David, C S

Abstract We have introduced HLA-DQ8 (HLADQB*0302 and HLA-DQA*0301) genes into A β 0 knockout mice. Transgenic animals were immunized with a whole body extract of Dermatophagoides pteronyssinus (Der p), one of the causative agents of house dust mite allergy. Transgenic mice expressing HLA-DQ8 genes elicited HLA-DQ8-restricted responses driven by CD4+ T cells. Synthetic-overlapping peptides representing a major allergen of house dust mite (Der p 2) were synthesized and used as immunogens. HLA-DQ8+ mice responded to three peptides: 8 (residues 61-80), 11 (residues 91-110), and 13 (residues 111-129). The mice produced IL-2, IL-4, and IL-6 response to Der p challenge, suggesting a mixed Th1/Th2 response. These mice represent a new model for studies of the immune basis of allergy.

Abstract Chick type II collagen (CII), a protein commonly found in joint cartilage, induces an autoimmune arthritis when administered to susceptible strains of mice. A cyanogen bromide fragment of CII, CB11, contains the requisite epitopes critical for inducing collagen-induced arthritis. If administered as a tolerogen, however, before immunization, CB11 prevents the onset of disease. Therefore, delineation of structural elements of CB11 that can regulate autoreactive T cells became the goal of this study. To delineate the structural elements of CB11 antigenic to T cells, 14 peptides containing overlapping sequences of CB11 were generated. Mononuclear cells from CII-immunized DBA/1 mice were cultured with these peptides and the resulting supernatants examined for the production of IFN-gamma. Two peptides, CII 181-209 and CII 245-270, generated the greatest responses. The ability of these two peptides to regulate arthritis was tested by administering them to neonatal DBA/1 mice as tolerogens before immunization with CII. Both peptides suppressed the incidence of arthritis whereas no other peptide used as a tolerogen significantly altered the course of the disease. T cells from four arthritis-resistant murine strains did not recognize either peptide when immunized with CII, whereas cells from the disease-susceptible B10.Q mice responded well to both. Thus, the coincidence of T cell responses to CII 181-209 and CII 245-270 in CIA-susceptible mice and the lack of response in disease-resistant strains or CII-tolerized mice identify these two peptides as containing important T cell epitopes that regulate CIA.

Abstract Recently, we have proposed that the combination of HLA-DQ and -DR alleles is responsible for the association of the HLA class II region with rheumatoid arthritis (RA). According to this model, some HLA-DQ alleles, namely DQ4, DQ7, DQ8, and DQ9, predispose carriers to severe RA, but a self peptide of sequence KDILEDERAAVDTYC from the third hypervariable (HV3) region of some DRB1 alleles, including DRB1*0402, can protect from the disease if presented by DQ molecules. This model implies that DQ4, DQ7, DQ8, and DQ9 should be able to present a set of common peptides, despite polymorphisms in their Ag binding groove. In the present study, we have further analyzed the immunogenicity of the DRB1*0402 HV3 peptide in DQ8-transgenic mice. We found that the motif DERAA guarantees DQ8-restricted immunogenicity, and that R is the main anchor residue for binding of the DRB1*0402 peptide to DQ. Interestingly, the p1 pocket that probably controls binding of the R residue is identical in all four RA-associated DQ molecules. Our results imply that the association of RA with some DR subtypes can be explained by their linkage with DQ alleles displaying a binding site for similar “arthritogenic” peptides.

Abstract A series of 101 peptides each 20 amino acids in length (10-residue overlap) spanning the helical portion of the mature α-chain of human type II collagen (CII) was synthesized. DBA/1 (H-2q) mice were immunized with individual peptides, and draining lymph node cells were challenged in vitro. Strong responses were measured to three peptides: peptide I (residues 74-93), peptide 14 (residues 254-273), and peptide 81 (residues 924-943). B10.Q (H-2q) mice were responsive to peptides I and 81 but not to peptide 14. B10.RIII (H-2r) mice, which are resistant to arthritis induction following immunization with human CII, were unresponsive to peptides I, 14, and 81. Using single amino acid truncated peptides, we determined minimal immunostimulatory lengths for peptides I and 81. Residues critical to antigenicity were identified by introducing alanine and glycine substitutions into minimal length immunostimulatory peptides. The determinants within peptides I and 81 are 100% homologous to mouse CII and are autoantigens. Peptide 81 has homology to viral proteins. Peptide 14 is 90% homologous to mouse CII and has homology to heat shock proteins.

Abstract SWR/J mice are resistant to collagen-induced arthritis (CIA) despite having a susceptible H-2q haplotype. We have earlier demonstrated the possible role of the V beta TCR mutation of SWR in the resistance to CIA. To investigate the influence of the C5 deficiency of SWR in this resistance, crosses were made between SWR and A/J (C5 deficient, TCRwild, H-2a), and between SWR and C3H.A (C5 sufficient, TCRwild, H-2a). Upon immunization with bovine type II collagen in adjuvant, there was a similar incidence and severity of arthritis in H-2q-bearing mice in the back-crosses A x (SWR x A) ad C3H.A x (SWR x C3H.A). The absence of hemolytic complement was confirmed in the arthritic A x (SWR x A) back-cross mice by standard SRBC hemolytic assays. In addition C57L (H-2b) mice, which were C5 sufficient but had a V beta TCR deletion mutation similar to SWR, could not complement for CIA susceptibility in H-2q-bearing C57L x (SWR x C57L) back-crosses and in (C57L x SWR)F2 hybrids. These studies show that complement C5 does not play a significant role in CIA susceptibility, and further implicate the V beta TCR mutation in the resistance to CIA in SWR mice.

Abstract Human type II collagen (HuCII) may be one of the autoantigens involved in human rheumatoid arthritis (RA). By using over-lapping peptides, we have previously described an immunodominant region (HuCII.250-270) on HuCII. In the present study, this 21-mer HuCII.250-270 peptide was used as tolerogen, and its effect on both early and effector phase of collagen-induced arthritis (CIA) was examined. Upon immunization with HuCII-derived peptide 250-270, HuCII.250-270-tolerized mice showed diminished T cell proliferation that was mediated by Th1 cytokine, IL-2. More interestingly, oral tolerance with HuCII.250-270 peptide diminishes primarily a Th1 type of immune response. Arthritis severity was reduced markedly in mice orally tolerized with HuCII.250-270 peptide both at early and effector phases. Suppression of CIA at the effector phase by oral administration of HuCII peptide suggests a potential immunotherapeutic use of collagen II peptide in the treatment of human RA.

Abstract The molecular basis for the defective expression of Ef alpha was determined by analysis of the 5' region of a full length Ef alpha gene. The gene was isolated from a genomic library prepared from the A.CA/SnDv mouse strain. DNA sequence analysis of the 5' portion of the Ef alpha gene, which encodes the 5' regulatory sequences and the signal peptide, revealed the presence of a stop codon in the exon encoding the signal peptide. The remainder of the sequences were highly related to sequences found in previously characterized, functional E alpha alleles. Previous studies indicate that the f allele is transcribed at rates comparable to the rates of functional alleles and that mRNA accumulates in the cytoplasm. Primer extension analysis demonstrated that Ef alpha transcripts initiate identically to the functional Ek alpha allele, mapping the defect in the Ef alpha gene 3' of the transcriptional initiation site. To determine whether the stop codon in the signal peptide was the only major defect in this gene, reciprocal chimeric genes were constructed in which the 5' regions, including the first exons, of the defective f allele and the functional k allele were exchanged. The hybrid genes were inserted into an SV40 promotor driven expression vector for cotransfection with an Ek beta gene. Surface I-E expression was demonstrated using I-E specific mAb in Cos-7 and L cell lines transfected with the hybrid gene consisting of the 5' region of the k allele and the 3' portion of the f allele. Therefore, the single stop codon present in the exon encoding the leader peptide of the Ef alpha gene appears to be the only defect preventing this gene from expressing a functional E alpha-chain.

Abstract A rare D-region recombination event which gave rise to the B10.RQDB major histocompatibility complex haplotype has been examined to ascertain the nature of the crossover and to determine which class I genes are present in the new alignment of D-region genes. Serologic analysis have shown that the B10 . RQDB major histocompatibility complex recombinant mouse inherited the H-2Dd gene from the B10.T(6R) parental line and the H-2Dd gene from the B10.A(2R) parental line, representing the first example of an intra-D-region crossover resulting from an intercross. Previous molecular genetic analyses of the d and b haplotypes revealed structural diversity in the organization of their D-region gene clusters. Hence, the D region is comprised of five class I genes in the d haplotype and only one in the b haplotype. Because allelic relationships among the various D-region genes are not defined, either a homologous or nonhomologous alignment of genes has generated the RQDB crossover. Therefore, the possibility that all three D-region antigen-presenting molecules (Dd, Ld, and Db) might be encoded by the RQDB haplotype was examined. Fluorescence-activated cell sorter and cytotoxic T lymphocyte analyses revealed no detectable levels of H-2Ld cell-surface expression, confirming earlier studies with antibody-mediated cytotoxicity and immunoprecipitation. Southern blot analysis localized the recombination point to within a 1-kb region at the centromeric end of the H-2Ld gene on the B10 . T(6R) chromosome in a region of high homology to the H-2Db gene on the B10 . A(2R) chromosome. Together, these studies define the D region of the RQDB haplotype as containing the five class I genes: Dd, D2d, D3d, D4d, and Db. In addition to providing insight into rare recombination events in the D region, the B10.RQDB mouse should be a useful tool for exploring the function of D-region genes.

Abstract Type II collagen from six mammalian species was investigated for the capacity to induce an immune response and collagen-induced arthritis (CIA) in C57/B10 congenic mouse strains. H-2q haplotype mice were susceptible to chick, bovine, deer, rat, and human type II collagen, but were resistant to arthritis induced by porcine type II collagen. H-2r haplotype mice only developed CIA in response to bovine, deer, and porcine collagen. High antibody responses in the absence of disease, directed against a specific type II collagen, were observed in many independent haplotypes. The cross-reactive capacity of different antisera to the various collagen species was studied. The data support the existence of two arthritogenic and multiple nonarthritogenic epitopes on the type II collagen molecule.

Abstract Intracerebral infection of certain strains of mice with Theiler’s virus results in chronic immune-mediated demyelination in spinal cord. We used mouse mutants with deletion of the V beta class of TCR genes to examine the role of TCR genes in this demyelinating disease which is similar to multiple sclerosis. Quantitative analysis of spinal cord lesions demonstrated a markedly increased number and extent of demyelinated lesions in persistently infected RIII S/J mice which have a massive deletion of the TCR V beta-chain (V beta 5.2, V beta 8.3, V beta 5.1, V beta 8.2, V beta 5.3, V beta 8.1, V beta 13, V beta 12, V beta 11, V beta 9, V beta 6, V beta 15, V beta 17) compared with B10.RIII mice which are of identical MHC haplotype (H-2r) but have normal complement of V beta TCR genes. In contrast, infection of C57L (H-2b) or C57BR (H-2k) mice which have deletion of the V beta TCR genes (V beta 5.2, V beta 8.3, V beta 5.1, V beta 8.2, V beta 5.3, V beta 8.1, V beta 13, V beta 12, V beta 11, and V beta 9) resulted in few demyelinating lesions. Genetic segregation analysis of (B10.RIII x RIII S/J) x RIII S/J backcrossed mice and (B10.RIII x RIII S/J) F2 mice demonstrated correlation of increased susceptibility to demyelination with deletion of TCR V beta genes. The increase in number of demyelinating lesions correlated with increase in number of virus-Ag+ cells in spinal cord. These experiments provide strong evidence that the structural diversity at the TCR beta-complex can influence susceptibility to virus-induced demyelination.