Promoter sequence and chromosomal organization of the genes encoding glycophorins A, B and E

Promoter sequence and chromosomal organization of the genes encoding glycophorins A, B and E

Gene, 95 (1990) 289-293 Elsevier 289 GENE 03766 P r o m o t e r sequence and chromosomal organization of the genes encoding glyeophorins As B and E...

501KB Sizes 0 Downloads 203 Views

Gene, 95 (1990) 289-293 Elsevier

289

GENE 03766

P r o m o t e r sequence and chromosomal organization of the genes encoding glyeophorins As B and E (Erythroid-specific regulation; cap site; c/s-acting sequences; blood-group antigens; sialoglycoproteins; recombinant DNA)

Main Vignal, Jacqueline London, C~ile Rahuel and Jean-Pierre Cartron INSERM U76, lnstitut National de Transfusion Sanguine, Paris (France) Received by G. Bernardi: 28 May 1990 Revised: 10 July 1990 Accepted: 16 July 1990

SUMMARY

The promoter and exon I sequences of the genes encoding erythrocyte glycophorins GPA, GPB and GPE were investigated in detail, both from a genomic clone sorted out of a human leukocyte library and from genomic clones obtained by polymerase chain reaction amplification of total genomic DNA from control individuals and from GPA and/or GPB deletion variarJts. The three exons 1 and upstream sequences were shown to be highly homologous with only a few point mutations that did not affect the potential cis-acting elements (CACCC, NF-EI and NF-E2) that are present in the same position ~thin the three genes. Moreover, these genes share the same transcription start point. Analysis of the exon l and promoter sequences together with the gene defects occurring in the GP variants indicate that unequal cross-overs between the three genes are responsible for deletiow, and the generation of hybrid gene structures in which the promoter of one gene is brought close to another gene of the family. On the basis of these studies, a model of the gene organization is proposed to explain the rearrangements occurring in the variants.

INTRODUCTION

GPA and GPB are the major surface GPs of human erythrocytes. The genes that encode these proteins have Correspondence to: Dr. J.-P. Cartron, INSERM U76, INTS, 6 rue Alexandre Cabanel, 75015 Paris (France) Tel. 33-(!)43067000, ext 330; Fax 33-(1)47347431. Abbreviations: aa, amino acid(s); bp, base pair(s); En(a-), genetic variant with no cell-surface synthesis of GPA; GP, glycophorin; GPA, GPB and GPE, glycophorins A, B and E, respectively; GPA, GPB and GPE, genes (DNA) encoding GPA, GPB and GPE, respectively; kb, kilobase(s) or 1000bp; Mi.V,genetic variant carrying a hybrid GPA-GPB gene structure; Mk, genetic variant with no cell.surface GPA and GPB; MMLV0 Moloney murine leukemia virus; MNSs, blood group antigens encoded by GPA (MN) and GPB (Ss) genes, respectively; NF-E, erythroid nuclear factor; nt, nucleotide(s); oligo, oligodeoxyribonucleotide; PCR, polymerase chain reaction; S-s-U-, genetic variant with no cell-surface synthesis of GPB; Stones (Sta), genetic variant carrying a hybrid OPB.GPA gene structure; tsp, transcription start point(s). 0378-1119190/$03.50O 1990Elsevier SciencePublishers B.V.(BiomedicalDivision)

been cloned and shown to e~hibit a high degree of homology, indicating that they have evolved from a common ancestral gene (for review see Cartron et el., 1990). Recently, a third GP-related gene, named GPE, has been characterized and its transcript isolated and sequenced (Vignal etal., 1990; Kudo and Fukuda, 1990). Various rearrangements of the GPA and GPB genes have been found by Southern-blot analysis of genomic DNA from rare variants presenting no cell-surface synthesis of GPA and/or GPB or producing unusual hybrid GP structures. For instance, deletions of the GPA or the GPB genes were found in homozygous En(a-) (Rahuel etal., 1988a,b) and S-s-U- (Huang etal., 1987) variants lacking GPA and GPB synthesis, respectively. Deletion of both the GPA and GPB genes was found in a homozygous Mk individual lacking synthesis of both proteins (Tate etal., 1989; Vignal oral., 1990). In other instances, unequal cross-overs between the GPA and GPB genes were found to be responsible for the presence of hybrid GP molecules in Stones

290 (Sta), Dantu and Mi.V variants (Huang and Blumenfeld, 1988; Huang et al., 1989; Vignal et al., 1989). Although the GP variants exhibit various alterations of the GPA and/or GPB genes, they apparently all carry the GPE gone. However, the limits ofthe deletions and rearrangements between the three genes remain uncertain. The exons 1 ofthe three GP genes are separated from the bulk of the genomic sequences by more than 30 kb (Kudo and Fukuda, 1990; Vignal et al., 1990), and no genomic clones overlapping exons I and 2 ofeither ofthe three genes have been isolated. It is therefore difficult to assign an exon 1 and its upstream region to a given GP gone. To address these issues, we have taken advantage of the deletions occurring in the GP variants to amplify and characterize the exon I and upstream sequences of selected genes by the PCR. This led us to characterize the promoter region of each gone and to propose a model for the organization of the GPA, GPB and GPE genes in control and variant individuals.

that the 431 clone, which we had previously attributed to the GPA gone by restriction fragment analysis (Rahuel et al., 1989), was in fact a GPB clone. From these findings, the gene deletion occurring in the GP variants could be defined more precisely. We found that the Mk genome, lacking GPA and GPB genes but containing the GPE gone, carries a GPA exon 1 together with its upstream sequence instead of the expected exon-I and upstream sequences of the GPE gone. In the En(a-) genome that lacks GPA, the exon-I and upstream regions found were those of GPA and GPE, whereas the S-s-U- genome that lacks GPB carries exon-1 and upstream sequences of GPA and GPB but not Of GPE. When the Mi.V genome that codes for a GPA/GPB hybrid protein was investigated, we found the GPA and GPE exon-1 and upstream sequences, and not the GPB and GPE ones as was postulated following Southern-blot analysis (Vignal et al., 1989). Accordingly, we concluded that a simple hybrid gene structure was present in this variant instead of the more complex GPB/GPA/GPB structure proposed previously (Rahuel et al., 1989).

EXPERIMENTAL AND DISCUSSION

(a) Isolation and sequence assignment of exon-I and upstream sequenees to the GPA, GPB and GP£ genes We have previously established that the human genome contains three GP.related genes, OPA, GPB and GPE (Vignal et al., 1990). In addition, Southern-blot analyses of GP variants presenting no cell-surface synthesis of GPA and/or GPB revealed the occurrence of gone deletions. To define the 5' limits of these deletions, we examined the exon-I region of each gene in control and variant individuals. Out of the three bands corresponding to the exon-I regions of the three genes visualized in some restriction patterns, one or two were absent together with the deletion of one or two OP genes, Our first assumption was that the missing restriction fragments were that of the exon 1 of the deleted gene(s). Since the exon-1 sequence of only one GP gone was known by the study of the ,131 clone (Rahuel et al., 1989), and because ofthe extensive homology existing between the three genes, we used the PCR to amplify and characterize the two other GP exon-I and upstream regions in the genome of GPA and/or GPB deletion variants. For each individual, amplification fragments were obtained from two different PCR experiments using ampfimers PI4 and PI8 or P19 and PIg flanking the exon-I structures (Fig. 1). These amplified products were cloned and sequenced and could be sorted out into three groups differing only by a few point mutations. Assignment ofthese clones to the GPA, GPB and GP£ genes could then be done by sequence comparison of their exon-I regions with those ofthe known GPA, GPB and GPE cDNAs. The respective sequences are shown in Fig. 1. It then turned out

(b) Organization of the GPA, GPB and GPF. genes and rearrangements in variants The GPA, GPB and GPE genes have been assigned to chromosome 4 (Cook et al., 1981; Rahuel etal., 1988a; Vignal et al., 1990). From these data and the above results of deletion analysis, we proposed that the three genes are organized in the order GPA.GPB-GPE, downstream from the transcription direction (Fig, 2). Accordingly, the breakpoints for the Mk deletion must reside on the 5' side between exons A-I and A-2 of GPA and on the 3' side between exons E.I and E-2 of GPE. This creates a GPA/GPE hybrid gene structure with exon A-I and upstream region of the GPA gone close to exons E-2 to E-4 of the GPE gone. Similarly, in the En(a-) and S-s-Ugenomes, equivalent deletions create GPA/GPB or GPB/GPE hybrid gene structures with the exon-1 and upstream region of one gene positioned close to the bulk of the next one (Fig. 2). The deletions observed in these variants may result from chromosome misalignment followed by an unequal cross-over within the first intron of two different GP genes. The model presented in Fig. 2 can be extended to the Mi.V variant which is known to carry a cell-surface GPA/GPB hybrid GP. Indeed, this variant carries a hybrid gone containing exons A-I to A-3 of GPA linked to exons B-3 to B-5 of GPB instead ofGPB/GPA/GPB gene structure proposed from previous Southern-blot analyses (Vignal et el., 1989). The mechanism leading to this structure is similar to that postulated above, except that the cross-over would take place within the introns separating exons A-3 to A-4 of GPA and exons B-2 to B-3 of GPB, These introns

291 gaaatgagaaggtccatggctccacaacagctacctca ................... ...................

B A E

-312 312 312

P19

B A E

-252 252

gcctggcacgtgccctggcctcagagattcacagtccagttctttgtccagttgggtggc ............................................................. P14

-192

B

tcctgtctaccaccttaccatgcccatttaacttatgcaaagttaatatcacaagtagcc

A

. . . . . . . . . . . . . . . . . . . . . . . . . .

c ......

g . . . . . . . . . . . . . . . . . . . . . . . . .

E

. . . . . . . . . . . . . . . . . . . . . . . . . .

c ......

g . . . . . . . . . . . . . . . . . . . . . . . . . . .

B A E

NFE-I NF a c c t gt t c c t t g c a g t g a a a a t t gt a c t t a c c a c t t t c a c a g c c c c a a g a t a t c c a t gta ....................... t ............... t ..................... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . t .....................

B

E-1 NFE-X tQtttattaacaggcgcttaacaacttgcatcatttaaaatgcctcccctgcctatcagc

A

a

............................................................. .............................................................

192

-132 132 132

-72

72 72

E

CACC NFE-I tgatgatggccgcaggaaggtgggcctggaagataacagctagtaggctaaggccagaca ........................................... c ......... t ....... ........... a ............................... c .................

B A

-192

-12 12 12

B

~tgacacttgcAGTTGTCTTTGGTAGTTTTTTTGCACTAACTTCAGGAACCAGCTCATGA

A E

................................ * ........................... ................................................ G .......

B A

TCTCAGG~TATGGARARATAATCTTTGTATTACTATTGTCAGgtaagtgattttattt ............................................................

109 109

g

............................................................

109

B

c a g t c t t g g t t c t q t tat at t g g g t a t g a g a t ........ ........

142 117

A

E

c

G---

49 49 49

117

P18 Fig, I. Comparison ofthe nt sequence ofthe exon-I and promoter regions ofthe GPA, GPB and GPE genes, The GPB ganomi¢ done ~31 was isolated from a human leukocyte library as described (Vignal et el. 1989), GPA and GPE clones were obtained by PCR with olign probes P19 and Pig or P14 and Pig on human total 8ChOraleDNA from control or GP-variant individuals deleted for GPA and/or GPB. PCRs were performed in a Perkin-Elmer Thermal DNA cycler as described before (Vignal et el,, 1990), Amplified fi'agments were purified on 1% egarose gels and sub¢loned into pUCI8 re©ton for sequencing, Restrietian enzymes, bacterial alkaline phnsphatase and pUC vectors were obtained from Appligene (Strasbourg, France).'14 DNA ligase and [~JSSldATP (400 Ci/mol) were from Amershem International (Bucks, U.K,), '17 sequencing kits were from Pharmaeia (Uppsala, Sweden~ Tbennus aquaticus (Taq) polymerase was from Perkin-Elmer {U,S.A.), Dashed lines represent the nt sequence homology existing between the GPA, GPB and OPE genes in the promoter and exon-1 regions, taking the ,%31GPB ganomic done as reference, The nt are numbered by taking the up of the three genes (downward arrowhead) as + I, with upstream nt indicated as negative numbers, The asterisk indicates a T nt that was only found in the 6PA clones amplified from Mk variant. Exons I are printed in bold capitals, The AT(} codon, used as translation start point for the GPA and GPB genes is overlined, Probes P19, P14 and Pig, used for the PCR amplifications are indicated as continuous lines. Putative cb.regulatory elements (NFE-I, CACC and NFE-2} on the 5' side of the Isp are indicated in bold-face characters and above the sequence. These tat sequences have been submitted to the GenBank dat~ base under accession numbers M37108, M37109 and M37110.

have been shown by restriction mapping and partial sequence analysis to be very homologous (Vignal et el,, 1989). All the variants described above represent the Leporetype result of unequal homologous recombinations such as those occurring in a ~-globin or haptoglobin gone families (Higgs et el., 1989; Maeda and Smithies, 1986). Hybrid gene structures of the anti-Lepore-type that could be the reciprocal exchange product of the Mi.V variant have also been documented in the Stones (St") variant carrying a GPB/GPA hybrid protein in addition to the normal GPA and GPB products (Blanchard et al., 1987; Huang et aL,

1989). Moreover, since unequal cross-overs within different homologous regions ofthe GPA, GPB and GPE gone family may occur, the generation of many other different Leporeand anti-Lepore-type variants could be predicted and identiffed in the near future among already known rare antigens of the MNSs blood group system. (e) Comparative analysis of the GPA, GPB and GPE promoter structures The GPA and GPB genes are both expressed in erythroid tissues only and appear at the same differentiation stage

292 A

"iV

"

S'

lS kid

$'--~)

lSkb/~ A.2toA-3



A*2 tO A.?

B

15kb

. . . . . . . . . . . . . . . .. .. ..

)

8,2 to B*S

e.310B.Sl



ff

s/~

E

1Skb

~.t)

E-2 to E.4

IS kbF~ S-2 'o [4

;-- 3' I

~-:i*

Fig. 2. Organization ofthe 6PA, GPB and 6PE genes on chromosome 4 as deduced by study of deletion variants. Fresh blood samples from common ~onors were collected on anticoagulant at the lnstitut National de Transfusion Sanguine (Paris, France). DNA from the En(a-) individual (G.W.) ofthe Finnish type was obtained from an Epstein-Barr virus (WES-2)-transformed cell line (gilt from Profs. C.G. Gahmberg and L.C. Anderson, Helsinki, Finland); blood from the homozyguus S-s-U donor (Fay.) was a gif~from Dr. M. Girard (C.T.S. Asni~res, France); the homozyguus Mk blood sample (R.S.) was a gilt from Dr. Y. Okubo (Osaka Red Cross Blood Center, Japan) and the homozyguus Mi.V blood sample (F.M.) was a gift from Dr. M.A. Blejchman (Canadian Red Cross Blood Transfusion Service, Hamilton, Ontario, Canada). The postulated deletions occurring in the En(a-), S-s-U-, Mk and Mi.V variants are indicated by dashed lines. This model is deduced from the present study (see text) and previous analysis by Southern-blot hybridization (Rahuel et al., 1988a,b; Vignal et al., 1989; 1990). For each gene, exon I is represented separately, since it is located at least 15 kb upstream from the hulk of the gene.

(Gal',mberg et al., 1978; Rahuel et al., 1989). To find out whether the GPE gene could have a similar transcription regulation as GPA and GPB, and to gain insight in the cis.acting sequences involved in the mechanisms leading to the specific expression of OPA and GPB, the promoter structures of the three genes were analyzed and compared. The GPE and tsp was investigated by primer extension analysis. A 32P.labeled (;PE.specific oligo probe, P25 (5'-AACACGAGCCATCGCCC-Y, located at nt positions 174-158 of the GPE cDNA, taking position + 1 as the equivalent of the first nt of the first codon of mature GPA or GPB), was hybridized to poly(A)÷RNA isolated from spleen erythroblasts and extended with MMLV reverse transcriptase as described previously (Rahuel et al., 1989). A major extension product 224 nt long was detected, suggesting a tsp similar to that previously found for GPA and GPB (Rahuel et al., 1989). This result was confirmed by PCR experiments using the P25 probe as the 3' amplimer and probes located on one or the other side of the postulated tsp as 5' amplimers (not shown). Amplification products were detected only with the 5' amplimers located downstream from the tsp determined by the primer extension analysis. Sequence comparison of the regions up to 330 nt upstream from the tsp of the three genes revealed great homology, with only a few point mutations (Fig, 1). Although transcription is initiated precisely at a single tsp for the three genes, with a sequence surrounding these sites following the weak consensus motif 5'-YYCAYYYYY-3' (Corden et al., 1980), the sequences upstream from the three tsp are devoid of TATA or CAAT boxes as already observed in the analysis ofthe GPB ~.31 clone (Rahuel et al.,

1989). Further examination of (;P,4 and GPE sequences revealed the existence of several consensus motifs present at the same location as in the GPB gene and often found in the promoter or regulatory elements of other erythroidspecific genes. First, at nt position -50 on the complementary strand lies a sequence 5'-CCCACC-3', which matches the consensus CAC-box sequence (Dierks et al., 1983) present in many erythroid-specific gene promoters, although it is bound by ubiquitous factors. Second, two perfect NFE-I consensus sequences (also named OF-l, Eryfl or EF.1) at nt positions -38 and -75 have been characterized and shown to be bound by an erythroidspecific trans-acting factor in the promoter and enhancer regions ofgiobin (De Boer et al., 1988; Evans et al., 1988; Plumb et al., 1989) and nongiobin (Mignotte et al., 1989) erythroid.specific genes. Two other NFE-I motifs with one mismatch are present at nt -130 and -143. Third, at nt -8 lies a sequence which displays 70~0 homology with the NFE-2 binding site found in the porphobilinogene deaminase .erythroid promoter (Mignotte et al., 1989). None of the mutations observed between the promoter regions of the three genes affect any of these potential regulatory elements (Fig. 1). The effect ofthe proximal promoter regions on transcription regulation should therefore be equivalent in the three genes, Comparison of the protein synthesis level in the deletion variants carrying hybrid gene structures should give some insight on the respective influence of the 5'-untranslated regions of the three genes. In control individuals, there are about ten times more cellsurface GPA than GPB molecules (Merry et al., 1986). Although the 6PB gene is under the control of the GPA

293 promoter in the GPA/GPB gene structure present in the En(a-) variant, the level of GPB protein on the cell surface remains similar to that of control individuals. Therefore, the region 5' upstream from the GPA tsp does not induce a higher GPB expression level in this variant. This suggests that other, still unidentified, regions of the GPA or GPB genes are responsible for the difference in ceil-surface synthesis of the two proteins in normal individuals. The effect of the GPA or GPB promoters on the GPE expression level is not known, as the GPE protein has not yet been detected in red cells and as erythroid tissues from the GP variants are not available for Northern-blot analysis. Further structural and functional analyses of the three promoters will be necessary to understand the molecular basis for the differential expression of the three genes. (d) Conclusions The study of the exon-I and promoter regions of the GPA, GPB and GPE genes in control individuals and GP variants by PCR analysis lead us to several conclusions: (1) the order of the GP genes on chromosome 4 is tsp-OPAGPB-GPE; (2)the Mk, En(a-), S-s-U- and Mi.V GP variants carry hybrid gene structures as the result of unequal cross-overs within the GPA, GPB and OP£ gene family; (3) the promoter regions of the three genes are highly homologous and differ only by a few point mutations in the sequence over 330 bp upstream from the identical tsp; (4) none of the mutations occur in one of the potential CACC, NFE-I or NFE-2 regulatory elements, indicating that all three promoters are potentially functional.

REFERENCES Bianchard D., Dahr W., Beyrcuther, K., Moulds, J. and Cortron, J.-P.: Hybrid gly¢ophorins from human erythrocyte membranes. Isolation and complete structural analysis of the novel sialoglycoproteinI~om St(a+ ) red cells, Eur. J. Biochem. 167 (1987) 361-366. Cortron, J.-P., Colin, Y., Kudo, S. and Fukuda, M.: Molecular genetics ofhuman erythrocyte sialnglycoproteinsglycophorins A, B, C and D. In Harris, J.R. (Ed.), Blood Cell Biochemistry,VoL I, Erythroid Ceils. Plenum, New York, 1990, pp, 299-335. Cook, PJ.L.o Lindenbaum, R.H., Salonen, R., de la Chapelle, A., Daker, M.G., Buekton, K.E., Noades, J.E. and Tippett, P.: The MNSs blood [groups of families with chromosome 4 rearrangements. Ann. Hum. Genet. 45 (1981) 39-47. Corden, J., Wasylyk, B., Buchwalder, A., Kedenger, C. and Chambon, P.: Promoter sequence of eukaryotic protein-coding gene. Science 209 (1980) 1406-1414. De Boer, E., Anmniou, M., Mignotte, V., Wall, L. and Grosveld, F.: The human /?.globin promoter; nuclear protein factors and erythroid specific induction of transcription. EMBO J. 7 (1988) 4203-4212. Dierks, P., Van Oogen, A., Cochran, M.D., Dobkin, C., Raiser, ~. and Weissmann,C.: Three regions upstream from the cap s~teare required for efficient and accurate transcription of the rabbit p.giobin gone in mouse 3T6 cells. Cell 32 (1983) 695-705. Evans, T., Reitman, M. and Felsenfeld, G.: An erythrecyte specil~cDNA

binding factor recognizes a regulatory sequence common to all chicken globin genes. Proe, Natl. Aead. Sci. USA 85 (1988) 5976-5980. Gahmberg, C.G., Jokinen, M. and Anderson, I-C.: Expression of the major sialoglycoprotein (glycophorin) on erythroid cells in human bone marrow. Blood 52 (1978) 379-387. Higgs, D.R,, Viekers,M.A., Wilkie,A.O.M., Pretorius,I.-M.,|arman, A.P. and WeatheruH DJ.: A review of the molecular genetics of the haman ~t-globingene cluster. Blood 73 (1989) 1081-1104. Huang, C.-H. and Blumenfeld. O.O.: Characterization of a genomic hybrid specifyingthe haman erythro~e antigen Dantu: Dantu gene is duplicated and finked to a &glyenphoringene deletion. Prec. Natl. Aced. Sci. USA 8S (1988) 9640-9644. Huang, C.-H., Johe, K., Moulds, JJ., Siebert, P.D., Fukuda, M. and Blumenfeld, O.O.: &Giycophorin (glycophorin B) gene deletion in two individuals homozygous for the S-s-U- blood group phenotype. Blood 70 (1987) 1830-193S. Huang, C.-H., Guizzo, M,L., Kikuchi, M. and Blamenfeld, O.O.: Molecular genetic analysis of a hybrid gene encoding St" glycophorin of the erythrocyte membrane. Blood 74 (1989) 836-843. Kudo, S. and Fukuda, M.: Identification of a novel human glycophorin, giyenphorin E, by isolation of genomic clones and complementary DNA clones utilizing polymerase chain reaction. J. Biol. Chum. 265 (1990) 1102-1! 10. Macda, N. and Smithies, O.: The evolution of multigene families:human haptnglobin genes. Annu. Ray. Genet. 20 (1986) 81-108, Merry. A.H., Hodson, C., Thomson, E., Mallinson, O. and Anstee, DJ.: The use of monoclonal antibodies to quantify the levels of sialogiycoproteins 0t and 6 and variant sialnglycoprotains in haman erythrocyte membranes. Biochem. J. 233 (1986) 93-98. Mignotte, V., Wall, L., De Boer, E., Grosveld, F. and Rom6o, P.-H.: Two tissue specific factors bind the crythroid promoter of the haman porphobilinogen deaminase gene. Nucleic Acids Res. 17 (1989) 37-72. Plumb, M., Frampton, J., Wainwright, H., Walker, M., Macleod, K., Goodwin, G. and Harrison, P.: GATAAG: a c/s.control region binding an erythroid speeifl©nuclear factor with a role in giobin and non-globin expression. Nucleic Acids Res. 17 (1989) 73-92. Rahuel, C., London, J., d'Auriol, L., Mallei, M.G., Toumamille, C., Skrzynia, C., Lebouc, Y., Gaiibert, F. and Cartron, J.-P.: Characterization of eDNA clones for human giycophorin A. Use for gene localization and for analysis of normal and glycophorin.A.defleient (Finnish type) genomic DNA. Eur. J. Biochem. 172 (198ga) 147-153. Rahuel, C., London, J., Vignal, A., Cherif-Zahar, B., Colin, Y., Siebert, P.D., Fukudu, M. and Cartron, J.-P.: Alteration of the genes for giycophorin A and B in glycophorin.A-deficientindividuals. Eur. J. Biochem. 177 (1988b) 605-614. Rahuel,C., Vignal, A., London, J., Hamel, S., Rom6o, P..H.,Colin,Y. and Cartron, J.-P.: Structure of the 5' flankingregion of the gene encoding human glycophorin A and analysis of its multiple transcripts. Gene 85 (1989) 471-477. Tare, C.G., Tanner, MJ.A., Judson, P,A. and Anstee, DJ.: Studies on human red-cell membrane giycophorin A and giycophorin B genesin giycophorin deficient individuals. Biochem J. 263 (1989) 993-996. Vignal, A., Rahuel, C., El Maliki, B., Le Van Kim, C., London, J., Blanchard, D., d'Auriol, L., Galibert, F., Blejchman, M.A. and Cartron, J.-P.: Molecular analysisofgiycophorin A and B gene structure and expression in homozygous lOJihenberger class V (Mi.V) human erythrocytes. Eur. J. Biochem. 184 (1989) 337-344. Vignal, A., Rahuel, C., London, J., Cherif-Zahar, B., Schaff, S., Hattab, C., Okubo, Y. and Cartron, 3.-P.: A novel member of the glycophorin A and B family. Molecular cloning and expression. Ear. J. Biocbem. 191 (1990) 619-625.