Association of Tyrosine Phosphatases SHP-1 and SHP-2, Inositol 5-Phosphatase SHIP with gp49B1, and Chromosomal Assignment of the Gene*

We have analyzed the molecules participating in the inhibitory function of gp49B1, a murine type I transmembrane glycoprotein expressed on mast cells and natural killer cells, as well as the chromosomal location of its gene. As assessed by SDS-polyacrylamide gel electrophoresis and immunoblot analysis, tyrosine-phosphorylated, but not nonphosphorylated, synthetic peptides matching each of the two immunoreceptor tyrosine-based inhibitory motif (ITIM)-like sequences found in the cytoplasmic portion of gp49B1 associated with the ∼65-kDa tyrosine phosphatase SHP-1 and ∼70-kDa SHP-2 derived from RBL-2H3 cells. In addition, the phosphotyrosyl peptide matching the second ITIM-like sequence also bound the ∼145-kDa inositol polyphosphate 5-phosphatase SHIP. Thus, it has been strongly suggested that the inhibitory nature of gp49B involves the recruitment of SHP-1, SHP-2, and SHIP for the delivery of inhibitory signal to the cell interior upon phosphorylation of tyrosine residues in their ITIMs. The gp49B gene has been found to be in the juxtaposition of its cognate gene, gp49A. The gene pair was shown to locate in the B4 band of mouse chromosome 10. In this region, no conserved linkage homology to human chromosome 19, where the genes for killer cell inhibitory receptors are found, has been identified.

The immunoreceptor tyrosine-based inhibitory motif (ITIM) 1 is a set of amino acid sequences that deliver an inhibitory signal upon phosphorylation of the specific tyrosine residues. The motif is found in cytoplasmic portions of Fc␥RIIB (1-3), CD22 (4,5), and inhibitory receptors expressed by human and mouse natural killer (NK) cells (6). The consensus sequence of ITIM is (V/I)-X-Y-X-X-(L/V) (7). In B cells, when F(abЈ) 2 fragments of anti-membrane immunoglobulin antibody induce cross-linkage of the surface B cell antigen receptors, a signal transduction cascade is elicited through the receptor that results in B cell proliferation and differentiation into antibodyproducing cells. In contrast, stimulation with intact anti-membrane Ig antibody results in co-cross-linkage between the B cell antigen receptor and Fc␥RIIB, in attenuated B cell signal transduction due to the phosphorylation of ITIM of Fc␥RIIB, and binding of src homology 2 (SH2)-containing proteins, which inhibit calcium signaling pathway (2,3). The importance of the inhibitory nature of Fc␥RIIB in the humoral immune response in vivo was demonstrated in the Fc␥RIIB-deficient mice generated by gene targeting (8). The proteins to be associated with the phosphorylated ITIM sequence of Fc␥RIIB was verified to be an SH2-containing tyrosine phosphatase, SHP-1 (9), and an SH2-containing inositol polyphosphate 5-phosphatase, SHIP, in B cells, whereas in mast cells SHIP was shown to be preferentially recruited (10).
On the other hand, human and mouse NK cells express two kinds of inhibitory molecules containing ITIMs. One of which, human killer cell inhibitory receptor (KIR), is a type I transmembrane glycoprotein belonging to Ig superfamily (6,11,12). In contrast, human CD94 and mouse Ly-49 molecules are type II transmembrane glycoprotein with C-type lectin structure (13,14). Despite these quite different structures, both Ly-49 and KIR recognize allelic groups of the major histocompatibility complex class I molecules on target cells. Engagement of these inhibitory receptors results in a dominant negative signal that prevents lysis of the target cells. Tyrosine phosphorylation of their ITIMs and recruitment of SHP-1 was shown to be critical for the inhibitory function (7,15).
Recently, it was shown that mouse NK cells as well as mast cells also express KIR-like molecules, gp49A (16) and gp49B1 (17), the latter of which contains two ITIM-like sequences in its cytoplasmic portion (17)(18)(19)(20)(21). Although the physiological function of gp49 molecules is unknown, it was demonstrated that the co-ligation of gp49B1 and a high affinity IgE receptor Fc⑀RI on the surface of mast cells suppresses Fc⑀RI-mediated exocytosis, suggesting that this molecule possibly functions as an inhibitory receptor (19). Moreover, transfection experiments have indicated that the cytoplasmic tail of gp49B1 inhibits lysis of major histocompatibility complex class I-negative cell line by mouse and human NK cell lines, showing the inhibitory nature of the gp49B1 molecule in NK cells (20). Thus, elucidation of the physiological ligand for gp49 as well as the biochemical characteristics of the inhibitory cascade of the molecule should be valuable for understanding the mode of action that regulates * This work was supported by research grants from the Ministry of Education, Science, Sports, and Culture of Japan (to Y. M. and T. T.), and Core Research for Evolutional Science and Technology, Japan Science and Technology Corporation (to T. T.). 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. activating and inhibitory signaling in cells of the immune system.
We report here that both of the phosphorylated ITIM-like sequences of gp49B1 bind SHP-1 and SHP-2, whereas the phosphotyrosyl second ITIM-like sequence associates with SHIP from RBL-2H3 cells. Moreover, we found that the gp49B gene and the cognate gp49A gene are co-localized on the B4 region of mouse chromosome 10, which is apparently not a syntenic position of human chromosome 19, where genes for the KIR family are present (6,22).

EXPERIMENTAL PROCEDURES
Screening and Isolation of Mouse Genomic DNA Clones for gp49B-A 1.0-kilobase pair (kbp) cDNA of gp49B was prepared by reverse transcription-polymerase chain reaction (PCR) of mRNA preparation from WEHI-3 cells using an oligo(dT) primer for a reverse transcription reaction, and the forward primer (5Ј-CGATAAGCTTGCCTGGACT-CACCATG-3Ј) corresponding to nucleotide residues 7-24 (17) containing a HindIII restriction site, and the backward primer (5Ј-CGATG-GATCCCTAGTTTTCATTCATGG-3Ј) corresponding to the residues 1075-1095 of gp49B1 containing a BamHI restriction site for a PCR reaction. After the PCR amplification and digestion with HindIII and BamHI, the gp49B cDNA was cloned into the HindIII/BamHI sites of pUC19. The cDNA insert was labeled with random primer labeling kit (Takara Shuzo Co., Otsu, Japan) and [␣-32 P]dCTP (specific activity, ϳ3,000 Ci/mmol, Amersham Corp.), and was used to screen a 129/Sv mouse genomic library (Stratagene): 8.3 ϫ 10 5 plaques were screened under stringent conditions (23). The resulting two positive clones were plaque-purified, and they were subcloned into the plasmid pUC19 or pBluescript (Stratagene) and sequenced by the dideoxy chain termination method (24) using a Cy5 AutoRead sequencing kit and an ALFexpress DNA sequencer (both from Pharmacia Biotech Inc.).
Cell Culture-Bone marrow-derived mast cells from C57BL/6 or B10.A mice were prepared as described previously (25). RBL-2H3 cells (obtained from the Japanese Cancer Research Resources Bank, Tokyo) were maintained in RPMI 1640 plus 10% fetal calf serum.
RBL-2H3 cells and bone marrow-derived mast cells were solubilized by adding extraction buffer (50 mM Tris-HCl, pH 7.4, 1% Nonidet P-40, 150 mM NaCl, 1 mM phenylmethyl sulfonyl fluoride, 10 g/ml aprotinin, 20 g/ml leupeptin, 1 mM disodium EDTA, 10% glycerol) to the cell pellet. The extract was then incubated on ice for 10 min, and the insoluble material was removed by centrifugation at 15,000 ϫ g for 15 min at 4°C. The soluble extract was precleared with 1.0 ml of avidin-Sepharose FF beads/10 7 cell eq (Pharmacia) in extraction buffer. The beads were removed by centrifugation at 15,000 ϫ g for 5 min, and the extract was then incubated for 18 h at 4°C with 0.5 ml of the biotinylated peptide avidin-Sepharose beads/10 7 cell eq, followed by precipitation. The pelleted beads were washed four times by centrifugation in extraction buffer containing 0.1% gelatin and devoid of aprotinin and leupeptin, resuspended in sodium dodecyl sulfate-polyacrylamide gel electrophoresis sample buffer (50 mM Tris-HCl, pH 6.8, 2% SDS, 0.1% bromphenol blue, 10% glycerol, and 150 mM 2-mercaptoethanol), boiled for 1 min to liberate immunoprecipitated materials, and centrifuged. The affinity-isolated materials or cell lysates were subjected to SDSpolyacrylamide gel electrophoresis separation using 7.5% gel and then transferred to a polyvinylidine difluoride membrane (Immobilon P, Millipore Corp.) using a Milliblot electroblotting system. The blot was first blocked with 2% nonfat dry milk (Carnation) in phosphate-buffered saline, 0.05% Tween 20, incubated with 0.5 g/ml rabbit anti-SHP-1 or anti-SHP-2 IgG (both from Santa Cruz Biotechnology, Santa Cruz, CA), or rabbit anti-SHIP antiserum (generously provided by Dr. K. M. Coggeshall, Ohio State University) and washed three times in phosphate-buffered saline/Tween. The blot was then treated with the peroxidase-conjugated mouse anti-rabbit IgG (clone RG-16, Sigma) and detected by enhanced chemiluminescence (Amersham Corp.).
Chromosome Preparation and in Situ Hybridizaiton-The direct Rbanding fluorescence in situ hybridization method was used for chro-mosomal assignment of the mouse gp49B gene. Preparation of R-banded chromosomes and fluorescence in situ hybridization were performed as described by Matsuda et al. (26) and Matsuda and Chapman (27). The chromosome slides were hardened at 65°C for 2 h, denatured at 70°C in 70% formamide in 2 ϫ SSC, and dehydrated in cold ethanol (5 min each in 70 and 100%).
The mouse 6.7-kb genomic DNA fragment inserted in SacI site of pBluescript were labeled by nick translation with biotin 16-dUTP (Boehringer Mannheim) following the manufacturer's protocol. The labeled DNA fragment was ethanol-precipitated with a 10-fold excess of mouse Cot-1 DNA (Life Technologies, Inc.), sonicated salmon sperm DNA and E. coli tRNA, and then denatured at 75°C for 10 min in 100% formamide. The denatured probe was mixed with an equal volume of hybridization solution to make final concentration of 50% formamide, 2 ϫ SSC, 10% dextran sulfate, and 2 mg/ml bovine serum albumin (Sigma). The mixed probe was placed on the denatured chromosome slides and incubated overnight at 37°C. The slides were washed in 50% formamide in 2 ϫ SSC at 37°C for 20 min and in 2 ϫ SSC and 1 ϫ SSC for 20 min each at room temperature. After rinsing in 4 ϫ SSC, they were incubated under coverslips with fluoresceinated avidin (Vector Laboratories) at a 1:500 dilution in 1% bovine serum albumin, 4 ϫ SSC for 1 h at 37°C. They were washed sequentially with 4 ϫ SSC, 0.1% Nonidet P-40 in 4 ϫ SSC and 4 ϫ SSC for 10 min each on a shaker, rinsed with 2 ϫ SSC, and stained with 0.75 g/ml propidium iodide (Sigma). Excitation at wavelength 450 -490 nm (Nikon filter set B-2A) and near 365 nm (UV-2A) was used for observation. Kodak Ektachrome ASA 100 films were used for microphotography.
Linkage Mapping with Interspecific Backcross Mice-Recombinant mice used in this study were generated by mating male feral-derived mouse stocks, Mus spretus, with female C57BL/6J, and backcrossing the F 1 female with male M. spretus (28). Genomic DNA derived from kidneys of each backcross mouse was digested with restriction endonucleases. The resulting fragments were separated on 0.8% agarose gels and transferred to a nylon membrane (Bio-Rad). The DNA was used to determine the genotype of each animal by Southern blot hybridization following the standard protocol. According to the result obtained by the cytogenetic mapping using fluorescence in situ hybridization, microsatellite DNA markers (Research Genetics, Huntsville, AL) for linkage analysis were chosen. All PCR reactions were performed in a total 15-l reaction mixture containing 125 ng genomic DNA and 15 pmol oligonucleotide primers. Amplification conditions were 95°C for 10 min, 30 cycles of 94°C for 1 min, 55°C for 1 min, and 72°C for 50 s, and 72°C for 10 min. The PCR products were visualized under UV light with ethidium bromide staining after separation on polyacrylamide gel. Mock, mock affinity isolation without any peptide but with avidin-Sepharose; Lysate, total cell lysate without any adsorption.

RESULTS AND DISCUSSION
Analysis of Cellular Components Associating with Cytoplasmic Domain of gp49B-gp49B has two possible ITIM sequences in its 74-amino acid cytoplasmic domain (17)(18)(19). One is IVYAQV, and the other is VTYAQL: they are separated by 18 amino acid residues. On the other hand, its cognate molecule gp49A has a short cytoplasmic region comprised of 42 amino acid residues, and has no such inhibitory sequences within this region (16). Transfection experiments have shown that the whole cytoplasmic portion of gp49B is sufficient for the inhibitory nature of the gp49B molecule in NK cells (20). A question arises whether the possible ITIM sequences of gp49B actually function as inhibitory motifs that interact with any cellular component or not. To address this issue, we attempted to detect cellular proteins associating with these sequences. Cellular proteins of RBL-2H3 cells and bone marrow-derived mast cells were lysed and then incubated with tyrosine-phosphorylated or nonphosphorylated synthetic peptide matching each ITIM-like sequence found in the cytoplasmic portion of gp49B. The bound proteins were separated by SDS-polyacrylamide gel electrophoresis and subjected to immunoblot analysis. As shown in Fig. 1, both anti-SHP-1 and anti-SHP-2 antibodies detected a ϳ65and ϳ70-kDa protein, respectively, bound to pY1 and pY2 but not to the corresponding nonphosphotyrosyl peptides. Unexpectedly, we found that anti-SHIP antiserum clearly detected a ϳ145-kDa protein associating with pY2 but not with pY1 and the corresponding nonphosphotyrosyl peptides. Detection of possible association of Syk kinase using anti-Syk antiserum was not successful (data not shown). Therefore, we concluded that each phosphorylated ITIM-like sequence mainly binds SHP-1 and SHP-2, whereas the second possible ITIM also associates with SHIP from RBL-2H3 extract. Inhibition of cellular activation signal by SHP-1 is associated with inhibition of early tyrosine phosphorylation events, release of Ca 2ϩ from intracellular stores, and secondary influx of extracellular Ca 2ϩ . On the other hand, SHIP inhibits only the secondary influx of extracellular Ca 2ϩ . Thus, we postulate that the two possible ITIM sequences in gp49B can function as inhibitory motifs within the cells. Since the recruitment of SHP-1, SHP-2 and SHIP is dependent on the presence of the phosphotyrosine residue, the phosphorylation of specific tyrosine residues in ITIM-like sequences of gp49B should be prerequisite for recruitment of SHP-1 and SHIP in stimulated cells such as mast cells. Since the gp49-specific antibody is not available at present, we cannot verify that the cytoplasmic tail of gp49B is tyrosine-phosphorylated within the cells. Possible candidates that phosphorylate the tyrosine residues in the cytoplasmic domain of gp49B are a src family kinase or a Syk/Zap-70 family kinase such as Lyn or Syk, but this issue remains to be clarified. A possibility remains that the preference of the association with SHP-1 or SHIP could be dependent on the cell type.
Juxtaposition of gp49A and gp49B Genes-As mentioned above, two related genes have been identified for mouse gp49, gp49A and gp49B, by cloning of cDNA for each molecule (16,17), cloning of genomic DNA for gp49B (17), and blot hybridization analysis of total genomic DNA (16). Recently, several groups of genes with almost the same structural characteristics but different in their cytoplasmic regions are revealed. One example is a group of KIR that comprises inhibitory KIR molecules and the molecules with nearly the same extracellular domains but with no inhibitory cytoplasmic domains such as some of the NK-associated transcripts (for review, see Refs. 29 -32). The genes for them have been found to be clustered in human chromosome 19q13 (22). Therefore, it is interesting to test a hypothesis that gp49 genes also locate in a limited area of a mouse chromosome.
The first step toward this issue, we have isolated the genomic clones for gp49B using a gp49B cDNA fragment as a probe. We initially characterized the exon/intron structures of one of the positive clones (clone 1, harboring a 16-kb insert and containing all of the exons for gp49B). Unexpectedly, we found that this genomic clone 1 also contained a 3Ј part of the gene for gp49A as shown in Fig. 2. An overlapping clone, clone 2, was also shown to contain all of the gp49A exons and a 5Ј part of the gene for gp49B by sequencing analysis. Thus, we found that gp49A and gp49B genes are in the juxtaposition and in the same orientation: they are apart by 4.4 kb (Fig. 2). In addition, sequencing analysis of the gp49A and gp49B genes from 129/Sv mouse revealed that there is no nucleotide substitution in the exon sequences when compared with the respective genes from BALB/c and C3H mice, respectively (data not shown), support- ing the notion that gp49 genes are not polymorphic within a species.
Chromosomal Mapping of the gp49B Gene-The chromosomal assignment of the mouse gp49B gene was performed by direct R-banding fluorescence in situ hybridization using a mouse genomic DNA fragment as a biotinylated probe. As shown in Fig. 3, the signals of the gp49B gene were detected on the R-band-positive B4 band of mouse chromosome 10 (33).
For fine linkage mapping of the mouse gp49B gene, genomic DNAs of C57BL/6J, M. spretus and their F 1 were digested with six different restriction endonucleases to find the restriction fragment length variants using Southern blot hybridization. The restriction fragment length variant with HindIII, 6.1 kb in C57BL/6J and 5.4 kb in M. spretus, was used to examine the concordance of segregation of the restriction fragment length variant with the segregation of microsatellite markers, D10Mit3, D10Mit38, D10Mit20, D10Mit31, and D10Mit7. Gene order was determined by minimizing the number of multiple recombinations among the loci on the same chromosome. Comparative pairwise loci analyses showed the gene order and recombination frequency for gp49B on mouse chromosome 10 (genetic distance in centimorgan Ϯ S.E. and the number of recombinations in parentheses) as follows: centromere, D10Mit3, 3.5 Ϯ 2.0 (3/86), D10Mit38, 3.5 Ϯ 2.0 (3/86), gp49B, 1.2 Ϯ 1.2 (1/86), D10Mit20, 3.5 Ϯ 2.0 (3/86), D10Mit31, 10.5 Ϯ 3.3 (9/86), D10Mit7, telomere (Fig. 4). It should be emphasized that the conserved linkage homology to human chromosome 19, where KIR genes have been localized, has not been identified in this region.
Recent progress in structural and functional aspects of KIR have proposed an intriguing hypothesis that there may be several unidentified groups of molecules with positive and negative regulatory functions in human as well as in mouse. In fact, human immunoglobulin-like transcript-1, -2, and -3 (34, 35), mouse gp49A and gp49B (17)(18)(19)(20)(21), and mouse p91 (36) and paired inhibitory receptors (37) are candidates for such regulatory groups or pairs that may play important roles in the initiation of cellular responses. In this report, we showed that the cytoplasmic ITIM sequences of gp49B1 are able to recruit SHP-1, SHP-2, and SHIP upon tyrosine phosphorylation, suggesting that the inhibitory function of gp49B1 is due to the recruitment of these phosphatases to the phosphorylated ITIM and inhibiting the activation signaling such as the increase in the intracellular Ca 2ϩ concentration. Studies on the ligand and activatory/inhibitory signaling cascades of gp49 molecules will facilitate the understanding of the mode of regulation of cells in the immune system.