Analysis of the mouse transcriptome based on functional annotation of 60,770 full-length cDNAs.

Only a small proportion of the mouse genome is transcribed into mature messenger RNA transcripts. There is an international collaborative effort to identify all full-length mRNA transcripts from the mouse, and to ensure that each is represented in a physical collection of clones. Here we report the manual annotation of 60,770 full-length mouse complementary DNA sequences. These are clustered into 33,409 'transcriptional units', contributing 90.1% of a newly established mouse transcriptome database. Of these transcriptional units, 4,258 are new protein-coding and 11,665 are new non-coding messages, indicating that non-coding RNA is a major component of the transcriptome. 41% of all transcriptional units showed evidence of alternative splicing. In protein-coding transcripts, 79% of splice variations altered the protein product. Whole-transcriptome analyses resulted in the identification of 2,431 sense-antisense pairs. The present work, completely supported by physical clones, provides the most comprehensive survey of a mammalian transcriptome so far, and is a valuable resource for functional genomics.

Efficient discovery of immune response targets by cyclical refinement of QSAR models of peptide binding.

Peptides that induce and recall T-cell responses are called T-cell epitopes. T-cell epitopes may be useful in a subunit vaccine against malaria. Computer models that simulate peptide binding to MHC are useful for selecting candidate T-cell epitopes since they minimize the number of experiments required for their identification. We applied a combination of computational and immunological strategies to select candidate T-cell epitopes. A total of 86 experimental binding assays were performed in three rounds of identification of HLA-A11 binding peptides from the six preerythrocytic malaria antigens. Thirty-six peptides were experimentally confirmed as binders. We show that the cyclical refinement of the ANN models results in a significant improvement of the efficiency of identifying potential T-cell epitopes.

Data warehousing in molecular biology.

In the business and healthcare sectors data warehousing has provided effective solutions for information usage and knowledge discovery from databases. However, data warehousing applications in the biological research and development (R&D) sector are lagging far behind. The fuzziness and complexity of biological data represent a major challenge in data warehousing for molecular biology. By combining experiences in other domains with our findings from building a model database, we have defined the requirements for data warehousing in molecular biology.

FIMM, a database of functional molecular immunology.

FIMM database (http://sdmc.krdl.org.sg:8080/fimm ) contains data relevant to functional molecular immunology, focusing on cellular immunology. It contains fully referenced data on protein antigens, major histocompatibility complex (MHC) molecules, MHC-associated peptides and relevant disease associations. FIMM has a set of search tools for extraction of information and results are presented as lists or as reports.

Residue 116 determines the C-terminal anchor residue of HLA-B*3501 and -B*5101 binding peptides but does not explain the general affinity difference.

HLA-B*3501 and -B*5101 molecules, which belong to the HLA-B5 cross-reactive group, bind peptides carrying similar anchor residues at P2 and the C-terminus, but differences are observed in the preference for a Tyr residue at the C-terminus and the affinity of peptides. A recent study of HLA-B*3501 crystal structure suggested that residue 116 on the floor of the F-pocket determines a preference for anchor residues at the C-terminus. In order to evaluate the role of the residue 116 in the peptide binding to both HLA-B*3501 and HLA-B*5101 molecules, we generated HLA-B*3501 mutant molecules carrying Tyr at residue 116 (B*3501-116Y) and tested the binding of a panel of nonamer peptides to the B*3501-116Y molecules by a stabilization assay with RMA-S transfectants expressing the mutant molecules. The substitution of Tyr for Ser at residue 116 markedly reduced the affinity of nonamer peptides carrying Tyr at P9, while it enhanced that of nonamer peptides carrying Ile and Leu at P9. On the other hand, the affinity of peptides carrying aliphatic hydrophobic residues at P9 to B*3501-116Y molecules was much higher than that to HLA-B*3501 and HLA-B*5101 molecules. These results indicate that residue 116 is critical for the structural difference of the F-pocket between HLA-B*3501 and HLA-B*5101 which determines the C-terminal anchor residues, while leaving other residues which differ between HLA-B*3501 and HLA-B*5101 may be responsible for the low peptide binding property of the latter.

Crucial role of N-terminal residue of binding peptides in recognition of the monoclonal antibody specific for the peptide-HLA-B5, -B35 complex.

The monoclonal antibody (mAb) 4D12 specific for the HLA-B5, -B35 cross-reacting group (CREG) bound to a fraction of HLA-B*3501 and HLA-B*5101 molecules carrying self-peptides. Analysis of the binding of mAb 4D12 to HLA-B*3501 and -B*5101 molecules pulsed with chemically synthesized peptides revealed that this mAb recognizes a restricted number of peptides and that P1 of the bound peptides critically influences its binding. The 4D12 mAb bound only to HLA-B*3501 molecules carrying peptides with Asn, Asp, Glu, Ser, and Val at P1. Analysis using an HLA-B*3501 crystallographic model suggested that 4D12 may recognize the side chain of the P1 residue that is pointing to the solvent. On the other hand, 4D12 bound only to HLA-B*5101 molecules carrying peptides with Asn or Asp at P1, suggesting that the 4D12 epitope formed by Glu, Ser, or Val at P1 and the A-pocket was changed by the substitution of His for Tyr at residue 171 of HLA-B*3501 molecules. This was confirmed by testing the binding of mAb 4D12 to HLA-B*3501 mutant molecules at residue 171 carrying these peptides. These results together suggest that the conformation of the A-pocket and its hydrogen bound network with the P1 residue is also critical for the binding of mAb 4D12. The present study shows the molecular basis of the specificity of 4D12 for the peptide-HLA class I complex.

Application of genetic search in derivation of matrix models of peptide binding to MHC molecules.

T cells of the vertebrate immune system recognise peptides bound by major histocompatibility complex (MHC) molecules on the surface of host cells. Peptide binding to MHC molecules is necessary for immune recognition, but only a subset of peptides are capable of binding to a particular MHC molecule. Common amino acid patterns (binding motifs) have been observed in sets of peptides that bind to specific MHC molecules. Recently, matrix models for peptide/MHC interaction have been reported. These encode the rules of peptide/ MHC interactions for an individual MHC molecule as a 20 x 9 matrix where the contribution to binding of each amino acid at each position within a 9-mer peptide is quantified. The artificial intelligence techniques of genetic search and machine learning have proved to be very useful in the area of biological sequence analysis. The availability of peptide/MHC binding data can facilitate derivation of binding matrices using machine learning techniques. We performed a simulation study to determine the minimum number of peptide samples required to derive matrices, given the pre-defined accuracy of the matrix model. The matrices were derived using a genetic search. In addition, matrices for peptide binding to the human class I MHC molecules, HLA-B35 and -A24, were derived, validated by independent experimental data and compared to previously-reported matrices. The results indicate that at least 150 peptide samples are required to derive matrices of acceptable accuracy. This result is based on a maximum noise content of 5%, the availability of precise affinity measurements and that acceptable accuracy is determined by an area under the Relative Operating Characteristic curve (Aroc) of > 0.8. More than 600 peptide samples are required to derive matrices of excellent accuracy (Aroc > 0.9). Finally, we derived a human HLA-B27 binding matrix using a genetic search and 404 experimentally-tested peptides, and estimated its accuracy at Aroc > 0.88. The results of this study are expected to be of practical interest to immunologists for efficient identification of peptides as candidates for immunotherapy.

Identification of HTLV-1-specific CTL directed against synthetic and naturally processed peptides in HLA-B*3501 transgenic mice.

Previous studies of CTL responses to influenza peptides in HLA single transgenic mice resulted in the identification of at most one immunodominant epitope. Since HLA-B*3501 is known to present multiple HIV-1-specific T cell epitopes we tested the cellular immune response of HLA-B*3501 transgenic mice to synthetic HTLV-1 peptides mixed with the lipohexapeptide N-palmitoyl-S-[2,3-bis(palmitoyloxy)propyl]cysteinyl-seryl-lysyl-l ysyl- lysyl-lysine, which is a biocompatible, Th-epitopeindependent adjuvant. Eleven of 37 tested HLA-B*3501 binding peptides mounted a CTL response after three in vitro stimulations. The HLA-B*3501 affinity of peptides correlated with their ability to induce CTL in HLA-B*3501 transgenic mice. Seven peptides derived from env-gp46 (VPSPSSTPLL, VPSSSSTPL, YPSLALAPH, and YPSLALAPA), pol (QAFPQCTIL), gagp19 (YPGRVNEIL), and tax (GAFLTNVPY) proteins induced peptide-specific CTL Bulk CTL generated by four peptides derived from env-gp46 (SPPSTPLLY, VPSPSSTPLLY, and VPSPSSTPLL) and pol (QAFPQCTILQY) killed peptide-pulsed and recombinant vaccinia-infected target cells. The latter peptides therefore present T-cell epitopes and are vaccine candidates for our transgenic mouse model.

Refined peptide HLA-B*3501 binding motif reveals differences in 9-mer to 11-mer peptide binding.

HLA-B*3501 is associated with subacute thyroiditis and fast progression of AIDS. An important prerequisite to investigate the T-cell recognition of HLA-B*3501-restricted antigens is the characterization of peptide-HLA-B*3501 interactions. In this study, peptide-HLA-B*3501 interactions were determined in quantitative peptide binding assays. The results were statistically analyzed to evaluate the influence of both anchor and nonanchor positions and the predictability of peptide binding. The binding data demonstrated that all anchor residues at position 2 and the C-terminus found in 9-mers functioned equally as anchors in 10-mers and 11-mers. These minimum requirements of peptide binding were refined by assessing positive and negative effects of nonanchor residues. Aliphatic hydrophobic residues at positions 3, 5, and 8 of 10-mers and position 3 of 11-mers significantly enhanced HLA-B*3501 binding. Similar effects rendered aromatic, bulky residues, acidic or polar residues of 11-mers at position 1 as well as at positions 4, 8, and 10, respectively. Negative effects were observed for residues carrying positively charged side-chains at position 7 of 11-mers. The refined HLA-B*3501 peptide binding motifs enhanced the identification of potential T-cell epitopes. The disparity between positive effects at the middle and C-terminal part (positions 5 - 8 and 10) of 11-mers and shorter peptides supports the extrusion of 11-mer residues at positions 5, 6, and 7, away from the HLA-B*3501 binding cleft.

Fine tuning of peptide binding to HLA-B*3501 molecules by nonanchor residues.

The prerequisites of peptide HLA-B*3501 interactions have been revisited by quantitative peptide binding assays with 190 chemically synthesized peptide possessing two anchor residues corresponding to the HLA-B*3501 peptide motif and a statistical residue-position analysis of binding and nonbinding peptides. According to the peptide motif of HLA-B*3501, aliphatic hydrophobic (Leu, Ile, and Met) or aromatic residues (Tyr and Phe) specify the main anchor at the C terminus, and position 2 renders an auxiliary anchor for proline. The importance of these residues was confirmed as a minimum requirement for peptide binding. Moreover, we demonstrated that high affinity peptide binding requires more than one favorable position of positions 3, 4, and 7. Aliphatic hydrophobic residues and residues that contain -OH or -SH side chains in position 3, 7, and 4 significantly enhance binding. Positions 1 and 5, or 7 may deteriorate peptide binding if these positions are held by proline and small residues (Ala and Gly) or basic residues carrying positively charged side chains (Arg and Lys), respectively. Positions 6 and 8 were statistically free of constrains. Yet, bulky aromatic residues and basic residues with a positively charged side chain at position 8 decreased the binding affinity. These findings were used to assess the predictability of binding and nonbinding peptides. Our binding predictions of 28 nonamers were verified by experimental data. Taking into account the importance of anchor and nonanchor positions in peptide binding and their practical value in peptide binding prediction, the search for peptide epitopes becomes more efficient.

HLA-B*3501-peptide interactions: role of anchor residues of peptides in their binding to HLA-B*3501 molecules.

Two HLA-B*3501 binding self-peptides, LPFDFTPGY (37F) and LPGPKFLQY (28H), were isolated from HLA-B*3501 molecules expressed by cultured human B lymphoid cells. Both sequences were consistent with previously reported motifs of HLA-B*3501 binding peptides which carry proline at position 2 and tyrosine at position 9 as anchor residues. Direct binding of these peptides to HLA-B*3501 molecules was quantitated by flow cytometry analysis of RMA-S cells. transfected with the HLA-B*3501 gene (RMA-S-B*3501). Both 37F and 28H peptides bound effectively to HLA-B*3501 molecules. Substitution of amino acids at position 2 and/or 9 of HLA-B*3501 binding peptides markedly reduced their binding to HLA-B*3501 molecules. These results indicate that two anchor residues, proline at position 2 and tyrosine at position 9 are critical in binding of peptides to HLA-B*3501 molecules. Insertion of up to four glycine residues at position 8 of the peptide 37F did not affect its binding affinity to HLA*3501 molecules. These results indicate that long peptides can effectively bind to HLA class I molecules provided that anchor residues are conserved.

Conservative evolution of the Mbc-DP region in anthropoid primates.

To determine the organization of the DP region in the Mbc of anthropoid primates, we constructed contig maps from cosmid clones of the chimpanzee and orangutan, representatives of the infraorder Catarrhini, as well as of the cotton-top tamarin, a representative of the infraorder Platyrrhini. We found the maps to be remarkably similar to each other and to the previously published map of the human DP region. In each of the four species, the DP region consists of four loci arranged in the same order (DPB2 . . . DPA2 . . . DPB1 . . . DPA1) and in the same transcriptional orientation (tail-to-tail). The regions in the four species are of approximately the same length and many of the restriction sites are shared between species. The inserts of most Alu elements, of a ribosomal protein pseudogene, and of an IgC epsilon-like pseudogene are found in corresponding positions in all four species. The data indicate that the human-type organization of the DP region was established before the divergence of the Catarrhini and Platyrrhini lines more than 37 million years ago and that it has remained principally intact since that time. This conservation of the DP region is in striking contrast to the evolutionary instability of certain other Mbc regions, in particular those occupied by the DRB or C4 and CYP21 loci. We interpret the stability of the DP region as an indication that the region is being phased out functionally.

Multiplication of Mhc-DRB5 loci in the orangutan: implications for the evolution of DRB haplotypes.

The beta chain-encoding (B) class II genes of the primate major histocompatibility complex belong to several families. The DRB family of class II genes is distinguished by the occurrence of haplotype polymorphism--the existence of multiple chromosomal forms differing in length, gene number, and gene combinations, each form occurring at an appreciable frequency in the population. Some of the haplotypes, or fragments thereof, are shared by humans, chimpanzees, and gorillas. In an effort to follow the DRB haplotype polymorphism further back in time, we constructed DRB contig maps of the two chromosomes present in the orangutan cell line CP81. Two types of genes were found in the two haplotypes, Popy-DRB5 and Popy-DRB1*03, the former occurring in two copies and one gene fragment in each haplotype, so that the CP81 cell line contains four complete DRB5 genes and two DRB5 fragments altogether. Since the four genes are more closely related to one another than they are to other DRB5 genes, they must have arisen from a single ancestral copy by multiple duplications. At the same time, however, the two CP81 haplotypes differ considerably in their restriction enzyme sites and in the presence of Alu elements at different positions, indicating that they have been separated for a length of time that exceeds the lifespan of a primate species. Moreover, a segment of about 100 kilobase pairs is shared between the orangutan CP81-1 and the human HLA-DR2 haplotype. These findings indicate that part of the haplotype polymorphism may have persisted for more than 13 million years, which is the estimated time of human-orangutan divergence.

Gorilla major histocompatibility complex-DRB pseudogene orthologous to HLA-DRBVIII.

The HLA-DR4 haplotype consists of four DRB genes: DRB1*04, DRBVII, DRBVIII, and DRB4*01, arranged in this order on the chromosome. The DRB1 and DRB4 genes code for beta chains of the alpha beta heterodimers expressed on the cell surface and bearing the HLA-DR4 and HLA-DRw53 determinants, respectively; the DRBVII and DRBVIII are pseudogenes. We found and sequenced a gene closely related to HLA-DRBVIII in the genome of the lowland gorilla "Sylvia." We designate this gene Gogo-DRB8. The close relationship between the two genes is indicated by the overall sequence similarity, the absence of recognizable exons 1 and 2 in both genes, the presence of two Alu repeats at corresponding positions, and high sequence similarity between corresponding repeats. The comparison with an outgroup (tamarin) gene and the functional counterparts of the DRB8 gene indicate that DRB8 emerged between 18 and 26 million years ago and became inactivated at the same time as or shortly after its creation. Hence DRB8 has probably existed as a pseudogene since the divergence of apes from Old World monkeys more than 20 million years ago.

Mhc-DRB genes of the pigtail macaque (Macaca nemestrina): implications for the evolution of human DRB genes.

The DRB family of human class II major histocompatibility complex (Mhc) loci is unusual in that individuals differ in the number and combination of genes (haplotypes) they carry. Indications are that both the allelic and haplotype polymorphisms of the DRB loci predate speciation. Searching for the evolutionary origins of these polymorphisms, we have sequenced five DRB clones isolated from a cDNA library of a pigtail macaque (Macaca nemestrina) B lymphocyte line. The clones represent five different genes which we designate Mane-DRB*01-Mane-DRB*05. The genes appears to be approximately equidistant from each other, so that allelic relationships between them cannot be established on the basis of the sequence data alone. If positions coding for the peptide-binding region of the class II beta chains are eliminated from sequence comparisons, the Mane-DRB genes appear to be most closely related to the human (HLA) DRB1 genes of the DRw52 group. We interpret this finding to indicate that the ancestral gene of the DRw52 group of human DRB1 alleles separated from the rest of the HLA-DRB1 alleles before the separation of the Old World monkeys (Cercopithecoidea) from the apes (Hominoidea) in the early Oligocene. After this separation, the ancestral DRB1 gene of the DRw52 group duplicated in the Old World monkey lineage to give rise to genes at three loci at least, while in the ape lineage this gene may have remained single and diverged into a number of alleles instead. These findings suggest that some of the polymorphism currently present at the DRB1 locus is greater than 35 Myr old.