Proc. Natl. Acad. Sci. USAVol. 85, pp. 1615-1619, March 1988Immunology

Exon-intron organization and sequence comparison of human andmurine Til (CD2) genes

(genomic maps/molecular cloning/T-cell receptors/DNA sequencing)

DON J. DIAMOND*t, LINDA K. CLAYTON*t, PETER H. SAYRE*, AND ELLIS L. REINHERZ*t*Laboratory of Immunobiology, Dana-Farber Cancer Institute and Departments of tPathology and tMedicine, Harvard Medical School, 44 Binney Street,Boston, MA 02115

Communicated by Barry R. Bloom, November 2, 1987 (received for review September 23, 1987)

ABSTRACT Genomic DNA clones containing the humanand murine genes coding for the 50-kDa T11 (CD2) T-cellsurface glycoprotein were characterized. The human T11 geneis =12 kilobases long and comprised of five exons. A leaderexon (L) contains the 5'-untranslated region and most of thenucleotides defining the signal peptide [amino acids (aa) -24to -5]. Two exons encode the extracellular segment; exon Exiis 321 base pairs (bp) long and codes for four residues of theleader peptide and aa 1-103 of the mature protein, and exonEx2 is 231 bp long and encodes aa 104-180. Exon TM is 123 bplong and codes for the single transmembrane region of themolecule (aa 181-221). Exon C is a large 765-bp exon encodingvirtually the entire cytoplas'mic domain (aa 222-327) and the3'-untranslated region. The murine T11 gene has a similarorganization with exon-ntron boundaries essentially identicalto the human gene. Substantial conservation of nucleotidesequences between species in both 5'- and 3'-gene flankingregions equivalent to that among homologous exons suggeststhat murine and human genes may be regulated in a similarfashion. The probable relationship of the individual T11 exonsto functional and structural protein domains is discussed.

mic domain rich in prolines and basic residues. cDNAsencoding the murine and rat equivalents of T11 have beencloned and shown to define a homologous structure bearinga 50% overall identity at the amino acid level (12-14). Incontrast to immunoglobulins whose variable and constantregion domains consist of antiparallel (3-sheets, protein mod-eling of murine and human T11 external segments indicatesthat these structures most likely belong to an a/,B-proteinfolding class (12). Furthermore, the cytoplasmic region ofT11 in each species is predicted to have a nonglobularconformation and is sufficiently elongated to allow for po-tential interactions with multiple other intracellular proteins.It is thus likely that the cytoplasmic region subserves a signaltransduction function.Here we report on the isolation and characterization of

human and murine T11 genes. DNA sequence analysisdemonstrates that the human and murine exon-intron orga-nization is virtually identical.§ Of note, the extracellularsegment of the mature T11 protein is encoded by two exons,whereas almost the entire intracellular segment and 3'-untranslated region is encoded on a single exon.

The 50-kDa T11 (CD2) surface glycoprotein plays a majorrole in T-lymphocyte function (1-5). Monoclonal antibodiesdirected against an epitope on the external segment of thehuman T11 molecule (T111) block T-lymphocyte activation,including that mediated through the T-cell receptor forantigen and major histocompatibility complex (Ti-T3) (6). Incontrast, antibodies defining two other extracellular segmentepitopes (T112 and T113) in concert give rise to antigen-independent human T-cell activation (1-3). Thus, interactionof specific monoclonal antibodies with the T11 moleculeproduces either profound antagonistic or agonistic effects onT lymphocytes in vitro. The ability of anti-T111 monoclonalantibodies to inhibit T-lymphocyte activation is predicated,at least in part, on their capacity to abrogate T11-mediatedcell-cell contact (7). In fact, the spontaneous interactionbetween human T lymphocytes and sheep erythrocytes isbut one manifestation of the role of T11 in facilitatingcell-cell adhesion (8).To more precisely define the biochemistry of this func-

tionally important molecule, we and others have elucidatedthe complete primary structure of the human T11 moleculeby protein microsequencing and cDNA cloning (9) or byantibody selection of expressed cDNA clones (10, 11). Thefollowing three segments of the protein have been defined:(i) an extracellular segment comprising more than half of themolecule and bearing only limited homology to members ofthe immunoglobulin gene superfamily including the T4 T-cellsurface protein (9, 10); (ii) a single hydrophobic transmem-brane segment; and (iii) a lengthy 117-amino acid cytoplas-

MATERIALS AND METHODSGenomic Library Screening and Restriction Mapping. A

human peripheral blood lymphocyte genomic library (kindlyprovided by S. Orkin, Children's Hospital, Boston, MA)containing DNA partially digested with Sau3aI and ligatedinto the EMBL3A vector (15) was screened with a 335-base-pair (bp) 5' fragment of the human T11 cDNA (PB2), endingat the unique EcoRV site (positions 8-335), and with a401-bp 3' fragment ending at the second internal Taq I site(positions 617-1017) (9). Genomic fragments containing re-gions that hybridized to PB2 cDNA were cloned into pUC18for further restriction analysis. Exons were identified withinthe human Til gene through hybridization with kinase-treated 17-base oligonucleotides whose sequences were de-rived from the human cDNA. All mapping distances areaccurate to within 100 bp. Oligonucleotide synthesis em-ployed standard cyanoethyl phosphoramidite chemistry onan Applied Biosystems model 381A (Applied Biosystems,Foster City, CA).BALB/c mouse liver DNA was partially digested with

Mbo I, and the size-selected 15- to 20-kilobase (kb) frag-ments were ligated into the EMBL3B vector and propagatedin the host bacterium LE392 to produce a murine genomiclibrary. The 819-bp 5' EcoRI fragment of the murine cDNAXB2 (12) was used to screen the library for murine T11

Abbreviation: aa, amino acid(s).§These sequences reported in this paper are being deposited in theEMBL/GenBank data base (Bolt, Beranek, and Newman Labora-tories, Cambridge, MA, and Eur. Mol. Biol. Lab., Heidelberg)(accession no. J03622 for the human T11 gene and J03623 for themurine T11 gene).

1615

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

Janu

ary

24, 2

021

Proc. Natl. Acad. Sci. USA 85 (1988)

sequences. Restriction analysis of the phage DNA andsubcloning of hybridizing fragments was performed asabove.

Sequence Analysis. Fragments obtained by electroelutionfrom agarose gels of bacteriophage A DNA digests weresubcloned into bacteriophage M13mpl8 or -mpl9 sequenc-ing vectors by standard procedures (16). Sequence analysiswas performed on these clones by the dideoxy chain-termination procedure of Sanger et al. (17) with either[a-32P]dATP or adenosine 5'[a-35S]thiotriphosphate (dATP[35S]) as a single radioactive nucleotide. In some cases,sequence information was obtained from fragments clonedinto pUC18 by utilizing avian myeloblastosis virus reversetranscriptase (New England Nuclear) on double-strandedDNA as described by Chen and Seeburg (18). All 5'- and3'-flanking sequences and coding regions were sequenced onboth strands.

RESULTS AND DISCUSSIONIsolation of Recombinant Phage Containing the Human and

Mouse Tl1 (CD2) Gene Segments. A genomic library con-structed in the A vector EMBL3A (15) by partial Sau3aIrestriction endonuclease digestion of human peripheralblood lymphocyte DNA was screened with 5'- and 3'-specific fragments from the human T11 cDNA. Three typesof phage were recovered from the library that encompassedall of the coding segments from the known cDNA sequence(9) as well as extensive 5'- and 3'-flanking regions (Fig. 1).The 5' fragment detected two classes of phage, such asABT2D2 and ABT2; whereas the 3' fragment detected ABT2and ABT1 (see Fig. 1). The genomic copy of the murinehomologue of T11 was obtained in a similar fashion byprobing a murine genomic partial library constructed in the Avector EMBL3B with cDNA fragments (12). Once again,three overlapping phage inserts were obtained that corre-sponded to the complete coding sequence of the cDNA aswell as extensive 5'- and 3'-flanking regions (Fig. 1). The 5'portion of the gene was contained within AYH2, whereasAYH16 and A8B contained the 3' segments (Fig. 1).

Structural Comparisons of the Human and Murine Homo-logue of the Tl1 Gene. Southern blotting experiments witholigonucleotide probes on the isolated recombinant phagefrom the human and mouse genomic libraries revealed that

Human T-1 1

L Ex1 Ex2 TM CIm m I I

SR Pv SBg At B Bg S Bg B SmN Sm S XhH S B

XBT2- I

I - AXBT2D2ABT 1 - I

Murine T-1 1

L Ex1 Ex2 TM CI. I

B E EHX E P X E Hp HIp X P P XE X H

1 4 I W 1 1i Xi W l

1- XYH2

AYH16 -CA8B-

1 Kb

the coding segments of T11 overlapped the derived phageclones such that the 5' and 3' ends of the correspondingcDNA were located on separate phage inserts (Fig. 1).Results of restriction analysis and DNA sequencing ofhuman and murine T11 genes indicate that, in each case, fiveexons are encoded within a gene -12 kb long (Fig. 1). Theexons are separated by introns that vary from 0.1-3.9 kblong. Both the sizes and positions of the exons and intronsare extremely similar between the human and murine T11genes (see Table 1). This conservation of the length of theindividual exons and introns and their relative positionssuggests that both genes evolved similarly to encode pro-teins with homologous function. Comparison of the murineand human genomic copies of interleukin 2 (19), the 6subunit of T3 (20), the a subunit of the T-cell receptor Ti (21,22), and the f8 subunit of the T-cell receptor Ti (23, 24) alsorevealed similar exon-intron organizations. These genespresumably have similar functions in both species. TheT-cell lineage restricted distribution of the human and mu-rine T11 mRNA, shown in RNA gel blot analysis (9, 10),provides further evidence for the similarity of function ofthese two genes. Chromosomal locations for the murine andhuman T11 genes (ref. 13 and L.K.C. and E.L.R., unpub-lished results) have been shown to be on chromosomes 3 and1, respectively. These chromosomes are known to be syn-tenic, which further implies a common origin and evolutionfor the murine and human T11 genes.

Nucleotide Sequence Analysis. By using a series of 17-baseoligonucleotide primers generated from the sequence of themurine and human T11 cDNAs, the exon-intron structure ofthe corresponding T11 genes was determined (Fig. 2). Thenucleic acid sequence of each exon bounded by its canonical5'- and 3'-splice junctions (26) is shown in Fig. 2. There is ahigh degree of similarity between the boundaries of therespective human and murine exons, such that in every casea precisely homologous codon is split between the first andsecond nucleotide in the corresponding exon boundary (seeFig. 2). For the T-cell receptor subunit genes (21-24),immunoglobulin genes (27), and T3 6-subunit genes (20), acodon triplet is also split between the first and secondnucleotides by an intron junction. However, introns of genessuch as y interferon (28), human (29), and mouse interleukin2 (19) split exons between codons. Table 1 shows that acodon yielding an identical amino acid is present (e.g.,

FIG. 1. Partial restriction maps of the hu-man (Upper) and murine (Lower) T11 (CD2)genes. Cleavage sites of restriction enzymes areas follows: A, Ava II; B, BamHI; Bg, Bgl II; E,EcoRI; H, HindIII; Hp, Hpa I; N, Nde I; P, PstI; Pv, Pvu II; R, EcoRV; S, Sac I; Sm, Sma I;X, Xba I; and Xh, Xho I. The dark boxes abovethe restriction maps represent exons L, Ex1,Ex2, TM, and C. The exons correspond to thefollowing positions in the human P131 cDNAsequence (9): bp 1-84 (exon L), bp 85-405(exon Exi), bp 406-636 (exon Ex2), bp 637-759(exon TM), and bp 760-1524 (exon C). Theoriginal EMBL3 recombinant phage used tosubclone the respective genes are shown asopen boxes under the restriction map.

1616 Immunology: Diamond et al.

Dow

nloa

ded

by g

uest

on

Janu

ary

24, 2

021

Immunology: Diamond et al. Proc. Natl. Acad. Sci. USA 85 (1988) 1617

-h A( ( T(.CTTAA(.CTCTCG(;GT:GTCT:GACTCCACCAGTCTCACTTCAGTTCCTTTTGCATTAAGAGCTCAG

m ACGTGC-t-TTCA(;CTTCCTGCeGTCTCGCTTCTGCCACCCTCACCACAG-TCC--TGACAGAAGMCTCAG

-24METSerPheProCysLysPheValAlaSerPheLeuLeuIle

AATCA-AAAGAGGAAACCAACCCCTAACATGAGCMCCATGTAAM GTAGCCAGCTTCCTTCTGATT

ACTCACCCCTGGGAAAAGAACTCTAAAGATGA ------AATGTAAATTCCTGCGTAGCTTCTITCTGCTCMETL ysCysLysPheLeuGlySerPhePheLeuLeu

-10

PheAsnValSerSerLysGTTCAATGTTTCTTCCAAAGGTAAG

TTCAGCCTTTCCGGCAAAGGTAAG ............ ..................................

PheSerLeuSerGlyLysG +1lyAlaValSerLysG

..................................... CTCTTTTGCTTTTTATAGGTGCAGTCTCCAAMG

....................................... TCTTACTTTTTTACAGGGGCGGACTGCAGAGlyAlaAspCysArgA

10 20luI leThrAsnAlaLeuGluThrTrpGlyAlaLeuGlyGlnAspIleAsnLeuAspIleProSerPheGlAGATTACGAATGCCTTGCAAACCTGGGGTGCCTTGGGTCAGGACATCAACTTGGACATTCCTAGTTTTCA

ACAATGAGA - -- CC ------ATCTGGGGTGTCTTGGGTCATGGCATCACCCTGAACATCCCCAACTTTCAspAsnGluT hr IleTrpGlyValLeuGlyHisGlyIleThrLeuAsnIleProAsnPheGl

30 40nMetSerAspAspIleAspAspIleLysTrpCluLysThrSerAspLysLysLysIleAlaGlnPheArgAATGAGTGATGATATTGACCATATAAAATGGGAAAAAACTTCAGACAAGAAAAAGATTGCACAATTCAGA

AATCACTCATGATATTGATGAGGTGCGATGGGTAAGGA ---GGGGC -ACCCTGGTCGCAGAGTTTAAAnMetThrAspAspIleAspGluValArgTrpValArgA rgGly ThrLeuValAlaGluPheLys

Exl 50 60 70LysGluLysGluThrPheLysGluLysAspThrTyrLysLeuPheLysAsnGlyThrLeuLysIleLysHAAAGAGAAAGAGACTTTCAAGGAAAAAGATACATATAAGCTATTTAMTGGAACTCTGAAAATTAAGC

AGGAAGAAGCCACCTTTTTTGATATCAGAAACGTATGAGGTCTTAGCAAACGGATCCCTGAAGATAAAGAArgLysLysProProPheLeuI leSerGluThrTyrGluValLeuAlaAsnGlySerLeuLysIleLysL

80 90isL euLysThrAspAspGlnAspIleTyrLysValSerIleTyrAspThrLysGlyLysAsnValLeATC --- TCAAGACCGATGATCAGCATATCTACAAGGTATCAATATATGATACAAAAGGAAAAAATGTGTT

ACCCCATGATGAGAAACGACAGTGGCACCTATAATGTAATGGTGTATGGCACAAATGGGATGACTAGGCTysProMetMetArgAsnAspSerGlyThrThrAsnValMetValTyrGlyThrAsnGlyMetThrArgLe

100uGluLysIlePheAspLeuLysIleGInGGCAAAAAATATTTGATTTGAAGATTCAAGGTAAG .........................

GGAGAAGGACCTGGACGTGAGGATTCTGGGTAAG...................................._ uoluLysAspLeuAspValArgI leLeuG

110luArgValSerLysProLysIleSerTrpThrCysIleAsn

.CTTACTTTCTTTTTAGAGAGGGTCTCAAAACCAAAGATCTCCTGGACTTGTATCAAC

................. TCTTTCTTTTAGAGAGGGTCTCAAAGCCCGTGATCCACTGGGAATGCCCCAACluArgValSerLysProValIleHisTrpGluCysProAsn

120 130 140ThrThrLeuThrCysGluValMetAsnGlyThrAspProGluLeuAsnLeuTyrGlnAspGlyLysHisLACAACCCTGACCTGTGAGGTAATGAATGGAACTGACCCCGAATTAAACCTGTATCAAGATGGGAAACATC

ACAACCCTGACCTGTGCGGTCTTGCAAGGAACAGATTTTGAACTGAAGCTGTATCAAGGGGAAACACTACThrThrLeuThrCysAlaValLeuGlnGlyThrAspPheGluLeuLysLeuTyrGlnGlyGluThrLeuL

150 160e uLys LeuSerGlnArgValIleThrHisLysTrpThrThrSerLeuSerAlaLysPheLysCysThT-AAAA- -CTTTCTCAGAGGGTCATCACACACAAGTGGACCACCAGCCTGAGTGCAAAATTCAAGTGCAC

TCAATAGTCTCCCCCAGAAGAACATGAGTTACCAGTGGACCA- -A-CCTGAGCGCACCATTCAAGTGTGAeuAsnSerLeuProGlnLysAsnMetSerTyrGlnTrpThrA s nLeuSerAlaProPheLysCysGl

170 180rAlaGlyAsnLysValSerLysGluSerSerValGluProValSerCysProGAGCAGGGAACAAAGTCAGCAAGGAATCCAGTGTCGAGCCTGTCAGCTGTCCAGGTGCG ............

************ * * ** ************* *

GGCGATAAACCCGGTCAGCAACGAGTCTAAGACGCGAAGTGTTAACTGTCCAGGTAAG ............

uAlaI leAsnProValSerLysGluSerLysThrGluValValAsnCysProG;

190luLysGlyLeuAspIleTyrLeuIleI leGlyIleCysCly

.CTCTCTTCCCTTTGCAGAGAAAGGTCTGGACATCTATCTCATCATTGGCATATGTGGA

.TTTTCTTCTTGCAGAGAAAGGTCTGTCCTTCTATGTCACAGTCGGGGTCGGTGCAluLysGlyLeuSerPheTyrValThrValGlyValGlyAla

200 210GlyGlySerLeuLeuMetValPheValAlaLeuLeuValPheTyrIleThrLysArgLysLysClnArgSGGAGGCACCCTCTTGATGGTCTTTGTGCCACTGCTCGTTTTCTATATCACCAAAGGAAAAAAACAGAGGA***** ******************** * ********** *********** ** * TMGGAGGACTCCTCTtGGTCCTCTTGGTGGCGCTTTTTATTTTCTGTATCTGCAAGAGGAGAAAACGGAACAGlyGlyLeuLeuLeuValLeueuValAlaLeuPheIlePheCysIleCysLysAr8ArgLysArSAsrLA

220erArgAr&AsnAGTCGGAGAAATGGTAAG

GGAGGAGAAAAGGTAAG ................................................rgArgArgLysA

............................................................TATTGAGCTT

............................................................TTAAAACGTC

230 240

TTGCCATTATAGATGAAGAGCTGGAAATAAAAGCTTCCAGAACAAGCACTGTGGAAAGGGGCCCCAAGCCspGluGluLeuGluIleLysAlaSerArgThrSerThrValGluArgGlyProLysPr

250 260oHisGlnIleProAlaSerThrProGlnAsnProAlaThrSerGlnHisProProProProProGlyHisCCACCAAATTCCAGCTTCAACCCCTCAGAATCCAGCAACTTCCCAACATCCTCCTCCACCACCTGGTCAT

GCACTCAACCCCAGCCCGCAGCAGCGCAGAATTCAGTGGCGCTCCAA - -- GCTCCTCCTCCACCTGGCCAToHisSerThrProAlaAlaAlaAlaGlnAsnSerValAlaLeuGln AlaProProProProGlyHis

270 280ArgSerGlnAlaProSerHisArgProProProProGlyHisArgValGlnHisGlnProGlnLysArgPCGTTCCCAGGCACCTAGTCATCGTCCCCCGCCTCCTGGACACCGTGTTCAGCACCAGCCTCAGAAGAGGC

CACCTCCAGACACCTGGCCATCGTCCCTTGCCTCCACGCCACCGTACCCGTGAGCACCAGCAGAAGAAGAHisLeuGlnThrProGlyHisArgProLeuProProClyHisArgThrArgGluHisGlnGlnLysLysA

290 300 310roProAlaProSerGlyThrGlnValHisGlnGlnLysGlyProProLeuProArgProArgValGlnPrCTCCTGCTCCGTCGGGCACACAAGTTCACCAGCAGAAAGGCCCGCCCCTCCCCAGACCTCGAGTTCAGCC

GACCTCCTCCATCAGGCACACACATTCACCAGCAGAAAGGCCCTCCTTTACCCAGACCCCGAGTTCACCCrgProProProSerGlyThrGlnlleHisGlnGlnLysGlyProProLeuProArgProArgVa1G1nPr

320oLysProProHisGlyAlaAlaGluAsnSerLeuSerProSerSerAsnAAAACCTCCCCATGGGGCAGCAGMAACTCATTGTCCCCTTCCTCTAATTAAAA.AGATAGAAACTGTCT

AAAACCTCCCTGTGGGAGTGGAGATGGTGTTTCACTGCCGCCCCCTAATT-AAGAAGGCAGAGTTCGTCAoLysProProCysGlySerGlyAspGlyValSerLeuProProProAsn

TTTCCA MAAGCTGTGTGGATTTAT- -CTTCTTCAGGTG

TGTGTGCAGAACATTGTCACCTCCTGAGGCTGTGGGCCACAGCCACCTCTGCATCTTCGAACTCAGCCATGTGGTCAACATCTGGAGTTTTTGGTCTCCTCAGAGAGCTCCATCACACCAGTAAGGAGAAGCAATATAAGTGTGATTGCAAGAATGGTAGAGGACCGAGCACAGAAATCTTAGAGATTTCTTGTCCCCTCTCAGGTCATGTGTAGATGCGATAAATCAAGTGATTGGTGTGCCTCGGTCTCACTACAAGCAGCCTATCTGCTTMGAGACTCTGGAGTTTCTTATGTGCCCTGGTGGACACTTGCCCACCATCCTGTGAGTAAAAGTGAMIAMGCTTTGACTAG

C

FIG. 2. Alignment of the individual exons from the human (h) and murine (m) T11 gene sequences. The canonical 5'- and 3'-splice sites arealso shown at the ends of the respective exons. The leader exon (exon L) begins with the putative mRNA transcription initiation site in thehuman sequence. Numbering refers to codon positions with respect to the initial amino acid of the mature protein (aa + 1). The canonicalpolyadenylylation signal sequence AATAAA is underlined. The exon sequences from mouse and human were aligned with the LOCALalgorithm (25) with gaps introduced (dashed lines) as a result of the alignment. Nucleotide identities are indicated by asterisks. Exons L, Exl,Ex2, TM, and C are described in the legend to Fig. 1.

glycine between exons L and Exl) at each of the junctions incomparison of the human and murine genes. The nucleotidehomology between species ranges from 62.1% to 71.5% incomparison of individual exons of the T11 genes (Table 1)(overall homology 67.8%).Exons often encode functional domains within a given

protein. The exon-intron organization of T11 is repre-sentative of a type I integral membrane protein (30) withexon L encoding the 5'-untranslated sequence and the hy-drophobic leader peptide and with exon TM encoding thesingle hydrophobic membrane-spanning segment. In addi-tion, exons Exi and Ex2 encode the hydrophilic extracellu-lar T11 segment, and exon C encodes virtually the entirecytoplasmic segment and 3'-untranslated sequence. Of note,11 putative cytoplasmic amino acid residues are also en-coded within exon TM (Fig. 2). However, these are predom-inantly charged amino acid residues that are presumably

needed to anchor T11 in the plasma membrane. Nucleotidehomology searches employing each of the human T11 exonsseparately failed to show any significant identities withknown genes other than T11 in GenBank.¶ This findingindicates that each of the T11 exons is unusual in nature anddistinct from immunoglobulin genes, T-cell receptor sub-units, or other members of the gene superfamily.

It is worthy of comment that the T11 gene organizationdiffers substantially from that of the 8 subunit of T3 (20)whose gene product is also selectively expressed in T-lineage cells and is thought to be involved in signal trans-duction. Unlike T11, the T3 6-subunit gene is comprised of asingle exon that encodes the extracellular segment of the

SNational Institutes of Health (1987) Genetic Sequence Databank:GenBank (Research Systems Div., Bolt, Beranek, and Newman,Cambridge, MA), Tape Release 50.

Ex2

Dow

nloa

ded

by g

uest

on

Janu

ary

24, 2

021

1618 Immunology: Diamond et al.

Table 1. Comparison of human and murine T11 exons

Exon Intron aa %Exon Source size, bp length, kb interrupted homologyL H 157* 0.77 Gly

M 149* 0.108 Gly 70.5Exi H 321 3.3 Glu

M 309 3.3 Glu 62.1Ex2 H 231 2.9 Glu

M 231 2.9 Glu 68.3TM H 123 3.9 Asp

M 123 3.9 Asp 71.5C H 765

M 374 70.8

H, human; M, mouse.*Exon size is based on putative cap site.

mature protein and several exons encoding cytoplasmicsegments and 3'-untranslated sequence. Moreover, unlikeT-cell receptor subunit variable or constant region exons,exon Exl of T11 lacks any codons for cysteine residues thatmight be involved in potential intrachain disulfide bondformation known to stabilize immunoglobulin domains(21-24).The extracellular T11 protein segment is involved in

facilitating cell-cell contact as well as T-cell activationevents (1-7). Although it is likely that these functions may bemediated by the same structural domain, the possibilityremains that the functions are compartmentalized into exonsExl and Ex2. In this regard, we have already observed thatthe T111 epitope, which presumably mediates conjugateformation (i.e., sheep erythrocyte-T-cell-rosette formationand cytotoxic T lymphocyte-target conjugate formation) byinteracting with the ubiquitous surface structure lymphocytefunction-associated antigen 3 (4, 7), resides on a proteinfragment encoded exclusively by exon Exl (N. E. Richard-son and E.L.R., unpublished observation). Whether exonEx2 encodes the domain responsible for initiating activationevents, however, remains to be determined.The human genomic T11 sequence agrees with our pub-

lished cDNA sequence (9) with the following four differ-ences: (i) a cytidine in the genomic sequence instead of athymidine reported at position 338 of the cDNA; (ii) theaddition of a cytidine after position 1037; (iii) TG instead ofGT at positions 1322 and 1323; and (iv) the addition of anadenosine after position 1336 (numbered according to PB1cDNA in ref. 9). These differences are the result of sequenc-ing errors in the published cDNA sequence (9). Only theadditional cytidine after position 1037 affects the predictedT11 protein sequence. It causes a frameshift leading to apredicted protein 13 amino acids shorter than reported andto a carboxyl terminus with better homology to the predictedmurine protein (12). This change also agrees with the se-quence reported by Seed and Aruffo (11).The murine T11 genomic sequence reported here is de-

rived from the BALB/cJ mouse strain. It differs at thefollowing five nucleotide positions from the B10.D2 T11cDNA sequence reported (12): (i) a guanosine in the genomicsequence instead of an adenosine at position 397 of theB10.D2 cDNA; (ii) an adenosine instead of a guanosine atposition 450; (iii) a guanosine instead of an adenosine atnucleotide position 539; (iv) a guanosine instead of a thymi-dine at position 588; and (v)-a cytidine instead of a thymidineat position 590 (numbered according to XB2 cDNA in ref.12). These alterations result in four predicted amino aciddifferences: at amino acid (aa) residue 106, valine instead ofmethionine (conservative change); at aa residue 153, serineinstead of asparagine (conservative change); at aa residue169, lysine instead of asparagine (nonconservative change);and at aa residue 170, threonine instead of methionine

(semiconservative change). These changes probably repre-sent genetic polymorphism between BALB/cJ and B10.D2mouse strain. Each of these is restricted to exon Ex2suggesting that there may be fewer evolutionary constraintsin this region of the gene.

Analysis of Til Gene Flanking Sequences. To compareputative regulatory regions of the T11 gene in man andmouse, both 5'- and 3'-flanking regions were analyzed. Weobtained -450 bp of upstream flanking DNA sequence fromeach of the T11 genes (Fig. 3). We provisionally determinedthe human cap site with a modification (31) of the Weaverand Weissman method (32) of S1 nuclease analysis (data notshown). Alignment of human and mouse 5' sequences withthe LOCAL algorithm (25) showed very significant homol-ogy (61.3%). Of note are several regions of near identitybetween the human and mouse genes (from positions - 249to - 229 including 18 of 21 residues and from positions - 211to -192 including 16 of 20 residues and from positions -1 to+ 25 including 21 of 26 residues) that may have somecommon T-lineage importance for gene regulation.We were unable to locate a canonical Goldberg-Hogness

box in the 5'-flanking sequence from either the murine orhuman T11 genes that was consistent with the known sizesof Til mRNA detected by RNA gel blot analysis (9-11).However, several upstream candidate sequences within thehuman gene (positions - 322, - 292, and - 280) and mousegene (position - 360, data not shown) are present within the5'-flanking DNA. In this regard, several other genes havebeen shown not to contain the canonical "TATA" sequence,and in some of these cases [8 subunit of T3 (20); Thy-i (33);hypoxanthine phosphoribosyltransferase (34)] multiplemRNA initiation sites have been described.

Results from RNA gel blot analysis (9) and cDNA cloninghave demonstrated that two distinct sizes (1.7 and 1.3 kb) ofT11 mRNA exist in the human T cell (9-11). In contrast,only a single 1.3-kb T11 mRNA is observed in murinethymocytes (9, 13). Sequence analysis identified two poly-adenylylation signals (AATAAA) within the 3'-untranslatedregion of the human T11 gene (Fig. 2, underlined), whereasthe mouse T11 cDNA contained only one such site. Thedifference in size of the two types of human cDNA clones(-400 bp) (9) can be most easily explained by differential use

-34 2 - 300h CTATTGGCTTGTGAACATTTACCTATATTTCTATGTGGTCTTGTTAGCGACAGTATACCTAAGTGCATAA

m CTATT- -CCTATGACCTTTTGCCGGTCAGTCTACTTGGTCCTGTTGAACCCAGCACAGCTCAGTGGCCA

-250

AAGGCTGTCTGGTTGAATTTGGCTTCTTGTTTACAAAAGAGTGATCCTTAGTGATCTACTTAGCCTCTCT

--- GCTATGTT----M-- -- GCC-CCTTGTTTACAAAAGGGT-AGC- -TA-T -CGCAACTGA-CCTCCCA

-200 -150GTTCCTTTTC-TCTTTCACTGAGATCAGAAAACCTATCCTTCCCAATTTTTTTGTGTGAGAATTAAAATG

GTTCCTCTTCTTCCTCTGGTGGGGTGCTAAAACCCA-ACACCCTAAGCTCTTTTTCTAAACATTAAAATG

-100CAGCA-AGAAAACACACACTCATAAACACATCTGCTTTGGCAAAGGACCACATCAGAAGG - -GCTGGCTT

TGACACAGACACCTATTTCGGTTAAGGAGGGCAGCAAATGCATGAGTCGTTTTGATAGGGTCTCTGTCTC

-50GTCCGCGCT-CTTGC --- -TCTCTGTGTATGTGTATTATGTT ---- -TTATGTTACTGT-AAAAGATGTAAA

TCTCTCTCTCCTTCCCCATCTCTACCTCTCCCTCTCCCCCTCCCCCTCCCCTACTGTGAACAGCAGGCAT

+1 +50CAGAGGCACGTGGTTAAGCTCTCGGGGTGTCGACTCCACCACTCTCACTTCAGTTCCTTTTGCATCAAGA

GAAAGACACGTGGTTCAGGTTGCTGGGTGTGGCTTCTGCCAGCCTCACCACAG TCC - -TGACAGAAACA

+99GCTCAGAATCA-AAAGAGGAAACCAACCCCTAAGATC

ACTCAGAGTCACCCCTGGGAAAAGAACTCTAAAGATG

FIG. 3. Comparison of the human (h) and murine (m) T11 gene5'-flanking regions. Numbering refers to the putative transcriptioninitiation site (position + 1) in human. Alignments were performedas in Fig. 2. The homology value was highly significant at 195;weights used were as follows: for gaps, - (1.01 + 0.90 x length),and for mismatches, -0.10 (ref. 25). The initial methionine codonhas been double underlined.

Proc. Natl. Acad. Sci. USA 85 (1988)

Dow

nloa

ded

by g

uest

on

Janu

ary

24, 2

021

Proc. Natl. Acad. Sci. USA 85 (1988) 1619

of polyadenylylation signals since both are present withinthe genomic exon C (Fig. 2). To determine the DNA se-quence at a comparable position flanking the mouse exon,we sequenced 360 bases downstream of the known murinepolyadenylylation site and compared mouse and humannucleotide sequences. Alignment of the human and mouseDNA sequences was done with the LOCAL algorithm (Fig.4). There is extensive homology between the two sequences(62.1%) with a minimum number of gaps introduced to alignsequences. Surprisingly, the level of similarity betweenhuman 3'-untranslated sequence and murine 3'-flanking se-quence is essentially equal to that between human andmurine T11 exons (Table 1).Most interesting is the presence of a second canonical

polyadenylylation signal in the murine 3'-flanking sequenceat a similar position to the downstream human site (Fig. 4).Why this fails to give rise to a 1.7-kb mRNA species inmurine thymocytes is unclear. Noteworthy is the substitu-tion of three adenosine residues in the human cDNA withthree guanosine residues in a similar position within themurine flanking sequence. These differences may allow thehuman polyadenylylation signal sequence to function as asite for poly(A) addition, resulting in the expression of anmRNA with additional 3' sequence. Alternatively, a tran-scriptional stop sequence may have formed within a 3'-flanking DNA of the murine T11 gene that precludes the useof the downstream AATAAA sequence. We have not founda unique functional role as yet for the longer 1.7-kb humanT11 cDNA, since both the 1.7- and 1.3-kb mRNAs encode aprotein showing properties of the known T11 (9). However,conservation of 3' sequence for at least several hundredbases (65%) suggests the possibility of a common down-stream murine and human regulatory element of the T11gene.Given that the level of nucleotide similarity between 5'-

and 3'-flanking regions of the T11 gene in both species is asgreat as the homology between murine and human T11exons, it would appear likely that the T11 gene is similarlyregulated in both species. Consistent with this notion is theT-lineage restriction of T11 mRNA expression (9-11, 13).

*1250h TGCATATCCC:TACTTCCATGAGGTGTTTTCTGTGTG(CAGAACATTGTCACCTCCT(A(;GC- Ts;TG(G( C(., A

**** ** ***** * ****** ** * ** ** ***** *** * *,F >>v- I:.

m AGCATGCCCACTCTTCCGTCTAGTGTTTAATGGAACTAGGACCCAAGTGCCTCCCCAG ACTTGCA(sA( A

t 1300CAGCCACCTCTGCATCTT- CC -AACTCAGCCATGTGGTC -AACATCTGGAGTTTTTGC;TCTCTC(A-AsA**** ** * ** *** ** ***** * ****** * * ** *** ** k***

AAGTGGTGTCTTTTGATTAAGCTACACAGTCACTTGGTCTGGCCTCTG(GAATCT(GA.GGCCTCTTCTTAG A

+1350 41400GCTCCATCACACCAGTAA-GGAGAAGCAATATAAGTGT(GATTGC;CAAGAATGGTAG;AGGACCG;A(;CA(-AC;A

- -TCTGTCGCGGCAGAAACACAGAAACCACAAACATGTGCACATAAGGATGTT --- -GCA- CAA(;- (AC;

+1450AATCTTAGAGATTTCTTGTCCCCTCTCAGGTCATGTGTAGATGCGATAAATCAAGTGATTCuTGT(GCCTG

----CATAAAGATTTCACCT-TCCTCTCAGGTCCTCTACCCATGTGAGCTACCACATGACC(;CT(;-G-(-T-

+1500 *1550GGTCTCACTACAAGCAGCCTATCTGCTTAAGAGACTCTGGAGTTTCTTATGTGCCCT(;GTG;GAC-Af.TT(,C

CATGTC CTAC -AGGCTACCTG-AGCAGAGGTTC -GAGTCCCCCAGGTGCCCCAATGGACAG T1G

+1597CCACCATCCTGTGAGTAAAAGTGAAATAAAAGCTTTGACTAG

CCA- CTTCCCTTGAGCAGGGGTAAAATAAAAATTTAACCCTG

FIG. 4. Analysis of T11 gene 3'-flanking regions. The polya-denylylation signal sequence is underlined. Note that the humansequence (h) is derived from exon C as shown in Fig. 2 beginning 30bp downstream of the first polyadenylylation signal sequence. Themurine sequence (m) begins immediately on the 3' side of exon C asshown in Fig. 2. The homology value was 165 with parametersidentical to those in Fig. 3.

Transfection analysis of the 12-kb T11 genes in cells of the Tlineage will ultimately determine which sequences are im-portant for tissue-specific expression.

Note Added in Proof. Further analysis of the human gene with themore sensitive RNase protection assay suggests multiple initiationsites (between positions + 16 and + 36 and positions + 57 and + 67),consistent with the absence of a canonical "TATA" sequence.

We acknowledge Forrest Nelson, Hema Ramachandran, andCindy Knall for excellent technical assistance. D.J.D. was a Fellowof the Leukemia Society of America during the course of most ofthis work. L.K.C. was supported in part by Biomedical ResearchSupport Grants 2S07RRO5526-24 and ACS 118G. This work wassupported by Grant A121226 from the National Institutes of Health.

1. Meuer, S. C., Hussey, R. E., Fabbi, M., Fox, D., Acuto, O., Fitzgerald,K. A., Hodgdon, J. C., Protentis, J. P., Schlossman, S. F. & Reinherz,E. L. (1984) Cell 36, 897-906.

2. Siliciano, R. F., Pratt, J. C., Schmidt, R. E., Ritz, J. & Reinherz, E. L.(1985) Nature (London) 317, 428-430.

3. Fox, D. A., Hussey, R. E., Fitzgerald, K. A., Bensussan, A., Daley,J. F., Schlossman, S. F. & Reinherz, E. L. (1985) J. Immunol. 134,330-335.

4. Plunkett, M. L., Sanders, M. E., Selvaraj, P., Dustin, M. L. &Springer, T. A. (1987) J. Exp. Med. 165, 664-676.

5. Vollger, L. W., Tuck, D. T., Springer, T. A., Haynes, B. F. & Singer,K. J. (1987) J. Immunol. 135, 358-363.

6. Martin, P. J., Longton, G., Ledbetter, J. A., Newman, W., Braun,M. P., Beatty, P. G. & Hansen, J. A. (1983) J. Immunol. 131, 180-185.

7. Shaw, S., Ginther Luce, G. E., Quinones, R., Gress, R. E., Springer,T. A. & Sanders, M. E. (1986) Nature (London) 323, 262-264.

8. Lay, W. H., Mendes, N. F., Bianco, C. & Nussenzweig, V. (1971)Nature (London) 230, 531-533.

9. Sayre, P. H., Chang, H. C., Hussey, R. E., Brown, N. R., Richardson,N. E., Spagnoli, G., Clayton, L. K. & Reinherz, E. L. (1987) Proc.Nat!. Acad. Sci. USA 84, 2941-2945.

10. Sewell, W. A., Brown, M. H., Dunne, J., Owen, M. J. & Crumpton,M. J. (1986) Proc. Nat!. Acad. Sci. USA 83, 8718-8722.

11. Seed, B. & Aruffo, S. (1987) Proc. Nat!. Acad. Sci. USA 84, 3365-3369.12. Clayton, L. K., Sayre, P. H., Novotny, J. & Reinherz, E. L. (1987) Eur.

J. Immunol. 17, 1367-1370.13. Sewell, W. A., Brown, M. H., Owen, M. J., Fink, P. J., Kozak, C. A.

& Crumpton, M. J. (1987) Eur. J. Immunol. 17, 1015-1020.14. Williams, A. F., Barclay, A. N., Clark, S. J., Paterson, D. J. & Willis,

A. C. (1987) J. Exp. Med. 166, 368-380.15. Frischauf, A. M., Lehrach, H., Poustka, A. & Murray, N. (1983)J. Mol.

Biol. 170, 827-842.16. Vieira, J. & Messing, J. (1982) Gene 19, 259-268.17. Sanger, F., Nicklen, S. & Coulson, A. R. (1977) Proc. Nat!. Acad. Sci.

USA 74, 5463-5467.18. Chen, E. Y. & Seeburg, P. (1985) DNA 4, 165-170.19. Fuse, A., Fujita, T., Yasumitsu, H., Kashima, N., Hasegawa, K. &

Taniguchi, T. (1984) Nucleic Acids Res. 12, 9323-9331.20. van den Elsen, P., Georgopoulos, K., Shepley, B. A., Orkin, S. &

Terhorst, C. (1986) Proc. Nat!. Acad. Sci. USA 83, 2944-2948.21. Davis, M. M., Chien, Y. H., Gascoigne, N. & Hedrick, S. M. (1984)

Immunol. Rev. 81, 235-258.22. Hayday, A. C., Diamond, D. J., Tanigawa, G., Folsom, G., Heilig,

J. S., Saito, H. & Tonegawa, S. (1985) Nature (London) 316, 828-832.23. Toyonaga, B., Yoshikai, Y., Vadasz, V., Chin, B. & Mak, T. W. (1985)

Proc. Nat!. Acad. Sci. USA 82, 8624-8628.24. Malissen, M., Minard, K., Mjolsness, S., Kronenberg, M., Goverman,

J., Hunkapiller, T., Prystowski, M. B., Yoshikai, Y., Fitch, F., Mak,T. W. & Hood, L. (1984) Cell 37, 1101-1110.

25. Smith, T. F., Waterman, M. S. & Burks, C. (1985) Nucleic Acids Res.13, 645-655.

26. Breathnach, R. & Chambon, P. (1981) Annu. Rev. Biochem. 50,349-383.

27. Ellison, J. & Hood, L. (1982) Proc. Natl. Acad. Sci. USA 79, 1984-1988.28. Gray, P. M. & Goeddel, D. Y. (1982) Nature (London) 298, 859-863.29. Fujita, T., Takaoka, C., Matsui, H. & Taniguchi, T. (1983) Proc. Nat!.

Acad. Sci. USA 80, 7437-7441.30. Wickner, W. T. & Lodish, H. F. (1985) Science 230, 400-407.31. Larsen, P. R., Harney, J. M. & Moore, D. D. (1986) Proc. Nat!. Acad.

Sci. USA 83, 8283-8287.32. Weaver, R. F. & Weissman, C. (1979) Nucleic Acids Res. 7, 1175-1192.33. Gigueve, Y., Isobe, K. I. & Grosveld, F. (1985) EMBO J. 4, 2017-2024.34. Melton, D. W., Konecki, D. S., Brennand, J. & Caskey, C. T. (1984)

Proc. Nat!. Acad. Sci. USA 80, 2147-2151.

Immunology: Diamond et al.

Dow

nloa

ded

by g

uest

on

Janu

ary

24, 2

021