determination of the physiological function of the reg protein have been limited by supply of the protein. While we have been able to isolate reg homologues from human and bovine pancreas as well as the AR42J acinar cell line, studies are needed using large volumes of a pure preparation of protein. The aim of this study was to clone the rat reg I gene into a bacterial expression vector, express a novel fusion protein with a leader sequence to facilitate isolation and identification, isolate the fusion protein, and determine its bioactivity using a mitogenic assay on pancreatic derived cell lines in culture.

Pancreatic regenerating protein (reg I) is expressed in acinar cells and is mitogenic to ␤- and ductal cells. Isolation of large amounts of endogenous reg I for in vivo and in vitro experiments has been difficult. The aim of this study was to develop a recombinant protein and determine its bioactivity on rat pancreatic derived cells. cDNA of the rat reg I coding region was created with unique BamHI flanking sequences using reverse transcriptase PCR. The coding region was then cloned into a bacterial expression vector in which expression is controlled by a T7 promoter. After transformation into the Escherichia coli strain B21(DE3) and induction by isopropyl-␤-D-thiogalactopyranoside, a fusion protein of 24 kDa in size, named reg-PET, was noted in the bacterial lysate. This protein contained a polyhistidine and S-peptide sequence to facilitate isolation and identification, respectively. This protein was purified using affinity chromatography, and identity was confirmed with gel electrophoresis and Western analysis. The reg-PET protein was mitogenic to both ARIP and RIN cells, rat pancreatic ductal and ␤-cell lines, respectively. Antibodies raised to the protein reacted against rat reg I in pancreas. The purified recombinant reg I fusion protein, like endogenous reg I, is mitogenic to pancreatic derived cells. It is more potent than reg I isolated from pancreatic tissue. This protein can be isolated rapidly, easily, and with a high amount of purity. © 2000 Academic Press

METHODS Vector preparation. A rat reg I cDNA clone was created by reverse transcriptase polymerase chain reaction (RT-PCR) of rat pancreatic mRNA, using primers which added a BamH1 restriction site to either end of the coding region. The upstream primer included the BamHI site and ATG start sequence [1] (sense: 5⬘GGATCCACAGTCTGCTGCTCATCATGACTC-3⬘), while the downstream primer (antisense: 5⬘-GGATCCTCAGATG ATTTCAGGCTTTGAAC-3⬘) was derived from the final sequence of the coding region and included the stop codon TGA and the BamHI site (sense: 5⬘-GTTCAAAGCCTGA AATCATCTGAGGATCC-3⬘). RT-PCR, performed as described in [2], yielded a 520-bp product, which was cloned into the pCRII vector using the TA-cloning technique (InVitrogen). The pCRII plasmid containing the reg I coding sequence was named reg-PCRII and transformed into NovaBlue Escherichia coli competent cells (Novagen, Inc, Milwalkee, WI), and plasmid DNA isolated using the Qiagen plasmid miniprep kit (Qiagen, Inc., Valen-

INTRODUCTION

The pancreatic regenerating gene and its protein product (reg I) are derived from the acinar cell of the pancreas. Reg mRNA is normally expressed in pancreatic acinar cells and is overexpressed during pancreatic 1 Presented in part at the Annual Meeting of the Association for Academic Surgery, Dallas, Texas, November 6 – 8, 1997.

0022-4804/00 $35.00 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.

60

LEVINE ET AL.: MITOGENIC ASSAY OF PANCREATIC REGENERATING PROTEIN cia, CA). To confirm the presence of the entire coding region, orientation, and reading frame, 400 ng of recombinant DNA plasmid was subjected to PCR sequencing. DNA templates were sequenced by the Albert Einstein College of Medicine DNA Sequencing Facility, using ABI 377 automated sequencers from The Perkin–Elmer Corporation, Applied Biosystems Division (PE/ABI, Foster City, CA). Samples were cycle sequenced with Perkin–Elmer 9600 thermocyclers using PE/ABI AmpliTaq DNA polymerase and PE/ABI dye terminator chemistry. The reaction products were washed with either spin columns (Princeton Separations) or sodium acetate/ethanol precipitation. Cleaned and dried sequencing reaction products were resuspended in 1 ␮l of loading buffer (deionized formamide, EDTA, and blue dextran) and denatured on a heat block. A total of 0.5 ␮l of denatured sample was loaded per lane using 0.2-mm 64-lane or 96-lane sharkstooth combs (The Gel Company). Sequencing reactions were run on 5% gels prepared with TBE (Tris– boric acid– EDTA) buffer using 50% stock Long Ranger (FMC BioProducts). ABI Sequencing Collection and Analysis software was used to track and extract the fluorescent data and for base calling analysis. T7 and SP6 primers for sequencing were synthesized by Life Technologies (Rockville, MD) using phosphoramidite chemistries and supplied through the Albert Einstein College of Medicine Oligonucleotide Facility (1 pM per reaction). Since the reg I coding region has no BamHI site [1], the clone of the entire coding region, flanked by BamHI, can be removed from regPCRII for subcloning using this restriction endonuclease. After restriction endonuclease digestion of the pCRII plasmid with BamHI, reg I cDNA was recovered from a 0.7% agarose gel by cutting out the 520-bp band and recovery of the cDNA by Qiaquick gel extraction kit (Qiagen, Inc.). The cDNA was then subcloned into the BamHI site of the pET-30c bacterial expression vector (Novagen, Inc.) and transfected into the NovaBlue competent cells. The orientation of reg I cDNA in the pET30c vector was then confirmed by restriction enzyme digestion with EcoRI and StyI (Boehringer Manheim) and gel electrophoresis (0.7% agarose). Bands were visualized with ethidium bromide (10 mg/ml). PCRbased DNA sequencing using the T7 promotor and T7 terminator (Novagen) was again performed to confirm the correct reading frame. This clone was named reg-PET. Protein expression. reg-PET plasmid was transformed into competent BL-21(DE3) cells, and transfectants were isolated on agar plates containing kanamycin (30 ␮g/ml). Cells were grown to log phase (OD ⫽ 0.6 – 0.7) in LB broth with kanamycin. Isopropyl-␤-Dthiogalactopyranoside (IPTG), 1 ␮g/ml, was then added to the medium to induce the T7 promoter. Cells were harvested after 2 h, and subjected to Western analysis using the S-Tag Western blot kit. In later experiments, cells were incubated for 0.5, 1,2, 2.5, 3, 4, and 5 h and in 1, 2, 5, and 10 ␮g/ml IPTG to determine the best concentration of IPTG and exposure time. For assay of expression, cells were lysed in NP-40 buffer (Sigma) and subjected to polyacrylamide gel electrophoesis (12%) and transfer to nitrocellulose. Western analysis was performed using the S-Tag Western blot kit (S-protein AP conjugate, Novagen). In this assay, S-protein is linked to alkaline phosphatase and development is performed with NBT/BCIP. For large-scale purification of recombinant fusion reg-PET protein, the His-Bind purification kit was used (Novagen). Two hundred milliliters of bacteria was lysed as per their protocol. Protein was isolated in high-salt buffer (500 mM NaCl, 20 mM Tris–HCl, pH 7.5) by affinity chromatography to a His-binding resin, a CL-6B resin with an amino-diacetate chelator, previously charged with nickel. The protein was eluted with imidazole (300 mM, pH 6.0) and dialyzed against water. We found that the best yield was obtained if the lysis buffer contained 6 M guanidine–HCl and 5 mM imidazole (pH 7.9). After binding to the nickel-charged column, elution was carried out in a buffer that contained 6 M guanidine–HCl and 200 mM imidazole (pH 7.9). After elution, the solution was dialyzed against water.

61

Mitogenesis assay. RIN 1046-38 (rat insulinoma) and ARIP (rat ductal) cells were cultured as described previously [3]. For each experiment ARIP (7.5 ⫻ 10 3 cells per well) and RIN (60 ⫻ 10 3 cells per well) were plated per well in a 96-well plate, in Dulbecco’s modified Eagle’s medium (DMEM) with 10% fetal calf serum, penicillin, and streptomycin (Gibco Life Technologies, Rockville, MD). After an overnight incubation, each well was washed three times with 1 ml of phosphate-buffered saline (Sigma, St. Louis, MO) and then inoculated with reg-PET protein at increasing concentrations in DMEM medium with 1% serum replacement medium (Sigma). Cells were incubated for 72 h, at which point mitogenesis was assayed by the MTS-tetrazoleum assay (CellTiter96, Promega, Inc., Madison, WI). Colorimetric assay was performed on a Bio-Rad microplate reader (Bio-Rad, Inc., Hercules, CA) at 490 nm. Each experiment was done in triplicate and repeated at least twice. Antibody formation. Polyclonal antibodies to the reg-PET fusion protein were raised in guinea pigs (Harlan Sprague–Dawley, Indianapolis, IN). One milligram of protein was injected subcutaneously in 1 cc complete Freund’s adjuvant. Six weeks later, the animals were boosted with another 1-mg injection in 1 cc of incomplete Freund’s adjuvant. Serum was obtained from vena caval aspiration using a 25-gauge needle as described in Ref. [5]. Western analysis was performed after confirmation of protein in the gel by Coomassie brilliant blue staining (Bio-Rad). Protein was transferred to nitrocellulose (Nylon membranes, positively charged, Boehringer Mannheim) electrophoretically at 100 V for 1 h. Membranes were washed with TBS-T (Tris-buffered saline–Tween 20, 20 mM Tris, 137 mM NaCl, 0.1% Tween, pH 7.4 ) three times. The guinea pig primary antibody was used at 1:500 dilution, and the second antibody, goat anti-guinea pig IgG linked to alkaline phosphatase, was used at 15,000 dilution (Sigma). Western blots were developed with 1-Step NBT/BCIP (Pierce, Rockford, IL).

RESULTS

Cloning RT-PCR of rat pancreatic RNA using the primers described under Methods yielded a product of 520 bp in length, which was cloned into pCRII (data not shown) and is referred to as reg-PCRII. DNA sequencing confirmed the orientation of the reg I cDNA within pCRII and defined the reading frame (Table 1). The predicted reading frame matched that available in the pET30c vector. Specifically, subcloning the coding region cDNA into this vector using the flanking BamHI restriction sites would result in a translated protein in which the reading frame would be intact. The reg I coding region from reg-PCRII was then subcloned into the pET30c using the BamHI restriction site. This plasmid was named reg-PET, and the vector was transformed into NovaBlue cells. The predicted size of the reg-PET plasmid was 5962 bp. Digestion of plasmid DNA from five colonies by EcoRI assessed the orientation for subsequent protein expression. In the correct orientation, digestion by EcoRI at its restriction sites generated a 333-bp fragment; the incorrect direction generated one of 219 bp. Similarly, after digestion with StyI, the correct orientation yielded a fragment of 609 base pairs, while the incorrect orientation would yield one of 482 bp (data not shown). Table 1 shows the nucleotide sequence of the rat reg

62

JOURNAL OF SURGICAL RESEARCH: VOL. 89, NO. 1, MARCH 2000

TABLE 1 Partial Sequence of reg I in pCRII and pET30c Plasmids reg-pCRII: reg protein reg-PET: reg-PET: protein

BamHI Reg start 5⬘-TTCGGCTTCCGGATCCACAGTCTGCTGCTCATCATGACTCGCAACAAATATTTCATTCTG-3⬘ M T R N K Y F I L3 5⬘-GGATATCTGTGGATCCACAGTCTGCTGCTCATCATGACTCGCAACAAATATTTCATTCTG-3⬘ 4G Y L W I H D L L L I M T R N K Y F I L3

Note. The sequence of the reg I coding region in pCRII (reg-pCRII) is shown at the BamHI restriction site just prior to the ATG codon. The known reg I protein sequence, starting at the ATG, is shown underneath. The reg cDNA was then cloned into the pET30c vector using BamHI. The predicted amino acid sequence of reg-PET protein in pET30c confirms that the BamHI site is in frame with the subsequent ATG of the endogenous reg protein.

I coding sequence within the pET30c plasmid (regPET), as determined by PCR-based sequencing using T7 polymerase. The correct orientation is confirmed, and the correct reading frame was conserved. Also, the sequence was identical to the reg I sequence originally cloned into pCRII (reg-PCR). Expression The reg-PET fusion protein contains a 41 amino acid N-terminal leader sequence. This includes a 6 amino acid poly-histidine tag to facilitate isolation [6], and a 15 amino acid S-peptide (S-tag) domain. The S-peptide fragment can bind to S-protein to form an active ribonuclease S. The latter was then used for Western analysis (using a synthetic S-protein linked to alkaline phosphatase) (see below) [7]. The fusion protein is pre-

dicted to be 221 amino acids in length, with a predicted size of 25 kDa. After isolation of the reg-PET vector plasmid DNA and transformation into the BL-21(DE3) cell line, protein expression experiments were carried out. Time course experiments using different concentrations of IPTG revealed that the best induction of recombinant protein was obtained with 5 ␮M IPTG at 2.5 h. Western analysis of bacterial lysate showed that a new protein, approximately 24 kDa in size, was present in the lysate (Fig. 1a). A positive control included the pet30c plasmid without the reg insert. This should generate a 5-kDa protein consisting of the pET30 leader sequence plus a terminal sequence (not included in our protein due to the reg stop codon intrinsic to the coding region).

FIG. 1. (a) Expression of reg I fusion protein. After transformation of reg-PET plasmid into BL21(DE3) and growth into log phase, 5 mM of IPTG was added to the broth for 2.5 h to induce expression of recombinant protein. Western analysis of cellular lysate using S-protein labeled with alkaline phosphatase shows that a recombinant protein is formed and that it is ⬃24 kDa in size. The -insert control (lane 2) is the pET-30c vector without the reg I insert. This has the poly-his tail, S-tag, but no reg I protein, and the predicted MW is 5–7 kDa. The -IPTG lane is a negative control—the reg-PET transformed bacteria without IPTG—recombinant protein is not expressed. A positive recombinant control, ␤-galactosidase, is shown in the lane labeled Cntl. (b) Isolation of recombinant protein. Using a charged nickel column, reg-pET protein was purified from the induced bacterial lysate. A single band at 24 kDa is seen from two separate isolates on polyacrylamide gel electrophoresis.

LEVINE ET AL.: MITOGENIC ASSAY OF PANCREATIC REGENERATING PROTEIN

63

Bioactivity of Bacterial Lysate Increasing concentrations of bacterial cell lysate were added to a culture of 2.5 ⫻ 10 4 ARIP cells, a rat ductal cell line. This cell line will divide when exposed to both human and rat reg I [3, 8]. After incubation for 24 h, mitogenesis was determined by using an MTStetrazolium assay. Cells were grown in 96-well plates in DMEM and 1% serum replacement medium (Sigma). All experiments were performed in triplicate. Uninduced lysate (⫺IPTG) added to the culture medium had no mitogenic activity. At a concentration of 10% induced lysate (⫹IPTG) the growth of the ARIP cells in culture increased significantly from an absorbance of 0.53 ⫾ .04 OD in controls to 0.70 ⫾ .02 OD (P ⬍ 0.05 compared to control, ANOVA). This suggested that the protein in the lysate was, in fact, reg-PET (not shown). Isolation and Bioactivity of Purified Recombinant Protein A charged nickel affinity column was then used to isolate pure reg-pET protein from the induced bacterial cell lysate. Identity was again confirmed with gel electrophoresis. A single band is seen at 24 kDa (Fig. 1b), with no evidence of contamination from the parental bacterial protein. This protein reacted with anti-rat reg I antibodies generated in Ref. [9] (data not shown). Mitogenesis was then assayed again on ARIP cells and on the rat ␤-cell line (RIN) using 7.5 ⫻ 10 3 cells/well and 60 ⫻ 10 3 cells/well, respectively. Cells were cultured on 96-well plates for 72 h. Cell growth was again measured with an MTS-tetrazolium assay. When exposed to the purified reg-pET fusion protein, the ARIP cells showed a dose-related increase in cell growth, which reached statistical significance at 100 pM of reg-pET (P ⬍ 0.05 compared to control, ANOVA) (Fig. 2a). At 10 nM concentration of reg-PET, an increase in MTS absorbance was noted from 100 ⫾ 3% to 117 ⫾ 3% (P ⬍ 0.05). Increase of RIN cell growth was significant at 10 pM of reg-pET (P ⬍ 0.05 compared to control, ANOVA) (Fig. 2b). After exposure to 10 nM of reg-PET, an increase in MTS absorbance was noted from 100 ⫾ 4% to 125 ⫾ 3% (P ⬍ 0.05). This response is about 10⫻ more potent when compared to the effect observed on ARIP and RIN cultures with reg I isolated from human and bovine pancreas [3, 8]. From a typical culture of 200 ml of induced bacteria, we were able to obtain 1–2 mg of purified protein. Confirmation of the Recombinant Fusion Protein as Reg I Antibodies to the reg-pET fusion protein were raised in guinea pigs. Western analysis of this antibody showed reaction with reg-pET protein as well as with a rat pancreatic protein at 15–16 kDa (Fig. 3). This is the size of endogenous reg I.

FIG. 2. (a) Bioactivity of the recombinant protein. The reg-pET fusion protein was mitogenic to ARIP cells (7.5 ⫻ 10 3 cells/well). Statistical significance was achieved at 100 pM and 100 nM. (b) Bioactivity of the recombinant protein (II). The reg-pET fusion protein was mitogenic to RIN ␤-cells (60 ⫻ 10 3 cells/well). Statistical significance was achieved at 10 pM.

DISCUSSION

The reg family of proteins has recently been classified by Unno et al. [10]. The family is made up of a number of proteins with a molecular weight of 15–19 kDa, with amino acid sequences similar to that of calcium-dependent lectins. Reg I, or pancreatic regenerating protein, is constituitively expressed in acinar cells and induced in islets during regeneration. It has also been isolated and identified as pancreatic stone protein [11, 12] and pancreatic thread protein in humans [13]. Reg II has only been recently described [10] and may be a marker of risk for the development of diabetes [14]. While reg III is normally not expressed in the acinar pancreas, it is expressed in the small intestine. The gene is induced in the pancreas after the induction of pancreatitis and has been therefore called pancreatitis-associated protein [15]. A homologue of reg III is found in bovine pancreatic extracts, and it is identical to pancreatic thread protein [16].

64

JOURNAL OF SURGICAL RESEARCH: VOL. 89, NO. 1, MARCH 2000

FIG. 3. Confirmation that the reg-pET protein is reg I. Antibodies were raised to reg-pET fusion protein in guinea pig. Western analysis of this antibody showed reaction with native reg-pET and with a protein in three different rat pancreas lysates (a, b, c) at 15 kDa. This is the size of native reg I (or PSP).

We have recently shown that the proteins human reg I (PTP, isolated from pancreas), rat reg I (isolated from the cell line AR42J), and bovine reg III (PTP, isolated from pancreas) are all mitogenic to pancreatic derived cell lines ARIP and RIN [3, 8]. Others have identified rat reg III as a potent Schwann cell mitogen [17]. We believe that reg proteins (I and III) are acinar derived proteins which are important in islet regeneration, function, and maintainence. In order to perform adequate in vivo and in vitro studies, a pure protein preparation is needed. Our initial protein experiments were performed with human reg I (hPTP/reg) [3], but this yielded quantities sufficient for only in vitro studies. While we have found that bioactive rat reg I can be isolated from AR42J in sufficient quantity for in vitro experiments, using similar techniques to the human isolation [8] this too is not efficient. For large-scale purification of rat reg I protein, we could use preparative HPLC but this is costly. In this study, we successfully cloned the coding region of rat reg I into the versatile plasmid pCRII, after adding a BamHI cleavage site to each end of the sequence. Since reg does not have a BamHI sequence, the entire coding region can then be easily subcloned to different vectors, depending on the experiment desired. We originally planned to clone the reg I coding region cloned into pCRII to subclone into a eukaryotic expression vector. By taking advantage of the flanking BamHI sites, we successfully subcloned it into a pAcMP2 vector (Pharmingen, San Diego) for transfection into the bacculovirus Autographi californica and confirmed its orientation by restriction endonuclease digest (unpublished data). Gentle lysis of infected Sf9 cells and subsequent gel electrophoresis revealed a new protein produced by the bacculovirus, which was not present in the wild-type control. The protein was

not secreted. While the lysate was mitogenic to ARIP cells, indicating that a bioactive protein was produced, the protein was not produced in a high enough concentration for our use. Others have successfully cloned rat reg I into a bacculovirus vector using a later viral promoter [4, 18] to enhance secretion. While they found it to be secreted at reasonable levels, polyacrylamide gel electrophoresis of the samples reveals a fair amount of foreign protein (likely viral) contamination [18, 19]. Therefore, we elected to clone the cDNA of the entire reg I coding region into a bacterial expression vector. This would facilitate rapid, large-scale isolation. In order to enhance expression, the coding region was linked to a T7 promotor induced by IPTG. In order to enhance purity of isolation, we created a fusion protein which contained a poly-histidine moiety. In order to facilitate identification and subsequent quantitation, we planned for a fusion protein which contained an S-peptide sequence, allowing rapid Western analysis. Recombinant fusion protein reg-PET, which contains the entire coding region of reg I, was then successfully expressed after induction with IPTG. As predicted, the cell lysate of the bacteria containing induced reg-PET protein was mitogenic to ARIP cells in culture. The protein was then further purified by affinity chromatography using an agarose– iminodiacetic acid column charged with nickel, which binds the poly-histidine domain of the recombinant protein. Although this technique has been advocated for isolation of protein under nondenaturing conditions, we obtained the best yield using 6 M guanidine. Gel electrophoresis revealed pure protein with minimal contamination. Dialysis against water yielded bioactive protein, and the purified reg I protein was found to be mitogenic to both ARIP and RIN cells in culture. The bioactivity was approximately 100-fold more than the protein isolated from pancreas—we believe this is due to contaminants present in the pancreatic isolates. We conclude that the recombinant reg-PET fusion protein, like endogenous pancreatic reg I, is bioactive. Interestingly, although the human (and presumably rat) reg I protein is heavily glycosylated [20], our data suggest that these eukaryotic posttranslational modifications, which do not take place on prokaryotes, probably do not confer reg bioactivity. This is supported by recent data from our laboratory, which strongly suggest that a specific peptide sequence of the molecule, away from the N-terminus, may confer the mitogenic bioactivity (manuscript in preparation). This novel fusion protein is potentially very versatile, and this is important with a protein with only mild mitogenic activity and which is difficult to measure with antibodies [3, 21–23]. The S-peptide sequence can be used for protein identification by Western analysis, specifically to study the protein in vivo, where endog-

LEVINE ET AL.: MITOGENIC ASSAY OF PANCREATIC REGENERATING PROTEIN

enous reg is present. Also, quantitation is possible by active ribonuclease assay with S-protein [7]. The regPET recombinant fusion protein has 221 amino acids. The 43 amino acid non-reg I N-terminus protein can be cleaved off by treatment with recombinant enterokinase. The resulting reg I protein will contain an extra 13 amino acids attached to the N-terminus of the native reg I sequence. This will yield a pure rat reg protein, with a 13 amino acid addition at the N-terminus. The fact that our protein was bioactive, however, has so far negated the need for this step.

11.

de Caro, A., Lohse, J., and Sarles, H. Characterization of a protein isolated from pancreatic calculi of men suffering from chronic calcifying pancreatitis. Biochem. Biophys. Res. Commun. 87: 1176, 1979.

12.

Giorgi, D., Bernard, J. P., Rouquier, S., Iovanna, J., Sarles, H., and Dagorn, J. C. Secretory pancreatic stone protein messenger RNA. J. Clin. Invest. 84: 100, 1989.

13.

Gross, J., Carlson, R. I., Brauer, A. W., Margolies, M. N., Warshaw, A. L., and Wands, J. R. Isolation, characterization, and distribution of an unusual pancreatic human secretory protein. J. Clin. Invest. 76: 2115, 1985.

14.

Baeza, N., Sanchez, D., Vialettes, B., and Figarella, C. Specific reg II gene overexpression in the non-obese diabetic mouse pancreas during active diabetogenesis. FEBS Lett. 416: 364, 1997.

15.

Iovanna, J., Orelle, B., Keim, V., and Dagorn, J. C. Messenger RNA sequence and expression of rat pancreatitis-associated protein, a lectin-related protein overexpressed during acute experimental pancreatitis. J. Biol. Chem. 266: 24664, 1991.

16.

Cai, L., Harris, W. R., Marshak, D. R., Gross, J., and Crabb, J. W. Structural analysis of bovine pancreatic thread protein. J. Prot. Chem. 9: 623, 1990.

17.

Livesey, F. J., O’Brien, J. A., Li, M., Smith, A. G., Murphy, L. J., and Hunt, S. P. A Schwann cell mitogen accompanying regeneration of motor neurons. Nature 390: 614, 1997.

18.

Bimmler, D., Frick, T. W., and Scheele, G. A. High level secretion of native rat pancreatic lithostatine in a bacculovirus expression system. Pancreas 11: 63, 1995.

19.

Gross, D. J., Weiss, L., Reibstein, I., van den Brand, J., Okamoto, H., Clark, A., and Slavin, S. Amelioration of diabetes in nonobese diabetic mice with advanced disease by linomideinduced immunoregulation combined with reg protein treatment. Endocrinology 139: 2369, 1998.

20.

De Reggi, M., Capon, C., Gharib, B., Wieruszeski, J. M., Michel, R., and Fournet. The glycan moiety of human pancreatic lithostatine. Structure characterization and possible pathophysiological implications. Eur. J. Biochem. 230: 503, 1995.

21.

Schmiegel, W., Burchert, M., Kalthoff, H., Roeder, C., Butzow, G., Grimm, H., Kremer, B., Soehendra, N., Schreiber, H. W., Thiele, H. G., and Greten. Immuochemical characterization and quantitative distribution of pancreatic stone protein in sera and pancreatic secretions in pancreatic disorders. Gastroenterology 99: 1421, 1990.

22.

Provansal-Cheylan, M., Mariani, A., Bernard, J. P., Sarles, H., and Dupuy, P. Pancreatic stone protein: Quantification in pancreatic juice by enzyme-linked immunosorbent assay and comparison with other methods. Pancreas 4: 680, 1989.

23.

Satomora, Y., Sawabu, N., Mouri, I., Yamakawa, O., Watanabe, H., Motoo, Y., Okai, T., Ito, T., Kaneda, K., and Okamoto, H. Measurement of serum PSP/reg protein concentration in various diseases with a newly developed enzyme-linked immunosorbent assay. J. Gastroenterol. 30: 643, 1995.

ACKNOWLEDGMENT This study was supported by NIH RO1 DK 54511-01 and Core Grant P60 DK20541. The DNA Sequencing Facility at Albert Einstein College of Medicine is supported by NCI Cancer Research Center Core Support Grant CA13330.

REFERENCES 1.

2.

3.

4.

5. 6.

7. 8. 9. 10.

Terazono, K., Yamamoto, H., Takasawa, S., Shiga, K., Yonekura, Y., Tochino, Y., and Okamoto, H. A novel gene activated in regenerating islets. J. Biol. Chem. 263: 2111, 1988. Zenilman, M. E., Perfetti, R., Swinson, K., Magnuson, T. H., and Shuldiner, A. R. Pancreatic regeneration (reg) gene expression in a rat model of islet hyperplasia. Surgery 119: 576, 1996. Zenilman, M. E., Magnuson, T. H., Swinson, K., Egan, J., Perfetti, R., and Shuldiner, A. R. Pancreatic thread protein is mitogenic to pancreatic-derived cells in culture. Gastroenterology 110: 1208, 1996. Watanabe, T., Yonemura, Y., Yonekura, H., Suzuki, Y., Miyashita, H., Sugiyama, K., Moriizumi, S., Unno, M., Tanaka, O., Kondo, H., Bone, A. J., Takasawa, S., and Okamoto, H. Pancreatic beta-cell replication and amelioration of surgical diabetes by Reg protein. Proc. Natl. Acad. Sci. USA 91: 3589, 1994. Reuter, R. E. Venapuncture in the guinea pig. Lab. Anim. Sci. 37: 245, 1987. Van Dyke, M. W., Sirito, M., and Sawadogo, M. Single-step purification of bacterially expressed polypeptides containing an oligo-histidine domain. Gene 111(1): 99, 1992. Kim, J. S., and Raines, R. T. Ribonuclease S-peptide as a carrier in fusion proteins. Protein Sci. 2(3): 348, 1993. Zenilman, M. E., Chen, J., and Magnuson, T. H. Effect of reg protein on pancreatic ductal cells. Pancreas 17: 256, 1998. Zenilman, M. E., Chen, J., and Zheng, Q. H. Characteristics of rat reg protein. Surgery 124: 855, 1998. Unno, M., Yonekura, H., Nakagawara, K., Watanabe, T., Miyashita, H., Moriizumi, S., and Okamoto, H. Structure, chromosomal localization and expression of mouse reg genes reg I and reg II. J. Biol. Chem. 268: 15974, 1993.

65