Differential tolerance is induced in T cells recognizing distinct epitopes of myelin basic protein.

Experimental allergic encephalomyelitis (EAE) is induced by T cell-mediated immunity to central nervous system antigens. In H-2u mice, EAE is mediated primarily by T cells specific for residues 1-11 of myelin basic protein (MBP). We demonstrate that differential tolerance to MBP1-11 versus epitopes in MBP121-150 is induced by expression of endogenous MBP, reflecting extreme differences in stability of peptide/MHC complexes. The diverse MBP121-150-specific TCR repertoire can be divided into three fine specificity groups. Two groups were identified in wild-type mice despite extensive tolerance, but the third group was not detected. Activated MBP121-150-specific T cells induce EAE in wild-type mice. Thus, encephalitogenic T cells that escape tolerance either recognize short-lived peptide/MHC complexes or express TCRs with unique specificities for stable complexes.

Sensitivity and selectivity in protein similarity searches: a comparison of Smith-Waterman in hardware to BLAST and FASTA.

To predict the functions of a possible protein product of any new or uncharacterized DNA sequence, it is important first to detect all significant similarities between the encoded amino acid sequence and any accumulated protein sequence data. We have implemented a set of queries and database sequences and proceeded to test and compare various similarity search methods and their parameterizations. We demonstrate here that the Smith-Waterman (S-W) dynamic programming method and the optimized version of FASTA are significantly better able to distinguish true similarities from statistical noise than is the popular database search tool BLAST. Also, a simple "log-length normalization" of S-W scores based on the query and target sequence lengths greatly increased the selectivity of the S-W searches, exceeding the default normalization method of FASTA. An implementation of the modified S-W algorithm in hardware (the Fast Data Finder) is able to match the accuracy of software versions while greatly speeding up its execution. We present here the selectivity and sensitivity data from these tests as well as results for various scoring matrices. We present data that will help users to choose threshold score values for evaluation of database search results. We also illustrate the impact of using simple-sequence masking tools such as SEG or XNU.

Hidden Markov models of biological primary sequence information.

Hidden Markov model (HMM) techniques are used to model families of biological sequences. A smooth and convergent algorithm is introduced to iteratively adapt the transition and emission parameters of the models from the examples in a given family. The HMM approach is applied to three protein families: globins, immunoglobulins, and kinases. In all cases, the models derived capture the important statistical characteristics of the family and can be used for a number of tasks, including multiple alignments, motif detection, and classification. For K sequences of average length N, this approach yields an effective multiple-alignment algorithm which requires O(KN2) operations, linear in the number of sequences.

Peptide mass maps: a highly informative approach to protein identification.

A computer searching algorithm has been used to identify protein sequences in the Protein Information Resource (PIR) database with peptide mass information (mass map) obtained from proteolytic digests of proteins analyzed by microcapillary high-performance liquid chromatography electrospray ionization mass spectrometry. A theoretical analysis of the cytochrome c family demonstrates the ability to identify protein sequences in the PIR database with a high degree of accuracy using a set of six predicted tryptic peptide masses. This method was also applied to experimentally determined peptide masses for a small GTP-binding protein, a protein from pig uterus, the human sex steroid binding protein, and a thermostable DNA polymerase. The results demonstrate that a set of observed masses which is less than 50% of the total number of predicted masses can be used to identify a protein sequence in the database. For the analysis presented in this paper, a mass matching tolerance of 1 amu is used. Under these conditions, mass maps created by fast atom bombardment mass spectrometry and matrix-assisted laser desorption time-of-flight would also be applicable. In cases where multiple matches are observed or verification of the protein identification is needed, tandem mass spectrometry sequencing can be used to establish sequence similarity.

Model genomes: the benefits of analysing homologous human and mouse sequences.

The human genome initiative has provided the motivating force for launching sequencing projects suitable for testing various DNA-sequencing strategies, as well as motivating the development of mapping and sequencing technologies. In addition to projects targeting selected regions of the human genome, other projects are based on model organisms such as yeast, nematode and mouse. The sequencing of homologous regions of human and mouse genomes is a new approach to genome analysis, and is providing insights into gene evolution, function and regulation which could not be determined so easily from the analysis of just one species.

Sequence accuracy of large DNA sequencing projects.

Very little information has been accumulated regarding the likely accuracy of final or consensus DNA sequence data. With the large-scale efforts anticipated for the Human Genome Project, the subjective determination of final sequence must eventually be replaced with more objective, automatic methods. This will require a much better understanding of the nature of error in raw sequencing data and its impact on the determination of the final sequence. This paper describes a start at defining the error model of large-scale sequencing efforts based on random subcloning strategies.

Large-scale and automated DNA sequence determination.

DNA sequence analysis is a multistage process that includes the preparation of DNA, its fragmentation and base analysis, and the interpretation of the resulting sequence information. New technological advances have led to the automation of certain steps in this process and have raised the possibility of large-scale DNA sequencing efforts in the near future [for example, 1 million base pairs (Mb) per year]. New sequencing methodologies, fully automated instrumentation, and improvements in sequencing-related computational resources may render genome-size sequencing projects (100 Mb or larger) feasible during the next 5 to 10 years.

Chimeric immunoglobulin-T cell receptor proteins form functional receptors: implications for T cell receptor complex formation and activation.

We constructed chimeric receptor chains in which an immunoglobulin heavy chain variable region (VH) from a phosphorylcholine-specific antibody is substituted for T cell receptor (Tcr) alpha and beta V regions. We demonstrate that the VH region joined to either the C alpha or the C beta region can form stable chimeric proteins in EL4 T cells. Both chimeric receptor chains associate with CD3 polypeptides in functional receptor complexes and respond to phosphorylcholine coupled to Sepharose beads. The VH-C alpha chimeric chain associates with the EL4 beta chain, while the VH-C beta chimeric protein appears to form either a homodimer or a heterodimer with the native EL4 beta chain. Thus, functional receptor complexes can be formed using two C beta regions, and the C alpha region may not be required for CD3 association and surface expression of Tcr complexes.

Comparison of exon 5 sequences from 35 class I genes of the BALB/c mouse.

DNA sequences of the fifth exon, which encodes the transmembrane domain, were determined for the BALB/c mouse class I MHC genes and used to study the relationships between them. Based on nucleotide sequence similarity, the exon 5 sequences can be divided into seven groups. Although most members within each group are at least 80% similar to each other, comparison between groups reveals that the groups share little similarity. However, in spite of the extensive variation of the fifth exon sequences, analysis of their predicted amino acid translations reveals that only four class I gene fifth exons have frameshifts or stop codons that terminate their translation and prevent them from encoding a domain that is both hydrophobic and long enough to span a lipid bilayer. Exactly 27 of the remaining fifth exons could encode a domain that is similar to those of the transplantation antigens in that it consists of a proline-rich connecting peptide, a transmembrane segment, and a cytoplasmic portion with membrane-anchoring basic residues. The conservation of this motif in the majority of the fifth exon translations in spite of extensive variation suggests that selective pressure exists for these exons to maintain their ability to encode a functional transmembrane domain, raising the possibility that many of the nonclassical class I genes encode functionally important products.

Nucleotide sequence of dengue 2 RNA and comparison of the encoded proteins with those of other flaviviruses.

We have determined the complete sequence of the RNA of dengue 2 virus (S1 candidate vaccine strain derived from the PR-159 isolate) with the exception of about 15 nucleotides at the 5' end. The genome organization is the same as that deduced earlier for other flaviviruses and the amino acid sequences of the encoded dengue 2 proteins show striking homology to those of other flaviviruses. The overall amino acid sequence similarity between dengue 2 and yellow fever virus is 44.7%, whereas that between dengue 2 and West Nile virus is 50.7%. These viruses represent three different serological subgroups of mosquito-borne flaviviruses. Comparison of the amino acid sequences shows that amino acid sequence homology is not uniformly distributed among the proteins; highest homology is found in some domains of nonstructural protein NS5 and lowest homology in the hydrophobic polypeptides ns2a and 2b. In general the structural proteins are less well conserved than the nonstructural proteins. Hydrophobicity profiles, however, are remarkably similar throughout the translated region. Comparison of the dengue 2 PR-159 sequence to partial sequence data from dengue 4 and another strain of dengue 2 virus reveals amino acid sequence homologies of about 64 and 96%, respectively, in the structural protein region. Thus as a general rule for flaviviruses examined to date, members of different serological subgroups demonstrate 50% or less amino acid sequence homology, members of the same subgroup average 65-75% homology, and strains of the same virus demonstrate greater than 95% amino acid sequence similarity.

L3T4 and the immunoglobulin gene superfamily: new relationships between the immune system and the nervous system.

L3T4 is a mouse cell surface protein expressed on most thymocytes and on the subset of mature T cells that recognizes class II MHC molecules. Its primary function on T cells is most likely that of increasing the avidity of the interaction between T cells and antigen-presenting or target cells. It may accomplish this by binding to a nonpolymorphic region on class II MHC molecules. The cDNA and gene encoding L3T4 have been isolated and sequenced. Analysis of the amino acid sequence predicted by the nucleotide sequence indicates that L3T4 is a member of the Ig gene superfamily. It is most closely related to Ig and Tcr V regions. Although the amino-terminal domain of L3T4 is the portion of the molecule that is most similar to V-regions, L3T4 is one of the polydomain members of the Ig gene superfamily. Studies of the expression of L3T4 mRNA in various tissues led to the surprising finding that this gene is transcribed not only in T lymphoid cells, but also in brain. The predominant form of L3T4 mRNA expressed in brain is foreshortened as compared to that in T lineage cells, and it is most likely the product of a distinct transcriptional start site. If translated, the protein encoded by this brain transcript would be 217 amino acids in length and would lack the signal peptide and the amino-terminal 214 amino acids of the mature protein. It is not known whether a stable protein product is synthesized from this mRNA or what its function might be. However, these findings place L3T4 in an intriguing class of Ig gene superfamily members characterized by coexpression in the immune system and the nervous system.

Diversity and structure of human T-cell receptor alpha-chain variable region genes.

The nucleotide sequences of 27 T-cell receptor alpha-chain variable region (V alpha)-containing cDNA clones isolated from a cDNA library derived from human peripheral blood lymphocytes were determined. Eighteen different V alpha and 26 different joining (J alpha) gene segments are utilized in these clones. The V alpha gene segments belong to 12 different subfamilies, each containing from one to seven members. Comparisons with the 16 different V alpha and 21 different J alpha sequences previously reported suggest that the germ-line repertoires for these gene segments are greater than previously estimated. Flexibility in the sites of gene segment joining and possibly N-region diversification also contribute to human alpha-chain diversity. Comparisons of human V alpha regions indicate a high degree of variability spread uniformly across the entire V alpha region without obvious hypervariable regions. However, amino acids important for the maintenance of V gene structure are conserved.

Isolation and sequence of L3T4 complementary DNA clones: expression in T cells and brain.

T lymphocytes express on their surface not only a specific receptor for antigen and major histocompatibility complex proteins, but also a number of additional glycoproteins that are thought to play accessory roles in the processes of recognition and signal transduction. L3T4 is one such T-cell surface protein that is expressed on most mouse thymocytes and on mature mouse T cells that recognize class II (Ia) major histocompatibility complex proteins. Such cells are predominantly of the helper/inducer phenotype. In this study, complementary DNA clones encoding L3T4 were isolated and sequenced. The predicted protein sequence shows that L3T4 is a member of the immunoglobulin gene superfamily. It is encoded by a single gene that does not require rearrangement prior to expression. Although the protein has not previously been demonstrated on nonhematopoietic cells, two messenger RNA species specific for L3T4 are found in brain. The minor species comigrates with the L3T4 transcript in T cells, whereas the major species is 1 kilobase smaller.