Identification of a Receptor for Reg (Regenerating Gene) Protein, a Pancreatic b -Cell Regeneration Factor*

Reg (regenerating gene) was isolated as a gene specifically expressed in regenerating islets (Terazono, K., Yamamoto, H., Takasawa, S., Shiga, K., Yonemura, Y., Tochino, Y., and Okamoto, H. (1988) J. Biol. Chem. 263, 2111–2114). Rat and human Reg gene products, Reg/REG proteins, have been demonstrated to stimulate islet b -cell growth in vitro and in vivo and to ameliorate experimental diabetes. In the present study, we isolated a cDNA for the Reg protein receptor from a rat islet cDNA library. The cDNA encoded a cell surface 919-amino acid protein, and the cells into which the cDNA had been introduced bound Reg protein with high affinity. When the cDNA was introduced into RINm5F cells, a pancreatic b -cell line that shows Reg-dependent growth, the transformants exhibited significant increases in the incorporation of 5 * -bromo-2 * -deoxyuridine as well as in the cell numbers in response to Reg protein. A homology search revealed that the cDNA is a homologue to a human multiple exostoses-like gene, the function of which has hitherto been unknown. These results strongly suggest that using an apoptosis screening kit (Wako Pure Chemical, Osaka, Japan). RNase Protection Assay regenerating islets were prepared as described rat cell as described The Pst Bgl II fragment of rat Reg receptor cDNA residues 755–1,064) was subcloned into the Pst Bam HI site of pBluescript SK linearized Hin dIII, and transcribed by T3 RNA polymerase 32 P]CTP. The resultant 0.45-kilobase complementary RNA used a probe. RNase protection assay performed using a Ribonuclease Protection Assay III kit (Ambion) according manufacturer’s recommendation.

Only pancreatic islets of Langerhans can produce insulin, but they have limited capacity for regeneration, which predisposes them to the development of diabetes mellitus (1). In 1984, we found that the administration of poly(ADP-ribose) synthetase inhibitors, such as nicotinamide (2)(3)(4)(5), to 90% depancreatized rats induced islet regeneration (6). By screening a re-generating islet-derived cDNA library we isolated a gene, Reg (regenerating gene), specifically expressed in regenerating islets (7). We have also isolated the human REG cDNA and gene (7,8). The Reg gene, which encodes a 165-166-amino acid protein (7), was activated during experimental regenerative processes of pancreatic islets, suggesting possible roles for the Reg gene in the replication, growth, and maturation of islet ␤-cells (7, 9 -13). In fact, rat Reg and human REG proteins have been demonstrated to induce the proliferation of ␤-cells to ameliorate the diabetes of 90% depancreatized rats and of non-obese diabetic (NOD) 1 mice (14,15). Thus, although Reg protein has been thought to act on pancreatic ␤-cells as an autocrine and/or paracrine growth factor (16 -18), the receptor for Reg protein has been elusive. In the present study, we isolated a Reg receptor cDNA, the product of which mediated a growth signal of Reg protein for ␤-cell regeneration.

EXPERIMENTAL PROCEDURES
Isolation of Reg-binding Protein cDNA-Rat and human Reg cDNAs (7) were inserted into pPIC3.5 (Invitrogen) and introduced into Pichia pastoris. The recombinant rat and human Reg proteins were purified from the culture media of Pichia by cation exchange chromatography as described (19). Purified rat Reg protein was labeled with 125 I-labeled Bolton-Hunter reagent (NEN Life Science Products). A ZAP II rat islet cDNA expression library (20,21) was screened by binding to 125 Ilabeled Reg protein. In brief, about 1.2 ϫ 10 6 independent phages (approximately 5 ϫ 10 4 phages on a 150-mm agar) were plated. After incubation for 5 h at 37°C, the plates were overlaid with nitrocellulose filters that had been immersed with 10 mM isopropyl-␤-D-thiogalactopyranoside. Incubation was continued for 4 h at 37°C. The filters were then removed, washed with PBST (phosphate-buffered saline with 0.05% Tween 20 (v/v)) at room temperature, and blocked in PBST with 5% nonfat dry milk (w/v) overnight at 4°C. Following the blocking, filters were incubated for another 2 h at room temperature in the presence of 125 I-labeled rat Reg protein (1 ϫ 10 6 cpm/ml). After three washes with PBST, the filters were exposed to x-ray film. The positive phages were isolated, and cDNA inserts were sequenced in both directions after in vivo excision.
Isolation of Full-length Rat Reg Receptor cDNA-The rat islet cDNA library (20, 21) (5 ϫ 10 6 clones) was screened by plaque hybridization with the cDNA fragment that had been isolated by the initial screening of Reg-binding protein, and 8 positive clones were obtained. The 8 clones largely overlapped each other and had complete nucleotide identity in the overlapping regions.
Immunoblot Analysis-The rat Reg receptor cDNA ligated with oligonucleotide encoding the hemagglutinin (HA) nonapeptide tag (YPY-DVPDYA) at the N terminus was inserted into a pCI-neo mammalian expression vector (Promega) and expressed in COS-7 cells as described (22,23). After a 48-h incubation, cells were collected, homogenized, and fractionated as described (22,23). The protein sample was electrophoresed on a 12.5% SDS-polyacrylamide gel (w/v) and transferred to Immobilon-P (Millipore). Western blot analysis was carried out as described (23,24) using a monoclonal antibody against HA (3F10, Roche Molecular Biochemicals).
Isolation of Stable Transformants Expressing Rat Reg Receptor-The rat Reg receptor expression vector with HA tag was introduced into CHO cells and RINm5F cells. Cells were cultured in RPMI 1640 with 10% fetal calf serum (BioWhittaker, Walersville, Maryland) and 250 g/ml neomycin (Life Technologies, Inc.) for 2 weeks (25). Stable transformants expressing high levels of the recombinant protein were * This work was supported in part by grants-in-aid from the Ministry of Education, Science, Sports and Culture, Japan and by the Research Fund for Digestive Molecular Biology. 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.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) AB033367.
§ This work was performed in partial fulfillment of the requirements for a doctoral degree by Tohoku University.
¶ Recipient of a fellowship from the Japan Society for the Promotion of Science.
Reg Binding Assay-The Reg receptor expression vector and the control vector were electroporated into COS-7 cells and expressed transiently (22,23). The CHO cells stably expressing the Reg receptor were isolated as described above. The cells (7.5 ϫ 10 5 cells) were washed with RPMI 1640 and incubated on ice in the presence of 125 I-labeled rat or human Reg protein in RPMI 1640 containing 1% fetal calf serum for 2 h. After washing with RPMI 1640 three times, the cells were lysed by 1 ml of 100 mM Tris-HCl (pH 7.6), 1 mM EDTA, and 1% Triton X-100. The radioactivity of the lysate was determined by a ␥-counter (Cobra, Packard).
5Ј-Bromo-2Ј-deoxyuridine (BrdUrd) Incorporation-Stable transformants expressing the Reg receptor were cultured (5 ϫ 10 4 cells/well) in RPMI 1640 with 1% fetal calf serum in the presence of increasing concentrations of rat Reg protein for 24 h. During the last 2 h, BrdUrd (10 M) was added in the culture medium, and BrdUrd incorporation was measured using a colorimetric cell proliferation enzyme-linked immunosorbent assay kit (Roche Molecular Biochemicals).
Measurement of Apoptosis-After a 24-h incubation of the stable transformants expressing the Reg receptor in RPMI 1640 with 1% fetal calf serum in the presence of increasing concentrations of rat Reg protein, apoptosis was detected by the TUNEL method (26) using an apoptosis screening kit (Wako Pure Chemical, Osaka, Japan).
RNase Protection Assay-Rat regenerating islets were prepared as described (6,7,9). RNAs were isolated from various rat tissues and cell lines as described (20,23,27). The PstI/BglII fragment of rat Reg receptor cDNA (nucleotide residues 755-1,064) was subcloned into the PstI/BamHI site of pBluescript SK (-), linearized with HindIII, and transcribed in vitro by T3 RNA polymerase using [␣-32 P]CTP. The resultant 0.45-kilobase complementary RNA was used as a probe. RNase protection assay was performed using a Ribonuclease Protection Assay III kit (Ambion) according to the manufacturer's recommendation.

RESULTS AND DISCUSSION
To isolate the Reg receptor, we constructed a ZAP II rat islet cDNA expression library and screened the library by the binding to 125 I-labeled rat Reg protein. A cDNA clone with a 1.6-kbp cDNA insert was obtained. Using the cDNA as a probe, we isolated the full-length cDNA of Reg-binding protein from the rat islet cDNA library. The cDNA had a 2,760-base pair open reading frame encoding a 919-amino acid protein (Fig. 1), and the amino acid sequence of the cDNA suggested that the protein is a type II transmembrane protein with a long extracellular domain (868 amino acid residues), a transmembrane domain (residues 29 -51), and a short intracellular region at the N terminus. We then constructed an expression vector for the rat cDNA and expressed it transiently in COS-7 cells. Immunoblot analysis revealed that the cDNA-encoded protein was expressed predominantly in the cell membrane fraction with an apparent molecular weight of 105,000 ( Fig. 2A), showing good agreement with its calculated molecular weight (104,682) from the deduced amino acid sequence. The COS-7 cells into which the expression vector had been introduced bound 125 I-labeled rat Reg protein, and the binding was attenuated by the addition of unlabeled Reg protein (Fig. 2B). A homology search against DNA/protein data bases revealed that the cDNA and its deduced amino acid sequences show significant homologies to those of multiple exostoses (EXT) family genes (28 -33), especially to human EXT-like gene 3 (EXTL3) (28)/EXT-related gene 1 (EXTR1) (29) (over 97% amino acid identity), indicating that the cDNA encodes a rat homologue to human EXTL3/EXTR1. The EXTL3/EXTR1 gene was isolated as a member of the EXT family genes by homology screening, but its physiological function and pathological significance have not yet been clarified. EXTL3/EXTR1 is thought to belong to the EXT family (28,29), because it shows homology to EXT2 and EXT1 at their C-terminal regions (52% in C-terminal 262 amino acids with EXT2 and 40% in C-terminal 247 amino acids with EXT1) (see Fig. 1). However, the N-terminal region (residues 1-656) of EXTL3/EXTR1 has no homology to any other members of the EXT family genes. Furthermore, the N-terminal region of EXTL3/EXTR1 contained the membrane-spanning domain, but the other members of the family lacked this domain and therefore were not thought to be cell surface proteins. In addition, the 1.6-kbp cDNA, which was initially isolated in the screening of the rat islet cDNA expression library as a Reg-binding protein, contained the N-terminal region alone (amino acid residues 1-332). Therefore, it is reasonable to assume that this region contains the Reg binding domain and that the EXT family members other than EXTL3/EXTR1 have no ability to bind to Reg protein. We introduced the expression vector into CHO cells, established cell lines overexpressing the Reg-binding protein (rat homologue to human EXTL3/EXTR1), and found that the CHO cells bound rat Reg protein with high affinity (K d ϭ 4.41 nM). As shown in Fig. 2C, the binding of 125 I-labeled rat Reg protein was displaced by increasing the concentrations of unlabeled rat Reg protein (K i ϭ 1.61 nM). The Hill coefficient (n H ) was estimated to be 1.18, indicative of interactions with a single, homogenous population of binding sites. Human REG protein (7), which shows a 70% FIG. 1. Amino acid sequence of rat Reg receptor. Alignment of the predicted protein sequences of rat Reg receptor (rEXTL3), human EXTL3/EXTR1 (hEXTL3) (28,29), human EXT2 (hEXT2) (31), human EXT1 (hEXT1) (30), human EXTL1 (hEXTL1) (32), and human EXTL2 (hEXTL2) (33). The transmembrane domain is underlined. Numbers in the right column correspond to amino acid residues. Residues identical to rat Reg receptor are indicated by dots. Hyphens denote the absence of corresponding residues in rat Reg receptor.
amino acid identity to rat Reg protein, also bound to the CHO cells (K d ϭ 14.0 nM), but higher concentrations of human REG protein were required for the displacement of the rat Reg protein binding to the CHO cells (K i ϭ 7.41 nM). These results suggest that EXTL3/EXTR1 is a cell surface Reg receptor that binds to Reg protein.
As shown in Fig. 3A, a single protected band of 309 bases was observed by RNase protection assay, and this was consistent with the size of the protected band derived from rat Reg receptor mRNA. The Reg receptor mRNA was expressed in normal pancreatic islets, regenerating islets, and RINm5F ␤-cells, a rat insulinoma-derived ␤-cell line. Reg protein increased the BrdUrd incorporation (1.5-2-fold, data not shown) and the number (Fig. 4A) of RINm5F cells in response to Reg protein when the cells were seeded at 20 -40 ϫ 10 4 cells/well. The Reg protein concentrations exhibiting the growth-stimulating effect on RINm5F cells were consistent with those for primary cul-tured rat islets (14), suggesting that islets and RINm5F cells use the same receptor for Reg protein. We therefore introduced the expression vector into RINm5F cells and established sev- Regenerating islets were isolated from 90% depancreatized rats receiving intraperitoneal administration of 0.5 mg/kg/day of nicotinamide for 1-3 months (6,7,9). Lane 1, normal pancreatic islets; lane 2, regenerating islets 1 month after the partial pancreatectomy; lane 3, regenerating islets 2 months after the partial pancreatectomy; lane 4, regenerating islets 3 months after the partial pancreatectomy; eral cell lines overexpressing the Reg receptor. The control RINm5F cells showed no increase of WST-1 cleavage in response to Reg protein when the cells were seeded at 5 ϫ 10 4 cells/well (Fig. 4A). In contrast to the control cells, the receptorexpressing cell lines (lines 1, 6, and 24) showed significant increases in BrdUrd incorporation when incubated with 0.3-300 nM rat Reg protein (Fig. 4B). Moreover, the numbers of RINm5F cells were increased in response to Reg protein (0.3-100 nM) as judged by the cleavage of WST-1 (Fig. 4C). With high concentrations of Reg protein (300 -1,000 nM), the cell numbers were decreased, and the decrease in the viable cells (Fig. 4C) and the increase in TUNEL-positive cells (Fig. 4D) were correlated, suggesting that the Reg-Reg receptor interaction may regulate the proliferation and apoptosis of pancreatic ␤-cells for maintaining the ␤-cell mass.
As shown in Fig. 3A, the expression of the Reg receptor was not increased in regenerating islets as compared with that in normal islets, suggesting that the regeneration and proliferation of pancreatic ␤-cells in the increase of the ␤-cell mass are primarily regulated by the expression of Reg protein. This idea is consistent with the observations that Reg gene was first identified as a gene specifically expressed in regenerating islets (7,9) and that Reg gene expression was also observed in the phase of transient ␤-cell proliferation such as in pancreatic islets of BB/Wor/Tky rats during the remission phase of diabetes (11), islets of NOD mice during active diabetogenesis (12), and pancreatic ductal cells, which are thought to be progenitor cells of ␤-cells, during differentiation and proliferation in a mouse model of autoimmune diabetes (13). ARIP cells, a pancreatic ductal cell line, which express the Reg receptor (see Fig.  3A, lane 6), were also reported to proliferate in a Reg proteindependent manner (34,35).
The expression of Reg receptor mRNA was also detected in liver, kidney, stomach, small intestine, colon, adrenal gland, pituitary gland, and brain, but not in heart (Fig. 3B), suggesting the possible involvement of the Reg-Reg receptor signal system in a variety of cell types other than pancreatic ␤-cells. Recently, several Reg and Reg-related genes have been isolated and revealed to constitute a multigene family, the Reg family (8,17,18,36,37). Based on the primary structures of the Reg proteins, the members of the family are grouped into three subclasses, type I, II, and III (17,18,36,37). Type I Reg proteins, which contain the rat and human Reg proteins described in the present paper, are expressed in regenerating islets (7, 9 -13, 16 -18). Type I Reg expression under pathological conditions has recently been reported in human colorectal carcinomas (8,38,39) and in rat gastric mucosa (40) and enterochromaffin-like cells (41), and type III Reg proteins have also been suggested to be involved in cellular proliferation in intestinal Paneth's granular cells (42), hepatocellular carcinomas (42), pancreatic acinar cells (42,43), and Schwann cells (44). Therefore, whether the Reg receptor identified in the present study functions in various tissues and cells in physiological and pathological conditions as a receptor for the Reg family gene products remains to be clarified.