Monoclonal antibody and ligand binding sites of the T cell erythrocyte receptor (CD2)

Eph receptors as cancer targets for antibody-based therapy

Receptor tyrosine kinases (RTKs) are integral membrane sensors that govern cell differentiation, proliferation and mobility, and enable rapid communication between cells and their environment. Of the 20 RTK subfamilies currently known, Eph receptors are the largest group. Together with their corresponding ephrin ligands, Eph receptors regulate a diverse array of physiologic processes including axonal guidance, bone remodeling, and immune cell development and trafficking. Deregulation of Eph signaling pathways is linked to cancer and other proliferative diseases and, because RTKs play critical roles in cancer development, the specific targeting of these molecules in malignancies provides a promising treatment approach. Monoclonal antibodies targeting RTKs represent a potentially attractive modality for pharmaceutical development due to their relatively high target specificity and low off-target binding rates. Therefore, new technologies to generate antibodies able to target RTKs in their native in vivo context are likely to facilitate pre-clinical and clinical development of antibody-based therapies. Our group has recently reported a platform discovery methodology termed Selection of Phage-displayed Accessible Recombinant Targeted Antibodies (SPARTA). SPARTA is a novel and robust stepwise method, which combines the attributes of in vitro screenings of a naïve human recombinant antibody library against known tumor targets with those features of in vivo selections based on tumor-homing capabilities of a pre-enriched antibody pool. This unique approach overcomes several rate-limiting challenges to generate human monoclonal antibodies amenable to rapid translation into medical applications.

CD2 Immunobiology

The glycoprotein CD2 is a costimulatory receptor expressed mainly on T and NK cells that binds to LFA3, a cell surface protein expressed on e.g., antigen-presenting cells. CD2 has an important role in the formation and organization of the immunological synapse that is formed between T cells and antigen-presenting cells upon cell-cell conjugation and associated intracellular signaling. CD2 expression is upregulated on memory T cells as well as activated T cells and plays an important role in activation of memory T cells despite the coexistence of several other costimulatory pathways. Anti-CD2 monoclonal antibodies have been shown to induce immune modulatory effects in vitro and clinical studies have proven the safety and efficacy of CD2-targeting biologics. Investigators have highlighted that the lack of attention to the CD2/LFA3 costimulatory pathway is a missed opportunity. Overall, CD2 is an attractive target for monoclonal antibodies intended for treatment of pathologies characterized by undesired T cell activation and offers an avenue to more selectively target memory T cells while favoring immune regulation.

Monolayers from Genetically Engineered Protein Pores

A selection of microscopic pores is being made by genetic manipulation of a bacterial channel protein, α-hemolysin (α-HL). It will include: pores with different internal diameters, with differential selectivity for the passage of classes of molecules, and with different gating properties. The pores will be made into monolayers and incorporated into materials such as thin films to confer novel permeability properties upon them. Such products will have several technological applications, for example as molecular filters in sensors or as components of optically gated devices in electronics.

Processing and localization of the african swine fever virus CD2v transmembrane protein

The African swine fever virus (ASFV)-encoded CD2v transmembrane protein is required for the hemadsorption of red blood cells around infected cells and is also required for the inhibition of bystander lymphocyte proliferation in response to mitogens. We studied the expression of CD2v by expressing the gene with a V5 tag downstream from the signal peptide near the N terminus and a hemagglutinin (HA) tag at the C terminus. In ASFV-infected cells, a full-length glycosylated form of the CD2v protein, which migrated mainly as a 89-kDa product, was detected, as well as an N-terminal glycosylated fragment of 63 kDa and a C-terminal nonglycosylated fragment of 26 kDa. All of these forms of the protein were localized in the membrane fraction of cells. The 26-kDa C-terminal fragment was also produced in infected cells treated with brefeldin A. These data indicate that the CD2v protein is cleaved within the luminal domain and that this occurs in the endoplasmic reticulum or Golgi compartments. Confocal microscopy showed that most of the expressed CD2v protein was localized within cells rather than at the cell surface. Comparison of the localization of full-length CD2v with that of a deletion mutant lacking all of the cytoplasmic tail apart from the 12 membrane-proximal amino acids indicated that signals within the cytoplasmic tail are responsible for the predominant localization of the full-length and C-terminal 26-kDa fragment within membranes around the virus factories, which contain markers for the Golgi compartment. Processing of the CD2v protein was not observed in uninfected cells, indicating that it is induced by ASFV infection.

Lamina propria T cell activation: role of the costimulatory molecule CD2 and its cytoplasmic tail for the regulation of proliferation and apoptosis

BACKGROUND/AIMS

Accumulation of T lymphocytes in the gut is a hallmark of inflammatory bowel disease probably caused by insufficient T cell apoptosis. Activated peripheral T cells, or "resting" lamina propria T lymphocytes (LPLs), are highly susceptible to apoptosis induction, e.g., using the mitogenic anti-CD2 monoclonal antibody (mAb) pair T11(2+3). It is, however, unknown how CD2-mediated LPL apoptosis is related to proliferation and whether the whole CD2 molecule is required for apoptosis induction.

MATERIALS AND METHODS

Mapping of anti-CD2 mAb was performed using erythrocyte rosetting assays and cross-blocking enzyme-linked immunosorbent assay (ELISA). Lamina propria mononuclear cells (LPMNCs) or phytohemagglutinin (PHA) blasts were stimulated with a panel of 18 anti-CD2 mAbs followed by apoptosis analysis [Annexin V expression on propidium iodide (PI)-negative cells, 4c6-diamidino-2-phenylindole x 2HCl (DAPI) staining]. Proliferation was measured by [(3)H]-thymidine incorporation. For structural analysis, EL4 cells were used which were transfected with human CD2 (wild type (WT), cytoplasmic-deficient, cytoplasmic CD28). Sorting was performed employing standard techniques

RESULTS

All three mitogenic anti-CD2 mAb pairs induced apoptosis of LPMNC and PHA blasts. Two out of four submitogenic anti-CD2 mAb, AICD2.M3, and ICRFCD2.3 lead to LPMNC proliferation but no apoptosis. Importantly, apoptosis was also detected in cytoplasmic-deficient CD2 tg or CD2/CD2/CD28 tg EL4 cells. Sorted CD45(high) huCD2 WT EL4 had higher apoptosis rates compared to WT huCD2tg EL4 cells

CONCLUSION

LPMNC apoptosis induction via CD2 is always associated with proliferation, although proliferation is not necessarily associated with apoptosis. The cytoplasmic tail of CD2 is not required, and CD45 appears to transmit apoptotic signals entering the T cell via CD2.

Peptide self-aggregation and peptide complementarity as bases for the evolution of peptide receptors: a review

This paper reviews the three major theories of peptide receptor evolution: (1) Dwyer's theory that peptide receptors evolved from self-aggregating peptides; (2) Root-Bernstein's theory that peptide receptors evolved from functionally and structurally complementary peptides; and (3) Blalock's theory that receptors evolved from hydropathically complementary sequences encoded in the antisense strand of the DNA encoding each peptide. The evidence to date suggests that the co-yevolution of peptides and their receptors is strongly constrained by one or more of these physicochemically based mechanisms, which argues against a random or frozen accident' model. The data also suggest that structure and function are integrally related from the earliest steps of receptor-ligand evolution so that peptide functionality is non-random and highly conserved in its origin. The result is a molecular paleontology' that reveals the evolutionary constraints that shaped the interaction of structure and function.

Analysis of human and primate CD2 molecules by protein sequence and epitope mapping with anti-human CD2 antibodies

A panel of anti-human CD2 monoclonal antibodies (mAb) and soluble human CD58 (LFA-3) were tested for binding to human peripheral blood mononuclear cells (PBMCs), recombinant human CD2 and mononuclear cells from Cynomolgus, Rhesus and African green monkey, Stump-tail, Pig-tail and Assamese macaque, Chimpanzee and Baboon. This analysis revealed that whilst some antibodies recognized all species, there were differential binding profiles with others. Three antibodies, MEDI-507, 6F10.3 and 4B2, recognized CD2 from human and Chimpanzee but not that from the other primates. We have cloned eight of the previously unknown primate CD2 molecules and report here their sequences for the first time. This analysis revealed that 12 amino acids formed a common set of residues in the extra cellular domain of human and Chimpanzee CD2. Using a "knock-in" mutagenesis approach starting with Baboon CD2, which does not bind MEDI-507, 6F10.3 and 4B2, we have identified three residues in the adhesion domain of human CD2 which are critical for its binding to these mAbs. These residues, N18, K55 and T59 define a region located outside of the previously described binding regions on CD2. Affinity measurements of the mutants revealed a variety of degrees of binding restoration for MEDI-507, 6F10.3 and 4B2, indicating that there are fine differences within a given epitope. Furthermore, the analysis of the competition of several of the anti-human CD2 antibodies with each other and CD58 demonstrated the existence of a continuum of overlapping epitopes on human CD2, which is in contrast to the commonly held belief that epitopes on human CD2 are clearly segregated.

A feline CD2 homologue interacts with human red blood cells

A cDNA encoding a feline homologue of CD2 (fCD2) was identified. Several amino acids (aa) important for ligand interaction, molecular folding or signal transduction, found in other mammalian CD2, were found to be highly conserved in the predicted fCD2 aa sequence. fCD2-expressing cells were able to form rosettes with human red blood cells (probably via human CD58), and the rosette formation was inhibited by an anti-fCD2 monoclonal antibody. These results are indicative of the similarity of feline and human CD2 structures. fCD2 was found to be expressed in feline peripheral blood T lymphocytes, monocytes and cultured lymphoid cells.

Expression, epitope analysis, and functional role of the LFA-2 antigen detectable on neoplastic mast cells

Recent data suggest that mast cells (MCs) in patients with systemic mastocytosis or mast cell leukemia express a CD2-reactive antigen. To explore the biochemical nature and function of this antigen, primary MCs as well as the MC line HMC-1 derived from a patient with mast cell leukemia were examined. Northern blot experiments revealed expression of CD2 messenger RNA in HMC-1, whereas primary nonneoplastic MCs did not express transcripts for CD2. In cell surface staining experiments, bone marrow (BM) MCs in systemic mastocytosis (n = 12) as well as HMC-1 cells (30%-80%) were found to express the T11-1 and T11-2 (but not T11-3) epitopes of CD2. By contrast, BM MCs in myelodysplastic syndromes and nonhematologic disorders (bronchiogenic carcinoma, foreskin phimosis, uterine myeomata ) were consistently CD2(-). All MC species analyzed including HMC-1 were found to express LFA-3 (CD58), the natural ligand of CD2. To study the functional role of CD2 on neoplastic MCs, CD2(+) and CD2(-) HMC-1 cells were separated by cell sorting. CD2(+) HMC-1 cells were found to form spontaneous aggregates and rosettes with sheep erythrocytes in excess over CD2(-) cells, and a T11-1 antibody inhibited both the aggregation and rosette formation. Moreover, exposure of CD2(+) HMC-1 cells to T11-1 or T11-2 antibody was followed by expression of T11-3. In addition, stimulation of neoplastic MCs through T11-3 and a second CD2 epitope resulted in histamine release. These data show that neoplastic MCs express functionally active CD2. It is hypothesized that expression of CD2 is associated with pathologic accumulation and function of MCs in systemic mastocytosis.

Selective deletion of antigen-specific, activated T cells by a humanized MAB to CD2 (MEDI-507) is mediated by NK cells

CD2 is a 50-kDa transmembrane glycoprotein that plays an important role in T and natural killer (NT) lymphocyte functions. CD2 serves as both an adhesion molecule and as a costimulatory molecule through interactions with its ligand, CD58, on antigen presenting or target cells. Consistent with earlier studies using a rat anti-CD2 mAb, we have shown that treatment of alloantigen stimulated T lymphocytes with a humanized mAb, MEDI-507 (IgG1, kappa), induced hyporesponsiveness to subsequent stimulation with alloantigen but not to mitogen (phytohemagglutinin). Fluorescence-activated cell sorting analysis of cells from mixed lymphocyte reaction (MLR) treated with MEDI-507 revealed pronounced deletion of T and NK cells, consistent with lack of proliferation in the MLR. MEDI-507 F(ab')2 fragments did not have inhibitory activity or induce deletion of lymphocytes in the MLR. Removal of the NK cell subset by magnetic bead depletion using anti-CD16 and anti-CD56 mAbs eliminated both the T cell deletion and the inhibitory effect. Reconstitution of NK depleted responder populations using autologous NK cells restored the MEDI-507-mediated deletion activity to levels measured in the original MLR. Formaldehyde-fixed NK cells failed to mediate the MEDI-507-induced deletion effect. Altogether, our studies indicate that activated T cells with MEDI-507 bound to CD2 are preferential targets for autologous NK cells through a nonapoptotic cytotoxic mechanism.

Aberrant expression and reverse signalling of CD70 on malignant B cells

In normal lymphoid tissues the tumour necrosis factor-receptor family member CD27 and its ligand CD70 have a restricted expression pattern. Previously, we reported that expression of CD27 is deregulated in B-cell leukaemias and lymphomas. Here we show that, although infrequently expressed by normal human B cells in vivo, CD70 is found on 50% of B-CLLs, 33% of follicle centre lymphomas, 71% of large B-cell lymphomas, and 25% of mantle cell lymphomas. Interestingly, in the majority of leukaemias and lymphomas examined, CD70 was found to have a capped appearance, a feature that coincided with co-expression of CD27. Functional analysis showed that a subset of B-CLLs could proliferate vigorously in response to CD70 mAb but not to CD27 mAb. This response was synergistically enhanced by ligation of CD40 but inhibited by the presence of IL-4. Additional experiments indicated that the proliferative response was due to an agonistic signal delivered via CD70, rather than blocking of negative signalling by CD27. Thus, next to its role as ligand, in a subset of malignant B cells CD70 can operate as receptor and as such might contribute to progression of these B-cell malignancies.

Structure of a heterophilic adhesion complex between the human CD2 and CD58 (LFA-3) counterreceptors

Interaction between CD2 and its counterreceptor, CD58 (LFA-3), on opposing cells optimizes immune recognition, facilitating contacts between helper T lymphocytes and antigen-presenting cells as well as between cytolytic effectors and target cells. Here, we report the crystal structure of the heterophilic adhesion complex between the amino-terminal domains of human CD2 and CD58. A strikingly asymmetric, orthogonal, face-to-face interaction involving the major beta sheets of the respective immunoglobulin-like domains with poor shape complementarity is revealed. In the virtual absence of hydrophobic forces, interdigitating charged amino acid side chains form hydrogen bonds and salt links at the interface (approximately 1200 A2), imparting a high degree of specificity albeit with low affinity (K(D) of approximately microM). These features explain CD2-CD58 dynamic binding, offering insights into interactions of related immunoglobulin superfamily receptors.

Functional glycan-free adhesion domain of human cell surface receptor CD58: design, production and NMR studies

A general strategy is presented here for producing glycan-free forms of glycoproteins without loss of function by employing apolar-to-polar mutations of surface residues in functionally irrelevant epitopes. The success of this structure-based approach was demonstrated through the expression in Escherichia coli of a soluble 11 kDa adhesion domain extracted from the heavily glycosylated 55 kDa human CD58 ectodomain. The solution structure was subsequently determined and binding to its counter-receptor CD2 studied by NMR. This mutant adhesion domain is functional as determined by several experimental methods, and the size of its binding site has been probed by chemical shift perturbations in NMR titration experiments. The new structural information supports a 'hand-shake' model of CD2-CD58 interaction involving the GFCC'C" faces of both CD2 and CD58 adhesion domains. The region responsible for binding specificity is most likely localized on the C, C' and C" strands and the C-C' and C'-C" loops on CD58.

Epitope mapping of monoclonal antibodies against N-domain of carcinoembryonic antigen

Monoclonal antibodies (MoAbs) against N-domain of carcinoembryonic antigen (CEA), C249, K348, K1338, and K1444, that inhibit CEA-mediated cell adhesion, did not crossreact with nonspecific cross-reacting antigen (NCA). To determine amino acid sequences involved in the adhesion, epitopes of the MoAbs were mapped with recombinant NCAs carrying CEA-NCA chimeric N-domain. The data showed that the epitopes of C249, K1338, K1444 are located within the regions 1-32, 1-32, and 33-59 of CEA, respectively, and that two discrete regions 1-32 and 60-93 may be related to the epitope of K348. Comparison of amino acid sequences between CEA and NCA suggested that four residues (21, 27-29), eight residues (21, 27-29, 66, 78, 79, 89), and three residues (43, 44, 46) are important for recognition by C249 (or K1338), K348, and K1444, respectively. These residues seem to participate in the cell adhesion mediated by CEA.

A cdc15-like adaptor protein (CD2BP1) interacts with the CD2 cytoplasmic domain and regulates CD2-triggered adhesion

A human CD2 cytoplasmic tail-binding protein, termed CD2BP1, was identified by an interaction trap cloning method. Expression of CD2BP1 is restricted to hematopoietic tissue, being prominent in T and natural killer (NK) cells, with long (CD2BP1L) and short (CD2BP1S) variants arising by alternative RNA splicing. Both CD2BP1 molecules are homologous to Schizosaccharomyces pombe cdc15, and include a helical domain, variable length intervening PEST sequence and C-terminal SH3 domain. Although the CD2BP1 SH3 domain binds directly to the CD2 sequence, KGPPLPRPRV (amino acids 300-309), its association is augmented markedly by the CD2BP1 N-terminal segment. Upon ligand-induced clustering of surface CD2 molecules, CD2BP1 redistributes from a cytosolic to a surface membrane compartment, co-localizing with CD2. In turn, CD2-stimulated adhesion is downregulated by CD2BP1, apparently through coupling of the protein tyrosine phosphatase (PTP)-PEST to CD2. These findings offer the first molecular view into the control processes for T cell adhesion.

A Phase II Study of BTI-322, a Monoclonal Anti-CD2 Antibody, for Treatment of Steroid-Resistant Acute Graft-Versus-Host Disease

BTI-322, a rat monoclonal IgG2b directed against the CD2 antigen on T cells and natural killer (NK) cells, blocks primary and memory alloantigen proliferative responses in vitro. We have evaluated the pharmacokinetics and safety of BTI-322 during treatment of 20 transplant recipients with steroid-refractory acute graft-versus-host disease (GVHD). Treatment consisted of BTI-322 by intravenous (IV) bolus or 30-minute infusion at approximately 0.1 mg/kg/d for 10 days in addition to continuing high-dose steroids and tacrolimus or cyclosporine. Pharmacokinetic sampling was performed in 10 patients; the t1/2 ± SE was 9.1 ± 1.3 hours, the Cmaxwas 2,549 ± 291 ng/mL, the Vd was 3.97 ± 0.95 L, and the Vd/kg was 0.05 ± 0.01 L/kg. Ten patients experienced transient dyspnea sometimes accompanied by nausea, vomiting, diarrhea, and tachycardia shortly after the initial bolus dose of drug, but serious drug-related adverse events were not seen during the remainder of the infusions. At the end of treatment (day 11), there were six patients with complete responses and five with a reduction in grade of GVHD for a total response rate of 55% (95% confidence interval [CI], 32% to 77%). Antibodies targeting CD2 may be active in the treatment of acute GVHD, and evaluation of a humanized form of BTI-322 is warranted.

A Phase II Study of BTI-322, a Monoclonal Anti-CD2 Antibody, for Treatment of Steroid-Resistant Acute Graft-Versus-Host Disease

BTI-322, a rat monoclonal IgG2b directed against the CD2 antigen on T cells and natural killer (NK) cells, blocks primary and memory alloantigen proliferative responses in vitro. We have evaluated the pharmacokinetics and safety of BTI-322 during treatment of 20 transplant recipients with steroid-refractory acute graft-versus-host disease (GVHD). Treatment consisted of BTI-322 by intravenous (IV) bolus or 30-minute infusion at approximately 0.1 mg/kg/d for 10 days in addition to continuing high-dose steroids and tacrolimus or cyclosporine. Pharmacokinetic sampling was performed in 10 patients; the t1/2 ± SE was 9.1 ± 1.3 hours, the Cmaxwas 2,549 ± 291 ng/mL, the Vd was 3.97 ± 0.95 L, and the Vd/kg was 0.05 ± 0.01 L/kg. Ten patients experienced transient dyspnea sometimes accompanied by nausea, vomiting, diarrhea, and tachycardia shortly after the initial bolus dose of drug, but serious drug-related adverse events were not seen during the remainder of the infusions. At the end of treatment (day 11), there were six patients with complete responses and five with a reduction in grade of GVHD for a total response rate of 55% (95% confidence interval [CI], 32% to 77%). Antibodies targeting CD2 may be active in the treatment of acute GVHD, and evaluation of a humanized form of BTI-322 is warranted.

Engineering an intertwined form of CD2 for stability and assembly

The amino-terminal domain of CD2 has the remarkable ability to fold in two ways: either as a monomer or as an intertwined, metastable dimer. Here we show that it is possible to differentially stabilize either fold by engineering the CD2 sequence, mimicking random mutagenesis events that could occur during molecular evolution. Crystal structures of a hinge-deletion mutant, which is stable as an intertwined dimer, reveal domain rotations that enable the protein to further assemble to a tetramer. These results demonstrate that a variety of folds can be adopted by a single polypeptide sequence, and provide guidance for the design of proteins capable of further assembly.

Regulation of IL-4 expression by activation of individual alleles

To study the in vivo role of IL-4-expressing cells, we developed a strategy to tag these cells, by generating mice in which one IL-4 allele was replaced with a cDNA encoding the human CD2 (huCD2) cell-surface molecule. Expression of the huCD2 reporter was, like IL-4, restricted to the appropriately polarized T helper 2 cells. However, most of the cells expressed only the IL-4 or the targeted allele. Analysis of the frequency of monoallelic versus biallelic expression suggests that the activation of each individual allele is regulated by a stochastic process whose probability can be augmented by increasing the strength of signal delivered through the TCR. Allele-specific activation may be a general feature of cytokine regulation that contributes to the functional diversity within T helper cell subpopulations.

CD2 and the nature of protein interactions mediating cell-cell recognition

Rapid progress has recently been made in characterising the structures of leukocyte cell-surface molecules. Detailed analyses of the structure and interactions of CD2 were the first involving a molecule that has not been directly linked to antigen recognition in the manner of antigen receptors or co-receptors. It seems highly likely that the properties of ligand binding by CD2 are relevant to the general mechanisms of cell-cell recognition. As an example of biological recognition, the defining characteristic of cell-cell contact is that it involves the simultaneous interaction of hundreds, if not thousands, of molecules. Affinity and kinetic analyses of ligand binding by CD2 indicated that the protein interactions mediating cell-cell contact, whilst highly specific, are much weaker than initially anticipated, probably due to the requirement that such contacts be easily reversible. Simultaneously, in addressing the mechanism of this mode of recognition, structural and mutational studies focussed on the role of charged residues clustered in the ligand-binding face of CD2, yielding the concept that electrostatic complementarity, rather than surface-shape complementarity, is the dominant feature of specific, low-affinity protein recognition at the cell surface by CD2. The crystallographic analysis of the CD2-binding domain of CD58 strongly supports this concept.

Modulation of the CD2 receptor and not disruption of the CD2/CD48 interaction is the principal action of CD2-mediated immunosuppression in the rat

CD48, the murine homolog of human CD58, binds to CD2 in rats and mice. Whereas inhibition of CD2 signaling leads to profound immunosuppression, no information is available on CD48-targeted therapy in the rat. We could show that anti-CD2 treatment (OX34) efficiently inhibited TCR-driven as well as CD2-mediated proliferation, whereas blocking of ligand binding (OX45) remained completely uneffective. Inhibition of allogeneic MLR by OX45 turned out to be due to induction of unspecific suppressive mechanisms. In vivo, OX45 failed to prolong rat heart allograft survival in contrast to that seen with OX34. Grafts were rejected despite persistent and complete downmodulation of CD48 on lymphocytes without any cell depleting effect, rendering receptor/ligand interactions physically impossible. Combined application of CD2 and CD48 mAb did not enhance immunosuppression induced by CD2 mAb alone. Provided that there is no alternative CD2 ligand in the rat, we conclude that CD2-directed immunotherapy is mediated by suppressive events induced by modulation of the CD2 receptor ("negative signaling") rather than by mere disruption of the CD2-CD48 interaction.

The molecular structure of cell adhesion molecules

Considerable advances have been made in our knowledge of the molecular structure of cell adhesion molecules, their binding sites, and adhesion complexes. For the cadherins, protein zero, and CD2, additional experimental data support the insights obtained from structural analysis of their domains and molecular models of their adhesion complexes. For neural cell adhesion molecules, L1, fibronectin, tenascin-C, integrins, and vascular cell adhesion molecules, the molecular structure of domains, and in most cases their binding sites, have been elucidated. The substrate recognition sites in some of these molecules possess rate constants for association and dissociation that permit both rapid cell migration and, through avidity, high-affinity cell-cell interactions.

Differential interaction of the CD2 extracellular and intracellular domains with the tyrosine phosphatase CD45 and the zeta chain of the TCR/CD3/zeta complex

T cell activation via CD2 requires interaction of CD2 with several signaling molecules. To investigate the structural requirements for an association of CD2 with the tyrosine phosphatase CD45 and the zeta chain of the T cell receptor (TCR)/CD3/zeta complex, we have expressed in mouse EL4 T cells a series of human CD2 chimeric and mutant proteins. Chimeric proteins in which the CD2 transmembrane and/or cytoplasmic domains were deleted or exchanged with analogous regions of CD4, CD28 or CD58 retained association with high levels of murine CD45 phosphatase activity, suggesting that the CD2 extracellular domain largely controls interaction with CD45. To a lesser extent, the cytoplasmic domain of CD2 was also shown to interact with CD45, as demonstrated by an increase in co-immunoprecipitated phosphatase activity observed following replacement of the CD58 cytoplasmic domain with that of CD2. In contrast, the cytoplasmic domain of CD2 was found to be responsible for the majority of CD2 interaction with the zeta chain of the TCR/CD3/zeta complex. Deletion of the CD2 cytoplasmic domain, excluding the first three amino acids, removed virtually all CD2 associated zeta chain and approximately sevenfold higher levels of zeta chain were found in association with a CD58/58/2 chimera than with control human CD58 wild type. This study suggests that the CD2 extracellular and intracellular domains are differentially involved in regulating T cell activation through interaction with the tyrosine phosphatase CD45 and the zeta chain of the TCR/CD3/zeta complex.

Short course single agent therapy with an LFA-3-IgG1 fusion protein prolongs primate cardiac allograft survival

The interaction of T cell costimulatory molecules with their ligands is required for optimal T cell activation. Interference with such interactions can induce antigen unresponsiveness and delay xeno- and allograft rejection. We have previously shown that LFA3TIP, a soluble human lymphocyte function-associated antigen (LFA)-3 construct, binds CD2 and inhibits responses of human T cells in vitro. This study reports the first use of a human fusion protein, LFA3TIP, to significantly prolong primate cardiac allograft survival. Based on our observations that LFA3TIP inhibits baboon allogeneic mixed lymphocyte reactions, we gave baboon recipients of heterotopic cardiac allografts injections of LFA3TIP, 3 mg/kg i.v., for 12 consecutive days, starting 2 days before transplantation. This regimen delayed graft rejection from an average of 10.6 +/- 2.3 days for human IgG-treated controls (n = 5) to an average of 18.0 +/- 5.3 days for LFA3TIP-injected animals (n = 7; P < or = 0.01). Grafts from LFA3TIP-treated animals showed markedly diminished coronary endothelialitis as compared with control animals. LFA3TIP reached peak serum levels of approximately 100 micrograms/ml after 7-9 injections and persisted in the 10-micrograms/ml range for 1 to 2 weeks after the final injection. Despite these blood levels, circulating antibodies to LFA3TIP were not detected in the serum. No renal or hepatic toxicity was noted. The possible mechanism by which LFA3TIP acts to inhibit graft rejection is discussed; success in prolonging graft survival when LFA3TIP is used as a single-agent therapy suggests great potential for this novel therapeutic agent.

Directed evolution studies with combinatorial libraries of T4 lysozyme mutants

Gene duplication with divergence to new functions has been an important mechanism in protein evolution. However, the questions of how many new functions can arise from a particular ancestral gene and how many mutational steps are typically required to generate new functions have been difficult to approach experimentally. We have addressed these questions using T4 lysozyme as a model system by synthesizing two combinatorial libraries of > 10(7) mutant T4 lysozyme genes: one library with an average of 14 missense mutations spread throughout the gene and one library in which 13 active site residues have been simultaneously randomized. These libraries were placed under selection in lacZ or pheA deficient strains of E. coli to investigate whether they sample sufficient diversity to contain mutants with acquired beta-galactosidase or prephenate dehydratase activities. Although neither selection yielded T4 lysozyme mutants with these new activities, a novel E. coli locus was cloned that weakly complements these mutants, allowing them to form 1 mm colonies in 4-6 weeks. This growth rate corresponds to a turnover number of approximately 1000 or 25 min-1 for the lacZ or pheA complementation systems, respectively, thus defining the limits of evolved enzymatic activity detectable in these selections. Thus, the strong selective pressure uncovered an unexpected solution to the biochemical blocks, a frequently observed phenomenon in selection experiments. The characterization of this locus will allow its elimination from future E. coli complementation schemes.

Visualization of CD2 interaction with LFA-3 and determination of the two-dimensional dissociation constant for adhesion receptors in a contact area

Many adhesion receptors have high three-dimensional dissociation constants (Kd) for counter-receptors compared to the KdS of receptors for soluble extracellular ligands such as cytokines and hormones. Interaction of the T lymphocyte adhesion receptor CD2 with its counter-receptor, LFA-3, has a high solution-phase Kd (16 microM at 37 degrees C), yet the CD2/LFA-3 interaction serves as an effective adhesion mechanism. We have studied the interaction of CD2 with LFA-3 in the contact area between Jurkat T lymphoblasts and planar phospholipid bilayers containing purified, fluorescently labeled LFA-3. Redistribution and lateral mobility of LFA-3 were measured in contact areas as functions of the initial LFA-3 surface density and of time after contact of the cells with the bilayers. LFA-3 accumulated at sites of contact with a half-time of approximately 15 min, consistent with the previously determined kinetics of adhesion strengthening. The two-dimensional Kd for the CD2/LFA-3 interaction was 21 molecules/microns 2, which is lower than the surface densities of CD2 on T cells and LFA-3 on most target or stimulator cells. Thus, formation of CD2/LFA-3 complexes should be highly favored in physiological interactions. Comparison of the two-dimensional (membrane-bound) and three-dimensional (solution-phase) KdS suggest that cell-cell contact favors CD2/LFA-3 interaction to a greater extent than that predicted by the three-dimensional Kd and the intermembrane distance at the site of contact. LFA-3 molecules in the contact site were capable of lateral diffusion in the plane of the phospholipid bilayer and did not appear to be irreversibly trapped in the contact area, consistent with a rapid off-rate. These data provide insights into the function of low affinity interactions in adhesion.

Hydrophobicity engineering to facilitate surface display of heterologous gene products on Staphylococcus xylosus

Protein engineering has been employed to investigate the effect of specific amino acid changes on the targeting of heterologous proteins to the outer cell surface of the Gram-positive bacterium Staphylococcus xylosus. Three different variants, corresponding to a 101 amino acid region of the major glycoprotein (G protein) of human respiratory syncytial virus (RSV), were generated in which multiple hydrophobic phenylalanine residues were either substituted or deleted. The different G protein fragments were expressed as one part of recombinant receptors designed for surface display on S. xylosus cells. The engineered variants of the RSV G protein hybrid receptors were, in contrast to a non-engineered fragment, efficiently targeted to the outer cell surface of recombinant S. xylosus cells as determined by different methods, including fluorescence-activated cell sorting. In addition, immunization of mice with live recombinant S. xylosus demonstrated that surface exposure was required to generate receptor-specific antibodies. The present strategy of hydrophobic engineering should be of general interest in surface-display applications and for secretion of proteins otherwise difficult to translocate through host cell membranes.

One sequence, two folds: a metastable structure of CD2

When expressed as part of a glutathione S-transferase fusion protein the NH2-terminal domain of the lymphocyte cell adhesion molecule CD2 is shown to adopt two different folds. The immunoglobulin superfamily structure of the major (85%) monomeric component has previously been determined by both x-ray crystallography and NMR spectroscopy. We now describe the structure of a second, dimeric, form present in about 15% of recombinant CD2 molecules. After denaturation and refolding in the absence of the fusion partner, dimeric CD2 is converted to monomer, illustrating that the dimeric form represents a metastable folded state. The crystal structure of this dimeric form, refined to 2.0-A resolution, reveals two domains with overall similarity to the IgSF fold found in the monomer. However, in the dimer each domain is formed by the intercalation of two polypeptide chains. Hence each domain represents a distinct folding unit that can assemble in two different ways. In the dimer the two domains fold around a hydrophilic interface believed to mimic the cell adhesion interaction at the cell surface, and the formation of dimer can be regulated by mutating single residues at this interface. This unusual misfolded form of the protein, which appears to result from inter- rather than intramolecular interactions being favored by an intermediate structure formed during the folding process, illustrates that evolution of protein oligomers is possible from the sequence for a single protein domain.

A human monoclonal antibody to a human self-antigen, CD2 derived from human peripheral blood lymphocytes engrafted in SCID mice

To establish human hybridoma lines, production of human immunoglobulin (Ig) and behavior of the implanted human peripheral blood lymphocytes (PBL) were characterized in severe combined immunodeficiency (SCID) mice. Human PBL from healthy donors were injected into the peritoneal cavity of SCID mice, and they were immunized with self-antigen, CD2. CD45+ cells (human PBL) migrated to lymphoid tissues in the mice as early as 4 days, accounting for more than half the lymph node cells and thymocytes. The number of cells releasing human IgG specific to the antigen increased 3.5 weeks after immunization without the usual constraint that production of the IgG, an autoantibody, is prohibited by immunological tolerance in humans. Therefore, we established several human hybridomas secreting human IgG to CD2, since splenocytes and lymph node cells from the implanted SCID mice at 3.5 weeks were fused with a human B lymphoblastoid cell line. A human anti-CD2 monoclonal antibody (MAb) was confirmed to bind to natural CD2 on human T cells by flow cytometric analysis. The epitope for the MAb was identical with a portion that the ligand LFA-3 binded, so that the MAb might reduce the inflammatory reaction caused by preventing activation of human T cells. Here, we report that the human immune system could be reconstituted in SCID mice to develop human hybridomas producing human MAb to a human self-antigen.

Mapping the homotypic binding sites in CD31 and the role of CD31 adhesion in the formation of interendothelial cell contacts

CD31 is a member of the immunoglobulin superfamily consisting of six Ig-related domains. It is constitutively expressed by platelets, monocytes, and some lymphocytes, but at tenfold higher levels on vascular endothelial cells. CD31 has both homotypic and heterotypic adhesive properties. We have mapped the homotypic binding sites using a deletion series of CD31-Fc chimeras and a panel of anti-CD31 monoclonal antibodies. An extensive surface of CD31 is involved in homotypic binding with domains 2 and 3 and domains 5 and 6 playing key roles. A model consistent with the experimental data is that CD31 on one cell binds to CD31 on an apposing cell in an antiparallel interdigitating mode requiring full alignment of the six domains of each molecule. In addition to establishing intercellular homotypic contacts. CD31 binding leads to augmented adhesion via beta 1 integrins. The positive cooperation between CD31 and beta 1 integrins can occur in heterologous primate cells (COS cells). The interaction is specific to both CD31 and beta 1 integrins. Neither intercellular adhesion molecule-1 (ICAM-1)/leukocyte function-associated antigen-1 (LCAM-1) nor neural cell adhesion molecule (NCAM)/NCAM adhesion leads to recruitment of beta 1 integrin adhesion pathways. Establishment of CD31 contacts have effects on the growth and morphology of endothelial cells. CD31(D1-D6)Fc inhibits the growth of endothelial cells in culture. In addition, papain fragments of anti-CD31 antibodies (Fab fragments) disrupt interendothelial contact formation and monolayer integrity when intercellular contacts are being formed. The same reagents are without effect once these contacts have been established, suggesting that CD31-CD31 interactions are critically important only in the initial phases of intercellular adhesion.

A proposed structural model of domain 1 of fasciclin III neural cell adhesion protein based on an inverse folding algorithm

Fasciclin III is an integral membrane protein expressed on a subset of axons in the developing Drosophila nervous system. It consists of an intracellular domain, a transmembrane region, and an extracellular region composed of three domains, each predicted to form an immunoglobulin-like fold. The most N-terminal of these domains is expected to be important in mediating cell-cell recognition events during nervous system development. To learn more about the structure/function relationships in this cellular recognition molecule, a model structure of this domain was built. A sequence-to-structure alignment algorithm was used to align the protein sequence of the fasciclin III first domain to the immunoglobulin McPC603 structure. Based on this alignment, a model of the domain was built using standard homology modeling techniques. Side-chain conformations were automatically modeled using a rotamer search algorithm and the model was minimized to relax atomic overlaps. The resulting model is compact and has chemical characteristics consistent with related globular protein structures. This model is a de novo test of the sequence-to-structure alignment algorithm and is currently being used as the basis for mutagenesis experiments to discern the parts of the fasciclin III protein that are necessary for homophilic molecular recognition in the developing Drosophila nervous system.

Intracellular mediators regulate CD2 lateral diffusion and cytoplasmic Ca2+ mobilization upon CD2-mediated T cell activation

CD2 is a T cell surface glycoprotein that participates in T cell adhesion and activation. These processes are dynamically interrelated, in that T cell activation regulates the strength of CD2-mediated T cell adhesion. The lateral redistribution of CD2 and its ligand CD58 (LFA-3) in T cell and target membranes, respectively, has also been shown to affect cellular adhesion strength. We have used the fluorescence photobleaching recovery technique to measure the lateral mobility of CD2 in plasma membranes of resting and activated Jurkat T leukemia cells. CD2-mediated T cell activation caused lateral immobilization of 90% of cell surface CD2 molecules. Depleting cells of cytoplasmic Ca2+, loading cells with dibutyric cAMP, and disrupting cellular microfilaments each partially reversed the effect of CD2-mediated activation on the lateral mobility of CD2. These intracellular mediators apparently influence the same signal transduction pathways, because the effects of the mediators on CD2 lateral mobility were not additive. In separate experiments, activation-associated cytoplasmic Ca2+ mobilization was found to require microfilament integrity and to be negatively regulated by cAMP. By directly or indirectly controlling CD2 lateral diffusion and cell surface distribution, cytoplasmic Ca2+ mobilization may have an important regulatory role in CD2 mediated T cell adhesion.

Localization of multiple functional domains on human PECAM-1 (CD31) by monoclonal antibody epitope mapping

PECAM-1, a cell adhesion molecule of the immunoglobulin gene (Ig) superfamily, has been implicated in white cell transmigration, integrin activation on lymphocytes, and cell-cell adhesion. The purpose of this investigation was to identify specific regions of the PECAM-1 extracellular domain mediating these functions by identifying the location of epitopes of bioactive anti-PECAM-1 monoclonal antibodies. The binding regions of mAbs important in PECAM-1-mediated leukocyte transmigration (Hec 7.2 and 3D2) were mapped to N-terminal Ig-like domains. The epitopes of monoclonal antibodies that activated integrin function on lymphocytes were dispersed over the entire extracellular region, but those that had the strongest activating effect were preferentially localized to the N-terminus of the molecule. The binding regions of mAbs that blocked PECAM-1-mediated heterophilic L-cell aggregation were located either in Ig-like domain 2 (NIH31.4) or Ig-like domain 6 (4G6 and 1.2). Site-directed mutagenesis further pinpointed the epitope of the 4G6 mAb to a hexapeptide, CAVNEG, within Ig-like domain 6. These results demonstrate that PECAM-1 contains multiple functional domains. Regions within N-terminal Ig-like domains appear to be required for transmigration. In contrast, two distinct regions were implicated in L-cell mediated heterophilic aggregation.

Ligand binding by the immunoglobulin superfamily recognition molecule CD2 is glycosylation-independent

The evolutionary success of the immunoglobulin superfamily (IgSF) is thought to reflect the ability of IgSF protein domains to form stable structural units. The role of glycosylation in stabilizing these domains is controversial, however. In this study a systematic analysis of the effect of glycosylation on the ligand-binding properties of the cell-cell recognition molecule CD2, which consists of two IgSF domains, was undertaken. A form of human soluble CD2 (hsCD2) with single N-acetylglucosamine residues at each glycosylation site was produced by inhibiting glucosidase I with N-butyldeoxynojirimycin during expression in Chinese hamster ovary cells and digesting the expressed hsCD2 with endoglycosidase H. The ligand and antibody binding properties of this form of hsCD2 were indistinguishable from those of fully glycosylated hsCD2 as determined by surface plasmon resonance analyses. The protein also formed diffraction quality crystals and analysis of the 2.5-A resolution crystal structure indicated that the single N-acetylglucosamine residue present on domain 1 is unlikely to stabilize the ligand binding face of hsCD2. A second, fully deglycosylated form of hsCD2 also bound the ligand and antibodies although this form of the protein tended to aggregate. In contrast to the results of previous studies, the current data indicate that the structural integrity and ligand binding function of human CD2 are glycosylation-independent.

A novel mutagenesis strategy identifies distantly spaced amino acid sequences that are required for the phosphorylation of both the oligosaccharides of procathepsin D by N-acetylglucosamine 1-phosphotransferase

A novel combinatorial mutagenesis strategy (shuffle mutagenesis) was developed to identify sequences in the propiece and amino lobe of cathepsin D which direct oligosaccharide phosphorylation by UDP-GlcNAc:lysosomal enzyme N-acetylglucosamine 1-phosphotransferase. Propiece restriction fragments and oligonucleotide cassettes corresponding to 13 regions of the cathepsin D and glycopepsinogen amino lobes were randomly shuffled together to generate a large library of chimeric molecules. The library was inserted into an expression vector encoding the carboxyl lobe of cathepsin D with a carboxyl-terminal myc epitope and a CD8 transmembrane extension. Transfected COS1 cells expressing the membrane-anchored forms of the cathepsin D/glycopepsinogen chimeras at the cell surface were selected with solid phase mannose 6-phosphate receptor or an antibody to the myc epitope. Plasmids were rescued in Escherichia coli and sequenced by hybridization to the original oligonucleotide cassettes. Two regions of the cathepsin D amino lobe (segments 7 and 12) were found to contribute to proper folding, surface expression, and selective phosphorylation of the carboxyl lobe oligosaccharide. Two different cathepsin D regions (the propiece and segment 5) cooperated with a previously identified recognition element in the carboxyl lobe to allow efficient phosphorylation of both the amino and carboxyl lobe oligosaccharides. Three general models for extending the catalytic reach of N-acetylglucosamine 1-phosphotransferase to widely spaced oligosaccharides are presented.

Topology of the CD2-CD48 cell-adhesion molecule complex: implications for antigen recognition by T cells

BACKGROUND

The T-lymphocyte cell-surface molecule, CD2, was the first heterophilic cell-adhesion molecule to be discovered and has become an important paradigm for understanding the structural basis of cell adhesion. Interaction of CD2 with its ligands. CD58 (in humans) and CD48 (in mice and rats), contributes to antigen recognition by T cells. CD2, CD48 and CD58 are closely related members of the immunoglobulin superfamily and their extracellular regions are predicted to have very similar structures. The three-dimensional crystal structure of this region of CD2 has been determined, revealing two immunoglobulin domains with the ligand-binding site situated on an exposed beta sheet in the membrane-distal domain. This GFCC'C" beta sheet is also involved in a homophilic 'head-to-head' interaction in the CD2 crystal lattice, which has been proposed to be a model for the interactions of CD2 with its ligands.

RESULTS

We show that the CD2-binding site on rat CD48 lies on the equivalent beta-sheet of its membrane-distal immunoglobulin domain. By making complementary mutations, we have shown that two charged residues in the CD48 ligand-binding site interact directly with two oppositely charged residues in CD2's ligand-binding site. These results indicate that the amino-terminal immunoglobulin domains of CD2 and CD48 bind each other in the same orientation as the CD2-CD2 crystal lattice interaction, strongly supporting the suggestion that CD2 interacts head-to-head with its ligand. Modelling CD48 onto the CD2 structure reveals that the CD2-CD48 complex spans approximately the same distance (134 A) as predicted for the complex between the T-cell receptor and the peptide-bound major histocompatibility complex (MHC) molecule.

CONCLUSIONS

Our results, together with recent structural studies of CD2, provide the first indication of the specific topology of a cell-adhesion molecule complex. The similar dimensions predicted for the CD2-CD48 complex and the complex between the T-cell receptor and the peptide-bound MHC molecule suggest that one of the functions of CD2 may be to position the plasma membranes of the T cell and the antigen-presenting (or target) cell at the optimal distance for the low-affinity interaction between the T-cell receptor and the peptide-bound MHC molecule.

Amino acid residues required for binding of lymphocyte function-associated antigen 3 (CD58) to its counter-receptor CD2

Efficient activation and regulation of the cellular immune response requires engagement of T cell accessory molecules as well as the antigen-specific T cell receptor. The lymphocyte function-associated antigen (LFA) 3 (CD58)/CD2 accessory pathway, one of the first discovered, has been extensively characterized in terms of structure and function of the CD2 molecule, which is present on all T lymphocytes and natural killer cells of the human immune system. The binding site of human CD2 for LFA-3 has been localized to two epitopes on one face of the first immunoglobulin (Ig)-like domain of this two-domain, Ig superfamily molecule. Human LFA-3 is genetically linked and is 21% identical in amino acid sequence to CD2, suggesting that this adhesive pair may have evolved from a single ancestral molecule. We have aligned the amino acid sequences of LFA-3 and CD2 and mutagenized selected amino acids in the first domain of LFA-3 that are analogous to those implicated in the binding site of CD2. The data show that K30 and K34, in the predicted C-C' loop, and D84, in the predicted F-G loop of LFA-3, are involved in binding to CD2, suggesting that two complementary sites on one face of the first domain of each molecule bind to each other.

Interaction between human CD2 and CD58 involves the major beta sheet surface of each of their respective adhesion domains

The CD58 binding site on human CD2 was recently shown by nuclear magnetic resonance structural data in conjunction with site-directed mutagenesis to be a highly charged surface area covering approximately 770A2 on the major AGFCC'C" face of the CD2 immunoglobulin-like (Ig-like) NH2-terminal domain. Here we have identified the other binding surface of the CD2-CD58 adhesion pair by mutating charged residues shared among CD2 ligands (human CD58, sheep CD58, and human CD48) that are predicted to be solvent exposed on a molecular model of the Ig-like adhesion domain of human CD58. This site includes beta strand residues along the C strand (E25, K29, and K30), in the middle of the C' strand (E37) and in the G strand (K87). In addition, several residues on the CC' loop (K32, D33, and K34) form this site. Thus, the interaction between CD2 and CD58 involves the major beta sheet surface of each adhesion domain. Possible docking orientations for the CD2-CD58 molecular complex are offered. Strict conservation of human and sheep CD58 residues within the involved C and C' strands and CC' loop suggests that this region is particularly important for stable formation of the CD2-CD58 complex. The analysis of this complex offers molecular insight into the nature of a receptor-ligand pair involving two Ig family members.

CD2 is involved in maintenance and reversal of human alloantigen-specific clonal anergy

Induction and maintenance of a state of T cell unresponsiveness to specific alloantigen would have significant implications for human organ transplantation. Using human histocompatibility leukocyte antigen DR7-specific helper T cell clones, we demonstrate that blockade of the B7 family of costimulatory molecules is sufficient to induce alloantigen-specific T cell clonal anergy. Anergized cells do not respond to alloantigen and a variety of costimulatory molecules, including B7-1, B7-2, intercellular adhesion molecule-1 (ICAM-1), and lymphocyte function-associated molecule (LFA)-3. However, after culture in exogenous interleukin (IL)-2 for at least 7 d, anergized cells can respond to alloantigen in the presence of LFA-3. LFA-3 costimulation subsequently restores responsiveness to alloantigen in the presence of previously insufficient costimulatory signals. Expression of CD2R epitope is downregulated on anergic cells and is restored after 7 d of IL-2 culture. The loss of the CD2R is temporally associated with the inability of anergized cells to respond to LFA-3. These results suggest that in addition to blockade of B7 family members, inhibition of CD2 and, potentially, other costimulatory pathways that might reverse anergy will be necessary to maintain prolonged alloantigen-specific tolerance.

Human cell-adhesion molecule CD2 binds CD58 (LFA-3) with a very low affinity and an extremely fast dissociation rate but does not bind CD48 or CD59

CD2 is a T lymphocyte cell-adhesion molecule (CAM) belonging to the immunoglobulin superfamily (IgSF) which mediates transient adhesion of T cells to antigen-presenting cells and target cells. Reported ligands for human CD2 include the structurally-related IgSF CAMs CD58 (LFA-3) and CD48 as well as, more controversially, the unrelated cell-surface glycoprotein CD59. Using surface plasmon resonance technology, which avoids several pitfalls of conventional binding assays, we recently reported that rat CD2 binds rat CD48 with a very low affinity (Kd 60-90 microM) and dissociates rapidly (koff > or = 6 s-1) [van der Merwe, P. A., Brown, M. H., Davis, S. J., & Barclay, A. N. (1993) EMBO J. 12, 4945-4954]. In contrast, a study using conventional equilibrium binding methods reported a much higher affinity (Kd 0.4 microM) for human CD2 binding CD58 which suggested that the weak binding of rat CD2 to CD48 may not represent a typical CAM interaction. In the present study we have used surface plasmon resonance to obtain definitive affinity and kinetic data on the interactions of a soluble, recombinant form of human CD2 with soluble forms of CD58, CD48, and CD59. Binding of CD2 to CD58 was readily detected but we were unable to detect any direct interaction between CD2 and either CD59 or CD48 under conditions in which very low affinity interactions (Kd approximately 0.5 mM) would have been detected. In contrast to previous reports we found that human CD2 bound CD58 with a very low affinity (Kd 9-22 microM) and dissociated with an extremely fast dissociation rate constant (koff > or = 4 s-1). The association rate constant (kon) could not be measured directly but was calculated to be > or = 400,000 M-1s-1. Taken together, these results provide conclusive evidence that CAM interactions can have very low affinities and extremely fast dissociation rate constants.

Crystal structure of the extracellular region of the human cell adhesion molecule CD2 at 2.5 A resolution

BACKGROUND

The T-lymphocyte antigen CD2 is an adhesion molecule implicated in immune responses in vivo. The extracellular regions of the human and rat homologues of CD2 share only 45% sequence identity and bind different protein ligands. Comparison of the human and rat soluble CD2 (sCD2) structures should provide insights into the structural basis of cell surface recognition.

RESULTS

We therefore determined the crystal structure of a form of human sCD2 with single N-acetylglucosamine residues at each glycosylation site to 2.5 A resolution with an R-factor of 19.3%. It is composed of two immunoglobulin superfamily domains similar to those of rat sCD2, but the relative orientation of the domains in the two homologues differs by up to 20 degrees. An interaction involving the flat, highly charged, ligand binding GFCC'C" faces of crystallographically related human sCD2 molecules duplicates, in a different lattice, that observed in the rat sCD2 crystals.

CONCLUSIONS

Intramolecular flexibility appears to be a conserved feature of CD2. The head-to-head interaction between molecules represents a general model for interactions between adhesion molecules of this structural class. Ligand specificity may be influenced by the distribution of charged residues on the binding face.

The DNA binding domain of retinoic acid receptor beta is required for ligand-dependent suppression of proliferation. Application of general purpose mammalian coexpression vectors

We have developed a family of mammalian coexpression vectors that permit identification of living or fixed cells overexpressing a gene of interest by surrogate detection of a coexpressed marker protein. Using these ‘pMARK’ vectors, a fluorescence-based, single cell proliferation assay was developed and used to study the effect of retinoic acid receptor beta (RAR-beta) on cell cycling. We demonstrate that transient overexpression of RAR-beta in the presence, but not absence, of all-trans retinoic acid results in a dramatic suppression of cell proliferation. We further show that this effect requires the DNA binding (C) domain of RAR-beta. It has been previously shown that RAR-beta expression is markedly altered in a variety of neoplasms and cell lines. Our data support the hypothesis that loss of RAR-beta may contribute to tumor progression by removing normal restraints on proliferation. The pMARK vectors should be useful for studying other genes that putatively suppress or enhance proliferation.

Expression cloning of an equine T-lymphocyte glycoprotein CD2 cDNA. Structure-based analysis of conserved sequence elements

An equine CD2 cDNA has been isolated by monoclonal antibody screening of a T-lymphocyte cDNA library. The cDNA contained an open reading frame of 1041 bp encoding a translated product of 347 amino acids. Northern blotting analysis revealed a single mRNA species expressed in spleen, thymus and activated peripheral lymphocytes. The predicted amino acid sequence has 50-65% identity with the human, rat and mouse CD2 sequences with greatest similarity shared with the human homologue. Evolutionarily conserved structural and functional domains in CD2 were identified by comparing the sequences of the equine, human, mouse and rat CD2 homologues in the context of the recently derived crystal structure of rat soluble CD2 [Jones, E. Y., Davis, S. J., Williams, A. F., Harlos, K. & Stuart, D. I. (1992) Nature 360, 232-239]. The key conserved features of the extracellular region included core residues necessary to preserve the structural integrity of the molecule, residues in the linker region likely to maintain the unique domain organization of CD2, an array of highly charged residues in the putative ligand-binding face of the molecule and glycosylation-signal distributions that render the putative ligand-binding GFCC'C" face of domain 1 relatively unhindered by glycosylation.