CA1340702C - Derivatives of soluble t-4 - Google Patents

Abstract

This invention provides a therapeutic agent capable of specifically forming a complex with human immunodeficiency virus envelope glycoprotein which comprises a polypeptide. In one embodiment of the invention, the amino acid sequence of the polypeptide comprises the amino acid sequence shown in Figure 6 from about +3 to about +185 fused to the amino acid sequence from about +351 to about +369. In another embodiment of the invention, the amino acid sequence of the polypeptide comprises the amino acid sequence shown in Figure 6 from about +3 to about +106 fused to the amino acid sequence from about +351 to about +369. In yet a further embodiment of the invention, the amino acid sequence of the polypeptide comprises the amino acid sequence shown in Figure 6 from about +3 to about +185.
This invention also provides a method for treating a subject infected with a human immunodeficiency virus.
The method comprises administering to the subject an effective amount of a pharmaceutical composition comprising an effective amount of a therapeutic agent of the invention and a pharmaceutically acceptable carrier.

Description

DERIVATI9E8 OF sOLOHLE T-4 BACKGROUND OF THE INVENTION
Within this application several publications are refer-enced by Arabic numerals within parentheses. Full citations f~~r these references may be found at the end of the specification immediately preceding the claims.
The disclosures of these publications are of interest herein to more fully describe the state of the art to which this :Lnvention pertains.
' The different functional classes of T lymphocytes rec-ognize antigen on the surface of distinct populations of target cells. Helper T cells interact largely with macrophages and B cells: cytotoxic T cells interact with a broader range of antigen-bearing. target cells.
j These cellctlar recognition events are likely to be 25 mediated by the specific association of surface mole cules on bath effector and target cells. The surface of T cells is characterized by a number of polymor phic, as well as nonpo3ymorphic, proteins which are restricted for the most part to T lymphocytes. A1 though most: of these molecules are common to all T
cells, two classes of surface proteins consistently differ on the different functional classes of T cells, and these proteins have been implicated in T cell-tar get cell initeractions.

One class o:E surface molecules distinguishes the major functional subsets of T lymphocytes: the surface glycoproteins T4 and T8. Early in thymic development, the glycoproteins T4 and T8 are coexpressed on the surface of thymocytes (1). In the peripheral immune system, the T4 and T8 molecules are expressed on mutu-ally exclusive subsets of T cells and are only rarely expressed on the same cell (2, 3). The T4 molecule is expressed on T cells that interact with targets bearing class II ma:ior hi.stocompatibility complex (l~iC) mole cules, whereas T8-bearing T cells interact with targets expressing class I I~iC proteins (4, 5, 6, 7, 8, g), The T4 population of T lymphocytes contains helper cells, whereas the T8 population contains the majority of cytotoxic and suppressor cells (6, 10). However, rare T4+ T cells can function as cytotoxic or supres sor cells (6,, 10), suggesting that the expression of T4 or T8 is more stringently associated with I~iC class recognition than with effector function. The signifi cance of thence molecules in T cell-target cell interac tions can beg demonstrated by studies with monoclonal antibodies. Antibodies directed against specific epitopes of 'the T4 molecule (or the murine equivalent L3T4) inhibit antigen-induced T cell proliferation, lymphokine release and helper cell function (7, 8, il, 12, 13). Similarly, monoclonal antibodies directed against T8 (or the murine equivalent Lyt2) inhibit cytotoxic T cell-mediated killing (14, 15). These observations, along with the fact that T4 and T8 do not reveal significant polymorphism, has led to the hypothesis that T4 and T8 recognize nonpolymorphic regions of class II and class I molecules, respective-ly.

i~~~70~

A second class of proteins thought to differ on differ-ent effecto:r T cells are the receptors that recognize antigen in .association with polymorphic regions of MHC
molecules (:L6, 17, 18). The interactions of helper T
lymphocytes are largely restricted to antigen-bearing target cells: expressing class II MHC proteins, whereas cytotoxic and suppressor T cells are restricted to tar-gets bearingf class I MHC molecules (4, 5, 6, 7, 8, 9) , These specific interactions may be mediated by the T
cell receptor (or receptors) that recognize antigen in the context of specific MHC molecules (17, 18). Thus, the T lymphocyte may have two independent receptors capable of recognizing both constant and polymorphic determinants of MHC proteins, and these receptors may be responsible for specific targeting of functionally distinct populations of T cells.
The human acquired immune deficiency syndrome (AIDS) is characterized by a depletion of T4+ lymphocytes. As a consequence, T cell-mediated immunity is impaired in AIDS patients, resulting in the occurrence of severe opportunistic; infections and unusual neoplasms. AIDS
results from the infection of T lymphocytes with a collection of closely related retroviruses (LAV, HTLV
III, or ARV), now termed human immunodeficiency virus (HIV). The range of infectivity of these agents is restricted to cells expressing the T4 glycoprotein on their surface.
Therefore, th;e T4 glycoprotein may serve not only as a receptor for molecules on the surface of target cells, but also as a receptor for the AIDS virus. Monoclonal antibodies directed against T4 block AIDS virus infec-tion of T4+ cells in vitro. Furthermore, recent stud-ies have demonstrated that when T4+ T lymphocytes are '~ ~3~~~' exposed to AIDS virus, the 110 kd envelope glyco-protein of the virus is associated with the T4 molecule on the host cell» The lymphotropic character of the virus could therefore be explained by the restricted expression of its receptor, T4, in subpopulations of T
lymphocytes.
The depletion of T4+ T lymphocytes in AIDS results in the impairment of the cellular immune response. In addition, AIDS is frequently accompanied by central nervous system (CNS) dysfunction, most often the conse quence of a subacute encephalitis. AIDS virus RNA and DNA has been identified in affected brains, and virus has been isolated from both brain and cerebrospinal fluid from patients with neurological disorders. These observations suggest that the AIDS virus infects brain cells and is directly responsible for the CNS lesions observed in AIDS patients. Thus, the AIDS virus may be neurotropic as well as lymphotropic. It is therefore important to detenaine whether T4 is also expressed :in the CNS or whether additional brain-specific surface molecules may serve as a receptor for the AIDS virus.
The elucidation of the specific interactions of T4 and T8 would be facilitated by the isolation of the T4 and T8 genes, the determination of their structure, and the ability to introduce them into different cellular en vironments. The isolation and sequence of a cDNA en coding the 'r8 molecule has recently been reported (19, 20, 21). The deduced protein sequence indicates that T8 is a membrane-bound glycoprotein with an N-terminal domain that bears homology to the variable region of immunoglobul.in light chains.

_~ -5- ~~~o~
SUMMARY OF TIME INVENTION
This invention pravides a therapeutic agent capable of specifically fonaing a complex with human immunodefi-ciency virus. envelope glycoprotein comprising a poly-peptide. In one embodiment of the invention, the amino acid aequence of the polypeptide comprises the amino acid sequence shown in Figure 6 from about +3 to about +185 fused to the amino acid sequence from about +351 to about +36'9. In another embodiment of the in-vention, the amino acid sequence of the polypeptide comprises the amino acid sequence shown in Figure 6 from about +3 to about +106 fused to the amino acid sequence from about +351 to about +369. In yet a fur-ther embodiment of the invention, the amino acid se-15 quence of t:he polypeptide comprises the amino acid sequence shown in Figure 6 from about +3 to about +185.
This invention also provides a method for treating a subject infected with a human immunodeficiency virus.

2~ The method comprises administering to the subject an effective amount of a pharmaceutical composition com-prising an eaffective amount of a therapeutic agent of the invention and a pharmaceutically acceptable carri-er.
LIST OF DRAWINGS
Figure 1. Cytofluorographic Patterns of Indirect Immunofluorescent Staining with OKT*4 and OKT*8;
3p Figure 2. Northern Blot Analysis of RNA Derived from T4+ and T4-L Cells and Human Cells;
Figure 3. Restriction Nuclease Maps of pT4B and the T4 Gene, Sequencing Strategy, and Recombinant Vectors:
Figure 4. Southern Blot Analysis of DNA from Untransforme:d and T4+ L Cells and T, B, and Nonlymphoid Human Cells:
* Trademarks, - ~~~070~
-5a-;,~'~ OF DRAWINGS (CONT' D) Figure 5. 7;mmunoprecipitation of the T4 Glycoprotein from NIH 3T3 Cells Transformed with the Retroviral Expression Constructs:
Figure 6. Nucleotide Sequence of the T4 cDNA AND
Translated Sequence of the T4 Protein:
Figure 7. In Vitro Translation RNA derived from SP6 Transcription:
Figure 8. Schematic diagram of the T4 glycoprotein spanning the cell membrane:
Figure 9. Alignment of the Variable, Joining, and Transmembrane Regions of T4 with Members of the Immunoglobulin Gene Family:
Figure 10. Fcestriction nuclease map of the T4 gene in human chromosomal DNA:
Figure 11. Recombinant Retroviral Expression Vectors and Construction of Transformed Cells:
Figure 12. The Efficiency of Infection of Naturally-Isolated and Transformed T4+ Cells:
Figure 13. Formation of Syncythia in T4+ HeLa Transformants:
Figure 14. Flow C'ytometry Analysis of AIDS Virus binding to T4+ Transformed Cells;
Figure 35. Northern Blot Analysis of RNA Derived from Human and Mouse Brain, Lymphoid, and Myeloid Cells;
Figure 16. F~lasmid map of a psT4DHFR:
Figure 17. Fluorescent histogram (cell number vs.
fluorescence inte:nsity):
Figure 18. Inhibition of HIV infectivity by sT4.
ri,, r~,. .:; ,:.

-6- 1~~0702 BRIEF DESCRIPTION OF THE FIGURES
Figure 1. C~rtofluoroaraphic Patterns of Indirect Im munofluorescent Staining with OKT~4*and OKTe8*
Cells (~~ x 105) where incubated with the mouse monoclonal antibodies OKT~4H or OKT~8, washed, and then incubated with FITC conjugated goat anti-mouse immuno-globulin. The cells were analyzed on a FRCS IV Cell Sorter and plotted by a VAX 11/780 computer as cell number vsc. log fluorescence. Untransformed NIB 3T3 cells an<i L cells gave identical cytofluorographic tracings. Fro 2.2 is a leukemic T cell line with phe-notype T3 ; T4+; T8+; T11+. LTD-4 is a T4+ primary L
cell tran,sformant obtained following transfer of total genomic DNA. 3A+ is an NIB 3T3 cell line that was ,transformed with the T4-pMV6tk/neo retroviral expres sion construct.
Figure 2. Northern Hlot Analysis of RNA Derived from T4+ and T4: I, Cells and Human Cells Three micrograms of poly(A)+ RNA or 12 ~g of .total RNA
(peripheral T cells and thymocytes) were electrophore sad through a 0.8% agarose-formaldehyde gel, blotted onto GenaScreeri (New England Nuclear), and probed with a 3ZP-labeled 0.6 kb T4 cDNA insert. T4+ cells include LTD-4 (T4+, T8 L cell transformant), SK-7 T cell hy-bridoma (T'4+, T8-), OT-CLL leukemia (T4+, T8 ), Fro 2.2 leukemia (T4+, T8 ), T4- enriched peripheral T
lymphocytes, and human thymocytes. T4 cells include untransfonned cells, tk7 (T8+ L cell transfonaant), HeLa cells, human neuroblastoma cells (Il~t), and T8 enriched peripheral T lymphocytes. The human thymo cyte lane was exposed four times longer and photo * Trademar~;s -graphed on high contrast film.
Figure 3. Restriction Nuclease MaDS of pT4B and the T4 Gene, SeguE:ncing Strategy, and Recombinant Vectors A. Alignment of the Bam HI restriction fragments of pT4B cDNA and the T4 gene. The order of Bam HI frag-ments in the T4 gene was determined by Southern blot analysis and genomic clone mapping. The alignment of the 5' end of pT4H and the T4 gene is shown by dotted lines, and the shaded region in pT4B corresponds to the coding sequence. The indicated sizes are in kilobases.
B. Sequencing strategy. Arrows indicate length of sequence determined by subcloning fragments into M13 and sequencing by the dideoxy termination procedure (36) .
C. Eukaryotic expression vectors. These constructs contain two Moloney murine leukemia virus long terminal repeats (L'rRs) whose orientations are indicated by arrows. The pT4B cDNA was subcloned into the Eco RI
site of each vector in the orientation indicated. (a) The T4-pVcos7 construct. (b) The T4-pMV6tk/neo con struct contains the neomycin phosphotransferase gene fused to the HSV thymidine kinase promoter.
Figure 4. Southern Blot Analysis of DNA from Untrans-formed and 'r4+ L Cells and T, B, and Nonlym hoid Human Cells Ten microgr<ims of cellular DNAs were digested with Bam HI, electrophoresed through a 0.8% agarose gel, blotted onto GeneScreen, and probed with a nick-translated pT4B
cDNA insert., The indicated size markers are in kilo ~.3~~'~~?
_$_ bases. Hybridizing bands of sizes 20 kb, 6.6 kb, 4 kb, 1.8 kb, and 1 kb appear in all human DNAs. DNAs from T4 , nonlymphoid origin include untransformed L cells, human fibroblasts (GM), human neuroblastoma cells (NB), and HeLa ~~ells. CB, CP58, and CP94 are DNAs derived from EBV-transformed human B cell lines. LTD-4 is the T4+ primary L cell transformant. RPMI and HSH2 are T4 human T cell leukemic lines; E+ cells and thymocytes (Thym.) contain T4+ T cells. OT-CLL, Jurkat (lurk.), Fro 2.2, c:EM, and Molt 4 are T4+ T cells. gaM4 is a genomic clone which contains sequences spanning the 3~
end of the T4 gene.
Figure 5. Cmmunoprecipitation of the T4 Glycoprotein from NIH 3T3 Cells Transformed with the Retroviral Expression Constructs L-[35S]-methionine labeled proteins from two indepen dent NIH 3T3 transformants, peripheral T lymphocytes, and untran~sformed 3T3 cells were subjected to lentil lectin chromatography to enrich for glycoproteins. 2.5 x 106 cpm of each sample was precleared and then im munoprecipitated with OKT~4 monoclonal antibodies and Protein A-S~epharose. The beads were washed, dissolved in sample buffer, and electrophoresed through a l0%
SDS-polyacrylamide gel under reducing (lanes a-d) and nonreducing (lanes a and f) conditions. Lane a, un transformed NIH 3T3 cells. Lane b, T4C2, an NIH 3T3 cell transformed with the T4-pVcos7 construct. Lanes c and e, 3A+,, an NIH 3T3 cell transformed with the T4 pMV6tk/neo construct. Lanes d and f, peripheral human T lymphocyi:es. Relative molecular masses (Mr) are given in ki:Lodaltons.

1~~~~~?
-g-Figure 6. Nucleotide Seguence of the T4 cDNA and Trans lated Sectu~ence of the T4 Protein The nucleotide and predicted amino acid sequences of the cDNA clone pT4B obtained according to the sequenc-ing strategy outlined in Figure 38. Numbers shown above the amina acid sequence designate amino acid residue positions, The numbers on the right show nu cleotide F~ositions. All extracellular cysteines are marked by (~) or (o). The leader sequence (L), vari able-like (V), joining-like (J), transmembrane (TM), and cytopla~smic (CYT) regions are indicated by horizon tal arrows below the sequence, although the exact boundaries are ambiguous. Two potential N-linked gly cosylation sites (Asn-Leu-Thr) are also indicated (cxo).
Figure 7. In Vitro Translation RNA derived from SP6 Transcri ti~on The full length T4 cDNA insert was subcloned into the RNA expression vector pSP65 (Promega Biotec). Linear-ized plasmid DNA was transcribed with SP6 polymerise (40), and RNA was translated in a wheat germ system (~thesda Research Laboratories) containing L-[35S]-methionine. The in vitro translation products were subjected to electrophoresis through a 10% SDS-poly-acrylamide gel (lane T4). Bovine pituitary RNA (BP) was used as a control. Relative molecular masses (Mr) are given in kilodaltons.
Figure 8: Schematic diagram of the T4 glycoprotein spanning the: cell membrane T4 consists of four tandem VJ-like domains (ViJi-V4J4), a hydrophobic membrane-spanning segment (shaded area), and a charged cytoplasmic region (CYT). Two potential N-linked glycosylation sites in the extracellular por-tion are :Lndicated ( ---). The positions of introns 2-8 in the T4 gene are also marked ( ).
Figure 9. i~lignment of the Variable, Joining, and Tran-smembrane Regions of T4 with Members of the Immuno-globulin Gene Family A. Alignment of the variable region amino acid se quence of T4 with a mouse kappa light chain immuno globulin J606 (66), T8 (20), a human T cell antigen receptor p-chain YT35 (97) , and a human T cell antigen receptor a--chain HPB-MLT a(98). The invariant resi dues in the light chain variable region are included (Inv.) in the alignment. The alignment was performed in order to maximize identities and structural homolo gies with T4, which appear as boxed residues. The lines below the sequence with letters A, B, C, C', D, E, F, and G indicate the residues which form ,B-strands (67). ,B-strand G continues into the J sequence.
B. Alignment of the joining region amino acid sequence o! T4 with the consensus J sequences of the T cell antigen receptor p-chain, immunoglobulin lambda and kappa light: chains, and the J sequence of the human T
cell receptor a-chain (99).
C. Alignment of the transmembrane regions of T4 and an MHC class II ,B-chain (100). The putative transmem-brane domain (TM) is indicated below the sequence.

Figure 10. Restriction nuclease map of the T4 Qene in human chromosomal DNA
The positions of the 9 axons were determined by genomic clone mapping, Southern blot analysis, and nucleotide sequencing. The leader sequence (L), variable-like (V), joining-like (J), transmembrane (TM), and cyto-plasmic (C:YT) regions are boxed. The position of the methionine codon surrounded by the intitiation consen-0 sus sequence is indicated (ATG) at the beginning of the leader axon (L) N the termination codon TGA is shown at the end o~f the second cytoplasmic axon (CYT). The indicated sizes are in kilobases.
~5 Figure 1l.:Recombinant Retroviral Expression Vectors and Construction of Transformed Cells A. Recombinant retroviral expression vectors. pMV7 contains t.wo directly repeated Moloney murine sarcoma 20 virus long terminal repeats (LTRs) in the orientation indicated by arrows. pMV7 also contains the bacterial neomycin plhosphotransferase gene (neo) fused to the HSV
thymidine kinase promoter (tk). Full length cDNA in-serts encoding T4 (T4H) (70) or T8 (T8F1) (20) were 25 subcloned :into the Eco RI site in the orientation indi-cated by arrows, generating T4-pMV7 and T8-pMV7, re-spectively,. The coding sequences are shown as shaded regions. '.Che indicated sizes are in kilobases.
8. Retrovirus-Mediated Gene Transfer Strategy.
Figure 12. The Efficiency of Infection of Naturally-Isolated and Transformed T4+ Cells ~. ~ f~ (~ '~l (~ ~

Calls were inoculated with serial 10-fold dilutions of AIDS virus, incubated for 18 hours at 37°C, washed, and plated in microculture. The frequency of infected cultures was determined by an enzyme-linked immunoab-sorbent assay (ELISA) 12 days post-infection (46). The results were plotted as % positive cultures vs. log virus dilution. Infectious virus titer (ID-50) is defined as 'the reciprocal of the dilution at which 50%
of the cultures are positive for virus (47). Naturally isolated T4+ cells include phytohemagglutinin (PHA) stimulated normal peripheral lymphocytes and the T cell line CEM ( U---O ) , T4+ transfected cell lines include HSH2-T4+ T cells ( ~--~1 ) and Ra j i-T4+ H cells ( ~----/ ) . The T8+ transfected cell lines HSB2-'T8+ and Raj i-T8+ ( D--.~ ) served as controls in these studies.
Figure 13. Formation of Syncythia in T4+ HeLa Tran-formants A. 2 x 105 monolayer HeLa-T4+ tranformants were mixed with 2 x 104 AIDS virus-producing H9 cells and incubat-ed at 37'C. Inspection of the cultures after 18 hours revealed that over 90% of nuclei in the monolayer sheet wars contained within syncytia.
H. Anti-T4A :monoclonal antibody (1:20) was added to the mixed cultures at the time of seeding. Inspection of the cultures. after 18 hours revealed a complete ab-sence of cell fusion.
Cultures were photographed at 160 X magnification.
Figure 14. 1?low Cytometry Analysis of AIDS Virus bind-ing to T4+ ~~ransformed Cells Column A. Cells (5 x 105) were incubated with fluores-cein-conjugated anti-T4A (-) or anti-T8 (---) monoclonal. antibodies, washed, and analyzed by cyto-f luoromet=y .
Column e. Cells (5 x 105) were incubated with buffer (---), or AIDS virus ( ), washed, incubated with fluorescei.n-conjugated anti-AIDS virus antibody, and analyzed by cytafluorometry.
Column C. Cells (5 x 105) were incubated with buffer (---), or with anti-T4A monoclonal antibody followed by AIDS virus: ( ), or with anti-T8 monoclonal antibody followed by AIDS virus (-'-'-). After a wash, fluores-cein-conjugated anti-AIDS virus antibody was added and the cells were analyzed by cytofluorometry.
Flouorescence histograms (cell number vs. fluorescence intensity) of each cell line are arranged horizontally.
Figure 15. Northern Blot Analysis of RNA Derived from Human and 1!iouse Brain, Lymphoid, and Myeloid Cells A. Northern blot analysis of human RNA samples. One microgram ~of poly (A) + RNA from Ra j i (T4 B cell line) , U937 T4+ monoc tic cell line ( y ) , and Jurkat (T4+ T cell line), and five micrograms of poly(A)+ RNA from cere-bral corte:~c, were electrophoresed through a 1% agarose-formaldehy<ie gel, blotted onto Hybond (Amersham), and probed with a 32F-labelled T4 cDNA insert, pT4B (70).
B. Northern blot analysis of mouse RNA samples. Five micrograms of paly(A)+ RNA from 3T3 cells (fibroblast cell line), forebrain, and hindbrain, and 20 micrograms of total RNA from thymocytes, were eletrophoresed ~~~~0~~~

through a 1% agarose-formaldehyde gel, transferred onto Hybond, and probed with a 32P-labelled L3T4 cDNA in-sert, pL3T4B.
Figure 16. Plasmid map of psT4DHFR
Plasmid psT4DHFR is a pUCl8 derivative containing by 1-1257 of the T4 cDNA clone pT4B which encodes the leader and extracellular segment of T4. This sT4 cDNA is inserted bE~tween an SV40 early promoter and a synthetic linker containing a TAA termination codon (inset) fol-lowed by the polyadenylation region of the bovine growth honnone gene. The sT4 expression cassette is linked to a mouse hdfr expression cassette consisting t5 of the ~-g:lobin promoter, mouse dhfr coding sequence, and the SV40 polyadenylation region.
Figure 17. Fluorescent histogram (cell number vs fluorescence intensity) sT4 inhibits HIV binding to T4+ CEM cells. Cells were incubated with buffer (~-.-.-), HIV preincubated with sT4 ( - ), or with HIV preincubated with concentrat-ed control supernatant from untransformed DXB-11 cells (- - -), waslhed, exposed to fluorescent-conjugated anti-HIV antibody, and analyzed by cytofluorometry. A fluo-rescent histogram (cell number vs. fluorescence inten-sity) is shown.
Figure 18. Inhibition of HIV infectivity by sT4 Infectivity titration of an HIV inoculum (ID-50 assay) was performead. Serial 10-fold dilutions of virus inocu lum are incubated with indicator cells (PHA-stimulated human lymphocytes) for 18 hrs. The cells are then washed and plated in microculture (1 X 105 cells per culture, :l0 cultures per dilution). At day 4, 8, and 12, supernatants are tested for HIV by the antigen capture assay. ID-50 titrations were performed in media coni:aining 8.6 ~g/ml sT4 which was added to the HIV dilution 30 minutes prior to inoculation of cells and maintained in the culture media throughout the experiment. ( ~ ), or in media containing sT4 introduced after the initial 18 hr inoculum ( O ) (delayed addi-0 tion control), or in media without sT4 ( O ) (con-trol).
A. Plot of percent of cultures positive for HIV at day 8 versus dilution of virus inoculum.
H. Plot of ID-50 (reciprocal of virus dilution at which 50% of cultures are positive) at days 4, 8, and 12.
C. Plot of percent of cultures positive for HIV at day _ 8 versus varying concentrations of sT4 using a 10 dilution of HIV.

DETAILED DESCRIPTION OF THE INVENTION
This invention provides a therapeutic agent capable of specifically forming a complex with human immunodefi-ciency virus envelope glycoprotein comprising a poly-peptide. :Cn one embodiment of the invention, the amino acid sequence of the polypeptide comprises the amino acid aequence shown in Figure 6 from about +3 to about +185 fused to the amino acid sequence from about +351 to about +369. In another embodiment of the in-vention, the amino acid sequence of the polypeptide comprises the amino acid sequence shown in Figure 6 from about +3 to about +106 fused to the amino acid sequence from about +351 to about +369. In yet a fur-ther embodiment of the invention, the amino acid se-~5 quence of the polypeptide comprises the amino acid sequence shown in Figure 6 from about +3 to about +185.
A pharmaceutical composition useful as a therapeutic agent for the treatment of a subject infected with a 20 human immunodeficiency virus is also provided. This pharmaceutical composition comprises an amino acid sequence of the present invention which is capable of specifically forming a complex with a human immunode-ficiency vi~:-us envelope glycoprotein and is soluble in 25 an aqueous solution and a pharmaceutically acceptable carrier. Such pharmaceutically acceptable carriers are known in the art to which the present invention per-tains and include, but are not limited to, 0.01-O.1M, preferably 0.05 M, phosphate buffer or 0.8% saline.
A method for treating a subject infected with a human immunodefic:Lency virus is also provided. This method comprises administering to the subject an effective amount of a pharmaceutical composition containing a 1340~I02 pharmaceutically acceptable carrier and an amino acid sequence of the present invention, capable of specifi-cally forming a complex with a human immunodeficiency virus envelope glycoprotein and soluble in an aqueous solution, so as to render human immunodeficiency virus-es (also referred to herein as AIDS viruses) with which the subject is infected incapable of infecting T4+
cells.
Characterization of the in vitro biological and immuno logical properties of the sT4 protein indicate the protein is of value in the prevention and treatment of AIDS. Studies indicate the sT4 protein acts as an inhibitor of extracellular and cell to cell spread of the virus. Because sT4 has been shown to block virus binding to T4+ target cells in culture, it is believed administrai:ion of sT4 to infected persons would' act to inhibit the extracellular spread of the virus. There-fore, sT4 is of value both as a prophylactic and thera-peutic agent for treatment of AIDS.
As a prophylactic, sT4 is administered to individuals at high-risk for the disease or individuals who show exposure to HIV by the presence of antibodies to virus.
Administration of an effective amount of sT4 at an early stages of the disease or prior to its onset acts to inhibit infection of T4+ lymphocytes by HIV. As a therapeutic., administration of sT4 to persons infected with HIV acts to inhibit extracellular spread of the virus.
Fusion between HIV infected cells and other T4+ lympho cytes also appears to be a route of virus spread.
Further, fusion may be responsible, in part, for the impainaent of T~+ lymphocyte function and ultimately ~.'~~0'~~~
the depletion of T4+ lymphocytes in infected individu-als. Ce7.1 fusion is dependent on both viral envelope gene products and the T4 receptor and can be inhibited by the 01tT4A, or similar monoclonal antibodies (blabs) (120). ::T4 interferes with cell fusion and therefore is expected to diminish the cell to cell spread of virus and the loss of T4+ lymphocyte function.
The T4 receptor is monomorphic and thus sT4 is believed to be a universal inhibitor of virus which recognize the surface domain of the T4 receptor, including all HIV.
sT4 may be used in combination with other agents, for example, in association with agents directed against other HIV proteins, such as reverse transcriptase, protease or tat. An effective therapeutic agent against HIV should prevent virus mediated as well as cell to cell transmission of infection. sT4 may also be used in combination with other anti-viral' agents, for examp7.e, azidothymidine (AZT).
The sT4 protein of this invention also has utility as an inhibitor of T4+ cell function. Numerous studies suggest a critical role for the CD4 receptor (CD4 is general tearminology for the human T4 receptor and its counterparts in other mammalian cells) in immune toler-ance, parl:icularly in the pathogenesis and progression of autoimmune diseases and in host specific graft re-jection. Of particular relevance to sT4 are the obser-vations with anti-C04 blabs. Through their association with the C:D4 receptor, certain of these blabs ameliorate autoimmune: responses and graft rejection. Examples of such action include inhibition of T-cell function _in vitro, for example, antigen induced proliferation, 1~4~'~U~

lymphokine secretion and helper cell function by cer-tain anti-CD4 Mabs; treatment of systemic lupus erythe-matosus by administration of anti-CD4 Mabs to retard the onset of murine lupus: and grafting studies in mice wherein a single dose of murine Mab directed against the murine CD4 receptor results in acceptance of the allograft.
The molecu:Lar consequence of the binding of Mab to CD4 is unclear. The Mabs may block the association of CD4 with its ligand, which ligand, evidence suggests, is a conserved ~epitope on MHC class II antigens (121,122).
However, at least some of these same Mabs inhibit CD4 cell activation by an apparent class II independent path.
sT4 is also believed to inhibit T-cell interactions as a competitor of cellular T4, perhaps by binding to extracellul.ar target molecules which normally interact with the surface domain of the T4 receptor. This dis tinction between Mabs and sT4 could have important functional consequences. For example, whereas some Mabs direcited against T4 elicit a response on the T4 cell, sT4 may elicit a response on cells expressing MHC
class II antigens. Also, the affinity of T4 for its presumed class II ligand appears to be quite low com-pared to t;he high affinities of Mabs directed against T4. Thus, although Mabs and sT4 may interfere with the same processes, they would affect different target molecules and different target cells.
As a prophylactic or therapeutic, sT4 is administered parenterally, preferably intravenously. The agent can be administered by infusion or by injection, e.g., daily, weekly or monthly. The amount and rate of sT4 administration is selected so as to maintain an effec-tive amount of sT4 in circulation. An alternative mode of administration would be extracorporal, employing sT4 as a dialysis agent.
The sT4 protein of this invention can also be used as a reagent t~o identify natural, synthetic or recombinant molecules which act as therapeutic agents or inhibitors of T4+ cell interactions.
For examp7le, sT4 protein can be used in screening as-says, such as assays for protein interaction measured by ELISA based methodologies, to provide a biochemical-ly pure, aqueous soluble reagent which may be used in combination with other reagents to assay for competi-tore of the T4 receptor surface domain interactions.
For example, since sT4 binds to HIV env protein or mixtures containing HIV env proteins, it can be used to screen for inhibitors of virus binding. Based on _in vitro data. showing that sT4 binds to cells expressing HIV env proteins, sT4 can also serve as a selective targeting molecule for HIV infected cells in vivo. As a target sspecific carrier protein, sT4 can serve, for example, as the carrier protein for delivery of cyto-toxic agents to the infected cells.
In addition, based on data showing that the T4 receptor specifically associates with I~iC class II antigens on antigen-presenting cells as suggested by the class re-striction of T4+ cells, sT4 can be used in combination with class II antigens to test for inhibitors of T4 lymphocyte - target cell interactions. In addition to these examples, which are based on direct binding as says between sT4 and its target molecules, more complex assays can be designed which rely on the biochemical -21- 1340'02 responses to sT4 recognition.
Further provided is an expression vector encoding a polypeptide, the amino acid sequence of which comprises the amino acid sequence shown in Figure 6 from about +3 to about +185 fused to the amino acid sequence from about +351 to about +369. In another embodiment of the invention, the expression vector encodes a polypeptide, the amino acid sequence of which comprises the amino acid sequence shown in Figure 6 from about +3 to about +106 fused t:o the amino acid sequence from about +351 to about +369. In yet an additional embodiment of the invention, the expression vector encodes a polypeptide, the amino acid sequence of which comprises the amino acid sequence shown in Figure 6 from about +3 to about ~5 +185.
A host cell comprising the expression vectors of this invention is also provided by the present invention.
In one embodliment of the invention, the suitable host is a bacterial cell. In another embodiment of the invention, the bacterial cell is an Escherichia coli cell. In yet another embodiment of the invention, the suitable host a eucaryotic cell. In a further embodi-ment of the invention, the eucaryotic cell is a mamma-25 lian cell. In yet a further embodiment of the inven-tion, the eucaryotic cell is a yeast cell. In still another embodiment of the invention, the suitable host is an insect cell.
30 The present :invention also provides a means for produc ing sT4 consisting of the predicted extracellular do main of the T4 receptor. Using that portion of the T4 cDNA which encodes the leader and extracellular domains of the T4 receptor, i.e., pre sT4, vectors are con -22- 13 4 0'~ 0 2 structed c:apable of overexpression of sT4 in mammalian cells.
The sequence of one sT4 is as follows:

~

~ - _ CAA GOC CAG AGC ixT GOC ATP TCT G'IG GGC T~C'A GGT CXx TAC 'IGC TCA GaC C)CT
'IOC TCc C'I~C GGC AAG GOC AC'A AZG AAC COG GGA GI~C OLT TIT AGG CAC TIG CTT CZC GZG
CIG C~F1 Met Asn Arg Gly Val F3~ Phe Arg Isis Ieu IEU Ieu Va1 Leu Gln * * * +1 cIG GC~ CiC CIC CX'A GCA GCC ACT C3~G GGA AAG AAA GTG GIG CIG OGC AAA AAA GOG
GAT
1p Leu Ala Leu Leu Faro Ala Ala 'Ihr Gln Gly Lys Lys Val Val Ieu Gly Lys Lys Gly Asp ACA GIG GAA CZG AOC TGT ACA GCT 'IOC (34G AAG AAG AGC ATA CAA TIC CAC 'It3G
AAA AAC
Thr Val Glu I~ Thr Cys Zhr Ala Ser Gln Lys Lys Ser Ile Gln F3~e His Trp Lys Asp * *
T~ AAC cAG ATA Ate ATr c~ Gc,~ AAT c~ Gcc Toc Toc TrA Acr ~ c~cr aA 'roc AAG
Ser Asn Gln Ile Lys Ile Leu Gly Asn Gln Gly Ser Ffie Leu 'Ihr Lys Gly Pro Ser Lys * * * *
CIG AAT GAT CGC C~:.T GA~C TCA AGA AGA AGC CIT 2CIG GAC CAA GGA AAC TIr CCC
CIG AZC
Leu Asn Asp Axg A:la Asp Ser Arg Arg Ser Leu Trp Asp Gln Gly Asn Pipe Pro Ieu Ile A2C AGG AAT CTr A~~G ATA GAA GAC TrA GAT ACT TAC ATC TGT GAA GTG GAG GAC C:yG
AAG
Ile Lys Asn Leu Lyg Ile Glu Asp Ser Asp Thr Tyr Ile Cys Glu Val Glu Asp Gln Lys * * * 104 GAG GAG G'IG CAA T1C CTA GTG TIC C~ T~ A(,~I, ~ p,~ ~ ~ ACC CAC C'IG CTT CMG
Glu Glu Val Gln L~u I~ Val Ptse Gly Ieu 'Ihr Ala Asn Ser Asp ~. ~ ~ ~ Gln GGG CMG AGC C1G ACC CIG ACC TIC GAG AGC OOC OCT GGT ALT AGC COC 'I~'A GTG CAA
TGT
Gly Gln Ser Leu Thr Ieu Thr Leu Glu Ser Pro Faro Gly Ser Ser Pro Ser Val Gln Cys AGG AGT CCA AGG GGT AAA AAC ATA CAG GGG GGG AAG ACC CTC T~OC GIG TrT CMG CIG
GAG
~ ~ ~ ~ Gly Lys Asn Ile Gln Gly Gly Lys Thr Leu Ser Val Ser Gln ~ Glu C'TC C~G GAT AGT G~3 AC7C TGG ACA TGC ACT G'DC TIG CAG AAC CAG AAG AAG GIG GAG
TIr Leu Gln Asp Ser Gly Zhr Trp Thr L~ ~ V~ ~ G~ ~ G~ Lys Ly5 Val Glu Phe * * * 183* *
AAA ATA GAC ATC GIG GTG CPA GCT Tit CAG AAG GOC TaC AOC ATA GTC TAT AAG A1~1 GAG
Lys Ile Asp Ile Val Val Ieu AIa plye Gln Lays Ala Ser Set Ile Val Tyr Lyg Lys Glu * * *
GGG GAA CAG GTG i;~4C TTC TOC TIC OCR CTC GaC TIT AC~~ GTT GA71 AAG CZG AID
GOC AGT
Gly Glu Gln Val ;4sp phe Ser Ffie pto L~ pea p~ ~ Val Glu Lys Leu ~1r Gly Set.
790 800 810 820 830 ' 840 * * * * * *
~ GAG CIG TGG '.tt3G CAG GCG GAG AOG GtZ' TCC TaC TrOC AJ~G TCt' TC~G ATC AiOC
TIT G~1C
Gly Glu Ieu Trp ~,~p Gln Ala Glu Arg Ala Ser Ser Ser L~ys Ser ~p Ile Thr pl~e Asp CIG AAG AAC AAG CAA GTG TCT GTA AAA CGG GTP AOC CAG GAC CCT AAG CTC CAG ATG
GOC
Leu Lys Asn Lys Gilu Val Ser Va1 Lys Azg Val ~r Gln Asp Pro Lyg ~ G~ ~ Gly * * * *
A~ Apc crc o~ rrc cAC cIr Aroc cIC ooc c~ ccc Trc ocT cac TAT ccT occ TcT Gc,~
Lys Lys Leu Pto Leu Hi.s Leu ~r Leu Pro Gln Ala Lai Pto Gln Tyr Ala Gly Ser Gly AAC CTC A(aC CIG GOC CIT GAA GCS AAA ACA GGp. AAG TTG CAT CAG GAA GTG AA~C CTG
GTG
Asn I,eu Thr LQU Ala LQU G:lu Ala Lys Thr Gly Lys LEU Hi.s Gln Glu Val Asn Leu Val * ~k * * * *
GTG ATG AGA GOC Fur' Q~G CTC CAG A~1A AAT TIG AOC TGT GAG GTG TGG GGA COC AG1C
TCC
Val Met Atg Ala ~r Gln Leu Gln Lys Asn Leu ~r Cys Glu Val Ttp Gly Pro Zhr Ser ttT AAG CIG ATG C'.LG AGC TTG AAA CIG GAG AAC AAG GAG Gc~ AAG GTC T~fT AAG CMG
GAG
~'° LYs ~u ~t Leu Ser LEU Lys Leu Glu Asn Lys Glu Ala Lys Val Ser Lys Arg Glu 1150 1160 1170 1180 1190 upp * *
AAG GCS GTG TGG G7~ CTG AAiC OCT GAG GOG c~G ATG 'IGG CAG TGT CTG CTG AGT GAC
TCG
Lys Ala Val Ttp Va~l Lau Asn pro Glu Ala Gly Met Ttp Gln Cys LQU Ieu Ser Asp Ser u10 1220 1230 1240 1250 1260 * ,w GGA CAG GTC CIG CIG GAA TtSC AAiC ATC AAG GTT CTG DOC ACA T3G TOC AOC CCG GIG
TAA
35~ G1n val Lai Less Glu Ser A~ Ile Lys Val Leu Pte 'tier Trp Ser 'Ihr Pro3 *
Tu~G ~C CT~C TAG A

....

The coding sequence for sT4 is obtained, for example, by synthesizing 'the gene using the known DNA sequence, by standard cloning techniques based on the sequence and by. reis~olation by detection of protein, i.e., tran sfection of cDNA clones from T4 expressing cell lines and identii'ication by antibodies directed against the protein. c:DNA clones carrying the sT4 coding sequence are identified by use of oligonucleotide hybridization probes. The probes are designed based on the known sequence of the T4 protein. Having identified a clone carrying th.e sT4 coding sequence, the coding sequence is excised by the use of restriction endonucleases and inserted into cloning and/or expression vectors. In an expression vector, the sT4 coding sequence is opera-tively linked to regulatory functions required or de-~5 sirable for transcription, translation and processing of the coding sequence.
Regulatory functions include, for example, functions required for RNA polymerase binding and transcription, as well as other functions such as polyadenylation and enhancement of transcriptional sequences. The promoter can be regu:latable so that, for example, expression is not induced until after transfection and selection of transformed clones. Promoters useful in the practice 0! the invention include, for example, the SV40 early promoter, anal the long terminal repeats (LTR's) of rows sarcoma virus, moloney sarcoma virus or cytomegalovirus (CMV) .
prior to transfection, the sT4 minigene, i.e., the gene encoding the leader and extracellular domains of the T4 receptor, preferably is incorporated into a larger DNA
molecule which comprises a genetic selection marker system. The selection marker system consists of any gene or genes which cause a readily detectable pheno-typic change in a transfected host cell. Such pheno-typic change may be, for example, foci formation, drug resistance, such as genes for 6418 or hygromycin B
resistanc~a: or other selectable markers such as xan-thine guanine phosphoribosyl transferase (xgprt), thy-midine kinase (TK) and galactokinase (galK). A selec-tion marker which permits gene amplification can be employed i.o increase copy number, whether by increased 0 transfectj.on efficiency or by enhanced intracellular replication of the gene of interest and the selection marker. Such markers which also serve to amplify gene copy number include genes for dihydrofolate reductase (methotrexate resistance), CAD (N-phosphonacetyl-L
aspartate resistance) and adenosine deaminase (2-de oxycoformycin resistance).
Following transcription and translation in mammalian cells, the: leader sequence appears to be cleaved and mature sT4 is secreted into the conditioned medium.
In the preferred practice of the invention, the sT4 minigene is linked with a human H-ras or a mouse dihy drofolate reductase (DHFR) minigene to create the ex Pression v~actors.
The sT4 minigene is linked, for example, with the human H-ras or naouse DHFR in order to provide a selection marker and means to selectively amplify gene expression through co~-transfection with these genes. Common se-lection markers include, for example, DHFR, 6418 or hygromycin which select for integration of as few as a single copy of the gene of interest. Amplification with, for example, methotrexate (mtx) in the DHFR sys-tem results. in overexpression of the gene.

An alternate means of increased expression of the gene includes u,se of the ras proto-oncogene. The ras gene family includes the H-ras, K-ras and N-ras genes. In a preferred application of the invention, the H-ras gene is used.
Other DNA l:unctions can be linked directly or indirect-ly to the sT4 minigene or such functions may be un-linked. See, for example, Axel, U.S. Patent No.
4,399,216.
Overexpression of gene products in mammalian cells can be achieved by transient or stable means. Transient ~5 overexpression can be achieved by viral methods such as the use of vaccinia virus vectors or by gene amplifica-tion methods such as with SV-40 based vectors in cells which support SV-40 replication. These methods ulti-mately lead to cell death. Stable overexpression can 20 be achieved by generation of multiple gene copies such as through selection for gene amplification or through the use of t:he ras proto-oncogenes.
Overexpressj.on of sT4 protein by co-transfection using 25 the H-ras gE:ne system can be achieved using a number of di!lerent cell lines, with the preferred cell line being the NI:H-3T3 cell line which is a contact inhibit ed mouse fi.broblast cell line. other cell lines in clude the na~rmal rat kidney (NRK) (ATCC 1571) cell line and the rat. embryo fibroblast 52 (REF-52) cell line ( 115 ) .
When using the selection marker system, for example, DHFR with avethotrexate (DHFR/MTX technique), wherein amplification of gene copy number is achieved by selec ~.r'~~~~1 ~~

tive amplification, the Chinese hamster ovary cell line (CHO) is preferred. In particular, CHO cells deficient in DHFR are used (116). Other cell types which may be used include, for example, any mammalian cell which has been modified so as to be DHFR .
Some DHFR+ cell types may be used in combination with mutant DHF'R genes which are less sensitive to metho-trexate than normal DHFR (Axel, U.S. Patent No. 4,399,-0 216). In principle, DHFR+ cells may be used in combi nation with normal DHFR genes and an additional domi nant selectable gene such as the gene for 6418 resis tance (117). Transfection is carried out using stan dard techniques (118, 119). These techniques include, for example, calcium phosphate precipitation, DEAE
dextrin induced pinocytosis, electroporation and viral transfection.
Following transfection, a cell which carries the sT4 minigene i:; cultured in a nutrient medium under condi tions which permit amplification of the selectable gene. Standard mammalian cell culture medium can be employed, for example, F12 medium (GIBCO, Grand Island, New York) 'without hypoxanthine and thymidine and con taining 10~% fetal bovine serum. Cell cultures are maintained at ambient pressure at 30 to 45'C. Cells which survive selection and express high levels of sT4 protein are selected for further culturing. Such cells are culturesd under selective conditions and the prod-uct, the sT4 protein, is collected and purified.
Cell culture methods which may be employed in the prac tice of th.e invention include, for example, use of adherent cells ar growth of cells in a suspension.
Conditioned medium (CM) can be collected from cells ~.~~o~~oz grown in suspension or adhered to solid supports, i.e., CM is prpeared from adherent cells grown in roller bottles or grown on solid supports and cultured in suspension or in fluidized or packed beds. CM is pre-y pared from suspension cells in stirred-tank vessels.
The sT4 of the invention includes derivatives of the extracellul.ar domain of T4. Such derivatives comprise addition, deletions or subsitutions which alterations do not significantly adversely affect secretion of the protein into the conditioned medium and the affinity of the protein for HIV env protein, i.e., gp120. For example, one or a few amino acids can be added to or deleted from the N- or C- terminus. Or, one or a few amino acids, preferably no more than four amino acids can be inserted into, deleted or substituted for inter nal amino acids. Alternatively, a hybrid protein, i.e., a translational fusion, can be constructed be tween sT4 and a protein carrier, another antigen or other sT4 molecules to prepare a poly-sT4 molecule. In yet another alternative, sT4 can be synthetically con-jugated to .3 carrier molecule.
One embodiment of a sT4 derivative is illustrated in th. Examples below (ses Example 3). Affinity of the sT4 for HIV env can be demonstrated by a competitive binding assay using a sT4 molecule having a known af finity or using antibodies which recognize the T4 re ceptor, such as OKT4 and OKT4A. Useful derivative sT4 molecules o1a the invention are selectively precipitated from conditioned medium by oKT4A as shown in Example 3.
Derivatives can be prepared chemically, after expres-sion, or genetically, prior to expression, by manipula-tion of the coding sequence for the leader and/or extracellula,r domain.

The sT4 of the invention can be purified from the spent culture media using various protein purification tech-niques, for example, affinity chromatography; ion ex-change chromatography; size exclusion chromatography;
hydrophobic chromatography or reversed phase chromatog-raphy.
sT4 can bas purified by affinity chromatography using general group-specific adsorbents, for example, carbo-hydrate binding or dye affinity ligands: or using ligands that specifically bind to sT4, for example, monoclonal antibody or HIV gp120 protein or portions thereof.
An exemplas.-y purification scheme comprises: (1) grow-ing cells in a serum-free selection growth media; (2) clarification of the conditioned media: and (3) separa-tion of sT4 of the invention from other proteins Present in the canditioned media.
In the prei'erred method, the sT4 is purified from the serum-free culture medium using a series of chromatog-raphy steps which are based on the physical properties 0! the sT4 molecule. The sT4 may also be purified from culture medium captaining serum using similar chroma-tography meithods .
In the pre:~erred method of purification of sT4, the culture medium is f first passed through an ion exchange column, preferably an S-Sepharosee (Sulpho-propyl Sepharose) column, which binds the sT4 while the major-ity of contaminating proteins flow through the column.
The protein sample is then eluted using a linear salt gradient. A second ion exchange column is used. This 140"l0~

column, preferably a Q-Separose~ (quarternary amino ethyl Sepharose~ column has properties such that the contaminating proteins present in the sample are bound to the column while the sT4 does not bind and is. recov-ered in the column flow-through buffer. Finally, a gel filtration column is used which acts to remove remain-ing contaminating materials.
An alternative method of purification of sT4 involves the use of monoclonal antibodies directed against sT4.
The sT4 protein can be purified in one step by passage of clarified culture media through an affinity gel support to which monoclonal antibody directed against sT4 is bound. The sT4 will bind to the column at the antibody binding site while all contaminating proteins wash through the column. The sT4 is then eluted from the column under conditions that prevent sT4 protein from being inactivated.
Further provided is a method of producing any one of the above described therapeutic agents capable of spe cifically forming a complex with human immunodeficiency virus envelope glycoprotein which comprises growing the host vector system of the invention under suitable conditions permitting production of the therapeutic agent and recovering the therapeutic agent so produced.
The sT4 can be used in diagnostic assays for the detec-tion of T4 proteins or the molecules with which they interact. For example, quantitation of T4 and T4+
cells and antibodies to T4 would be of diagnostic value for AIDS.
In addition, sT4 can be used to generate new diagnostic reagents, for example, Mabs or other types of molecules for use i.n the standard immunologic assays, i.e., ELISA, capture immunoassays, radioimmune assays. Be-cause sT4 displays the OKT4, the OKT4A and most if not all of the other surface epitopes of the T4 receptor, sT4 is especially useful in immunodiagnostic assays as it can be used for absolute quantitation of T4 levels in a system. Ourrently there are no standards for quantitating the T4 receptor.
The T4 receptor resides in three diverse chemical envi-ronments: the oxidizing, hydrophilic cell surface; the hydrophobic membrane: and the reducing, hydrophilic cytoplasm. These diverse environments would most like ly preclude the isolation of the receptor in its fully native state. sT4, which consists of only the extra cellular domain, is secreted as a soluble protein into the cell supernatant and its conformation appears to mimic the surface of the receptor surface domain.
Thus, sT4 ins suitable for detailed structural analysis, in particular for x-ray crystallography. Determination of the three-dimensional structure of sT4 alone or in a complexed form with other interactive molecules could provide a basis for the rational design of selective antagonists and agonists for sT4.
The various prophylaxis and immunization methods for AIDS provided by the present invention are based upon the abilities of the novel peptides, antibodies, and DNA molecules disclosed herein to form complexes with, or hybridize to, specific molecules and to invoke an i~unologica,l response effective for neutralizing the AIDS virus. These molecules, methods for their prepa ration, and methods of AIDS treatment will be better understood by reference to the following experiments and examples which are provided for purposes of illus '' ~ 1.3~~~~~~

tration and are not to be construed as in any way lim-iting the scope of the present invention, which is defined by the claims appended hereto.

a.~~o~oz Materials and Methods Cells and Antibodies Peripheral blood leukocytes isolated by Ficoll-Hypaque density gradient centrifugation were fractionated into sheep erythrocyte rosette-positive (E+) cells. T4+ and T8+ subset: within the E+ population were isolated by positive selection of T8-bearing cells with anti-T8 antibody anal human erythrocytes conjugated with affini ty-purified rabbit anti-mouse IgG (10). Cytofluoro-metric analysis of these subsets demonstrated that the T4+ cells were 95% T4+ and 2% T8+, whereas the T8+
cells were 95% T8+ and 2% T4+.
_ The Fro 2.2 T cell line (T3 , T4+, T8+, T11+) was de-rived from .an adult patient with undifferentiated acute leukemia. ;Jurkatt is T3 , T4+, T8+, T11 , RPMI 8402 is T3 , T4 , T'8 , T11+. OT-CLL is a chronic lymphocytic leukemia which is T3+, T4+, T8 , and T11+ (22j. The T4+ cell lines CEM and Molt 4 were obtained from the American Type Culture Collection. All leukemic T cell lines were continuously grown in RPMI 1640 medium con taining 5% :fetal calf serum. Transformed B cell lines C8, CP58 an<i CP94 were derived as previously described (23) .
Affinity-purified rabbit anti-mouse IgG was conjugated to human erythrocytes by the chromium chloride method (24) .

' -35-Cotransformation of L Cells and NIH 3T3 Cells Murine L tk aprt- cells were maintained in Dulbecco's modified Eagle's medium (DME) supplemented with 10%
calf serum (Gibco) and 50 micrograms/ml diaminopurine (DAP). L cells were plated out at a density of 5 x 104 cells per 10 cm dish, 1 day before transformation.
Calcium phasphate precipitates were prepared by the method of Graham and van der Eb (25), as modified by Wigler et al. (26), using 100 ng of pTK and 20 micro-grams of hj.gh molecular weight T cell or L cell DNA
pez dish. The L cells were placed under selection in DME with 10% calf serum, 15 micrograms/ml hypoxanthine, 1 microgram/ml aminopterin and 5 micrograms/ml thy-~5 midine (HAT medium (27)) on the following day. After 12-14 days of HAT selection, tk+ transformants were screened using the rosetting assay.
Murine NIH 3T3 cells were maintained in DME supplement-20 ed with 10% newborn calf serum (Gibco) . NIH 3T3 cells were plated out at a density of 5 x 104 cells per 10 cm dish, 2 days before transformation. A calcium phos-phate p;ecipitate was applied to the cells using 10 micrograms of carrier DNA and either 10 micrograms of 25 T4-pMV6tk/neo or 10 micrograms of T4-pVcos7 and 500 ng of pSV2neo. After 2 days, the cells were placed under selection in DME with 10% calf serum and 500 micro-grams/ml 6418 (Geneticin°; Gibco). Rosetting assays were performed on surviving colonies one week after 30 gr°~"~h in selective medium.
RosettinQ Asst After one rinse with phosphate-buffered saline (PBS), 35 the plates were incubated with 2.5 ml of tha purified J.

monoclonal antibody OKT'4A (1 mg/ml) diluted at 1/500 in PBS containing 5% fetal calf serum for 45 minutes at room temperature. Free antibody was removed from the plates with three gentle rinses in PBS. Six millili-tens of human erythrocytes conjugated with purified rabbit anti-mouse IgG antibody (2% v/v stock suspen-sion, diluted 1,/10 in PBS/5% fetal calf serum) were added and the plates were left at room temperature.
After 45 minutes, free erythrocytes were gently aspi-rated and PBS was added prior to inspection for ro-sette-positive colonies.
Cytofluorometric Analysis ~5 Adherent cells ware removed with 0.005 M EDTA in PBS
and washed once with PHS containing 1% bovine serum albumin (BSA) and 0.01% sodium azide (cytowash). Cells (5 x 106) i.n 0.1 ml were added to tubes with appropri-ate dilutions of OKT~4, OKTs8 or control antibodies.
20 The cell-antibody mixture was incubated for 45 minutes at 4'C and then washed twice in cytowash. Fluorescein isothiocyanate (FITC)- conjugated goat anti-mouse IgGJ+
A + M (Cappel) was added to the cells and incubated for 1 hour at 4'C. The cells were then washed three times in cytowash and resuspended in 0.5 ml of PBS with 0.01%
sodium azid,e. The cells were analyzed on a Becton Dickinson F~ACS IV Cell Sorter and the data was stored and plotted using a VAX 11/780 computer (Digital Equip-ment Co.) 13~~'~02 RNA and DNA Isolation Total RNA was isolated from cells by homogenation in 4 M guanidinium thiocyanate, followed by ultracentri-fugation through a 5.7 M CsCl cushion (28). Poly(A)+
selection was achieved by oligo(dT)-cellulose chroma-tography (Type 3, Collaborative Research) (29). High molecular weight genomic DNA was prepared as described by Wigler et al. (26).
cDNA and Genomic Libraries Double-stranded cDNA was synthesized from poly(A)+ RNA
derived from peripheral human T cells (20). After treatment with EcoRI methylase and T4 DNA polymerise, the double-stranded cDNA was cloned into the EcoRI site of agtl0 (30) using EcoRI linkers. The Charon 4 human genomic library was generously provided by Dr. Tom Maniatis (Harvard University) (31).
Synthesis o:f a Subtracted cDNA Probe 32p-labeled cDNA was synthesized from poly(A)~ RNA
derived from the primary transformant, LTD-4, as de-scribed by Davis et al. (32). After annealing the cDNA
to an excess of untransformed L cell poly(A)+ RNA (Rot = 3000), single-stranded sequences, which were enriched for human cDNAs, were isolated by hydroxyapatite chro matography (32). Prior to filter hybridization, the subtracted c:DNA probe was concentrated with sec-butanol and desalted on a G-50 Sephadex column equilibrated in TE.

-3$_ ~.~40'~02 Screening of cDNA and Genomic Libraries The peripheral human T cell library was plated on _E.
cola C600/HFL and the human genomic library was plated on E. cola LE392. Screening of duplicate filters was carried out: according to the standard procedure (33), with the hybridization performed in 50% formamide and 5x SSC at 42'C. In the screen of the cDNA library, 6 x 104 cpm of subtracted probe was applied per 137 mm nitrocellulose filter. Filters from the genomic li brary were hybridized to a nick-translated (34) cDNA
insert. The washes were performed at 68'C, with a final wash in 0.2 x SSC. Autoradiography was performed at -70'C in the presence of intensifying screens for 1-2 days.
DNA Sequencing Restriction fragments of pT4B were subcloned into the M13 vectors mpl8 and mpl9 (35). Sequencing reactions were performed using the dideoxy chain termination technique (:36). The sequencing strategy is depicted in Figure 3H..
Southern and Northern Hlot Hybridizations High molecular weight cellular DNAs were digested with 5 units of restriction nuclease per microgram of DNA
according to the manufacturer's recommendation (Hoeh-ringer Mannheim). Samples (10 micrograms) were sub-jected to electrophoresis on a 0.8% agarose gel. DNA
fragments were transferred to GeneScreen (New England Nuclear: (37)) and hybridized as described by Church and Gilbert i;38).

-- I~~O'~~2 RNA was run on a 0.8% agarose-formaldehyde gel (39) and transferred to GeneScreen. Northern hybridization was performed according to the procedures supplied by the manufacturer. Hoth Southern and Northern blots were hybridized. to nick-translated probes.
Synthesis and In Vitro Translation of SP6 RNA
The kb T4 cDNA was subcloned into the EcoRI site of pSP65 (Promega Biotec) and linearized with HindIII.
Transcription of linearized plasmid DNA (1 microgram) with SP6 polymerise in the absence of radiolabeled nucleotides was perfonaed as described (40), except that GpppG and unlabeled CTP were added to the tran-~5 scription buffer. One-tenth of the reaction mixture was translated in a wheat germ system (Bethesda Re-search Laboratories) containing L-[32S]-methionine (Amersham) and 1 micromolar S-adenosylmethionine. The in vitro translation products were subjected to SDS-20 Polyacrylamide electrophoresis under reducing condi-tions as described below.
Cell Labeling, Lectin ChromatoQrachy and Immunopre cioitation fills were grown for 12 hours in methionine-free DME
medium containing 10% dialyzed calf serum and 1 mCi of L-[32S]-methionine (Amersham) as previously described (41). The cells were solubilized in 10 mM Tris (pH
W 4)~ 150 mM NaCI (TBS) containing 0.5% Nonidet P-40 (Shell) and 0.2 mM phenylmethylsulfonyl fluoride (Sig-ma). The lysates were centrifuged for 1 hour at 100,000 x ~~, and the supernatants were subjected to lentil lectin chromatography (Pharmacia) according to the procedures of Hedo et al. (42) . Eluates were pre-~. ~ 1340'02 absorbed once with a mixture of control mouse ascites and protein A-Sepharose (Pharmacia) for 1 hour at 4'C
and twice with protein A-Sepharose alone for 1 hour at 4'C. Of each supernatant, 2.5 x 104 cpm were then mixed with 10 microliters monoclonal antibody (approxi mately 1 mg/ml) and protein A-Sepharose and incubated on a turntable overnight at 4' C. The beads were then washed four times with cold TBS containing 0.5~ NP-40 and 0.2t SDS and were resuspended in electrophoresis sample butler.
Gal Electrochoresis SDS-polyacrylamide gel electrophoresis was performed ~5 according to the procedure of Laemmli (43). The im-munoprecipatates and in vitro translation products were dissolved in sample buffer with or without 2-mer-captoethanol and then were applied to 10~ poly-acryl-amide gala.. Autoradiography was performed on Kodak XAR-5* film in the presence of intensifying screens (DuPont Chemical Company).
Costranaformation and Rosettina Assav 25 Mouse ~6-2 cells (44) were maintained in Dulbecco~s modilied Eagle's medium (DME) supplemented with 10~
calf serum (CS) (Gibco) . ~b-2 cells were plated out at a density of 5 x 105 cells per 10 cm dish, 2 days be-fore trans:~ormation. Calcium phosphate precipitates were prepared by the method of Graham and van der Eb (25) , as modified by Wigler et al. (27) . Precipitates were applied to the cells using to micrograms of carri-er DNA and either 10 micrograms of T4-pMV7 or 10 micro-grams o! T8-pMV7. After 2 days, the cells were placed 35 under selection in DME/10~ CS and 500 micrograma/ml * Trademark 6418 (Genet:icin~: Gibco).
Rosetting assays to identify T4+ or T8+ colonies were performed on surviving colonies 1 week after growth in selective medium. After one rinse with phosphate-buff eyed saline (PBS), the plates were incubated with 2.5 ml of the purified monoclonal antibody OICT~4A or OKTeg (img/ml; Ori:ho) diluted at 1/500 in PHS containing 5%
fetal calf serum (FCS) for 45 minutes at room tempera ture. Free antibody was removed from the plates with three gentle. rinses in PBS. 6 ml of human erythrocytes conjugated with purified rabbit anti-mouse IgG antibody (2% v/v stock suspension, diluted 1/10 in PBS/5% FCS) were added and the plates were left at room tempera-ture. After 45 minutes, free erythrocytes were gently aspirated and PBS was added prior to inspection. T4+
and T8+ ,~-2 clones were purified by colony isolation and characterized by flow cytometry and Northern blot analysis.
Recombinant Retrovirus Production and Infection T4+ and T8+ ,~-2 clones were isolated which produce recombinant retrovirus stocks with titers of 105 cfu/
ml. Viral stocks were prepared by adding 10 ml of fresh DME/10% CS to a near confluent monolayer of the T4.+ or T8+ ~~-2 clones. After 24 hours, the medium was removed and filtered through a 0.45 micrometer filter (Millipore). For infection, 5 x 105 cells were incu-bated with 2 ml of viral supernatant (or a dilution) in the presence of 8 micrograms/ml polybrene (Aldrich).
After 3 hours, 8 ml of fresh medium was added. 3 days after infection the cells were reseeded into DME/10% CS
containing 500 micrograms/ml 6418, grown for 2 weeks, scored for G4~18r colonies, and screened for surface T4 134~'~t72 or T8 expression using the in situ rosetting procedure or flow cytometry.
~b-2 culture supernatants were used to infect mouse ~-AM
cells as described above. T4+ or T8+ adherent trans-formants were purified by the in situ rosetting assay followed by colony isolation; T4+ or T8+ non-adherent transforma:nts were purified by fluorescence-activated cell sorting (FRCS). Non-adherent human lymphoid cell lines (HS82, RPI~I-T cells; Raji - B cells) and adherent epithelial cells (HeLa) were infected by co-cultivation with T4+ or T8+ ,~-AM clones (pretreated with 10 micro-grams/ml m.itomycin-C for 2 hours; Sigma) and were puri-fied.
Cell lines were selected for 6418 resistance at a con-centration of 1.5 mg/ml, except for HeLa cells which require 1 mg/ml, and fibroblasts which require 0.5 mg/ml. All cell cultures producing recombinant ampho-trophic viruses (4-AM) were maintained under P3 con-tainment conditions.
AIDS Virus The prototype LAV strain of HTLV-III/LAV was obtained from J.-C. Cherman (Institut Pastuer, Paris; (45)).
Virus inocula used in these studies were from the sec-ond to fifth passages of virus in our laboratory.
Inocula are cu7.ture supernatants from HTLV-III/LAV-infected, phytohemagglutinin (PHA)-stimulated peripher al lymphocytes which were harvested by sequential cen trifugatior~ (300 x g for 7 minutes followed by 1500 x g for 20 minutes), and were stored in liquid nitrogen.
For binding studies, virus was concentrated from cul ture supernatants, harvested as above, by ultracentri fugation at 90,000 x g for 90 minutes over a 15% cush-ion of Renagraffin (E.R. Squibb) in 0.01 M Tris, 0.15 M
NaCl, 1 mM EDTA, pH 8Ø
Anti-HTLV-III/LAV Reagents Serum with high levels of antibody to HTLV-III/LAV was obtained from a hamosexual man with chronic lymphadeno-pathy, and its specificity by immunofluorescence (46), Western blot analysis (47), and radioimmunoprecipi-tation (48) has been described. Portions of the IgG
fraction were coupled with fluorescein isothiocyanate (FITC; FIZ'C:protein ratio of 10.7 micrograms/ml), horseradish peroxidase (HPO; type VI: Sigma) and aga-rose as described (47, 49, 50, 51) . Conjugates of IgG
from a nonimmune serum were prepared in parallel.
Reverse Transcriptase Assay 20 Magnesium-dependent, particulate reverse transcriptase (RT) activity was measured with a template primer of (A)n(dT)12-1.8 (or (dA)n(dT)12-18 as the negative con-trol) in the: presence of 7.5 mM Mg2+ (52).
25 Immunofluorescence Detection of C to lasmic AIDS Virus Cultured ce:Lls (1 x 105 in 0.1 ml) were centrifuged onto glass slides (Shandon Cytocentrifuge), fixed in 95% ethanol and 5% acetic acid at -20'C for 30 minutes, and rehydra~ted with three 10 minute changes of PBS
(0.01 M P04, 0.15 M NaCl, pH 8.0). Slides were exposed to a 1/500 dilution of FITC-anti-HTLV-III/LAV (19 mi crograms/ml) for :30 minutes at room temperature. The slides were then washed (three changes, 10 minutes each) and mounted under a coverslip with 50% glycerol ......

in PBS. T'he slides were examined with an epi-illumi-nated LeitZ; Orthoplan microscope at 630 x power. Under these conditions, the FITC-anti-HTLV-III/LAV reagent is specific for HTLV-III/LAV. Uninfected PHA-stimulated cells, Epsi:ein Barr (EB) virus-infected B cell lines, an adenovir~us-infected cell line, several T cell lines, and HTLV-I and HTLV-II infected cell lines were not stained.
AIDS Virus Immunoassay (Antigen Capture Assay) This is a sandwich immunoassay that has been described in detail (47). Briefly, culture supernatant is added to microtiter plate wells coated with anti-HTLV-III/LAV
~5 IgG. After the plates are washed, bound virus antigen is detectedt with HPO-anti-HTLV-III/LAV. This assay, which is at least as sensitive as the RT assay, is negative with culture supernatants from PHA-stimulated lymphocytes from numerous donors, EB virus-infected B
20 cell lines, several T cell lines, polyclonal and cloned IL-2 dependent T cell lines, the myeloid line K562, as well as cell lines that harbor HTLV-I or HTLV-II. The cutoff OD49~~ for discriminating a positive from a nega-tive supernatant was determined in each run from the 25 mean plus 2 SD of at least 10 replicative determina-tions on control (uninfected cell culture) supernatants harvested at: the same time.
AIDS Virus Infectivity (ID-50) Assay The microcul.ture assay for the titration of infectious HTLV-III/LAV' has been described in detail (47). Brief-ly, PHA-stimulated lymphocytes or cell lines (2 x 106 cells/ml) are inoculated with serial 10-fold dilutions of virus inoculum and incubated for 18 hours at 37'C.

-45- 1~~~J~~2 Tha cells were then washed and plated in microculture (10 to 20 cultures per dilution: 1 x 105 CBIIS rlor culture in 0.25 ml medium). Every 4 days, 100 microli-ters of supernatant was removed and replaced with fresh medium. Supernatants were then assayed for viral anti-gen by the antigen capture assay as described above.
Infectious virus titer (ID-50) is defined as the recip-rocal of tlhe dilution at which 50~ of the cultures are positive for virus (47).
VSV Pseudotype Assay Vesicular stomatitis virus (VSV, Indiana strain, wild type) was propagated in cells producing the retrovirus ~5 required for the envelope pseudotype as described (53).
Hyperimmune neutralizing sheep anti-VSV serum was added to the harvested VSV to inactivate non-pseudotype virions. The pseudotype titers ranged between 104 and 105 PFU/ml. For the assay, 2 x 105 cells to be infect-ed with VSV pseudotypes were plated in 30 mm diameter tissue culture wells. HeLa, NIH 3T3, and L cells were naturally adherent: all other cells types were attached by pretreatment of the substratum with 50 micrograms/ml poly-L-lysine. After virus adsorption for 1 hour, the cells were washed and 106 mink CCL64 or bovine MDBK
calls were added to each well. These cells provide excellent plaques for secondary VSV infection but are resistant to infection by pseudotype virions. After allowing the plaque indicator cells to settle and spread (app.roximately 90 minutes), the monolayers were overlaid with agar medium. VSV plaques were counted 2 days after infection. Anti-T4A monoclonal antibody (1:20), ani_i-HTLV-III serum (1:10), or anti-HTLV-I
serum (1:10) were used to inhibit pseudotype plaque formation by pretreatment of cells 30 minutes before addition of pseudotypes as described by (54).
Syncytium Induction Assay 2 x 105 cells were co-cultivated with 2 x 104 H9 cells infected by and producing HTLV-III (55) in 10 mm diame ter wells" The cultures were incubated at 37~C and examined for syncytia formation after 18 hours as pre viously described (54, 56). Cells were five or more syncytia were scored as positive. Syncytium inhibition was assayed by adding anti-T4A monoclonal antibody (1:20) to the mixed cultures at the time of seeding.
~tofluorometric Analysis and AIDS Virus Binding The method has been described in detail (46). Briefly, cell surfa<:e T4 or T8 expression was detected by direct immunofluorescence with fluorescein-conjugated anti-T4A
or anti-T8 monoclonal antibodies (OKTs4A, OK'i'~8). The diluent/was.h buffer was 0.01 M P04, 0.15 M NaCl, pH
7.4, conta:Lning 0.1% bovine serum albumin, 2% v/v AB+
human serum, and 0.01% NaN3. All reagents were pre titered for optimal (saturating) binding. Cells (5 x 105) were incubated in a 25 microliter dilution of monoclonal antibody for 30 minutes at 4'C. The cells were washed; by centrifugation (300 x g for 7 minutes) , resuspended in 0.5 ml of 1% paraformaldehyde in saline, and analyzed with a fluorescence-activated cell sprter (FRCS IV, Hecton Dickinson). For HTLV-III/LAV binding, 5 x 105 cells were incubated with HTLV-III/LAV (500 ng in 10 microliters) for 30 minutes at 37'C. Washed cells were resuspended in 25 microliters of fluoresce in-conjugated anti-HTLV-III/LAV for 30 minutes at 4'C.
The cells were washed, resuspended in 1% paraformal dehyde, and analyzed by FACS as above. For inhibition of HTLV-II:C/LAV binding, cells were preincubated with anti-T4A o:r anti-T8 (20 ng in 20 microliters) for 30 minutes at 4'C followed by addition of HTLV-III/LAV
(500 ng in 10 microliters) for 30 minutes at 37'C. The cells were washed incubated with fluorescein-conjugated anti-HTLV-III/LAV, washed, resuspended in paraform-aldehyde, and analyzed by FACS as above.
Cell Surface Radioiodination, Immunooreci itation, and Gal Electrophoresis T4+ NIH 3T:~ transformants were surface radioiodinated by the lactoperoxidase technique (18) as follows: 4 x 107 cells were suspended in 1 ml of PBS containing 0.5 mM EIYrA, 2 mCi Na125I, and 20 micrograms lactoperoxi-dase. At times 0, 1, 5, 10, and 15 minutes, 10 micro-liters of 0.03% H202 were added. The reaction was carried out at 23°C and was stopped at 20 minutes by f centrifugati.ons in 50 volumes of cold PBS containing 10 20 ~ NaI. Labeled cells were split into 4 tubes and incubated, .as indicated, with HTLV-III/LAV (2 micro-grams in 20 microliters) for 30 minutes at 37'C. Sub-sequent washes and manipulations were performed at 0' to 4'C. Washed cells were lysed by adding 1 ml of 25 detergent lysing buffer (LB; 0.02 M Tris, 0.12 M NaCl, pH 8.0, containing 0.2 mM phenylethlsulfonylfluoride, 5 micrograms/ml aprotinin, 0.2 mM EGTA, 0.2 mM NaF, 0.2%
sodium deoxycholate, and 0.5% (v/v) Nonidet P-40).
Tubes were held on ice for 15 minutes, and nuclei were 30 removed by centifugation at 3000 x g for 20 minutes.
For absorpti.ons, Sepharose conjugates of human anti-HTLV-III/LAV IgG, human nonimmune IgG, anti-T4A, and anti-TS antibodies were prepared as described (48).
35 LYsates were preabsorbed with 200 microliters of Sepha-.

rose-nonimmuna human IgG for 1.5 hours with rotation, and then immunoprecipitated with 20 microliters of Sepharosa conjugates (as indicated) for 3 hours with rotation. Sepharose absorbents were washed 3 times:
once with L8: once with LB containing 0.5 M NaCl; and once with LB captaining 0.1% sodium dodecyl sulfate (SDS). Absorbed material was eluted at 65'C for 30 minutes with 20 microliters of sample buffer (0.01 M
Tris, pH 8.,0, containing 2% SDS, 5% 2-mercapto-ethanol (v/v), 25 micrograms bromphenol blue, and 10% glycerol (v/v). E:lectrophoresis was performed in a 3.3-20%
gradient p~olyacrylamide gel with a 3% stacking gel (57), and autoradiographs were developed with Kodak XAR-5 film.
Virus Inhibition Assay 2 x 105 T4+~ JM T cells were exposed to AIDS virus at 0 minutes. The inhibitors ammonium chloride (20 mM) or amantadine (20 mM) were added at various times during the course of virus infection (0 minutes, 30 minutes, and 60 minutes). After 6 hours, cells were washed and replated in fresh medium (RPMI/10%FCS). The effect of these agents on AIDS virus infection was determined 5 days post infection. The fraction of infected cells in the cultures expressing viral antigens was determined by immunof7.uorescence microscopy as described above (58) .
RNA Isolation and Northern Blot Hybridizations Total RNA was isolated from cells by homogenation in 4M
guanidinium thiocyanate, followed by ultracentrifu gation through a 5.7 M CsCl cushion (28). Poly(A)+
selection w,as achieved by oligo(dT)-cellulose chroma tography (Type 3, Collaborative Research) (29).
RNA was electrophoresed through a 1% agarose-formald-hyde gel (39) and transferred onto Hybond (Amersham).
Northern blot hybridization was performed according to the procedures supplied by the manufacturer. Probes were nick-translated to a specific activity of 0.5-1 x 109 cpm/microgram with a32P-labeled deoxynucleotide triphosphavtes (59).

r--RESULTS
Isolation of a T4 cDNA
The strategy used to isolate a T4 cDNA initially in-volved constructing L cell transformants that express T4 on their- surface. cDNA synthesized from the mRNA of a T4+ transformed. fibroblast was enriched by substrac-tive hybridization and used as a probe to isolate a cDNA encoding T4 from a cDNA library made from the mRNA
of peripheral T lymphocytes. The identity of T4+ cDNA
clones was determined by Northern and Southern blot analyses, and ultimately by the ability of these clones to transfer the T4+ phenotype to recipient cells.
~5 Similar te~~hniques have previously been employed to isolate the gene encoding the T8 protein (20).
Mouse L cells deficient in thymidine kinase (tk) were cotransformed with genomic DNA from the T cell leuke-20 mic cell line I~IUUT-102 along with the tk-containing plasmid, pTK (25, 26). tk+ L cell transformants ex-pressing T cell surface proteins were identified by an in situ rosetting assay. tk+ colonies were exposed to mouse monoclonal antibodies directed against T4 and were then incubated with red blood cells coupled with rabbit anti-mouse immunoglobulin. T4+ transformants ara visibly red by virtue of their specific associa tion with red blood cells. In this manner, one primary T4+ transformant, LTD-4, was obtained: The expression of the T4 molecule by this clone was independently verified by cytofluorometric analysis (Figure 1).
The mRNA papulatian of the T4+ transformant, LTD-4, should differ from that of an untransformed L cell only in the expression. of newly transformed genes. These ~ ~-~ ~M~~2 sequences were enriched for by annealing highly radio-active cDN~A prepared from poly(A)+ RNA of the T4+
transformant with a vast excess of RNA from an untrans-formed L cell (32, 60). cDNA incapable of hybridiz-ing, even at high Rot values, was isolated by hydroxy-apatite chromatography and used to screen a human pe-ripheral T cell cDNA library constructed in the lambda cloning vector gtl0. Four weakly hybridizing plaques were identified, plaque-purified and analyzed for the presence of T4 sequences.
To determine whether any of these clones encoded T4, Northern blot analyses were initially performed with RNA from TS4+ and T4 peripheral T cells, leukemias, ~5 thymocytes, L cell transformants and nonlymphoid cells (Figure 2). One of the four clones hybridized to an RNA present only in T4+ cells. This clone detects a 3 kb RNA present in the T4+ transformant, LTD-4, which is also present: in a population of T4+ peripheral lympho-20 cytes, a variety of T4+ leukemic cell lines, and thymo-cytes. No hybridization was observed with RNA from untransformed fibroblasts, T4 peripheral lymphocytes, HeLa cells, or human neuroblastoma cells.
Tha pattern of expression of RNA detected by this clone is consistent with the possibility that it encodes T4.
However, this cDNA, is only 0.6 kb in length but hybrid izes to a 3 kb mRNA. Therefore, the human peripheral T
cell cDNA library was rescreened and one clone (pT4B) was obtained which contained a 3 kb insert, close in size to that of the mature messenger RNA. Restriction maps of this clone are shown in Figures 3A and 3B.

Genomic Blot Analysis Southern blot experiments (37) were next performed to demonstrate that the isolated cDNA clone hybridized with DNA from the T4+ transformant as well as human DNA, but not with untransformed mouse L cell DNA (Fig-ure 4). Genomic DNA from a variety of human cells reveals a set of five hybridizing fragments after cleavage with the enzyme BamHI. As expected, T4 se-quences can be detected in the transformant LTD-4, but not in unt:ransfarmed L cell DNA. The BamHI fragment closest to the 3' end of the gene (6.6 kb) is not present in LTD-~, presumably as a consequence of the integration event. Moreover, no gross rearrangements ~5 are apparent at this coarse level of analysis when comparing L~NA from lymphoid and nonlymphoid cells. The sum of the moleaular_weights of the hybridizing frag ments is 33 kb, suggesting that the T4 gene is quite large. A complete set of genomic clones spanning this region was obtained (see below) and the BamHI fragments were ordered by restriction analysis of these clones (Figure 3A), confirming that the gene is large and must contain introns of significant lengths.
Expression of the T4 cDNA in Transformed Mouse Fibro-blasts Further evidence that the isolated cDNA encodes T4 would be provided if this clone could convert fibro blasts to t;he T4~ phenotype after transformation. The T4 gene in chromosomal DNA is large and spans several genomic clones. Therefore, the cDNA clone was intro duced into two retroviral expression vectors, pVcos7 and pMV6kt/neo, which contain the Moloney murine leuke mia virus long terminal repeats (LTRs) flanking a sin gle EcoRI cloning site (Figure 3C). The 5'-LTR pro-motes transcription through the cloning site and the 3'-LTR contains sequences necessary for cleavage and polyadenylation. The vector pMV6tk/neo also contains the tk promoter fused to the coding region of the neo-mycin phosphotransferase gene. The construct employing pVcos7 requires transformation with an unlinked selec-table marker, whereas pMV6tk/neo carries the neomycin resistance marker, which permits linked cotransformat-ion. Neo+ colonies of NIH 3T3 cells obtained after transformation were selected by their ability to grow in media containing the neomycin analogue 6418, and were screened using the rosetting procedure to detect the expression of T4 on the cell surface. Approximate ly 50% of the 6418 colonies obtained with pVcos7 and 75% of the colonies obtained with pMV6tk/neo were posi-tive for T4 in this assay. Rosette-positive colonies were further analyzed by cytofluorometry to confirm that T4 is. expressed on the transformed cell surface (Figure 1).
Metabolic protein labeling experiments were performed which demonstrate that the T4+ transformed fibroblast and the T :lymphocyte express a T4 protein of identical molecular weight. Untransformed NIH 3T3 cells, T4+
transformants and T lymphocytes were labeled for 12 hours in the presence of L-[35S]-methionine (41). The cells were detergent solubilized and the lysate was passed over lentil lectin columns to enrich for glycoproteins (42). The bound glycoprotein fraction was eluted and immunoprecipitated with monoclonal anti bodies directed against T4 (Figure 5). Under reducing conditions, a glycoprotein migrating at a relative molecular mass of 55 kd is detected in extracts from T
lymphocytes and two independent T4+ transfonaants.

l~~~v~~

This protein is not detected in control 3T3 fibro-blasts. Under nonreducing conditions, a 51 kd glyco-protein is immunoprecipitated with anti-T4 in T cells and in the transformed fibroblasts.
These experiments demonstrate that the transformants express a 55 kd glycoprotein immunoprecipitated with anti-T4 which is identical in size to that expressed on the surfaces of T lymphocytes. Thus, Northern and Southern analyses using the isolated cDNA, taken to-gether with the ability of this cDNA to confer the T4+
phenotype to mouse fibroblasts, indicate that the en-tire coding sequence of the T cell surface protein T4 had been cloned.
Nucleotide Secruence of the T4 cDNA and the Deduced Protein Sequence The complete nucleotide sequence of the T4 coding re-gion was determined by sequencing both strands of the 3 kb cDNA insert using the dideoxy termination method (35, 36). The complete nucleotide sequence and the predicted protein sequence are shown in Figure 6. The longest open reading frame begins at position 76 with a methionine c;odon surrounded by the initiation consen-sus sequence PurNNATGPur (61). This reading frame extends 13744 nucleotides, encoding a polypeptide con-taining 458 amino acids. The contiguity of this read-ing frame was confirmed by inserting this cDNA into the ~A expression vector pSP6 (40). RNA synthesized from this vector,, when translated in vitro, directs the synthesis of an unmodified 51 kd protein, the precise molecular weight predicted from the nucleotide sequence (Figure 7).

T4 is comprised of a leader sequence, four tandem vari-able-joining (VJ)-like regions, and a membrane-spanning domain each sharing homology with corresponding re-gions of different members of the immunoglobulin gene family (62, 63) (Figures 6 and 8). A stretch of hydro-phobic residues, corresponding to a leader peptide predicted by a Kyte-Dolittle (64) hydropathicity plot, immediately follows the initiation codon. Although the exact position at which the native T4 protein is pro-cessed cannot be determined, it is contemplated that cleavage occurs just after the threonine at positions -1 based on known cleavage patterns (65). Therefore, the signal. peptide contains 23 amino acids and the processed 't4 protein consists of 435 residues.
Residues 1-94 of the mature protein share both amino acid and ;structural homology with the immunoglobulin light chain variable domain (Figure 9A). The overall homology o~f this domain with immunoglobulin variable regions is. 32%. Sequence comparison between the V
regions of light chain immunoglobulins and the N-ter minal V-like region (V1) of T4 demonstrates that eight out of 14 invariant residues are conserved (66). This domain contains two cysteine residues, separated by 67 amino acids, whose positions and spacing are analogous to that found in light chain immunoglobulins and relat ed molecules (67). These cysteines may be capable of forming the conserved intrastrand disulphide bond characteristic of V domains. This suggestion is sup ported by our observation that T4 migrates more rapidly under nonreducing conditions than under reducing condi-tions, con:aistent with the formation of at least one intrastrand. linkage (Figure 5, lanes a and f).

1~~:~'~~~

Aside from homologies at the level of individual amino acids, the V1 domain of T4 shares structural features with immu:noglobulin variable regions. Immunoglobulin variable and constant domains fold in a characteristic pattern in which a series of antiparallel ~-strands fold to form two ~9-sheets (67, 68). These ~-sheets are held together both by a disulphide bridge and by characteristic hydrophobic interactions. To determine how the predicted secondary structure of the V-like 0 domain of T4 compares with the structure of the V do-mains of light chain immunoglobulins, two-dimensional structural alignments were performed. Also, a plot of probable ~9-strands and ,B-turns in these sequences using the empirically derived algorithm of Chou and ~5 Fasman (69~) was obtained. These analyses suggest the presence of seven ~-strands within the V-like domain of T4 which closely match those found in the immuno-globulin V' domain (Figure 9A). The two conserved cys-teines of T4 are found within ~-strands B and F, matching exactly the positions of the cysteines in the V region known to form the conserved disulphide bond in immunoglobulin. A tryptophan residue lies 12 amino acids downstream of the first cysteine and a tyrosine residue is. situated two amino acids before the second 25 cysteine. These residues are highly characteristic o! ~9-strands C and F, respectively in light chain V
regions. In addition, an aspartate residue is found six amino acids before the second cysteine, and an arginine residue lies at the base of p-strand D. These charged residues are highly characteristic of V domains (67). Finally, patches of alternating hydrophobic residues a,re present throughout the ~-strands, which strengthen the interaction of the two ~B-sheets.

1~ ~:~°~~~

Tha V1 domain of T4 is followed by a stretch of amino acid residues bearing significant homology to the join-ing (J)regi.ons of immunoglobulins and T cell antigen receptors. In Figure 9B, this J-like region of T4 is aligned with the consensus joining sequences of immuno-globulin light chains and the two chains of the T cell antigen receptor. This J-like region is followed by a 265 amino acid stretch which may be structurally divid-ed into thrcae additional VJ-like domains with statisti-cally significant sequence and structural homology to prototype immunoglobulin VJ regions (Figures 6 and 8).
Additionally, this sequence contains two potential N-linked glycosylation sites (Asn-Leu-Thr: Figure 6).
~5 The extracellular domain is followed by a putative transmembrane sequence, predicted by a hydropathicity plot (64), which contains only hydrophobic and neutral amino acid ~~esidues. This segment bears striking ho-mology to t:he transmembrane exon of the ~9-chains of 20 class II major histocompatibility proteins (Figure 9C).
Alignment of the transmembrane regions of T4 and I~IC
class II p-chains reveals 48% homology without gaps.
Following the membrane-spanning segment, a highly charged sequence of 40 amino acids comprise the cyto-25 plasmic doma:Ln (Figures 6 and 8).
Tha T4 Gene: Chromosomal Location and Intron-Exon Positions The T4 cDNA was used to determine the chromosomal loca-tion of the T4 gene by analyzing its segregation pat tern in a panel of mouse-human somatic cell hybrids and by in situ h~rbridization to human metaphase chromosomes (lol). Genomic blot experiments and in situ hybridiza tion indicate that the T4 gene resides on the short arm of human chromosome 12, between regions 12p12 and l2pter.
A set of overlapping genomic clones spanning the T4 gene was obtained by screening human genomic libraries constructed in the lambda cloning vectors Charon 4 and EMLB-3 (31) with a radiolabeled pT4B cDNA insert (70).
Characterization. of these clones by both restriction and Southearn blot analyses indicated that they con-twined the entire T4 coding sequence. The complete intron-exan organization of the T4r~gene was then deter-mined by ;sequencing specific fragments of the genomic clones us~lng the dideoxy termination procedure (35, 36) .
The T4 gene is comprised of 9 exons split by 8 introns as shown in Figures 8 and 10. The first exon contains the 5'-untoanslated region and the leader segment. The first variable-like domain, V1, is split by a large intron located at nucleotide position 289 (Figure 6).
Therefore, the V1J1 domain is encoded by the second and third exon:: and 'the V2J2, V3J3, V4J4, and transmembrane (TM) domains are each encoded by separate exons (axons 4-7). The cytoplasmic domain (CYT) is split by an intron and the last portion of the cytoplasmic domain and the 3'-~untranslated region are encoded by the ninth axon.
The Construction of T4+ and T8+ Transformed Cells The experimental approach used to study the role of T4 in AIDS virus infection initially involved the intro-duction of the T~ gene into T4 cell lines incapable of supporting viral infection. The transformed cells were than tested. for susceptibility to AIDS virus, followed .--.

by studies on the mechanism by which T4 mediates viral infection.
A full length cDNA clone encoding the surface protein T4 was subc:loned into the retroviral expression vector, pMV7. The expression vector, pMV7 (Figure 11A), con-tains two directly repeated Moloney murine sarcoma virus long terminal repeats (LTRs) which flank a single EcoRI cloning site. The 5'-LTR constitutively promotes transcription through the cloning site, whereas the 3'-LTR provides sequences necessary for cleavage and poly-adenylation of the RNA. In addition, pMV7 contains the herpesvirus thymidine kinase promoter (tk) fused to the coding regj.on of the bacterial neomycin phosphotrans-ferase genes (neo), a dominant selectable marker, per-mitting linked cotransformation and infection.
T4-pMV7 was introduced into ~-2 and ~-AM cells,. NIH
3T3 cell lines containing defective ecotropic and am-photropic proviruses, respectively (Figure 11B) (44,59). Both cell lines are incapable of encapsidating endogenous viral RNA but can provide all obligate traps viral funct:fons. Stable transfection of these cell lines with T4-pMV7 results in the production of recom-25 binant retroviral stocks encoding T4 which are free of helper viru:a. These pure viral stocks can then be used to efficiently introduce T4 sequences into both mouse and human cealls without the production of retrovirus by the target cell.
Briefly, T4.-pMV7 DNA was introduced into ,~-2 cells using the procedure of DNA-mediated gene transfer (Fig-ure 11B) (2fi, 27). Neo+ positive colonies were select-ed by their ability to grow in media containing the neomycin analog 6418 (Geneticin~) and screened for the expression of T4 on the cell surface using an in s-rosetting assay (20, 70). Colonies of transfected ~6-2 cells expressing T4 were then identified which produce recombinant retrovirus in titers of 105 cfu/ml.
T4+ ~-2 clones were then used to generate retroviruses capable o:~ infecting mouse ,~-AM cells. T4 express-ing ~-AM clones were isolated which yield recombinant retroviral titers of 104 cfu/ml. T4+ human transfor-mants were generated by co-cultivation of cells with mitomycin-~~ treated or ~-AM clones (Figure 11B). T4+
transformants were subsequently analyzed by Northern blot analysis and flow cytometry to confirm that T4 is expressed .and is present on the cell surface. Control cell lines expressing the surface protein T8 were con-~5 structed in an analogous manner.
T4 is Essential for AIDS Virus Infection To initially determine whether the presence of the T4 protein on the surface of a human lymphocyte is suffi-cient to :render the cell susceptible to AIDS virus infection, transformants of the primitive T cell leuke-mic line, HSB2 (71), which expresses only the early T
lymphocyte proteins T1 and T11 on its surface, were 25 constructedl. HS82 expresses neither T4 nor T8, nor does it express the T cell antigen receptor or the associated complex of T3 proteins. Transformants of HSB2 which express either the T4 or T8 proteins on the cell surface were selected and used to determine the susceptibility of these cell lines to AIDS virus infec tion. Several different experimental approaches were employed to assess AIDS virus infection, including expression of reverse transcriptase (52), expression of virus in t:he cytoplasm of the cell by immunofluores cence microscopy (46), detection of viral antigens in .... ~.3~~r1 ~?

the culture supernatant using an immunoassay (47), as well as production of infectious virions by supernate subculture with phytohemagglutinin (PHA)-stimulated peripheral lymphocytes (46). Using these assays, evi-deuce of AIDS virus infection of the HSB2 cell line was not observed (Table I).

en~~~.. ~ Ci . + I t + Q t + t t +
ctt N y ~ ~ tp ~ ~ ..~ ~ > v .., ~ ~r w to v ... ~ + 1 I + t t + I ~ + U C ~
C H n ~ W O y H e~~ ip C C
y t~.~ ..1 ~ v > .~'~ y . ' 'C ..~, ~ 'C
1 O ~ + 1 1 + i t + t t + ~ ,O > > ~ ~.0~~
~ ~ 'O ~ O
O C ~ ~ v vN O V
+ I I i0 U ~ v O
+ ~ t + I t + ~ ~ ~ ~ C
w '' Cry ' CJ > ~~ ~ . G
Vj 'C 1~ v C Z
~' p v N to ~r ~.
1 t + ~ 1 + I I + 4 c~oc~ ~O ~ E
~ y ~sr CD
.~ W ~ NU ~ .~ G4 V 3 ~1 m .O
~ ~ a y ~~ N ~.1 ~0 + + t t + ~ t + ~o 'v .~ v o t I t + ~ y v it 'C
a w o; ° c v ..:
44~J ~ v ~ °: w . ~ ,.) ,~~ ~ O ie1 y C M .. 'p y Q' wo .~ Q ~ ~ am ~n 3 '~ ~ C ~f v ~ C ~ ~ ,gyp ~ ~ ~ . >) ro ~ '~
a p ,D ip G0 w O
~~r ~y ~ v Z 'O ~ O pp ~.
V7 voU Z R O y ~ +
..
.) ~ + 1 ~ ~ ~ > U
H t p" ~» ~ it ~'' O v ~yw c~a~a~ '~.oC ~w > Hw ~.~-~~P~~2 In addition, it has been previously demonstrated that extensive cell fusion occurs when uninfected human cells bearing receptors for AIDS virus are co-cultivat-ed with cells producing AIDS virus (54). In this as-say, there is no induction of syncytia when HS82 cells are mixed with AIDS virus-producing H9 cells (Table I), although abundant syncytia are formed With HTLV-I and HTLV-II producing cells (data not shown).
Finally, viral entry.~was tested for using pseudotypes of vesicular stomatitis virus (VSV) bearing the enve-lope glycoproteins of the AIDS virus (Table I) (53, 54). When cells infected with AIDS virus are super-infected with VSV, a proportion of the progeny VSV
assemble sufficient AIDS virus envelope glycoprotein to resist neutralization by hyperimmune anti-VSV serum.
The host range of these VSV (AIDS) pseudotype virions is restrict~sd to cells expressing receptors specific to the AIDS virus. Following penetration of the cell and uncoating o:E the virion, the transcapsidated VSV genome replicates ito produce non-pseudotype particles. During the secondary infection, progeny VSV released from infected cells penetrate and destroy neighboring indi-cator cells resistant to VSV (AIDS) pseudotype infec-tion (mink CCL64 or bovine MDBK cells), resulting in the formation of VSV plaques which are then scored.
Thus, infeci:ion with VSV (AIDS) pseudotypes provides a quantitiative cytopathic plaque assay for viral entry (54). In this assay, no plaques over background were observed when HSH2 cells were exposed to VSV (AIDS) pseudotypes (Table I). In control experiments with pseudotypes of VSV RNA encapsidated in an HTLV-I enve-lope (VSV (HTLV-I)), numerous plaques were observed, demonstrating that the HSH2 cell, which bears HTLV-I

1~~~'~~~

receptors, is capable of replicating VSV efficiently.
These observations demonstrate that the VSV genome encapsidated in an AIDS virus envelope is incapable of entering HSB2 cells.
Whether th.e introduction of a functional T4 cDNA into HSH2 would. render this cell susceptible to AIDS virus infection was next studied (Table I). Exposure of HSH2-T4+ transformants to AIDS virus results in a pro-ductive viral infection as determined by expression of reverse t:canscriptase activity (52), expression of virus in t:he cytoplasm of the cell by immunofluores-cence microscopy (46), detection of viral antigen in the culture supernatant using an immunoassay (47), as ~5 well as the production of infectious virus by supernate subculture with PHA-stimulated lymphocytes (Table I) (46). Control HS82-T8+ cells were consistently nega-tive in each of the assays.
In addition, the efficiency with which different T4+ T
cells are :infected with AIDS virus was also examined.
HSB2-T4+ and HSB2-T8+ transformants, the naturally-isolated T4+ T cell line CEM, as well as PHA-stimulated peripheral lymphocytes were exposed to serial 10-fold 25 dilutions c~f AIDS virus, washed, and plated in micro culture. The frequency of infected cultures was then determined using an immunoassay 12 days after exposure to virus (Figure 12) (47). In this manner, the titer of AIDS virus required to infect 50% of the exposed cul tures (ID-50) was defined. The ID-50 of PHA-stimulated peripheral lymphocytes is 2-3 orders of magnitude greater than that observed for either naturally-isolat-ed or trans;fonaed T4+ cell lines. The efficiency of infection of HSB2-T4+ cells is about 10 fold higher than that observed for the naturally-isolated T4+ T

l~~Or~~

cell line C7~M (Figure 12). Control HSB2-T8+ cells are not susceptible to infection even at the highest virus titers examined.
The ability of HSB2-T4+ cells to support both syncytia formation and the replication of VSV (AIDS) pseudotypes was also studied. when HSB2-T4+ cells are co-cultivat-ed with AID~i virus producing H9 cells, syncytia forma-tion is readily observed within 18 hours (Tables I and II). Moreover, syncytium induction is abolished by pretreating cultures with anti-T4A monoclonal antibody (Table II). Finally, when HSB2-T4+ cells are exposed to VSV (AIDS) pseudotypes, infectious VSV particles are produced which destroy neighboring indicator cells ~5 (Tables I and III). Furthermore, plaque formation is inhibited by pretreatment with either anti-AIDS virus antibody or anti-T4A monoclonal antibody (Table III).
Control HSB~!-T8+ cells are consistently negative in each of the seven assays employed to detect AIDS virus 20 infection (Tables I, II, and III). These observations provide genetic evidence that in an immature human T
lymphocyte, the mere presence of the T4 protein pro-vides an essential function required for AIDS virus infection.

m:~~~~z Table II
Ir~trtion of S~rn~yt.ia in T1+ Min '~ansloaronts S~CN
H9~ Fi9/~
a JM (TI+) 8166(T~+) ~
IiSB~ ~
H582-T8+ ~
~H2~T4+
Ra7i ~
Raj i-rT8+ ~ ' Raji~TI+ ,.
H~'T8+ I~
Hela-T4+
2 x 105 cells wee co-cultivated with 2 x 10 AIDS virus-producing H9 cells (H9/AIDS) and incubated at 37oC, The cultures were examined for syncytia formation after 18 hours. The resin is are expressed as the approximate percentage of nuclei cor;tained within syncytia: - (no syncytia); ++ (25%); +++ (50%) +++++(g0%);ND (not determined) .
Syncyti um inhibition was assayed by adding anti-T4A
monoclonal antibody (aT4A; 1:20) to the mixed cultures at the time of seeing. The natuarlly-i sol aced T4 T cel l 1 ine s JM and 8166 served as positive controls in these studies.

'~ ~ 1~~40'~02 Table III
y5V Pseudotype Cytop~athic Pla4ue Assay cn T~+ and T8~ fitnmn Transformants V5y PSSC1D0~'Yp~ ~g (?'FZJ/ntl) FitI~1 C~~L.S V5y ( HTLV-I V5~1 ~
) (AZ 1 + LV-I + aAI D + aT~
aHT

CES'!IT+) 20, 000 50 42, 000 50 200 ~ 0 H582~1'8~ 10, 000 50 0 HS82~T4~ 12,080 50 1,000 100 300 Raji~f8+ 5,000 I~ 0 Raji~T4 5,000 50 1,500 25 150 10,000 :m 0 ~~4 10,000 50 17,000 50 200 2 x 105 cells were incubated with V SV
(ADS) pseudotypes (53, 54) for 1 hour at 37 C. The cell s were then washed and 1 x 106 mink CCL64 or bovine MDHK plaque indicator cells, permissive to VSV
infection but resistant to vSV (AIDS) , were adc9ed to each well. The cultures were theen overl aid w ith aga r medi um and scored for VSV plaques two days post i of ect i on. Ant i-T4 A monocl onal ant i body (tiT4At 1:20) or anti-AIDS virus serum (aAIDS; 1:10) were used to inhibit VSV
(AIDS) pseudotype plaque formation be pretreatment of cells 30 minutes before exposure to pseudotypes (54) . VSV (HTLV-I) pseudotypes, which plate on a wide variety of human cell types (54), were used as controls in these experiments.
Anti-HTLtJ=I serum ( 1:10) was used to block VSV (HTLV-I) pseudotype plaque formation.
The results are expressed as PFU/ml; ND
(not determined) .

AIDS Virus :Cnfection Is Not Restricted to T Lymphocytes A functiona:L T4 cDNA was introduced into two human non-T cell lines: HeLa, an epithelial cell line derived from a cervical carcinoma (72), and Raji, a B lympho-blastoid cell line derived from a patient with Burkitt~s lymphoma (73) (Figure 11B). Prior to retro-virus-mediated gene transfer, these~cell lines do not express surface T4 protein or T4 mRNA, nor are they susceptible to AIDS virus infection (Table I). In addition, the parental cell lines do not support the induction oi.° syncytium nor the plating of VSV (AIDS) pseudotypes (Tables I, II and III).
In contrast,. T4+ Raji and HeLa transformants support AIDS virus infection by all of the criteria previously described ('fable I). The efficiency with which Raji T4+ cells ca.n be infected with AIDS virus approximates that of HSB2-T4+ cells and is about 10 fold higher than the efficiency of infection of the naturally-isolated T4+ T cell line CEM (Figure 12). Moreover, upon co cultivation with AIDS virus-producing H9 cells, Raji T4+ and HeLa-T4+ cells support the induction of syncy tia which is abolished by pretreating cultures with anti-T4A monoclonal antibody (Tables I and II: Figure 13). In addition, exposure of these cells to VSV
(AIDS) pseudotypes results in the production of infec-tious VSV and the formation of plaques which are inhib-ited by pretreatment with anti-AIDS virus antibody or anti-T4A monoclonal antibody (Tables I and III). Con-trol Raji-T8+ and HeLa-T8+ transfonaants are consis-tently negative in each of these assays (Tables I, II, and III).

~~~o~o~

Therefore, the introduction of a functional T4 gene into either human T lymphocytes, B lymphocytes, or epithelial cells is sufficient to render such cells susceptible to AIDS virus infection. Taken together, these observations indicate that the T4+ T cell tropism observed in vivo is a consequence of the restricted expression of the T4 molecule and not the nature of the cell type i.n which it is expressed.
AIDS Virus Binds to Surface T4 Protein The previous experiments provide genetic evidence that T4 expression is required for AIDS virus infection but do not provide information on the role of this molecule in the Viral life cycle. The observation that surface ~5 expression of T4 is necessary for AIDS virus infection suggests that T4 is the AIDS virus receptor. Cyto-fluorometry was therefore used to examine the binding of AIDS Vi~:vs to the surfaces of T4+ and T8+ trans-fonaed human cells (Table I: Figure 14). HSH2, Raji, and HeLa cells, and the T4+ or T8+ transformants, were incubated with AIDS virus. Following viral absorption, the cells were washed, exposed to fluorescein-conjugat-ed anti-AIDS virus antibody, and analyzed by flow cyto-matry. This assay indicated that the AIDS virus binds efliciently and specifically to the human transformants expressing surface T4, but not to the T4 parental cells nor to the T8+ transformants (Figure 14, column B: Table I). The binding of AIDS virus to the T4+
cells is abolished by preincubation with anti-T4A
monoclonal antibody but not by preincubation with anti T8 monoclonal antibody (Figure 14, column C). More over, when T4+ transformed cells are exposed to AIDS
virus, the T4 glycoprotein coprecipitates with the viral envelope glycoprotein , suggesting a direct phys ical association between these molecules (data not shown). These results indicate that the AIDS virus binds to the T4 molecule on the cell surface and that this binding is independent of other T cell-specific proteins since binding occurs to all T4+ cell types examined.
Previous studies have described two distinct pathways of entry for enveloped viruses (74, 75, 76, 77). Some viruses fuse directly with the plasma membrane, releas ing their nucleocapsids into the cytoplasm, whereas others are internalized by receptor-mediated endocyto sis. The acidic environment of the endosome then fa cilitates fusion of the viral envelope with the limit ing membrane of the vacuole. Infection by viruses which enter cells via the endocytic pathway can be inhibited b:Y treating cells with agents such 'as weak bases which deacidify the endosome (58, 78, 79, 80), In the press:nce of ammonium chloride, fusion is blocked in the endc~some but lysosomal degradation still pro ceeds at a reduced rate (80).
The effect o~f ammonium chloride on AIDS virus infection of the T4+ '.C cell line JM was therefore examined. In the absence of ammonium chloride, over 50% of JM cells exposed to ,IDS virus express viral antigens five days after infection as determined by immunofluorescence microscopy. If JM cells are exposed to ammonium chlo ride (for 6 hours) either at the time of addition of virus or witlhin 30 minutes after the addition of virus, greater than 95% inhibition of viral infection was observed. however, if cells were treated with ammonium chloride one hour after the addition of virus, no inhi bition of infection was observed, a finding consistent with the kinetics of viral entry described for other ,,,.. , viruses which enter cells via receptor-mediated endocy-tosis. Finally, the ammonium chloride effect was com-pletely reversible. Cells exposed to ammonium chlo-ride for one hour, and then washed free of the compound and exposed to AIDS virus, supported control levels of viral infection. These results are consistent with previous observations that upon removal of ammonium chloride, th~a pH of the endosome returns to the origi-nal low values within 1-2 minutes (78, 80). Similar results with amantadine, a compound which deacidifies the endosome) were abtained.
These results are consistent with a mechanism of viral entry which involves endocytosis of the T4-AIDS virus complex and l.ow pH-induced fusion of the viral envelope with the limiting membrane of the endosome, releasing the viral nucleocapsid into the cytoplasm of the cell.
T4 mRNA is Expressed in the Brain In addition to the disruption of the cellular immune system, AIDS is frequently accompanied by central ner-vous system (CNS) disorders which are thought to be the consequence of the direct infection of brain cells by the AIDS virus (81). It was therefore of interest to determine whether T4 is expressed in cells within the CNS, thereby providing an explanation for the neurotro pic properties of the virus. Northern blot analyses of RNA prepared from both human and mouse brains were performed to determine whether T4 mRNA sequences are expressed in the CNS (Figure 15). Poly(A)+ RNA derived from human cerebral cortex contains two distinct T4 mRNAs with molecular weights of approximately 3 and 1.8 kb (Figure 15~A) . The weaker 3 kb RNA is identical in size to the mRNA expressed by two T4+ leukemic cell lines, U93~i (monocytic cell line) and Jurkat (T cell line), as well as by peripheral T lymphocytes. The smaller, more abundant 1.8 kb mRNA absent from T lym-phocytes could result from alternative splicing or alternative 5' or 3' termini.
A more careful analysis of the localization of T4 mRNA
was performed by isolating poly(A)+ RNA from specific regions of 'the mouse brain (Figure 15B). Hybridization with radiolabeled cDNA encoding the murine homologue of T4, L3T4, r~aveals an intense 2.2 kb mRNA in mouse fore brain which is absent from hindbrain samples. The 2.2 kb L3T4 mRNA is detectable in the cortex, hypothalamus, and is most: abundant in the striatum, but is absent from the cerebellum, brain stem, or spinal cord (data not shown). This 2.2 kb mRNA detected in the CNS is approximately 1 kb smaller than the 3.2 kb mRNA encod ing L3T4 in thymocytes (Figure 15B). These results indicate that the neurotropism displayed by the AIDs virus is likely to be the result of surface expression of the T4 molecule on brain cells. The level of mRNA
detected in forebrain is about 1/30th the level in thymocytes. This may reflect low level expression by a large number of cells or higher levels of expression by a small sub;population of cells. It is not known at present whether T~ is expressed by neurons or support-ing cells. The presence of a variant transcript in the CNS, however, makes it unlikely that the T4 mRNA in brain is expressed by the rare invading T lymphocyte.
Discussion The segregation of T4 and T8 with functionally distinct subsets of T cells suggests that these molecules may be important in the interaction of T lymphocytes with ,.-..

appropriate target cells. As a first step in under-standing the specific role of these proteins, cDNA
clones were obtained of both the T4 and T8 molecules and their nucleotide sequences were determined (20, 70). Comparison of the deduced protein sequences of T4 and T8 indicates that these molecules share signifi cant sequence and structural homology with immunoglo bulin variable (V) domains and as members of the im munoglobulin supergene family. However, the N-ter minsl V-like domains of T4 and T8 are quite different:
they share .only u8% homology and are therefore less homologous to each other than each is to immunoglobulin light chains. (Figure 9A). Moreover, the regions of maximum conservation between T4 and T8 are also the regions of strongest homology to immunoglobulin and T
cell receptor V regions. Thus, the immunoglobulin-like domains of these two molecules, although structurally similar, show significant sequence divergence consis-tent with the hypothesis that they recognize different molecules on different subsets of target cells.
The V-like region structural homology shared by the N-terminal domains of T4 and T8 may be of particular relevance to the functions of these proteins. Virtu-ally all members of the immunoglobulin supergene family pa~icipate :in the immune res onse p (62). Moreover, the individual members of this family show a strong tendency to associate with each other to form dimers.
This association is apparent in the interaction of the heavy and light chains of immunoglobulin, the alpha and beta chains of the T cell antigen receptor, p2-micro globulin and class I MHC proteins and the alpha and beta chains of class II MHC molecules. The T8 glyco protein formsc a disulphide bond with T6, a presumed MHC-like molecule, on the surface of thymocytes (82), and exists as multimers of the 32 kd subunit on periph-eral T lymphocytes (83). The presence of four V-like domains in T4 indicates that these regions associate with one another as well as with specific ligands on the surface of other cells or viruses. These specific affinities of immunoglobulin-like molecules may be essential for the recognition functions of T4 and T8.
Evolution of T4 In the immunoglobulin and T cell antigen receptor genes, the 'V and J exons are widely separated and be-come juxtaposed only after a somatic recombination event (62, Ei3). The T4 mRNA encodes four contiguous V-and J-like elements without the requirement for DNA
recombination events. It is therefore possible that T4 reflects a more primitive gene that evolved before the emergence of rearrangement mechanisms. Further support for this derives from recent observations that the first V--like region of T4 (V1) is split by an in tron not present in the V genes encoding either the immunoglobulins or T cell antigen receptors. Accumu lating evidence suggests that it is far more likely for introns to be precisely removed during evolution that for introns to be inserted in ~ a previously intron-free environment. Thus, T4 may represent an ancestral im munoglobulin gene which underwent duplications, diver gence, and :rearrangement to generate the current im munoglobulin gene family. Although functional in a far more complex immune system at present, T4 may reflect receptors operative in more primitive cellular immune responses. Primitive immune responses, such as those of invertebrates, do not appear to involve a diverse repertoire of receptor molecules, but in the simplest cases are restricted to a distinction between self and 13~~"l~~

nonself (8!i, 86) and are likely to be accommodated by a "static" set of genes that do not undergo rearrange-ment.
Whatever the order of appearance of T4 in evolutionary time, the organixation of this genes reveals an inter-esting example of exon shuffling. T4 consists of four V-J-like domains, a J-like region and a transmembrane segment, each sharing homology with different members of the immunoglobulin supergene family. The V- and J-like domains are homologous to the equivalent regions of both immunoglobulins and the T cell antigen receptor chains: the transmembrane domain shows considerable homology to this region in the p-chains of class II
t5 I~iC molecules (Figure 9C) . T4, therefore, consists of a collection of axons conserved in several members of the immunog:lobulin supergene family which are shuffled in different: ways to generate a large number of differ-ent molecules which participate in the immune response.
T4 is the AIDS Virus Receptor The data provided herein suggest a mechanism of AIDS
virus infection which initially involves the specific association of the AIDS virus with T4 molecules on the cell surface. This association may be demonstrated on T lymphocytes, B lymphocytes, and epithelial cells, and therefore does not require the participation of addi tional T cell-specific proteins. Additionally, the data provided herein indicates that the T4-AIDS virus complex is .internalized via receptor-mediated endocy tosis and the viral envelope then fuses with the limit ing membrane of the endosome, releasing the nucleocap sid into thEa cytoplasm. Viral replication and tran scription can then occur in both lymphoid and non-lym phoid cell lines. Moreover, the T4 gene is expressed in the brain as well as in lymphocytes, providing an explanation for the dual neurotropic and lymphotropic character ~of the AIDS virus. In this manner, a T lym-phocyte surface protein important in mediating effector cell-target: cell interations has been exploited by a human retrovirus to specifically target the AIDS virus to populations of T4+ cells.
Cell surface receptors have been identified for a num-0 ber of enveloped viruses and the pattern of expression of these receptors is often responsible for the host range and tropic properties of specific viruses (74, 76). Some viruses will infect only a narrow range of cell types, reflecting the expression of the viral ~5 receptor cn specific populations of target cells.
Rabies virus, for example, interacts with the nicotinic acetylcholine receptor (87) and infects largely skele-tal muscle and neurons, whereas the Epstein-Harr virus interacts with the C3d complement receptor type 2 (88) 20 and infects H lymphocytes. Other viruses, such as the myxoviruses, interact with ubiquitously distributed sialic acid residues on the cell surface and infect a much broader range of cell types.
25 ~m restricaed expression of cell surface receptors provides only one explanation for viral tropism. Some viruses will replicate only in a restricted set of differentiated cell types whereas others will only be efficiently transcribed in specific cell types. Hence, 30 the Moloney marine leukemia virus (Mo-MuLV) induces T
cell lymphomas in newborn mice " yet the closely-relat ed Friend helper marine leukemia virus (Fr-MuLV) induc es primarily erythroleukemias (89, 90, 91). This tro pism is thought to result from differences in the LTRs . -"- 1340~~0~
which facilitate tha efficient transcription of the Mo-MuLV genome in T lymphocytes and the Fr-MuLV genome in erythroid precursors (92, 93, 94).
As indicated herein, the primary tropic determinant of the AIDS vi~:vs is the expression of the T4 protein on the surface of the target cell. In vivo infection is restricted t,o lymphoid cells and myeloid cells as well as brain cells: three populations which express T4.
-In vitro demonstrations indicate that the introduction of T4 into T4 human ,B lymphocytes and epithelial cells, cellsc which are not natural targets for AIDS
virus, renders these cells susceptible to productive infection by AIDS 'virus.

Example 1: Soluble T4 Frarnnents Soluble T4 glycoprotein fragments are prepared using limited protease digestion from cell preparations.
Alternatively, DNA expression vectors encoding T4 fragments which lack the transmembrane domain, a region containing neutral and hydrophobic residues, may be constructed and used to produce such T4 fragments.
These fragments are soluble in aqueous solutions and contain leader (signal) sequences. When expressed in mammalian cells, these fragments are transported to the rough endop:lasmic reticulum/golgi complex and eventual-ly secreted from the cells.
Example 2: Treatment of AIDS Patients Soluble T4 c~lycoprotein fragments as described iri Exam-ple 1, tyica~lly in a pharmaceutically acceptable carri-er, are administered to patients infected with a human immunodeficiency ~rirus so as to bind to virus present in the the subject's blood and other body fluids and block infection of T4+ cells in vivo. Alternatively or additionally, a patient's blood is cycled through a column containing either immobilized T4 glycoproteins or soluble T'4 fragments so that the virus may be sepa-rated from the blood. Such measures permit the immune system to mount a more effective immunologic response against the virus, i.e., allow uninfected T4+ T cells to proliferate.
Soluble T4 fragments are used as a therapeutic, i.e., an inhibitor of extracellular and cell-cell spread of HIV infection. Applicants have shown that soluble T4 fragments inhibit in vitro HIV binding to, and infec-tion of, T4+ target cells (see Example 4). Administra-~.:~~U°~~?

tion of soluble T4 fragments to persons infected with HIV inhibits extracellular spread of the virus infec-tion. Additionally, fusion of HIV-infected T4+ cells and nonin:~ected T4+ cells, which is also a route by which the virus spreads, are inhibited by administra-tion of soluble T4 fragments.
Therefore, administration of soluble T4 fragments slows the course of disease, alleviates several symptoms associated with AIDS, and prevents occurrence of new pathologic changes.
Soluble T4 fragments, biochemically pure, aqueous solu ble reagents, are used in combination with other re agents to assay for competitors of the T4-HIV interac tion. Thus, soluble T4 fragments, in combination with HIV envelope prateins or biochemical mixtures contain-ing HIV envelope proteins, are used to screen for in-hibitors o:E viral binding.
Example 3: Production of Soluble T4 Fractments A cDNA encoding the membrane-bound T4 protein (pT4B) has been isolated, characterized, and expressed in a variety o! mammalian cell types (70). Soluble T4 frag manta are produced in bacterial, yeast, insect, and mammalian ;systems. Because the T4 protein likely folds in a complex manner and is glycosylated, expression in mammalian ;systems is preferred. Soluble T4 fragments are produced by truncating pT4H after the V4J4 domain.
Such DNA fragments terminate before the transmembrane segment, which begins at approximately nucleotide posi tion 1264 (Figure 6). This recombinant DNA molecule lacks both the transmembrane and the cytoplasmic do mains. The EcoRI-HpaII fragment, which encompasses ~.~4~7~~

nucleotides 1 through 1252, is isolated by assemblying it from smaller fragments of pT4B. Alternatively, the membrane-spanning segment alone is deleted, leaving the V4J4 domain fused to the cytoplasmic domain. One ap-proach is to delete the fragment spanning the HpaII
sites from nucleotides 1252-1342 from pT4H. Such con structs maintain the T4 signal sequence necessary for entry into the rough endoplasmic reticulum/golgi com plex and eventual secretion from the cell. In addi tion, these: constructs maintain the extracellular por tion of the: T4 protein in which the binding domain for the human :immunodeficiency virus envelope glycoprotein exists.
In order to express soluble T4 fragments in mammalian -systems, the modified T4 cDNA fragment is subcloned into vectors containing strong eukaryotic promot ers/enchancers as well as polyadenylation sites re quired for cleavage and polyadenylation of the RNA.
For example, the simian virus (SV40) early promoter and enchancer is positioned upstream from the soluble T4 cDNA fragment. Transcription termination and RNA poly-adenylation are achieved by placing the polyadenylation site of either SV40 or the human growth hormone gene downstream from the soluble T4 cDNA fragment. Intro-duction of a construct containing these elements to-gather with a selectable marker into eukaryotic cells by any of several methods known in the art leads to the stable integratian of exogenous DNA. Transformants selected by virtue of their ability to grow in selec-tive media secrete soluble T4 fragments into the cul-tune supernatant. Soluble T4 fragments are detected in the supernatant by one of several assays, e.g. radioim-munoprecipit,ation) and then purified.

-.
-81- 1~4~'~~~
Purification and characterization of soluble T4 frag-ments is greatly enhanced by constructing a cell line which overexpresses the secreted protein fragment.
Strategies which allow the overexpression of proteins have been employed in bacteria, yeast, insect, and mammalian systems. Inducible expression systems have also been employed in bacteria and yeast to overproduce proteins which may be toxic if constitutively ex pressed. Overexpression of soluble T4 fragments is accomplished by amplifying a soluble T4 expression vector, resulting in constitutive overexpression. The amplificaticm of dihydrofolate reductase (dhfr) genes by growth in progressively increased concentrations of the drug mei:hotrexate, an antagonist of dhfr, has been widely employed. Since the amplified unit is not limit ed to dhfr coding sequences, this approach results in the coamplification of sequences adjacent to them.
Therefore, dhfr is used as a selectable marker and as a means of coamplifying newly introduced sequences. This strategy has been successfully employed to increase the expression ~~f several different genes cotransformed with dhfr plasmids. An alternative amplification scheme involves cotransfection of the soluble T4 cDNA expres-sion vector with the plasmid pdLAT-3 followed by a selection scheme as previously described (102).
Therefore, t;he soluble T4 cDNA expression construct is cotransfecte<i with a dhfr expression plasmid. Alterna-tively, the dhfr gene resides on the same plasmid as the soluble T4 cDNA fragment, allowing linked cotrans-formation. ~rransfection of these constructs into dhfr deficient (dhfr ) Chinese hamster ovary (CHO) cells and subsequent selection in methotrexate permits the isola tion of stable transformants that express newly intro duced sequen<:es. Several clones are purified and cul tuts supernatants collected and assayed for the pres ence of soluble T4 fragments. Clones which produce the greatest levels of soluble T4 fragments are further characterized by Northern and Southern blot analyses.
These cell lines are then cultivated in selective media containing stepwise increased concentrations of metho trexate. This selective pressure results in the ampli fication of the newly introduced dhfr gene and adjacent T4 sequences. After reaching the highest methotrexate concentration, surviving cells are subjected to North ern and Southern blot analyses to determine the extent of amplification and culture supernatants are examined for the presence of soluble T4 fragments.
In order to characterize the soluble T4 fragments in the culture supernatant, several transformants are metabolically labeled with (35S)-methionine. Cell lysates and supernatants are then analyzed by radioim-munoprecipil:ation and Western blot analysis using com-mercially available anti-T4 antibodies. SDS-polyacryl-amide gel electrophoresis is performed on the precipi-tates and a~ band of the predicted relative ,molecular mass (Mr) of the secreted, truncated form of T4 is observed. Because it is synthesized in a mammalian system, this protein is properly glycosylated and fold-ed, i.e. disulphide bridges are formed. In order to purify the soluble T4 fragments from culture superna-tant, immunoaffinity column chromatography is performed using anti-'f4 antibodies. Protein bound to the column is eluted at high salt concentrations and low pH. SDS-polyacrylamide gel electrophoresis of the eluted mate rial is performed in order to determine the Mr and purity of tlhe eluted protein fraction. In addition, radioimmunoprecipitation and Western blot analysis is also performed in order to further characterize the ''-..

affinity-purified material.
Similar approaches may be undertaken in bacteria, yeast and insects to produce soluble T4 fragments. In addi tion, fragments smaller in size than the one described herein, e.g. containing only the V1J1 domain may be produced.
Construction of vectors Using recombinant DNA manipulations, base pairs (bp) 1-1257 of the :human T4 cDNA seguence were placed between an SV-40 early promoter and a TAA termination codon followed by the polyadenylation region of the bovine growth hormone gene. This sequence of the T4 cDNA
encodes the leader and predicted extracellular domain of the T4 receptor. This sT4 minigene was joined with a human H-ras, or a mouse dihydrofolate reductase mini gene, to cr~aate the vectors pST4CHras and pST4DHFR
respectively. The construction of these vectors was as follows:
Construction of pST4sal: Plasmid pST4sa1 was con-structed from two other plasmids, JRT4 and pUCsT4. The construction of these plasmids is detailed below.
Construction of plasmid JRT4: To create plasmid JRT4, plasmid DSP1 (103) was cut with XhoI, the SV40 polyA
early region was deleted and the Xho I sites were filled in using Klenow fragment of DNA polymerise. The bovine growth hormone polyadenylation region (113) was cut with PwII and KpnI, and the KpnI site was blunted by treatment 'with T4 DNA polymerise. This 230 by frag-ment was ligated to DSP1 to create DSP1BGH.
.

.. ~.'~~0~0~

DSP1HGH was cut with SmaI and SalI and the galK cas-sette (consisting of the SV40 early promoter, galK
coding region and BGH polyA region) was ligated into pUCl9 (107) at the SalI site by using a synthetic link-s er consisting of a SalI end, a BglII site, and a SmaI
end. This. three part ligation resulted in plasmid DSP18ZHGH.J'T.
DSP18ZBGH.JT, which was cut with StuI and BclI to de-lets the galK coding region, was ligated to a 1.7 kb EcoRi (fil:led in)-BamHI fragment, containing the T4 cDNA from plasmid pT4B (70) to create plasmid JRT4.
Construction of plasmid pUCsT4: To create plasmid pUCsT4, a HaeII and HpaII fragment (1125 bp) of the T4 cDNA from plasmid pT4B was ligated through the use of synthetic linkers to vector pUCl8 which had been cut with KpnI a,nd Xbal. The HaeII end of the T4 cDNA was ligated to the KpnI site of pUCl8 using a synthetic linker with a Kpn I end and a HaeII end. The HpaII end of the T4 cDNA was ligated to the XbaI site of pUCl8 using a synthetic linker with a HpaII end and an XbaI
end. This linker also inserted a TAA stop codon after nucleotide 1257 of the T4 coding region. The resulting plasmid was pUCsT4.
To create plasmid pST4sal, plasmied JRT4 was cut with BglII and Sacl and 959 by fragment (consisting of the SV40 early promoter and the first 602 nucleotides of the T4 cDNA) was isolated. Plasmid pUCsT4 was cut with SacI and XbaI and a 660 by fragment (consisting of the T4 cDNA from nucleotides 603-1257 followed by the TAA
codon from the synthetic linker) was isolated. These two fragments were ligated into DSP1HZBGH.JT which had been cut with Bg:LII and XbaI to delete the SV40 early .--, -g5-promoter and full length T4 coding region. The re-sulting plasmid was pST4sal.
Construction of pST4DHFR: To create plasmid pST4DHFR, a BglII-HamHI fragment containing a ~-globin DHFR ex-pression cassette was ligated into the BamHI site of pST4sal. The ~-globin DHFR expression cassette con sists of: the mouse ~-globin promoter (550 by HincII
fragment from plasmid pPK288 (108) modified at its 5' end with a synthetic linker to contain a BglII site;
the mouse I)HFR coding region (735 by HindIII (fill-in) fragment from plasmid pSV2-DHFR (109); NheI (fill-in)-BamHI (fill-in) SV40 polyA early region from DSP1 (103); and the mouse DHFR terminator region (907 by HindIII (fill-in fragment from plasmid mDH9_ (110), modified at its 3' end with a synthetic linker to cre-ate a BamHI site. The plasmid map of pST4DHFR is shown.
Construction of pST4cHras: Plasmid pMERcHras was cre-ated by cutting plasmid pSVK (111) with EcoRV and HindIII (fill-in), to remove the galK region, and li-gating in an 870 by NdeI (fill-in)-SalI (blunted by mung bean nuclease) fragment, containing the coding region for ~~Hras, from plasmid pSKcHras (112).
The soluble T4 transcription cassette was removed from pST4sal via a BglII-BamHI fragment and ligated into the BamHI site (3' to the SV40 polyA early) of pMERcHras to create pST4cHras.
Expression ~of soluble T4 (sT4) miniQenes in mammalian cells Expression of psT4cHras in NIH-3T3 cells: Plasmid pST4c8ras (10 y~g) was co-precipitated by the calcium 1~~~0'~02 phosphate precipitation method with 10 ~g of plasmid pTKneo, a vector conferring 6418 resistance, in the presence of .10 ~g carrier DNA (NIH-3T3 genomic DNA) onto NIH-3T3 cel7:s (seeded at 5 X 105 cells per 60 mm culture dish. on the preceding day). The cells were incubated with the precipitated DNA for 6 hours at 37~
C. The DNA precipitate was removed and fresh media 5% Nu-Serums (Collaborative Research, Inc., Lexington, M;assachusetts)) was added to the dishes.
The cells were trypsinized 16 hours later and seeded into three 100 mm dishes and maintained in the above media. Foci (approximately 50 per dish) appeared in 12-14 days. Eleven of the transfonaed foci were se lected, expanded and then seeded at 5 x 105 cells per 100 mm dish for selection in the above media plus 500 ~g/ml GENETICINe 6418 (Gibco Laboratories, Grand Is-land, New York). All 11 clones survived 6418 selection (500 ~g/ml) and were screened for H-ras (p21) levels by standard protein immunoblot analysis.
The clones which expressed the highest levels of p21 (approximately 2 ng p21/~g Triton-soluble protein) were assayed for e:~cpression of sT4. Confluent cultures were incubated for 18 hours With 35S-labelled methionine and cysteine. Culture supernatants and cell lysates were immunoprecipitated with monoclonal antibodies specific for the T4 (OKT4, OKT4A) and the T8 (OKT8) receptors, polyclanal antibody specific for ras proteins, or nonspecific mouse IgG. A protein of about 45 kd, the predicted size of sT4, was specifically precipitated from the culture medium by both of the monoclonal antibodies directed against the T4 receptor.
The sT4 band was not observed in cell lysates. As expected, p21 was precipitated from the cell but not from the culture supernatants. Subsequent quantitation ~.~~1~'~~2 shows, compared to purified sT4, these cells produce relatively low levels of sT4, i.e., approximately 100 fold lower than with CHO cells as described in Example 2B.
Expression ~of pST4DHFR in Chinese Hamster Ovary (CHO) cells: DXB-11 cells, a DHFR deficient CHO cell line (104) were transfected by calcium phosphate precip-tiation with 10 to 30 ~g of pST4DHFR in the presence of ~g carrier DNA (NIH-3T3 genomic DNA), one day after seeding 60 :mm dishes with 5 X 105 cells. Cells were incubated with the DNA precipitate for 6 hours at 37' C, the media was removed, and fresh media (F12, 10%
FBS, 100 units/ml penicillin and streptomycin) was added to the dishes. The media was changed again after 16 hours and the cells were incubated for another 24 hours. The cells were then trypsinized, seeded into three 100 mm dishes and selected in nucleoside-free media (F-12 without hypoxanthine and thymidine, 10%
dialyzed FBS, and 100 units/ml penicillin and strepto-mycin). Colonies (approximately 100 per dish) appeared in 7-10 days. Colonies from each dish were pooled, expanded and. then seeded at 5 x 103 and at; 5 x 104 cells per well in 24 well culture plates, or at 5 x 105 cells per 100 mm dish. The cells were allowed to re-cover for 3 days before beginning selection in nucleo-side free mEadia containing 20 nM methotrexate (mtx).
Individual wells ar clones were assayed at confluence for sT4 expressian, and those selected for further amplification were seeded into 24 well culture plates at the densities described above. Selection at 800 nM
mtx in nucleoside-free media was started 3 days after seeding. This selection procedure was repeated for.
selections at, 8 ~M mtx and 80 ~M mtx.

Several cell lines were derived using this method which express soluble T4 at a minimum of 3 pg/cell/24 hrs.
Purification of sT4: Conditioned medium (CM) was pre-pared serum-free from adherent cell cultures expanded into 850 cm2 roller bottles under mtx selective condi-tions. At confluence, the cells were washed twice with phosphate buffered saline (PBS) without Mg2+ and Ca2+
and the growth medium (Ham s F12 without hypoxanthine and thymid~ine, 10% fetal bovine serum, 100 units/ml penicillin and streptomycin and mtx at the selective concentration) was replaced with the same medium minus serum and mtx and plus 1 x ITS (insulin, transferrin and selenium (Collaborative Research Inc.). After 24 48 hrs., tt~e medium was removed and replaced with se lective growth medium. Serum-free medium was then reapplied within 3-5 days and this cycle repeated in definitely, i.e., for more than two months. CM was clarified by centrifugation at 8,000 x g. A protease inhibitor PMSF (phenylmethylsulfonylfluoride) was add ed to 0.5 ar,M and the CM was concentrated about 10-fold by pressure membrane filtration. This concentrated CM
was clarified by centrifugation at 2000 x g and Aprotinin, protease inhibitor (Sigma Chemical, St.
Louis, Missouri) was added to a final concentration of 5 ~g/ml. The sample was processed directly or after storage at ~-7 0' C .
The concentrated CM sample was diluted 2-fold with 50 mM MES [2-~(N-morpholino)-ethanesulfonic acid], pH 6.0 and filtered through a o.45 micron filter. The sample was then treated with 100 ~m pAPMSF (p-amidinophenyl-methylsulfonylfluoride) (Cal8iochem-Behring, San Diego, California) and applied to a S-Sepharosea (Sulpho-pro-pyl) (Pharmacia F-L Biochemicals, Piscataway, New Jer-~.~4~'~~~

say) column equilibrated in 50 mM MES, pH 6.0 at a protein concentration of 1.5-2.0 mg/ml gel. The sample was eluted using a linear gradient of 0-0.5 M NaCi in 50 mM MES, pH 6Ø Peak fractions which eluted at approximately 0.2 M NaCl were pooled and treated with pAPMSF to 100 ~Mf. Fractions containing sT4 were con-firmed by ;SDS-PAGE and immunoblot assays. After sit-ting at 4'C: for 1 hour the sample was dialyzed against 50 mM Bis-~Tris propane [1,3-bis[tris-(hydroxymethyl)-methyl amino]propane), pH 6Ø
The sample was treated with 100 ~M pAPMSF and 0.1% thio-diglycol, pH 9.0 was then applied to a Q-Sepharose°
(quarternary amino ethyl) column (Pharmacia) (5 ml ~5 sample/ml c~el) equilibrated in 50 mM Bis-Tris propane (BTP), pH 9Ø 'rhe sT4 sample does not bind to the Q-Sepharose° and was recovered in the unbound fraction and column wash. The unbound sample was immediately adjusted to pH 6Ø
The final step was chromatography on a 30 micron Supe-rose° 12 column (2.5 x 46 cm) (Phanaacia) equilibrated in 50 mM phosphate, 0.15 M NaCl pH 7Ø The column was run at a flow rate of 3.0 ml/min. Ten ml injections were made and the 42 minute peak was collected batch-wise. The process yielded approximately 1Ø mg of product per' 20.0 mg of total protein for a cell line producing approximately 3 pg/cell/day.

1~~~'~U~

Characterization of the sT4 Physical Properties: Total protein concentration was determined using the colormetric BCA protein assay (Hicinchoninic Acid, Pierce Chemical Co., Rockford, Illinois). Absolute concentrations were determined by quantitiativ~e amino acid analysis. The measured amino acid composition of the purified sT4 was performed using standard amino acid analysis techniques and was found to agree with the predicted sequence for the molecule to within experimental error (+/- 15%).
Through the first 20 residues the sequence was as pre dicted except that it begins lys-lys-val-val----.
Thus, the mature amino terminus begins at position +3 with respect to the predicted leader clip site and dif fers from the predicted sequence at that position by an asn to lys change. The position of the mature amino terminus agrees well with the determined termini of the mouse and sheep CD4 proteins. The asn to lys change may represent a ;sequencing error (single base change) or a mutation which arose during recombinant procedures.
Immuno-epito es: The monoclonal antibodies OKT4 and OKT4A recognize non-interfering surface epitopes of the T4 receptor (114). These antibodies are specific for the native conformation in that they will not bind to reduced, SDS denatured protein in immuno-blot assays.
It was shown that both antibodies specifically precipi-tate sT4 from 35S-labeled culture supernatants using the following immunoprecipitation procedure:
Cultures of sT4-producing cells containing 1 x 106 cells per 60 mm culture dish were labeled for 16 hours at 37~C in 1.5 ml methionine and cysteine free F12 medium containing ITS, and 170 ~Ci/ml [35S]methionine and 30 ~Ci/ml (35SJcysteine (ICN Biomedicals, Inc., Costa Mesa, California). Clarified medium (100 ~l) was diluted with an equal volume of precipitation buffer (10 mM sodium phosphate pH 7.5, 100 mM NaCl, 0.1% NP-40°, 0.5% non-fat dry milk) and incubated with 3 ~g rabbit IgG for 15 min. at 4°C followed by 30 ~l (packed volume) of protein A sepharose beads (Pharmacia P-L
Biochemical;s) for 30 min. at 4'C. The precleared su-pernatant was incubated with 5 ~g of OKT4, OKT4A and OKTS (gift: of P. Rao, Ortho Pharmaceuticals Corp., Raritan, N'ew Jersey), mouse IgG (Cooper Biomedical, Malvern, Pennsylvania), or rabbit a-mouse IgG (Cooper Biomedical) for 30 min at 4'C. OKT4, OKT4A, OKT8, mouse IgG, and rabbit a-mouse IgG were precipitated by incubation with 20 ~1 (packed volume) of protein A
sepharose beads :Eor 30 min. at 4°C. Following precipi tation, thE: beads were washed twice with 200 ~1 precip itation buffer and once with 200 ~1 precipitation buff er minus NP-40° and non-fat dry milk. The washed beads were boiled for 5 min. in 20 ~1 sample buffer (125 mM
Tris-HC1 pH 6.8, 20% glycerol, 1.4 M p-mercaptoe thanol), a;nd the supernatants were analyzed by elec trophoresis on a 12.5% SDS-polyacrylamide gel. Similar results were obtained with OKT4B, OKT4C, OKT4D, OKT4E, OKT4h and other Mabs specific for T4. These results suggest that the conformation of sT4 accurately mimics the surface: domain of the T4 receptor.
To determine whether sT4 can associate with HIV gp120 and whether this association can inhibit the binding of HIV to T4 cells, approximately 5 micrograms of purified sT4 were absorbed to Sepharose beads coated with OKT4 or control antibody. The beads were then mixed with a lysate of 35S-methionine labeled HIV. The complex of sT4 with OKT4 caprecipitates only the 120 kd envelope glycoprotei,n. No viral proteins are precipitated by ORT4 beads in the absence of sT4 or in the presence of control supernatants from the untransfected CHO cells.
Furthermore, viral protein is not precipitated if Sepharose beads coated with control mouse immunoglo-bulin (isotype matched to OKT4) are incubated with sT4.
These studies indicate that the sT4 obtained, which is free of other cell surface components present on the surface of T lymphocytes, is capable of specifically associating with the envelope glycoprotein of the AIDS
~i~s.
Cytofluorometry was performed to demonstrate that the interaction of T~ with gp120 of intact HIV abolishes the binding of AIDS virus to the surface of T4+ cells.
T4+ CEM cells were exposed to HIV in the presence or absence of ;sT4. Following viral absorption, the cells were washed,, exposed to fluorescein conjugated anti-HIV
antibody, and analyzed by flow cytometry (Figure 17) (48). In t:he absence of sT4, HIV binds efficiently to T4+ CEM cells. If HIV is preincubated with sT4, the binding of Virus to T4+ cells is abolished (Figure 17).
Ten nanograms of purified sT4 is sufficient to inhibit the binding of 100 nanograms of viral protein. If the envelope glycoprotein comprises 5% of the total viral protein, an estimated a 5:1 molar ratio of T4 to gp120 is capable of complete inhibition of HIV binding to T4+
c~lls.
The ability of sT4 to inhibit the infection of T4+
cells by H7:V was also studied. Phytohemagglutinin-stimulated ;human 1 hoc tes were a Ymp y xposed to serial ten-fold dil.utions of an HIV inoculum in the presence or absence of sT4, washed, and plated in microculture.
The frequency of infected cultures was determined using an immunoassay 4.8 and 12 days after exposure to virus ~ ~~o~fl~

(47). In this manner the infectious virus titer, ID-50 is defined as the reciprocal of the dilution required to infect 50~% of the exposed cell cultures at day 12.
In the absence of sT4, the ID-50 observed with the viral inoculum is approximately 105. However, in the presence of .e micrograms/ml of purified soluble T4, the infectively is diminished by almost 4 logs to an ID-50 of 101'5 (Fi.gure 18). This dramatic reduction in in-fectivity by HIV is observed throughout the entire course of infection. As a control for nonspecific inhibition or toxic effects of sT4, sT4 was added to cultures 18 hours after the initial exposure to virus.
Cultures exposed to sT4 18 hours after infection show only a 1 loci inhibition in the ID-50 which presumably results from inhibition of virus spread following the initial inoculation. Thus, the 4 log reduction in virus infectivity observed when virus 'is preincubated with sT4 is :Likely to result from the specific associa tion of sT4 with gp120 on the surface of virus. These particles are therefore no longer capable of interact ing with thE: T4 receptor on the cell surface. Four logs of inhibition were observed when 105 infectious particles/ml are preincubated with 8 micrograms/ml of sT4. The viral preparations of 105 infectious parti cles/ml were estimated to contain 109 particles/ml. If each particle contains 1,000 envelope glycoproteins, then the inhibition is obtained at a ratio of 100 T4 molecules/mol.ecule of envelope protein.
The availability of relatively large quantities of structurally intact sT4 facilitates the study of the mechanism of interactions of T4 with the surface of both antigen-presenting cells as well as with HIV vi rus. The :specificity of interaction of T4+ helper cells with antigen-presenting cells (B cells and macro ,.
1~~Q~~~

phages) may result, at least in part, from the associa-tion of T4 with class II l~iC molecules (105, 106j. The availability of significant amounts of purified sT4 permits a direct demonstration of a physical associa-tion between T4 and class II I~iC molecules.
The ability of sT4 to bind gp120 and inhibit viral infection in vitro indicates that sT4 is an effective anti-viral agent for treating AIDS patients.

1340'02 Example 4: Production of Soluble V1V2J4 Soluble T-4 Fragments Construction of Vectors Construction of psT4BBVIDHFR: To create plasmid psT4B-BVIDHFR, pl<ismid pST4DHFR (described in example 3) was cut with EcoRI arid XbaI and the smaller fragment con-taining the sT-4 coding region was deleted. Plasmid sT4sa1 was cut with XbaI and BbvI and a 1120 base pair coding fragment containing the sequence for soluble T-4 minus the leader region was isolated. This fragment was ligated to the EcoRI/XbaI-cut pST4DHFR, using a synthetic linker with an EcoRI end, a KpnI site, and a BbvI end. This fragment is compatible with the BbvI
~5 overhang on the sT-4 fragment isolated above. The resulting pl.asmid was called pST4BBVIDHFR.
Construction of OMPAST4: To create plasmid OMPAST4, plasmid OMPA.GS was digested with NcoI and SaII and the 20 smaller fragment containing the linker region was de-leted. OMP,~.GS is a derivative of pASI (Rosenberg et al., Meth. Enzymo:l., 101:123 (1983): U.S. Patent 4,578,-355) having a synthetic sequence inserted into the NdeI
site at the: 3' end of the cII ribosome binding se-25 quence. T:he synthetic sequence comprises the OMPA
leader followed by a multiliner sequence. The synthet-ic sequence was substantialy as follows:
5' -T ATG AAA AAG ACA GCT ATC GCG ATT GCA GTG GCA CTG

CTA GTT AAC TAG-3' Plasmid pUCsT4 was cut with Ncol and SaII and the 1149 base pair fragment containing SCD-4 sequence [T-4 1~44'~0~

nucleotides 124-1.257) was isolated. This fragment was ligated to 'the NcoI/SaII cut OMPA.GS to create OMPAST4.
Construction of OMPAST4BbvI: To create plasmid OMPAS-T4BbvI, plasmid OMPAST4 was cut with NaeI and XBaI.
The smaller fragment resulting from this cut, contain-ing the s'.~-4 coding region was deleted. Plasmid ST4BBVIDHFR was cut with KpnI and the resulting 3' overhang was blunt ended with T4 DNA polymerase. The blunt ended. DNA was then cut with XBaI and the 1124 base pair fragment containing the nucleotides 145-1257 of the CT4 cDNA was isolated. The isolated fragment was ligated to the NaeI/XbaI cut OMPAST4 plasmid to create plasmid OMPASTI4BbVI.
~5 Construction of pucGT4184: To created plasmid pucSt-4184, an EcoRI-NheI fragment from the T-4 cDNA [a 682 by fragment. encoding aminoacids (-23) to (+178)] was ligated to the synthetic linker sk727/725 (NheI and Aval ends) at the NheI site. sk727/725 codes for T-4 20 amino acids 179-:L85). The AvaI end of sk727/725 was ligated to ;gin AvaI-Xba 1 fragment of pUcST4 (comprising by 1198-125'7 of the human cDNA and encoding amino acids 351-369 of the T-~4 receptor followed by a TAA termina-tion codon). This sequence, which is flanked by EcoRI
25 and XbaI ends, was inserted into pUCl9 at the EcoRI and Xba 1 ends in the. puCl9 polylinker (sk727/725). sk727-/725 is substantially as follows:
5'gaccagaaggaggaggtgcaattgctagtgttcggattgactgccaac 3' 30 gtcttcci~cctccacgttaacgatcacaagcctaactgacggttgagc 5' Construction of pucST4106: To create plasmid pucST-4106, an Ec~oRI-Ava II fragment (comprising by 1-413, and encoding amino acids (-) 25-87) of the T-4 cDNA was joined to the synthetic linker sk791/792 (Ava II-AvaI
ends) at the AvaI:I site. sk791-792 codes for T-4 amino acids 88-104. The AvaI end of sk791/792 was ligated to an AvaI-XbaI fragment of pucST4 (comprising by 1198-1257 of the human cDNA, and encoding amino acids 351-369 of the T-4 receptor followed by a TAA termination codon). This sequence, which is flanked by EcoRI and XbaI, ends was inserted into pUCl9 at the EcoRI and XbaI ends in the puCl9 polylinker. sk791/792 is sub-stantially ass follows:
5' gaccagaaggaggaggtgcaattgctagtgttcggattgactgccaac gtcttcct:cctccacgttaacgatcacaagcctaactgacggttgagc 5' Construction of sT4184DHFR: To create plasmid sT41-84.DHFR, an EcoRI-XbaI fragment of pucST4184 (encoding amino acids -25 to 183 fused to 355 to 373) was ligat-ed to the EcoRI and XbaI ends of psT4DHFR in place of the EcoRI-Xx~aI fragment encoding sT-4.
Construction of sT4106DHFR: To create plasmid sT4106.-DHFR an Ec:oRI-XbaI fragment of pucST4106 (encoding amino acids -23 to 106 fused to 351 to 369) was ligated to the EcoR:C and XbaI ends of psT4DHFR in place of the EcoRI -Xba 1. fragment encoding sT-4.
Expression of pST4184DHFR in Chinese Hamster Ovary (CHO) cells:
DXB-11 cells, a DHFR deficient CHO cell line (Urlaub, et al., sera) were transfected by calcium phosphate precipitation with 10 to 30 ~cg of pST-4184DHFR in the presence of 10 ~g carrier DNA (NIH-3T3 genomic DNA), one day after Needing 60 mm dishes with 5 x 105 cells.
Precipitate:c were removed after 6 hrs. and replaced with F12 medium without hypoxanthine or thymidine, containing 10% dialyzed fetal bovine serum. Colonies (approximat:ley 100 per dish) appeared in 16 days.
Colonies from each dish were pooled, expanded and then seeded at :c x 104 cells per well in 24 well culture plates. These cells were grown in nucleoside free media conta~lning 80 nM methotrexate (mtx) to select for potential amplification of the transfected dhfr and V1V2J4 minigene. After two weeks in 8 nM mtx, actively growing cells were clearly visible. Individual wells or clones were assayed at confluence for expression of the deletion mutants and those selected for further amplification were seeded into 24 well culture plates at the density described above. A number of subpopu-lations which expressed high levels of V1V2J4 were grown in increasing levels of mtx to select for further amplification of the dhfr transcription unit and the T4184s(V1V2-J4) mi.nigene.
Several cells lines were derived using this method which expresses the deletion mutants in amounts of at least about 2 pg/cell/24 hrs.
Conditioned medium (CM) was prepared serum-free from 25 adherent cell cultures expanded into 850 cm2 roller bottles under mtx selective conditions. At confluence, tha cells were washed twice with phosphate buffered saline (PBS) without Mg2+ and Ca2+ and the growth medi um (Ham's F7.2 without hypoxanthine and thymidine, 10%
fetal bovine serum, 100 units/ml penicillin and strep tomycin and mtx at the selective concentration) was replaced witlh the same medium minus serum and mtx and plus 1 x ITS (insulin, transferrin and selenium (Col laborative Research Inc.). After 24-48 hrs. the medium was removed and replaced with selective growth medium.

.~, 1340'02 Serum-free medium was than reapplied within 3-5 days and this cy<:le repeated indefinitely) i.e., for more than two months. CM was clarified by centrifugation at 8,000 x g. A protease inhibitor PMSF (phenylmethyl-sulfonylfluoride) was added to 0.5 mM and the CM was concentrated about lOx fold by pressure membrane fil-tration. This concentrated CM was clairified by cen-trifugation at 2000 x q and Aprotinin, protease inhibi-tor, (Sigma chemical, St. Louis, Missouri) was added to a final concentration of 5 ~g/ml. The sample was pro-cessed directly or after.storage at -70'C.
Western Blot Analysis: Conditioned media from ViV2J4 producing CHO cells was concentrated and run on a 15~
~5 polyacrylamide reducing gel. Protein was transferred to nitrocellulose paper and probed with polyclonal antisera raised against a denatured E. coli derived T-4-NSI fusion molecule. This antisera specifically recognizes a V1V2J4 doublet migrating at approximately 20 25 Kd. This corresponds to the predicted size of the V1V2J4 coding sequence after leader processing. The physical basis for the doublet is not known. It is not likely to result from the differences in N-linked gly-cosylation, since the two consensus sequences in T-4 25 for this glyc~osylation are not present in V1V2J4.
Immuno-epitopes: The monoclonal antibodies OKT-4 and OKT-4A, OKT4B, OKT4C, OKT4D, OKT4E and OKT4F specifi cally recognize surface epitopes on the native, mem brane bound ~.C-4 receptor (Rao, et al., Cell Immunol.
80:310(1983)). These antibodies are specific for the native conformation. in that they will not bind to re duced, SDS-denatured protein in immuno-blot assays. It was shown that both antibodies specifically precipitate V1V2J4 from 3~'S-labelled culture supernatants using the ,.:..,,.

following immunoprecipitation procedure.
Cultures of ViV2J4 producing cells containing 1 x 106 cells per X50 mm culture dish, were labelled for 16 hours at 37'C in 1.5 ml methionine and cysteine free F12 medium containing ITS, and 170 ~Ci/ml [35S]methio nine and 30 ~Ci/ml.[35S]cysteine (ICN Biomedicals, Inc., Costa Mesa, CA). Clarified medium (100 ~1) was diluted with an equal volume of precipitation buffer lOmM so dium phosphate pH 7.5, (100 mM NaCl, 0.1$ NP-40°, 0.5%
non-fat dry milk) and incubated with 3 ~g rabbit IgG for min at 4'C followed by 301 (packed volume) of pro tein A sepha.rose beads (Pharmacia P-L Biochemicals) for 30 min at 4'C. The precleared supernatant was incu bated with !i ~g of each of the OKT4 antibodies refer enced above, mouse IgG (Cooper Biomedical, Malvern, PA), or rabbit a-mouse IgG (Cooper Biomedical) for 30 min at 4'C. The UKT4 antibodies, mouse IgG, and rabbit a-mouse IgG were precipitated-by incubation with 20 ~1 (packed volume) of protein A sepharose beads for 30 min at 4'C. Following precipitation, the beads were washed twice with 200 ~1 precipitation buffer and once with 200 ~1 precipitation buffer minus NP-40° and non-fat dry milk;. The washed beads were boiled for 5 min in 20 ~1 sample buffer (125 mM Tris-HCL pH 6.8, 20%
glycerol, 1.4 M ~-mercaptoethanol), and the supernatants were analyzed by electrophoresis on a 12.5% SDS-polyacrylamide gel. With the exception of OKT4, each of the monoclonal antibodies specifically precipitated the V1V2J4 from 35S-labelled cultured supernatants..
Competition of HIV binding: The recognition of V1V2J4 by OKT4A indicated that V1V2J4 would inhibit the bind ing of the AIDS virus to susceptible cells. CM con w 1340'~0~

taining or lacking ViV2J4 was used in the initial as-says. HIV was incubated with the CM and virus binding to the T-4~ CEM cell line was quantitated by incuba-tion with a FITC conjugated, anti-HIV antibody and FAGS
(Fluorescein activated cell sorter) analysis as de-scribed by McDougal, et al., supra (1985). CM from the V1V2J4 producing cell line inhibited HIV binding in a dilution dependent manner, whereas no response was seen with CM from matched non-producer (DXB-11) cells.
The protein V1J4 was similarly expressed in CHO cells.
However, in these preliminary experiments, the protein apparently was not exported into the media. Based on other studies showing that V1J4 produced in other re-combinant organisms binds OKT4A, it appears that V1J4 expressed in mammalian cell culture would inhibit HIV
binding.
ZO
Example 5: Preparation of Anti-Soluble T4 Fragment Antibodies Eight week old Balb/c mice are injected intraperito-neally with 50 micrograms of a purified soluble T4 fragment of the present invention (prepared as de-scribed above) in complete Freund's adjuvant, 1:1 by volume. Mice are then boosted, at monthly intervals, with the soluble T4 fragment mixed with incomplete Freund's ad:fuvant, and bled through the tail vein.
Zmmunoglobul.in cuts of sera are generated by ammonium sulfate precipitation and specific anti-soluble T4 fragment antibodies are purfified by affinity chroma-tography using an immobilized T4 fragment.

,... , )~~~~~

Example 6: Preparation of Soluble T4 Fragment Anti-Idiotypic Antibodies Syngenic and congenic mice are injected intraperito-neally with 50 micrograms of a purified anti-soluble T4 fragment antibody) of the present invention (prepared as described above) in complete Freund's adjuvant and boosted with the anti-soluble T4 fragment antibody in incomplete Freund's adjuvant monthly. On days 4, 3, and 2 prior to fusion, mice are boosted intravenously with 50 micr~~grams of immunoglobulin in saline. Splen-ocytes are then fused with P3X63 AG8.653 non-secreting myeloma cells accarding to procedures which have been described and are :known in the art to which this inven-~5 tion pertains. Two weeks later, hybridoma supernatants are screened for binding activity against anti-soluble T4 fragment antibodies by radioimmunoassay. Positive clones are then assayed for the ability to bind a human immunodeficie:ncy virus envelope glycoprotein and AIDS
20 virus. Alternatively, using the "one-step" procedure, mice are injected intraperitoneally with a soluble T4 fragment in <:omplete Freund's adjuvant, boosted intra-venously with the soluble T4 fragment in saline, and mice spleen ~~ells fused with myelomas as above. Hy-25 bridoma supernatants are then assayed directly for soluble T4 fragment. anti-idiotypic antibodies.

r.

REFERENCES
1. E.L. Reinherz et al., "Discrete Stages of Human Intrathymic Differentiation: Analysis of Normal Thymocyte and Leukemic Lymphoblasts of T Cell Lineage", Proc. Natl. Acad. Sci.
USA 77: 1588-1592 (1980).
2. E.L. Reinherz and S.F. Schlossman, "The Dif-ferentiation and Function of Human T Lympho-cytes", Cell 19: 821-827 (1980).
3 . M. L. Hlue et al . , "Coexpression of T4 and T8 on Peripheral Blood T Cells Demonstrated by Two-color Fluoroescence Flow Cytometry", J.
Immunol. 134: 2281-2286 (1985) 4. E.is. Engleman et al., "Activation of Human T
Lymphocyte subsets: Helper and Suppres sor/Cytotoxic T Cells Recognize and Respond to Distinct Histocompatibility Antigens", J.
Immunol. 127: 2124-2129 (1981) 5. A.M. Krensky~et al., "Long-term Human Cyto-lytic T-cell Lines Allospecific for HLA-DR6 Antigen Are OKT4+", Proc. Natl. Acad. Sci.
USA 79: 2365-2369 (1982).

6. S.C:. Meuer, S.F. Schlossman and E. Reinherz, "Cl.onal Analysis of Human Cytotoxic T Lympho cyt:es T4~ and T8+ Effector T Cells Recognize Products of Different Major Histocompatibili-ty Complex Regions", Proc. Natl. Acad. Sci.
USA, 79: 4395-4399 (1982).

,... .
-l04-7. W.E. Biddison et al., "Possible Involvement o:f the OKT4 Molecule in T Cell Recognition of Class II HLA Antigens", J. Exp. Med. 156:
1065-1076 (1982).

8. D.R: Wilde et al., "Evidence Implicating L3T4 in Class II MHC Antigen Reactivity Monoclonal Antibody GK 15 (Anti-L3T4) Blocks Class II
MHC Antigen-Specific Proliferation, Release oiF Lymphokines and Binding by Cloned Murine Helper T Lymphocyte Lines", J. Immunol. 131:
27L78-2183 (1983) .

9. S.L. Swain, "T Cell Subsets and the Recogni-tion of MHC Class", Immunol. Rev. 74: 129-142 -(7.983) .

10. Y. Thomas et al., "Functional Analysis of Human T' Cell Subsets Defined by Monoclonal Antibodies. IV. Induction of Suppressor Cells ZO
Wj.thin the OKT4+ Population", J. Exp. Med.
15~4: 459-467 (1981).

11. E.G. Engleman et al., "Antibodies to Membrane Structures that Distinguish Suppressor/Cyto-tc~xic and Helper T Lymphocyte Subpopulations Block the Mixed Leukocyte Reaction in Man", J. Exp. Med. 154: 193-198 (1981).

12. P. Marrack et al., "The Mayor Histocompati-bility Complex-restricted Antigen Receptor on T Cells. II. Role of the L3T4 Product", J.
Exp. Med. 158: 1077-1091 (1983).

13~~'702 13. L.. Rogozinski et al., "The T4 Surface Antigen is Involved in the Induction of Helper Func-tion", ;J) Immunol. 132: 735-739 (1984).

14. S.L. Swain, "Significance of Lyt Phenotypes:
Lyt2 Antibodies Block Activities of T Cells that Recognize Class I Major Histocompatibil-ity Complex Antigens Regardless of Their Fun-ction", Proc. Natl. Acad. Sci. USA 78: 7101-7105 (1981).

15. U. Landegren et al., "Selective Inhibition of Human T Cell Cytotoxicity at Levels of Target Recognition of Initiation of Lysis by Mono-~5 clonal OKT3 and Leu-2a Antibodies", J. Exp.
Med. 155: 1579-1584 (1982).

16. R.M. Zinkernagel and P.C. Doherty, "MHC-re-stricted Cytotoxic T Cells: Studies on the 20 Biological Role of Polymorphic Major Trans-plantation Antigens Determining T Cell Re-striction, Specificity, Function, and Respon-siveness"', Adv. Immunol. 27: 52-177 (1979).
25 17. J. Kappler et al., "The Major Histocompati-bil.ity Complex-Restricted Antigen Receptor on T Cells in Mouse and Man: Identification of Constant and Variable Peptides", Cell _35:
295-302 (' 1983 ) .
18 ~ O. Acuto et al . , "The Human T Cell Receptor:
Appearance in Ontogeny and Biochemical Rela-tionship of Alpha and Beta Subunits on IL-2 Dependent Clones and T Cell Tumors", Cell 34: 717-726 (1983).

19. P. Kavathas et al., "Isolation of the Gene Coding for the Human T Lymphocyte Antigen Leu-2 (T8) by Gene Transfer and cDNA Subtrac-tion", Proc. Natl. Acad. Sci. USA 81: 7688-7692 (1984).
20. D.R. Littman et al., "The Isolation and Se-quence of the Gene Encoding T8: A Molecule Defining Functional Classes of T Lymphocy tes", Cell 40:_237-246 (1985).
21. V.P.~Sukhatma et al., "The T Cell Differenti-ation Antigen Leu-2/T8 is Homologous to Im-munoglobulin and T Cell Receptor Variable Regions",, Cell 40: 591-597 (1985).
22. S.M. Friedman et al., "OT-CLL: A Human T
Cell Chronic Lymphocytic Leukemia That Pro-duces IL-2 in High Titer", J. Immunol. 128:
935-940 (1982).
23. D.A. Thurley-Lawon, L. Chess and J.L. Stro minger, "Suppression of In Vitro Epstein-Barr Infection: A New Role for Adult Human T
Cells", J. Exp. Med. 146: 495-508 (1977).
24. J.W. Goding, "The Chromic Chloride Method of Coupling Antigens to Erythrocytes: Defini-tion of Some Important Parameters", J. Im-munol. Methods 10: 61-66 (1976).
25. F.L" Graham and A.J. van der Eb, "A New Tech-niqi,ie for the Assay of Infectivity of Human Adenovirus DNA", Virology 52: 456-467 (1973).
-I~34~'~~2 26. M. Wigler et al., "Biochemical Transfer of Single-Copy Eucaryotic Genes Using Total Cel-lular DNA as Donor", Cell 14: 725-731 (1978).
27 ~ M., Wigler et al . , "Transfer of Purified Her-pes Virus Thymidine Kinase Gene to Cul-tured Mouse Cells", Cell 11: 223-232 (1977).
28. J.M. Chirgwin et al., "Isolation of Biologi-cally Active Ribonucleic Acid from Sources Enriched in Ribonuclease", Biochemistry 18:
5294-5299 (1979).
~5 29~ H. Aviv and P. Leder, "Purification of Bio-logically Active Globin Messenger RNA by Chromatography on Oligothymidylic Acid-Cellu-lose", Proc. Natl. Acad. Sci. USA _69: 1408-1412 (1972).
30. T. Huynh, R.A. Young and R.W. Davis, "Con-st:ruction and Screening cDNA Libraries in gtl0 and gtll", DNA Cloning Technigues - A
Practical Approach, D.M. Glover, ed. (Oxford:
IR~L Press) , in press.
31. T. Maniatis et al., "The Isolation of Struc-tural Genes from Libraries of Eucaryotic DNA", Cell 15: 687-701 (1978).
32. M.M. Davis et al., "Cell-Type-Specific cDNA
Probes and the Murine I Region: the Local-ization and Orientation of Ad Alpha", Proc.
Nat:l. Acad. Sci. USA 81: 2194-2198 (1984).

33. T. Maniatis, E.F. Fritch and J. Sambrook, Molecular Cloning (Cold Spring Harbor, NY;
Cold Spring Harbor Laboratory) (1982).
34. P.W.J. Rigby et al., "Labeling Deoxyribo-nucleic Acid to High Specific Activity In Vitro by Nick Translation with DNA Polymerise I", J. Mol. Biol. 113: 237-251 (1977).
35. J. Vieira and J. Messing, "The pUC Plasmids, an M13mp7-Derived System for Insertion Muta-genesis and Sequencing with Synthetic Univer-sal Primers", Gene 19: 259-268 (1982).
36., F. Singer, S. Nicklen and A. Coulson, "DNA
Sequencing with Chain-Terminating Inhibi-tors", proc. Natl. Acid. Sci. USA _74: 5463-5467 (1977).
20 37' E~ Southern, "Detection of Specific Sequences Among DNA Fragments Separated by Gel Electro-phoroesis", J. Mol. Biol. 98: 503-517 (1975).
38. G.;M. Church and W. Gilbert, "Genomic Sequenc 25 ing", Proc. Natl. Acid. Sci. USA 81: 1991 19~95 ( 1984 ) .
39. R.H. Scheller et al., "A Family of Genes that Codes for ELH, a Neuropeptide Eliciting a 30 Stereotyped Pattern of Behavior in Aplysia", Ce:Ll 28: 707-719 (1982).
40. K. Zinn, D. DiMaio and T. Maniatis, "Identi-fic:ation of Two Distinct Regulatory Regions 35 Ad:lacent to the Human Beta-Interferon Gene", 13~~'~02 Cell 34.: 865-879 (1983).
41. C. Terhorst et al., "Further Structural Stud-ies of the Heavy Chain of HLA Antigens and It.s Similarity to Immunoglobulins", Proc.
Natl. Acad. Sci. USA 74: 4002-4006 (1977).
42. J..A. Hedo, L.C. Harrison and J. Roth, "Bind ing of Insulin Receptors to Lectins: Evi dence for Common Carbohydrate Determinants on Se~~eral Membrane Receptors", Biochemistry _20:
3385-3393 (1981).
43. U.K. Laemmli, "Cleavage of Structural Pro tej.ns During the Assembly of the Head of Bac teriophage T4", Nature 227: 680-685 (1970).
44. R. Mann, R.C. Mulligan, and D. Baltimore, Construction of a retrovirus packaging mutant and its use to produce helper-free defective retrovirus, Cell 33, 153-159 (1983).
45. F. Barre-Sinoussi,et al., Isolation of a T
lymphotropic retrovirus from a patient at risk for acquired immune deficiecny syndrome (AI17S), Science 220, 868-871 (1983).
46. J.S" McDaugal, et al., Cellular tropism of the human retrovirus HTLV-III/LAV. I. Role of T cell activation and expression of the T4 antj.gen, ~T. Immol. 135, 3151-3162 (1985).
47. J.S. McDougal, et al., Immunoassay for the detection and quantitation of infectious human retrovirus, lymphadenopathy-assoicated I3~~'~

virus (LAV), J. Immunol. Meth. 76, 171-183 1;1985) .
48. J.S. McDougal, et al., Binding of HTLV-I:II/LAV to T4+ T cells by a complex of the l,lOK viral protein and the T4 molecule, Sci-ence 231, 382-385 (1986).
49~ C.B. Reimer, et al., Standardization of ligand binding assays for alpha-fetoprotein.
In Immunofluorescence and Related Staining Techniques. W., Knapp, K. Holubar, and G.
Wick, eds. (Amsterdam:Elsevier/North Holland Press) p. 189 (1978).
50. M.B. Wilson and P.K. Nakane, Recent develop-ments i.n the periodate method of conjugating horseradish peroxidase (HRPO) to antibodies.
In Immunofluorescence and Relating Staining Techniques, W. Knapp, K. Holubar, and G.
Wick, eds. (Amsterdam:Elsevier/North Holland Press), p. 215 (1978).
51. J,. Porath, R. Axen, and S. Ermback, Chemical coupling of proteins to agar, Nature 215, 14491-1493 (1967) .
52. B.J. Poiesz, et al., Detection and isolation of type C retrovirus particles from fresh and cultured lymphocytes of a patient with cuta-neous T cell lymphoma. Proc; Natl. Acd. Aci.
USA 77, 7415-7419 (1980).
53. P. Clapham, K. Nagy, and R.A. Weiss, Pseudo-types o:E human T-cell virus types 1 and 2:

1:~ ~ 0 '~ ~ 2 -ill-Neutralization by pateints' sera, Proc. Natl.
l~,cad. Sci. USA 81, 2886-2889 (1984).
54. A..G. Dalgleish, et al., The CD4 (T4) antigen is an essential component of the receptor for the AIDS retrovirus, Nature 312, 763-766 (1984) .
55. M. Popovic, et al., Detection, isolation, and continuous production of cytopathic retor-v:iruses (HTLV-III) from patients with AIDS
and Pre-AIDS, Science 224, 497-500 (1984).
56. K., Nagy, et al . , Human T-cell leukemia virus ~5 type 1: induction of sycytia and inhibition by patients' sera, Int. J. Cancer 32, 321-328 (7.983) .
57. D.M. Neville and H. Glossman, Molecular weight determination of membrane protein and glycoprotein subunits by discontinuous gel electrophoresis in dodecyl sulfate, Methods Enzymol. 32, 92-102 (1974)) 25 58 ~ A. Helenius, et al . , On the entry of Semliki Forest virus into BHK-21 cells, J. Cell.
Biol. 84,, 404-420 (1980).
59. R.D. Cone and R.C. Mulligan, High-efficiency 30 gene transfer into mammalian cells: Genera-tion of helper-free recombinant retroviruses with broad mammalian host range, Proc. Natl.
Acad. Sci. USA 81, 6349-6353 (1984).

60. F.W. Alt et al., "Probes for Specific mRNAs by Subtractive Hybridization: Anomalous Ex-pression of Immunoglobulin Genes", in Eucary-otic Gene ReQUlation, R.Axel, T.Maniatis and C.F. Fox) eds. (New York: Academic Press), pp. 407-419 (1979).
61. M. Kozak) "Comparison of Initiation of Pro-tein Synthesis in Procaryotes, Eucaryotes and Organelles", Microbiol. Rev. 47: 1-45 (1983).
62. L. Hood, M. Kronenberg and T. Hunkapiller, "T
Cell Antigen Receptors and the Immunoglobulin Supergene Family", Cell 40: 225-229 (1985).
63. S. 'ronegawa, "Somatic Generation of Antibody Diversity", Nature 302: 575-581 (1983).
64. J. Kyte and R.F. Doolittle, "A Simple Method for Displaying the Hydropathic Character of a Protein", J. Mol. Biol. 157: 105-132 (1982).
65. G. von Heijne, "Patterns of Amino Acids Near Signal-Sequence Cleavage Sites", Eur. J. Bio-chem,. 133: 17-21 (1983).
66. E.A. Kabat et al., "Sequences of Proteins of Immunologi.cal Interest" (Washington, D.C.;
U.S. Department of Health and Human Servic-es), p. 281 (1983).
67. A.F. Will.iams et al., "Cell Surface Glyco proteins and the Origins of Immunity", In the Proceedings of the Sigrid Juselius Symposi um, L.C. Andersson, C.G. Gahmberg and P.

Ekblom, eds. (New York: Academic Press), in press (1984).
68. L~.M. Amzel and R.J. Poljak, "Three-Dimension-al Structure of Immunoglobulins", Ann. Rev.
Biochem. 48: 961-997 (1979).
69. P.Y. Chou and G.D. Fasman, "Empirical Predic-tions of Protein Conformation", Ann. Rev.
-Biochem. 47: 251-276 (1978).
70. P.J. Maddon, et al., The isolation and nu cleotide sequence of a cDNA encoding the T
cell surface protein T4: A new member of the immunoglobulin gene family, Cell 42, 93-104 (:1985) .
71. R.A. Adams, A. Flowers and B.J. Davis, Direct implantation and serial transplatation of human acute lymphoblastic leukemia in ham st:ers, SB-2, Can. Res. 28~ 1121-1125 (1968).
72. G.O. Gey, W.D. Coffman and M.T. Kubicek, Tissue culture studies of the proliferative capacity of cervical carcinoma and normal epithelium, Cancer Res. 12, 264-265 (1952).
73. R.J.V. Pulvertaft, Cytology of Burkitt's tumor (African lymphoma), Lancet _I, 238-240 (1964) .
74. N.~J. Dimmock, Initial stages of infection with animal viruses, J. Gen. Virol. 59, 1-22 ( 1'982 ) . .

75. J. White, M. Kielian and A. Helenius, Mem-brane fusion proteins of enveloped animal viruses, Quart. Rev. Biophys. _16, 151-195 (1983).
76 ~ M'. Marsh, The entry of enveloped viruses into cells by endocytosis, Biochem. J. 2I8, 1-10 (1984).
77~ M. Kielian and A. Helenius, Entry of alpha-viruses. In the Togaviridae and Flaviviri-dae, S. Schlesinger and M.J. Schlesinger, a<is., (Plenum Publishing Corp.), pp. 91-119 (:1986) .
78. S.. Ohkuma, and B. Poole, Fluorescence probe measurements of the intralysosomal pH in living cells and the pertubation of pH by various agents, Proc. Natl. Acad. Sci. USA
3327-3331 (1978).
79~ F.R. Maxfield, Weak bases and ionophore rap-idly and reversibly raise the pH of endocytic vesicles in cultured mouse fibroblasts, J.
Cell. Biol. 95, 676-681 (1982).
-80~ A. Helenius, M. Marsh, and J. White, Inhibi-ti~~n of Semliki Forest virus penetration by ly~sosomotropic weak bases, J. Gen. Virol. _58, 47~-61 (1982) .
81. R.'.C. Johnson and J. C. McArthur, AIDS and the brain, TINS 9, 91-94 (1986).

.., 82. P.M. Snow, M. Van de Rijn and C. Terhorst, "Association Between the Human Thymic Differ-entiation Antigens T6 and T8", Eur. J. Im-munol., in press (1985).
83~ P.M. Snow and C. Terhorst, "The T8 Antigen is a Multimeric Complex of Two Distinct Subunits on Human Thymocytes but Consists of Homo-mu.ltimeric Forms on Peripheral Blood T Lym-ph~~cytes", J. Biol. Chem. 258: 14675-14681 (1!83) .
84. C. Terhorst et al., "Biochemical Analysis of Human T Lymphocyte Differentiation Antigens T4 and T5", Science 209: 520-521 (1980).
85. W.H. Hildemann, "Immunocompetance and Allo-geneic Polymorphism Among Invertebrates", Transplantation 27: 1-3 (1979).
86. V.L. Scofield et al., "Protochordate Allore-cognition is Controlled by a MHC-like Gene System", Nature 295: 499-502 (1982).
87) T.L. Lentz, et al., Is the acetylcholine receptor a rabies virus receptor ?, Science 215 182-184 (1982).
88. J.D~. Fingeroth, et al., Epstein-Barr virus receptor of human B lymphocytes is the C3d receptor CR2, Proc. Natl. Acad, Sci. USA 81, 4510-4514 (1984).
89~ P.E. Tambourin, et al., The physiopathology of Friend leukemia, Leukemia Res. 3, 117-129 _ ('1979) "
90. A,. Oliff, et al., Isolation of transplantable erythraleukemia cells from mice infected with helper-independent Friend murine leukemia virus, Hlood 58, 244-254 (1981).
91~ J.E. Silver and J.M. Fredrickson, Suscep-tibility to Friend helper virus leukemias in C:?CB recombinant inbred mice, J. Exp. Med.
1.'58, 1693-1702 (1983) .
92~ P.A. Chatis, et al., Role for the 3' end of the genome in determining disease specificity of Friend and Moloney murine leukemia virus-es, Proc. Natl. Acad. Sci, USA. 80, 4408-4411 (7.983) .
93~ P.A. Chatis, et al., A 3' end fragment encom passing the transcriptional enhancers of nondefective Friend virus confers erthyro-leukemogenicity on Moloney leukemia virus, J.
Virol. 52, 248-254 (1984).
94. A. Bosze, H.J. Thiesen and P. Charnay, A
transcriptional enhancer with specificity for erythroi.d cells is located in the long termi-nal repeat of the Friend murine leukemia virus, EMBO J. 5, 1615-1623 (1986).
95~ A.N. Barclay, et al., Immunoglobulin-related structures associated with vertebrate cell surfaces, in press.

1340'l0~

96. M.O. Dayhoff, W.C. Barker and L.T. Hunt, Establishing homologies in protein sequences.
In Methods in Enzymology. Enzyme Sturcture Part I, C.H.W. Hirs and S.N. Timasheff, eds.
(;New York: Academic Press), pp. 524-545 (1983).
97~ Y. Yanagi, et al., A human T cell-specific cl7NA clone encodes a protein having extensive homology to immunoglobulin chains, Nature 308, 145-149 (1984).
98~ G.K. Sim, et al., Primary structure of human T-cell receptor -chain, Nature 312, 771-775 ( 1.984 ) .
99~ H. Saito, et al., A third rearranged and expressed gene in .a clone of cytotoxic T
lymphocytes, Nature 312, 36-40 (1984).
ZO
100. H. Saito, et al., Complete primary structure of the E chain and gene of the mouse major histocompatibility complex, Proc. Natl. Acad.
USA 80, 5520-5524 (1983).
101. M. Isobe, et al., The gene encoding the T-ce:Ll surface protein T4 is located on human chromsome 12, Proc. Natl. Acad. Sci. USA, _83, 43Si9-4402 (1986) .
102. J.M. Roberts and R. Axel, Gene Amplification and Gene Correction in Somatic Cells, Cell 29, 109-119 (1982).

.,-.. .
. .

103. D.t3. Pfarr, G. Sathe and M.E. Reff, A Highly Modular Cloning Vector for the Analysis of Euc:aryotic Genes and Gene Regluatory Ele-ments, DNA, 4, 461-467 (1985).

104. G. Urlaub and L.A. Chasin, Isolation of Chi-nese Hamster Cell Mutants Deficient In Dihy-drofolate Reductase Activity, Proc. Natl.

Acad. Sci.. USA, 77, 4216-4220 (1980).

105. Gay, D. et al., Nature 328: 626-629 (1987).

106. Sleckman, B.D. et al., Nature 328: 351-353 (1987) .

107. Yan:isch-Perron, et al., Gene 33: 103 (1985).

108. Berg, P.,. et al. Mol. Cell. Hiol. _3: 246 (1983) .

109. Subramani" et al., Mol. Cell Biol. _1: 854 (198.1) .

110. Frayne, et al., Mol. Cell Biol. _4: 2921 (1984) .

111. Schumperli., et al., Proc. Natl. Acad. Sci.

USA ,79: 257 (1982).

112. Gross, et. al., Mol. Cell Biol. 5: 1015 _ (198'5) .

113. Pfarr, et al., DNA 5: 115 (1986).

114. Rao, et al., Cell Immunol. 80: 310 (1983).
-,.~.. .
1 ~~0'~02 115. Mr.Clure, et al., Cold Spring Harbor Confer-ence on Cell Proliferation 9: 345 (1982).
116. Urlaub, et al., Cell 33: 405 (1983).
117. K~.m, et al., Cell 42: 129 (1987).
118. Wi.gler, et al., PNAS USA 76: 1373 (1979).
119. Copeland, et al., Cell 17: 993 (1979).
120. Sattentau, et al., Science 234: 1120 (1986).
121. Greenstein, et al., Ann. Inst. Pastuer 138:
134 (1987) .
122. Gay, et al., Ann. I9nst. Pasteur 138: 127 ( 1'9 8 7 ) .

35~

Claims (16)

1. A polypeptide capable of specifically forming a complex with human immunodeficiency virus envelope glyco-protein, the amino acid sequence of which comprises the amino acid sequence shown in Figure 6 from about +3 to about +185 fused to the amino acid sequence from about +351 to about +369.

2. A polypeptide capable of specifically forming a complex with human immunodeficiency virus envelope glyco-protein, the amino acid sequence of which comprises the amino acid sequence shown in Figure 6 from about +3 to about +106 fused to the amino acid sequence from about +351 to about +369.

3. A polypeptide capable of specifically forming a complex with human immunodeficiency virus envelope glyco-protein, the amino acid sequence of which comprises the amino acid sequence shown in Figure 6 from about +3 to about +185.

4. A pharmaceutical composition which comprises an effective amount of the polypeptide of any of claims 1, 2 or 3 and a pharmaceutically acceptable carrier.

5. The use of a polypeptide in accordance with any one of claims 1 to 3, in the control of a human immuno-deficiency virus infection, or for the manufacture of a medicament therefor.

6. An expression vector encoding the polypeptide of any one of claims 1, 2 or 3.

7. A host cell comprising the expression vector of claim 6.

8. A bacterial host cell of claim 7.

9. An Escherichia coli host cell of claim 8.

10. A eucaryotic host cell of claim 7.

11. A mammalian host cell of claim 10.

12. A yeast host cell of claim 10.

13. An insect host cell of claim 10.

14. A method of producing a polypeptide capable of specifically forming a complex with human immuno-deficiency virus envelope glycoprotein comprising growing a host cell comprising an expression vector encoding a polypeptide, the amino acid sequence of which comprises the amino acid sequence shown in Figure 6 from about +3 to about +185 fused to the amino acid sequence from about +351 to about +369, under suitable conditions permitting production of the polypeptide and recovering the polypeptide so produced.

15. A method of producing a polypeptide capable of specifically forming a complex with human immuno-deficiency virus envelope glycoprotein comprising growing a host cell comprising an expression vector encoding a polypeptide, the amino acid sequence of which comprises the amino acid sequence shown in Figure 6 from about +3 to about +106 fused to the amino acid sequence from about +351 to about +369, under suitable conditions permitting production of the polypeptide and recovering the polypeptide so produced.

16. A method of producing a polypeptide capable of specifically forming a complex with human immuno-deficiency virus envelope glycoprotein comprising growing a host cell comprising an expression vector encoding a polypeptide, the amino acid sequence of which comprises the amino acid sequence shown in Figure 6 from about +3 to about +185, under suitable conditions permitting production of the polypeptide and recovering the polypeptide so produced.