Binding of monoclonal antibodies inhibits dissociation but not exchange of HLA and beta 2-microglobulin

Intrabody-mediated phenotypic knockout of major histocompatibility complex class I expression in human and monkey cell lines and in primary human keratinocytes

Cultured keratinocyte allografts from unrelated donors can be readily grown as sheets in large-scale cell culture and have been used as an immediate skin cover for severely burned patients. Despite the absence of passenger leukocytes and the unlimited amount of material that can be obtained for permanent skin coverage, the allografts are susceptible to rejection. Since MHC class I (MHCI) antigens serve as targets for allograft rejection, we investigated whether 'phenotypic knockout' of human MHCI could be achieved through expression of an ER-directed anti-human MHCI single-chain intrabody (sFvhMHCI) that is directed against a monomorphic, conformational epitope, expressed across species lines, on the MHCI heavy chain. Co-immunoprecipitation of both MHCI heavy chain and beta2-microglobulin occurred in transfected monkey COS-1 cells, while Jurkat T cells stably expressing the ER-directed sFvhMHCI intrabody showed that complete phenotypic knockout of MHCI cell surface expression could be achieved. Infection of several human cell lines of divergent tissue sources and different HLA haplotypes resulted in marked down-regulation of MHCI expression, even under conditions where inflammatory cytokines (eg gamma-IFN) which up-regulate MHCI expression were used. Finally, when adenovirus encoding the anti-human MHCI intrabody was used to transduce primary human keratinocytes, a marked reduction of surface MHCI expression was observed. These in vitro studies set the groundwork for in vivo studies to determine if intrabody-mediated knockout of MHCI can impair alloantigen expression and prolong the survival of keratinocyte allografts.

Dissociation of beta 2-microglobulin leads to the accumulation of a substantial pool of inactive class I MHC heavy chains on the cell surface

A large pool of free class I heavy chains is detected in situ on the plasma membrane of living cells. These chains are present on cells of different MHC genotypes and appear to exist under physiological conditions in vivo. These molecules arise from the dissociation of previously assembled class I heterodimers at the cell surface. The ratio of intact to dissociated heterodimers is strongly affected by the occupancy of the peptide-binding site of the class I molecule. Upon dissociation of the heterodimer, the class I molecule is functionally inactive. These findings may help to explain why class I molecules on the cell surface are unreceptive to binding peptides yet readily associate with peptides in the presence of exogenous beta 2-microglobulin. These results have implications for understanding the distinct functions of class I versus class II molecules and how the immunological identity of cells is preserved.

Excess beta 2 microglobulin promoting functional peptide association with purified soluble class I MHC molecules

T lymphocytes expressing alpha beta receptors recognize antigenic peptide fragments bound to major histocompatibility complex class I or class II molecules present on the surface membranes of other cells. Peptide fragments are present in the two available HLA crystal structures and recent data indicate that peptide is required for the stable folding of the class I heavy chain and maintenance of its association with the class I light chain, beta 2-microglobulin (beta 2m), at physiological temperature. To explain how the exogenous peptide used to create targets for cytotoxic cells bearing CD8 antigen could associate with apparently peptide-filled extracellular class I molecules, we hypothesized that stable binding of exogenous peptide to mature class I molecules reflects either the replacement of previously bound peptide during the well documented beta 2m exchange process or the loading of 'empty' class I heavy chains dependent on the availability of excess beta 2m. In either case, free beta 2m should enhance peptide/class I binding. Using either isolated soluble class I molecules or living cells, we show here that free purified beta 2m markedly augments the generation of antigenic complexes capable of T-cell stimulation.

A unique epitope on the CD2 molecule defined by the monoclonal antibody 9-1: epitope-specific modulation of the E-rosette receptor and effects on T-cell functions

Monoclonal antibody (MoAb) 9-1 shows unique binding properties to CD2 resulting in peculiar epitope specific changes. 9-1 shows competitive binding with MoAb to D66 epitope, and gives a similar staining pattern with T-cell populations, at a low density on resting T cells, and high density on thymocytes and activated T cells. However, 9-1 has the opposite effect on anti-D66 MoAbs on rosette formation, namely, 9-1 increases the stability of rosettes, but 9-1 plus anti-mouse Ig bound to T-cell surface blocks rosettes. 9-1 plus anti-mouse Ig, like anti-D66 MoAbs, induces further appearance of D66 and 9-1 epitopes but, contrary to anti-D66, induces appearance of T11(3) epitopes. Thus, binding 9-1 results in unique "epitope-specific modulation" events that are not solely artificial, but appear to mimic events naturally occurring during T-cell differentiation/activation. The effects of binding 9-1 on T-cell functions also display peculiarities. 9-1, like anti-D66 MoAbs, activates T cells when added in combination with anti-9.6/T11(1) MoAbs but not with anti-T11(3). To obtain full activation, monocytes are required; however, adding 9-1 alone do not inhibit specific T-cell cytotoxicity contrary to anti-D66 or anti-9.6/T11(1), although 9-1 inhibits NK activity of peripheral cells. Given the apparent complexities of the functions exerted by CD2, these data show that definite conformational changes or reorientation, which would be naturally produced by soluble and/or cell surface ligand(s), would be key events in determining how CD2 will influence T-cell functions.

Subunit interactions of class I histocompatibility antigens

The kinetics of dissociation of iodinated beta 2-microglobulin (beta 2m) from the papain-solubilized class I histocompatibility antigen HLA-B7 have been investigated. In the presence of unlabeled beta 2m, most of the HLA dissociates according to a single rate constant, whereas in the absence of unlabeled beta 2m, the system approaches an equilibrium dependent upon the initial HLA concentration. When iodinated beta 2m is incubated with unlabeled HLA-B7, the rate of incorporation of beta 2m into the complex is much less dependent on the concentration than is expected for a simple association/dissociation system; instead, the system behaves as if the "activity" (in a thermodynamic sense) of the HLA heavy-chain intermediate cannot surpass a critical concentration. The dissociation rate for each class I specificity is a function of temperature, ionic strength, pH, and the status of the heavy chain (papain solubilized vs. detergent solubilized). High temperature, high ionic strength, and extremes of pH promote dissociation. The intact molecule dissociates about 10 times more slowly than the papain-solubilized molecule. In contrast, the rate of dissociation of all papain-solubilized class I antigens tested falls within the range of about a factor of 2. The presence of the carbohydrate has no effect on the rate of dissociation. The possibility that HLA class I antigen dissociation may occur in vivo within acidic internal vesicles is discussed.

On the heterologous interaction between beta 2-microglobulin and the heavy chain of rat major histocompatibility complex class 1 antigens

The heterologous interaction between beta 2-microglobulin (beta 2m) and rat major histocompatibility complex (MHC) (RT1) antigens was measured in a two-step binding assay consisting of binding of radiolabelled beta 2m to RT1 antigens and immunoprecipitation of beta 2m-RT1 antigen complexes with RT1 antisera. The effects of varying the concentrations of the three reactants involved were studied. The molecular events taking place in the two steps were analysed by gel chromatography. The beta 2m-RT1 antigen complex had the apparent size of albumin and reacted completely with specific alloantisera. RT1 antigens prepared from Wistar/Furth (RT1u) and Brown Norway (RT1n), respectively, both effectively bound heterologous beta 2m. The times for association and dissociation, respectively, at 37 degrees C, were of the same order, but dissociation was slightly slower. Association was markedly temperature-dependent and was considerably slower at low temperatures. All these processes were slower for RT1n than for RT1u antigens. The association constant for the interaction between RT1u antigens and 125I-human beta 2m was estimated by Scatchard analysis to be about 10(9) M-1. Contribution to the heterologous interaction by products from various rat MHC subloci (A, B, and C) was investigated by the introduction of sublocus-specific antisera in step 2. The reaction apparently involved neither class 2 antigens (sublocus B) nor the presumed rat Qa homologue (sublocus C). Classical class 1 antigens (sublocus A) clearly contributed to the binding. However, a monoclonal antibody against products from rat MHC class 1 genes only precipitated less than half of the RT1 antigen-complexed beta 2m. Thus, at least two RT1u class 1 alloantigen molecules seem to participate in the reaction. This, in turn, indicates that the rat genome may contain multiple class 1 genes, as is the case for most other mammals investigated.