CA1341042C - Y, & t cell receptor and methods for detection - Google Patents

Abstract

The present invention provides purified polypeptides which comprise at least a portion of a .delta. T cell receptor polypeptide, a .gamma. T cell receptor polypeptide, a .gamma. , .delta. T
cell receptor complex or a .gamma. , .delta. T cell receptor complex.
Substances capable of forming complexes with these polypeptides are also provided.
Additionally, methods for detecting T cells which have within them or on their surfaces a polypeptide of the present invention are provided. Moreover, methods for diagnosing immune system abnormalities are provided which comprise measuring in a sample from a subject the number of T cells which have within them or on their surfaces a polypeptide of the present invention.

Description

a34~ ~4 Z-7 , d T CELL REC PTOR AND METHODS FOR DETECTION
BACRGROUND OF 7.~I
Within this application several publications are refer-enced by Arabic numerals within parentheses. Full citations for these references may be found at the end of the specification immediately preceding the claims.
Understanding T cell recognition of antigen and the restriction of the process by major histocompatibility complex (MHC) encoded antigens ~as been an important lc~ goal in immunology. A major step forward occurred with the immunochernical identification of clone specific disulfide-linked het:erodimers on T cells, composed of subunits terme~9 T cell antigen receptors (TCR) a ands .
The TCR a and B subunits have a relative molecular l,; mass (Mr) of approximately 50,000 and 40,000 daltons, respectively (.1, 2, 3). Genes that rearrange during T
cel l ontogeny and encode the TCR s ( 4 , 5 ) and TCR a ( 6 , 7, 8) subunit:s were isolated either by subtractive hybridization or by probing with oligonucleotides.

2.0 A unique feature of the human TCR a, a was the observed comodulation (2), coimmunoprecipitation (9, 10) and required coexpression (11) of the TCR a ,s molecules -2- 1 341 04 2~
with the T3 glycoprotein, which suggested that these two structures were rErlated. Subsequently, the direct physical association of the two protein complexes was demonstrated by chemically cross-linking the TCR a , s molecules to the T3 glycoprotein and identifying the components of the cross-linked complex as the TCR s subunit and the T3 dlycoprotein (Mr 28,000) subunit (12) . A T3 counterpart is similarly associated with murine TCR a , 8 ( 13 , :14) .
0 A third gene that rearranges in T cells, designated TCR Y, has been :identified in mouse (15, 16, 17) and in man (18, 19) . However, there are major differences between the human and mouse TCR Y gene in terms of its genetic structure; for example, the cDNA of the human TCRYgene indicates five potential sites for N-linked glycosylation in the TCR r gene (36) product, which con-trasts with the notable absence of such sites in the m a r i ne TCR Y gee ( 15) . Thus, the human TCR Y gene pmd-uct will have a high molecular weight which is not 20 predictable from its genetic sequence.
The TCR Y gene rearrangements occur in lymphocytes with suppressor-cytotoxic as well as helper phenotypes and may produce a large number of TCR Y chains (18, 19, 25 20, 21, 22, 23) . However, the function of the TCR Y
gene is unknown. Furthermore, neither the protein encoded by the TCR Y gene nor its possible association with other structures (as occurs with TCR a , s and T3 glycoproteins) have been def fined. In humans, the 30 multiple glycosylation sites render it impossible to predict with accuracy the nature and size of .the TCR T
polypeptide structure. Additionally, the published literature does not teach or suggest the util ity of TCR Y with regard to diagnosing, monitoring or staging 35 h~an diseases.

1~41042~
_3_ It appears increasingly likely that the TCR a , B
molecule alone determines both antigen recognition and MHC restriction on at least some T cells (24, 25).
However, it is not clear that TCR a, s accounts for the process of T cell selection during T cell ontogeny or for all antigen specific recognition by mature T cells.
For example, auppressor T lymphocytes remain an enigma;
in some case; they delete or fail to rearrange TCRs genes (26,27). Thus, it is of great importance to determine if a second T cell receptor exists, to define its structure (part.icularly with regard to the possible use of the TCR Y gene product) and ultimately to under-stand what fur.~ction or f unctions it serves.

:35 SUMMARY OF TfiE INVENTION
The present i:nvent:ion provides a purified polypeptide which comprises at least a portion of a a T cell recep-tor polypeptide. Additionally, a substance capable of specifically forming a complex with at least one a T
cell receptor polyp~eptide is provided.
Also provided is a method fo~~ detecting T cells, each of which has <3 a T cell receptor polypeptide. This meth-od comprises contacting a sample which contains T cells with substances capable of forming complexes with a T
cell receptor polypeptides so as to form cellular com-plexes between the substances and the a T cell receptor polypeptides. These cellular complexes are detected ~5 and thereby T' cells, each of which has a a T cell receptor polyF~eptidE', are detected.
The invention further provides a method for diagnosing an immune system abnormality in a subject. This method 20 comprises determining the number of T cells in a sam-ple from the subject and contacting the sample with substances capable of forming complexes with at least one a T cell receptor polypeptide so as to form cellular complexes between the substances and the a T cell recep-25 for polypeptides. The percentage of T cells in the sample which have a~ a T cell receptor polypeptide is determined and compared with the percentage of T cells which have a a T cell receptor polypeptide in a sample from a normal subject who does not have the immune 30 system abnormality. A difference in the percentage of T cells so determined would be indicative of the immune system abnormal ity.

1341042.
A further method for diagnosing an immune system abnor-mality in a subject: is provided by the present inven-tion. This method comprises determining the number of d T cell receptor polypeptide bearing T cells in a sample from t:he subject and determining the amount of a T cell receptor polypeptides in the 6 T cell receptor bearing T cells. The amount of a T cell receptor polypeptides so determined is compared with the amount of s T cell receptor polypeptides in an equal number of s T cell receptor polypeptide bearing T cells in a sample from a norrnal subject who does not have the immune system abnormal ity. A difference in the amount so determined would be indicative of the immune system abnormal ity.
~5 A further method fo;r diagnosing an immune system abnor-mality in a subject is provided. This method comprises determining in a sample from the subject the number of T cells which have a a T cell receptor polypeptide and the number of T cells consisting of the group of T
20 cells which have one of the surface markers T4, T8 and a . s T cell receptor. The numbers of T cells so determined arse compared with the number of T cells which have a a T cell receptor polypeptide and the number of T cells in the group which have the same 25 surface marker as t:he group of T cells determined in the sample from the subject, in a sample from a subject who does not have the immune system abnormality. A
difference in the n~unber of T cells so determined which have a a T cell receptor polypeptide relative to the .30 n~ber of T cells in the group so determined would be indicative of 'the immune system abnormal ity.
The present invention also provides a purified polypeptide which comprises at least a portion of a T
;35 -s- 1341Q42 cell receptor polypeptide. Additionally, a substance capable of specifically forming a complex with at least one Y T cell receptor polypeptide is provided. Further-more, a method for detecting T cells, each of which has a Y T cell receptor pol.ypeptide is provided. This meth-od comprises contacting a sample which contains T
cells with substances capable of forming complexes with Y T cell receptor polypeptides so as to form cellu-lar complexes be~twEen the substances and the Y T cell receptor polypeptides. These cellular complexes are 0 detected and thereby 'T cells, each of which has a Y T
cell receptor pol.ypept:ide, are detected.
A further method for diagnosing an immune system abnor-mality in a subject is provided by the present inven-~5 tion. This method comprises determining the number of T cells in a sample from the subject and contacting the sample with substances capable of forming complexes with at least one Y T cell receptor polypeptide so as to form cellular complexes between the substances and 20 the YT cell receptor polypeptides. The percentage of T
cells in the sample which have a r T cell receptor polypeptide is determined and compared with the per-centage of T cells which have a Y T cell receptor polypeptide in a sample from a normal subject who does 25 not have the immune system abnormality. A difference in the percentage of T cells so determined would be indicative of the immune system abnormality.
Still another method for diagnosing an immune system 30 abnormality is provided. This method comprises deter-mining the number of Y T cell receptor polypeptide bearing T cells in a sample from the subject~and the amount of Y T cell receptor polypeptides in the Y T
cell receptor polypeptilde bearing T cells. The amount ' 1 341 04 2 of Y T cell receptor polypeptides so determined is com-pared with they amount of Y T cell receptor polypeptides in an equal number of Y T cell receptor polypeptide bearing T cells in a sample from a normal subject who does not have the immune system abnormality. A differ-s ence in the amount so determined would be indicative of the immune system at~normality.
Yet another method is provi3ed by the present invention for diagnosing an immune system abnormality in a sub-ip ject. This method comprises determining in a sample from the subject the number of T cells which have a Y T
cell receptor polyF~eptide and the number of T cells consisting of the croup of T cells which have one of the surface markers T4, T8 and a , sT cell receptor.
i5 The numbers of T cells so determined are compared with the number of T cells which have a Y T cell receptor polypeptide and then number of T cells in the group which have the same surface marker as the group of T
cells determined in the sample from the subject, in a sample from a subject who does not have the immune system abnormality. A difference in the number of T
cells so determinecl which have a Y T cell receptor polypeptide relativE~ to the number of T cells in the group so determined would be indicative of the immune :25 $Ystem abnormality.
The invention further provides a purified complex which comprises at least a portion of a a T cell receptor polypeptide and at least a portion of a Y T cell recep-,30 for polypeptide. Also provided are substances capable of specif ically forming a complex with .at least one r , a T cell receptor complex. Moreover; a method for detecting 'T cells, each of which has a .Y , d T cell receptor complex, i;s provided. This method comprises ;35 ~, s 1341A42 contacting a sample which contains T cells with sub-stances capable of forming complexes with r , s T cell receptor comF~lexes so as to form cellular complexes between the substances and the Y ,d T cell receptor complexes. These cellular complexes are detected and thereby T cells, each of which has a Y , s T cell recep-tor complex, are detected.
Still further, the present invention provides another method for diagnosing an immune system abnormality in a subject. This method comprises determining the number of T cells in a sarnple from the subject and contacting the sample with substances capable of forming complexes with at least one y , a T cell receptor complex so as to form cellular complexes between the substances and ~5 the Y , 6 T cell receptor complexes. The percentage of T cells which haves a r , d T cell receptor complex is determined and comF>ared with the percentage of T cells which have a Y , d T cell receptor complex in a sample from a normal subject who does not have the immune 20 system abnormality. A difference in the percentage of T cells so determir,~ed would be indicative of the immune system abnorm~3lity.
The invention provides yet another method of diagnosing 25 an immune system abnormality in a subject. This method comprises determining the number of Y , a T cell recep-tor complex tearing T cells in a sample from the sub-ject and the amount. of Y , 6 T cell receptor complexes in the r, 6 T cell receptor complex bearing T cells. The 30 amount of Y ,~ 6 T cell receptor complexes so determined is compared with the amount of Y , a T cell receptor complexes in an e<;ual number of Y .a T cell receptor complex bearing T cells in a normal subject who does not have the immune system abnormality. A difference 1 341 Q4 2~'~
in the amount: so determined would be indicative of the immune system abnormal ity.
Yet another method for diagnosing an immune system abnormality is provided by the present invention. This method comprises determining in a sample from the sub-ject the number o!: T cells which have a Y , d T cell receptor complex and the number of T cells consisting of the group which have one of the surface markers T4, T8 and a , s T cell, receptor complex. The numbers of T
cells so determined are compared with the number of T
cells which have a Y , a T cell receptor complex and the number of T cells in the group which have the same surface marker as the group of T cells determined in the sample from they subject, in a sample from a subject ~5 who does not have the immune system abnormality. A
difference in the number of T cells so determined which have a Y , s T ce7.1 receptor complex relative to the number of T cells in the group would be indicative of the immune system abnormal ity.
20 In another aspect t:he invention provides a method of identifying a. monoclonal antibody directed against a T cell antigen recE~ptor comprising contacting a viable cell expressing a ~' cell antigen receptor and a CD3 antigen on ita cell surface with the antibody, for a 25 period of tide Sufi=icient to effect co-modulation of the CD3 antigen amd dei:ecting the co-modulation of the CD3 antigen.

1341p42 BRIEF DESCRIPTION OF THE FIGURES
Figure 1: Reactivity of framework monoclonal anti-bodies recognizing TCR a , s .
A. Lane l: Control antibody, normal mouse serum.
Lane 2: Anti-framework TCR a , s mono-clonal antibody (9F1).
B. Lane l: Control antibody, normal mouse serum.
Lane 2: Anti-T3 monoclonal antibody ( UCHT-1 ) .
Lane 3: Anti-framework TCR a, s mono-clonal antibody (WT31).
C. Three dimensional display of flow cyto-metry analysis of normal adult peripheral blood lymphocytes. Red and green fluores-cence were measured and compared to non-spe-cific control FITC- and biotin- conjugated monoclonal antibodies. Cells unreactive with either monoclonal antibody were non T
cells (lower left corner); cells that were double positive, i.e. reacting with both ORT'3 and s F1, make up the large papulation of lymphocytes in the center ~ region of the grid; cells that were 8 F1-but ORT~3+ comprise a small but distinct group of lymphocytes (4$ of the T3+
cells) observed along the X-axis.
Figure 2: SDS-PAGE analysis of cell surface T3 and T3-associal:ed (cross-linked) molecules by immunopre-cipitation from. im~munodeficiency patient (mP) mPl and mP2 cell lines.

1341p42-.
-1, A. IDP1 cell line 2 (WT31+) and cell line 3 (WT31-) .
Lanes 1, 2, 7, 8 . Normal mouse serum.
Lanes 3, 4, 9, 10: Anti-T3 monoclonal antibody (UCHT-1) .
Lanes 5, 6, 11, 12: Anti-framework TCR a , s mono-clonal antibody ( sFl) .
B. IDP2 cell line 7 (88% WT31-T3+) Lanes 1, 4, 7, 10: Normal mouse serum.
Lanes 2, 5, 8, 11: Anti-framework TCR
monocl oval anti bo dy (sFl).
Lanes 3, 6, 9, 12: Anti-T3 monoclonal antibody (UCHT-1) .
1251-labeled samples XL-cross-linked with DSP.

C. IDP2 cell l ine 5 ( WT31+ T3+) and cell l ine 7 ( 8 8% 31-T3 +) .
WT

Lanes 1, 3 . Normal mouse serum.

Lanes 2, 4 . Anti-T3 monoclonal antibody (UCHT-1) .

Figure 3: Northern blot analysis of RNA isolated f rom IDP2 cel:L 1 ine~susing TCR a , TCR s and TCR Y
cDNA

probes.

A. Lane 1 . IDP2 cell line 6 (WT31-) .

Lane 2 . T leukemic cell line HBP-I~T.

B. Lane 1 . IDP2 cell line 5 (WT31+T3+).

Lane 2 . IDP2 cell line 7 (88% WT31-T3+) .

Lane 3 . Cell line HPB-MLT.

Figure 4: Ani~i-V Y and anti-C Y peptide sera immunopre-cipitations from IDP2 cell line 7.

-l~- ~~41a42 A. Lane 1 . Normal mouse serum.

Lane 2 . Anti-V Y peptide mouse serum.

Lane 3 . Normal rabbit serum.

Lane 4 . Anti-C Y peptide rabbit serum.

B. Lane 1 . Normal mouse serum.

Lane 2 . Anti-T3monoclonal antibody (UCHT-1 ).

Lane 3 . Normal rabbit serum.

Lane 4, : Anti-C Y peptide rabbit serum.

Figure 5 Immunoprecipitations of TCRY , a and T3 from a human tumour and peripheral blood lymphocyte lines.
Immur,~oprecipitations from 125I_labelled cell lysates were analysed by SDS-PAGE
~5 (10$ acrylamide) under reducing (R) or nonre~ducing (N or NR) conditions. Size markers, M
r in thousands.
A. TCR Y , a and T3 subunits on IDP2 and PEER
20 cells. Immunoprecipitations were performed using 1 ug control mAb P3 (mAb secreted by the P3X63.Ag8 myeloma lanes 1, 3 , 5 and 6 ) : 1 a g UC HT1 ( an t i-T3 ) ( 4 0 ) (lanes 2, 4, 7 and 8); 10 ul normal rabbit 25 serum, (NRS lane 9) and 10 a 1 anti-C
peptide sera (anti-TCRY ) (lane 10).
Arrows indicate positions of TCR d subunits which change mobility under R and NR conditions.
B. TCRY,a and T3 subunit on peripheral blood lymphocyte (PBL) T cell clone, PBL C1 and the W T31-pBL
LINE. Immunoprecipitations were performed using control mAb P:3 (lanes 1, 4, 9 and s 12) , 1 ug sFl (anti-TCRS ) (lanes 2,5, 10 a.nd 13'1. NRS (lanes 7 and 15) and anti-C
peptides sera (lanes 8 and 16) . Open arrow indicates disulphide-linked F1 and unreact:ive T3-associated species; solid arrow :indicates non-disulphide-linked, T3-associated material that displays increased SDS-PAGE mobility under r~onreducing conditions (like TCR s in A).
Figure 6 Nlorthern blot analysis of RNA isolated f rom PBL C1.
Total RNA pr epa rations f rom the T
leukaernic cell line HPB-1~T (lane 1 for each probe and PBL C1 ( lane for each probe) were analysed on Northern Blots using '.CCRa, TCR s and TCRY cDNA probes.
Figure 7 ~:lao-dimensional gel analysis of TCR
F>olypeptides and precursors.
Panels A-D C:ompar:ison of reduced (separated) and nonreduced (dimeric) T3-associated poly-peptides from PBL C1 cells were lysed in CHAPS and immunoprecipitated with anti-T3 mAb. ~t~ao-dimensional gel electrophoresis was carried out under reducing conditions (A, C) or nonreducing conditions (B, D).
The T3Y , a and a positions are labelled and focused to similar positions under both R. and N conditions. After cleaving t:he disulphide bond, the T3 associated polype;ptides (40K and 36R) migrated to i:ocusing positions close to T3 Y R
<:onditions, but shifted to a more acidic position (close to T3 s ) under N

_1~_ 1 3 41 p 4 2 _ conditions (when both components of the dimes were present (68R)). Size markers, M
r.
Panels E-H Analysis of glycosylated and non-glycoslated IDP2 and PBL C1 TCR peptide precursors. IDP2 and PBL C1 cells were pulse'-labelled with 35S-methionine, lysed under: denaturing conditions and immunoprecipitated with anti-C Y peptide sera.. Immunoprecipitations were then either treated and with endo-H or mock treated and analysed by two-dimensional gel electrophoresis. Glycosylated TCR Y
peptides are denoted by open arrows, and nong7.ycosylated TCR Y peptides by solid arrows. Apparent relative molecular masses were calculated from migration of standards used in panels A and B (not shown). A small amount of contaminating actin, denoted by a diamond in each panel, served as an internal marker. E, IDP2 TCR Y , mock incubated; F, IDP2 TCR r , endo H treated; G, PBL C1 TCR Y mock incut>ated; H, PBL C1 TCR Y endo-H
treated.
Figure 8 Rearrangement of the Y and ~ genes in T
cells expressing the TCR Y polypeptide.
Genonnic DNAs isolated from the IDP2 cell line, PBL C1, PBL C2, WT31 PBL LINE, feta:L thymus, newborn thymus, PBL and a B
cell line (JY for germline) were examined in Southern blot analysis for TCR Y (A, B) and TCR s (C) gene rearrangements.
Genornic DNAs were digested with BamHI (A, C) or EcoRi (B) fractionated on agarose gels and transferred to nitrocellulose filters for hybridization with 32p-labe7.led J Y 1,3 (A~B) or Cs2 Probes (C) .
Arrows and roman numerals denote TCR Y
rearrangements. Size markers in kb.
Figure 9 Cyto7.ysis by IDP2 and PBL C1 cells.
Panels A,C IDP2 or PBL C1 effector cells were incubated (at effector:target, (E: T) ratios indicated) with 5lCr-labelled targE~t cells 8562 (erythroid line) , U937 (monocytic line), MOLT-4, CEM (T leukaemic lines), Daudi (Burkitt's lymphoma line) or allogeneic or autologous PBL (3-day PHA
blasts of human PBL) . The % specific rele<ise of 5lCr for each target is shown.
The same assays were carried out after prebinding anti-T3 mAb UCHT1 to the effe<:tor cells for 30 minutes at OoC
(+ant:i-T3 ) .
Panel B. Inhibition of IDP2 cytolysis of MOLT-4 target cells by various mAb. IDP2 and 5lCr--labelled MOLT-4 cells were incubated together at a 40:1 E:T ratio in the presE~nce of various dilutions of anti-MHC
Class 1 mAb W6/32 (anti-HLA-A, B, C mono-morphic determinant) (58), anti-HLA-A, B, C (monomorphic determinant) (59), 4E
(anti-HLA-B and C locus) (60), 131 (anti-HLA-A locus)(61) or anti-MHC Class II mAb LB3.7L (anti-DR specific)(62), anti-Leu 10 (ant:i-DQ specific) (63) or anti-T3 mAb UCHT7L. Higher dilutions were used for mAb as ascites (W6/32, 4E, 131, LB3.1 and UCHT1) and lower dilutions were used for commerical mAb(anti-Leu 10) or culture supernatant (anti-HLA-A, B, C) .
Figure 10. Immunoprecipitation of a TCR Y chain derived from Peer cells fi Lanes I-8 are various hybridoma culture sup~ernatates f rom the same hybridoma fusion experiment; Lane 4 is anti-Y chain monoclonal antibody 34D12; Lane 9 is control P3x63Ag8.653 culture supernate;
Lane 10 is control normal mouse serum.
Figure 11. Immunoprecipitation of a TCR a chain derived f rom IDP2 cells 1 ~i Lane 1 is control P3x63Ag.8.653 culture sup~'rnate; Lanes 2 and 5 are monclonal antibody 4A1 culture supernatate; Lane 3 is Leu 4; Lane 4 is control normal rabbit serum.
2t) IDP2 cells were 1251 labeled using lactoperoxidase and solubl ized in 2%
Tri~ton* X100. In Lanes 4 and 5, this lys~ite was boiled for 3 minutes in 1% SDS, diluted with 4 volumes of 1% Triton X100 and renatured overnight at 4°C. This resulted in separation of the TCR Y and b chains. After such chain separation, mon~~clonal antibody 4A1 specifically immunoprecipitated the TCR a chain (Lane 5) His described hereinabove. ' * pegisl~Qxed Trade-mark 3:5 1341p42~

DETAILED DESCRIPTION OF THE INVENTION
The present :invention provides a purified polypeptide which comprises at least a portion of a s T cell recep tor polypeptide. This polypeptide may have at least one intrachain, covalent, disulphide bridge. Addition ally, the polypeptide may comprise a d T cell receptor nolypeptide having a molecular weight of about 40,000 daltons. Furthermore, the a T cell receptor polypeptide may be a human s T cell receptor poly peptide.
A substance capable of specifically forming a complex with at least one a T cell receptor polypeptide is also provided by t:he .invention. In one embodiment of the invention, tt;e substance is capable of specifically forming a complex with one s T cell receptor polypeptide. In another embodiment of the invention, the substances is capable of specifically forming a complex with more than one a T cell receptor polypeptide. The substance may be an antibody. The antibody may be a polyclonal antibody or a monoclonal antibody.
Also provided is a method for detecting T cells, each of which has ,~ s T cell receptor polypeptide. This meth-od comprises contacting a sample containing T cells with substanc~as capable of forming complexes with d T
cell receptor polypeptides so as to form cellular com-plexes between the substances and the a T cell recep-for polypepticfes. These cellular complexes are detect-ed and thereby T cells, each of which has a, d T cell receptor polyF~eptide, are detected.

~341Q42i Accordingly, in one embodiment of the invention, the d T cell receptor polypeptides are present on the surfaces on t:he T cells. In another embodiment of the invention, t:he a T cell receptor polypeptides are present in the cytoplasm of the T cells.
This method may be performed by forming complexes with a specific a T cell receptor polypeptide. In one em-bodiment of the invention, the specific a T cell recep-tor polypeptide is present only in suppressor T cells.
The inventions further provides a method for diagnosing an immune system abnormality in a subject. Within this appl ication, immune system abnormal ity means a condi-tion of immunological responsiveness to antigens char-acterized by an increased or a decreased immune re-sponse compared to a normal or standard immune re-sponse. Accordingly, immune system abnormality in-cludes, but is not limited to, immunodeficiency con-ditions and ~~iseases, e.g. acquired immune deficiency syndrome and congenital immunodeficiencies and hyper-immune conditions and diseases, e.g. allergies and hayfever. The method of the present invention compris-es determining the number of T cells in a sample from the subject and contacting the sample with the sub-stances capable of torming complexes with at least one d T cell receptor polypeptide so as to form cellu-lar complexes between the substances and s T cell re-ceptor polypeptides. The percentage of T cells in the sample which have a d T cell receptor polypeptide is determined and compared with the percentage of T cells which have a 6 T cell receptor polypeptide in a sample from a normal subject who does not have the immune system abnormality. A difference in the percentage of T
cells so determined would be indicative of the immune system abnormal ity.

In one embodiment of the invention, the immune system abnormality is a cancer. The cancer may be a leukemia or a lymphoma. In another embodiment of the invention, the immune system abnormal ity is acqui red immune def i ciency syndrome. In yet another embodiment of the invention, the immune system abnormality is congenital immunodeficiency. In still a further embodiment of the invention, the immune system abnormality is an autoim-mune disease.
The subject :gin whom the immune system abnormality is diagnosed may be an animal. In one embodiment of the invention they subject is a human. Furthermore, the sample from the subject may comprise blood or tissue.
Yet another method for diagnosing an immune system abnormality i;s provided by the present invention. This method comprises determining the number of a T cell receptor polypeptide bearing T cells in a sample from the subject and the amount of a T cell receptor polypeptides :in the T cell receptor polypeptide bear-ing T cells. The amount of a T cell receptor polypeptides so determined is compared with the amount of a T cell receptor polypeptides in an equal number of a T cell receptor polypeptide bearing T cells in a sample f rom a normal subj ect who does not have the immune system abnormality. A difference in the amount so determined would be indicative of the immune _ system abnormality. In one embodiment of the invention, the amount of a single a T cell receptor polypeptide is determined.
A f ur ther method f or di agnosing an immune system abnor-mality in a subject is provided. This method comprises 1341042-.

determining ~.n a sample from the subject the number of T cells which have a aT cell receptor polypeptide and the number of T cells consisting of the group of T
cells which have one of the surface markers T4, T8 and a , s T ~~ell receptor. The numbers of T cells so determined is compared with the number of T cells which have a a T cell receptor polypeptide and the number of T
cells in the group which have the same surface marker as the group of T cells determined in the sample from the . subj ect, in a sampl a f rom a subj ect who does not have the immune system abnormality. A difference in the number of: T cells so determined which have a a T
cell receptor polypeptide relative to the number of T
cells in the group so determined would be indicative of the immune system abnormality.
The present invention also provides a nucleic acid molecule encoding a a T cell receptor polypeptide having a molecular weight of about 40,000 dal tons. In one embodiment of: the invention, the molecule is a DNA
molecule. Further provided is a nucleic acid molecule which is complementary to the nucleic acid molecule which encodes a a T cell receptor polypeptide.
A purified polypeptide which comprises at least a por-tion of a Y T cell receptor polypeptide is also provid-ed by the present invention. This polypeptide may comprise a Y T cell receptor polypeptide having a mo-lecular weight of about 55,000 daltons. In one embodiment of the invention, the polypeptide has a peptide sequence with a molecular weight within the range from about 31,000 daltons to about 40,000 daltons. Additionally, the polypeptide may be a human Y T cel:~ receptor polypeptide.

The present invention further provides a purified complex which comprises two Y T cell receptor polypeptides of the present invention associated with each other. In one embodiment of the invention, the two r T cell receptor polypeptides are associated with each other through at least one interchain, covalent, disulphide linkage. In another embodiment of the invention, the two r T cell receptor polypeptides are noncovalently associated with each other. In still another embodiment of the invention, the two Y T cell receptor polypeptides have the same constant domain.
In yet a further embodiment of the invention, the two Y
T cell receptor polypeptides have different constant domains.
~5 The present invention also provides a substance capable of specifically forming a complex with at least one Y T
cell receptor polypeptide. In one embodiment of the invention, the substance is capable of specifically forming a complex with one Y T cell receptor 20 polypeptide. In another embodiment of the invention, the substances is capable of specifically forming a complex with more than one Y T cell receptor polypeptide. 'Phe substance may be an antibody. In one embodiment c~f the invention, the antibody is a poly-25 clonal antibody. In another embodiment of the inven-tion, the antibody is a monoclonal antibody.
A method for detecting T cells, each of which has a Y T
cell receptor polypeptide, is further provided. This 30 method comprises contacting a sample which contains T
cells with substances capable of forming complexes with Y T cell receptor polypeptides so as to form cel lular complexE~s between the substances and the Y T cell receptor polypeptides. These cellular complexes are - 1341042) detected and thereby T cells, each of which has a r T
cell receptor polypeptide, are detected. In one embod-iment of t:he invention, the r T cell receptor polypeptides are present on the surfaces of the T
cells. In another embodiment of the invention, the r T
cell rece f or of a tides are p p yp p present in the cytoplasm of the T cells. In yet another embodiment of the in-vention, the substances are capable of firming complex-es with a sp~~citic 7 T cell polypeptide. The specif-ic Y T cell receptor polypeptide may be present only in suppressor T cells. Furthermore, the r T cell receptor polypeptide may be associated with another Y T cell receptor polypeptide. In one embodiment of the invention, the Y T cell receptor polypeptide is associated with another r T cell receptor polypeptide.
~5 In another embodiment of the invention, the Y T cell receptor polypeptide is associated with another Y T
cell receptor polypeptide only in non-major histocompatibility restricted cytotoxic T lymphocytes.
Furthermore, the non-major histocompatibility complex 20 restricted c;rtotoxic T lymphocytes may be T killer cells or natural killer-like cells.
The present invention further provides a method for diagnosing an immune system abnormal ity in a subj ect.
25 This method comprises determining the number of T cells in a sample from the subject and contacting the sample with substances capable of forming complexes with at least one Y T cell receptor polypeptide so as to form cellular complexes between the substances and Y T cell 30 receptor polypeptides. The percentage of T cells in the sample wr,ich have a Y T cell receptor polypeptide is determined, and compared with the percentage of T
cells which have a Y T cell receptor polypeptide in a normal subj eca who does not have the immune system abnormality. A difference in the percentage of T cells so determined would be indicative of the immune system abnormality. In one embodiment of the invention, the immune system abnormal ity is a cancer. The cancer may be a leukemia or a lymphoma. In another embodiment of the invention, the immune system abnormality is ac-quired immunE~ deficiency syndrome. In yet another embodiment of the invention, the immune system abnor-mality is congenital immunodeficiency. In still a further embodiment of the invention, the immune system abnormality is an autoimmune disease.
The subject in which the immune system abnormality is diagnosed may be an animal. Additionally, the subject in which the immune system abnormal ity is diagnosed may ~5 be a human. Furthermore, the sample of which the per-centage of T cells which have a r T cell receptor polypeptide is determined may comprise blood or tissue.
Yet another method for diagnosing an immune system 20 abnormality in a subject is provided by the present invention. This method comprises determining the num-ber of Y T cell receptor polypeptide bearing T cells in a sample f rom the subj ect and the amount of Y T cell receptor polypeptides in the r T cell receptor 25 polypeptide bearing T cells. The amount of Y T cell receptor polypeptides so determined is compared with the amount of YT cell receptor polypeptides in an equal number of Y T cell receptor polypeptide bearing T cells in a sample from a normal subject who does not have the 30 rune system abnormality. A difference in the amount so determined would be indicative of the immune system abnormality. In one embodiment of the invention, the amount of a single T T cell receptor polypeptide is determined.

Further provided i;s another method for diagnosing an immune system abnormality in a subject. This method comprises determining in a sample from the subject the number of T cells which have a Y T cell receptor polypeptide and the number of T cells consisting of the group of T cells which have one of the surface markers T4, T8 and a , a T' cell receptor. The numbers of T
cells so detez~mined are compared with the number of T cells which have a Y T cell receptor polypeptide and the number of T cells in the group which have the same surface marker as the group determined in the sample f rom the subj ect , :in a sampl a f ran a subj ect who does not have the immune system abnormality. A difference in the number of T cells so determined which have a Y
~5 T cell receptor polypeptide relative to the number of T
cells in the group so determined would be indicative of the immune system abnormality.
A purified complex which comprises at least a portion 20 of a a T cell. receptor polypeptide and at least a por-tion of a Y T cell receptor polypeptide is further provided by 'the present invention. This complex may comprise a a T cell receptor polypeptide having a mo-lecular weight of about 40,000 daltons and a Y T cell 25 receptor polypeptid~e having a molecular weight of about 55,000 dal tons. Furthermore, the a T cell receptor polypeptide may be a human a T cell receptor polypeptide and the Y T <:ell receptor polypeptide may be a human Y
T cell recep~:or polypeptide. Moreover, the a T cell 30 receptor pol.ypeptide and the Y T cell receptor polypeptide may be associated with each other through at least one inter chain, covalent, disulphide linkage, or may be noncovale~ntly associated with each other.
T

13~10~2 Also provided is a substance capable of specifically forming a complex with at least one Y , s T cell receptor com~~lex» This substance may be capable of forming a complex with one Y , a T cell receptor complex. Furthermore, the substance may be capable of f orming a conipl ex with more than one Y , a T cell re-ceptor complex.
In one embodiment of the invention, the substance is an antibody. In another embodiment of the invention, the substance is a polyclonal antibody. In yet another embodiment of the invention, the substance is a monoclonal antibody.
The present invention further provides a method for ~5 detecting z' cells, each of which has a Y, d T cell receptor complex. This method comprises contacting a sample containing T cells with substances capable of forming complexes with Y , a T cell receptor complexes so as to form cellular complexes between the substances 20 and the Y s Z' cell receptor complexes. These cellular complexes area detected and thereby T cells, each of which has a Y , a T cell receptor complex, are detect-ed. In one embodiment of the invention, the Y , a T
cell receptor complexes are present on the surface of 25 the T cells. In another embodiment of the invention, the Y , d T cell receptor complexes are present in the cytoplasm of the T cells. In yet another embodiment of the invention., the substances are capable of forming complexes with a specific Y , a T cell receptor complex.
30 The specific r, 6 T cell receptor complex may be present only in suppr~~ssor T cell s.
A method for diagnosing an immune system abnormality in a subject is further provided by the present invention.

134'042_ This method comprises determining the number of T cells in a sample from the subject and contacting the sample with substances capable of forming complexes with at least one Y , 6 T cell receptor complex so as to form cellular comple~;es between the substances and Y , 6 T
cell receptor complexes. The percentage of T cells in the sample which have a Y , 6 T cell receptor complex is determined and compared with the percentage of T cells which have a Y, d T cell receptor complex in a sample from a normal aubject who does not have the immune system abnormality. A difference in the percentage of T cells so determined would be indicative of the immune system abnormality. In one embodiment of the inven-tion, the immune system abnormal ity is a cance r. The cancer may be a leukemia or a lymphoma. In another embodiment of the invention, the immune system abnor-mality is acquired immune deficiency syndrome. In yet another embodiment of the invention, the immune system abnormality is congenital immunodeficiency. In yet a further embodiment of the invention, the immune system 20 abnormality is a;n autoimmune disease.
The subject in which the inatuu~e system abnormality is diagnosed may be an animal. Furthermore, the subject in which the immune system at~normal ity is diagnosed may be a human.
25 Moreover, the sample in which the percentage of T cells which have a r, d T cell receptor complex is deter-mined may comprise blood or tissue.
Still another method for diagnosing an immune system 30 abnormality in a subject is provided by the present invention. This. method comprises determining ,the num ber of Y , 6 T cell receptor complex bearing T cells in a sample f rom the subj ect and the amount of T , d T
cell receptor complexes in the Y , a T cell receptor complex bearing T cerlls. The amount so determined is compared with I:he amount of Y , d T cell receptor com plexes in an e<~ual number of Y , 6 T cell receptor com pl ex bea r ing T cel 1 s in a sampl a f rom a normal subj ect who does not have immune system abnormality. A differ s ence in the amount so determined would be indicative of the immune system abnormality. In one embodiment of the invention, the amount of a single Y ,6 T cell receptor complex is c9etermined.
Yet a turther method for diagnosing an immune system abno-rmality is provided. This method comprises deter-mini ng i n a sampl a f rom the subj ect the numbe r of T
cells which have a Y . 6 T cell receptor complex and the number of T cells consisting of the group of T
~5 cells which have one of the surface markers T4, T8 and a , 9 T cell receptor. The numbers of T
cells so determined are compared with the number of T
cells which have a Y , a T cell receptor complex and the number of T cells in the group which have the same 20 surface marker as the group of T cells determined in the sample from the subject, in a sample from a subject who does not have 'the immune system abnormal ity. A
difference in ithe number of T cells so determined which have a Y, s T cell receptor complex relative to the 25 number of T cells in the group so determined would be indicative of l:he immune system abnormality.

__ ._ ....

-2g- i34i~42_ The various methods for diagnosing abnormalities and for detecting T cells provided by the present invention are based upon they novel polypeptides and substances capable of f o.rming complexes with these polypeptides as described more f ull.y hereinabove. The methods util ize methods for detecting and quantifying T cells, includ ing but not limited to, fluorescence activated cell sorting and autora~diography, which are well known to those skilled in the art to which this invention per tains.

Example s Example 1 Materials and Methods.
Lymphocyte culture ar,~d cell population analysis Viable lymphocytes were isolated by Ficoll'-hypaque density centrifugation and stained with 0.5 micrograms of a speci f is monocl oval anti body, e. g. WT31 ( 2 8, 2 9) or ORT'3, ORT''4 or ORT'8 (Ortho Diagnostic Systems, Inc. , Raritan, NJ) . for 30 minutes at 4oC. After washing, the cell pellets were stained again with fluorescein isothiocyanate (FITC)-conjugated goat anti-mouse IgG(ab)'~, fragments. Fluorescence activated cell ~~5 sorter (FRCS) analyseas were performed on an Ortho cytofluorograph or a Coulter Epics~ as previously described (37) .. Specifically stained positive cells were determined relative to a negative control pr of ile for each cell :line (stained with a nonspecific control monoclonal antibody) . Cells having fluorescence inten-sity channel numbers greater than the intercept of the negative control pr of:ile with the baseline were counted as positive, anal the % positive was calculated relative to the total number of cells counted.
;25 All IL-2 depenc9ent cell lines were propagated in vitro in media composed o:E RPMI 1640, 10% human serum and conditioned media containing 2-5 units of interleukin-2 activity as previously described (34).
Alloantigen (a~.lo) activated cultures were stimulated with irradiatec9 allogenic peripheral blood lymphocytes at weekly intervals. Mitogen, i.e. phytohemagglutinin (PHA), activatead lines were stimulated with a 1:1,000 dilution of PHp, (Difco, Detroit, MI) at culture initia-tion.
* Trademark Reactivity and characterization of cell culture using monoclonal antibodies Immunoprecipit~ates from 1251-labeled lymphocyte lysates were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). The radioiodinated T
leukemia cell linea HPB-MLT and Jurkat, the HTLV-1 transformed cell line ANITA and resting peripheral blood lymphocytes were solubilized in 1% Trito ~X100 (TX-100) and immunoprecipitated with a control anti-body, normal mouse aerum (NMS) or a framework antibody to TCR a , a i.e., ;f1 (54) . The sFl monoclonal anti-body was prepared according to standard procedures (46, 47, 52). Spleen cells from mice immunized with purified TCR a , s as described in (28) were used for the fusion experiments. A pos;~ive clone, sFl, was obtained by immunoprecipitation with T cell lines and peripheral blood lymphocytes as described above.
125I-labeled lymphocytes were solubilized in 0.1% TX-2C~ 100 and immunoprecipitated with NMS, the anti-T3 anti-body UCHT-1 (40) anf~ a framework antibody to TCR i.e., WT31. The eff:icienc:y of immunoprecipitation with WT31 was improved at the lower TX-100 concentration used here and the monoclonal antibody 187.1 (53) was used as a second antib~~dy.
Two-color FACS analysis of normal adult peripheral blood lymphocytes was performed using an anti-TCR a , a monoclonal antibody and an anti-T3 monoclonal antibody.
Peripheral blood lymphocytes were stained first with an FITC-conjugated anti-T3 monoclonal antibody (ORT~3) and then with a biotinyl-anti-TCR a. B monoclonal antibody (BF1) followed by phycoerythrin-conjugated avidin (PE-avidin, Becton Dickinson Mt. View, CA).

»

41p42 Viable lymphocytes were isolated by ficoll-hypaque density centrifugation for SDS-PAGE and FACS analyses.
For SDS-PAGE analysis, lymphocytes were radioiodinated by the lactoperoxidase technique, solubilized in 1% TX-100 and immunoprec;ipitated using 1 microgram of a spe-cific antibody, :i.e. monoclonal antibody sFl or monoclonal antibody UCHT-1, or 1 microliter of NMS.
The immunoprecipitates were then analysed by 10.5% SDS-PAGE under reducing conditions. The 125I-labeled mole-cules were vi.suali.:ed by autoradiography as previously described (28) .
Two-colored cytofluorographic analysis was performed by first staining with FITC-ORT'3 monoclonal antibody for ~5 45 minutes at 4°C. After washing, the lymphocytes were fixed in 1% paraformaldehyde for 15 minutes at 23°C
then incubated in 70% ethanol in phosphate buffered saline (PBS) for 5 minutes at -20°C. After further washing, the cells were stained with the biotinyl-sFl 20 monoclonal antibody followed by PE-avidin. Analysis was performed on an. Ortho' cytofluorograph (Ortho Diagnostic Systems, Inc., Westwood, MA).
Analysis of cell surface protein molecules associated 25 with T3 molecules on IDP1 and IDP2 cell lines IDP1 cell line 2 (1~1T31+) and cell line 3 (WT31 ) were 1251-labeled as described above. Radioiodinated, in-tact lymphocyi=es were then either cross-linked by incu-30 bation in PBS (pH 8) containing 50 micrograms/ml di-thio-bis-succ:inimid~yl propionate (DSP) or mock incu-bated. The cells were then solubilized in 1% TX-100 and immunopre~cipitated as previously described (12) .
T3 associated molecules (Mr 40,000-55,000) in the anti-T3 immunoprecipitations were detected at low levels in the noncross-linked samples and at higher levels in the cross-linked samples.
IDP2 cell line 7 (88% WT31-T3+) was I25I-labeled and treated with DSP or mock incubated. Immunoprecipita-tions were performed using NMS, the anti-T3 monoclonal antibody UCHT-1 and the anti-TCR a, ~ monoclonal antibody BFI either without or with preclearing TCR a , s molecules with t:he monoclonal antibody B F1. A small fraction of radiol<sbeled TCR a , B was detected in samples which were not precleared but not in samples which were precleared with BF1.
IDP2 cell l ine !i (WT3:1+T3+) and cell line 7 ( 88% WT31-~5 T3+) were 1251-labeled, solubilized in 1% TX-100 and immunoprecipitated using NMS or the anti-T3 monoclonal antibody UCHT-1. The T3 heavy subunit (Mr 27,000) appeared similar' on these two cell lines, while the T3 light subunits (Mr 19,.000-25,000) did not.
1251-labeling, solubil ization in 1% TX-100, immunoprecipitation and visualization after 10.5% SDS-PAGE analysis by autoradiography were performed as previously described (28) . Chemical cross-linking was performed for 30 minutes at 23°C on intact radiola-beled lymphocytes using DSP (50 micrograms/ml) in PHS
(pA 8) as previously described (12) . After immuno-precipitation, all samples were examined by SDS-PAGE
under reducing conditions using 5% 2-mercaptoethanol, which cleaved both the disulfide bonds between protein subunita and the DSP chemical cross-link.
Northern blot a:nalys:is of RNA isolated from IDP2 cell lines using TCgn T 'lZa and TCRY cDNA probes Total RNA (15 micrograms) isolated from IDP2 cell l ins 6 ( WT31-) and f rom T 1 eukemic cel l l ins HBP-MLT
was fractionated on a 1.5% agarose gel containing 2.2M formal dehydEa, transferred to nitrocellulose and hybridized with T'CR a , TCR s and TCR Y probes.
Total RNA (3 micrograms) isolated from IDP2 cell line 5 (WT31+T3 ~) . II)P2 cell line 7 (88% WT31-T3+) and HPB-MLT was analyzed as described above.
RNA preparation, electrophoresis, transfer to nitro-cellulose and hybridization with 32P-labeled, nick-translated probes; (1-3 X 108 cp~/microgram) were as described previously (41) . a-chain probes were ei-~5 then the human cDNA clones pGAS(8) or Ll7a (42). s chain probes were either the human cDNA
clones 12A1 (43) or L17 (43). The Y-chain probe was an EcoRI to AccI fragment derived from human cDNA
clone TY-1 (36) . Radioactive bands were visualized by autoradiography using intensifying screens. All probes were labeled to nearly identical specific activity, and identical exposure times are presented.
Immunoprecipitation of IDP2 cell line 7 surface mole-pules using anti-Y antiserum TX-100 solubilize~d 125I_labeled IDP2 cell line 7 (88%
WT31 T3+) was denatured (see below) and then immuno-precipitate~d witla NMS or normal rabbit serum and with anti-V Y peptide serum or anti-C r peptide serum.
A specific band was observed at Mr 55,000 in both the anti-V Y and anti-C r immunoprecipitations. The addi tional band at h1r 90,000 was not reproducibly ob served in the anti-C Y immunoprecipitations (see be 1 ow ) .

DSP cross-linked native lysates (1% TX-100) from 1251-labeled IDP2 cell line 7 were immunoprecipitated with NMS ~or with the anti-T3 monoclonal antibody UCHT-1. Alternatively, the lysate was denatured (as described below) and immunoprecipitated with either normal rabbit serum or with anti-CY peptide serum.
An additional aliquot of lysate was subjected to a two stage immunoprecipitation. Polypeptides were immunoprecipitate~d with the anti-T3 monoclonal anti-body UCHT-1, and were eluted from the immunoabsorbent under denaturing and reducing conditions, in order to break the DSP ci:oss-link. Immunoprecipitation from this eluate~ was then performed using anti-C Y peptide ~5 serum.
1251-labeling, ~solubilization in 1% TX-100 and immuno-precipitation were performed as described above. Native l;ysates (1% TX-100) were denatured by 20 the addition of SDS (final concentration of 1%) and dithiothreitol ( i:inal concentration of ZmM) followed by heating the mixture for 5 minutes at 680C. After cool ing, io~doacet:amide was added ( 20mM f final concen-tration) ar,~d samples were diluted with the addition 25 of 4 volumEas of 1.5% TX-100 in Tris buffered saline (pH 8) . The inii:ial immunoprecipitate in the experi-ment was denatured and subsequently partially re-natured (2F~). Samples were immunoprecipitated with microliters ot: anti-C Y or anti-V Y peptide sera, 1 30 microgram of UCHT-1 or 1 microliter of NMS or normal rabbit serum and analyzed by 10.5% SDS-PAGE under reducing conditions (5% 2-mercaptoethanol).

Peptides corresponding to deduced V Y or C ramino acid sequences (residue numbers noted below in the Experi-mental Result section) were synthesized on a Beckman 990 peptide synthesizer using the method of Erickson and Merrifield (44) . Peptide purity was assessed by high pressure liquid chromatography and peptide se-quence was conf i rmed by amino acid analysis. Pep-tides were coupled to keyhole limpet hemocyanin (RLA) at a ratio of 50 peptides per RLH molecule (45) .
Mice and rabbits were immunized with the V Y peptides or C Y peptides, respectively. Animals were injected at three week intervals and the antisera screened for binding reactivity on peptide-RLH and peptide-bovine serum albumin conjugates to ascertain the presence of peptide-specific antibodies.
Monoclonal ant:ibodi~es against the Y chain were gener ated by standard procedures as described in (47, 50).
BALB/c mice were immunized with the IC.A-coupled peptide to the variable region Y chain peptide de scribed above using the method of Erickson and Merri field (44) . After four immunizations at two week intervals, spleen cells were fused with P3-X63-Ag8U1 myeloma cells. Positive hybridana clones were screened and identified by the enzyme immunoassay (EIA) described in (48) .
Isolation of DNA sequences of the a polypeptide using sequences from purii°ied proteins DNA sequences of th,e TCR a gene may be isolated and determined by strategies utilized to isolate the TCR s gene as described in (49, 50) . Briefly, the amino acid sequence of the TCR a gene may be deter-mined following isolation of the TCR a polypeptide which is described hereinafter. After the amino acid sequence is determined, short, synthetic DNA
sequences may be prepared using a commercial DNA
synthesizer (.Applied Biosystems, Inc. , Foster City, CA) . The synthetic DNA sequences may be used as probes for the isolation of the complete sequence of DNA from a cDNA library of cell lines containing the T CR ~ p o 1 y pe p t i d a . The complete primary structure of the protein may then be determined (51) .
Preparation oi: monoclonal antibodies against the a polvoeptides and against r , a comel~xes Monoclonal antibodies against the a polypeptide may be generated by standard procedures (47, 50).
Peptides derived from the TCR a polypeptide may be prepared from nucleic acid sequences determined by the methods described above. Methods for the selection of such Peptides useful for immunization have been described in detail (55, 56, 20 57) .
Monoclonal antibodies directed against Y,a complexes may be prepared according to published procedures (47, 50) . Y , a complexes may be isolated from the T
cell lines described above and used to immunize HALH/c mice as d~escr ibed in pr ev iously publ fished procedures (28). Alternatively, BALB/c mice may be immunized with cell l fines, e. g. , the IDP1 cell l fine or the IDP2 cell line.
Methods for the fusion, generation and maintenance of hybridoma cell lines have been widely published and are known to those skilled in the art. Hybridoma cells that producEr monoclonal antibodies which are directed against apecific TCR Y , a cell lines but which do not cross react with other T cell lines may be selected and recovered.
Immunoprecip:itations of TCR Y, a and T3 from a human tumour aid c~:richeral blood lvmchocyte lines Viable lymphocytes were isolated by Ficoll-Hypaque density gradient centrifugation and 2 x 107 cells were radio-iodinated by the lactoperoxidase technique as described (28) . Labelled cells were lysed in 5 ml of T8S (10 naM Tris pH B, 140 mM NaCl) with 0.3% 3-f (3- cholamidopropyl) dimethylammonio] -1-propane-sulphonate (CSAPS~) , which preserves the TCR-T3 association (5~1) , containing 2mM phenyl methylsulphonyl fluoride (P~!SF) and 8 mM
iodoacetamidE~ (IAA). Immunoprecipitation was carried out using fixed ,~t~~rlococcus aureus Cowan I (SACI) as described (12), and the immune complexes were washed x 5 i.n T8S containing 0.1% Triton X-100 (TX-;00 100) . Rec9uced samples were boiled in 2 mM
dithiothreitol (DT'T) and all samples incubated for 10 min at 23oC :in 10 mM IAA before analysis by SDS-PAGE.
Immunoprecipi.tations using anti-C Y sera were performed on 1% T~!;-100 lysates that were dialysed to remove IAA and then denatured by the addition of one tenth vol um~e of sodi um dodecy 1 sul phate ( SDS ) containing 3 mM DTT with boiling for 3 min. After partial renaturation by the addition of 4 vols of 1.5% TX-100 i.n TeS containing 30 mM IAA. anti-C Ysera 3.0 or NRS were added and the immunoprecipitates were washed in TBS containing 0.5% TX-100, 0.5%
deoxycholate, 0.05% SDS before analysis by ~ SDS-PAGE.
Rat anti-mousse a chain-specific mAb 187.1 ~ 15 ug) was added as a second antibody to provide protein A
35 binding of Ic~GI ~~~ s F1, UCHT and P3 ( 53) .

13~~~42_ Northern blot ana.lvsis of RNA isolated from PBL C1 Approximately 1.5 ~g RICA. was loaded per lane, probes were labelled to similar specific activity, and identical autoradiographica:L exposures are presented. RNA sizes were determined based on previously published lengths for TCRa and TCRY transcripts (36, 4) .
Two-dimensional aE~l analysis of TCRY polyp~tides and presursors After radioiodinai~ion with lactoperoxidase, lymphocytes were treated with 100U of neuraminidase (Gibco) in phosphate-buffered saline (PBS) 1 mg ml-1 bovine serum ablumin, 1 mg ml-1 glucose for 90 minutes at 23°C, washed in PBS and solubi=Lized in 0.3o CHAPS. Immunoprecipitates were prepared as in Fig. 1 and NEPHGE (charge separation) was carried out u:~ing p:H 3.5-10 ampholines (LKB, Sweden), or IEF using pH 3.5-10, 4-6, 9-11 amphollines (2:15.5:1.5) followed by 10.5% SDS-P,~GE gels for size separation as described (12). PJEPHGE was carried out (A, B) applying the iodinated IEF sampl~= at the acidic end, while IEF (C, D) was carried out. for 20 hours at 400 V applying the sample at the other (basic) end. Brackets enclosed the T3-associated species.
Cells (2x10') were preincubated for 1 hour at 37°C in 4 ml methionine-free RI?MI 1640 supplemented with 10% fetal bovine serum. 35S-methionine was added to 250~,Ci ml-1 and incubation was continued for 1 hour at 37°C. Cells were collected, washed and lysed in 0.4m1 of boiling solution of to SDS, lOmM T~__~is-HCl (pH 8.0), 0.1 mM PMSF and 10 mM
IAA. Lysates were diluted with 1.6 ml of 2.5o Nonidet-P40, 1% gelatin, .LO mM 'Tris-HCl (pH
., T Ji.

8) and 0.2 ml, of 1 mg ml-1 DNase. 0.5 mg ml-1 RNase, 0.5M Tris-HCl (pH i') , 50 mM MgCl2 and incubated at 0°C
for 2-4 hours. Afi:er centrifugation for 15 minutes at 12,000 x g to pellet insoluble debris, immuno-precpitations with anti- Y serum were performed using protein A sepharose preincubated with 1% gelatin and washing as described (65) . Elution from the immunoabsorbent and treatment with endo-H (Miles Scientific, rlaperv:ille, IL) were as described (65) .
Samples were ana7Lysed with anti-serum by two dimensional gel electrophoresis employing NEPHGE in the first dimension and 10% SDS-PAGE in the second dimension, followed by fluorography (66) .
Rearrangments of the Y and s genes in T cells exnressinq the T$R 'r polypgptides Genomic DNA was isolated as described (BamHI or EcoRI), size-fractionated o~n 0.7& agarose (BamHI digests) or 0.9% agarose (EcoRI digest), and transferred to Filters were nitrocellulose as described (67) .
hybridized to a nick-translated 32P-labelled 0.8 kb HindIII-EcoRI JY 1~~3 pr°be (20) or a 1.1 kb EcoRI-HindIII Cs2 probe (68) . Filters were washed in 2x SSC
and 0.1% SDS tollowed by 0.2% SSC and 0.1% SDS at 55oC
before autora<9iography with intensifying screens.
Cytolysis by IDP2 a:nd PBL C1 cells Cytolytic assays were performed in round-bottom 96-well tissue culture plates with 5lCr-labelling, harvesting and calculation of' % specific release as described (34) . IDP2 or PSL C1 cells were either preincubated with UCHTl (1:300 dilution) (+anti-T3) for 30 minutes at 0°C, washed x 3 or mock incubated and placed a34~o42 _ together with labelled target cells. Anti-HLA Class I
and Class II mAb and anti-T3 mAb were placed in wells containing then SlCr-labelled MOLT 4 cells for 30 minutes at 0°C, then IDP2 cells were added at a 40:1 E:T ratio. All samples were assayed in triplicate, each experiment: was ;performed at least three times, and one representai:ive experiment of each is shown in Fig.
9.
'I 0 ;~ xa '34042 Ez~erimental ;Results A murine framework antiserum that recognizes the major-ity of human TCR ~l , s molecules has previously been reported (28). Subsequently, a murine monoclonal anti-s body, designated Framework 1 (sFl), that is reactive with shared determinants on the human TCR s chain was obtained (46). The sFl monoclonal antibody reacts with the majority of T3 positive (T3+) human peripheral blood lymphoc~rtes (PHLs) and is capable of immunopreci-pitating the 'rCR a , d heterodimer front all human T cell lines examined that have a , s T cell receptors and express the T3 glycoprotein. Immunoprecipitations from a panel of Z' cell lines using this monoclonal antibody demonstrate this reactivity as well as the heterogeneity of the TCR a and TCR s subunits from dif-ferent recept~~rs (fig. lA) . Like the framework anti-serum (28), this monoclonal antibody does not stain the surface of living T' cells, but will specifically react with both membrane and cytoplasmic T cell receptors 2p after partial solution of the lymphocyte plasma mem-brane with 70% ethanol. Double staining of human PHLs with a fluorescein- anti-T3 monoclonal antibody and a biotinyl- sFl monoclonal antibody followed by PE avidin reveals that t:he sFl monoclonal antibody recognizes 95-97% of peripheral blood T3+ lymphocytes. However, it clearly def in~es a small population of T lymphocytes that is ~F1 negatives ( ~E'1 ) , yet T3+ (Fig 1C) .
A second framework monoclonal antibody designated WT31, initially thought to recognize the T3 antigen (29) , has recently been shown to react with a common epitope of human TCR n , ~~ (30) . While double staining with an anti-T3 monoclonal antibody (ORT~3) and WT31 revealed that each of these monoclonal antibodies cross-block binding of the other, one-color fluorescence indicated that WT31 typically recognized 1-38 fewer cells in peripheral blood than. do anti-T3 monoclonal antibodies.
The WT31 monoc7.ona1 antibody efficiently binds to the surface of T cells (such as in FACS analyses) and is capable of immunoprErcipitating the TCR a ,s molecules, albeit inefficiently, from radiolabeled detergent lysates (30) (F:ig 1H, lane 3) . Thus, the sFl monoclonal antibody and the WT31 monoclonal antibody appear to recognize all but a small fraction of human peripheral blood T3+ cells,, and define a subpopulation that is T3+
but unreactive with twth of these framework monoclonal antibodies against the TCR a , ~ molecules. Ebidence that the T3+ lymphocytes that are unreactive with the monoclonal antibody ~sFl are also unreactive with the monoclonal antibody WT31 is shown below. WT31 was used primarily for FRCS analyses and eFl was used primarily for immunoprecipitation studies.
Efforts at growing the WT31 T3+ population from normal adult PBLs proved difl;icult, since the wT31+T3+ lympho-cytes usually overgrew the WT31 T3+ cells following mitogenic stimulation. However, growth of the WT31-T3+
population from the fBLs of immunodeficiency patients was successful. Immunodeficiency patient 1 (IDP1) suffered from the bare lymphocyte syndrome and lacked class II MHC antigen expression on lymphoid cells (31, 32) , while immunodeficiency patient 2 (IDP2) suffered from an ectodermal d!~rsplasia syndrome (33) and dis played poor ~ vitro 9C cell proliferative responses to mitogens.
After activation of PSLs f ran IDP1 with alloantigen and propagation in c~~nditioned media containing . interleu-kin-2 ( IL-2 ) act:iv ity ( 34 ) , the resul taut cell 1 ine was ~34T042 observed to be approximately 50% WT31+T3+ and 50% WT31 T3+ (see Table I below, cell line 1) . Subsequent sort-ing of this cell :line yielded homogeneous populations of WT31+T3+ cells and WT31 T3+ cells (see Table I be-low, cell Iin~es 2 and 3, respectively).

% POSITIVE
CELL LINE OSOS URCE CELL LINES WT31 ~ T4 T8 NUhBER DESCRIPTION
1 IDP1 alto 50 100 11 50 2 IDP1 sort 100 100 70 28 WT31+

3 IDP1 _ 0 100 0 62 WT31 sort 4 IDP2 fresh PHL 61 63 38 16 6 IDP2 allo 2 100 0 43 1Ce11 line description indicates the conditions for activation or source of lymphocytes. WT31+ and WT31-sorted cell liners 2 and 3 (sort) were obtained by flourescence~ activated cell sorting of IDP1 cell line 1.
Cell lines were also obtained from IDP2. Fresh PBLs from IDP2 revealed that 63% of the PBLs were T3+ and 1-3% fewer cells (61'd) were wT31+, which is typical of normal PBLs (Table I, cell line 4) . Activation of these IDP2 PBLs witlh either phytohemagglutinin (PHA) or alloantigen and propagation ~n vitro with conditioned media resulteci in several cell lines. These included a homogeneous W~~31+T3~~ cell line (Table I, cell line 5) , a homogeneous WT31-T3+ cell l ine (Table I, cell l ine 6) and on a third occasion, a cell line that was 88% WT31 T3+ (with 12% contaminating WT31+T3+ cells) (Table I, - 1 34 ~ 04 2 cell line 7) . The W'T31 T3+ population both contained T4 T8+ and T4 T~9 cells (Table cell lines 3, and I, 6 7). Further phenotypic analysis revealed that this population was T11+ but negative for natural ki ller cell markers such as l~eu 7, Leu and ORM1 and the 11 for c~ immature thymocyte marker T6.

The sFl monoclonal antibody immunochemically defined a heterodimeric structure on the surface of 1251-labeled WT31+T3+ IDP1 lymphocytes (Fig. 2A, lane 5) , yet failed 0 to recognize a >>imilar protein on the WT31-T3+ popula-tion from this same individual (Fig. 2A, lane 11) .
Similar analysis of II)P2 cell lines revealed a trace of TCR a , s on the 88% wr31-T3+ cell line 7 (Fig. 28, lane 2) consistent ~~rith the 12% contamination with the ~5 WT31+T3+ cells. Thus., the WT31 T3+ cells, identified by the lack of cell surface reactivity with the WT31 monoclonal antibody in FRCS analysis, were also BF1-, as determined by the lack of TCR a , a on immunoprecip-itation. All WT31+T'3+ and WT31 T3+ cell lines ex-20 pressed similar amounts of T3 by FRCS analysis and by immunoprecipitat,ion with an anti-T3 monoclonal antibody (Fig. 2A, lanes 3 and 9; Fig. 2C. lanes 2 and 4) .
However, the T3 molecule found on WT31- BF1-T3+ lympho-cytes was not identical to the T3 molecule found on 25 WT31+BF1+T3+ cells by SDS-PAGE. One-dimensional (Fig.
2C) and two-dimensional gel analysis indicated that the difference in 't3 was restricted to the light T3 subunits, which reproducibly displayed different SDS-PAGE mobilities (Fig. 2C, arrowhead) .

To determine if the WT31-sFl-T3+ population lacked TCR a , a molecules, or alternatively expressed- TCR a , s molecules that failed to react with these monoclonal antibodies, the presence of mRNAs encoding the TCR a ,~

1341 p42 and s proteins was investigated. 32P-labeled cDNA
Clones encoding TCRa, TCR(3, and TCR~ were used to probe Northern blots containing whole cell RNA from WT31- sFl-T3+ and WT31.+ BF1+T3+ IDP2 cell 1 fines and f rom HPB-hQ,T, which is known to contain mRNA for TCR a , TCR a and TCR 'r, No TCR a or TCR a mRNA transcripts could be detected in the RNA from the WT31 sFl-T3+ IDP2 cell line 6 (Fig. 3A-~ probe, lane 1; or s-probe, lane 1) , whereas exF~ression of both was clearly detectable in RNA from HPB-MLT (Fig. 3A o-probe, lane 2; and a-probe, lane 2) . Notably, TCR r mRNA was present in the WT31-T3+ cells <it levels comparable to that in APB-lrB.T
(Fig. 3A r-probe,. lanes 1 and 2) . Thus, the WT31 sFl T3+ lymphocytes lacked TCR a and s mRNA. Subsequent experiments on cell lines that were mostly WT31 T3+
~5 corroborated -these results. For example, Northern blot analysis performed on IDP2 cell line 7 (88% WT31-T3+) and compared with IDP2 cell line 5 (WT31+T3+) , as well as with HPB-t~,T cells, revealed only a trace of TCRa or TCR s mRNA in the 8E3% WT31-T3+ cells (consistent with 20 the 12% contamination with WT31+T3+ cells) (Fig. 3B, lane 2 for each probe) . Further, the majority of the a transcripts that could! be detected were 1.0 and not 1.3 kb and were prot>ably nonf unctional ( 35) . In contrast, the IDP2 cell 1. fine .°> (WT31+T3+) expressed level s of 25 both RNA species which were comparable to HPB-MLT (Fig.
3B, lane 1 for each probe). However, like the WT31-T3+
cell line shown in Fig. 3A, both the WT31-T3+ and the WT31+T3+ cell lines showed TCR ~ RNA levels comparable to HPB-IrQ,,T (Fig.3B, y~-probe) . Thus, the WT31-T3+ cells 30 lacked a and a T cell receptor mRNA (Northern analysis) and a and s T cell receptor proteins (immuno-precipitation and FAC;S analysis) . The presence of TCR~ mRNA in WT31-T3+ cells, while consistent with TCR Y
protein expression, could not be taken as strong evi-1341p42 dence for this, since many human cell lines that ex-p r a s s TCR ~ mRNA of normal size may express full length transcripts that. are out of frame due to defective V-J
joining (36) .
c; To determine if proteins analogous to the TCR a , s molecules existed on the WT31 eFl T3+ cells, the technique of chernical cross-linking was utilized.
This procedure has been used to show directly the physical association of the TCR a , 9 molecules with the T3 glycoprotein (12) . The bifunctional, cleavable reagent, dithio-bis-succinimidyl propionate (DSP) was employed to cross-link 1251-labeled surface proteins of viable T lymphocytes. After cross-linking, the lympho-cytes were solubil iz~ed in a non-ionic detergent and ~5 immunoprecipitated with an anti-T3 monoclonal antibody.
As expected, they WT3l~~sF1+T3+ lymphocytes revealed that the TCR a and s chair,~s were cross-linked to T3. For example, TCR a , a molecules and T3 were found in anti-T3 or sFl monoclonal antibody immunoprecipitates from 20 cross-linked IDP1 cell line 2 (WT31+T3+) (Fig. 2A, lanes 4 and 6) . However, despite the lack of reactivi-ty with the eFl mono<:lonal antibody and lack of TCR
or TCR smRNA, II)P1 cell l ine 3 (WT31-T3+) and IDP2 cell line 7 (88% WT31-T:3+) both expressed two protein 25 subunits (Mr °_.5,000 and 40,000) that specifically cross-linked to T3 (Fig. 2A, lane 10; Fig. 2B, lane 6). The mobilities of these T3 associated molecules were clearly different from those of the TCR a ands chains f rom WT31.+T3+ cell l fines (compa re Fig. 2A, lanes 30 4 and 10; or Fic~. 2B, lanes 5 and 6).
Since IDP2 cell. lines 7 (88% wT31-T3+) contained 12%
WT31+T3+ cells, accounting for the weak sFl immuno-precipitates noted (fig. 2B, lane 2) , the lysate f rom _4,_ 1 3 41 Q 4 2 these cells was F>recleared of TCR a, a protein using the sFl monoclona7l antibody. After preclearing, no residual sFl reactive material could be detected (Fig.
2B, lanes 8 and 11). When this sFl- precleared lysate from cross-linked ~~ells was immunoprecipitated with an anti-T3 monoclonal antibody, Mr 55,000 and 40,000 subunits were still detected (Fig. 2B, lane 12).
Since these WT31- ~?1 T3+ cell lines display undetect-able levels of TCR a and TCR s mRNA, the molecules found specifically cross-linked to T3 on their cell surfaces cannot represent proteins encoded by the known TCR or TCRS genes.
cDNA clones representing the rearranging human TCRY
~5 gene would encode a~ polypeptide with a predicted molec-ular weight of 40,,000 dal tons (36) . However, unlike the murine TCR r gene, which does not reveal any N-linked glyco~;lyation sites (15) , the human TCR Y gene reveals five potential sites for N-linked glycosyl-20 ation, four of which are located in the constant region (36). Since a TCR r protein has not previously been isolated, it is not know n how many of these potential sites may be used. However, a fully glycosylated hu-man TCR Y protein may have a Mr of about 55,000. The 25 heavy chain of the non-a-non-sT3-associated subunits identified on the WT31-sFl T3+ IDP1 and IDP2 cell lines has a relative mobility on SDS-PAGE of 55,000 daltons (Fig. 2A and :2B) .
30 In order to determine if this T3-associated heavy chain was serologically cross-reactive with or identical to the TCR r protein, antisera were raised to a synthetic peptide having the sequence RTRSV TRQ TG SSAE ITC

1341042 _ (representing a 17 amino acid stretch of residues 5-21 from the variable region; anti-V r peptide serum) and to a synthetic peptide having the sequence:

(representing a 20 amino acid stretch of residues 117 136 from the constant region; anti-C Y peptide serum) of the TCR Y amino acid sequence deduced from a human cDNA clone (?.6) . Both the anti-C r peptide serum and anti-V r pepti.3e serum immunoprecipitated a molecule with Mr 55,000 from the denatured lysate of 1251 labeled WT31 sFl T3+ cells ( Fig. 4A, lanes 2 and 4) .
Such molecules cou~.ld not be immunoprecipitated from lysates of 1'~5I-labeled HPB-MLT cells, which express only nonf uncti.onal 'PCR Y mRNA ( 36) To demonstrate that the 55,000 dalton molecule immuno-precipitated by the anti-C Y and anti-V Y peptide sera was, in fact, the heavy chain subunit that cross-linked to T3, an additional experiment was performed (Fig.
4B) . A sample of DSP cross-linked lysate from the WT31- sFl T3+ cells was f first immunoprecipitated with an anti-T3 monoclonal antibody, again demonstrating the presence of Mr 55,000 and 40,000 subunits associated with T3 (Fig. 4B, lane 2) . In parallel, another ali-quot of the cross-linked lysate was immunoprecipitated with an anti-T3 monoclonal antibody, and the immunopre-cipitated T3 cross-:linked polypeptides were eluted f rom the immunoabs<>rbent under denaturing and reducing con-ditions in order to break the DSP cross-link. This eluate was then reprecipitated with anti-C r peptide serum. The Mr 55,000 subunit that cross-linked to T3 was re-precipitated by anti-rpeptide serum (Fig. 4B, lane 5), indicating that the Mr 55,000 subunits defined by these two approaches were identical.

1341042 -, Immunoprecipitations from lysates of surface-iodinated IDP2 lymphocytes using anti-T3 mAb (under conditions that do not dissociate TCR subunits from T3, see Fig.
5) yielded two species (55R and 40R) in addition to the T3 subunits (Fig. 5A). This result is identical to the one reported previously using chemical cross-linking.
The 55R species was shown to react specifically with anti-C Y and anti-V r peptide sera. The 40R polypeptide was unreactive with these anti-Y peptide sera and is thus likely t:o rep;resent a non TCR a , s or Y subunit, namely s . To determine if these subunits are covalently linked, like the TCR a and s subunits, the T3 co-immunoprecip~itated polypeptides were examined under reducing and nonreducing conditions. In striking contrast i.o the TCR a~s subunits, which exist in a heterodimeric disulphide-linked form under nonreducing c~~nditions, the TCR r and 6 subunits on the IDP2 cell line are not covalently linked (Fig. 5A). A
small increase in relative mobility on SDS-20 polyacrylamide gel electrophoresis (PAGE) under nonreducing conditions was observed for the diffuse, heavily glycosylat~ed (see below) TCR Y , whereas a dramatic increase i.n moblility was observed for the a subunit, suggesting the presence of one or more 25 intrachain disulphide loops (compare species at arrows, lanes 2 and 4) .
Weiss, et al. suggeasted that the PEER cell line might express the TCR Y polypeptide since it lacked expression 30 of TCR mRNA yet expressed a T3-associated 55-60R
polypeptide (69) . On further examination, this cell line was found to lack reactivity with a mAb recognizing framewrork determinants on the TCR 9 chain, sFl (Fig. SA) and to express a strongly iodinated 38R polypeptide. The 55-60R polypeptide was specifically immunoprecipitated with anti-C Y peptide sera and thus appears to represent a further example of the TCR r protein (Fig. 5A) . The TCR Y and a polypeptides on PEER were of similar size to those on the IDP2 cell line and similarly were not disulphide-linked. Like the d subunit on IDP2 cells, the counterpart molecule on PEER underwent a marked shift in SDS-PAGE mobility when compared under reducing and nonredu~ing conditions (compare species a!t arrows, lane 7 and 8) . Thus the IDP2 and the PEER <:ell lines appear to express similar types of TCR r ~ a -T3 complexes, in which the TCR r and d subunits are :not covalently linked.
We wished to determine if this second TCR was also ~5 expressed as a component of the T cell population in normal peripheral blood. Two-colour cytofluorographic analysis comparing staining of human peripheral blood lymphocytes ;PBL) with mAb BF1 and ORT~3 showed a discrete population representing 2-5% of the T3+ PBL
20 that appeared to bEa TCR a, s negative. To examine this lymphocyte population, normal adult PBL were subjected to cytofluorographic cell sorting of ter staining with mAb WT31. Unstained PBL were isolated and propagated vitro in I1~-2-containing conditioned media receiving 25 biweekly additions of irradiated autologous feeder cells and phytohaemagglutinin (PHA-P) . The cell line derived, WT31~~PBL :LINE was cloned by limiting dilution with plating at 0.5 cell well and the cloned cells were propagated as for the polyclonal cell line. Several 30 such peripheral blood derived T cell clones were obtained, and PBL Clone 1 (PBL C1) was studied in detail. By cytofluorographic analysis, this clone was T3+Tll but T4 T8~ and WT31 . The expression of TCR a , ,~ and'. Y mI~NA f rom PBL C1 was determined by ~ 34? 042 _ Northern blot analysis (Fig. 6). By comparison with the WT31+ BFli' T cell tumor HPB-l~T, only very low levels of TCR a and. s mRNA were detected. In contrast, abundant TCR r rnRNA was noted (Fig. 6 Y probe);
interestingly,, the TCR r mRNA was slightly smaller than the 1.6 kilot~ase (;kb) message found in HPB-hff.T and in the TCR r-expressing IDP2 cell line (Fig. 6).
Consistent with these observations, WT31 reactivity was not detected in cytofluorographic analysis (data not shown) and only scant levels of TCR a and s polypeptides were found by immunoprecipitation using mAb sFl (Fig. SB, lanes 2, 5). In contrast it is likely that the trace levels of TCR a and s protein detected in PBL C:1 are accounted for by the 1-2$
contamination with irradiated autologous feeder cells ~5 used in the propagation of this clone. Two abundant chains (40R and 3610 were observed associated with T3 under reducing conditions in SDS-PAGE analysis (Fig.
5B, lane 3). Anti-C Y sera immunoprecipitated both of these polypeptides from reduced and denatured PBL C1 20 lysates (Fig. 5B, 1<ine 8) .
To determine if these 40R and 36R TCR Y polypeptides were part of .a disulphide-linked dimer, co-immunoprecipit:ation~a with anti-T3 were examined under 25 nonreducing conditions. A single band of Mr (FOR) was observed indicating that, unlike IDP2 and PEER cells, PBL C1 expresses a T3-associated TCR r gene product that is part of a d:LSUlphide-linked dimeric complex.
30 As a TCR r partner ( a) was present on the non-disulphide linked form of this receptor complex on IDP2 and PEER cel:Ls, we examined whether the disulphide-linked form of the receptor on PBL C1 was composed of a homo- or a heterodimer. Immunoprecipitates were analysed by two-dimensional gel electrophoresis (nonequilibriwn p13 gel electrophoresis (NEPHGE) followed by ;SDS-PAGE; Fig. 7A, B) . Under reducing conditions, both the TCR Y species (40R and 35R) were found to have identical charges, and displayed heterogeneity typical for a sialylated glycoprotein.
These characteristics are like those described previously for f~ifferentially glycosylated TCR a polypeptides having the same amino-acid backbone.
Thus, these species may represent differentially glycosylated i:orms of the same TCR Y peptide. This conclusion is supported by the results of metabolic pulse-labelling (below) which reveal only a single precursor TCR Y species in PBL C1.
~5 A disulphide-7. inked dimer composed of one or both of these TCR Y species should have a focusing position similar to eiither of the two components alone when analysed by N~:PHGE or equilibrium isoelectric focusing (IEF). But a heterodimer compose d of TCR r and a 20 distinct polyF~eptide~ might have a different charge and focusing position. The position of the disulphide-linked dimer was therefore examined by carrying out NEPHGE under nonreducing conditions, followed by SDS-PAG E under nonreducing conditions (Fig. 7B).
25 Strikingly, the position of the disulphide-linked dimer was substanti,311y more acidic than that of the TCRr polypeptides examined under reducing conditions (compare the ~~OR and 36R species in Fig. 7A with the 70R species in Fig. 7B) . This result suggests that the 30 TCR 'r species were covalently linked to a polypeptide of distinct N1~PHGE mobility. Thus, although a TCR
partner could not be directly visualized (either because it w,as inadequately labelled with 1251 or because it didl not resolve in the focusing system used here) , the TC'R Y polypeptide on PBL C1 appeared to be expressed as .a part. of a disulphide-linked heterodimer.
Experiments using eaquilibrium IEF (rather than NEPHGE) confirmed thia observation (Fig. 3D).
A further distinction between the disulphide-linked and non-linked forms was the size of the mature TCR Y
glycopeptide (55-60R on IDP2 and PEER versus 40R and 36R on PBL C1) . To assess how much of this radical size difference is due to differential glycosylation and how much to different peptide backbones, TCR Y
peptides were analysed in cells pulse-labelled with 35S-methionine~. After solubilization under denaturing and reducing con ditions, the lysates were immuno-precipitated 'with anti-C Y sera and examined by two-~5 dimensional gel electrophoresis (Fig. 7E-H) . Immuno-precipitates were either treated with endoglycosidase H
(endo-H) to remove immature high-mannose glycans from pulse-labelled material, or were mock treated. Two TCR Y polypeptides (46R and 43R) of identical NEPHGE
20 mobility were synithesized by the IDP2 cell line.
Treatment with endo-H reduced both forms to a 40R form, suggesting that the 46R and 43R forms carried different numbers of ~~arbohydrates, and that a single TCRY
polypeptide backbone (40R) was synthesized by IDP2 25 cells (Fig. 7E, F). In contrast, a more basic, 38R
glycosylated form was synthesized by PBL C1, which after endo-H digestion displayed a nonglycosylated 31R
peptide backbone (Fig. 7G, H). Thus the TCR Y
polypeptides on the non-disulphide-linked (IDP2) and 30 the disulphide-linked (PBL C1) forms characterized here have radicall;r different peptide backbone sizes (40R
and 31R respec:tivel;~r) . The fact that the glycosylated TCR Y peptides observed by pulse-labelling are of different molecular weight than those found by cell -54- ~ ,3 4 ~ Q 4 2 surface iodination presumably results from the different tyF~es of carbohydrates they carry, namely high-mannose versus complex.
We next wished to determine if both a disulphide-linked and a noncovalently associated form occurred in normal adult peripheral blood. The polyclonal peripheral blood cell line (wT31-PBL LINE) from which PBL C1 had been cloned was therefore studied in greater detail.
WT31 PBL LINE was ;homogeneously T3+T11+ and contained 95% WT31 T4 T8 with 5% contaminating WT31+ cells.
When examined by immunoprecipitation from iodinated, solubilized cealls, weak but detectable reactivity with mAb sFl was observed (Fig. 5B, lanes 10 reduced and 13 nonreduced), consistent with the expected 5%
TCR a~s positiv~a lymphocytes. In contrast, anti-T3 mAb immunoprecipitated large amounts of both T3 and associated polypeptides of 35-45R under reducing conditions (Fig. 5B, lane 11). To determine what fraction of these were disulphide-linked, the T3 immunoprecipitate was examined under nonreducing conditions (Fi.g. 5B, lane 14) . Less than half of the T3-associated polypeptides were disulphide-linked.
This material included disulphide-linked TCRn,apeptides located above the open arrow, lane 14 (size identified by the sFl precipitate, lane 13) and disulphide-linked TCR Y peptides of smaller size (open arrow, lane 14) .
Strikingly, the majority of the T3-associated species were not disulphide-linked and migrated with the same mobility under both reducing and nonreducing conditions. Notably, a fraction of these non-linked species displayed a marked increase in SDS-PAGE
mobility under' nonreducing conditions, similar to the TCR a on the IDP2 and PEER cells (see Fig. 58, lane 14, solid arrow) . Reac;tivity with anti-C Y sera confirmed that most of the labelled material associated with T3 expressed on wT3l~-PBL LINE was TCR Y gene products (lane 16) .
Thus, the protein product of the TCR Y gene occurs on T3+ 1 m hoc ~tes in adult y p y peripheral blood in both disulphide-linked and unl inked molecular forms.
Moreover, the non disulphide-linked form of TCR Y may be further divided into 55-60R glycosy fated (IDP2 and PEER) or 35-4.5R glycosylated (thymic T cell clone C11 (70) and WT31~-PBL LINE) species.
TCR Y and s gene rearrangements were examined in T
cells known to express the TCR Y polypeptide on their cell surfaces. Southern blot analysis were carried out ~5 using the 0.8 kb EcoRI-AindIII human J
Y1,3 Probe (nomenclauturn according to Quertermous et al. (71) ) .
This probe detects germline bands of 23kb and l2kb in a BamHI digest of genomic DNA. The 23 kb band encompasses C n and the 12 kb band encodes C Y2. Using 20 this probe, IDP1, PBL C1 and PBL C2 (also derived from the WT31-PBL LINE) showed rearrangements of the TCR Y
gene (both PBL C1 and PBL C2 displayed an identical rearrangement,; Fig. 8A).
25 Seven rearrangements in PBL using the J probe and Y1,3 EcoRI-digestec9 genomic DNA in Southern blot analyses have been detected (20) . Six (I, II, III, IV, VII and V) of these seven rearrangements are shown in PBL, fetal thymus, and newborn thymus genomic DNA (Fig. 8B;
30 see arrows and rearrangement numbers). Four re arrangements (I, II, VI and VII) either are not used by peripheral blood lymphocytes which express the TCR Y
polypeptide o~: cells demonstrating them were lost under the propagation conditions used for the WT31 PBL LINE.

13~~ X42 _ Nevertheless, the 1aT31 PBL LINE DNA revealed at least three of there rearrangements (III, IV and VI) (Fig.
8B, lane 3) and these same rearrangements were used by IDP2, PBL C1 and PBL C2 (data not shown for the EcoRI
digest) and <ill of these rearrangements are displayed i n f etal thymus .
The TCR s gene was also rearranged in IDP2, PBL C1 and PBL C2 cells. The 1.1 kb EcoRI-HindIII C s2 probe detects a ger;mline band of 20 kb which encompasses both C s constant regions in a BamHI digest of genomic DNA
(68) . One predominant TCR s rearrangement for IDP2 and two identical rearrangements for PHL C1 and PBL C2 were observed (Fig. 8C). It is assumed that these TCR s rearrangements are nonproductive based on the ~5 immunoprecipi~tations and Northern analyses f or these cell lines. As both PBL C1 and PBL C2 have the same TCR Y and s rearrangements, they appear to be clonal and derived from the same cell within the WT31 PBL
LINE.
As TCR Y-e:Kpressing cells were found in adult peripheral blood, :functional studies were carried out to determined whether they have effector capabilities.
When IDP2 and PBL C'1 were examined for their ability to lyse target cells in 5lCr release assays, they proved to have spontaneous effector cytotoxic capability (Fig.
9). Although the IDP2 cell line did not lyse the majority of natural. killer (NR) targets or PHA blasts of allogeneic: PBL, they were selectively capable of lysing 5lCr-labelled MOLT-4 cells (Fig. 9A top). In two of six aimilar assays, weak lysis (10-15$ 5lCr release) of 8:562 targets was also observed. Lysis of MOLT-4 cells was .not inhibited by a variety of mAb directed against monomorphic MHC Class I (W6/32 anti-1 3 41 04Z _ -ALA-A, B, C, 4E and 131) or Class II (LB3.1 and anti-Leu 10) determinants (Fig. 9B) , although we have previously found that these mAb efficiently block killing by both MAC Class I and Class II allospecific CTL (34) . ".these data suggest that lysis of MOLT-4 cells was MHC class I and II independent. Only anti-T3 mAb partially blocked the specific lysis of MOLT-4 cells (Fig. 9B) . On the other hand, when triggered by prcbinding o:E ant.i-T3 mAb to IDP2, as has been previously reported for thymic-derived CII7, 5lCr-labelled target cells that express Fc receptors for IgG
(for example, U937) , were efficiently lysed (Fig. 9A, +anti-T3). Such killing could be completely inhibited by aggregatec9 human IgG, conf firming that this T3-mediated lysis occurred through a mechanism of enhanced ~5 conjugate formation via IgG Fc receptors (data not shown) . The paradoxical augmentation of lysis by anti-T3 mAb for :>ome targets (U937) and the blocking of lysis for specifically recognized targets (MOLT-4) might result from the competing effects of triggering 20 and increasing conjugate formation via T3 but sterically blocking antigen recognition via the TCR.
PBL C1 proved a more efficient killer cell than IDP2.
PBL C1 displayed spontaneous cytolytic activity against 25 8562 cells (MFiC Class I and II negative) showing nearly 508 specific ~'1Cr relsease when examined at an effector target (E: T) ratio of 20:1 (Fig. 9C top). Moreover, PBL C1 also lysed MOLT-4 cells and to a lesser extent, CEM cells. No lysis of Daudi, 0937, or either 30 autologous or allogeneic PBL was detected. Triggering with anti-T3 mAb induced PHL C1 to lyse the 0937 cell line. Further, lysis of 8562 was slightly augmented while that of MOLT-4 was partially inhibited (Fig. 9C).
Taken together, the spontaneous cytolytic activity of IDP2 and P8L C1 on tumour targets such as 8562 and MOLT-4 and the fai).ure to block such activity by anti-MHC mAb indicates that these TCR Y lymphocytes are non-MHC class I and class II restricted cytotoxic T
lymphocytes.

Framework monclonal antibodies against the TCR a, B
molecules, a F1 and wT3l, were used to identity and isolate the wT3l~- BF1 T3+ lymphocyte population f rom the peripheral blood lymphocytes of two immunodeficiency patients. By the criteria of both immunoprecipitation analysis with framework monoclonal antibodies and Northern blot analysis using TCR a and TCR a specific cDNA probes, polyclonal human T cell lines of this 0 phenotype were chown to express neither TCR a , s mRNA
transcripts nor polypeptides. Nevertheless, chemical cross-linking studies using the cleavable DSP reagent revealed the existencer of a protein complex associated with the T3 glycoprotein on the surface of these cells.
The heavier of the two subunits that cross-linked to T3 (Mr 55,000) was also immunoprecipitated by two differ-ent antisera, one generated against a 17 amino acid synthetic peptide corresponding to a part of the vari-able region and another generated against a 2G ammo 20 acid synthetic F~eptidea corresponding to a part of the constant region ~of the deduced amino acid sequence of a rearranged TCR y gene (19, 36) Thus , the Mr 55,000 protein is the TCR r F>rotein encoded by the rearranged TCR ~ gene (15) . The !!!r 40,000 polypeptide is a fourth 25 T3-associated protein designated TCR a (Fig. 2A and 2B) .
The TCR Y and TCR a polypeptides form a T3-associated heterodimeric structure on these cells (TY,6 -T3) that is analogous to the previously described T cell recep-tor complex (TCR a , B ) .
The TCR Y lymphocytes examined here exhibit non-MBC
restricted cytolytic activity and may be similar to other T3+ NR-like cells whose T-cell receptors have not yet been detinit:ively characterized (39, 72, 73, 74).

As NR-like lymphoyctes, they may participate in host immune surveillance against malignancy. The specificity of lysis observed suggests that the possibility of TCR Y mediated antigen-specific recognition of some but not all tumour targets. As anti-T3 mAb could trigger nonspecific lysis of some target cells or alternatively block specific lys:is of other targets, the T3 molecule on these cells appears to be functional.

Example 2 Materials and Methods Cultura method The Peer cell line described hereinabove was cultured in vitro in a medium composed of RPMI 1640, 10$ fetal calf serum, penicillin-streptomycin, and L-glutamine.
The culture w~~s feel twice a week and was kept at 37°C
in a humidif ie~d incubator with 5$ C02 .
Hybridoma production for monocloanl antibodies specific f or the TCR Y chaff n A BALB/c mouse was immunized intraperitoneally (I. P.) with 2 x 107 Peer cells suspended in 0.2 ml of phos-phate buffered saline (PBS) . The mouse was boosted by I. P. injection every 10 days with 2 x 107 Peer cells for a total of 20 injections. Three days before fu-sion, the mouse was boosted by intravenous (I.V.) in jection with :? x 107 Peer cells for 3 sequential daily I.V. injections. The mouse was sacrificed and the spleen was rernoved at the last I.V. injection. Immune spleen cells were :Fused with mouse myeloma cell P3 x 63Ag8.653 in the presence of polyethylene glycol 1500 at the ratio of 5:1 by standard procedures. After fusion, cells were suspended in the culture medium containing hypoxanthine ( 1 x 10-4, ) , aminopterin ( 8 x 10-8M) , and tlaymidine (1.6 x 10-5M) and plated at 2 x 10 cells per well in microliter plates which contained 2 x 105 BALB/c thymocytes per well as feeder cells.
The cultures were fed with the same medium on day 7.
Beginning on day 14, cultures were fed with the same medium lacking amino pterin.

134~~42 ~y_bri,~oma scree;nin4 ffor monoclonal antibodies snecif ~
for the TCR ~ chain Since both the r andl s chains of the T cell antigen !i receptor protein on 1?eer cells are complexed with CD3 antigen, antibodies against the Peer T cell antigen receptor should be able to co-modulate these surface proteins with a,n anti-CD3 monoclonal antibody such as ORT~3 (2). Such co-modulation was employed as the primary screen of desi rable hybridomas as follows.
Each of the hybridonna culture supernatants was har vested and screened for its ability to co-modulate surface CD3 protein complexes with an anti-CD3 monoclonal antibbdy. One hundred microliters of a cul ture supernatant: were added to each well of a 96-well microtiter plate containing 5 x 105 Peer cells per well. After overnight incubation at 37°C, fluorescein isothyiocyanate-conjugated ORT~3 was added to each well and cultured for an additional 30 minutes at 0°C.
Samples were then analyzed by flow cytometry.
Supernatants which induced a significant decrease in fluorescence intensity were selected and further characterized by the immunoprecipitation methods described below" The cells in selected wells which secreted anti-human 1' cell antigen receptor proteins were subsequently cloned by the limiting dilution method.
3() 3a Immunoprecipitation Peer cells were radio;labeled with 1251 and sotubilized in Tris bufferec9 saline (TBS) containing 18 Nonidet P-40 as described herei.nabove. Immunoprecipitation was performed by incubating 1251-labeled Peer cell lysates C.

_ -63- 1 341 Q4z _ with each of the selected supernatants under reducing conditions. Afvter immunoprecipitation, the samples were analyzed by 10$ sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). The gels were dried a.nd autoradiographed, and the molecular weight of ~~roteins was determined by comparison with molecular weight standards (75).
Production and screening of hybridomas which produce a monoclonal antibody specific for the TCR a chain Monoclonal antibodies were made by immunizing BALB/c mice with immunol?recipitated TCR Y. ~ -CD3 from the Peer cell line. Briefly, 1 gram of Peer cells was solubilized iri 0.3$ CHAPS detergent and immunopreci;pitated with 5 microliters of UCHT1 ascites and fixed Staphylococcus aureas C.Jwan I strain bacteria as the irnmunoadsorbant similar to the procedure described in (46). The washed immune complexes were injected intraperitoneally at 4 week intervals for a total of 5 immunizations ( 46 ) . The mice were then sacrificed and the spleen cells fused to P3X63Ag8.653 myeloma cells. The hybridomas were grown in HAT
selection, screened and characterized by immunoprecil?itati~on on 1251-labeled Peer cells and other cells as described in (75) .
Results As shown in Fig. 10, Lane 4, the antibody in hybridoma 34D12 supernate immunoprecipitated a 55 Kd protein and a 20Rd protein under reducing conditions from the iodinated lysate of Peer cells. This 55 kd protein corresponds to the Y chain of the T cell antigen receptor and the 20 kd protein of the T3 protein on Peer cells.

-64- 1 3 4 1 0 4 Z _ In a separate experiment a monoclonal against the T
cell antigen receaptor s chain, i.e., 4A1, was produced and characterized. As shown in Figure 11, lane 5, 4A1 specifically reimmunoprecipitated a T cell antigen receptor a chain 1:~0 kd) from IDP2 cells (75) . 4A1 has also been shown t:o immunoprecipitate the T cell antigen receptor Y , a complex from several other T cell antigen receptor Y,d positive cell lines, including IDP2, Molt-13 and PBL line 2.

._ 1s41p42 REFERENCES
1 . All iso:n, J. P. , McInty re, B. W. & Bloch, D. J.
Immunol. 129, 2293-2300 (1982) .
2. Meuer, S. C. , Fitzgerald, K.A. , Hussey, R. E. , Hodgdon, J.C., Schlossman, S. F. & Reinherz, E. L. J'.Exp. Med. 157, 705-719 (1983) .
3. Haskin;s, K. , Kubo, R. , White, J. , Pigeon, M. , Kapple~r, J. & Marrack, P. J. Exp. Med. 157, 1149-1169 (7.983) .
4. Yanagi, Y. , Yoshikai, Y. , Leggett, R. , Clark, S, p,, Aleks,ander, I. & Mak, T.W. Nature 308, 145-19.9 ( 1984) .
5 . Hedr is k, S,. M. , Niel son, E. A. , Kaval er, J. , Cohen, D.I. & Davis, M. M. Nature 308, 153-158 (1984) .
6. Chien, Y. , l3ecker, D.M. , Lindsten, T. , Okamura, M., Ccrhen, D.I. & Davis, M. M. Nature 312, 31-35 (1984) .
7. Saito, H., Kranz, D.M., Takagaki, Y., Hayday, A. C. , Eisen, H. N. & Tonegawa, S. Nature 312, 36-40 ( 1984) .
g , Sim, C.. K. , '.vague, J. , Nel son, J. , Ma rrack. , P. , Palmer, E., Augustin, A. & Kappler, J. Nature 312, 771-775 (1984).
9. Reinherz, E. L. , Meuer, S. C. , Fitzgerald, R.A. , Hussey, R.',E. , Hodgdon, J. C. , Acuto, 0. &

1 3 41 p 42 Schl ossman, S. F. Proc. Natl . Acad. Sci U. S. A.
80, 47.04-4108 (1983) .
. Oettgen, H. C. , Kappl er, J. , Tax, W. J. M. &
5 Terhorst, C.J. Biol. Chem 259, 12,039-12,048 (1984) .
11. Weiss, A. & Stobo, J.D. J. Exp. Med. 160, 1284-7.299 (:L984) .
12. Brenner, M.B., Trowbridge, I.S. & Strominger, J. L. (:el l 417 , 183-190 ( 1985) .
13. Allison, J. P. & Lanier, L. L. Nature 314, 107-109 (7.985) .
14. Samel son, L. E. & Schwartz, R. H. Immunol . Rev.
81, 1_fl-144 (1984) .
15. Saito, H., Dranz, D.M., Takagaki, Y., Hayday, A. C., Eisen, H. N. & Tonegawa, S. Nature 309, 757-7fi2 (1984) .
16. Rranz,~ D. M. , Saito, H., Heller, M. , Takagaki, Y., Haas, W., Eisen, H. N. & Tonegawa, S. Nature 313, ',~52-755 (1985) .
17 . HaydaSr, A. C. , Sai to, H. , Gil l i es, D. , Kr anz , D. M., Tanigawa, G., Eisen, H. N. & Tonegawa, S.
Cell 40, 259-269 ( 1985) .
18. Lefranc, M-P & Rabbitts, T. H. Nature 316, 464-466 (:L985) .

19. Murre, C., Waldmann, R.A., Morton, C.C., Bongiovanni, K. F. , Waldman, T.A. , Shows, T. B. &
Seidman, J.G. Nature 316, 549-552 (1985) .
20. Quertermous, T., Murre, C., Dialynas, D., Duby, A. D. , Str ominge r, J. L. , Wal dman, T. A. &
Seidman, J.G. Science 231, 252-255 (1986).
21. LeFranc, M-:P, Forster, A., Baer, R., Stinson, M.A. & Rabbi.tts, T. H. Cell 45, 237-246 (1986) .
22. Iwamoto, A., Rupp, F., Ohashi, P. S., Walker, C. L. , Pi rch~er, H. Joho,R. , Henga rtner, H.
, &

Mak, T'.W. J. Exp. Med. 163, 1203-1212 (1986) .

23. Zauderer, M. , Iwamoto, A. & Mak, T.W. J. Exp.
Med. 163, 1314-1318 (1986) .
24. Vague, J., 'White, J., Coleclough, C., Rappler, J., P,~lmer, E. & Marrack, P. Cell 42, 81-87 (1985) .
25. Dembic, Z., Haas, W., Weiss, S., McCubrey, J., Keifec, H., von Boehmer, H. & Steinmetz, M.
Nature 320, 232-238 (1986) .
26. Hedrick, S.M. et al. Proc. Natl. Acad. Sci.
U.S.A. 82, 531-535 (1985).
27. Blanckmeiste~r, C.A., Yamamoto, K., Davis, M.M.
& Hammerling, G.J. J. Exp. Med. 162, 851-863 (1985) .
28. Brenner, M. B. , Trowbridge, I. S. , McLean, J. &
Strominger, J.L. J. Exp. Med. 160, 541-551 (1984) .

29. Tax, W. J. M. , Willens, H. W. , Reekers, P. P. M. , Capel, P"J.A. ~ Koene, R. A. P. Nature 304, 445-447 (198:!) .
3 0 . ~ Spi ts, H. , Bor st, J. , Tax, W. , Capel . P. J. A. , Terhorst, C. & de Vries, J.E. J. Immunol. 135, 1922-1228 (1985).
0 31. Griscel l i, C. , Durandy, A. , Vi rel iz ier, J. L.
et al. 1.1980) In: Seligmann, M. & Hitzig, H.
(eds) Primary immunodeficiencies, Elsevier, North-Holland pp 499-503.
~5 32. Hadman, M. R. , Dopfer, R. Hans-Harmut, P. &
Neithamme~r, D. (1984) In: Griscelli, C., Vossen, J. (eds) Progress in immunodeficiency research and therapy I. Elsevier Science Pub-lishers B.V., Amsterdam pp 43-50.
33. Levin, L. S., Perrin, J. C. S, Ose, L., Dorst J.P., Miller, J.D. & McKusick, V.A. J. Ped.
90, 55-67l (197'7) .
34. Brenner, M. B. , McLean, J. , Yang, S. Y. , van de r Poel, J.J., Pious, D. ~ Strominger, J.L. J.
Immunol. 135, :384-390 (1985) .
35. Yasunobu, Y., Anatoniou, D., Clark, S.P., Yanagi, Y. , S~angster, R. , Van den Elsen, P. , Terhorst" C. ~ Mak, T. Nature 312, 521-524 (1984) .
36. Dialynas,, D. P. , Murre, C. , Quertermous, T. , Boss, J,. M., Leiden, J.M., Seidman, J.G. &

Strominge;r, J.L. Proc. Natl. Acad. Sci. U.S.A.
83, 2619-:?623 (1986) .

37. Raulet, D.H., Garman, R.D., Saito, H. &

Tone~gawa, S. Nature 314, 103-107 (1985).

38. Snodgrass, H.R., Dembic, Z., Steinmetz, M. &

von Boehme~r, H. Nature 315, 232-233 (1985) .

39. De .La Hera, A., Toribio, M.L., Marquez, C. &

Martinez-Pv., C. Proc. Natl. Acad. Sci. U.S.A.

82, 6268-6271 (1985) .

40. Beverley, P.C. & Callard, R.E. Eur. J. Immunol.

11, 329-334 (1981).

41. Krangel, Mf.S. EMBO J. 4, 1205-1210 (1985).

42. Leiden, J.M., Fraser, J.D. & Strominger, J.L.

In press, Immunogenetics (1986).

43. Leiden, J.M. & Strominger, J.L. In press, Proc.

Natl. Acad. Sci. U.S.A. (1986).

44. Erickson, B.W. & Merrifield, R.B. (1976) In:

Neur.ath, H. & Hill, R.L. (eds) The proteins.

Academic Press, N.Y. pp 255-527.

45. Liu, F-T, Zinnecker, M. Hamaoka, T. & Ratz, D.H. Biochem. 18, 690-697 (1979).

46. Brenn er, M.B., et al. J. Immunol. 138, 1502-1509 (1987 ) .

134104 _ 47. Kenneth, R.1Y., McKern, T.J. and Bechtol, K.B.
(eds) Monoclonal Antibodies: a new dimension in biological anlaysis. Plenum Press, N. Y.
( 1980) .
48. Ischimori, Y'. , Kurokawa, T. , Honda, S. , Suzuki, N., Waakimasu, M, and Tsukamoto, R., J. Immun.
Method. 80, 55-66 (1985).
49. Royer, H. D., Bensussan, A., Acuto, 0. and Reinherz, E. L., J. Exp. Med. 160: 947-953 ( 1984) .
50 . Acuto, 0. , :Fabbi, M. , Smart, J. , Poole, C. B. , Protentis, .J., Royer, H. D. , Schlossman, S. F.
and R~°inherz, E. L. , Proc. Natl . Acad. Sci .
U. S.A. , 81, 3851-3855 (1984) .
51. Maniat:is, T.. , Fritsch, E. and Sambrook, F. , Molecular C:lonina. A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, N. Y. 1:L724 ( 1982) .
52. Schrei~er, M. , Rohler, G. , Hengartner, H. , Berek, C., Trucco, M. and Formi, L., Hybridoma Techniques, Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 11724 (1980).
53. Yelton,, D. E. , Desaymard, C. and Scharff, M. D. , Hybridoma 1, 5-11 (1981) .
54. Brenner, M. B. , McLean, J. and Strominger, J. L. , Feb. Proc. 4:5, 1292 (1986) .

55. Hopp, T. P., and Wood, K.R., Proc. Natl. Acad.

Sci . USA 7 8, 3 82 4-3 82 8 ( 1981 ) .

56. Ryte, J. and Doolittle, R. F., J. Mol. Biol.

157, 105-132 (1982) .

57 . Geysen, H. M. , Bartel ing, S. J. and Meloen, R.
H. , Proc. Natl. Acad. Sci. USA 82, 178-182 (1985) .

58. Barnstable, C.J. et al., Cell 14, 9-20 (1978).

59. Bushkin, Y., et al., J. Exp. Med. 164, 458-473 ( 19 86 ) .

~5 60. Yang, S.Y., et al., Immunorenetics 19, 217-231 ( 1984) .

61. Spear, B. T. , et al . , J. Exp. Med. 162, 1802-1810 (:1985) .

62. Gorga, J.C., et al., Meth. Enzym. 106, 607-613 ( 1984) .

63. Chen, Y.X., et al., Hum. Immun. 10, 221-235 25 (1984) .

64. Samelson, L. E., et al., PNAS USA 82, 1969-1973 ( 1985) ,.

30 65. Rrange7., M. S. , et al . , Cel l 18, 97 9-991 ( 197 9) .

66. Bonner, W.J. and Laskey, R.A., Eur. J. Biochem 46, 83--88 (1:974) .

~34~042 67. Southern, E.M., J. Molec. Biol. 98, 503-517 ( 1975) .
6 8 . Duby, A. D. and S ei dman, J. G. PNAS U SA 83 , 4 890-4894 (1986) .
69. Weiss, A., et al., PNAS USA 83, 6998-7002 (1986) .
70. Bank, I. , et al. , Nature 322, 179-181 (1986) .
71. Quertennous, T.., et al., 1987 J. Immunol., 138:2687-2690.
~5 72. Now ill, A. , et al . , J. Exp. Med. 163, 1601-1606 (1986) .
73. Lanier, L.L., et al., J. Exp. Med. 164, 339-344 (1986) .
74. Moingeon, P'. , et al . , Nature 323, 638-640 ( 1986) .
75. Brenner, M., et al., Nature 325, 689-694 (1987) .

SDPPLEMENTARY DIBCLOSUItE
The present invention is further directed to a form of the human y T cell antigen receptor polypeptide termed Form 2bc, which has a molecular weight of about 40,000 daltons, and a constant region which contains a sequence encoded by only two C~2 C:CI exon copies: The invention also relates to T cell antigen receptor heterodimers comprising y Form 2bc, and to nucleic acid sequences encoding ~ Form 2bc and portions thereof. The invention also provides monoclonal antibodies specifically reactive with an epitope of the y or d T cell antigen receptor polypeptides.
2. BACKGROUND OF THE INVENTION
The T cell antigen receptor (TCR) was shown to be a clone specific disulfide-linked heterodimer on T cells, composed of two glycosylated subunits, one of which is designated the a chain and the other of which is designated the ~ chain. 'The a .and ~TCR subunits have a relative molecular mass (Mr) o~ approximately 50,000 and 40,000 _ daltons, respectively (Allison et al., 1982, Immunol.
129:2293-2300; Meuer et al., 1983, J. Exp. Med. 157:705-719;
Haskins et al., 1983,, J. Exp. Med. 157:1149-1169). Genes that rearrange durir.~g T cell ontogeny and encode the pTCR
(Yanagi et al., 1984, Nature 308:145-149: Hedrick et al., 1984, Nature 308:153-158) and aTCR (Chien et al., 1984, Nature 312:31-35; Saito et al., 1984, Nature 312:36-40, Sim et al., 1984, Nature: 312:771-775) subunits were isolated either by subt.ractiwe hybridization or by probing with oligonucleotid,es.
The. alpha and beta chains of the T cell antigen receptor of a T cell. clone are each composed of a unique combination of domains designated variable (V), diversity (D), joining (J), and constant (C) (Siu et al., 1984, Cell 37:393; Yanagi. et a7.., 1985, Proc. Natl. Acad. Sci. USA 82:
3430). Hyperv~ariab7.e regions have been identified (fatten 1341042r et al., 1984, ;Nature 312:40; Becker et al., 1985, Nature 317:430). In .each T cell clone, the combination of V, D and J domains of both the alpha and the beta chains participates in antigen recognition in a manner which is uniquely characteristic of that T cell clone and defines a unique binding site, also known as the idiotype of the T cell clone. In contrast, the C domain does not participate in antigen binding.
A unique feature of the human a,pTCR was the observed comodulatio:n (Meuer et al., 1983, J. Exp. Med.
157:705-719), ~~oimmu:noprecipitation (PCT International Publication No. WO 88/00209, published January 14, 1988;
Oettgen, et al., 1984, J. Biol. Chem. 259:12,039-12,048) and required coexpressio:n (Weiss et al., 1984, J. Exp. Med.
!5 160:1284-1299) of the a,~TCR molecules with a CD3 glycoprotein complex. Subsequently, the direct physical association of the two protein complexes was demonstrated by chemically cro:as-linking the a,~TCR molecules to the T3 qlycoprotein and identifying the components _of the _cross-20 linked complex as the TCR subunit and the T3 glycoprotein (Mr 28,000) subunit (Brenner et al., 1985, Cell 40:183-190).
A T3 counterpart is similarly associated with murine a,~9TCR
(Allison et al", 198'5, Nature 314:107-109; Samelson et al., 1984, Immunol. Rev. 81:131-144).
25 A third gene that rearranges in T cells, designated yTCFt, was identified, first in mice (Saito et al, 1984, Nature 309:757-762: Kranz et al., 1985, Nature 313:762-755: Hayday yet al., 1985, Cell 40:259-269) and then in humans (Lefranc et al., 1985, Nature 316:464-466; Murre 30 et al., 1985, nature 316:549-552). The human ~TCR locus appears to con:~ist o:E between five and ten variable, five joining, and two con:~tant region genes (Dialynas et al., 1986, Proc. Nat:l. Acad. Sci. U.S.A. 83: 2619). Although the total number o!: funcitional variable and joining regions is 35 limited, signil:icant diversity is introduced during the process of V-J joining (Kranz et al., 1985, Nature 313:752-755: Lefranc et al., 1986, Cell 45:237-246; Quertermaus et al., 1986, Nature 32.2:184). The ~TCR gene rearrangements occur in lymphocytes; with suppressor-cytotoxic as well as helper phenotypes (Lefranc et al., 1985, Nature 316:464-466;
Murre et al., 1985, Nature 316:549-552, Quertermaus et al., 1986, Science 231:2°~2-255; Lefranc et al., 1986, Cell 45:237-246, Iwamoto et al., 1986, J. Exp. Med. 163:1203-1212: Zauderer et al.., 1986, J. Exp. Med. 163:1314-1318).
0 The products of the 7TCR gene have been identified in T3 coi.mmunoprecipitates from a~TCR CD3+ T
(Brenner et al., 198.6, Nature 322:145-149: Bank et al., 1986, Nature 322:179-181: Borst et al., 1987, Nature 325, 683-688: Moingeon et: al., 1987, Nature 325, 723-726, PCT
~5 International Publication No. WO 88/00209, published January 14, 1987). The yTCR. polypeptides were identified by use of monoclonal antibodies directed against ~TCR peptide sequences; these pol.ypeptides were found to be incorporated into heterodim.ers with another polypeptide called dTCR
20 (Brenner et al., 1986, Nature 322:145-149). The y,d heterodimer was reported to be associated noncovalently with CD3.
Use of antisera directed against yTCR-specific peptides has led to the identification of CD3-associated 25 '~TCR polypeptides on cells originating in peripheral blood, thymus, and a leukemic cell line (Hrenner et al., 1986, Nature 322:145-149: Bank et al., 1986, Nature 322:179-181, Weiss et al., 1986, Proc. Natl. Acad. Sci. U.S.A. 83:6998-7002; Brenner et al., 1987, Nature 325:689-694 Lew et al., 30 1986, Science 234:1401-1405). Bank et al. (supra) disclosed a 44 kd ~ form which was associated with a 62,000 kD peptide and T3 on the surface of a human thymocyte clone. A similar yTCR polypeptide was. also identified on murine T
lymphocytes, a.nd thE: expression of this peptide during 35 thymocyte diff'erent~.ation is the subject of much current ~ 341 042 study (Rbulet et al., 1985, Nature 314:103-107; Snodgrass et al., 1985, Nature 315:232-233; Lew et al., 1986, Science 234:1401-1408; Parda.u et al., 1987, Nature 326:79-81;
Bluestone et al., 1987, Nature 326:82-84).
With the study of ~TCR+ human cell lines, two different ~TCR polypeptides have been identified that differ in their molecular weight and in their ability to form disulfide linkages (Borst et al., 1987, Nature 325:683-688;
Brenner et al., 1987, Nature 325:689-694; Moingeon et al., 1987, Nature 325:723-726; Lanier et al., 1987, J. Exp. Med.
165:1076). Two different ~TCR constant region gene segments, called C~1 and Cy2, respectively, have been compared; a cysteine residue encoded by the second exon of C~1 appears to be absent in C~2 exon segments, and its absence has been suggested to explain the inability of some yTCR peptides 'to form disulfide bonds (Krangel, et al., 1987, Science, 237:1051-1055; Littman et al., Nature 326:85088).
In contrast to the multiple forms of ~TCR, the 20 sTCR molecule :is relatively invariant and it appears that there is only one aTCR constant region (Hata et al., 1987, Science 238:678-682).
During T cell ontogeny, it has been shown that 7TCR gene rearrangement precedes ~ and aTCR gene 25 rearrangement (Roulet et al., 1985, Nature 314:103-107;
Snodgrass et al., 1985, Nature 315:232-233: Sangoter et al., 1986, J. Exp. :Med. 163:1491-1508).
Of :mature, circulating T lymphocytes, a relatively small proportion are y6TCR+, and exhibit either 30 CD3+4 8 (double negative) or CD3+4 8+ surface antigens.
CD3+4 8 T cells constitute approximately two percent of mature CD3+ T cells. Unlike most mature CD3+4+ or CD3+g+
major histocom;patibility locus (MHC) restricted cytotoxic T
cells, but similar to CD3 natural killer cells, ydTCR+
35 CD3+4 8 cloned lymphocytes have been shown to exhibit MHC-G

sD77 ~ ~ 4 ~ 0 4 2 nonrestricted cytolytic activity: however, unlike natural killer cells, these yd+ CD3+4 8 T cells did not consistently kill natural killer cell targets, such as K-562 (Borst et al., 1987, 325:683-688: Brenner et al., 1987, Nature 325:689-694: Moingeon et al., 1987, Nature 325:723-726: Bluestone; et al.., 1987, Nature 326:82-84).
3. SUMMARY OF THE INVENTION
The: present invention is directed to a form of 0 the human 7 T cell antigen receptor (TCR) polypeptide termed Form 2bc. The: Form 2bc yTCR chain has a primary amino acid sequence subst:antial.ly as depicted in Figure 25. The Form 2bc 7TCR chain has a: molecular weight of about 40,000 daltons, and compri:~es a constant region containing a 5 sequence encodled by only two Cy2 CII exon copies. The invention also relates to TCR heterodimers comprising the ~TCR polypeptide Fozzn 2bc.
They invention is also directed to nucleic acid sequences encoding_~~TCR Form_2bc, and to nucleic acid-20 'sequences comF~risinc~ a Cy2 constant region having only two CII exons. Ire a spE:cific embodiment, the nucleic acids of the invention comprise at least a portion of the nucleic acid sequences. shown in Figure 25.
The: invention also provides monoclonal antibodies 25 specifically reacti~re with an epitope of the ~ or dTCR
polypeptides. Such antibodies can be identified by detecting the3.r abi7Lity to co-modulate the CD3 antigen on a cell which expresses both the ~dTCR and a CD3 antigen. In a specific embodiment,, the invention relates to antibodies 30 reactive with the variable region of the bTCR chain. In a particular embodiment, such an antibody can be used to detect functional 6TCR variable gene rearrangements in a cell. In another embodiment, the invention relates to antibodies reactive with the constant region of the STCR

1~41p42 polypeptide. In yet. another embodiment, the invention relates to ant:ibodiea reactive with the constant region of the yTCR polyp~eptidE..
In anothE:r aspect of the invention, a method is provided for producing expression of a ~dTCR in a cell.
3.1. DEFINITIONS
As used herein, the following terms will have the meanings indicated:
0 TCR = T cell antigen receptor V = variable D = diversity J - joining C = constant ~5 mAb = monoclonal antibody 4. DESCRIPTION OF THE FIGURES
Figuire 12.. Cytofluorographic analysis of T cell l fines with ant: i-..TCRd~ 1.. --_ 20 Figure 13. Immunochemical analysis of the specificity of' mAb anti-TCRdl. Surface 125I-labeled IDP2 cells were ima~unopre:cipitated using control mAb P3 (lanes 1 and 2), anti-l.eu 4 (lanes 3-5), anti-TCR61 (lanes 6-8), or anti-Cy serum (lane 9) and were then resolved by SDS-PAGE
25 (polyacrylamicte gel electrophoresis) and visualized by autoradiography. N = nonreducing conditions: R = reducing conditions.
Figure 14.. N-glycanase digestion of dTCR.
Figure 15.. Map of pGEM3-0-240/38.
30 Figmre 16.. Immunoprecipitation of in vitro translation products of cDNA clone IDP2 0-240/38 by mAb anti-TCRbl.
Figure 17.. Immunoprecipitation and SDS-PAGE
analysis of T cell antigen receptor. Open arrowheads 35 indicate the position of the b chains. The solid arrowheads indicate the position of the y chains. Lysates were immunoprecipit:ated using bTCAR-3 antibody (odd numbered lanes) or ~9F1 antibody (even numbered lanes).
Figure 18. Immunoprecipitation of b chain by bTCAR-3 antibody. Molt-13 cells solubilized in Tris-buffered saline (pH 8) containing 0.3% CHAPS (lane 1) or in 1% Triton X-100 (lanes 2-7). Lane 1, bTCAR-3 immunoprecipit:ates -r,bTCR heterodimer with the CD3 proteins.
Lane 2, bTCAR-3 immunoprecipitates y,bTCR heterodimer 0 without the CI)3 proiteins. Lanes 3 and 4, STCAR-3 immunoprecipit:ates :single b chain from denatured lysates (N) and reducing I;R) conditions, respectively. Lane 5, UCHT-1 immunoprecipit:ates ithe CD3 proteins. Lane 6, ~Fl antibody does not immunoprec:ipitate a heterodimer from MOLT-13 cells.
one ~~ anti-c:7 antiserum immunoprecipitates a single y chain.
Figure 19.. Analysis of cell surface staining by flow cytometry. y,d~TCR-positive cells (MOLT-13) PEER, IDP2) and a.BTCR-posit.ive_cells (HPH-ALL, J_urkat)_were incubated with bTCAR-3, OKT3, WT31 and normal mouse serum (NMS) antibodies and anal.~zed by flow cytometry. The B cell line, Daudi, was the negaitive control.
Figure 20. Two color cytofluorographic analysis of bTCAR-3+ arid OKT'.3+ peripheral blood lymphocytes. The fluorescein i:aothiocyanate (FITC) fluorescence is depicted on the Y axis and pllycoerythrin (PE) fluorescence on the X
axis. The CD:3+ y,bTCR+ cells in this sample represent 2.4%
of CD3+ lymphocytes.
Figure 21. Measurement of intracytoplasmic Ca2+
concentration ([Ca2~Ji) versus time. Top panel: bTCAR-3.
Bottom panel: Anti-'.Leu antibody. Arrows indicate the time of addition oi: antibody.
Figure 22. Immunoprecipitation of the three forms of y,bTCR. Fair parts A-E, the antibodies used for l~unoprecipit:ation are anti-Leu4 (anti-CD3), ~F1 (anti *Trade-mark r TCR~); anti-bl'.rCR (anti-bTCR), anti-Cyb serum (anti-~TCR) and P3 (unlabelled lanes, control). Immunoprecipitations from 1251-labelled cell lysates were analyzed by SDS-PAGE
(10% polyacrylamide) under reducing (R) or nonreducing (N) conditions. An open arrow (D) indicates the position of TCR b under reducing conditions, whereas the solid arrow (,) denotes the position of bTCR under nonreducing conditions. Size markers, Mr in thousands, are shown on the left.
A) Nond.isulfide-linked yTCR (40kD) on PBL-L2.

In lanes 1-6 the radiolabelled cells were solu,bilized in 0.3% CHAPS detergent which preserves the TCR-CD3 association, whereas in lanes 7 and 8, immunoprecipitations were performed after chain separation (see methods).

B) Nond.isulfide-linked yTCR (55kD) on IDP2 cells. In lanes 1-4 radiolabelled cells were.solubilized in 0.3% CHAPS detergent, whereas in lanes 5 and 6 imunoprecipitations were: carried out after chain separation.

C) Disulfide-linked ~TCR (40kD) on WM-14 cells.

All lanes correspond to immunoprecipitations from 1% digitonin solubilized radiolabelled cel l s .

D) Nond.isulfide-linked yTCR (40kD) on thymic Clone II cells. Radiolabelled cells were solu,bilized in 1% digitonin (lanes 1-4) or in 0.1% Triton X-100 (lanes 5 and 6), whereas in lanes 7 and 8 immu.noprecipitations were carried out after chain separation.

E) Nond,isulfide-linked yTCR (40kD) on MOLT-13 leukemia T cell:. In lanes 1-4 immu.noprecipita.ions were carried out after SD81 1 3 41 p 4 2 solubilization of cells in 0.3% CHAPS
detergent, whereas in lanes 5 and 6 immunoprecipitations were carried out after chain separation.
Figure 23'.. Zmmunoprecipitation of ~TCR and bTCR
chain by anti-~C~rmi antibody and anti-TCRbl antibody, respectively. Cell surface radiolabelled MOLT-13 cells were solubilized in 0.3% CHAPS detergent and the y,bTCR-CD3 complex was isolated with anti-CD3 monoclonal antibody.
Immunoprecipit:ates were analyzed by 10% SDS-PAGE under reducing conditions"
Lame 1: Immunoprecipitation with anti-Leu4 (anti-CD3) mAb Lame 3: Immunoprecipitation with anti-Cym1 (anti-TCRy) mAb after separating chains of isolated y,bTCR-CD3 complexes.
Lame 4: Immunoprecipitation with anti-TCR bl (anti-TCRb) mAb after separating chains of isolated ~,bTCR-CD3 complexes.
Figure 29~. Determination of peptide backbone sizes and glyc:osylai:ion of y and bTCRs from PEER and MOLT-13 cells. Monoc7.ona1 antibodies used for immunoprecipitation are anti-Cyml (anti-~TCR~r) , anti-TCRbl (anti-TCRb) and P3 (labelled control) as shown at the tap of each lane. The labelled cell lines used are shown at the bottom of each 10%
SDS-PAGE autoradiograph or fluorograph. All samples were resolved under- reducing conditions. Size markers, Mr in the thousands.
A) Pepi;.ide backbone sizes of ~TCR from PEER and MOL'.~-13 cells. Cells were biosynthetically labcalled with 35S-cysteine and 35S-methionine for 15 minutes. Samples were either treated with Endo H (+) or mock treated (-). Immunoprecipitation with anti-C7m1 shows the positions of immature *Trade-mark SD82 1 3 4 1 a 4 Z -yTCR of PEER cells (lane 3) and of MOLT-13 cells (lane 7), while the corresponding pol~~peptide backbone sizes are visualized after treatment with endo H (lanes 4 and 8).
B) Glyc:osylation of TCR b from MOLT-13 cells.
125.x-labelled cells were immunoprecipitated with anti-CD3 mAb and the dTCR polypeptides were. gel purified (see methods) before incubation with N-glycanase (lane 4), endo H
0 _ (lane 2), or mock treated (lanes 1, and 3).
Figure 2°_i. Nucleotide sequence of MOLT-13 yTCR
(Form 2bc). F>art A;; Sequencing strategy of clone Ml3k. A
partial restriction map of the 1.1 kb cDNA clone Ml3k is shown. Part Et: Nucleotide and deduced amino acid sequence ~5 of clone Ml3k. Signal sequence (S), variable (V), N-region (N), joining (J) and constant (CI, CIIb, CIIc and CIII) region gene se::gment:a are indicated by arrows and were identified by comparison to genomic sequences, described by Lefranc et .al. , (19F36., Cell ~L5:237 246) (for S and V) , 20 Lefranc et al., (19F36, Nature, 319:420-422) and Quertermous et al., (1987, Immunol: 138:2687-2690) (for J) and Lefranc et al., (1986, Proc. Natl. Acad. Sci. U.S.A. 83:9596-9600) and Pellicci eat al.,, (1987, Science 237:1051-1055) (for C).
The deduced amino acid sequence beginning at the initiator 25 methionine is presented below the nucleotide sequence.
Extracellular cysteines are highlighted by boxes, and potential N-linked carbohydrate attachment sites (N-X-S or N-X-T; Marshall, 1977, Ann. Rev. Hiochem. 41:673-702) are indicated by ~sracketa.
30 Ficrure 26. Preferential use of y, bTCR Form 1.
Freshly isolated peripheral blood mononuclear cells from three healthy donor:: were 1251-labelled and solubilized in '1% Triton X-100. h~nunoprecipitates with P3 (control, lanes 1 341 p42 1 and 3)~, and anti-'.rCRbl (anti-TCRb, lanes 2 and 4), were analyzed unde:- nonreducing (N) and reducing (R) conditions.
Mr markers in the thousands are shown on the left.
Figure 2'7. Schematic representation of the three ~,bTCR forms in man. The CII exon encoded connector peptides are Highlighted by filled areas (~ ) as Cyl CII
exon encoded peptide; ~ , ~, m , as C~2 CII exon copy a, copy b, an~i copy c encoded peptides, respectively).
Potential N-linked glycan attachment sites (o), and 0 sulfhydryl groups (~-SH) and putative disulfide bridges (-S-S-) are indicated.
Figure 213. Map of the rearranged bTCR gene. A
map of r119b1 including EcoRI (RI), Hinc II (Hc), ScaI (S) and PvuII (P) sites and probes used in Southern blot ~5 analysis is shown.
Figure 2!9A. Schematic representation of the 7TCR
chains used for transfection into MOLT-13 cell line. The schematic is based on reported analyses (Brenner, M.B., et al. , 1986., Nature 322:145-_149.:_ Brenner, 1~I. B. , et al . , 1987, -20 Nature 325:689-694; Krangel, M.S., et al., 1987, Science 237:64-67). ~?red., predicted; Obs., observed. The predicted glyc:osylaited polypeptide size assumes that all available N-linked glycosylation sites (shown as lollipops), each containing 3 klDa of attached carbohydrate, are used, 25 and that no significant size differences are introduced by other post-translat:ional modifications. The intra-chain disulfide linkages typical of Ig-like molecules are shown.
Note that a cystein~e residue (cys) is encoded by the CII
exon in P8L C7l ~TCR" but such a cysteine is absent from all 30 the copies of CII e:KOn used in the two other yTCR chains.
~TCR constant region in the ~,b T leukemia cell line PEER
(Littman, D.R., et al., 1987, Nature 326:85-88) that also expresses a 5..°i kD y~.~CR protein (Hrenner, M.B., et al., 1987, Nature 325:689-694; Weiss, A, et al., 1986, Proc. Natl.
35 Acad. Sci. USA 83:6!398-7002) is identical to that of IDP2.

Figure 25iB. The expression plasmid constructs pFneo.PBL Cly and pF'neo,IDP2~ were used to introduce 7TCR
clones into the MOL7"-13 cell line. PBL Cl yTCR cDNA clone (PBL C1.15) and repaired IDP2 yTCR cDNA clone (IDP2.llr) (Krangel, M.S., et al., 1987, Science 237:64-67) were cleaved from their parent plasmid vector (pUC 18) by EcoRI
digestion, the: ends were made blunt with Klenow fragment of DNA polymerise: I, and the cDNAs were then ligated into a SalI-cut, and Klenow-treated pFneo mammalian expression 0 vector. Clonea containing the cDNA inserts in appropriate orientation with respect to the spleen focus forming virus (SFFV) LTR were selE:cted based on restriction mapping.
pFneo (Saito, T., ei~ al., 1987, Nature 325:125-130) is a derivative of pT~Fne:o (Ohashi, P., et al., 1985, Nature ~5 316:606-609) obtainead by BamHI digestion, to delete the murine ~9TCR cDNA inscert, followed by ligation with T4 DNA
ligase. As shown, i~his vector contains a bacterial neomycin resistance gene (near) under the control of SV40 promoter, thus conferring resistance to the antibiotic 6418 on.the 20 mammalian recipient cells. The restriction sites within parentheses were destroyed during construction.
Figure 317 A-E. Immunoprecipitation analysis of y,bTCR on Molt-13 yT'CR transfectants. Surface 125I-labeled cells were sol.ubili;sed in 0.3% CHAPS detergent to preserve 25 the chain association, immunoprecipitated with mAb P3 (control), anti-leu~-4 (anti-CD3), anti-TCRbl (anti-6TCR), or anti-Ti-7A (anti-V72), and were then resolved by SDS-PAGE
under nonreduc:ing (1!1) or reducing (R) conditions and visualized by autoradiography as described earlier (Brenner, 30 M~B., et al., 1986, Nature 322:145-149: Brenner, M.B., et al., 1987, Nature 325:689-694). Anti-Ti-yA mAb shows a pattern of reactivity on different T cell clones consistent with its recognition of Vy2 segment. M13.PBL Cly: MOLT-13 cells transfecaed with the PBL C1-derived yTCR cDNA: Clone 35 #7 was used for this analysis. M13.IDP2y: MOLT-13 cells transfec~ed with the IDP2-derived ~TCR cDNA: Clone #10 was used for this analysis. Size markers, Mr (molecular weight) in thousands of dalitons. Open arrow, resident MOLT-13 yTCR
chain: solid arrow, transfected (PBL C1- or IDP2-derived) yTCR cDNA: ast:erisk,, MOLT-13 bTCR chain under nonreducing conditions. tJpon reduction, the dTCR chain undergoes a mobility shift: and comigrates with the 40 kD yTCR chain.
However, bTCR chain is distinctly visualized as a 40 kD band under reducing conditions when the 40 kD ~TCR protein is not 0 coimmunoprecipitate<i, as is seen in anti-V-~2 immunoprecipit:ates of M13.IDP2y (Fig. 30C, lane 8).
Figure 3:1. Two-dimensional gel analysis of 7TCR
polypeptides of transfectants. 2x107 cells were surface 1251 labeled, treatead with neuraminidase (150 units in 1.5 ml PBS with 1 mg/ml each of glucose and bovine serum albumin for 1.5 hours at 23"C), solubilized in 0.3% CHAPS detergent, immunoprecipit:ated with 1 ~g anti-leu-4 mAb, and subjected to 2D gel ana7.ysis under reducing conditions. Non-equilibrium pH gradient-gel eletrophoresis (NEPHGE)-was -carried out using pli 3.5 to 10 Ampholines (LKB, Sweden) followed by SI)S-poll~acrylamide gel electrophoresis on 10.5% acrylamide ge:L (Brenner, M.B., et al., 1987, Nature 325:689-694). Positions of the CD3 components (not shown) were used to ~:denti:Ey and compare the yTCR species expressed in different cell lanes. M13.PBL Ch transfectant clone #10 and M13.IDP2y transi:ectant clone #10 were used. Open arrows, MOLT-J.3 yTCR species: solid arrows, IDP2 yTCR
species; asterisk, 1PBL Cl yTCR species.
Figure 3a. Analysis of backbone polypeptide sizes of 7TCR chains of transfectants. Cells were pulse labeled with '~5S-me~thionine and 35S-cysteine for 15 minutes and immunoprec:ipitalted with P3 (control) , anti-C~rml (anti-yTCR), or anti.-Ti-7A (anti-Vy2) mAb, as indicated.
Immunoprecipit:ates were treated with endoglycosidase-H
(Endo-H, +), or were mock-incubated (-), resolved by SDS-~,._.~.

PAGE,- and visualized by fluorography. All samples were run reduced. Mr, molecular weight markers in thousands of daltons. The 43 kD contaminating actin band serves as an additional internal marker. M13.IDP2~: MOLT-13 cells transfected with then IDP2 yTCR cDNA, clone #10: M13.PBL C17:
MOLT-13 cells transfected with the PBL C1 yTCR cDNA, clone #10.
Figrure 3::. Southern blot analysis of yd T cell clones and po7.yclonal human T cell populations. Genomic DNA
was digested with Ec:oRI and probed with the V-J probe. DNA
sources are: PBL T~-cell clones, (lanes 1, 3-9), PBL (lane 10), newborn t:hymocytes (NBT-lane 11), fetal thymocytes (FT-lane 12), and B cells (germline-lane 2). The germline 3 kb Vb and 6.7 kb Jd fragment are indicated on the left of the blot, while the 5 common rearrangements, numbered I-V
are indicated on the. right. The sizes of the rearrangements from I-V are 2.9 kb,, 3.5 kb, 4.2 kb, 6.2 kb and 7.1 kb respectively.
Figure 3~4A. Northern -blot analysis- of - group O -hybridizing transcripts. RNA sources were: JY, B cell line; HL60, myeloid cell line: HPB-ALL and SRW3, a,pTCR and surface TCR T cell :Lines, respectively: fresh and PHA
peripheral blood mononuclear cells (PBMC), fresh and 2 days PHA activated PBMC: IDP2, PEER, Molt-13 and PBL-L1 (identical to WT31 ~- PBL Line, (Brenner et al., 1987, Nature 325:689-694), ybTCR T cell lines. Arrowheads mark the positions of t:he four major transcripts detected: 18S and 28S rRNA servE:d as ~aarkers.
Figure 34B. Northern blot analysis using IDP2 ~A~ IDP2 RNA treaited as described in materials and methods was probed with niclk-translated 0-240, a 330 by EcoRI-ScaI
fragment of 0~~240/38 (V probe: see Figure 36) labelled by hexanucleotidea priming, or a 550 by HaeIII fragment of .0-240 (3'UT: see Figure 36) labeled by nick-translation.

~4~ 042 Figure 35A. Southern blot analysis of genomic DNA (XbaI digest). 5 ~g high molecular weight genomic DNA
samples were .digested with XbaI, electrophoresed through 0.7% agarose, transferred to nitrocellulose, and probed with nick-translated 0-240. DNA sources were: SB and 8392, B
cell lines; H:L60, myeloid cell line; 2B5, 2D6, Anita, Jurkat, and H:BP-MLT, a,~TCR T cell lines; Molt-4, CEM, 8402 and HSB, surface TCR T cell lines; PBMC, fresh peripheral blood mononuclear cells: Molt-13, IDP2, PEER and PHL L1, 1~ y6TCR T cell 7Lines. Bacteriophage lambda DNA digested with HindII was used as a molecular weight standard. Germline bands are marked by arrowheads.
Figure 35B. Southern blot analysis of genomic DNA (EcoRI and PwII digests). Genomic DNA samples digested with EcoRI or PvuII and probed with the 425 by 5' EcoRI
fragment of clone 0-240/38 (VJC probe: see Figure 36) labelled by nick-translation. PBL Cl is a ~dTCR T cell clone (Brenner et al., 1987, Nature 325:689-694): FET LIV 2 and FET THY 2-are~fetal liver and-thymus-samples-from the -2~ same fetus: F:E THY 4 is from a distinct fetus: other DNA
samples are those used in Figure 35A. Germline bands in each digest a:re marked by arrowheads.
Figure 36. Organization and sequencing strategy of group 0 cD:NA clones.
-Ficrure 3'~A and B. Composite nucleotide sequence of group O cDNA clones encoding dTCR protein. Amino acid residues are numbered from the presumed amino terminal processing point. Cysteine residues are boxed, potential N-linked glycosylat.ion sites are bracketed, and Polyadenylation signals used in the clones are underlined.
The composite nucleotide sequence is compared with that of the coding region a~f murine cDNA clone DN-4 (Chien, et al., 1987, Nature 327:677). (-) denotes identity and (*) denotes a gap.

1341042 ;

Figure 3.B. Amino acid sequence comparisons to consensus hum<in TCR V region sequences. Blanks indicate consensus assignment at that position. (-) indicates a gap.
Identities bei:ween 'the 0-composite sequence and consensus residues are boxed.
Figure 39. Amino acid sequence comparisons to consensus human TCR J region sequences.
Figure 40. Amino acid sequence comparisons to human TCR and immunoglobulin C region sequences.
_ Figure 4:1. Distribution of charged and uncharged amino acids in the region flanking and including the presumed tran:~membr<ine region of the 0-composite sequence compared with those of Ca, Cp, and Cy.
~5 5. DETAINED DESCRIPTION OF THE INVENTION
5.1. THE ~TCR FORM 2bc POLYPEPTIDE AND NUCLEIC ACIDS
The: invention is directed to a form of the human ~TCR polypepti.de teamed Form 2bc -(detailed -infra in Sections 8 and 9). The: invention is also directed to nucleic acids encoding y Form 2bc, such as DNA and RNA, and their complementary nucleic acids.
Foam 1 and Form 2abc ~TCR polypeptides are previously reported forms of the human yTCR (see PCT
International Publication No. WO 88/00209, Published January 4, 1988). The: Form 1 yTCR polypeptide has a molecular weight of about 40,000 daltons. The Form 2abc yTCR
polypeptide has a molecular weight of about 55,000 daltons.
Form 2abc yTCP: chain has a slightly larger peptide backbone and contains one exi~ra potential N-linked glycan than Form 1. In contrast, then ~TCR chain of the invention, Form 2bc, has a molecular weight of about 40,000 daltons.
Furthermore, t:he Fo~:~m 2bc yTCR polypeptide possesses a slightly smaller peptide backbone and 2-3 less potential N-linked glycansc.

yTCR chain Form 2bc differs in size by more than 15 kD (40 kD versus 55 kD) compared to the previously described Form 2abc. This difference is accounted for by a kD smaller polypeptide backbone size (35 kD versus 40 kD) 5 and by a reducaion in the amount of carbohydrates (5 kD
versus 15 kD). The approximately 35 kD polypeptide backbone size of Form 2;bc also serves to distinguisr it from Form 1;
Form 1 has a 9~0 kD backbone size.
yTf.R polypeptide Form 2bc also differs from Form 0 1 and Form 2abc in constant region (Cy) gene segment usage.
Form 1 yTCR chains have a constant region encoded by the Cyl gene segment (Krangeal et al., 1987, Science 237:64-67) containing a aingle CII exon. The Form 2abc y polypeptide is encoded by C~2 dene segments containing three CII exon ~5 copies, namely copy a, copy b and copy c (Krangel et al., 1987, Science 237:64-67; Littman et al., 1987, Nature 326:85-88). 7:n coni~rast, Form 2bc lacks one copy of the sequence encoded by the Cy2 second exon that is present in the cDNA of Form 2aba. This, Form -2bc- contaims -two -Cy2 ° CII
20 exon copies, namely copy b and copy c. Copy a of CII, which is missing in Form abc, encodes a part of a connector region between the membrane spanning region and the extracellular constant domain.
Six potential N-linked carbohydrate attachment 25 sites exist on the :Form 2bc polypeptide. Since the biochemical daita suggest that only 2-3 N-linked glycans are attached to the pol;ypeptide chain, it indicates that not all potential sites are used.
In specific embodiments, ~TCR polypeptide Form 30 2bc can be obi~ained from cells of the MOLT-13 (Loh et al., 1987, Nature 330:56'9-572) T cell line or thymus-derived Clone II (Banlc et al., 1986, Nature 322:179-181). ~TCR
chain Form 2bc can .also be obtained from T lymphocytes of a human subject which express that -~TCR form.

1 341 p42 Foz~ 2bc yTCR polypeptide comprises the primary amino acid sequence of the ~TCR polypeptide shown in Figure 25, or any portion i:hereof comprising a constant region consisting of copy b and copy c of Cy2 CII.
The: present invention also provides a nucleic acid molecule encoding a ~TCR Form 2bc polypeptide having a molecular weight of about 40,000 daltons. The constant region of yTCP: Form 2bc polypeptide results from translation of a nucleic acid sequence which has only two of the three C'r2 cII exons. The invention is also directed to nucleic acid sequences comprising a C~2 constant region having only two cII exons. The. nucleic acid can be a DNA, cDNA, RNA, and complementary nucleic acids and derivatives thereof. In a specific embodiment of the invention, the DNA molecule comprises at least a portion of the nucleic acid sequence shown in Figure 25.
In an ex~unple to be discussed in Section 8, the 2bc yTCR polyp~eptide: and its encoding nucleic acid sequence are described. _In an.example.to be discussed in Section 9, it is shown that then ability of the ~TCR polypeptide to form disulfide bonds or be glycosylated is determined by its constant region primary sequence.
5.2. POLS!PEPTIDE COMPLEXES CONTAINING yTCR FORM 2bc The: presE:nt invention also relates to polypeptide complexes which comprise the yTCR chain Form 2bc. In a specific embodiment" the polypeptide complex consists of a T
cell antigen receptor dimer. In particular, such a dimer can be a heterodimer (including but not limited to a y,b heterodimer, a ~r,~ heterodimer, and a a,7 heterodimer, or a y,~' heterodim~er in which y' can be yTCR polypeptide Form 1, 2abc, or 2bc), or a homodimer.
In a pari:icular embodiment of the invention, the polypeptide complex comprising ~TCR Form 2bc is a ~6TCR
heterodimer. Thus, a purified complex which comprises at ? 34? p42 _ least a portion of a dTCR polypeptide and yTCR Form 2bc polypeptide i~: provided by the present invention. The a polypeptide many havE: at least one intrachain, covalent, disulphide bridge. Additionally, the polypeptide may comprise a dTCR polypeptide having a molecular weight of about 40,000 dlaltona.
As detailed in the examples infra, the yTCR Form 2bc chain is r,~oncovalently associated in a complex with the sTCR chain. Thus, 7 Form 2bc forms a nondisulfide-linked 0 TCR complex. yTCR chain Form 2abc also forms a nondisulfide-linked complex with a dTCR chain (e. g., on IDP2 cells), while yTCR chain Form 1 forms a disulfide-linked complex with a~ 6TCR polypeptide.
As shown in the example of Section 9, infra, yTCR
~5 constant region CII exon usage (and thus the primary sequence of the yTCFt chain) determines not only the presence or absence of disuli°ide linkage between TCR ~ and d, but also the amount of carbohydrate attached to ~TCR, which is largely-responsible for the differences in sine of the cell-20 surface yTCR p~roteir~s. Thus, the present invention also provides a method for producing expression of ~aTCR
heterodimers crf defined intermolecular linkage (disulfide or nondisulfide-linked) and extent of ~TCR glycosylation, which comprises introducing a ~TCR gene encoding a particular y 25 polypeptide form inl:o a cell capable of expressing the gene, which cell expresses the dTCR chain.
The: preseant invention further provides a purified complex which comprises a ~TCR Form 2bc polypeptide of the present invention associated with another yTCR polypeptide 30 (e~gw Form 1, 2abc" or 2bc). In one embodiment of the invention, the: two -~TCR polypeptides are associated with each other through at least one interchain, covalent, disulfide linl~:age. In another embodiment of the invention, the two ~TCR p~olypeptides are noncovalently associated with 35 each other. I:n sti:Ll another embodiment of the invention, ' 34' 042 _ the two yTCR polypeptides have the same constant domain. In yet a further embodiment of the invention, the two yTCR
polypeptides have different constant domains.
5.a. MO1~10CLONAL ANTIBODIES REACTIVE
WI'.rH THE yaTCR POLYPEPTIDES
A monoclonal antibody (mAb) to an epitope of the y or b T cell antigen receptor can be prepared by using any technique which pro~~ides for the production of antibody molecules by continuous cell lines in culture. These 0 include but are nat limited to the hybridoma technique originally de:~cribe~i by Kohler and Milstein (1975, Nature 256:495-497), and the more recent human B cell hybridoma technique (Ko::bor et al., 1983, Immunology Today 4:72) and EBV-hybridoma technique (Cole et al., 1985, Monoclonal Antibodies and Cancer Therapy, Alan R. Liss, Inc., pp. 77-96).
In one embodiment, the monoclonal antibodies may be human monoclonal antibodes or chimeric human-mouse (or other species) monoclonal antibodies. Human monoclonal 20 antibodies may be made by any of numerous techniques known in the art (e,~, Teng et al., 1983, Proc. Natl. Acad. Sci.
U.S.A. 80:7308-7312: Kozbor et al., 1983, Immunology Today 4:72-79: Olsson et al., 1982, Meth. Enzymol. 92:3-16).
Chimeric antibody molecules may be prepared containing a 25 mouse (or rat,, or other species) antigen-binding domain with human constani: regions (Morrison et al., 1984, Proc. Natl.
Acad. Sci. U.S.A. 8:1:6851; Takeda et al., 1985, Nature 314:452).
The invention is also directed to a method of 30 identifying a monoclonal antibody reactive with a T cell antigen recepi:or. ;such a mAb can be identified by detecting its ability to comodulate the CD3 antigen upon binding of the mAb to a cell wlhich expresses both a T cell antigen receptor and t:D3 complex. The CD3 comodulation can be 1341 p42 detected, for example, by measuring the amount of labeled anti-CD3 antibody which is bound by the cell. This method is illustrated by way of example in Section 7.1.1, infra, in which it is a:red to identify hybridomas secreting anti-Va mAb dTCAR-3.
A molecular clone of an antibody to an epitope of a ~ or 3TCR polypept:ide can be prepared by known techniques.
Recombinant D1JA methodology (see e-g., Maniatis et al., 1982, Molecular Cloning, A Laboratory Manual, Cold Spring Harbor Laboraitory, Cold Spring Harbor, New York) may be used to construct nucleic acid sequences which encode a monoclonal anitibody molecule, or antigen binding region thereof.
Anitibody molecules may be purified by known techniques, e.c~., i:mmunoabsorption or immunoaffinity chromatography, chromatographic methods such as HPLC (high performance liquid chromatography), or a combination thereof, etc.
P.ntibody fragments which-contain the idiotype of-the molecule can be generated by known techniques. For example, such fragments include but are not limited to: the F(ab')2 fragment which can be produced by pepsin digestion of the antibody molecule; the Fab' fragments which can be generated by :reducing the disulfide bridges of the F(ab')2 fragment, and the Fab fragments which can be generated by treating the .antibody molecule with papain and a reducing agent.
One embodiment of the invention is directed to monoclonal antibodies reactive with the variable region of the 6TCR chain. Such an antibody is dTCAR-3 (aka TCSbl)(see Section 7, infra), which recognizes an epitope expressed from a specific 6 gene rearrangement. As described in Section 7.2, infra, mAb dTCAR-3 is capable of stimulating the proliferation o~f a ~,6+ T lymphocyte. Monoclonal _.

134 X42 _ antibody dTCAR-3 is also able to stimulate a rise in cytoplasmic free calcium ion concentration of y,6+ T
lymphocytes.
In another embodiment, the invention relates to antibodies reactive with the constant region of the ~ or sTCR polypeptide. In a specific embodiment, the invention is directed to mAb TCRbl which is reactive with the constant region of the dTCR chain (see Section 6, infra). In another specific embodiment, the invention relates to nAb anti-Cymi, 0 which is-reactive with the constant region of the yTCR chain (see Section 8.1.7, infra).
6. GENERATION OF MONOCLONAL ANTIBODY
ANTI:-TCRdl SPECIFICALLY REACTIVE WITH
THE TCR DELTA SUBUNIT CONSTANT REGION
6.1. EXPERIMENTAL PROCEDURES
6.1.1. CYTOFLUOROGRAPHIC ANALYSIS
~F T-CELL LINES WITH ANTI-TCRal The. anti-~TCRdl mAb, which is specifically reactive with the 6Z'CR chain constant region, was made as follows: One gram of PEER cells were solubilized in 50 ml of 0.3% CHAPS (3-[3--cholamidopropyl)dimethyl-ammonio]1-propanesulfona.te) dEaergent and were immunoprecipitated with 1 ~1 of UCHT1 (Beverley, P.C., and Callard, 1981, Eur. J.
Immunol. 11:329-334) ascites, 500 ~l of mAb 187.1 culture supernatant arid Staphylococcus aureus Cowan I strain (SACI).
Four intraperi.toneal: injections at six week intervals were carried out, followEad by a final boost of 7dTCR (without CD3) isolated by se7lective elution of y6TCR from the immune complexes using 2% 7Criton X-100. The eluted material was administered both intravenously and intraperitoneally: four days after this boo:at, the mice were sacrificed and fusion carried out a~; previously described (Brenner, M.B., et al., 1987, J. Immunol. 1;18:1502-1509).
~, , 1341042 _ y6TCR ce:Ll lines PEER and IDP2 or a~TCR cell lines HPB-MLT and JURKAT were stained with 50 ~l of anti-TCRdl culture supernatant followed by staining with FITC-conjugated goat anti-mouse Ig F(ab)'2 fragments with analysis on an Ortho*cytofluorograph (see Fig. 12). Control was the mAb secreted by P3X63.Ag8 hybridoma (P3) and anti-CD3 mAb was anti-Leu 4 (Ledbetter, J.A., et al., 1981, J.
Exp. Med. 153:310).
0 6.:1.2. IMMUNOCHEMICAL ANALYSIS OF THE
SPECIFICITY OF mAb anti-TCR61 Surface '125I-labeled IDP2 cells were solubilized and their provteins immunoprecipitated using control mAb P3, anti-Leu 4, anti-TCRal, or anti-C~ serum. Precipitated samples were analyzed by SDS-PAGE followed by ~5 autoradiograplay. In CHAPS detergent, ybTCR and CD3 remained associated and were immunoprecipitated as a complex by anti-Leu 4 (F:ig.l3, lanes 3 and 4). However, after solubilization in 2% Triton X-100 detergent, anti-TCR61 immunoprecipitated ybTCR as a dimeric complex without CD3 20 (lane 6) and anti-Leu 4 immunoprecipitated CD3 as a trimeric complex without ydTCR (lane 5). After separation of the ySTCR-CD3 component chains, anti-TCR61 immunoprecipitated TCRa alone (lane 7 .and 8), while anti-C~ serum immunoprecipitated TCR y alone (lane 9). For chain 25 separation experiments (lane 7-9), anti-Leu 4 immunoprecipitates from CHAPS solubilized IDP2 cells were boiled in 1% ,3DS and were then diluted with 4 volumes of 2%
Triton X-100 :followed by immunoprecipitation with anti-TCRS1 or anti-C~ se~:wa. 'this follows procedures used previously 30 (grenner, M.B., et al., 1986, Nature 322:145).
*Trade-mark 1341042 _ 6.1.3: N-L:CNKED ~sLYCOSYLATION OF THE TCR b POLYPEPTIDE
1251-labeled IDP2 cells were solubilized in 0.3%
CHAPS, immu~noprecipitated with anti-Leu 4 and resolved by SDS-PAGE (Fig. 14). Control lane is mock-digested IDP2 dTCR
polypeptide. N-glycanase digestion of bTCR polypeptide was performed as follows: aTCR was eluted from a gel slice followed by N-glycanase (Genzyme Corp.) digestion (10 U/ml) carried out in 30 ~1 0.17% SDS, 1.25% Nonidet P-40, 0.2 M
sodium phosphate buffer pH 8.6 for 16 hours at 37'C
(Tarentino, A.L., et al., 1985, Biochemistry 24:4665). The digested or mock-incubated bTCR samples were analyzed by SDS-PAGE and visualized by autoradiography.
6.1.4. RECOGNITION OF IN VITRO TRANSLATION PRODUCTS
_OF cDNA CLONE IDP20-240/38 BY mAb ANTI-TCRbl ~5 A plasmid, designated pGEM3-O-240/38 was constructed as follows and used for in vitro transcription-translation (Fig. 15). The IDP2 O-240/38 (bTCR) cDNA clone 1.5 kb insert begins within codon 7 of the composite Group O
sequence and includea the remaining coding region and most 20 of the 3' untranslat.ed region. This insert Was cleaved as a single EcoRI fragment from agtl0 arms by partial EcoRI
digestion (to prevent cleavage of the internal EcoRI site).
This fragment was su~bcloned into a Bluescript~+ vector (Stratagene). The insert was then removed from the vector 25 as a single BamHI-Sa:lI fragment (ends are from the Bluescript vector polylinker) facilitating directional . cloning into pGEM-3 (Promega Biotech) downstream of the T7 promoter. The resultant pGEM3-O-240/38 plasmid was linearized with SalI: and capped transcripts synthesized 30 using T7 RNA polymerase (Krangel, M.S., et al., 1987, Science 237:64). Integrity and size of the transcripts were monitored via an aliquot of the reaction mixture containing 32P-ATP. A single FtNA species of 1.5 kb was observed. _In vitro translation in the presence of 35S-methionine was *Tr. ade-mark .. 1 341 042 performed in a rabbit reticulocyte lystate. After in vitro translation, l:he samples were boiled in 1% SDS with 2 mM
dithiothreitoll followed by the addition of 10 volumes of 2%
Triton X-100 in Tri;s buffered saline pH 7.5. Samples were immunoprecipii:ated with control mAb P3 (Fig.l6, lanes 1 and 3) or with ani:i-TCR~51 mAb (lanes 2 and 4) and analyzed by SDS-PAGE followed by fluorography (Bonner, W.J. and Laskey, R.A., 1974, Eur. J. Biochem. 46:83-88).
6.:2. EXPERIMENTAL RESULTS
We have .generated a monoclonal antibody (mAb), anti-TCRbl, that is specifically reactive with the dTCR
constant region.
The 7bTCR-CD3 complex from the PEER cell line (Weiss, A., ei~ al., 1986, Proc. Natl. Acad. Sci. U.S.A.
83:6998-7002; Brenn~er, M.B., et al., 1987, Nature 325:689-694) was used as immunogen in the production of antibody-secreting hybridoma cell lines. Hybridomas were screened both by cell :surface-binding (cytofluorographic analysis) -and by immunoprecipitation of PEER cell proteins followed by SDS-PAGE analysis. Two hybridoma supernatants (5A6 and 4A1) bound to the aurface of PEER cells. After subcloning, one mAb (5A6.E9) was characterized further. This mAb bound to the surface o:f ybTClt lymphocytes (PEER, IDP2) but failed to react with a~TCR cells (HPH-MLT, JURKAT) or with non-T
leukocytes (Fig.l2 and data not shown). Although the immunogen was composed of a complex of ~dTCR and CD3, the greater affinity of the mAb for ybTCR cell lines suggested the mAb was rnot directed against CD3 determinants.
The specificity of the mAb was determined in immunoprecipitation studies using various detergents which affect the association of the proteins comprising the receptor complex. After 1251-labeled IDP2 cells were solubilized in CHAP'S detergent, TCR ~ and d, and CD3 y,b, and a subunita remained part of an associated complex 1 341 042 _ immunoprecipii:ated ;by anti-CD3 antibody (Fig.l3, lanes 3, 4). However, if radiolabeled IDP2 cells were solubilized in 2% Triton X-100 detergent, ~dTCR and CD3 became largely dissociated, <ind the use of anti-CD3 mAb resulted in selective pre<:ipitation of CD3 (Fig. l3, lane 5). Under these latter conditions, mAb 5A6.E9 immunoprecipitated -~aTCR
as a heterodiiaer without associated CD3 (Fig.l3, lane 6).
This observation provided the first direct evidence that TCR y and TCR d exist as a non-disulfide-linked heterodimer.
0 To determine whether mAb 5A6.E9 reacts with a yTCR chain, STCR chain or a combinatorial determinant, immunoprecipii:ation of separated polypeptide chains was performed. A~~ anti-Leu 4 immunoprecipitate from radiolabeled, CHAPS-solubilized IDP2 cells was boiled in 1%
SDS to dissociate tlhe ~TCR, 6TCR, and CD3 proteins. After dilution with four volumes of 2% Triton X-100, mAb 5A6.E9 specifically :immunoprecipitated the 40 kD (6TCR) species (Fig. 13, lane 7). ~idhen an aliquot of the same immunoprecipii:ate was analyzed under-reducing conditions (Fig.l3, lane 8), a dramatic shift in SDS-PAGE mobility was observed. This phenomenon is characteristic of dTCR from the IDP2 and 1?EER cell lines (Hrenner, M.B., et al., 1987, Nature 325:689-694). In contrast, when the separated chains were immunoprecipit;ated with anti-C7 sera, the 55 kD species (~TCR), but not the 40 kD species (bTCR) was immunoprecipii:ated (Fig. l3, lane 9). Based on these biochemical and surface binding studies, mAb 5A6.E9 is referred to as anti-TCRbl.
In addition to PEER and IDP2, anti-TCRbl also l~unoprecipii:ated 'rCR d from other ybTCR cell lines including MOL'.C-13 a:nd PBL line 2. Further experiments have shown that ani~i-TCR~61 reacts with a determinant encoded by a TCR d constant: (C) gene segment.
G

1341042 .

We have :isolated cDNA clones from the IDP2 cell line (e. g., II)P2 O-x40/38) by the subtractive approach representing a gene which encodes the TCR d subunit. Genes to which IDP2 group O cDNA clones hybridize in Southern blotting exileriment:a are expressed and rearranged in ~6TCR lymphocytes bu.t are typically not expressed (and are often deleted) in a~~TCR cells. By sequence comparison with other TCR genEa, these cDNA clones appear to be composed of novel V, D (?), J, and C gene segments. The IDP2 Group 0 0 composite DNA sequence contains a long open reading frame predicting a polypeptide with two potential asparagine-linked glycosylation sites and a molecular weight of 31.3 kilodaltons. To deitermine the molecular weight of the unglycosylated 6TCR protein and the number of asparagine-~5 linked carbohydrates that are present on the mature IDP2 dTCR polypepti.de, gEel purified dTCR was either treated with N-glycanase or mock-incubated and analyzed by SDS-PAGE (Fig.
14). Removal of N-linked carbohydrates resulted in a 5 kD
decrease in apparewt-molecular weight (40 -kD to 35 kD),-20 suggesting the presence of two (2.5-3 kD) N-linked glycans on the IDP2 dTCR. '.this correlates well with the number of N-linked glyc<ins predicted by the translated amino acid sequence in Figure :37 A and B. The apparent molecular weight of the protein is in general agreement, differing 25 from that predicted by 3.7 kD.
Given the reactivity of anti-TCRbl on IDP2 cells, the specificiity for the dTCR polypeptide, and the recognition o:E partially denatured (SDS boiled) sTCR, we tested whether this mAb would recognize directly polypeptide 30 encoded by this bTCR cDNA clone. Thus, the insert from cDNA
clone IDP2 O-:240/38 was subcloned into the pGEM-3 expression vector downstream of the T7 promoter (Fig. 15). Transcripts generated in yitro with T7 RNA polymerase were then used in a rabbit reticulocyte lysate system to direct the synthesis 35 of protein in the presence of 35S-methionine. Following in '~ ..~~.

1341 p42 vitro transcription-'translation, the reaction mixtures were boiled in 1% SDS, diluted with ten volumes of 2% Triton X-100, and then :immuno;precipitated with either an isotype-matched control mAb ~or with anti-TCRS1. Anti-TCRd1 mAb specifically immunoprecipitated a predominant species (34 kD) (Fig. l6,lane 4). No such band was observed in immunoprecipitates when control mAbs were used (lane 3), when RNA transcripts were omitted (lanes 1 and 2), or when yTCR constructs were used. Thus, the radiolabeled species immunoprecipitated by mAb anti-TCRS1 corresponds to a d polypeptide whose synthesis was specifically directed by the IDP2 O-240/38 cDNA clone. This polypeptide (34 kD) is very similar in size to the N-glycanase treated IDP2 bTCR chain (35 kD). The :IDP2 O-240/38 clone lacks a natural ATG
~5 initiation codon as well as the leader sequence. There are two potential internal ATG codons (at residues 12 and 44) within the V region of this clone (Fig. 37A). Use of these codons to initiate synthesis could result in more than one polypeptide species-possibly accounting for-the--minor -20 species noted (Fig.l.6, lane 4). Thus, there is direct serological recognition by mAb anti-TCR61 of the IDP2 3TCR
subunit encoded by clone IDP2 O-240/38.
7. GENERATION OF MONOCLONAL ANTIBODY
' sTCAR-3 SPECIFICALLY REACTIVE WITH

7.1. EXPERIMENTAL PROCEDURES
7.1.1. 7:MMUNOIPRECIPITATION AND SDS-PAGE ANALYSIS.
OF T CELL ANTIGEN RECEPTOR
30 The dTCAR-3 mAb, specifically reactive with the variable region of the ~TCR chain, was generated as follows:
One mouse was immunized with 2 x 107 Molt-13 cells by intraperitoneal injection. One month later, the mouse was boosted with 1 x 107 Molt-13 cells by intravenous injection _.. SD101 ~ ~ ~ ~ O 4 z _ each day for :3 sequential days, and then immune splenocytes Were fused wi1_h mouse myeloma P3x63Ag8.653 cells in the presence of 50% polyethylene glycol 1500. The hybridomas were screened by analyzing the CD3 co-modulation with flow cytometry. The ana:Lysis of CD3 co-modulation was based on the observation that antibody to a,p T cell antigen receptor, when incubated with the cells, caused the internalization of ithe CD3 complex (Lanier, L.L., et al., 1986, J. Immunol. 1:37:2286; Meuer, S.C., et al., 1983, J.
E~~ Med. 157;;705) .
Mo7lt-13, PEER, and HPH-ALL cell lines were iodinated using the lactoperoxidase technique. The 1251-~F1 labeled cells were solubilized in Tris-buffered saline (pH
8) containing 1% Triton X-100. Lysates were immunoprecipit:ated using aTCAR-3 antibody or ~9F1 antibody.
~F1 is a framework monoclonal antibody to the pTCR chain and is described e.lsewh~ere (Brenner, M.B., et al., 1987, J.
Immunol. 138::L502-1!509). All samples were analyzed by SDS-PAGE-under seducing or non-reducing-conditions (Fig. 17).
Molt-13 and PEER are both CD3+4 8 WT31 . HPB is CD3+4+8+WT31+., As shown in Figure l7,bTCAR-3 immunoprecipitated non-disulfide--linked y and 6 chains from Molt-13 and PEER
cells, while ~3F1 immunoprecipitated disulfide-linked a and ~
chains from H1?B-ALL cells. The difference in autoradiographic intensity between the bands corresponding to the b and ~~ chains represents differences in the extent of iodination of these two proteins.
7~1~2. IMMUNOPRECIPITATION OF bTCR
CHAIN BY dTCAR-3 ANTIBODY
Figure l8;shows 1251-labeled Molt-13 cells solubilized in Tris~-buffered saline (pH 8) containing 0.3%
CHAPS (3-[(3-<:holam.idopropyl)dimethylammoni]1-propanesulfonate) o:r in 1% Triton X-100. In 1% Triton X-G

SDloa 100, the~ydTCR. dissociates from the CD3 complex, while in 0.3% CHAPS, the ~bTC;R remains associated with the CD3 complex. Prior to :immunoprecipitation, the 1251-labeled lysates used in lanes 3, 4, and 7 of Figure l8were denatured by adding SDS to a :final concentration of 1% followed by heating for 5 minutes at 68'C. After cooling, iodoacetamide was added to a final concentration of 20 mM. The mixture was then diluted with 4 volumes of 1.5% Triton X-100 in Tris-buffered saline (pH 8). This denaturing process 0 completely dissociaites ~ chain, d chain, and CD3 proteins from one another. All samples were analyzed by SDS-PAGE
under non-reducing conditions (N) except for the sample in lane 4 which is under reducing conditions (R). Note the difference in mobility of d chain under reducing and non-5 reducing conditions. The anti-C~ antiserum was generated by immunizing a rabbit with a 20 amino acid synthetic peptide from the 7 constant region (residues 117-136).
?.1.3. ANALYSIS- OF'-CELL SURFACE STAINING BY FLOW CYTOMETRY
0 5 7~ 105 cells were incubated with the appropriate antibodies (NMS (no:rmal mouse serum), STCAR-3, OKT3, or WT31) at 4'C iEor 30 minutes and then Washed two times with 0.2% BSA in PBS (pH 7.4). Following incubation with fluorescein-conjugated goat anti-mouse IgG for 3o minutes at 25 4°C, cells were analyzed on an Ortho cytofluorograph (Fig.
19) .
7.1.4. TWO COLO~t CYTOFLUOROGRAPHIC ANALYSIS OF bTCAR-3+

Thca peripheral blood lymphocytes were first 30 incubated with dTCAI~-3 at 4°C for 30 minutes. After washing, cell:a were incubated with phycoerythrin (PE)-conjugated goat anti-mouse IgG for an additional 30 minutes at 4'C. After washing, the cells were incubated with 35 *Trade-mark ~ 341 442 fluorescein (FITC)-conjugated OKT3 for 30 minutes at 4'C and then cells were analyzed on an Ortho cytofluorograph (Fig.
20) .
7.1.5.. MEASUREMENT OF INT~CYTOPLASMIC Ca2+
CONCENTRATION ([Ca ]i) VERSUS TIME
Mo7.t-13 cells were labeled with the acetoxymethyl ester form of the Ca2+-sensitive probe fura-2 (2 ~M from a 1 mM stock in dimethy:l sulfoxide, Molecular Probes, Eugene, Oregon) at a c:oncenitration of 107 cells/ml in RPMI 1640 plus ~ 10% fetal bovine serum for 30 minutes at 37'C. Cells were then washed and resuspended at 107 cells/ml in Hanks balanced salt solution (HBSS) plus 5% fetal bovine serum and kept in the dark at room temperature until use. Immediately prior to fluorescent measurement, 2 x 106 cells were 5 centrifuged then resuspended in 2 ml of fresh HBSS and placed in a quartz cuvette at 37'C and constantly stirred.
Fluorescence was measured on the cell suspension in a SPF-500C fluoromei:er (S:LM Aminco, Urbana, Illinois), the excitation wavelength alternating between 340 (+2) and 380 ~ (+2) nm and emission was detected at 510 (+5) nm. The ratio of 350/380 wa:~ automatically calculated (1 ratio every 2 seconds), ploi~ted, .and stored in an IBM PC AT. Quantitation of [Ca2+]i from the fluorescence ratio was performed as described by csrynki~ewicz, et al. (1985, J. Biol. Chem.
25 260:3440). A<iditio:n of irrelevant antibodies did not alter [Ca2+]i, while. cell lysis resulted in a [Ca2+]i of 1 ~M.
7.2. RESULTS
We have .generated a monoclonal antibody, 30 gTCAR-3, that is directed against a variable region of the TCR 3 chain and which can be used to characterize the b polypeptide. This :monoclonal antibody binds to T cells bearing the y6~TCR and also elicits a fura-2 Ca2+ signal upon binding to Molt-13 cells.
G

SD104 ~ 3 4 ~ 0 4 2 The. dTCAIt-3 monoclonal antibody was generated by immunizing a mouse with the Molt-13 cell line which has a CD3+4 8 WT31 phenotype. The hybridomas were first screened by CD3 co-modulation. The positive clones were further screened by iramunop:recipitation. dTCAR-3 immunoprecipii:ation of 7bTCR heterodimer from 1251-labeled Molt-13 and PhER ly;sates is shown in Figure 17. 6TCAR-3 does not immunoprec:ipita,te any polypeptide from HPB-ALL (Fig. 17).
In contrast, E3F1, a framework monoclonal antibody specific to the ~ chain (Hrenner, M.B., et al., 1987, J. Immunol.
138:1502-1509), immnsnoprecipitates the a~ heterodimer from the HPH-ALL cell line (Fig.l7, lanes 10 and 12). The immunoprecipii:ated ~rd receptor from both Molt-13 and PEER
cells, when analyzed under either reducing or non-reducing conditions, displays a heterodimeric structure indicating a non-disulfide--linked ySTCR in these two cell lines. There is a slight shift in mobility of the d chain under reducing conditions re7lative to that observed under non-reducing conditions (Fig. l7,lanes 1-and 3; 5 and a), a phenomenon which has been noted previously in IDP2 and PEER cell lines (Brenner, M.B., et al., 1987, Nature 325:689-694), suggesting the: existence of intrachain disulfide linkages.
In order to de~monst:rate that the dTCAR-3 antibody recognizes a CD3-associai:ed ~aTCR, immunoprecipitations were performed using 125I-labeled lMolt-13 cell lysates solubilized in 0.3%
CHAPS detergent (Fig. l8, lane 1). Under these conditions, the CD3 complex remains associated with the receptor, and both ~d heterodimer and the CD3 complex are immunoprecipii:ated lby 6TCAR-3. However, when 1251-labeled lysates were :~olubi.lized in 1% Triton X-100 detergent which largely disso<:iates the CD3 complex from the ~d receptor, only ~b heterodimer is immunoprecipitated by 6TCAR-3 (Fig.
l8,lane 2). As a control, the anti-CD3 antibody, UCHT-1 ~r-~
~'~.

SD105 ~ 3 4 ~ 0 4 2 y (Beverley, P.C. and Callard, R.E., 1981, Eur. J. Immunol.
11:329-334) immunoprecipitates only the CD3 complex, but not the yb heterod:imer (Fig.lB, lane 5).
The specificity of 6TCAR-3 was further analyzed by using immunopreci.pitations of denatured, 1251-labeled Molt-13 lysates in which y,bTCR and CD3 proteins were completely dissociated. dTCAR-3 specifically immunoprecipitated t:he d chain which has an apparent molecular weight of 38 kD under non-reducing conditions (Fig. 18, lane 3) andl 40 kD under reducing conditions (lane 4). The anti-Cy antiserum immunoprecipitated the ~ chain with molecular weight 42 kD under reducing conditions (Fig.
18, lane 7). These data indicate that dTCAR-3 is d chain specific.
~5 6TCAR-3 n.ot only immunoprecipitates 7,dTCR
heterodimer from the: PEER and Molt-13 cell lines, it also binds to the surface: of these cell lines and to the IDP2 clone (Brenner, M., et al., 1987, Nature 325:689-694). It Sees -not-bind to thE: apTCR~bearing HPH-ALL and -Jurkat -cell 20 lines (Fig. 19). In contrast, WT31 (Tax, W.J.M., et al., 1983, Nature 304:44_-'i-447), a framework monoclonal antibody to the a~TCR, reacts with a~TCR-positive HPB-ALL and Jurkat cell lines, bu,t not with ~dTCR-positive Molt-13, PEER, and IDP2 cells (Fi.g. 19).. When normal peripheral blood 25 l~phocytes ( F~BL) wE:re examined, a subpopulation ( 0 . 9-2 . 4 % ) of CD3+ lymphocytes were positive with dTCAR-3 (Fig. 20).
When dTCAR-3, immobilized on tissue culture plates was used for culture of normal human PBL, it selectively stimulated the proliferation o1' the ybTCR-positive subpopulation.
30 After 45 days in cu7Lture, the ~dTCR subpopulation represented 9E.% of t:he total cell count.
Ant:ibodiEa to the a~ T cell antigen receptor stimulate a rise in the cytoplasmic free calcium ion concentration [Ca2+;~i (Weiss, A., et al., 1986, Ann. Rev.
35 I~°unol. 4:593.). Incubation of Molt-13 cells with bTCAR-3 elicited~a rapid inc;rease in [Ca2+]i similar to the response induced by anti-T3 antibodies (Fig. 21). Moreover, 6TCAR-3 similarly stimulated a Ca2+ flux in PEER cells and in the 76TCR-positive cell line generated from PBL as described above. We have also observed that incubation of Molt-13, PEER, and IDP2 cell=~ with bTCAR-3 causes the co-modulation of the CD3 protein c;omplex.
Further characterization of the epitope specificity of mAb dTCAR-3 (also termed mAb TCSSl) is 0 presented in Section 11.2.2, infra.
8. THREE FORMS OF THE HUMAN T CELL RECEPTOR ~d:
PREF'ERENT7:AL USE OF ONE FORM IN SELECTED
HEAhTHY INDIVIDUALS
In the examples herein, the structure of a new form of the human T cell receptor ~3 (ydTCR), consisting of a 40 kD TCR y glycoprotein noncovalently associated with a TCR d chain, is presented. The newly identified 7TCR
glycoprotein, termed Form 2bc, differs in size by more than 15 kD (40 kD versus 55 kD) compared to the previously described nondlisulfi~de-linked TCR y form (Form 2abc). This difference is accounted for by a 5 kD smaller polypeptide backbone size (35 kI) versus 40 kD) and by a reduction in the amount of carbohydrates (5 kD versus 15 kD). Nucleotide sequence analysis o1: cDNA clones corresponding to Form 2bc revealed that Form :!bc cDNA clones lacked one copy of the constant region (C~2} second exon that is present in the cDNA of the other nondisulfide-linked TCR y subunit (Form 2abc). This CII exon copy encodes part of a connector region between the membrane spanning region and the extracellular constant domain. Since the number and localization of the potential N-linked carbohydrate attachment sites is the same in both nondisulfide-linked forms, we conclude that the connector region influences the amount of attached carbohydrates, probably by affecting the conformation of the: protein. In contrast, the dTCR subunits of these ybTCl2 forms show little variability in peptide backbone sizes or peptide mapping analyses.
We also examined the usage of the three forms of the 7bTCR complex i:n peripheral blood. Nearly exclusive use of the disulfide-linked form, Form l, was observed in certain healthy subjects. In some individuals, Form 1 was expressed together with Form 2bc. Form 2abc was not identified in the subjects tested.
8.1. EXPERIMENTAL PROCEDURES
8.1.1. ANTIBODIES
Monoclonal antibodies used were anti-Leu4 (anti-CD3) (Ledbetter et al., 1981, J. Exp. Med. 153:310-323), ~F1 (anti-~TCR) (:Brenner et al., 1987, J. Iminunol. 138:1502-1509), anti-TCRb1 (anti-6TCR) (described in Section 6, supra; reactive with the dTCR chain constant region), P3 !control)- (secreted. by_P3X63.Ag8;. Koehler and Milstein, 1975, Nature 256:495-497), 187.1 (rat anti-mouse K light chain) (Yelton et al., 1981, Hybridoma 1:5-11), and WT31 (stains a~TCR lymphocytes brightly) (Spits et al., 1985, J.
Immunol. 135:1922-1928). Anti-Cyb peptide serum (anti-7TCR) was generated against a 22 amino acid synthetic peptide (Gln-Leu-Asp-.Ala-Asp-Val-Ser-Pro-Lys-Pro-Thr-Ile-Phe-Leu-Pro-Ser-Ile-Ala-Glu.-Thr-Lys-Cys) (PCT International Publication No. WO 88/00209, published January 14, 1988).
8.1.2. CELL LINES
PEER (We.iss et al., 1986, Proc. Natl. Acad. Sci.
U.S.A. 83:6998-7002) and MOLT-13 (isolated by J. Minowada, Loh et al., 1987, Nature 330:569-572) are T leukemic cell lines. Umbilical cord blood derived clone WM-14 (Alarcon et a.'., 1987, Proc. Natl. Acad. Sci. U.S.A. 84:3861-3865) and F~rlPheral blood derived cell line IDP2 (Brenner et al., G

~34~~42 -1986, Nature 322:145-149; PCT International Publication No.

9, publisrued January 14, 1988) and thymus-derived Clone II (Bank. et al.., 1986, Nature 322:179-181) were cultured as described earlier. Peripheral blood derived cell line 2 (P~BL-L2) was isolated by sorting peripheral blood isolated. lymphocytes that did not stain with mAb WT31.
The isolated cells were then expanded in vitro in RPMI 1640 medium supplemented with 10% (v/v) conditioned medium containing IL-2 and 10% (v/v) human serum, and stimulated 0 every 3 weeks with irradiated autologous feeder cells.
8.1.3. _IOI)INATION AND IMMUNOPRECIPITATION
2 x: 107 cells were isolated by Ficoll diatrizoate (Organon Tekni.ka Co:-p.) centrifugation and iodinated on ice 5 in 0.5 ml of phosphate-buffered saline, pH 7.4 (PBS) containing 1 aiM MgC7.2, 5 mM glucose by adding 100 ~g of lactoperoxidasce (80-~100 U/mg, Sigma) and 1 mCi of Na125I
(New England Nuclear). Ten ~l of a 0.03% hydrogen peroxide solution was added ~tt- 5 minute intervals over- a reactor.
20 period of 30 a~inute:~. Cells were solubilized overnight in detergent supplemeni_ed TBS (50 mM Tris-Base pH 7.6, 140 mM
NaCl) containing 1 mM phenylmethylsulfonyl fluoride (PMSF, Sigma) and 8 mM iodoacetamide (IAA, Sigma). As indicated, different detE:rgents used in this study were 0.3% (w/v) 3-25 [(3-cholamidopropyl;l dimethylammonio] 1-propane-sulfonate (CHAPS, Signma), 1% (w/v) digitonin (Aldrich) and Triton X-. 100 (TX-100, :4igma). After 20 minutes of centrifugation at 10,000 x g to remove insoluble material, detergent lysates were precleared by a 30 minute incubation with 4 ~1 of 30 normal rabbit serum (NRS) and 400 ~1 of 187.1 hybridoma culture mediuDn, followed by addition of 200 ~1 of a 10%
(w/v) cell su:~pension of fixed Staphylococcus aureus Cowan I
(Pansorbin, Calbioclzem). After a one-hour incubation, Pansorbir~ was removed by centrifugation. Specific 35 precipitation: were carried out by adding 0.25 gal ~9F1 *Trade-mark ~ 341 042 ascites,~l ~1 1 mg/ml anti-Leu4 or 0.25 ~cl P3 ascites, together with 150 ~7. of 187.1 culture supernatant to each sample, followed by a one-hour incubation. 100 ~1 of 10%
(v/v) Protein A-Sepharose*(Pharmacia) was added and the mixture was rocked it:or 1 hour at 4'C. Immunoprecipitates were washed five times with 0.1% (v/v) Triton X-100 containing TBfc and analyzed by sodium dodecyl sulfate-polyacrylamide: gel electrophoresis (SDS-PAGE) (Laemmli, 1970, Nature 227:680-685).
0 For immunoprecipitations with the anti-Cyb peptide serum, iodinated cells were solubilized in 1% (w/v) sodium dodecyl. sulfate (SDS) containing TBS and then boiled for 3 minutes. AftE:r cooling, 5 volumes of 2% (v/v) Triton X-100 in THS containing PMSF and IAA was added, together 5 with 200 ~1 of a mi~,;ture of 1 mg/ml deoxyribonuclease (DNAse) and 0.5 mg/ml RNAse in 50 mM MgCl2. Preclearing and immunoprecipit:ation.a were performed as described above, omitting the addition of 187.1 mAb. Immunoprecipitates were washed in THS containing 0~5%-(v/v) TritonX-100; 0.5% -(w/v) 20 deoxycholate (DOC), 0.05% (w/v) SDS.
8.1.4. BIOSYNTHETIC LABELLING
4 ~: 107 Exponentially growing cells were resuspended in 4 ml of methionine and cysteine-free RPMI
25 1640 (Select-Amine kit, Gibco) supplemented with 10%
dialyzed fetal. calf serum (FCS) and 20 mM Hepes. After a 30 minute starvation pEariod at 37°C, 1 mCi of 35S-methionine and 1 mCi of 355-cy:ateine were added, allowing a 15 minute labelling period. Cells were harvested and solubilized in 30 2% (v/v) Triton X-100, TBS. Preclearing and immunoprecipit:ation:a were performed as described above. The immunoprecipit:ates were washed four times in 0.5% (v/v) Triton X-100, 0.5% I;w/v) deoxycholic acid, 0.05% (w/v) SDS, TBS followed by three washes in 0.5% (v/v) Triton X-100, 0.5 35 M NaCl, 5 mM E;DTA, _-'i0 mM Tris, pH 7.6. The samples were *'I'ra~e-~~ark .w analyzed by SDS-PAGE and visualized by standard fluorography procedures (Banner .and Laskey, 1974, Eur. J. Biochem.
46:83-88).
8.1.5. GF:L PURIFICATION OF bTCR PROTEINS
Surface .iodinated cells were solubilized in 0.3%
(w/v) CHAPS-T13S and immunoprecipitated using 50 ~1 of anti-Leu4-coupled Sepharose beads. The immunoprecipitated species were resolved by SDS-PAGE under nonreducing conditions anti the wet gel was exposed for 24 hours at 4°C
on XAR-5 film (Koda:k) to visualize radiolabelled dTCR
proteins. The. gel :regions corresponding to STCR were excised, incubated .in 5% (v/v) 2-mercaptoethanol containing sample buffer and resolved a second time by SDS-PAGE.
Because of they characteristic SDS-PAGE mobility shift upon reduction, 6TC:R proi~ein could be separated and then purified from contaminants. TCR proteins were eluted from gel slices by overnight :incubation in 0.05% (w/v) SDS, 50 mM ammonium bicarbonate-buffer ;at-37°C and lyophilized.
8.1.6. ENDOGLYCOSIDASE DIGESTION
For endoglycosidase H (Endo H) digestions, immunoprecipii:ated material or gel purified protein was boiled for 3 minutes in a 40 ~l 1% (w/v) SDS solution containing O.:L4 M 2~-mercaptoethanol. After cooling, the mixture was diluted with 360 ~1 of 0.15 M acetate buffer, pH
5.5 containing 1 mM PMSF. Five ~1 Endo H (1 U/ml- Endo-~-N-acetylgluco:aaminidase H, Genzyme) was incubated with half of the above :~oluti~on for 14 hours at 37°C, while the other half was mock treated.
For N-gl;ycanase (N-GLY) digestion, gel purified material was boiled for 3 minutes in 35 ~1 of 0.5% (w/v) SDS, 0.10 M 2~-merca~ptoethanol. Then, 100 ~1 of 0.2 M sodium phosphate (pH 8.6), 1.25% (v/v) Triton X-100 was added.
Half of the mixture was incubated with 1 ~1 N-Glycanase (250 *Trad~-mark G

1341p42 U/ml~, peptide-N-[N-acetyl-~-glucosaminyl]asparagine amidase:
Genzyme) and incubated for 16 hours at 37°C, while the other half was mock treated.
After digestion, 10 ~g bovine serum albumin was added as carrier anct samples were recovered by trichloroacetic acid precipitation. Protein pellets were taken up in sample buffer containing 5% (v/v) 2-mercaptoethana~l.
1~ 8.1.7. PRODUCTION OF MONOCLONAL ANTIBODY anti-Cyml Pazt of t:he C~ CI and CII exons of HPB-MLT pTy-1 was isolated using t:he BamHI and PstI sites at nucleotide positions 571 and 84E8 (Dialynas et al., 1986, Proc. Natl.
Acad. Sci. US1~, 83:2Ei19-2623) and was cloned into expression vector pRIT2T (Pharmacia). The resulting Protein A fusion protein was ex:pressE:d in E. coli N4830. Bacteria were lysed with lysozyme and the fusion protein was isolated by purification over a IgG Sepharose column. Mice were injected intra:peritoneally pith 100 ~g of fusion protein in ~ Freund's adjuvant at: days 0, 7 and 28. Twenty-eight days later 100 ~g of fusion protein in PBS was injected intravenously. AftE:r three days, splenocytes were isolated and fused with the hybridoma P3X63Ag8.653 as described (Brenner et al.., 19F37, J. Immunol. 138:1502-1509).
Hybridomas were scrE:ened by enzyme-linked immunoabsorbent assay (ELISA). Ninety six-well flat bottom plates (LINBRO, Flow Laborato=vies) were incubated overnight with 0.4 ~g of fusion protein or nonfused protein in PBS. Nonspecific binding sites were blocked at 23°C with 0.25 mg/ml normal rabbit IgG (Si.gma) :in PHS containing 50% (v/v) FCS. 50 ~1 of hybridoma :supernatant was added for 1 hour at 4°C, followed by a similar incubation in 50 ~cl of a 5 ~g/ml solution of ps:roxidase-conjugated anti-mouse IgG (Cappel).
All described incubations were interspersed with washing steps, using 7.0% (v,iv) FCS, 0.1% (w/v) BSA, PBS. The ELISA
'__ y 1 341 p42 was developed with 0.08% (w/v) O-phenylenediamine (Sigma) in 0.012% (w/v) :hydrogen peroxide containing phosphate-citrate buffer, pH 5Ø
Although anti-Cyml (IgGl) does not recognize the native ybTCR/c:D3 co~aplex in cytofluorographic analysis nor the ybTCR hetE:rodim~ar from Triton X-100 solubilized cells in "
immunoprecipitation, it does recognize biosynthetically labelled ~TCR precursor and mature ~TCR proteins after separation of CD3/~~6TCR proteins into individual chains. In this way, anti-Cyml was shown to recognize the -~TCR protein after separating CD3/y6TCR complexes into individual chains by boiling anti-CD3 immunoprecipitates in 1% (w/v) SDS in TBS (Fig. 23, lane 3).
8.1.8. ISOLATION AND SEQUENCING OF A
MOLT-13 ~TCR CDNA CLONE
Poly (A)+ RNA was prepared from MOLT-13 cells by urea/lithium .chloride precipitation followed by oligo (dT) cellulose affinity chromatography. A agt 10 cDNA library was prepared from poly(A)+ RNA by the method of Huynh et al., 1985 (DN,A Cloning, Glover, D.M. ed. IRL Press, Oxford, I:49-78) using Mung Bean Nuclease for the hairpin loop cleavage (McCutcham et al., 1984, Science 225:626-628). The cDNA library 'was amplified on the E. coli strain C600 Hfl and screened :by plaque filter hybridization with 32P-labelled PTyI (Dial;ynas et al., 1986, Proc. Natl. Acad. Sci.
U.S.A. 83:2619-2623). Positive clones were analyzed for size and restriction enzyme map, and cDNA clone Ml3k was selected for sequencing. The cDNA of Ml3k was excised from agt 10 phage with t:he endonuclease EcoRI and further digested with appropriate restriction enzymes. The fragments were subcloned into M13 vectors and sequenced by the dideoxy chain termination method (Sanger et al., 1977, Proc. Natl. Ac:ad. Sci. U.S.A. 74:5463-5467) using the modified T7 polymerise (Sequenase, United States Biochemical Corp.).
Clone Ml:3k corresponds to a full length, in frame, ~TCR transcript, including 36 nucleotides of 5' untranslated region and 72 nucleotides of 3' noncoding region (Fig. 25). The nucleotide sequence ~f the V region is identical t:o the genomic V71.3 sequence (nomenclature Lefranc et al., 1986a, Cell 45:237-246: Strauss et al., 0 1987, Science 237:1;217-1219), except for a C to T (Ile to Val) change o1° nucleotide 53 in the putative signal sequence. ThE: J region is identical to the J72.3 sequence (nomenclature based on Lefranc et al, 1986b, Nature 319:420-422: Querte:rmous et al, 1987, J. Immunol. 138:2687-~5 2690). Intere=stingily, 8 nucleotides occur at the V-J
junction which do not appear to be encoded by the genomic V
or J sequence: and presumably represents an N-region. The C
region sequences mavtch the corresponding genomic sequence (Lefranc et ail.,- 19~B6c,-Proc. Natl. Acid.- Sci. U.S.A.
20 83:9596-9600),. with the exception of nucleotide 559 (G to C;
Val to Ile) and nucleotide 908 (T to C; Met to Thr).
8.2. RESULTS
8.~2.1. NOVEL y3TCR PROTEIN COMPLEX
Preliminary studies of peripheral blood 7aTCR
lymphocytes r~aveale~d the presence of a CD3-associated complex that was different from the known human y6TCR forms.
In an attempt to delineate this form, we produced and 30 characterized a number of cell lines derived from normal human donors. Peripheral blood lymphocytes were stained with monoclonal antibody (mAb) WT31, which brightly stains resting apTCR lymphocytes. Cells that did not stain were isolated by c~_11 sorting and then expanded in vitro in IL-2 35 containing medium. Peripheral blood lymphocyte line 2 G

a_ X34?~42 (PBL-L2)~obtai.ned in this way, proved to be homogeneously CD3+CD4 CD8 , a cell surface phenotype characteristic of 7dTCR lymphocytes.
To visua7.ize ~dTCR complexes on PBL-L2 cells, immunoprecipit:ations with an anti-CD3 mAb were carried out from cell surface 1'~5I-labelled cells solubilized in CHAPS
or digitonin. In these detergents, the physical association between the CD3 complex and ydTCR subunits is preserved.
SDS-PAGE of anti-CDa immunoprecipitates from PBL-L2 cells 0 resolved 40 kL~ and X44 kD proteins (referred to as 40 kD) that were ider,~tifie~t as yTCR subunits by anti-Cyb serum, an antiserum directed against a yTCR constant region peptide (Fig. 22A: see: methods section).
Theae yTC:R proteins on PBL-L2 are noncovalently ~5 associated with a bZ'CR subunit, which is visible as a weakly iodinated protein in the anti-CD3 immunoprecipitate analyzed under nonreduc:ing conditions (Fig. 22A, lane 6, closed arrow). This weakly iodinated protein represents the dTCR
subunit on PBL-L2 cealls, since it i~ not recognized by -20 anti-Cab serum (Fig. 22A, lane 8). In addition, it displays the same SDS-mobility shift comparing analysis under nonreducing and reducing conditions as was noted for the 3TCR proteins on IDF~2 and PEER cells (see infra: see also PCT International Publication No. WO 88/00209, published 25 January 14, 1988). The bTCR protein could not be visualized after reduction (Fig. 22A, lane 3), because it migrated with . a mobility of 40 kD (see infra) and then was obscured by the similar sized yTCR protein (open arrow).
This -~bTC:R form is not only present on normal 30 peripheral blood T :Lymphocytes, but is also observed on thymus-derived Clone II cells (Fig. 22D) and on the T-leukemic cell line 1KOLT-13 (Fig. 22E)~ These three cell lines possess yTCR species that display differential glycosylation resuliting in a ~TCR protein doublet observed 35 on PBL-L2 (40 kD and 44 kD; Fig. 22A, lane 8) and Clone II

~ 341 042 cells (4b kD and 44 kD; Fig. 22D, lane 8) or a diffusely labelled.yTCR protein band observed on MOLT-13 cells (40 to 46 kD: Fig. 22E, lane 6). Two-dimensional gel analysis [nonequilibrium pH gradient electrophoresis (NEPHGE) followed by SL)S-PAG1E] of the MOLT-13 yTCR protein band resolved two parallel yTCR species (40 kD and 44 kD), of which the 44 )EcD yTCR species contained an additional high mannose (or hybrid) N-linked glycan compared to the 40 kD
yTCR species. Thus,, the yTCR subunits of this receptor 0 complex isolated from three different cell sources (peripheral b7Lood, thymus, and leukemia) revealed cell surface specieas of ~40 kD that are noncovalently associated with dTCR partner chains.
For comparison to the y6TCR form on PBL-L2, Clone ~5 II and MOLT-1:3 cells, we examined the previously known forms on the IDP2 and WM-:14 cell lines. The IDP2 cell line (see PCT International Publication No. WO 88/00209, published January 14, 1988: B:renner et al., 1986, Nature 322:145-149) contains a-larger, 55-60-kD-7TCR protean (referred to as 55 20 kD), which is recognized by anti-Cyb serum (Fig. 22B). When the anti-CD3 :immunolprecipitate is examined under nonreducing conditions, ii. is evident that the IDP2 7TCR protein is associated noncoval~ently with its aTCR partner chain (Fig.
22B, lane 4, :solid .arrow). Upon reduction, the 6TCR protein 25 displays a decrease in SDS-PAGE mobility to a relative molecular mas:a of 40 kD (compare Fig. 22B, lane 4, closed arrow, with F:ig. 221B, lane 2, open arrow) .
In contrast to the noncovalently associated ~bTCR
forms, the peripheral blood-derived T cell clone, WM-14, 30 bears a disul:Eide-linked TCR dimer of 70 kD (Fig. 22C, lane 7), that was :recognized by anti-Cy serum (Alarcon et al., 1987, Proc. Natl. Acad. Sci. U.S.A. 84:3861-3865). This dimer is also recognized by anti-TCRSl, a mAb directed against the 6'.CCR sulbunit (Fig. 22C, lane 5), and therefore 1341 p42 ~, represents a ~~bTCR heterodimer. Analysis under reducing conditions re~~eals i~hree ~TCR proteins of 36 kD, 40 kD and 43 kD (referre:d to <is 40 kD) .
Thus, the. CD3-associated complex on PBL-L2, Clone II and MOLT-13 cell:a constitutes a novel ~dTCR heterodimer compared to the pre~riously known forms, since its TCR 7 subunit is 40 kD (s:imilar in size to the disulfide-linked Cyl encoded yTCR protein on WM-14 cells), yet it is not disulfide-linl~;ed to its partner chain (similar to the 55 kD, C'r2 encoded yT'CR protein on IDP2 cells). To understand the molecular basis of ithis complex, more detailed structural analysis of ita yTCR and dTCR subunits was carried out as described infra, using the MOLT-13 cell line as an example.
8.2.2. CORE POLYPEPTIDE SIZE OF MOLT-13 yTCR SUBUNIT
To detenaine the size of the yTCR core polypeptide o1: MOLT'-13 cells (40 kD yTCR glycoprotein), and compare it with that of PEER cells (55 kD yTCR
glycoprotein)" both- cell lines were biosynthetically -labelled for :l5 mirnutes in the presence of 35S-methionine and 35S-cyste:ine, solubilized in Triton X-100 and then immunoprecipii:ated with anti-C~ml,~a monoclonal antibody that specifically recognizes the yTCR chain (Fig. 24A, see methods section). Immunoprecipitated material was subsequently <iigest~ed with endoglycosidase H (Endo H) to remove the immature N-linked glycans. The MOLT-13 yTCR
polypeptide backbone has a relative molecular mass of 35 kD
(Fig. 24A, lane 8), which is 5 kD smaller than the PEER yTCR
core polypept:ide (40 kD: Fig. 24A, lane 4) or the IDP2 yTCR
core polypept:ide (40 kD: See PCT International Publication No. WO 88/00209, published January 14, 1988). It now can be concluded ithat MOLT-13 cells express a yTCR core polypeptide that is distinct from the IDP2 and PEER ~TCR
core polypept:ides based on its being 5 kD smaller in size.
In addition, only 5-11 kD of size on the mature MOLT-13 yTCR
G

_ ~34~~42 _ cell surface c~lycoprotein are accounted for by post-translational processes (40-46 kD surface size minus 35 kD
core size), where 1!5-20 kD of relative molecular mass can be accounted for by post-translational processes on the PEER
and IDP2 ~TCR glycoproteins (55-60 kD surface size minus 40 kD core size).. Assuming that all post-translational processes are N-linlted glycans and that each glycan chain accounts for approximately 3 kD of relative molecular mass, We predict that 2 to 3 N-linked glycans are attached to the 0 MOLT-13 ~TCR protein, while 5 N-linked glycans are added to the polypeptides on PEER and IDP2 cells. Experiments using N-glycanase to remo~~e N-linked carbohydrates from cell surface ~TCR F>roteins showed that the majority of the post-translational processes that are added to the core 5 polypeptide are indeed N-linked glycans.
8.2.3. PF~IMARY SEQUENCE OF MOLT-13 7TCR
To understand the structure of the constant region gene se=gment encoding the MOLT-13 7TCR-subunit, the -20 sequence of a cDNA clone representing the MOLT-13 yTCR
transcript wa:a dete:rmined. A agtl0 library from MOLT-13 derived poly-A+ RNA was constructed and probed with a human ~TCR cDNA clone, pT-~-1 (Dialynas et al., 1986, Proc. Natl.
Acad. Sci. U.;i.A. 83:2619-23). Based on size and limited 25 restriction enzyme mapping, one clone, Ml3k, was selected and its nucleotide ;sequence determined (Fig. 25). Clone Ml3k represenia a full length, in-frame yTCR transcript, using a V~1.3 gene segment joined to a J~2.3 gene segment (Lefranc et a:L., 1986, Cell 45:237-246: Lefranc et al., 30 1986, Nature :319:420-422: nomenclature based on Strauss et al., 1987, Science 237:1217-1219; Quertermous et al., 1987, J. Immunol. 1:38:2687-2690). The constant region sequence was found to be nearly identical to a recently reported non-functional yTCR (Pellici et al., 1987, Science 35 287~1051-1055) and 'to the C~2 genomic sequence containing two CII exon copies b and c (Lefranc et al., 1986, Proc.
Natl. Acad. Sci. U.S~.A. 83:9596-9600) (see methods section for detailed account:). This represents the first in-frame transcript encoding a 7TCR protein expressed on the cell surface that utilizea a Cy2 gene segment with two CII exon copies.
The deduced amino acid sequence of this cDNA
clone predicts a pol.ypeptide backbone size of 34.8 kD which is in good agreement: with biochemical data described above.
0 Surprisir;gly, six potential N-linked carbohydrate attachment sites are encoded by this transcript. Since the biochemical data suggest that only 2 to 3 N-linked glycans are attached to the polypep~tide chain, it indicates that not all potential sitea are used.
i5 To reflects C7 gene segment usage, we have denoted the disulfide-~linkect ydTCR form expressed by PBL-C1 and WM-14 as "'Form 1", sinc:e such disulfide-linked yTCR chains utilize the Cy1 gene: segment (Krangel et al., 1987, Science 237:64-67). The-lax-ge {55-kD), nondisulfide-linked ~TCR
20 s~unit of the: ydTCR. form expressed on IDP2 and PEER cells is encoded by Cy2 gene segments containing three CII exon copies, namely copy a, copy b and copy c (Krangel et al., 1987, Science 237:64-67; Littman et al., 1987, Nature 326:85-88) andl therE:fore this ydTCR form is called herein 25 ~Form 2abc"'. In concordance, the form characterized on MOLT-13 cells is rei:erred to as "'Form 2bc".
8.2.4E. PREFERENTIAL Cy GENE SEGMENT USAGE
To determine the presence of these three ~r,bTCR
30 forms in freshly isolated peripheral blood we analyzed the mononuclear cells from ten healthy subjects, using biochemical analysis with mAb anti-TCRbl (described in Section 6, ~~ra; reactive with the bTCR constant region).
This antibody reacts with the great majority, if not all 35 '~~dTCR lymphocytes. Representative results from this panel j- , r are shown in Figure 26. In subject 1, anti-TCRdl immunoprecipitates (analyzed under nonreducing conditions) demonstrate the presence of both disulfide-linked y,sTCR
complexes as a 70 kD~ protein band (Form 1) and nondisulfide-linked ~,dTCR complexes as a broad 40 kD
protein band (Form 2bc) (Fig. 26, lane 2). This indicates that the Cyl a:nd C72 constant regions are voth used by the expressed y, 6TC:R of ithis individual . However, the amount of Form 2bc varied among individuals. Note the smaller fraction of Form 2bc in subject 2 compared to subject 1 by comparing the intensity of the 40 kD protein bands in both individuals (compare: lane 2 of subject 2 with lane 2 of subject 1). Even more strikingly, only disulfide-linked ~,dTCR complexes could be detected on the mononuclear cells of three of the ten individuals examined, even after long exposure of the autoradiographs (see subject 3). None of the analyzed individuals revealed the 55 kD, nondisulfide-linked y,dTCR c:omple:K (Form 2abc) in peripheral blood.
20 8~2.5. CHARACTERIZATION OF THE 6TCR SUBUNIT
In contrast to the striking structural differences in size and glycosylation of the 7TCR proteins, dTCR subunits from different cell sources proved to be markedly similar. Z'he relative molecular mass of the bTCR
25 glycoprotein o~n MOLT-13 cells was directly determined to be 40 kD using th.e anti.-3TCR mAb (Fig. 23, lane 4), confirming that it is similar i.n size to the 6TCR glycoprotein on IDP2 cells (Fig. 22;B, lame 2, open arrow).
To also compare bTCR polypeptide backbone sizes, cell surface 1~25I-labelled aTCR protein from MOLT-13 cells was digested with N--glycanase to remove asparagine-linked glycans (of the high mannose, hybrid, and complex-type;
Tarentino et a.l., 15185, Hiochem. 24:4665-4671; Hirani et al., 1987, Anal. Biochem. 162:485-492). The dTCR core p°l~eptides of MOLT-13 cells has a relative molecular mass G

of 35 kD (Fig. 24B, lane 4), which is similar to that of the bTCR backbone of IDP2 cells (35 Kd) (Band et al., 1987, Science 238:682-684).
In ~3ddition, digestion of cell surface 1251-labelled MOLT-:13 dTC~R protein with endoglycosidase H (Endo H, removing only high mannose and certain hybrid N-glycans;
Tarentino et a:l., 1974, J. Biol. Chem. 249:811-817; Trimble and Maley, 1984, Anal. Biochem. 141:514-522) caused a decrease in relative molecular mass of 2.5 kD, (Fig. 24B, lane 2) consistent with the presence of one carbohydrate moiety, leaving a relative mass of 2.5 kD of Endo H
resistant carbohydrates attached to the polypeptide. Since there are two ;potential N-glycan attachment sites present in the dTCR constant domain (Hats, S., et al., 1987, Science ~5 238:678-682: Loh et al., 1987, Nature 330:569-572), these data show that both are used, but that their N-glycans are processed differently, namely one as a high mannose N-glycan (Endo H-sensitive) and the other as a complex N-glycan (Endo H-resistant, but N-g~lycanase sensitive). In contrast to the 20 3ifferent-amounts of attached N-linked carbohydrate on- TCR -polypeptide chains, the dTCR subunits expressed on PEER, IDP2 and MOLT-13 cells all revealed~the same peptide core sizes and the presence of two N-linked glycans (Fig. 24B and data not shown).
8.3. DISCUSSION
. In this example, three protein forms of the human ~TCR glycoprotein are compared, namely the disulfide-linked 40 kD yTCR protein (Form 1), the nondisulfide-linked 55 kD
'~TCR protein (Form 2abc) and the nondisulfide-linked 40 kD
~TCR protein (Form 2bc). All three forms are shown to be associated with a dZ'CR subunit. Complementary DNA sequences representing the first two ~TCR forms have been reported previously (Krangel et al., 1987, Science 237:64-67; Littman et al., 1987, Nature: 326:85-88. The constant region of -~TCR

SD121 ~ 3 4 ~ 0 4 2 polypeptide Form 1 (on PBL-Cl) is encoded by the C~1 gene segment containing .a single CII exon, while yTCR polypeptide Form 2abc (on IDP2 .and PEER cells) utilizes the C72 gene segment containing CII exon copy a, copy b and copy c. The cDNA sequence corresponding to a ~TCR chain of Form 2bc was shown to contain a C~2 gene segment utilizing only two CII
exon copies, namely copy b and copy c. Similarly, it seems likely that the gene structure of the yTCR connector region of Clone II and PBL-L2 (nondisulfide-linked, 40 kD yTCR
0 protein) will also lbe like the MOLT-13 structure determined here, namely of Form 2bc. Since the dTCR constant region used is the same fo:r all these forms (Rata, S., et al., 1987, Science 238:6'78-682: Loh, E.Y., et al., 1987, Nature 330:569-572), a complete comparison of the structures of the 5 three ydTCR forms in man now can be made (Fig. 27).
Two Cy2 polymorphic genomic forms exist in man (Lefranc et a:l., Nature 319:420-422: Pellici et al., 1987, Science 237:1051-10':55). The two transcript forms (Form 2abc and Form 2bc) are probably the product of these different 20 allelic tlrpes. - To ~3ate, no allelic -form of- yTCR -polypeptides have been found in mice. We conclude that the dramatic difference in yTCR cell surface protein size between Form aabc (:55 kD) and Form 2bc (40 kD) is largely determined by the amount of attached N-link carbohydrates, 25 most likely reflecting the number of N-linked glycans.
Backbone size:a of I:DP2 yTCR (Form 2abc) and MOLT-13 7TCR
(Form 2bc) proteins have been measured to be 40 kD and 35 kD
respectively, on the basis of SDS-PAGE, which correlates well with their predicted molecular masses of 36.6 kD and 30 34.8 kD respectivel;y, calculated on the basis of cDNA
sequences. Ii;. is clear that this small difference in backbone size (5 kD in SDS-PAGE), accounted for mainly by one CII exon ~sncode~d peptide of 16 amino acids, contributed to, but could not solely explain the observed difference in 35 molecular masa between the 55 kD and 40 kD nondisulfide-linked yTCR surface forms. Form 2abc yTCR polypeptides possess 5 potential N-linked glycan attachment sites that are probably all used, in contrast to the MOLT-13 ~TCR
polypeptide which bears one additional potential attachment site, while carrying only 2 to 3 N-linked glycans. The reason for this limited use of potential attachment sites is unknown, but 'may result from the influence of the CII exon encoded peptides on the conformation of the yTCR protein.
The CII exon encoded peptides and their neighboring amino acids make up a connector region between the plasma membrane and the immunoglobulin-like constant domain. This region contains most of the N-linked glycan attachment sites (Fig.
28). We conclude that the CII exon copies appear to determine the protein form not only by determining Polypeptide backbone size, and by creating the ability to disulfide-link chains, but also by influencing the amount of attached carbohydrates.
aTCR complementary DNAs of IDP2 (Hata et al., 1987, Science 238:678-682), PEER (Loh et al., 1987, Nature 330:569-572)-~:and MOLT-13-cells'have been sequenced and were found to be identical, except for the diversity/N-region interspacing the variable and constant region gene segments.
The dTCR protein on WM-14 cells has a relative molecular mass of 43 kD, which is similar to the aTCR protein described previously (Borst et al., 1987, Nature 325:683-688: Lanier et al., 1987, J. Exp. Med. 165:1076-1094) but is 3 kD larger than the other dTCR chains. These 43 kD dTCR
proteins might indicate the presence of an additional N-linked glycosylation site in a different 6 variable domain.
Structural differences comparable to those described for ~TCR constant region segments have not been observed for ~~TCR and pTCR genes (Yoshikai et al., 1985, Nature 316:837-840; Toyonaga et al., 1985, Proc. Natl.~Acad.
Sci. U.S.A. 82:8624-8628: Royer et al., 1984, J. Exp. Med.
160:947-952: Kronenberg et al., 1985, Nature 313:647-653).
G

There is~possi:ble structural similarity in the number of human CII exon repeats with the length in murine Cy regions, of which the C~yl, Cy;t and Cy4 constant regions encode for 15, 10 and 33 .amino acid connector region respectively (Garman et al, 1986, Cell 45:733-742: Iwamoto et al., 1986, J. Exp. Med. 163:1203-1212). The connector regions in mouse, however, reflect a difference in the size of the relevant exon, not the multiple use of exons as is seen in Form 2abc 7TCR and Form 2bc yTCR in humans. Also, the murine yTCR on:Ly exist in disulfide linked forms in contrast to the two non~disulfide linked human forms.
Importantly, the human ~,bTCR forms do not appear to be used equally. In some individuals (selected for high percentages of y,dTCR lymphocytes), a single form (Form 1) ~5 predominates, suggesting that either positive selection occurs for this form or that there is selection against other y,6TCR forms.
9. TH3tEE T CELL RECEPTOR ~,d ISOTYPIC FORMS RECON-.~'.TITTJTED~ BY PA3RIriG OF DISTIiJCT- TRAaSFECTED 'yTCR -20 CHAINS WITH A SINGLE dTCR SUHUNIT
As described in the example herein, the role of the ~TCR polyp~eptide in the formation of the ys heterodimer was explored. We examined by transfection the 7dTCR
complexes formed by the association of yTCR chains 25 corresponding to the: three ~sTCR forms (Forms 1, 2abc, and 2bc) with a single resident bTCR chain. yTCR DNA encoding either Form 1 or Form 2abc of the ~TCR polypeptide was transfected into the: MOLT-13 cell line, which constitutively expresses a yb heterodimer comprised of fona 2bc yTCR
30 polypeptide noncoval,ently associated with bTCR polypeptide.
Transfected cells were capable of expressing, together with the ~aTCR characteristic of the MOLT-13 cell line, yd heterodimers comprised of either Form 2abc ~rTCR
noncovalently associated with dTCR or Form 1 yTCR covalently linked to bTCR. Furthermore, the glycosylation of the transfected ~TCR gene products was identical to the glycosylation of theae genes in their native cell lines.
Thus, the degree of glycosylation and the ability to form disulfide linkages acre properties determined by the yTCR
gene. yTCR constant: region CII exon usage determines not only the presence or. absence of disulfide linkage between TCR 7 and b polypeptides, but also the amount of carbohydrate a.ttache:d to the yTCR chain, which is largely responsible for the differences in size of the cell surface yTCR proteins.
9.1. MATERIALS AND METHODS
9.1.1. CELL LINES
MOLT-13, a TCR yb+ T leukemia cell line (Bata, S., et al., 1987, Science 238:678-682: Loh, E.Y., et al., 1987, Nature 330:569-572), and peripheral blood derived TCR
~b+ cell lines PBL C'1 (Brenner, M.B., et al., 1987, Nature 325:689-694) and ID1?2 (Hrenner, M.B., et al., 1986, Nature 322:145-149) were cultured as previously described.
9.1.2. ANTIBODIES
The: monoclonal antibodies (mAb) used were:
Anti-leu-4 (anti-hwaan CD3; IgGl) (Ledbetter, J.A. et al., 1981, J. Exp. Med. :153:310-323), anti-TCRbl (anti-human TCRb chain constant: region: IgGl) (See Section 6; Band, H., et al., 1987, Science ;138:682-684), anti-Ti-~A (anti-V~2:
IgG2a) (Jitsulcawa, S., et al., 1987, J. Exp. Med. 166:1192-1197), anti-C-,~ml (anti-human ~TCR contant region: see Section 8.1.7;1, P3 (IgGl secreted by the P3X63Ag8 myeloma) (Koehler, G., and Milstein, C., 1975, Nature 256:495-497), and 187.1 (rai: anti-mouse ~c light chain-specific) (Yelton, D.E., et al., 1981, Hybridoma 1:5-11).

9.1.3. ISOLATION AND SEQUENCING OF
MOLT-13 dTCR cDNA CLONES
A <:omple~mentary DNA (cDNA) library prepared from MOLT-13 poly J~+ RNA in the vector agtl0 (Huynh, et al., 1985, in DNA Clonin~~, ed. Glover, D.M. (IRL Press, Oxford), Volume 1, pp. 49-78) was screened by hybridization with 32P-labeled h~unan d'.CCR cDNA clone IDP2 O-240/38 (Hats et al., 1987, Science :238:678-682). Clones were selected for detailed analysis o:n the basis of size and limited restriction enzyme :mapping. Nucleotide sequence was determined in M13 vectors by dideoxy chain termination method (Sanger et al., 1977, Proc. Natl. Acad. Sci. U.S.A.
74:5463-5467) using the modified T7 polymerase (Sequenace, United States Hioch~emical Corp.) (Potter et al., 1984, Proc.
Natl. Acad. Sci. U.,S.A. 81:7161-7165).
9.1.4. CONSTRUCTION OF EXPRESSION PLASMIDS
AND TRANSFECTIONS
yT(:R cDNAs (PHL C1.15 and IDP2.llr) (Krangel, iii. S . , - et -al . ; 1987 ; ~ Science 237 : 64-67 ) were- cloned _ into pFneo mammalian expression vector (Saito, T., et al., 1987, Nature 325:12'.5-130: Ohashi, P., et al., 1985, Nature 316:606-609) downstream from Friend spleen focus forming virus (SFFV) long terminal repeat (LTR), as shown schematically in Figure 21B. The plasmid constructs were transfected into MOLT-13 cells by electroporation (Potter, H., et al., 1'984, Proc. Natl. Acad. Sci. 81:7161-7165).
Transfectants were selected and maintained in medium containing 2 :mg/ml of 6418 (480 ~g/mg solid by bioassay;
GIBCO), and cloned by limiting dilution.
9.1.5. _IODINATION AND IMMUNOPRECIPITATION
Cell surface labeling with 1251 using lactoperoxidase, solubilization in 3-[(3-cholamidopropyl) dimethylammonio] 1-propanesulfonate (CHAPS: Sigma Chemical Co., St.'Louis, MO), immunoprecipitation with various antibodies, nonequilibrium pH gradient gel electrophoresis (NEPHGE), and ;SDS polyacrylamide gel electrophoresis (SDS-PAGE) were performed as described (See Section 8.1.3;
Brenner, M.B., et al., 1986, Nature 322:145-149: Brenner, M.B., et al., 1987, Nature 325:689-694). Specific immunoprecipit~ations were carried out with 1 ~g anti-leu-4, 0.1 ~1 anti-TC1~61 ascites, or 1 ~1 P3 ascites, together with 150 ~1 of 187.:1 culture supernatant. For anti-T9-yA, 1 ~1 0 of ascites was used without 187.1.
9.l.ti. BIOSYNTHETIC LABELING
Exponentially growing cells were incubated for 30 minutes in met:hionine- and cysteine-free medium followed by ~5 a 15 minute pulse labeling at 37°C with 35S-methionine and 35S-cysteine, and immunoprecipitations were carried out as described in Section. 8.1.3, supra. Immunoprecipitates were either treated with endoglycosidase-H (endo-H) or were mock-incubated, separated by SDS-PAGE,_and.visualized by 20 fluorography (Bonner, W.M. and Laskey, R.A., 1974, Eur. J.
Biochem. 46:83-88).
9.2. RESULTS
We investigated the products resulting from association of structurally distinct ~TCR gene products with a single dTCR ;protein in order to demonstrate the role of . the yTCR gene, and in particular the TCR CII exons, in determining the structural differences between various yTCR
isotypes. For this purpose, MOLT-13, a T leukemia cell line that expresses the 9~0 kD nondisulfide-linked ~TCR
polypeptide (form 2bc), was used as a recipient for yTCR
chain cDNA clones corresponding to the other two forms of the receptor (Forms 1 and 2abc). Complete sequences of the cDNA clones representing these yTCR chains are described in G

SD127 ~ 3 4 ~ 0 4 2 Figure 25 (Fon~ 2bc), and in Krangel et al. (1987, Science 237:64-67) (Fo:cms 1 and 2abc) and they are schematically represented in Figure 29A.
9.2.J.. A SINGLE FUNCTIONAL dTCR CHAIN IS

bTCR gene rearrangement studies of the MOLT-13 cell line (Hat,a, S., et al., 1987, Science 238:678-682) suggested that only a single functional 6TCR gene product was expressed in this cell line. However, to demonstrate 0 directly that .a single functional transcript for dTCR is made in MOLT-13 cells, cDNA clones cross-hybridizing with a 6TCR cDNA probe (Hata, S., et al., 1987, Science 238:678-682) were isolated from a MOLT-13 cDNA library prepared in agtl0 and the ;sequence of selected cDNA clones was ~5 determined. This analysis revealed that MOLT-13 cells express transcripts corresponding to one functionally rearranged and one aberrantly rearranged 6TCR gene. The cDNA clone corresponding to the functionally rearranged 6TCR
gene has the same V (Vd1), J (J61), and C gene segments 20 described earlier for the IDP2 cell line (Hats, S., et al., 1987, Science 238:678-682). The MOLT-13 dTCR cDNA clone, however, possesses a. distinct nucleotide sequence between the V and J gene segrments arising from D segment utilization (MOLT-13 probably uses only D~2), imprecise joining, and N-25 region diversity at the V-D and D-J junctions (Rata, S., et al., 1988, Science 240:1541-1544). The MOLT-13 STCR cDNA
also predicts a cyst:eine residue in the membrane proximal connector region of the constant gene segment that would be available for disulf°ide linkage to ~TCR gene products that 30 utilize the C~1 gene. segment. Although the MOLT-13 cell line expresses a nondisulfide-linked y,3TCR receptor, the presence of a cystei.ne residue in the membrane proximal connector region of its bTCR chain leaves open the SD128 ~ 3 4 ~ 0 4 2 possibility that this dTCR subunit might be capable of participating in either a nondisulfide-linked or a disulfide-linked complex.
9 ,. 2 . 2 . ~~TCR GENE PRODUCT DETERMINES
THE FORM OF THE RECEPTOR
The MOLT-'13 cells transfected with ~TCR cDNA
constructs were abbreviated as M13.P8L Ch (for MOLT-13 cells transfected with PHL C1-deriWed yTCR cDNA) and M13.IDP2y (for MOLT-13 cells transfected with IDP2-derived ~TCR cDNA). T:he bulk transfectant cell lines and representative subcl.ones derived from these lines were analyzed by Northern blot analysis with ~TCR (VJC) or Vy2-specific cDNA probes. In addition to the resident 1.6 kb MOLT-13 ~TCR transcript, a second ~TCR transcript of about ~5 1.8 kb (expected size for a yTCR transcript initiating in the SFFV LTR of the expression plasmid) was observed in transfectant lines and their clones. The 1.8 kb transcript specifically hybridized with a V~2 probe that does not cross-hybridize with the V~1.3 present in the resident MOLT-13 yTCR transcript, and thus the 1.8 kb transcript represents the. transcript of the transfected -~TCR cDNAs (which utilize. a V~2 segment).
To biochemically characterize the ~TCR proteins) expressed on the surface of the transfectants, a representative: clone derived from each line was analyzed by immunoprecipit.ation of surface iodinated cells with P3 (control), anti-leu-4 (anti-CD3), anti-TCRdl (anti-dTCR), or anti-Ti-yA mAbS. Anti-Ti-yA (Jitsukawa, S., et al., 1987, J. Exp. Med. 1.66:11512-1197) appears to specifically recognize the y,dTCR. cells that utilize the Vy2 gene segment as the Variable portion of their ~TCR chains. Untransfected (Fig. 30A) as well as the transfected MOLT-13 cells (Fig.
3oB and 30C) express the expected parental -~TCR (40 kD: see open arrow) and dTCR subunits (see asterisk). Note that SD129 ~ 3 4 ~ 0 4 2 anti-V~2-speci;Eic mAb (anti-Ti-7A) fails to react with the resident MOLT-13 yTC:R chain (Fig. 30A, lanes 7 and 8).
Anti-CD3 immunoprecipitates of M13.PBL Cly transfectant cells revealed an additional CD3-associated species (68 kD) when examined under nonreducing conditions (Fig. 30B, lane 3, see solid arrow). On both PBL C1 cells (Fig. 30D, lanes 3 and 4) and the M13.PBL Cl~ transfectant cell line (Fig.
308, lanes 3 and 4), the 68 kD complex yielded 40 and 36 kD
species upon reduction (in the case of the M13.PBL C17 0 transfectant, these bands are clearly visualized in the anti-Vy2 immunoprecipitate, see Fig. 30B, lanes 7 and 8, solid arrows). These 40 and 36 kD species represent differentially glycosylated ~TCR palypeptides (Brenner, M.B., et al., 1987, Nature 325:689-694). In these 5 immunoprecipitates, the dTCR chain (40 kD reduced) comigrates with the 40 kD yTCR polypeptide and is therefore not visualized (however, see below).
Importantly these experiments show that the resident dTCR-cha.n of the MOLT-13-cell liner normally part -20 of a nondisulfide-linked complex, associates with the PBL C1 ~TCR protein to form a disulfide-linked ~,3TCR heterodimer in the transfectant cell line. In contrast, the IDP2-derived ~TCR protein (55 kD) in the M13.IDP2 transfectant cell line formed a nondisulfide-linked complex with the 25 resident MOLT-13 6TCR chain (Fig. 30C, lanes 3 and 4).
Immunoprecipitates carried out with anti-TCRbl mAb (specific for 6TCR peptide) confirmed that the endogenous (Figs. 30A, D and E, lanes 5 and 6) as well as the transfected ~TCR
chains (Fig.30B, and C, lanes 5 and 6) from all these cell 30 lines were associated directly with the dTCR chain. Anti-Ti-~A specifically i.mmunoprecipitated the 68 kD disulfide-linked y,6TCR ;heterodimer from the M13.PBL Cl transfectant cells (Fig. 30B, lanes 7 and 8), and the 55 kD yTCR chain, along with the: 40 kI) dTCR chain, from the M13.EDP2~
35 transfectant cells (Fig. 30C, lanes 7 and 8), confirming G

SD130 ~ ~ 4 ~ 0 4 Z
that the yTCR ~~hains that are part of these complexes correspond to 'the transfected PBL C1 and IDP2-derived yTCR
cDNAs, respectively.
To further characterize the various yTCR proteins biochemically, two-dimensional (2D) gel analyses (NEPHGE
followed by SD~S-PAGE) of surface 1251-labeled cells were carried out. .Superimposition of the 2D patterns of resident (MOLT-13) and transfected (PHL C1 ar IDP2) yTCR chains relative to the positions of the CD3 components allowed 0 comparison of 'the relevant ~TCR species. In the immunoprecipit~ates from MOLT-13 cells, the ~TCR chains resolved as two discrete parallel series of iodinated species (Fig. 31A, see open arrows). MOLT-13 cells transfected with the PBL C1 or the IDP2 TCR cDNAs revealed ~5 the resident M~DLT-13 ~TCR polypeptide series, but in addition, showed radiolabeled species that were identical in 2D gel patterns to the ~TCR polypeptides of PBL C1 (compare Fig. 31B and D; see asterisks in Fig. 31B) or IDP2 (compare ~'i g. 31C -and D; - sea closed ar-rows in Fig. 31C) - cell s,-20 respectively. Thus, the 2D gel biachemical analyses confirmed that the transfected 7TCR chains were expressed and processed similarly in MOLT-13 transfectants and in the parental cell lines, PBL Cl and IDP2.
25 9.2.a. POLYPEPTIDE BACKBONE SIZES OF THE
TRANSFECTED yTCR CHAIN PROTEINS
The peptide backbone sizes of the transfected ~TCR chains were determined by endoglycosidase-H treatment of the material immunoprecipitated from metabolically .
pulse-labeled cells. Immunoprecipitates carried out with 30 anti-CyMl (specific for yTCR chain) identified 35.5 and 34 kD species in untransfected MOLT-13 cells (Fig. 32, lane 4) that represent endogenous MOLT-13 ~TCR polypeptides. The smaller of these two polypeptides (see open arrows) corresponds to the expected polypeptide core size of the 134' 042 MOLT-13 ~TCR polypeptide, whereas the larger polypeptide appears to represent: a partially processed intermediate. In addition to these resident MOLT-13 ~TCR polypeptides, a polypeptide with a deglycosylated size of 41 kD was immunoprecipit,ated by anti-C7M1 from the M13.IDP2y transfectant, (Fig. 32, lane 8, see solid arrow). The size of this transfectant:-specific yTCR polypeptide agrees well with the deglycosylated IDP2 ~TCR polypeptide core size determined earlier i.n IDP2 cells (Brenner, M.B., et al., 0 1987, Nature 325:689-694). As expected, M13.PBL C17 transfectant cells revealed an additional ~TCR protein with a deglycosylat.ed size of 32 kD (Fig. 32, lane 12, see solid arrow) which compares well with the yTCR polypeptide backbone size reported earlier for the PBL C1 cell line 5 (Brenner, M.B., et al., 1987, Nature 325:689-694). This 32 kD species was. specifically immunoprecipitated by the V~2-specific mAb, anti-Ti-yA (Fig. 32, lane 13, see solid arrow), thereby allawing unambiguous assignment of resident and-transfecte~d yTCR species in this cell-line. Thus, the-20 determined backbone sizes of the transfected yTCR chains, derived from I:DP2 and PBL C1 cell lines, match the backbone sizes of these: polypeptides in their parent cell lines. By comparing the yTCR polypeptide core sizes with those of the cell surface proteins, we infer that the MOLT-13, IDP2, and 25 PHL C1 derivedl-~TCR chains carry 6, 14, and 8 kD N-linked carbohydrate, respectively.
9.3. DISCUSSION
Three biochemically distinct forms of the human 30 '~dTCR subunit structure occur. In the present work, we show that a single bTCR polypeptide can associate with yTCR
chains representing each of the three receptor forms to reconstitute t:he appropriate y,bTCR heterodimers. The resident ~TCR polypeptide of MOLT-13 (form 2bc) is 40 kD and 35 is noncovalent:ly associated with the bTCR subunit. When the 1 34' X42 yTCR cDN~r clonEas corresponding to the disulfide-linked receptor of PB:L C1 (Form 1), or the 55 kD non-disulfide-linked receptor of t;he IDP2 cell line (Form 2abc) were transfected into the MOLT-13 cell line, the ~dTCR forms corresponding 'to those found in the cDNA-donor cell lines were reconstituted. The present transfection studies provide direct evidence that disulfide linkage is dictated by yTCR constant segment usage, since the resident MOLT-13 6TCR chain was shown to participate in a disulfide-linked 0 receptor complex with the PHL C1-derived yTCR chain (Form 1), and a nond.isulfide-linked receptor complex with the IDP2-derived y'.~CR chain (Form 2abc).
We lhave shown that the remarkable difference in size between tlhe 55 kD (Form 2abc) and 40 kD (Form 2bc) ~5 non-disulfide-linked yTCR polypetides is primarily due to different amounts of N-linked carbohydrate attached to the yTCR polypeptide backbone (See Section 8.2.2, supra). Thus, either 15 kD ('Form 2abc on IDP2 or PEER) or only 5 kD (Fona 2bc on MOLT 13)_of N-linked carbohydrate is_attached to -20 these 7TCR pol;~peptides even though the same number (five each) of N-linked glycan acceptor sites are encoded by the constant region gene segments used in both of these forms.
Four of these :N-linked glycosylation sites are present in or around the CiI exon-encoded connector region. In the 25 example herein, we show that the amount of N-linked carbohydrate attached to the transfected 7TCR proteins is . identical to that seen in their parent cell lines, based on a comparison of peptide core size and mature cell surface size of the protein products of transfected ~TCR cDNA
30 clones. Thus, the conformation of the two C~2 encoded protein segments must differ sufficiently to result in drastic differences in glycosylation. The major difference between Cy segments of these two forms is that copy "a"' of the CII exon is present in the 55 kD yTCR chain of Form 2abc 35 and it is absent from the 40 kD ~TCR chain of Form 2bc.
G

Thus the~prese:nce or absence of this CII exon copy may be largely responsible for the glycosylation differences that account for th,e ~TCR polypeptide sizes.
The: variation in structure of human ~bTCR
isotypic forms. is unprecedented among T cell receptors as no such parallel is observed in a~TCR.
10. T CELL RECEPTOR yS COMPLEX, NOT
ASSOCIATED WITH CD3, IS IDENTIFIED
IN HUMAN ENDOMETRIAL GLANDULAR EPITHELIUM
0 In the early stages of placentation, infiltration of mononuclear cells is abundant at the proximity of spiral arteries and e:ndometrial glands in maternal uterine tissues.
These include an unusual population of T lineage cells of unknown function. Many extravillous trophoblasts express a ~5 novel type of class I MHC antigens which is different from that expressedL on most somatic cells. We have tested a panel of monoclonal antibodies to TCR ~d heterodimer (~STCR) in pregnant & non-pregnant uteri. Surprisingly, y6TCR
complex was not detected in leukocytes, but was localized in 20 the cytoplasm of the endometrial glandular epithelium from pregnant uteri.. These antibodies also reacted with the glandular epithelium from non-pregnant uteri, and the reactivity wa~~ stronger in the secretory phase than that in the proliferat:ive phase of the menstrual cycle. However, 25 ~6TCR was not associated with the CD3 complex, as shown by examining immunoprecipitates using three different monoclonal antibodies to CD3 (OKT3, anti-leu-4, UCHT-1).
The 76TCR-positive glandular epithelial cells did not react with monoclonail antibodies to a~TCR; the cells were also 30 CD4- and CDS-r~egati~e. Moreover, the glandular epithelial cells lose the: class I MHC antigens in early pregnancy.
These data suggest that these ~6TCR bearing endometrial glandular cells undergo, at least, phenotypic alterations under local regulation of gene expression.

? X41 04~ _ 11. EXAMPLE: CHARACTERIZATION OF A HUMAN b T CELL
RECE'~PTOR~GENE AND A Vb SPECIC MONOCLONAL ANTIBODY
We have isolated bTCR cDNA and a rearranged bTCR
gene from a human yb T cell clone, AK119. From these DNA
clones, a Kb p:cobe was obtained, and used to determine the diversity of bTCR gene rearrangements in a panel of 13 human y,6 T cell clones and 3 y,b human T cell tumor lines.
Altogether five different rearrangements were detected, which corresponded to rearrangements using 2 to 5 different Xb genes. One particular rearrangement was always seen in human y,b T cells that reacted with. mAb TCSbl (bTCAR-3). In addition, TCSbl immunoprecipitated the bTCR polypeptide from a human 7,b tmnor cell line, Molt 13. We provide evidence that monoclonal antibody TCSbi recognizes an epitope encoded in the AK119 V'b gene: or in a combination epitope of the rearranged AK119 gene Vb-Jb gene.
11.1. MATERIALS AND METHODS
11.1.1. ISOLATION AND SEQUENCING OF AK119 bTCR cDNA CLONES
A cDNA library was generated from the PBL T-cell clone, AK119, by the method of Gubler and Hoffmann (Gubler and Hoffman, 1.983, Gene 25:263). About 100,000 plaques of an amplified library were screened using a 32P-labelled nick-translated Cb probe, isolated from a bTCR clone called O-024 (Hats, 5,., et al., 1987, Science 238:678). The longest hybrif,izing cDNA clone (1.3 kb clone C119b3) was selected for sequence analysis by the dideoxy chain termination meahod.
11.1.2. CLONING A REARRANGED bTCR GENE
A 3.5 kb genomic DNA clone containing the rearranged Vb gene was obtained from AK119 cells as follows:
EcoRI digested, DNA was size fractionated on a preparative c,.

1341 p42 agarose gel, ligated. into agtl0, packaged and transfected into E. coli. Recombinant phage were screened with a 32P-labeled nick translated 550 by EcoRI fragment derived from the cDNA clone, c11963. A rearranged clone called r119d1 which contains a 0.8 kb HincII fragment (V region specific) and a 1 kb HincII-EcoRI fragment (V-J region) was isolated.
11.1.3. DNA PREPARATION
Fetal and newborn thymic tissues were collected 0 in accordance 'with accepted guidelines regarding patients' rights and approval. T cell clones. were obtained from peripheral blood, pleural exudate ar cerebrospinal fluid by limiting dilution and were cultured in vitro (Hafler et al., 1985, Ann Neurol. 18:451; Van de Griend et al., 1987, J.
5 Immunol. 138:1627). In all cases, DNA was prepared by digestion with proteinase K in 1% sodium dodecyl sulfate, followed by extraction with phenol~'chloroform and ethanol precipitation.
20 11.1.4. SOUTHERN BLOT ANALYSIS
Genomic DNA was digested with EcoRI, size fractionated on a 0.9% agarose gel and transferred to nitrocellulose. Hybridization was carried out with 32P-nick translated probes as previously described (Maniatis, 1982, 25 Molecular Cloning: A Laboratory Manual (Cold Spring Harbor Laboratory; Cold Spring Harbor, New York).
11.1.5. CYTOFLUOROMETRIC ANALYSIS
Normal peripheral blood monoclear cells (PBMC), 30 obtained from volunteers were isolated by fractionation on a Ficoll gradient. PBMC and PBL T cell clones were stained by indirect immunofluorescence using TCS61 mAb (referred to previously as bTCAR-3; See Section 7, supra) and fluorescein-conjugated goat antimouse IgG (Becton Dickinson) 35 and analyzed in a fluorescence activated flow cytometer.
~~e.~

~34~042 11.2. RESULTS
11.2.1. DIVERSITY OF dTCR GENE REARRANGEMENTS
Using a Cb probe, we isolated a 1.3 kb bTCR cDNA
clone, termed c11963,, from a agtl0 cDNA library of the T-cell clone AK1:19. The 5' end of c119a3 was sequenced and found to use previously identified Vs and Jd genes (Rata et al., 1987, Sci~ance 238:678; Loh et al., 1987, Nature 330:569). The sequence of the V-J junction indicated that 011963 has an in-frame V-J joint.
A 5!50 basepair (bp) EcoRI fragment encoding all the variable and joining region and part of the constant region (V-J-C ~~robe) was obtained from c119d3 and used in Southern blot analysis of EcoRI digested genomic DNA from X119. This probe detects a germline 3.2 kb Vb and a germline 1.0 kb Ca band. AK119 showed an extra rearranged 3.5 kb band that is identical to the common dTCR
rearrangement (described in Hata et al., 1987, Science 238:673;~oh-e1t al.; 1987,-Nature~330:569)w(rearrangement-II-in Fig. 33). 'this 3.5 kb band was cloned from a EcoRI
size-fractionated agtl0 genomic library using the V-J-C cDNA
probe. A partial ma;p of the cloned rearranged bTCR gene, called r11961, is shown in Figure 2'8. The localization of the variable and joining region was determined using J
oligonucleotide probes and variable region specific probes.
From r11961, a 1 kb 'V-J probe was isolated by digestion with HincII and Eco:RI enzymes (see Fig. 28)~
This V-J probe was used to determine the diversity of a'.~CR gene rearrangements in a panel of I3 human 7dTCR positive T-cell clones and 3 human ~-dTCR positive tumor cell lines. As shown in Figure 33, five common rearrangements, numbered I-V, are seen in the polyclonal newborn thymoc;yte sample (lane 11). These rearrangements are representative of rearrangements used by the human yb T
cell clones. ~Dnly rearrangement II hybridizes to the HincII

SD137 ~ ~ 4 1 0 4 2 - HincII~Vd specific probe. Although we do not know that all these rearrangements represent V-D-J rather than D-J
rearrangements, some of them must represent rearrangements of new variable regions to the previously characterized Jd gene segment because these cells express a functional bTCR
polypeptide chain on their cell surfaces. We have not ruled out the possibility that these new rearrangements represent rearrangements of a single new Va gene to other Js genes, yet to be identified. Our data is consistent with the fact 0 that there must be 2-5 variable region genes that can be used in dTCR gene rearrangements.
11.2.2. DETERMINING THE SPECIFICITY OF mAb TCSdl TCS,61 previously referred to as 6TCAR-3, was ~5 generated by fusing splenocytes from of nice immunized with the human tumor ydTCR cell line MOLT-13 with a mouse myeloma line. When used in fluorescent activated cell sorter analysis, TCSd:L reacted only with some but not all human ~6 T cells: The-resultsware given-in Table 2. 'here is a 20 perfect correlation with usage of the AK119 Vd gene (rearrangement II) with positive staining by TCSbl. This data provides strong evidence that the epitope recognized by bTCSl is encode=d in 'the AK119 Vb gene or in combinatorial epitope of the rearranged AR119 Vs-Jb gene.

G

.~ SD138 ~ 3 4 '~

CORREL~rTION BETWEEN STAINING BY TCSls mAb AND
-_____-_- A-SPECIFIC Va REARRANGEMENT

human y6TCR T cell clones 3 rearranqementl TCSS1 AK4 V/~2 -4 AK925 III/? nd 1015 I/? -1018 IV/? -Wi.l I/?

1019 II/IV nd AK119 II/? +

Wi.K II/III +

human 7aTCR T cell tumor lines Peer II/? +

Molt-13 ~II/V +

DND41 II/VI3 +

1 dTCR rearrangements detected with V-J probe, numbered I-V
2 as in Figure 33.
Only 1 rearrangement was identified in each case even 3 though no germline J was detected.
A new rearrangement ~s observed which is not seen in newborn or fetal thymocytes. This rearrangement has been 4 assigned rearrangement VI.
+ means positive staining, - means negative staining.
nd means not determined.

12. CLONING OF THE T CELL ANTIGEN RECEPTOR DELTA GENE
12.1., EXPERIMENTAL PROCEDURES
1.2.1.1. NORTHERN BLOT ANALYSIS OF
GROUP O HYBRIDIZING TRANSCRIPTS
5 Ng total RNA samples were electrophoresed through 1.5% a~garose gels containing 2.2 M formaldehyde and transferred to nitracellulose. Filters were probed with nick-translated 0-240 or chicken actin (Oncor) (Figure 34A), or with nick-translated 0-240 (Figure 34B), a 330 by EcoRI-SacI fragment of 0-240/38 (V probe: see Fig. 36) labelled by hexanucleotide: priming, or a 550 by HaeIII fragment of 0-240 (3' UT: see Figure 36) labeled by nick-translation. Filters were washed with lxSSC, 0.5% SDS at 23'C followed by O.IxSSC
at 50°C.
12.1.2. 80UTHERN BLOT ANALYSIS OF GROUP 0 HYBRIDIZING GENOMIC DNA
Genomic I)NA samples were digested with restriction enzymes, electrophoresed through 0.7% agarose, transferred to nitracellulose, and probed with nick-translated group 0 clones (Figures 34A and 35B). Filters were washed with lxSSC, 0.5% SDS at 23'C followed by 0.lxSSC, O.lxS~DS at 68'C (Figure 35A) or 0.2 x SSc, O.IxSDS
at 55'C (Figure 35B). In Figure 35A, note that PBMC and PBL
L1 are derivedl from the same individual. The diminished signal in PBMC: presumably results from deletion in most T
cells in the sample. The remaining signal (largely B cells and monocytes) serves as a germline control for PBL L1. On this basis the: 9.0 kb fragment is interpreted as a polymorphism rather than a rearrangement.
G

SD140 ~ 3 4 1 0 4 2 12.1.3. SEQUENCE ANALYSIS OF GROUP O cDNA CLONES
Nuc;leotide sequences of clones 0-240, 0-254, 0-240/38 and 0-x!40/47 were determined using the dideoxy chain termination mEahod Wia the strategy outlined in Figure 36.
12.2. RESULTS
12.,2.1. SELECTION OF dTCR cDNA CLONES
A ~' cell--specific cDNA probe was generated by synthesizing high specific activity, 32P-labeled first 0 strand cDNA from IDP2 poly-A+ RNA, and subjecting this material to tyro cycles of hybridization with human B cell line JY poly-~,+ RNA followed by hydroxylapatite chromatography (Dav:is, M.M., et al., 1984, Proc. Natl. Acad.
Sci. U.S.A. 8~~:2194). The twice-subtracted single-stranded ~5 material was used to probe 40,000 plaques of an IDP2 agtl0 cDNA library I;Krangel, M.D., et al., 1987, Science 237:64), and 391 (1%) hybrid:i.zing plaques were obtained. Subsequent analysis organized these clones into 14 cross-hybridizing groups; composed~of ~s many as 139, arid as few as 2 members.
20 Three groups were identified as TCR 7 (10 members), TCR ~
(20 members), and CD3 6/E (7 members), based upon hybridization with appropriate probes. Representative members of the: remaining il groups (A,H,C,D,E,G,I,K,M,O,R) were 32P-labelled and used to probe Northern blots. One 25 group (O, consisting of 6 members) detected transcripts expressed in '.CDP2 and yaTCR cell line PEER (Weiss, A., et al., 1986, Proc. Natl. Acad. Sci. U.S.A. 83:6998-7002:
Brenner, M., cat al., 1987, Nature 325:689-694: Littman, D.R., et al., 1987, Nature 326:85), but not expressed in JY
30 and the apTCR. cell line HPH-ALL. Based on this result, two group O clone:a (O-240 and 0-254) were selected for further study.

12.2.2. NORTHERN BLOT ANALYSIS
Northern blot analysis of a larger panel of RNA
samples using O-240 as a probe (Figure 34A) revealed the expression of cross-hybridizing transcripts in four TCR ~b cell lines (IDP2, PEER, Molt-13, and PBL L1 (Brenner, M., et al., 1987, Nature 325:689-694). Four distinct transcripts, of 2.2kb, l.7kb, l.3kb, and 0.8kb (arrows, Figure 34A) were detected. However, transcripts were undetectable in B cell line JY, myeloid cell line IiL60, a~TCR T cell line HPH-ALL
and surface TC'.R T cell line SKW3. Transcripts were barely detectable in RNA from fresh or phytohemagglutinin-activated peripheral blood mononuclear cells (PHA PBMC), of which only a small fraction express TCR ~b.
12.2.3. EVIDENCE FOR REARRANGEMENT OF THE LOCUS
DEFINED BY THE GROUP O CLONES
Analysis of genomic DNA digested with a variety of enzymes revealed no evidence for rearrangement of O-24o hybridizing ss:quences in xdTCR T cells. However although a 9.5 kb XbaI fragment (and a 9.0 kb polymorphic fragment; see Figure 35A) was detected in B cells, myeloid cells and 7dTCR
T cells, this fragment was deleted on both chromosomes in all other T cE:ll examined. This represents somatic deletion rather than polymorphism, since pairs of B and T cells lines derived from t:he same individual were analyzed (SB and HSB;
8392 and 8402). These results suggest that deletion of sequences detected by O-240 may accompany rearrangement at either the TCR a or TCR ~ locus.
In~~.tial sequence analysis of clones O-240 (1.5 kb) and O-254 (0.7 kb) revealed that they both extend from an endogenous EcoRI site at the 5' terminus through a poly-A
tail and an Ec:oRI sate in the linker at the 3' terminus.
These clones were derived from a cDNA library constructed without methy7Lation of EcoRI sites. In order to obtain information 5" to the natural EcoRI site, O-240 was used to 134 042 _ probe an~EcoRI methylated IDP2 agtl0 cDNA library. Two clones that spanned the EcoRI site, O-240/38 (1.3 kb) and O-240/47 (1.4 h:b), were selected for detailed study. In contrast to the: results obtained using probes derived from the 3' end of t:he group O end of 0-240/38 detected discrete rearrangements in both EcoRI and PvuII digests of genomic DNA from five out of five TCR ys cell lines (Figure 35B).
Of the three germ line fragments in each digest detected by this probe (arrows, )h'igure 35B), rearrangements of the 3.3 kb EcoRI and 2'_x.0 kb PvuII fragment appeared to be shared by the five ydTCR cell lines, whereas rearrangments of the 6.6 kb EcoRI and 2.0 kb PvuII fragments distinguished the different cell lines. As opposed to these discrete rearrangements,. a heterogeneous smear of rearrangements was detected in EcoRI digests of two samples of fetal thymus DNA.
12.2.4. SEQUENCING OF GROUP O CLONES
The comparative organization and sequencing strategies used to characterize clones O-240, O-254, 0-240/38 and O-240/47 are presented in Figure 36. Partial restriction ma~~s and the locations of probes V, VJC and 3'UT
(hatched bars) are presented. Poly-A tails are noted.
Figure 37 A anc! B shows the composite nucleotide and deduced amino acid sequences of the group 0 cDNA clones O-240/38 begins within codon 7 of the composite sequences, whereas O-240 and O-254 begin wii~h codon 150. Within the coding region, sequences agree= at all positions except for codon 161 (GTG
in O-254 and O-240/38, TTG in O-240). This discrepancy is presumed to re.>ult from a reverse transcriptase error in 0-240. The composite sequence contains a long open reading frame of 293 a~aino acids clearly composed of V-, J- and C-like elements ;similar to those of TCR and immunoglobulin genes. Strkingly, the putative C region sequence is 79%
identical at the nucleotide level, and 73% homologous at the G

amino acid level, to the sequence of a novel murine TCR
constant region gene (Cx) recently reported by Chien et al.
(1987, Nature :327:677) to reside within the TCR a locus.
The high degree to sequence homology indicates that the group O clones reported here represent the human homologue of murine Cx. Thus, the deletion of this sequence in TCR a~
T cells suggests that the human constant region, like its murine counterpart, maps 5' to Ca within the human TCR a locus.
The 5' ends of O-240/38 and O-240/47 define a partial putative leader (L) sequence and a variable (V) region sequence. The precise processing point between these segments defining the amino terminus of the mature protein is unknown. However, processing of the TCR a chain in HPB-MLT has been suggested to occur between A(-1) and Q(+1) since the amino terminus of TCR a is blocked (Sim, G.K., et al., 1984, Nature 312:771-775). By analogy, we have tentatively assigned. the processing point to this location in our sequence; sinc~~in the~region-from =4 to +8 the'two sequences are identical in 11/12 residues.
The putative V region displays 57% amino acid sequence identity with a human Va sequence (PGAS; Sim, G.K., et al., 1984, Nature 312:771-775), 26% identity with a human V~ sequence (Y'T35: Yanagi, Y., et al., 1984, Nature 308:145-149), and 21% identity with a human V7 sequence (V~2, LeFranc, M-P., et al., 1986, Cell 45:237-246).
Comparisons among Va subgroup sequences and among V~
subgroup sequences can be used to identify consensus residues that occur in 50% or more of Va or Vp subgroups.
In Figure 38, the deduced O-composite V region amino acid sequence is ca~mpared to Va and Vp subgroup consensus sequences. Ca~nsensus residues were assigned based upon their appearance in 50% or more of Va or V~ subgroups, using the data compiled in Toyonaga and Mak (1987, Ann. Rev.
I~°unol. 5:585) .
'_..._ 1 3 41 p 42 The V region sequence reported here matches the Va consensus in 75% of these residues (30/40). By contrast, it only matches the V~8 consensus in 49% of theses residues (17/35). For .comparison, the randomly selected Va sequences 1.1, 6.1, and 12.1 match the Va consensus in 70%, 73% and 73% of these p~csitions, respectively, whereas the V~
sequences 2.2, 5.4 and 8.1 match at 40%, 53% and 60%. Thus, this V region is clearly Va-like, since it is as close to the consensus as other Va sequences.
0 In Figure 39, the deduced O-composite J region amino acid sequence is compared to Ja, J~ and J7 consensus residues. Consensus residues were assigned based upon their appearance in 40% or more of Ja, Jp and J~ sequences. Amino acids 112-125 display significant homology to human TCR
~5 consensus J region sequences and with the J region associated with muri.ne Cx (Figure 37A). However, amino acids 94-111 are homologous to neither V nor J sequences, and homology with the marine clone is minimal in this region as well (Figure 3'7A) . i~hether and --how -much of- this area is 20 encoded by a separate D element or results from so called N-region diversity (Tonegawa, S., 1983, Nature 302:575) remains to be determined. Clearly as the amino acid sequence remains in frame across the V(D)J junction, the IDP2 group O cDNA clones correspond to transcripts from a 25 productively rearranged gene.
The, putative constant region sequence includes an immunoglobulin,-like region with two cysteine residues separated by 5.1 amino acids, a connector region carrying a cysteine residue which is typically believed to mediate 30 interchain disulfide bonds between TCR components, and an intramembraneous region. Two potential sites of N-linked glycosylation are situated immediately amino-terminal to the first cysteine: and carboxy-terminal to the second cysteine.
In Figure 40, the deduced O-composite C region amino acid 35 sequence is compared with Ca (Yoshikai, et al., 1985, Nature G

316:837 ), C~1 iToyonaga, B., et al., 1985, Proc. Natl Acad.
Sci. U.S.A. 82..8624), Cyl (Lefranc, M.P., et al., 1986, Proc. Natl. Acad. Sci. U.S.A. 83:9596) and Ca (Langer, B., et al., 1968, S~. Physiol. Chem. 349:945). In Figure 41, the distribution of charged and uncharged amino acids in the region flanking and :including the presumed transmembrane region of the c)-composite sequence is compared with those of Ca, C~ and Cy. Within the first 91 amino acids of the constant region amino acid sequence identity is highest with C'r and Ca (22% and 20%, respectively) and Iawer with Ca and C~9 (15% and ll~t, respectively). The connector region shares elements with each of the other TCR chains. However, the 40 amino acids including and flanking the presumed transmembrane region show a significantly higher number of identities with the homologous region of Ca (3D%) than with either C~ (8%) or C7 (13%). These relationships are underscored by comparison of the number and distribution of charged and uncharged residues throughout this region.
Similar to that-of-~a the O-group constant region appears-to -have at least "two positively charged residues which may be buried within 'the membrane. Such charged residues are thought to be :important in mediating interactions with CD3 components, which display acidic residues within their transmembrane :regions (Van den Elsen, P., et al., 1984, Nature 312:413: Gold, D.P., et al., 1986, Nature 321:431:
Krissansen, G.l;~., et al., 1986, EM80 J. 5:1799). Also, as in the case of Ca, a:n intracellular tail (if it exists at all) would be extremely short. Whereas C~ and Cy display putative intracellular tails which are highly charged, the IDP2 group O sequence contains a single basic residue followed by four hydrophobic amino acids. The corresponding Ca sequence is of equal length. Regardless of how the membrane proximal sequences are disposed relative to the G

1341 p42 lipid bilayer and to CD3 components, it appears likely that this portion of the constant region is involved in interactions highly analogous to those of Ca.
The 3' untranslated (3'UT) sequences indicate the use of alternative polyadenylation sites. Whereas the O-240 3' UT extends some 950 by to an ATTAAA polyadenylation signal, that of O-254 extends only 26a bp, with polyadenylatio~n following the sequence TATAAA. Both sequences differ from the consensus AATAAA by a single ~ nucleotide. P~otenti.al for additional heterogeneity exists, since the sequence TATAAA occurs twice more within the O-240 sequence (13 b~p 3' to the signal used in O-254 and 130 by 5' to the signal used in O-240). Variation in the site of polyadenylatic~n is at least partially responsible for the 5 transcript heterogeneity observed in Northern blots (Figure 34B). Whereas, the 2.2 and 1.3 kb transcripts are selectively detected by a V region probe, an O-240-specific 3' UT probe deaects only the 2.2 and 1.7 kb transcripts.
Thus in-iDP2, -PEER and PBL L1 the two most abundant species ~ (2.2 and 1.3 Icb) represent differentially polyadenylated trasncripts. The minor 1.7 and 0.8 kb species therefore represent transcripts lacking V regions and are presumably transcribed from partially rearranged genes. By contrast, TCR p mRNA heterogeneity primarily results from the latter 25 mechanism (Yoshikai, Y., et al., 1984, Nature 312:521).
12.3. DICLTSSION
The: group O cDNA clones appear to encode the IDP2 TCR a peptide. They detected transcripts that are expressed 30 specifically i.n ~dTCR T cells and are encoded by genes specifically rearranged in the same cells. Transcript levels correlated well with the level of expression of cell surface TCR d polypeptide, which is lower in PEER than-in IDP2, and lows:r still in Molt-13.

Furthermore, the sequence of the group O clones is composed of V, J and C elements which are homologous to those of other TCR and Ig genes. The cDNA clones derived from IDP2 mRNA remain in frame across the V-J junction, indicating thai: they would encode a functional polypeptide in these cells., The predicted molecular weight of the polypeptide is 31.3 kd, with two potential N-linkd glycosylation rites. As demonstrated below, these predictions ag~:ee well with the properties of the TCR d Peptide of IDP:? cells. Furthermore, we have demonstrated by _in vitro transcription oand translation analysis that clone O-240/38 encodes a polypeptide immunologically crossreactive with the IDP2 '.SCR d protein (see Section 6.1.4).
Hum~in TCR y and 6 peptides can exist in a disulfide-linked form or an unlinked form in different cell lines (Brenner et al., 1987, Nature 325; 689-694; Borst et al., 1987, Nature 32.5:683: Moingeon et al., 1987 Nature 325:723; Lanier et al., 1987, J. Exp. Med. 165:1076; Lefranc at -al: , 1986; Proc.--Natl. Acad-: Sci,. -U.S-.A. -83:9596; -Toyonaga and Mak, 1987, Ann. Rev. Immunol. 5:585). This structural heterogeneity is controlled at least in part by TCR y constant region usage, since the C~-1 gene encodes a cysteine in th~~ membrane proximal connector region which is absent in Cy-2 (Krangel, M.S., et al., 1987, Science 237:64;
Littman, D.R., et al., 1987, Nature 326:85; Lefranc, M.P., et al., 1986, :Proc. Natl. Acad. Sci. U.S.A. 83:9596). IDP2 uses the Cy-2 gene, lacks this cysteine, and displays a nondisulfide linked receptor on the cell surface (Brenner, M., et al., 1987, Nature 325:689-694; Krangel, M.S., et al-., 1987. Science 237:64). One might have predicted that the IDP2 TCR b peptide would lack the analogous cysteine as well. However our cDNA sequences predict that IDP2 TCR s carries a cysteine in the membrane proximal connector that would be available for disulfide linkage. Moreover, Southern blots (Figure 35A and data not shown) provide evidence for only a single TCR b constant region gene.
Thus, it appears that a single TCR d gene product could interact with ~'CR ~ peptides encoded by Cy-1 to form a disulfide-linked complex, or with TCR -y peptides encoded by Cy-2 to form a nonlinked complex.
In contrast to TCR a and TCR ~, only a limited number of functional TCR y V regions exist (Lefranc, M.-P., a tla., 1986, (:ell 45:237-246). Thus the TCR d V gene pool size will be important in determining the number of antigens that may be recognized by TCR ~d lymphocytes. The V region used by IDP2 is clearly related to TCR a V regions, but whether TCR a and TCR d draw from the same or distinct pools of V regions i:a not known. Recent nucleotide sequence analysis indic~ites that the IDP2, PBL C1 and Molt-I3 TCR S
chains all use the same V region, an observation consistent with genomic r~aarrangement data (Figure 35B). This result suggests a limited TCR s V repertoire.

13. DEPOSIT OF HYBRIDOMAS -The following hybridoma cell lines, producing the indicated monoclonal antibody, have been deposited with the American Type Culture Collection (ATCC), Rockville, Maryland, on t:he indicated dates, and have been assigned the listed accession numbers:
Date of Accession Hybridoma Monoclonal Antibody Deposit Number bTCAR-3 TCSdl (dTCAR-3) 10/29/87 HB 9578 ( anti-~Va ) 5A6.E9 anti-TCR~l (anti-Ca) 7/27/88 HB 9772 #3 anti-C~ml (anti-C,~) 7/27/88 HB 9773 The present invention is not to be limited in scope by the cell lines deposited since the deposited embodiments are intended as single illustrations of one aspect of the invention and any cell lines which are functionally equivalent are within the scope of this invention. Indeed, various modifications of the invention in addition to those shown and described herein will become apparent to those skilled in the art from the foregoing description and accompanying drawings. Such modifications are intended to fall within the scope of the appended claims.

G

Claims (116)

1. An isolated polypeptide which is a .delta. chain of a T cell antigen receptor, which .delta. chain is characterized by (a) being associated in a complex with the T3 antigen when found on the surface of a T
cell, (b) not being reactive with antibodies to the alpha, beta T cell antigen receptor, (c) not being reactive with antibodies to the .gamma. chain of the T cell antigen receptor, (d) having at least one intrachain, covalent, disulphide linkage, and (e) having a molecular weight of about 40,000 daltons as determined by denaturing polyacrylamide gel electrophoresis.

2. The isolated polypeptide of claim 1, wherein the .delta. chain is a .delta. chain of a human T cell antigen receptor.

3. An antibody capable of specifically forming a complex with the polypeptide of claim 1.

4. The antibody of claim 3, which is polyclonal.

5. The antibody of claim 3, which is monoclonal.

6. A method for detecting T cells, each of which has the polypeptide of claim 1, which comprises, contacting a sample containing T cells with a substance capable of specifically forming a complex with such polypeptide so as to form complexes between the substance and such polypeptide, detecting such complexes and thereby detecting such T cells.

7. The method of claim 6, wherein the polypeptide is present on the surfaces of the T cells.

8. The method of claim 6, wherein the polypeptide is present in the cytoplasm of the T
cells.

9. The method of claim 6, wherein the polypeptide is a specific .delta. T cell antigen receptor polypeptide.

10. The method of claim 9, wherein the polypeptide is present only in suppressor T cells.

11. A method for diagnosing an immune system abnormality in a subject which comprises:
(a) determining the number of T cells in a sample from the subject;
(b) contacting the sample with a substance capable of forming a complex with the polypeptide of claim 1 so as to form a complex between such substance and such polypeptide;
(c) determining the percentage of T cells in the sample which have such polypeptide; and (d) comparing the percentage so determined with the percentage of T cells which have such polypeptide in a sample from a normal subject: who does not have the immune system abnormality, a difference in the percentage of T cells so determined being indicative of the immune system abnormality.

12. The method of claim 11, wherein the immune system abnormality is a cancer.

13. The method of claim 12, wherein the cancer is a leukemia.

14. The method of claim 12, wherein the cancer is a lymphoma.

15. The method of claim 11, wherein the immune system abnormality is acquired immune deficiency syndrome.

16. The method of claim 11, wherein the immune system abnormality is congenital immunodeficiency.

17. The method of claim 11, wherein the immune system abnormality is an autoimmune disease.

18. The method of claim 11, wherein the subject is an animal.

19. The method of claim 11, wherein the subject is a human.

20. The method of claim 11, wherein the sample comprises blood or tissue.

21. A method for diagnosing an immune system abnormality in a subject which comprises:
(a) determining the number of T cells bearing the polypeptide of claim 1 in a sample from the subject;
(b) determining the amount of the polypeptide in the T cells bearing the polypeptide; and (c) comparing the amount so determined with the amount of the polypeptide in an equal number of T cells bearing the polypeptide in a sample from a normal subject who does not have the immune system abnormality, a difference in the amount so determined being indicative of the immune system abnormality.

22. A method of Claim 21, wherein the amount of a single such polypeptide is determined.

23. A method for diagnosing an immune system abnormality in a subject which comprises:
(a) determining in a sample from the subject the ratio of the number of T cells which have the polypeptide of claim 1 relative to the number of T cells which have a surface marker selected from the group consisting of T3, T4, T8, .gamma., .delta. T cell antigen receptor and .alpha., .beta. T cell antigen receptor; and (b) comparing the ratio of (a) to the ratio determined in a sample from a subject who does not have the immune system abnormality, where a difference in the ratios so determined is indicative of the immune system abnormality.

24. An isolated polypeptide which is a .gamma. chain of a T cell antigen receptor, which .gamma. chain is characterized by (a) being associated in a complex with the T3 antigen when found on the surface of a T
cell, (b) not being reactive with antibodies to the .alpha.
or .beta. chain of the T cell antigen receptor; (c) being reactive with an antibody to the .gamma. chain of the T cell antigen receptor; and (d) having a molecular weight of about 55,000 daltons as determined by denaturing polyacrylamide gel electrophoresis.

25. An isolated polypeptide which is a .gamma. chain of a T cell antigen receptor, which .gamma. chain is characterized by (a) being associated in a complex with the T3 antigen when found on the surface of a T
cell, (b) not being reactive with antibodies to the .alpha.
or .beta. chain of the T cell antigen receptor; (c) being reactive with an antibody to the .gamma. chain of the T cell antigen receptor; and (d) having a molecular weight of about 40,000 daltons as determined by denaturing polyacrylamide gel electrophoresis.

26. The isolated polypeptide of claim 24, wherein the chain is a .gamma. chain of a human T cell antigen receptor.

27. The isolated complex which comprises two polypeptides of claim 24 associated with each other.

28. The isolated complex of claim 27, wherein the two polypeptides are associated with each other through at least one interchain, covalent, disulphide linkage.

29. The isolated complex of claim 27, wherein the two polypeptides are noncovalently associated with each other.

30. The isolated complex of claim 27, wherein the two polypeptides have the same constant domain.

31. The isolated complex of claim 27, wherein the two polypeptides have different constant domains.

32. A substance capable of specifically forming a complex with at least one polypeptide of claim 24.

33. The substance of claim 33 capable of specifically forming a complex with one .gamma. T cell antigen receptor polypeptide.

34. The substance of claim 33 capable of specifically forming a complex with more than one .gamma. T
cell antigen receptor polypeptide.

35. The substance of claim 33, which comprises an antibody.

36. The antibody of claim 35, which is polyclonal.

37. The antibody of claim 35, which is monoclonal.

38. The method for detecting T cells, each of which has a polypeptide of claim 24, which comprises contacting a sample containing T cells with a substance capable of specifically forming a complex with the polypeptide so as to form cellular complexes between the substance and the polypeptide, detecting such cellular complexes and thereby detecting such T
cells.

39. The method of claim 38, wherein the polypeptide is present on the surface of the T cells.

40. The method of claim 38, wherein the polypeptide is present in the cytoplasm of the T
cells.

41. The method of claim 38, wherein the substance is capable of forming complexes with a specific polypeptide.

42. The method of claim 41, wherein the specific polypeptide is present only in suppressor T cells.

43. The method of claim 38, wherein the polypeptide is associated with another .gamma. T cell antigen receptor polypeptide.

44. The method of claim 38, wherein the polypeptide is present only in non-major histocompatibility complex restricted cytotoxic T
lymphocytes.

45. The method of claim 44, wherein the non-major histocompatibility complex restricted cytotoxic T lymphocytes are T killer cells or natural killer-like cells.

46. A method for diagnosing an immune system abnormality in subject which comprises:
(a) determining the number of T cells in a sample from the subject;
(b) contacting the sample with a substance capable of forming a complex with at least one polypeptide of claim 24 so as to form a complex between the substance and the polypeptide;
(c) determining the percentage of T cells in the sample which have the polypeptide; and (d) comparing the percentage so determined with the percentage of T cells which have the polypeptide in a sample from a normal subject who does not have the immune system abnormality, a difference in the percentage of T cells so determined being indicative of the immune system abnormality.

47. The method of claim 46, wherein the immune system abnormality is a cancer.

48. The method of claim 47, wherein the cancer is a leukemia.

49. The method of claim 47, wherein the cancer is a lymphoma.

50. The method of claim 46, wherein the immune system abnormality is acquired immune deficiency syndrome.

51. The method of claim 46, wherein the immune system abnormality is congenital immunodeficiency.

52. The method of claim 46, wherein the immune system abnormality is an autoimmune disease.

53. The method of claim 46, wherein the subject is an animal.

54. The method of claim 46, wherein the subject is a human.

55. The method of claim 46, wherein the sample comprises blood or tissue.

56. A method for diagnosing an immune system abnormality in a subject which comprises:
(a) determining the number of T cells bearing the polypeptide of claim 24 in a sample from the subject;
(b) determining the amount of the polypeptide in the T cells bearing the polypeptide; and (c) comparing the amount so determined with the amount of the polypeptide in an equal number of T cells bearing the polypeptide in a sample from a normal subject who does not have the immune system abnormality, a difference in the amount so determined being indicative of the immune system abnormality.

57. The method of claim 56, wherein the amount of a single such polypeptide is determined.

58. A method for diagnosing an immune system abnormality in a subject which comprises:
(a) determining in a sample from the subject the ratio of the number of T cells which have the polypeptide of claim 24 relative to the marker of T cells which have a surface marker selected from the group consisting of T3, T4, T8, .gamma., .sigma. T cell antigen receptor and .alpha., .beta. T cell antigen receptor; and (b) comparing the ratio of (a) to the ratio determined in a sample from a subject who does not have the immune system abnormality, where a difference in the ratios so determined is indicative of the immune system abnormality.

59. An isolated receptor complex which comprises, a .delta. chain of a T cell antigen receptor, which .delta. chain is characterized by:
(a) being associated in a complex with the T3 antigen when found on the surface of a T
cell;
(b) not being reactive with antibodies to the alpha, beta T cell antigen receptor; and (c) is not reactive with antibodies to the .gamma.
chain of the T cell antigen receptor; and a .gamma. chain polypeptide of the T cell antigen receptor, which .gamma. chain is characterized by:
(a) being associated in a complex with the T3 antigen when found on the surface of a T
cell;
(b) not being reactive with antibodies to the alpha, beta chain of the T cell antigen receptor; and (c) being reactive with antibodies to the .gamma.
chain of the T cell antigen receptor.

60. The receptor complex of claim 59 in which the .delta. T cell antigen receptor polypeptide has a molecular weight of about 40,000 daltons and the .gamma. T
cell antigen receptor polypeptide has a molecular weight of about 55,000 daltons.

61. The receptor complex of claim 59, wherein the .delta. T cell antigen receptor polypeptide is a human .delta.
T cell antigen receptor polypeptide and the .gamma. T cell antigen receptor polypeptide is a human .gamma. T cell antigen receptor polypeptide.

62. The receptor complex of claim 59, wherein the .delta. T cell antigen receptor polypeptide and the .gamma. T
cell antigen receptor polypeptide are associated with each other through at least one interchain, covalent, disulphide linkage.

63. The receptor complex of claim 59, wherein the .delta. T cell antigen receptor polypeptide and the .gamma. T
cell antigen receptor polypeptide are noncovalently associated with each other.

64. A substance capable of specifically forming a complex with at least one receptor complex of claim 59.

65. The substance of claim 64 capable of specifically forming a complex with one such receptor complex.

66. The substance of claim 64 capable of specifically forming a complex with more than one such receptor complex.

67. The substance of claim 64, which comprises an antibody.

68. The antibody of claim 67, which is polyclonal.

69. The antibody of claim 67, which is monoclonal.

70. A method of detecting T cells, each of which has the receptor complex of claim 59, which comprises contacting a sample containing T cells with a substance capable of forming a complex with the receptor complex, detecting such formed complex and thereby detecting such T cells.

71. The method of claim 70, wherein the receptor complex is present on the surfaces of the T cells.

72. The method of claim 70, wherein the receptor complex is present in the cytoplasm of the T cells.

73. The method of claim 70, wherein the substance is capable of forming a complex with a specific receptor complex.

74. The method of claim 73, wherein the specific receptor complex is present only in suppressor T
cells.

75. A method for diagnosing an immune system abnormality in a subject which comprises:
(a) determining the number of T cells in a sample from the subject;
(b) contacting the sample with a substance capable of forming a specific complex with at least one receptor complex of claim 59;
(c) determining the percentage of T cells in the sample which have the receptor complex; and (d) comparing the percentage so determined with the percentage of T cells which have the receptor complex in a sample from a normal subject who does not have the immune system abnormality, a difference in the percentage of T cells so determined being indicative of the immune system abnormality.

76. The method of claim 75, wherein the immune system abnormality is a cancer.

77. The method of claim 76, wherein the cancer is a leukemia.

78. The method of claim 76, wherein the cancer is a lymphoma.

79. The method of claim 75, wherein the immune system abnormality is acquired immune deficiency syndrome.

80. The method of claim 75, wherein the immune system abnormality is congenital immunodeficiency.

81. The method of claim 75, wherein the immune system abnormality is autoimmune disease.

82. The method of claim 75, wherein the subject is an animal.

83. The method of claim 75, wherein the subject is a human.

84. The method of claim 75, wherein the sample comprises blood or tissue.

85. A method for diagnosing an immune system abnormality in a subject which comprises:
(a) determining the number of T cells bearing a receptor complex of claim 59 in a sample from the subject;
(b) determining the amount of the receptor complex in the T cells bearing the receptor complex; and (c) comparing the amount so determined with the amount of the receptor complex in an equal number of T cells bearing the receptor complex in a sample from a normal subject who does not have the immune system abnormality, a difference in the amount so determined being indicative of the immune system abnormality.

86. The method of claim 85, wherein the amount of a single such receptor complex is determined.

87. A method for diagnosing an immune system abnormality in a subject which comprises:
(a) determining in a sample from the subject the ratio of the number of T cells which have the receptor complex of claim 59 relative to the number of T cells which have a surface marker selected from the group consisting of T3, T4, T8, .gamma., .delta. T cell antigen receptor and .alpha., .beta. T cell antigen receptor; and (b) comparing the ratio of (a) to the ratio determined in a sample from a subject who does not have the immune system abnormality, where a difference in the ratios so determined is indicative of the immune system abnormality.

88. A method of identifying a monoclonal antibody directed against a .gamma., .delta. T cell antigen receptor comprising:
(a) contacting a viable cell expressing a .gamma., .delta. T
cell antigen receptor and a CD3 antigen on its cell surface with an antibody, for a period of time sufficient to effect comodulation of the CD3 antigen; and (b) detecting the co-modulation of the CD3 antigen.

89. The method according to claim 88 in which the detection of co-modulation is carried, out by:
(a) contacting the cell with a labeled antibody directed against the CD3 antigen, for a period of time sufficient to allow binding of the labeled antibody to the CD3 antigen;
and (b) measuring the amount of bound labeled antibody.

90. The monoclonal antibody of claim 5, which is characterized by the ability to co-modulate a CD3 antigen.

91. The monoclonal antibody of claim 37, which is characterized by the ability to co-modulate a CD3 antigen.

92. The monoclonal antibody of claim 69, which is characterized by the ability to co-modulate a CD3 antigen.

93. A composition of substantially purified cells which express the polypeptide of claim 1.

94. A method for diagnosing an immune system abnormality in a subject which comprises:
(a) determining the number of T cells in a sample from the subject;
(b) contacting the sample with a labeled nucleic acid probe capable of forming a complex with RNA encoding the polypeptide of claim 1;
(c) detecting the RNA;
(d) determining the percentage of T cells in the sample which have the RNA; and (e) comparing the percentage so determined with the percentage of T cells which have the RNA
in a sample from a normal subject who does not have the immune system abnormality, a difference in the percentage of T cells so determined being indicative of the immune system abnormality.

95. A composition of substantially purified cells which express the polypeptide or complex of any one of claims 24-31.

96. A method for diagnosing an immune system abnormality in a subject which comprises:
(a) determining the number of T cells in a sample from the subject;
(b) contacting the sample with a labeled nucleic acid capable of forming complex with RNA
encoding the polypeptide of claim 24;
(c) detecting the RNA;
(d) determining the percentage of T cells in the sample which have the RNA; and (e) comparing the percentage so determined with the percentage of T cells which have the RNA
in a sample from a normal subject who does not have the immune system abnormality, a difference in the percentage of T cells so determined being indicative of the immune system abnormality.

97. A composition of substantially purified cells which express the receptor complex of claim 59.

98. The isolated polypeptide of claim 24, wherein the chain comprises a nonglycosylated polypeptide backbone having a molecular weight of about 31,000 daltons as determined by denaturing polyacrylamide gel electrophoresis.

99. The isolated polypeptide of claim 24, which is glycosylated.

100. The isolated polypeptide of claim 25 in which the .gamma. chain is glycosylated.

101. The isolated polypeptide of claim 26 in which the .gamma. chain is glycosylated.

102. An isolated peptide fragment of a .delta. chain of a T cell antigen receptor, which .delta. chain is characterized by:
(a) being associated in a complex with the T3 antigen when found on the surface of a T
cell;
(b) not being reactive with antibodies to the .alpha.,.beta. T cell antigen receptor; and (c) not being reactive with antibodies to the .gamma.
chain of the T cell antigen receptor;
such isolated peptide fragment being characterized by:
(a) being antigenic;
(b) not being reactive with antibodies to the .alpha.,.beta. T cell antigen receptor; and (c) not being reactive with antibodies to the .gamma.
chain of the T cell antigen receptor.

103. The isolated fragment of claim 102, which has at least one intrachain, covalent, disulfide linkage.

104. The isolated fragment of claim 102, wherein the b chain of a T cell antigen receptor has a molecular weight of about 40,000 daltons as determined by denaturing polyacrylamide gel electrophoresis.

105. The isolated fragment of claim 102, wherein the b chain of a T cell antigen receptor is a .delta. chain of a human T cell antigen receptor.

106. A.n isolated peptide fragment of a .gamma. chain of a T cell antigen receptor, which .gamma. chain is characterized by (a) being associated in a complex with the T3 antigen when found on the surface of a T
cell;
(b) not being reactive with antibodies to the alpha, beta T cell antigen receptor; and (c) being reactive with antibodies to the .gamma.
chain of the T cell antigen receptor;
such isolated peptide fragment being characterized by:
(a) being antigenic;
(b) not being reactive with antibodies to the .alpha.,.beta. T cell antigen receptor; and (c) being reactive with antibodies to the .gamma.
chain of the T cell antigen receptor.

107. The isolated fragment of claim 106, wherein the .gamma. chain of a T cell antigen receptor has a molecular weight of about 55,000 daltons as determined by denaturing polyacrylamide gel electrophoresis.

108. The isolated fragment of claim 106, wherein the .gamma. chain of a T cell antigen receptor is a .gamma. chain of a human T cell. antigen receptor.

109. The isolated fragment of claim 106, wherein the .gamma. chain of a T cell antigen receptor has a molecular weight of about 40,000 daltons, as determined by denaturing polyacrylamide gel electrophoresis.

110. The isolated fragment of claim 109, wherein the .gamma. chain of the T cell antigen receptor comprises a nonglycosylated polypeptide backbone having a molecular weight of about 31,000 daltons as determined by denaturing polyacrylamide gel electrophoresis.

111. The isolated fragment of claim 109, wherein the .gamma. chain of the T cell antigen receptor comprises a nonglycosylated polypeptide backbone having a molecular weight of about 35,000 daltons as determined by denaturing polyacrylamide gel electrophoresis.

112. The isolated fragment of claim 106, wherein the .gamma. chain of the T cell antigen receptor comprises a nonglycosylated polypeptide backbone having a molecular weight of about 40,000 daltons as determined by denaturing polyacrylamide gel electrophoresis.

113. The isolated fragment of claim 106, which is glycosylated.

114. The isolated fragment of claim 107, which is glycosylated.

115. A hybridoma which produces the monoclonal antibody of claim 5.

116. A hybridoma which produces the monoclonal antibody of claim 37.