The phenotype of H-2M-deficient mice is dependent on the MHC class II molecules expressed.

For a broader view of the role of H-2M as an accessory molecule in antigen presentation, we investigated the degree to which different MHC class II isotypes and alleles depend on H-2M to function in vivo. We generated H-2M-deficient animals expressing Ek/b or Ak molecules in addition to the Ab molecules already present in the mutant strain, and compared the ability of the different MHC class II molecules to present antigen at the cell surface for recognition by T cells, and contribute to positive selection of CD4+ T cells in the thymus. Biochemical analyses were performed to assess MHC class II maturation, and to determine the peptide content of the molecules. In the absence of H-2M, Ek/b molecules contained a more heterogeneous set of class II-associated invariant chain peptides (CLIP) than Ab did, which, unlike Ab-CLIP complexes, were not SDS-stable. Unlike Ab molecules, both Ek/b and Ak efficiently presented exogenously added peptides to T cells in the absence of H-2M. In addition, epitopes from some proteins, especially those known to be invariant chain independent, were presented by Ak molecules in the mutant animals. To our surprise, expression of Ek/b overcame the positive selection defect observed in H-2M-deficient mice expressing Ab alone. In contrast, Ak expression did not augment positive selection of CD4+ T cells in the mutant animals. Some of these findings in vivo contrast significantly with findings from in vitro studies on murine MHC class II molecules in human DM-deficient cell lines.

Essential role for cathepsin S in MHC class II-associated invariant chain processing and peptide loading.

Destruction of li by proteolysis is required for MHC class II molecules to bind antigenic peptides, and for transport of the resulting complexes to the cell surface. The cysteine protease cathepsin S is highly expressed in spleen, lymphocytes, monocytes, and other class II-positive cells, and is inducible with interferon-gamma. Specific inhibition of cathepsin S in B lymphoblastoid cells prevented complete proteolysis of li, resulting in accumulation of a class II-associated 13 kDa li fragment in vivo. Consequently, the formation of SDS-stable complexes was markedly reduced. Purified cathepsin S, but not cathepsin B, H, or D, specifically digested li from alpha beta li trimers, generating alpha beta-CLIP complexes capable of binding exogenously added peptide in vitro. Thus, cathepsin S is essential in B cells for effective li proteolysis necessary to render class II molecules competent for binding peptides.

The class I-b molecule Qa-1 forms heterodimers with H-2Ld and a novel 50-kD glycoprotein encoded centromeric to I-E beta.

Recent biochemical characterization of the T23-encoded Qa-1 molecule revealed an additional higher molecular mass species of 50 kD coprecipitated with the 48-kD Qa-1 molecule in H-2b and H-2d mouse strains. We now demonstrate that the 50-kD protein coprecipitated with Qa-1 is the class I-a antigen Ld in all H-2Ld-positive mouse strains examined. Furthers analyses of a panel of recombinants revealed that the 50-kD protein coprecipitated with Qa-1 in H-2b haplotype mouse strains is encoded or controlled by a gene centromeric to major histocompatibility complex class II I-E beta. We have designated this gene and corresponding protein product as Qsm, Qa-1 structure modifier. Both Ld and Qsm can interact with Qa-1 to form cell surface-expressed heterodimers in vivo. These Qa-1 heterodimers are not expressed in H-2k haplotype cells. The Qa-1/Ld and Qa-1/Qsm heterodimers are associated by noncovalent interactions and occur only between fully processed proteins. In addition, we show that the Qsm-encoded protein can form heterodimers with Ld as well, and that the Ld molecules participating in these interactions with Qa-1 and Qsm may be devoid of beta 2-microglobulin and/or peptide. These data represent the first demonstration that class I molecules can be expressed as heterodimers (Qa-1/Ld) on the cell surface, and map a gene (Qsm) that may potentially encode a novel class I molecule, or another protein, that associates with both Qa-1 and Ld. These interactions may enable increased levels of Qa-1 to reach the cell surface and may subsequently influence T cell recognition of Qa-1 and/or Ld molecules.

How MHC class II molecules acquire peptide cargo: biosynthesis and trafficking through the endocytic pathway.

The antigen-specific receptors of T lymphocytes rely on products of the major histocompatibility complex (MHC) to recognize and engage antigen. MHC molecules display antigen on the cell surface in the form of small peptides, generated intracellularly by fragmentation of the intact protein antigen. They acquire these peptides at distinct intracellular locations: In the endoplasmic reticulum (ER), class I molecules bind peptides derived from cytosolic proteins, whereas class II molecules acquire their peptide cargo in an endocytic compartment. Sequestration of class II molecules from the constitutive secretory pathway is mediated by their interaction with an additional polypeptide, the invariant chain (Ii). The Ii contains sorting signals in its cytoplasmic tail that target class II molecules to the endocytic pathway where they encounter peptides generated from protein antigens that have also accessed this route.

The TL region gene 37 encodes a Qa-1 antigen.

Of all the biochemically defined mouse MHC class I molecules, the Qa-1 antigens are the only ones for which a gene has not been identified. Recent evidence has suggested that Qa-1 antigens are functional class I molecules and can function as restriction elements for gamma/delta T cells. We have examined the relationship between Qa-1 and the product of gene 37, a presumed novel class I antigen encoded within the TL region. Immunoprecipitation and polyacrylamide gel electrophoresis analysis of the molecules reactive with anti-Qa-1 and anti-37 sera show that the Qa-1 molecule of Qa-1b (Qa-1.2) mouse strains is identical to the product of gene 37 on the basis of molecular weight, pI, and strain distribution. Immunodepletion, biosynthetic labeling, and tunicamycin treatment confirm that the protein encoded by gene 37 in Qa-1b mice is Qa-1.2. In contrast, the anti-37 serum was unable to recognize the Qa-1 molecule in Qa-1a strains. Given the fact that the only allele to gene 37 thus far identified in a Qa-1a strain (A/J) has a termination codon in the alpha 3 domain, our data lead us to conclude that the Qa-1 molecule expressed in Qa-1a mice is not a true allelic product of the gene 37 encoded antigen of Qa-1b mouse strains.