Title: Genomic Organization and Chromosomal Localization of the Gene Encoding Human P-selectin Glycoprotein Ligand
Abstract: The gene for P-selectin glycoprotein ligand (PSGL-1) has been cloned from a human placenta genomic DNA library. A single intron of approximately 9 kilobases was found in the 5′-untranslated region and the complete coding region resides in exon 2. The genomic clone differs from the cDNA clone isolated from HL-60 cells in that it encodes an extra copy of the decameric repeat located in the extracellular domain of PSGL-1. Further analysis indicated that the PSGL-1 genes of HL-60 and U-937 cells contain 15 repeats, whereas the PSGL-1 genes of polymorphonuclear leukocytes, monocytes, and several other cell lines contain 16 repeats. Transfection experiments did not indicate a functional difference between these two variants of PSGL-1. The two previously observed PSGL-1 mRNA species of 2.5 and 4 kilobases most likely arise from differential utilization of polyadenylation signal sequences. The organization of the PSGL-1 gene closely resembles those of CD43 and human platelet glycoprotein GPIbα, both of which have an intron in the 5′-noncoding region, a long second exon containing the complete coding region, and TATA-less promoters. The gene for human PSGL-1, which has been designated SELPLG by the Human Gene Nomenclature Committee, was mapped to chromosome 12q24 using Southern blot analysis of DNA from a set of human-mouse cell hybrids, and fluorescent in situ hybridization on metaphase chromosome spreads. The gene for P-selectin glycoprotein ligand (PSGL-1) has been cloned from a human placenta genomic DNA library. A single intron of approximately 9 kilobases was found in the 5′-untranslated region and the complete coding region resides in exon 2. The genomic clone differs from the cDNA clone isolated from HL-60 cells in that it encodes an extra copy of the decameric repeat located in the extracellular domain of PSGL-1. Further analysis indicated that the PSGL-1 genes of HL-60 and U-937 cells contain 15 repeats, whereas the PSGL-1 genes of polymorphonuclear leukocytes, monocytes, and several other cell lines contain 16 repeats. Transfection experiments did not indicate a functional difference between these two variants of PSGL-1. The two previously observed PSGL-1 mRNA species of 2.5 and 4 kilobases most likely arise from differential utilization of polyadenylation signal sequences. The organization of the PSGL-1 gene closely resembles those of CD43 and human platelet glycoprotein GPIbα, both of which have an intron in the 5′-noncoding region, a long second exon containing the complete coding region, and TATA-less promoters. The gene for human PSGL-1, which has been designated SELPLG by the Human Gene Nomenclature Committee, was mapped to chromosome 12q24 using Southern blot analysis of DNA from a set of human-mouse cell hybrids, and fluorescent in situ hybridization on metaphase chromosome spreads. P-selectin glycoprotein ligand (PSGL-1)1( 1The abbreviations used are: PSGL-1P-selectin glycoprotein ligand-1PMNpolymorphonuclear leukocytekbkilobase(s)SELPLGgene symbol for PSGL-1ntnucleotidesPCRpolymerase chain reactionRACErapid amplification of cDNA endsRT-PCRreverse transcription PCRbpbase pair(s).) 1The abbreviations used are: PSGL-1P-selectin glycoprotein ligand-1PMNpolymorphonuclear leukocytekbkilobase(s)SELPLGgene symbol for PSGL-1ntnucleotidesPCRpolymerase chain reactionRACErapid amplification of cDNA endsRT-PCRreverse transcription PCRbpbase pair(s). is the high affinity counter-receptor for P-selectin on myeloid cells and stimulated T lymphocytes2( 2G. Vachino, X.-J. Chang, G. M. Veldman, R. Kumar, L. A. Fouser, M. C. Berndt, and D. A. Cumming, manuscript in preparation.) 2G. Vachino, X.-J. Chang, G. M. Veldman, R. Kumar, L. A. Fouser, M. C. Berndt, and D. A. Cumming, manuscript in preparation. (1Sako D. Chang X.-J. Barone K.M. Vachino G. White H.M. Shaw G. Veldman G.M. Bean K. Ahern T.J. Furie B. Cumming D. Larsen G.R. Cell. 1993; 75: 1179-1186Abstract Full Text PDF PubMed Scopus (643) Google Scholar, 2Moore K.L. Stults N.L. Diaz S. Smith D.F. Cummings R.D. Varki A. McEver R.P. J. Cell Biol. 1992; 118: 445-456Crossref PubMed Scopus (419) Google Scholar, 3Norgard K.E. Moore K.L. Diaz S. Stults N.L. Ushiyama S. McEver R.P. Cummings R.D. Varki A. J. Biol. Chem. 1993; 268: 12764-12774Abstract Full Text PDF PubMed Google Scholar, 4Moore K.L. Eaton S.F. Lyons D.E. Lichenstein H.S. Cummings R.D. McEver R.P. J. Biol. Chem. 1994; 269: 23318-23327Abstract Full Text PDF PubMed Google Scholar, 5Moore K.L. Patel K.D. Bruehl R.E. Fugang L. Johnson D.A. Lichenstein H.S. Cummings R.D. Bainton D.F. McEver R.P. J. Cell Biol. 1995; 128: 661-671Crossref PubMed Scopus (624) Google Scholar). As such it plays a critical role in the tethering of these cells to activated platelets or endothelia expressing P-selectin(5Moore K.L. Patel K.D. Bruehl R.E. Fugang L. Johnson D.A. Lichenstein H.S. Cummings R.D. Bainton D.F. McEver R.P. J. Cell Biol. 1995; 128: 661-671Crossref PubMed Scopus (624) Google Scholar). A cDNA clone for human PSGL-1 has been isolated from HL-60 cells and encodes a unique, mucin-like glycoprotein(1Sako D. Chang X.-J. Barone K.M. Vachino G. White H.M. Shaw G. Veldman G.M. Bean K. Ahern T.J. Furie B. Cumming D. Larsen G.R. Cell. 1993; 75: 1179-1186Abstract Full Text PDF PubMed Scopus (643) Google Scholar). P-selectin glycoprotein ligand-1 polymorphonuclear leukocyte kilobase(s) gene symbol for PSGL-1 nucleotides polymerase chain reaction rapid amplification of cDNA ends reverse transcription PCR base pair(s). P-selectin glycoprotein ligand-1 polymorphonuclear leukocyte kilobase(s) gene symbol for PSGL-1 nucleotides polymerase chain reaction rapid amplification of cDNA ends reverse transcription PCR base pair(s). To date little is known about the organization and attributes of the human PSGL-1 gene (SELPLG). In a previous study, two PSGL-1 mRNA transcripts (a predominant species of approximately 2.5 kb and a minor 4-kb species) were identified in PMNs and HL-60 cells(1Sako D. Chang X.-J. Barone K.M. Vachino G. White H.M. Shaw G. Veldman G.M. Bean K. Ahern T.J. Furie B. Cumming D. Larsen G.R. Cell. 1993; 75: 1179-1186Abstract Full Text PDF PubMed Scopus (643) Google Scholar). Furthermore, Southern blot analysis indicated that the PSGL-1 gene is present as a single copy in the human genome, suggesting that the two mRNAs are alternative transcripts derived from the same gene, resulting from either alternative splicing events or differential utilization of promoter and/or polyadenylation signal sequences(1Sako D. Chang X.-J. Barone K.M. Vachino G. White H.M. Shaw G. Veldman G.M. Bean K. Ahern T.J. Furie B. Cumming D. Larsen G.R. Cell. 1993; 75: 1179-1186Abstract Full Text PDF PubMed Scopus (643) Google Scholar). The presence of multiple mRNA transcripts for PSGL-1 raises questions about the organization of the PSGL-1 gene, in particular with respect to the number and placement of introns. Although the coding regions of the majority of eukaryotic genes are interrupted by introns, a number of integral membrane proteins lack introns in their coding region. In this regard, the glycoproteins genes can be divided into two classes. The first class represents genes whose coding region is contained on one long exon. Examples of this class are human and mouse CD43 and human GPIbα, which also share several other common features such as a promoter lacking the typical TATA box(6Shelley C.E. Remold-O'Donnell E. Rosen F.S. Whitehead A. Biochem. J. 1990; 270: 569-576Crossref PubMed Scopus (26) Google Scholar, 7Cyster J. Somoza C. Killeen N. Williams A. Eur. J. Immunol. 1990; 20: 875-881Crossref PubMed Scopus (49) Google Scholar, 8Wenger R.H. Kieffer N. Wicki A.N. Clemetson K.J. Biochem. Biophys. Res. Commun. 1988; 156: 389-395Crossref PubMed Scopus (89) Google Scholar). The second class represents genes whose coding region consists of multiple exons, and the promoters of these genes generally (albeit with a few exceptions) resemble the typical eukaryotic promoter containing a TATA consensus sequence. Examples of this class include glycophorin, CD4, and the L-selectin ligands GlyCAM-1 and CD34(9Kudo S. Fukuda M. Proc. Natl. Acad. Sci. U. S. A. 1989; 86: 4619-4623Crossref PubMed Scopus (81) Google Scholar, 10Siu G. Wurster A.L. Lipsick J.S. Hedrick S.M. Mol. Cell. Biol. 1992; 12: 1592-1604Crossref PubMed Scopus (114) Google Scholar, 11Dowbenko D. Andalibi A. Young P.E. Lusos A.J. Lasky L.A. J. Biol. Chem. 1993; 268: 4525-4529Abstract Full Text PDF PubMed Google Scholar, 12Brown J. Greaves M. Molgaard H. Int. Immunol. 1991; 3: 175-184Crossref PubMed Scopus (145) Google Scholar). Another potentially interesting question concerns the chromosomal localization of the PSGL-1 gene. The genes for P-selectin, E-selectin, and L-selectin as well as the L-selectin ligand CD34 all map to chromosome 1(13Watson M.L. Kingsmore S.F. Johnston G.I. Siegelman M.H. Le Beau M.M. Lemons R.S. Bora N.S. Howard T.A. Weissman I.L. McEver R.P. Seldin M.F. J. Exp. Med. 1990; 172: 263-272Crossref PubMed Scopus (113) Google Scholar, 14Molgaard H. Spurr N. Greaves M. Leukemia. 1989; 3: 773-778PubMed Google Scholar). The clustering of the selectins and their counter-receptor(s) to the same chromosome has obvious evolutionary implications and therefore it is of interest to ascertain whether PSGL-1 also maps to chromosome 1. Here we report the isolation, genomic organization, and chromosomal mapping of the human PSGL-1 gene. The tumor cell lines HL-60, U-937, THP-1, Ramos, CEM, MOLT-4, and SB were obtained from the American Type Culture Collection and maintained in RPMI 1640 containing 10% fetal calf serum (Sigma). Human polymorphonuclear leukocytes (PMNs) were isolated from whole blood by centrifugation over a Mono-Poly Resolving Medium Ficoll-Hypaque density gradient (ICN Biomedicals, Costa Mesa, CA). Peripheral blood mononuclear cells were isolated from heparinized blood of normal donors by centrifugation over a Ficoll/Hypaque density gradient (Sigma), and monocytes were subsequently isolated by adherence to plastic. A human placenta genomic library in the λ Fix II vector (Stratagene, La Jolla, CA) was plated at ~20,000 plaques/plate, and nitrocellulose replicas of the plates were hybridized according to standard procedures(15Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York1989Google Scholar). The PSGL-1 cDNA probe comprising nucleotides (nt) 60-389, was amplified by polymerase chain reaction (PCR) and labeled to a specific activity of 5 × 108 cpm/μg with [32P]dCTP (Amersham Corp.) using the random-primed DNA labeling kit (Boehringer Mannheim). Filters were hybridized for 16 h at 65°C with the labeled probe in 6 × SSC, 10 × Denhardt's solution, 0.1% SDS, and 50 μg/ml denatured, sheared salmon sperm DNA. The filters were washed three times for 30 min in 0.3 × SSC, 0.1% SDS at 65°C and exposed to film. Duplicate positive plaques were rescreened until purified plaques were obtained, and the λ DNAs were prepared using a Qiagen lambda kit (Qiagen, Chatworth, CA). Ten independent λ clones were obtained and rescreened with PSGL-1 sequence-specific oligonucleotide probes, which were end-labeled using [γ-32P]ATP and T4 polynucleotide kinase. Clones λ2 and λ4 were characterized by restriction mapping, and suitable restriction fragments were subcloned into the pBluescript vector (Stratagene). Subclones containing PSGL-1 gene exons and flanking regions were sequenced using the Applied Biosystems Taq DyeDeoxy Terminator Kit and ABI 373 autosequencer(16Smith L.M. Sanders J.Z. Kaiser R.J. Hughes P. Dodd C. Connell C.R. Heiner C. Kent S.B.H. Hood L.E. Nature. 1986; 321: 674-679Crossref PubMed Scopus (1250) Google Scholar). DNA sequences were analyzed using the Sequencher software package (Gene Codes, Ann Arbor, MI). Total RNA was purified using the guanidine isothiocyanate/cesium chloride cushion method(15Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York1989Google Scholar). Poly(A+) mRNA was purified using the PolyATtract® mRNA isolation system II (Promega, Madison, WI). RNA was fractionated by 1.0% formaldehyde-agarose gel electrophoresis and transferred to nitrocellulose via capillary blotting. The PSGL-1 cDNA or gene fragments to be used as probes were amplified by PCR and labeled to a specific activity of at least 1 × 109 cpm/μg with [P]dATP and [32P]dCTP and hybridized to filters as described above. Rapid amplification of cDNA 5′ and 3′ ends was performed as described by Innis et al.(17Innis M.A. Gelfand D.H. Sninsky J.J. White T.J. PCR Protocols: A Guide to Methods and Applications. Academic Press, Inc., San Diego, CA1990Google Scholar). Briefly, three primers were used for the 3′-RACE experiment. The dT17-adapter, a 35-mer containing an oligo(dT) stretch (17 residues) linked to a unique 18-nt adapter, was used to prime reverse transcription of total RNA isolated from PMNs, using the AMV Reverse Transcriptase System (Life Technologies, Inc.). The 18-nt adapter primer and a PSGL-1-specific primer, a 19-mer corresponding to nt 1564-1582, were subsequently used to amplify the cDNA by PCR. PCR products were analyzed on a 1% agarose gel. For 5′-RACE experiments, HL-60 poly(A+) RNA and PMN total RNA were reverse-transcribed using a PSGL-1-specific 18-mer (antisense to nt 294-311 in exon 2). The first-strand reaction products were extended with dATP and terminal transferase and subsequently amplified by PCR using a mixture of the dT17-adapter and the 18-nt adapter as the 5′ primers and a PSGL-1-specific 21-mer (antisense to nt 243-263) as the 3′ primer. The PCR products were analyzed on a 1.5% agarose gel and identified by Southern hybridization with a PSGL-1 oligonucleotide probe corresponding to nt 1-14 in exon 1. Primer extension was done according to described methods(15Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York1989Google Scholar). In brief, a 30-mer 5′-ACCACCGTGCTCAGCAGAGCATGGGACAGC-3′ (antisense to nt 27-56 in exon 1) was end-labeled to a specific activity of 8 × 106 cpm/pmol as described above. Approximat ely 5 ng of labeled primer was hybridized to 3 μg of poly(A+) RNA (HL-60) or 3 μg total RNA (PMNs) at 48°C overnight. The extension reaction was done using the AMV Reverse Transcriptase System (Life Technologies, Inc.) with 20 units of AMV reverse transcriptase (Life Sciences, St. Petersburg, FL) for 2 h at 48°C. The extension products were sized by electrophoresis on a denaturing 12% polyacrylamide gel using Polyoma/DdeI fragments as size markers. The number of repeats in PSGL-1 from different cell types was determined by PCR using primers that flank the repeat region. The 5′ primer corresponds to nt 381-400, and the 3′ primer is antisense to nt 880-903 of the PSGL-1 gene, resulting in PCR products of 522 bp and 492 bp, representing 16 and 15 copies of the repeat, respectively. mRNA templates were analyzed by reverse transcription PCR (RT-PCR), using the reverse transcriptase reaction described above to generate a cDNA template for PCR amplification. The PCR reactions contained ~1 ng of cDNA or genomic DNA template and were performed according to standard procedures (17Innis M.A. Gelfand D.H. Sninsky J.J. White T.J. PCR Protocols: A Guide to Methods and Applications. Academic Press, Inc., San Diego, CA1990Google Scholar) using Taq DNA polymerase (Promega). The PCR products were sized by electrophoresis on a 2.5% NuSieve (FMC Bioproducts, Rockland, ME), 0.5% agarose gel. Mapping of the human PSGL-1 gene was achieved by probing Southern blots of DNAs from 29 mouse-human somatic cell hybrids that were made from 17 unrelated human cell lines and 4 mouse cell lines(18Shows T.B. Sakaguchi A.Y. Naylor S.L. Harris H. Hirschhorn K. Advances in Human Genetics. Plenum Press, New York1982: 341-352Crossref Scopus (177) Google Scholar, 19Shows T. Eddy R. Haley L. Byers M. Henry M. Fujita T. Matsui H. Taniguchi T. Somatic Cell Mol. Gen. 1984; 10: 315-318Crossref PubMed Scopus (122) Google Scholar, 20Shows T.B. Rattazzi M.C. Scandalios J.G. Whitt G.S. Isozymes: Current Topics in Biological and Medical Research. Alan R. Liss, New York1983: 323-339Google Scholar). The hybrids were characterized by karyotypic analysis and by mapped enzyme markers (18Shows T.B. Sakaguchi A.Y. Naylor S.L. Harris H. Hirschhorn K. Advances in Human Genetics. Plenum Press, New York1982: 341-352Crossref Scopus (177) Google Scholar, 20Shows T.B. Rattazzi M.C. Scandalios J.G. Whitt G.S. Isozymes: Current Topics in Biological and Medical Research. Alan R. Liss, New York1983: 323-339Google Scholar). Chromosome translations with no intact chromosome present were not tabulated for the percent discordance. Cell hybrid DNA was digested with XbaI and subjected to Southern analysis using the 32P-labeled PSGL-1 cDNA probe comprising nt 60-389 (21Naylor S.L. Sakaguchi A.Y. Shows T.B. Law M.L. Goeddel D.V. Gray P.W. J. Exp. Med. 1983; 157: 1020-1027Crossref PubMed Scopus (133) Google Scholar). Control DNA from human and mouse gave a single human XbaI hybridizing fragment of 8 kb, with no cross-hybridization with mouse DNA. In situ fluorescent hybridization on metaphase chromosome spreads was done according to the method of Trask(22Trask B. Methods in Cell Biology: A Laboratory Guide. Vol. 35. Academic Press, New York1992: 1-35Google Scholar). The PSGL-1 gene fragment XP2.4 was labeled with digoxygenin-11-dUTP (Boehringer Mannheim) by random priming(23Feinberg A.P. Vogelstein B. Anal. Biochem. 1983; 132: 6-13Crossref PubMed Scopus (16562) Google Scholar). Digital images were obtained using a Nikon epifluorescence microscope coupled to a cooled CCD camera (Photometrics). The images were merged and enhanced using the IP Lab Spectrum image analysis software supported on an Apple Macintosh IIci. The PSGL-1 gene was isolated by hybridization of a human placenta genomic library using a 32P-labeled cDNA probe comprising nucleotides 60-389 of PSGL-1(1Sako D. Chang X.-J. Barone K.M. Vachino G. White H.M. Shaw G. Veldman G.M. Bean K. Ahern T.J. Furie B. Cumming D. Larsen G.R. Cell. 1993; 75: 1179-1186Abstract Full Text PDF PubMed Scopus (643) Google Scholar). Ten independent genomic clones were isolated after screening approximately 106 phages. Hybridization of these 10 clones with a 32P-labeled 14-mer derived from the 5′ end of the PSGL-1 cDNA revealed that only two clones were positive, suggesting that a large intron interrupted the PSGL-1 gene separating the sequences defined by the two probes. Two overlapping clones, denoted λ2 and λ4, were found to contain inserts of approximately 13 and 16 kb, respectively, and were selected for further fine mapping of the gene. The λ2 clone, which hybridized only to the PSGL-1 probe used in the primary screen, and the λ4 clone, which hybridized with both probes, were digested with HindIII, PstI, or XbaI and subcloned into the pBluescriptII vector. By this procedure several subclones, designated pPst4.5A, pPst4.5B, pHind5.0, pHind5.2, and pXba8, were generated (Fig. 1). Restriction enzyme mapping and hybridization of the generated fragments with oligonucleotides derived from the 5′ terminus (nt 1-14) and 3′ terminus (nt 1636-1649) of the PSGL-1 cDNA clone further established the genomic map for human PSGL-1 as presented in Fig. 1. Indeed, the PSGL-1 gene contains an approximately 9-kb intron inserted in the 5′-noncoding region between nt 54 and nt 55, based on the numbering system of the original PSGL-1 cDNA clone(1Sako D. Chang X.-J. Barone K.M. Vachino G. White H.M. Shaw G. Veldman G.M. Bean K. Ahern T.J. Furie B. Cumming D. Larsen G.R. Cell. 1993; 75: 1179-1186Abstract Full Text PDF PubMed Scopus (643) Google Scholar). The coding region which starts at nt 60 is fully encoded by exon 2 and does not contain any introns (Fig. 2).Figure 2:Nucleotide sequence of the human PSGL-1 gene. Position +1 corresponds to the first nucleotide of the PSGL-1 cDNA (Sako). The sequences shown include 440 nt of 5′-flanking sequence, exon 1 (nt 1-54), part of the 9-kb intron (lowercase letters), exon 2 (nt 55-2075, including the complete coding region), and additional 3′-flanking sequences. The amino acid sequence is shown in the one-letter code. The potential binding sites for Ets (dashed line) and Sp1 (….) are indicated. The extra repeat encoding 10 amino acids is boxed. The major polyadenylation signal is underlined and the Alu sequence is indicated between brackets.View Large Image Figure ViewerDownload Hi-res image Download (PPT) Correct splicing of the intron and linking of exons 1 and 2, as observed in the PSGL-1 cDNA clone derived from an HL-60 library, also occurs in RNA from different sources. RT-PCR on freshly isolated human PMN, monocyte, and HL-60 RNA with oligonucleotides spanning the intron yielded the expected PCR product in all three cases (data not shown). The putative transcription initiation region of the PSGL-1 gene was defined by 5′-RACE and by primer extension on RNA isolated from human PMNs and HL-60 cells. 5′-RACE was performed with a primer complementary to nt 243-263 in exon 2 as described under “Experimental Procedures.” Hybridization of the PCR products with an oligonucleotide probe derived from exon 1 identified a broad band of 280-340 bp from PMN as well as HL-60 RNA templates (data not shown). Considering that the requisite dT17 adapter is 35 nt long, this tentatively places the transcription initiation region between −42 and +18. Primer extension analysis of human PMN and HL-60 RNA was performed using a 30-mer complementary to nt 27-56 in exon 1. Several primer extension products were observed including two predominant bands between 55 and 70 nt in length, suggesting initiation from multiple sites in the region from −14 to +1 (Fig. 3). The region immediately upstream from these transcriptional start sites contains no TATA or CAAT box, but is GC-rich and contains putative binding sites for the regulatory elements Sp1 and Ets (Fig. 2). Two PSGL-1 transcripts, a predominant mRNA of approximately 2.5 kb and a minor 4-kb mRNA, have previously been identified in PMNs and HL-60 cells(1Sako D. Chang X.-J. Barone K.M. Vachino G. White H.M. Shaw G. Veldman G.M. Bean K. Ahern T.J. Furie B. Cumming D. Larsen G.R. Cell. 1993; 75: 1179-1186Abstract Full Text PDF PubMed Scopus (643) Google Scholar). The 3′-noncoding region of the PSGL-1 gene contains a polyadenylation signal at position 2075 as indicated in Fig. 2. Mapping of the PSGL-1 mRNA species from PMNs by 3′-RACE with an oligonucleotide corresponding to nt 1564-1582 identified a predominant PCR product of approximately 520 bp, indicating that the major PSGL-1 mRNA species utilizes this polyadenylation signal sequence (data not shown). Additional mapping of the minor 4-kb mRNA species was performed by Northern blot analysis of HL-60 poly(A+) RNA with two different PSGL-1 probes. Fig. 4 shows that the PSGL-1 cDNA probe, comprising nt 60-389, hybridized with the 2.5-kb mRNA and also (albeit very weakly in this exposure) with the 4-kb mRNA (lane A). The genomic DNA probe, comprising nt 2080-2459 located 3′ of the polyadenylation signal, only hybridized to the 4-kb mRNA (lane B), suggesting that the 4-kb mRNA utilizes a downstream, yet unidentified, polyadenylation signal sequence. The region of the PSGL-1 gene that is unique for the 4-kb transcript contains an Alu sequence (24Batzer M.A. Deininger P.L. Genomics. 1991; 9: 481-487Crossref PubMed Scopus (229) Google Scholar) as shown in Fig. 2. The PSGL-1 cDNA clone (pPL85) derived from an HL-60 cDNA library encodes 15 consecutive repeats of a 10-amino acid consensus sequence(1Sako D. Chang X.-J. Barone K.M. Vachino G. White H.M. Shaw G. Veldman G.M. Bean K. Ahern T.J. Furie B. Cumming D. Larsen G.R. Cell. 1993; 75: 1179-1186Abstract Full Text PDF PubMed Scopus (643) Google Scholar). Our genomic clone derived from human placenta, however, contains an extra sequence of 30 nucleotides (boxed in Fig. 2) located in the second repeat unit and thus encodes 16 such repeats. RT-PCR analysis with primers spanning the repeat region allowed for rapid analysis of the number of repeats in PSGL-1 mRNA from different cell types. Similarly, the number of repeats in PSGL-1 genomic DNA from different sources was determined by PCR analysis. Fig. 5 shows that freshly isolated HL-60 RNA contains 15 repeats (lane 3) similar to the original PSGL-1 cDNA clone pPL85 (lane 1), eliminating the possibility that the cDNA clone had lost one repeat due to a cloning artifact. Identical results were obtained for RNA isolated from the monocytic cell line U-937 (lane 7). In contrast, 16 repeats were found in the genomic clone λ4 (lane 2) and in mRNA isolated from human PMNs (lane 5) and monocytes (lane 6) and several human cell lines, such as THP-1 (lane 8) and Ramos, CEM, MOLT-4, and SB (not shown). To investigate whether splicing was responsible for this phenomenon, genomic DNA isolated from HL-60 cells was analyzed by PCR. Fig. 5(lane 4) shows that the HL-60 PSGL-1 gene encodes just 15 repeats. Similarly, the PSGL-1 gene of U-937 cells only encodes 15 repeats (data not shown). Sequence analysis of the repeat region in the PSGL-1 gene from U-937 cells confirmed that the U-937 and HL-60 PSGL-1 genes are identical and are therefore lacking the same repeat unit. There does not appear to be a direct functional role for the extra decameric repeat. The extra repeat was introduced into the PSGL-1 cDNA expression vector pPL85, generating pPL85-R16. COS cells cotransfected with vectors encoding the α(1,3/1,4)-fucosyltransferase gene, and pPL85 or pPL85-R16 bound equally well to P-selectin in vitro (data not shown). Human chromosomal localization of the PSGL-1 gene was performed in a two-step procedure employing analysis of a human-mouse somatic cell hybrid panel and chromosomal fluorescent in situ hybridization. Southern analysis of XbaI-digested DNA from 29 human-mouse hybrids with the human PSGL-1 cDNA probe comprising nt 60-389 detected a single 8-kb fragment in human DNA. This human PSGL-1 probe did not cross-hybridize with mouse DNA. The results in Table I show that only chromosome 12 gave 0% discordance, indicating that SELPLG localizes to chromosome 12.Table ISegregation of PSGL-1 with human chromosomes in XbaI-digested human-mouse cell hybrid DNA Open table in a new tab Fluorescent in situ hybridization on metaphase chromosome spreads was done with a 2.4-kb PSGL-1 genomic DNA probe (XP2.4 in Fig. 1) and the results are shown in Fig. 6. The double fluorescent signals on both chromosomes were found only at 12q24 in 17 out of 20 metaphase spreads (85%) examined and on no other chromosome. Three out of 20 spreads showed double fluorescent signals on only one chromosome. These results localize SELPLG to 12q24. A single 23-kb human PSGL-1 gene was isolated on two overlapping clones derived from a human placenta genomic library. No clones were isolated from any additional genomic loci, consistent with the existence of a single human PSGL-1 gene and in agreement with previously described Southern hybridization analysis of human genomic DNA(1Sako D. Chang X.-J. Barone K.M. Vachino G. White H.M. Shaw G. Veldman G.M. Bean K. Ahern T.J. Furie B. Cumming D. Larsen G.R. Cell. 1993; 75: 1179-1186Abstract Full Text PDF PubMed Scopus (643) Google Scholar). The human PSGL-1 gene contains a single intron of approximately 9 kb located 5 bp upstream of the ATG initiation codon, and the complete coding region is contained within exon 2. In contrast, many members of the mucin gene family, such as glycophorin, episialin, or MUC-1, and the L-selectin ligands CD34 and GlyCAM-1, are encoded by multiple exons(9Kudo S. Fukuda M. Proc. Natl. Acad. Sci. U. S. A. 1989; 86: 4619-4623Crossref PubMed Scopus (81) Google Scholar, 11Dowbenko D. Andalibi A. Young P.E. Lusos A.J. Lasky L.A. J. Biol. Chem. 1993; 268: 4525-4529Abstract Full Text PDF PubMed Google Scholar, 12Brown J. Greaves M. Molgaard H. Int. Immunol. 1991; 3: 175-184Crossref PubMed Scopus (145) Google Scholar, 25Lancaster C. Peat N. Duhig T. Wilson D. Taylor-Papadimitriou J. Gendler S. Biochem. Biophys. Res. Commun. 1990; 173: 1019-1029Crossref PubMed Scopus (122) Google Scholar). Although unusual, the lack of introns in the coding region has been reported for other integral membrane proteins, such as the human and mouse sialomucin CD43 and the human blood platelet membrane glycoprotein GPIbα(6Shelley C.E. Remold-O'Donnell E. Rosen F.S. Whitehead A. Biochem. J. 1990; 270: 569-576Crossref PubMed Scopus (26) Google Scholar, 7Cyster J. Somoza C. Killeen N. Williams A. Eur. J. Immunol. 1990; 20: 875-881Crossref PubMed Scopus (49) Google Scholar, 8Wenger R.H. Kieffer N. Wicki A.N. Clemetson K.J. Biochem. Biophys. Res. Commun. 1988; 156: 389-395Crossref PubMed Scopus (89) Google Scholar). These genes, like the PSGL-1 gene, lack introns in the coding region yet also have an intron in the 5′-untranslated region. Analysis of possible transcription initiation sites also indicated features shared between the PSGL-1 gene and the genes for CD43 and GPIbα. Primer extension and 5′-RACE experiments were employed to identify the transcription initiation sites of the PSGL-1 gene in PMNs and HL-60 cells. These preliminary analyses indicated that transcription may initiate from multiple sites just upstream of exon 1. Our data does not exclude the possibility that transcription of the PSGL-1 gene originates from other promoters as well, since those 5′-RACE PCR products would fail to hybridize with the exon 1 probe. A complete understanding of possible other promoter usage will require sequence analysis and mapping of cloned RACE PCR products. The PSGL-1 promoter region tentatively identified in this study lacks consensus sequences such as TATA and CAAT, which are general components of eucaryotic promoters. Several features of the PSGL-1 promoter, e.g. absence of TATA and CAAT boxes, multiple transcription initiation sites, high G + C content (55%), and a potential Sp1 binding site, resemble those of “housekeeping” genes(26Dynan W.S. Trends Genet. 1986; 2: 196-197Abstract Full Text PDF Scopus (369) Google Scholar). However, several tissue-specific promoters that lack TATA and CAAT boxes, but are nonetheless regulated, have been described recently, including the promoters for human GPIbα and CD43(8Wenger R.H. Kieffer N. Wicki A.N. Clemetson K.J. Biochem. Biophys. Res. Commun. 1988; 156: 389-395Crossref PubMed Scopus (89) Google Scholar, 27Wenger R.H. Wicki A.N. Kieffer N. Adolph S. Hameister H. Clemetson K.J. Gene (Amst.). 1989; 85: 517-524Crossref PubMed Scopus (54) Google Scholar, 28Kudo S. Fukuda M. J. Biol. Chem. 1991; 266: 8483-8489Abstract Full Text PDF PubMed Google Scholar). Other examples of TATA-less promoters include those for the integrins CD11a and CD11b and for CD4(29Cornwell R.D. Gollahon K.A. Hickstein D.D. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 4221-4225Crossref PubMed Scopus (41) Google Scholar, 30Hickstein D.D. Baker D.M. Gollahon K.A. Back A.L. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 2105-2109Crossref PubMed Scopus (75) Google Scholar, 31Salmon P. Giovane A. Wasylyk B. Klatzmann D. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 7739-7743Crossref PubMed Scopus (102) Google Scholar). Potential binding sites for transcriptional activators, such as Sp1, PU.1, Ets, and AP-2, have been identified in these promoter regions and are thought to direct tissue-specific gene expression(30Hickstein D.D. Baker D.M. Gollahon K.A. Back A.L. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 2105-2109Crossref PubMed Scopus (75) Google Scholar, 31Salmon P. Giovane A. Wasylyk B. Klatzmann D. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 7739-7743Crossref PubMed Scopus (102) Google Scholar). The PSGL-1 promoter has potential binding sites for Sp1 and Ets and therefore may very well fall in this category of promoters(32Faisst S. Meyer S. Nucleic Acids Res. 1992; 20: 3-26Crossref PubMed Scopus (938) Google Scholar, 33Wasylyk B. Hahn S.L. Giovane A. Eur. J. Biochem. 1993; 211: 7-18Crossref PubMed Scopus (809) Google Scholar). Indeed PSGL-1 expression appears to be regulated in a cell type-specific manner, since PSGL-1 transcripts have been found in myeloid and lymphocytic cells2 but not in the hepatoblastoma cell line HepG2 (1Sako D. Chang X.-J. Barone K.M. Vachino G. White H.M. Shaw G. Veldman G.M. Bean K. Ahern T.J. Furie B. Cumming D. Larsen G.R. Cell. 1993; 75: 1179-1186Abstract Full Text PDF PubMed Scopus (643) Google Scholar) or the human lung fibroblast line MRC-5.3( 3G. M. Veldman and G. Vachino, unpublished observation.) This is consistent with obser vations by others that cell surface PSGL-1 can be detected on leukocytes, but not on fibroblasts and keratinocytes(5Moore K.L. Patel K.D. Bruehl R.E. Fugang L. Johnson D.A. Lichenstein H.S. Cummings R.D. Bainton D.F. McEver R.P. J. Cell Biol. 1995; 128: 661-671Crossref PubMed Scopus (624) Google Scholar, 34Alon R. Rossiter H. Wang X. Springer T. Kupper T.S. J. Cell Biol. 1994; 127: 1485-1495Crossref PubMed Scopus (130) Google Scholar). The two previously identified PSGL-1 mRNA transcripts of 2.5 and 4 kb appear to arise from the single gene by differential utilization of polyadenylation sequences. Analysis of the 3′ end of PSGL-1 mRNAs indicated that the major mRNA species utilizes the polyadenylation signal at position 2075. This finding (together with the 5′ start mapping data) would predict a mRNA species of approximately 2.1 kb, not including the poly(A) tail. This is in agreement with the major PSGL-1 mRNA on Northern blots which was sized at approximately 2.5 kb(1Sako D. Chang X.-J. Barone K.M. Vachino G. White H.M. Shaw G. Veldman G.M. Bean K. Ahern T.J. Furie B. Cumming D. Larsen G.R. Cell. 1993; 75: 1179-1186Abstract Full Text PDF PubMed Scopus (643) Google Scholar). The minor 4-kb PSGL-1 mRNA hybridized to sequences located 3′ of the polyadenylation signal at position 2075, suggesting that the size difference of the two mRNA species may primarily arise from utilization of different polyadenylation sequences. A similar situation has been reported for the 1.9- and 4.3-kb mRNA species observed for CD43(35Shelley C.S. Remold-O'Donnell E. Davis III, A.E. Bruns G.A.P. Rosen F.S. Carroll M.C. Whitehead A.S. Proc. Natl. Acad. Sci. U. S. A. 1989; 86: 2819-2823Crossref PubMed Scopus (117) Google Scholar). The PSGL-1 gene in HL-60 and U-937 cells encodes 15 repeats of a 10-amino acid consensus sequence(1Sako D. Chang X.-J. Barone K.M. Vachino G. White H.M. Shaw G. Veldman G.M. Bean K. Ahern T.J. Furie B. Cumming D. Larsen G.R. Cell. 1993; 75: 1179-1186Abstract Full Text PDF PubMed Scopus (643) Google Scholar). This is in contrast to our human placenta genomic PSGL-1 clone, as well as the PSGL-1 in all other cells and cell lines tested, including the recently reported leukocyte PSGL-1 cDNA(5Moore K.L. Patel K.D. Bruehl R.E. Fugang L. Johnson D.A. Lichenstein H.S. Cummings R.D. Bainton D.F. McEver R.P. J. Cell Biol. 1995; 128: 661-671Crossref PubMed Scopus (624) Google Scholar), which contain 16 repeats. However, the number of consensus repeats does not appear to correlate with PSGL-1 function. HL-60 and U-937 cells bind P-selectin as do freshly isolated PMNs, monocytes, and THP-1 cells, all of which contain 16 copies of the repeat. In contrast, the lymphocytic cell lines Ramos, CEM, MOLT-4, and SB, which also express PSGL-1 with 16 repeats, do not bind P-selectin.2 Moreover, COS cells cotransfected with α(1,3/1,4)-fucosyltransferase and recombinant PSGL-1 with either 15 or 16 copies of the repeat exhibited indistinguishable binding to P-selectin. The PSGL-1 gene in the promyelocytic leukemia cell line HL-60 and the histiocytic lymphoma cell line U-937 are lacking the same copy of the 10-amino acid repeat in both alleles, yet these cell lines are unrelated. It is not clear how this loss of a repeat has occurred in the PSGL-1 genes in both cell lines, although this may have been caused by some form of polymorphic variation. Other examples of polymorphic and allelic variations in the repeat regions of mucin-type glycoproteins, such as the carcinoma-associated mucin episialin and human intestinal mucin, have been reported(36Ligtenberg M.J.L. Vos H.L. Gennissen A.M.C. Hilkens J. J. Biol. Chem. 1990; 265: 5573-5578Abstract Full Text PDF PubMed Google Scholar, 37Gum J.R. Byrd J.C. Hicks J.W. Toribara N.W. Lamport D.T.A. Kim Y.S. J. Biol. Chem. 1989; 264: 6480-6487Abstract Full Text PDF PubMed Google Scholar). Mapping of the PSGL-1 gene to chromosome 12 prompted a search of the genome data base for colocalized disease markers or genes. Although the genes for interferon-γ, protein phosphatase 1, and aldehyde dehydrogenase 2 also map to 12q24, the search did not reveal any readily apparent phenotypes that could be attributed to abnormal PSGL-1 expression(21Naylor S.L. Sakaguchi A.Y. Shows T.B. Law M.L. Goeddel D.V. Gray P.W. J. Exp. Med. 1983; 157: 1020-1027Crossref PubMed Scopus (133) Google Scholar, 38Barker H.M. Craig S.P. Spurr N.K. Cohen P.T.W. Biochim. Biophys. Acta. 1993; 1178: 228-233Crossref PubMed Scopus (59) Google Scholar, 39Hsu L.C. Yoshida A. Mohandas T. Am. J. Hum. Genet. 1986; 38: 641-648PubMed Google Scholar). The chromosomal loci for several other mucin-type glycoproteins and cell adhesion molecules have been mapped and appear to be clustered. In particular, the genes that encode the selectin family of leukocyte adhesion molecules are clustered on chromosome 1(13Watson M.L. Kingsmore S.F. Johnston G.I. Siegelman M.H. Le Beau M.M. Lemons R.S. Bora N.S. Howard T.A. Weissman I.L. McEver R.P. Seldin M.F. J. Exp. Med. 1990; 172: 263-272Crossref PubMed Scopus (113) Google Scholar). The gene for P-selectin (SELP) maps to chromosome 1q21-24, as do genes for L-selectin (SELL), E-selectin (SELE), and episialin(13Watson M.L. Kingsmore S.F. Johnston G.I. Siegelman M.H. Le Beau M.M. Lemons R.S. Bora N.S. Howard T.A. Weissman I.L. McEver R.P. Seldin M.F. J. Exp. Med. 1990; 172: 263-272Crossref PubMed Scopus (113) Google Scholar, 40Swallow D. Gendler S. Griffiths B. Kearney A. Povey S. Sheer D. Palmer R. Taylor-Papadimitriou J. Ann. Hum. Genet. 1987; 51: 289-294Crossref PubMed Scopus (101) Google Scholar). In addition, CD34, a ligand for L-selectin, also maps to chromosome 1q(14Molgaard H. Spurr N. Greaves M. Leukemia. 1989; 3: 773-778PubMed Google Scholar). The PSGL-1 gene, however, does not map to this cluster of selectin and selectin ligand genes, and neither is the PSGL-1 gene linked to the organizationally related genes for CD43 and GPIbα. The gene for human CD43 is located on chromosome 16p11.2, close to the genes encoding the α subunits of the leukocyte integrins LFA-1, Mac-1, and p150,95, and human GPIbα maps to chromosome 17p12-ter(29Cornwell R.D. Gollahon K.A. Hickstein D.D. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 4221-4225Crossref PubMed Scopus (41) Google Scholar, 41Pallant A. Eskenazi A. Mattei M.-G. Fournier R.E.K. Carlsson S.R. Fukuda M. Frelinger J.G. Proc. Natl. Acad. Sci. U. S. A. 1989; 86: 1328-1332Crossref PubMed Scopus (116) Google Scholar, 42Corbi A.L. Larson R.S. Kishimoto T.K. Springer T.A. Morton C.C. J. Exp. Med. 1988; 167: 1597-1607Crossref PubMed Scopus (70) Google Scholar). In summary, the organization of the PSGL-1 gene is strikingly similar to those of CD43 and platelet glycoprotein GPIbα. Each of these genes contains a single intron in the 5′-untranslated region and a long exon containing the complete coding region. The promoter regions lack the typical TATA and CAAT elements, and the genes are flanked by Alu sequences. Although PSGL-1, CD43, and GPIbα are all sialic acid-carrying O-glycosylated proteins, their amino acid sequences are unrelated and their genes do not appear to be linked in the genome. The table is compiled from Southern analysis of 29 mouse-human hybrids as described under “Experimental Procedures.” Scoring was determined by the presence (+) or absence (-) of a human PSGL-1 band on the blots compared with the presence or absence of human chromosomes in each hybrid. A 0% discordance indicates matched segregation of the probe with a human chromosome. We thank Gloria Vachino for preparing RNA samples, Dave Merberg for computer analysis of the DNA sequence, and Ron Kriz and Ravi Kumar for critical reading of the manuscript.