Knowledge-based structure prediction of MHC class I bound peptides: a study of 23 complexes

BACKGROUND: The binding of T-cell antigenic peptides to MHC molecules is a prerequisite for their immunogenicity. The ability to identify binding peptides based on the protein sequence is of great importance to the rational design of peptide vaccines. As the requirements for peptide binding cannot be fully explained by the peptide sequence per se, structural considerations should be taken into account and are expected to improve predictive algorithms. The first step in such an algorithm requires accurate and fast modeling of the peptide structure in the MHC-binding groove. RESULTS: We have used 23 solved peptide-MHC class I complexes as a source of structural information in the development of a modeling algorithm. The peptide backbones and MHC structures were used as the templates for prediction. Sidechain conformations were built based on a rotamer library, using the 'dead end elimination' approach. A simple energy function selects the favorable combination of rotamers for a given sequence. It further selects the correct backbone structure from a limited library. The influence of different parameters on the prediction quality was assessed. With a specific rotamer library that incorporates information from the peptide sidechains in the solved complexes, the algorithm correctly identifies 85% (92%) of all (buried) sidechains and selects the correct backbones. Under cross-validation, 70% (78%) of all (buried) residues are correctly predicted and most of all backbones. The interaction between peptide sidechains has a negligible effect on the prediction quality. CONCLUSIONS: The structure of the peptide sidechains follows from the interactions with the MHC and the peptide backbone, as the prediction is hardly influenced by sidechain interactions. The proposed methodology was able to select the correct backbone from a limited set. The impairment in performance under cross-validation suggests that, currently, the specific rotamer library is not satisfactorily representative. The predictions might improve with an increase in the data.     

Stenotic origin of an aberrant left subclavian artery from a right-sided aortic arch. A case report

OBJECTIVE: To refine the algorithms governing peptide presentation by HLA-B*2705 by analyzing: (i) the specificity of the human transporter associated with antigen processing (TAP) for HLA-B27 binding peptides; and (ii) the peptide binding affinity to HLA-B*2705. METHODS: TAP-translocation was measured with a labeled reporter peptide containing an N-linked glycosylation acceptor site in Streptolysin O-permeabilized cells for a panel of HLA-B27 binding peptides. Peptide binding affinity was determined by peptide-induced stabilization of empty HLA-B*2705 expressed by the TAP-deficient cell line T2-B27. RESULTS: Human TAP preferentially translocated analogues with residues leucine, isoleucine, methionine and arginine as the carboxy-terminal amino acids, whereas analogues with aspartic acid and serine were translocated poorly. The binding affinity to HLA-B*2705 of the poorly translocated aspartic acid and serine analogues was about 100-fold less compared to the parent HLA-B27 binding peptide. CONCLUSIONS: Human TAP shows considerable specificity for the C-terminus of potential HLA-B27 ligands. Nonamer peptides with aspartic acid and serine at the C-terminus are poorly translocated by the TAP and have low binding affinity for HLA-B*2705, and are therefore unlikely to become presented by HLA-B*2705.     

MR imaging of symptomatic osteochondromas with pathological correlation

Presentation of antigenic peptides by major histocompatibility complex (MHC) class I molecules depends on translocation of cytosolic peptides into the endoplasmic reticulum (ER) by transporters associated with antigen processing (TAP). Peptide transport by TAP is thought to include at least two steps: initial binding of peptide to TAP, and its subsequent translocation requiring ATP hydrolysis. These events can be monitored in peptide binding and transport assays. Previous studies have shown that the efficiency of peptide transport by human, mouse and rat transporters varies according to the C-terminals of peptide substrates in an allele and species-specific manner. However, it has not been clear during which step of peptide interaction with TAP selection occurs. We used an assay monitoring the peptide binding step to study the binding affinity of a library of 199 peptides for human TAP and the two major allelic rat TAP complexes. We observed a dominant influence of the C-terminus on peptide binding affinity for all transporters, and highly restrictive selection of peptides with aliphatic and aromatic C-terminals by rat TAP1/TAP2u complexes. The selectivity of peptide binding to rat TAP complexes is in full accordance with published data on selective peptide transport and on control of antigen presentation by rat TAP. These results strongly suggest that (i) peptide selection by TAP occurs exclusively in the initial binding step; (ii) all factors involved in peptide selection by TAP are present in insect cells.     

Von Recklinghausen''s disease (neurofibromatosis) and pregnancy: apropos of a clinical case

Using a random combinatorial synthetic peptide library method based on a one-bead one-peptide concept for ligand identification (Lam et. al, Nature 1991, 354, 82-84.), idiotype specific peptides were retrieved and optimized for interaction with the cell surface immunoglobulins [IgM(kappa)] of two murine B lymphoma cell lines. Several of the identified peptides were characterized with respect to cell binding and signal transduction. These peptides were able to bind specifically to the surface immunoglobulins of these lymphoma cells. In addition to binding, when synthesized in tetrameric or multimeric forms, the peptides were able to trigger signal transduction resulting in an increase in protein tyrosine phosphorylation. Since D-amino acid peptide libraries were used in some of our efforts to identify binding ligands, several of the idiotype-specific peptides are composed of all D-amino acids (e.g. wGeyvmvnG). These findings may have important therapeutic implications for targeted-therapy of B-cell lymphoma as these D-amino acid ligands are more resistant to proteolysis resulting in a prolonged pharmacokinetic disposition in vivo.     

Peptide design by artificial neural networks and computer-based evolutionary search

A technique for systematic peptide variation by a combination of rational and evolutionary approaches is presented. The design scheme consists of five consecutive steps: (i) identification of a "seed peptide" with a desired activity, (ii) generation of variants selected from a physicochemical space around the seed peptide, (iii) synthesis and testing of this biased library, (iv) modeling of a quantitative sequence-activity relationship by an artificial neural network, and (v) de novo design by a computer-based evolutionary search in sequence space using the trained neural network as the fitness function. This strategy was successfully applied to the identification of novel peptides that fully prevent the positive chronotropic effect of anti-beta1-adrenoreceptor autoantibodies from the serum of patients with dilated cardiomyopathy. The seed peptide, comprising 10 residues, was derived by epitope mapping from an extracellular loop of human beta1-adrenoreceptor. A set of 90 peptides was synthesized and tested to provide training data for neural network development. De novo design revealed peptides with desired activities that do not match the seed peptide sequence. These results demonstrate that computer-based evolutionary searches can generate novel peptides with substantial biological activity.     

Temporal and spatial expression of the E5a protein during the differentiation-dependent life cycle of human papillomavirus type 31b

Nalpha-for-Nle-Leu-Phe-Nle-Tyr-Lys, a chemotactic peptide that binds with high affinity to the chemoattractant receptor on granulocytes and monocytes, was labeled with 99mTc using the diaminedithiol (DADT) chelating system to coordinate the Tc. 99mTc labeling of the DADT-coupled peptide was accomplished in 84% overall yield (room temperature for 10 min) using [99mTc]glucoheptonate as the donor of prereduced Tc. HPLC analysis showed two major 99mTc-labeled peptide peaks, 99mTc-DADT-Pep-I and 99mTc-DADT-Pep-II, were obtained in a ratio of 1:0.85. Using an iodoacetamide-derivatized gel to remove unlabeled peptide from the 99mTc labeling mixtures, essentially no-carrier-added (nca) high-specific activity 99mTc-labeled chemotactic peptides were obtained. The 99Tc analogues of the peptides were synthesized (72% yield) in a similar fashion and correlated with 99mTc complexes I and II by HPLC. In vitro competitive receptor binding assays of the isolated 99Tc analogues were performed against the tritiated chemotactic peptide [3H]N-for-Met-Leu-Phe ([3H]fMLF) using isolated granulocytes. The 99Tc-derivatized peptides showed similar binding affinities to the chemoattractant receptor as the unlabeled Nalpha-for-Nle-Leu-Phe-Nle-Tyr-Lys. The nca 99mTc-labeled peptides gave high contrast images of experimental inflammation in rabbits without causing neutropenia. Thus, it is feasible to attach the Tc-DADT chelate to low-molecular weight receptor binding chemotactic peptides and retain substantial binding to the receptor. Chemotactic peptides labeled with 99mTc via the DADT ligand system have the potential for imaging focal sites of inflammation without toxic effects, an important consideration in the successful utilization of chemotactic peptide agonists.     

Probing the basis of antibody reactivity with a panel of constrained peptide libraries displayed by filamentous phage

The structural requirements for peptide binding to an antibody may be elucidated by probing it with a variety of peptides having different constraints. To this end, we have constructed and screened a panel of peptide libraries displayed by filamentous bacteriophage. The peptides in most of the libraries have the potential for constraint by fixed Cys residues, which have been placed at different sites within a randomized amino acid sequence of varying length. When taken together, the binding data obtained from screening the panel with a given antibody allow one to determine the types of constraints that promote binding, as well as the residues that are critical for binding. We describe the construction of 11, pVIII-displayed, peptide libraries, whose sizes range from 150 million to 10 billion clones. The libraries were screened with a number of polyclonal and monoclonal antibodies against peptides, proteins and carbohydrates. Cross-reactivity with peptides was always found for antibodies produced against peptides, linear epitopes on folded proteins and, surprisingly, carbohydrates, whereas antibodies against discontinuous epitopes on proteins were found less frequently. The implications of these results are discussed in terms of the structural basis for cross-reactivity with peptides.     

Characterization of peptides that bind the tumor-associated Thomsen-Friedenreich antigen selected from bacteriophage display libraries

Peptides with high affinities and specificities for numerous proteins and nucleic acids have been previously identified from random peptide bacteriophage display libraries. Here, random peptide bacteriophage display libraries were used to identify sequences that bound the cancer-associated Thomsen-Friedenreich glycoantigen (T antigen). The T antigen, present on most malignant cells, contains an immunodominant Gal beta1 --> 3GalNAc alpha disaccharide unmasked on the surfaces of most carcinomas. This antigen has been postulated to be involved in tumor cell aggregation and metastasis. Two 15 amino acid random peptide bacteriophage display libraries were affinity selected with glycoproteins displaying T antigen on their surfaces. Sequence analysis revealed that many of the peptides shared homology with sugar recognition sites in several carbohydrate-binding proteins. A comparison of affinity selected sequences from both libraries yielded a common motif (W-Y-A-W/F-S-P) rich in aromatic amino acids. Four peptides, corresponding to the affinity selected sequences, were chemically synthesized and characterized for their carbohydrate recognition properties. The synthetic peptides exhibited high specificities and affinities to T antigen displayed on asialofetuin or conjugated to bovine serum albumin (Kd = 5 nM for MAP-P30 binding to asialofetuin) as well as free T-antigen disaccharide in solution (Kd = 10 microM for MAP-P30, 20 microM for P10). Two peptides, P30 and P10, demonstrated high affinities and specificities for both asialofetuin and T antigen in solution. Iodination of a lone tyrosine residue in each sequence dramatically reduced their abilities to bind T antigen, suggesting that the tyrosine residue plays an important role in carbohydrate recognition. That these peptides are of functional significance is evidenced by the ability of both P30 and P10 to inhibit asialofetuin-mediated melanoma cell aggregation in vitro and to compete with peanut lectin for binding to T antigen displayed on the surface of MDA-MB-435 breast carcinoma cells in situ.     

Contribution of the hydrophobicity gradient of an amphipathic peptide to its mode of association with lipids

A class of peptides that associate with lipids, known as oblique-orientated peptides, was recently described [Brasseur R., Pillot, T., Lins, L., Vandekerckhove, J. & Rosseneu, M. (1997) Trends Biochem. Sci. 22, 167-171]. Due to an asymmetric distribution of hydrophobic residues along the axis of the alpha-helix, such peptides adopt an oblique orientation which can destabilise membranes or lipid cores. Variants of these oblique peptides, designed to have an homogeneous distribution of hydrophobic and hydrophilic residues along the helical axis, are classified as regular amphipathic peptides. These peptides are expected to lie parallel to the polar/apolar interface with their hydrophobic residues directed towards the apolar and their hydrophilic residues towards the polar phase. An hydrophobic, oblique-orientated peptide was identified at residues 56-68 in the sequence of the lecithin-cholesterol acyltransferase (LCAT), enzyme. This peptide is predicted to penetrate a lipid bilayer at an angle of 40 degrees through its more hydrophobic C-terminal end and thereby induce the destabilisation of a membrane or a lipid core. The LCAT-(56-68) wild-type peptide was synthesised together with the LCAT-(56-68, 0 degrees) variant, in which the hydrophobicity gradient was abolished through residue permutations. In two other variants, designed to keep their oblique orientation, the W61 residue was shifted either towards the more hydrophilic N-terminal at residue 57, or to position 68 at the hydrophobic C-terminal end of the peptide. Peptide-induced vesicle fusion was demonstrated by fluorescence measurements using pyrene-labeled vesicles and by monitoring of vesicle size by gel filtration. The interaction between peptides and lipids was monitored by measurement of the intrinsic tryptophan fluorescence emission of the peptides. Fluorescence polarisation measurements, using diphenyl hexatriene, were carried out to follow changes in the lipid fluidity. The LCAT-(56-68) wild-type peptide and the two oblique variants, induced fusion of unilamellar dimyristoylglycerophosphocholine vesicles. Tryptophan fluorescence emission measurements showed a 12-14 nm blue shift upon addition of the wild-type peptide and of the W61-->68 variant to lipids, whereas the fluorescence of the W61-->57 variant did not change significantly. This observation supports the insertion of the more hydrophobic C-terminal residues into the lipid phase, as predicted by the theoretical calculations. In contrast, the 0 degrees variant peptide had no fusogenic activity, and it associated with lipids to form small discoidal lipid/peptide complexes. The phospholipid transition temperature was decreased after addition of the wild-type, the W61-->68 and W61-->57 fusogenic peptides, whereas the opposite effect was observed with the 0 degrees variant. The behaviour of the wild-type and variant LCAT-(56-68) peptides stresses the contribution of the hydrophobicity gradient along the axis of an amphipathic peptide to the mode of association of this peptide with lipids. This parameter consequently influences the structural modifications occurring to lipids upon association with amphipathic peptides.     

Specificity and promiscuity among naturally processed peptides bound to HLA-DR alleles

Naturally processed peptides were acid extracted from immunoaffinity-purified HLA-DR2, DR3, DR4, DR7, and DR8. Using the complementary techniques of mass spectrometry and Edman microsequencing, > 200 unique peptide masses were identified from each allele, ranging from 1,200 to 4,000 daltons (10-34 residues in length), and a total of 201 peptide sequences were obtained. These peptides were derived from 66 different source proteins and represented sets nested at both the amino- and carboxy-terminal ends with an average length of 15-18 amino acids. Strikingly, most of the peptides (> 85%) were derived from endogenous proteins that intersect the endocytic/class II pathway, even though class II molecules are thought to function mainly in the presentation of exogenous foreign peptide antigens. The predominant endogenous peptides were derived from major histocompatibility complex-related molecules. A few peptides derived from exogenous bovine serum proteins were also bound to every allele. Four prominent promiscuous self-peptide sets (capable of binding to multiple HLA-DR alleles) as well as 84 allele-specific peptide sets were identified. Binding experiments confirmed that the promiscuous peptides have high affinity for the binding groove of all HLA-DR alleles examined. A potential physiologic role for these endogenous self-peptides as immunomodulators of the cellular immune response is discussed.     

Nonspecific protease-catalyzed hydrolysis/synthesis of a mixture of peptides: product diversity and ligand amplification by a molecular trap

We sought to develop a peptide library in solution and dynamically screen this library for peptides that would bind to macromolecules of interest. Peptide diversity was achieved in an initial stock solution of peptides by using proteases under conditions in which both hydrolysis and synthesis occurred. As an example, a simple reaction containing YGG, FL and thermolysin resulted in the synthesis of YGGFL as well as many other undefined products. When low molecular weight products of a reaction containing VA, AL, and thermolysin were subsequently exposed to dipeptidase, 7 out of 9 potential dipeptides were observed. Incubation of protease with an hydrolysate of albumin and a radiolabeled peptide resulted in the radiolabel participating in reactions other than simple hydrolysis and, after 24 h, the specific activity of radiolabel was shown by high performance liquid chromatography to disperse to a level that would be necessary in the event of maximum theoretical diversity. When a binding macromolecule was exposed to this system, ligand production was amplified relative to reactions run in the absence of binding macromolecule. This protease-based peptide scrambling and binding system was utilized for the discovery of novel peptides that bind to fibrinogen.     

Characterization and biological significance of immunosuppressive peptide D2702.75-84(E --> V) binding protein. Isolation of heme oxygenase-1

This is the first report on peptidic inhibitors of heme oxygenase. Such peptides were originally developed from the immunomodulatory peptide 2702.75-84 which corresponds to amino acid residues 75 to 84 of the alpha1-helix of HLA-B2702 (2702.75-84) and has been shown to be immunosuppressive in vitro and in vivo. In vitro, 2702.75-84 inhibited cytotoxic T- and natural killer cell- mediated target cell lysis, and in vivo peptide therapy resulted in prolongation of heart and skin allograft survival in mice. The peptide was also shown to bind to heat shock protein 70. However, D-enantiomers of 2702.75-84 and derivatives thereof, while still being immunosuppressive, did not bind to heat shock protein 70. This study was designed to identify proteins binding to peptide D2702.75-84(E --> V) (rvnlrialry) consisting of D-amino acids. Compared with 2702.75-84 (RENLRIALRY), glutamic acid residue 76 (E) was replaced with valine (V). Affinity chromatography using immobilized D2702.75-84(E --> V) and mouse and human cell extracts, resulted in the isolation of heme oxygenase-1 (HO-1). Peptide D2702.75-84 inhibited HO activity in vitro in a dose dependent manner. Similar to what has been observed with other inhibitors of HO, administration of peptide into mice resulted in an up-regulation of HO-1 mRNA and protein, as well as enzyme activity in liver, spleen and kidney. Other peptides derived from 2702.75-84 with similar immunomodulatory activity displayed similar effects. In contrast, inactive derivatives of 2702.75-84 had no effect on HO activity. Therefore, the immunosuppressive effects of the described immunomodulatory peptides are similar to those of cobalt-protoporphyrin, a known up-regulator of HO-1. Our results suggest that HO-1 modulation may be a novel mechanism of immunomodulation.     

Comparative Study of SVMs and ANNs in Aquifer Water Level Prediction

In this research, a data-driven modeling approach, support vector machines (SVMs), is compared to artificial neural networks (ANNs) for predicting transient groundwater levels in a complex groundwater system under variable pumping and weather conditions. Various prediction horizons were used, including daily, weekly, biweekly, monthly, and bimonthly prediction horizons. It was found that even though modeling performance (in terms of prediction accuracy and generalization) for both approaches was generally comparable, SVM outperformed ANN particularly for longer prediction horizons when fewer data events were available for model development. In other words, SVM has the potential to be a useful and practical tool for cases where less measured data are available for future prediction. The study also showed high consistency between the training and testing phases of modeling when using SVM compared to ANN. While for the proposed SVM model the relative error of mean square error increased by an average of 42% from the training phase to testing the phase, the corresponding testing error of the ANN model raised by approximately seven times the training error.     

High affinity binding of TAR RNA by the human immunodeficiency virus type-1 tat protein requires base-pairs in the RNA stem and amino acid residues flanking the basic region

The binding site for tat protein on TAR RNA has been defined in quantitative terms using an extensive series of mutations. The relative dissociation constants for the mutant TAR RNAs were measured using a dual-label competition filter binding assay in which 35S-labelled wild-type TAR RNA (K1) was competed against 3H-labelled mutant TAR RNA (K2). The error in the self-competition experiment was usually less than 10% (e.g. K2/K1 = 1.07 +/- 0.05, n = 19) and the experimental data accurately matched theoretical curves calculated with fitted dissociation constants. Mutations in U23, a critical residue in the U-rich "bulge" sequence, or in either of the two base-pairs immediately above the "bulge", G26.C39 and A27.U38 reduced that affinity by 8- to 20-fold. Significant contributions to tat binding affinity were also made by the base-pairs located immediately below the bulge. For example, mutation of A22.U40 to U.A reduced tat affinity 5-fold, and mutation of G21.C41 to C.G reduced tat affinity 4-fold. The binding of a series of peptides spanning the basic "arginine-rich" sequence of tat was examined using both filter-binding and gel mobility shift assays. Each of the peptides showed significantly reduced affinities for wild-type TAR RNA compared to the tat protein. The ADP-2 (residues 43 to 72), ADP-3 (residues 48 to 72) and ADP-5 (residues 49 to 86) peptides were unable to discriminate between wild-type TAR RNA and TAR RNA mutants with the same fidelity as the tat protein. For example, these peptides showed no more than 3-fold reductions in affinity relative to wild-type TAR RNA for the U23-->C mutation in the bulge, or G26.G39-->C.G mutation in the stem of TAR RNA. By contrast, the ADP-I (residues 37 to 72), ADP-4 (residues 32 to 62) and ADP-6 (residues 32 to 72) peptides, which each carry amino acid residues from the "core" region of the tat protein have binding specificities that more closely resemble the protein. The ADP-4 and ADP-6 peptides showed between 4- and 7-fold reductions in affinity for the U23-->C or G26.C39-->C.G mutations. The ADP-1 peptide most closely resembles the protein in its binding specificity and showed 9-fold and 14-fold reductions in affinity for the two mutants, respectively. Chemical-modification interference assays using diethylpyrocarbonate (DEPC) and ethylnitrosourea (ENU) were also used to compare the binding properties of the tat protein and the tat-derived peptides.(ABSTRACT TRUNCATED AT 400 WORDS)     

Epitope mapping of C1 inhibitor autoantibodies from patients with acquired C1 inhibitor deficiency

We report six patients with acquired C1 inhibitor (C1-inh) deficiency associated with serum C1-inh autoantibodies and circulating cleaved (96 kDa), functionally inactive C1-inh. In three patients, all of whom had IgG-kappa paraproteins in their sera, the Abs were IgG-kappa. In the remaining three patients, the Abs were IgM (2 kappa, 1 lambda). These data suggest that all the Abs were monoclonal. The autoantibodies recognized two synthetic peptides (peptides 2 and 3), which spanned the reactive center of C1-inh. Binding to peptide 3 (residues 448-459) was greater than to peptide 2 (residues 438-449), suggesting that the epitope recognized by the autoantibodies was expressed principally by peptide 3. Both peptides inhibited the binding of the autoantibodies to C1-inh. None of the autoantibodies recognized peptide 1 (residues 428-440), and this peptide did not inhibit the binding of the autoantibodies to C1-inh. The use of substituted peptides suggested that residues Q452 and Q453 made significant contributions to the epitope, and computer modeling studies showed their side chains to be surface exposed in the intact molecule. However, computer modeling also showed that none of the side chains of the polar residues in peptide 2 were sufficiently close to Q452 and Q453 to be able to contribute to a shared epitope. As peptide 2 could inhibit the binding of C1-inh autoantibodies to peptide 3 and vice versa, we conclude that an autoepitope also exists in peptide 2. Computer modeling and the use of substituted peptides suggested that the sequence LLVF (residues 446-449) in peptide 2 is structurally similar to the sequence QQPF (residues 452-455) in peptide 3. We therefore conclude that there are two potential epitopes in the intact C1-inh molecule that are capable of binding to C1-inh autoantibodies.     

American Heart Association Report on the Public Access Defibrillation Conference, December 8-10, 1994. American Heart Association Taskforce on Automatic External Defibrillation

Experimental conditions (pH 6.5, 24 h reaction, peptide:biotin ratio 1:5) were defined for preferential incorporation of the biotin molecule in the N-terminal alpha-amino group of peptides. This strategy could be helpful in numerous applications when an entire peptide chain must remain accessible for antibody or receptor binding. We illustrate this advantage in a solid-phase enzyme immunoassay designed to detect antibodies specific for bovine beta-lactoglobulin present in rabbit or human sera. This test involves synthetic peptides biotinylated in different positions and immobilized on a solid phase. The use of biotin/streptavidin interactions permitted more efficient detection of specific anti-peptide antibodies than solid phases prepared using conventional passive-adsorption techniques. The highest levels of antibody binding were measured when biotinylation occurred at the N-terminal extremity of immobilized peptides.     

Naturally processed peptides from HLA-DQ7 (alpha1*0501-beta1*0301): influence of both alpha and beta chain polymorphism in the HLA-DQ peptide binding specificity

Self peptides bound to HLA-DQ7 (alpha1*0501-beta1*0301), one of the HLA molecules associated with protection against insulin-dependent diabetes mellitus, were characterized after their acid elution from immunoaffinity-purified HLA-DQ7 (alpha1*0501-beta1*0301) molecules. The majority of these self peptides derived from membrane-associated proteins including HLA class I, class II, class II-associated invariant chain peptide and the transferrin-receptor (TfR). By in vitro binding assays, the specificity of these endogenous peptides for HLA-DQ7 (alpha1*0501-beta1*0301) molecules was confirmed. Among these peptides, the binding specificity of the TfR 215-230 self peptide was further examined on a variety of HLA-DQ and DR dimers. Several findings emerged from this analysis: (1) this peptide displayed HLA-DQ allelic specificity, binding only to HLA-DQ7 (alpha1*0501-beta1*0301); (2) when either the DQalpha or DQbeta chain was exchanged, little or no binding was observed, indicating that specificity of HLA-DQ peptide binding was determined by polymorphic residues of both the alpha and beta chains. (3) Unexpectedly, the TfR 215-230 self peptide, eluted from DQ, was promiscuous with regard to HLA-DR binding. This distinct DR and DQ binding pattern could reflect the structure of these two molecules as recently evidenced by crystallography.     

Identification of T-cell epitopes using allele-specific ligand motifs

In order to facilitate the identification of T-cell epitopes as useful components of synthetic vaccines, we investigated the role of MHC molecules as the restriction element for the recognition of epitopes by the alpha beta receptor of T cells. MHC molecules are able to present thousands of different peptides to T cells, with all the peptides presented by one distinct type of MHC sharing common structural features. Our group analyzed these common characteristics concerning peptide length (only MHC I ligands) and anchor positions (MHC I and II ligands) occupied by a small set of closely related amino acids. Until now, for more than fifty MHC proteins allele-specific "peptide motifs" have been defined. The exact knowledge of MHC I peptide motifs allows for a prediction of CTL epitopes, and this kind of prediction has been successful in many cases over the last three years.     

Rapid identification of comigrating gel-isolated proteins by ion trap-mass spectrometry

In the search for novel nuclear binding proteins, two bands from a sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gel were analyzed and each was found to contain a number of proteins that subsequently were identified by tandem mass spectrometry (MS/MS) on a quadrupole ion trap instrument. The bands were digested with trypsin in situ on a polyvinylidene difluoride (PVDF) membrane following electroblot transfer. Analysis of a 2.5% aliquot of each peptide mixture by matrix assisted laser desorption/ionization-mass spectrometry (MALDI-MS) followed by an initial database search with the peptide masses failed to identify the proteins. The peptides were separated by reversed-phase capillary high performance liquid chromatography (HPLC) in anticipation of subsequent Edman degradation, but mass analysis of the chromatographic fractions by MALDI-MS revealed multiple, coeluting peptides that precluded this approach. Selected fractions were analyzed by capillary HPLC-electrospray ionization-ion trap mass spectrometry. Tandem mass spectrometry provided significant fragmentation from which full or partial sequence was deduced for a number of peptides. Two stages of fragmentation (MS3) were used in one case to determine additional sequence. Database searches, each using a single peptide mass plus partial sequence, identified four proteins from a single electrophoretic band at 45 kDa, and four proteins from a second band at 60 kDa. Many of these proteins were derived from human keratin. The protein identifications were corroborated by the presence of multiple matching peptide masses in the MALDI-MS spectra. In addition, a novel sequence, not found in protein or DNA databases, was determined by interpretation of the MS/MS data. These results demonstrate the power of the quadrupole ion trap for the identification of multiple proteins in a mixture, and for de novo determination of peptide sequence. Reanalysis of the fragmentation data with a modified database searching algorithm showed that the same sets of proteins were identified from a limited number of fragment ion masses, in the absence of mass spectral interpretation or amino acid sequence. The implications for protein identification solely from fragment ion masses are discussed, including advantages for low signal levels, for a reduction of the necessary interpretation expertise, and for increased speed.