Journal of Molecular Biology
Volume 295, Issue 4, 28 January 2000, Pages 879-890
Journal home page for Journal of Molecular Biology

Regular article
Sequence signals for generation of antigenic peptides by the proteasome: implications for proteasomal cleavage mechanism1

https://doi.org/10.1006/jmbi.1999.3392Get rights and content

Abstract

Proteasomal cleavage of proteins is the first step in the processing of most antigenic peptides that are presented to cytotoxic T cells. Still, its specificity and mechanism are not fully understood. To identify preferred sequence signals that are used for generation of antigenic peptides by the proteasome, we performed a rigorous analysis of the residues at the termini and flanking regions of naturally processed peptides eluted from MHC class I molecules. Our results show that both the C terminus (position P1 of the cleavage site) and its immediate flanking position (P1′) possess significant signals. The N termini of the peptides show these signals only weakly, consistent with previous findings that antigenic peptides may be cleaved by the proteasome with N-terminal extensions. Nevertheless, we succeed to demonstrate indirectly that the N-terminal cleavage sites contain the same preferred signals at position P1′. This reinforces previous findings regarding the role of the P1′ position of a cleavage site in determining the cleavage specificity, in addition to the well-known contribution of position P1. Our results apply to the generation of antigenic peptides and bare direct implications for the mechanism of proteasomal cleavage. We propose a model for proteasomal cleavage mechanism by which both ends of cleaved fragments are determined by the same cleavage signals, involving preferred residues at both P1 and P1′ positions of a cleavage site. The compatibility of this model with experimental data on protein degradation products and generation of antigenic peptides is demonstrated.

Introduction

Cytotoxic T lymphocytes recognize short peptides of eight to ten amino acid residues, the processing products of protein antigens, presented on the surface of an antigen presenting cell. The main processing pathway involves degradation of cytosolic proteins, mostly by the proteasome, and transport of the degradation products by the transporter associated with antigen processing (TAP) to the endoplasmic reticulum (ER), where peptide binding to major histocompatibility complex (MHC) class I molecules takes place. This is followed by trafficking of the MHC-peptide complexes to the cell surface for presentation to cytotoxic T lymphocytes (reviewed by Lehner and Cresswell 1996, York and Rock 1996, Pamer and Cresswell 1998). The specificity of each of these processing stages influences the selection of T-cell epitopes along a protein sequence. Here, we concentrate on the first processing stage, the protein cleavage by the proteasome, attempting to identify sequence signals that may play a role in the determination of protein cleavage sites.

There is accumulating evidence suggesting that the proteasome is involved in the generation of most antigenic peptides (reviewed by Koopmann et al 1997, Pamer and Cresswell 1998, Rock and Goldberg 1999). The proteasome, a multimeric proteinase that is ubiquitous in all eukaryotic cells, is the central enzyme of protein degradation in both the cytosol and nucleus (for a review, see Baumeister et al., 1998). The proteolytic core of this complex is formed by the 20 S proteasome, whose general architecture is conserved from archaebacteria to eukaryotes Lowe et al 1995, Groll et al 1997, but the identity of its subunits and their refined structure differ. The 20 S proteasome is composed of two copies of 14 subunits organized in four rings. In the yeast 20 S proteasome the outer two rings contain subunits of the form (α1…α7) and the inner two rings, where the active sites reside, contain subunits of the form (β1…β7).

Three major specific peptidase activities attributed to three active sites have been defined (e.g. Orlowski et al 1993, Groll et al 1997, Heinermeyer et al 1997, Dick et al 1998, Nussbaum et al 1998): trypsin-like activity for subunit β2 (cleavage after basic residues); chymotrypsin-like activity for β5 (cleavage after hydrophobic residues); and peptidylglutamyl-peptide hydrolytic activity (PGPH) for β1 (cleavage after acidic residues). Recent studies with yeast 20 S proteasome have indicated that the β2 and β1 also contribute substantially to the chymotrypsin-like activity Dick et al 1998, Nussbaum et al 1998. In vertebrates there are three other β subunits that are induced by increased levels of interferon-γ upon induction of an immune response: β1i (LMP2), β5i (LMP7), and β2i (MECL1) that replace β1, β5 and β2 (Y, X, and Z in human), respectively (reviewed by Pamer & Cresswell, 1998). Assembly of this “immuno-proteasome” affects the hierarchy of proteasomal cleavage, enhancing cleavage after basic and hydrophobic residues and inhibiting cleavage after acidic residues (e.g. Gaczynska et al., 1994). This is in accord with the type of residues required at the C terminus of antigenic peptides for binding to the F pocket of MHC class I molecules (Madden, 1995).

From studies of the sequences of degradation products it has become clear that the nature of the peptidase activities by themselves cannot explain the cleavage specificity (e.g. Niedermann et al 1995, Niedermann et al 1996, Ehring et al 1996, Nussbaum et al 1998). In the studied cases, the amino acid residues at the P1 position † of many cleavage sites conformed to the peptidase activities described above, but other residues were also observed, indicating a large variation. Furthermore, residues that were cleaved preferentially at one sequence location behaved differently at other positions, suggesting that the sequence context of the cleavage sites may also play a role. Indeed, Nussbaum et al. (1998) have recently demonstrated some correlations between the amino acid residues at position P1 of a cleavage site and residues in its vicinity, in a span of ± five residues. Holzhutter et al. (1999) performed a statistical analysis on the flanking regions of a small sample of sequences of degradation products reported in the literature, and also described correlations between the type of residues at P1 and specific residues in nearby positions. However, the nature of the correlations were different than those reported by Nussbaum et al. (1998). Also, several experimental studies have indicated the effect of changes in the peptide flanking residues on antigen presentation (e.g. Bergmann et al 1994, Shastri et al 1995, Bergmann et al 1996, Yellen-Shaw et al 1997, Vijh et al 1998). Still, the precise location of the residues which affected presentation has been somewhat controversial. The inconsistencies in these results are probably due to the small number of experimentally studied sequences, which may be insufficient for deducing general rules.

An alternative approach would be to use the available information on naturally processed peptide sequences that were eluted from MHC molecules, and analyze their flanking residues. These peptides were cleaved by the proteasome, and if specific cleavage signals exist they should present them at their termini and/or flanking sequences. Therefore, the question that we address here is, can we identify significant signals at the termini and flanking regions of naturally processed peptides that can be attributed to proteasomal cleavage? A first attempt in this direction was carried out by Niedermann et al. (1996), who experimentally determined the amino acid preferences at the C and N termini and flanking residues of degradation products, and compared these results with the frequency distribution of amino acids at the terminal and immediate flanking positions of a set of 134 MHC class I-bound peptides. They demonstrated strong conformity in the distribution of residues at the C termini of the two sets and a weaker similarity between the distributions of residues at the N termini. Based on these results they suggested that the proteasome is frequently involved in the generation of the C termini of antigenic peptides, and in the generation of some but not all of their N termini.

Here, we have carefully organized a database of 286 naturally processed peptide sequences from 215 different proteins, and have applied systematic computational analyses to search for potential cleavage signals at the amino and carboxy-terminal residues, and in the flanking regions of the peptides, spanning 50 residues on each side. The signals that we have identified in a completely objective way are consistent with reported experimental results and emphasize the importance of position P1′ in determining the cleavage specificity of both the N and C termini. We propose a model for proteasomal cleavage mechanism and generation of antigenic peptides that incorporates the computer analysis implications with those derived from experimental results. Our results and conclusions can serve as the basis for a computer algorithm that will predict the boundaries of the initial degradation products along a protein antigen.

Section snippets

N-terminal flanking region

The 274 N-terminal flanking fragments were divided randomly into two equal-size groups (see Methods). In each group the sequences were aligned by the first (PN) peptide residue (refer to Figure 1 for nomenclature of sequence positions). First, one group of 137 sequences was analyzed. The amino acid frequency distribution at each of the N1 to N50 positions was compared to the amino acid frequency distribution in the data set of non-homologous source proteins (background frequency distribution)

Discussion

Unraveling the rules for generation of antigenic peptides by the proteasome is essential for understanding the selection of immunodominant T-cell epitopes along the sequence of a protein antigen. Availability of a quantitative measure that reflects the likelihood of cleavage between any pair of residues within a given sequence may enable the prediction of the protein cleavage pattern. This may provide the means for detection of potential protein fragments that will be further processed as

Organization of peptide sequence database

Peptide sequences associated with MHC class I of mouse and/or human were taken from a compilation of MHC ligands which includes known T-cell epitopes and naturally processed peptides eluted from MHC molecules (Rammensee et al., 1995; see http://www.uni-tuebingen.de/uni/kxi). In the current analysis only naturally processed peptides longer than six amino acid residues were included. Each peptide sequence was verified with the original paper, keeping the peptides that were individually sequenced,

Acknowledgements

We are grateful to Klaus Eichmann for valuable and inspiring discussions. We thank Norman Grover for statistical consultation, Jutta Bachmann for her help in the data organization, and Ora Schueler-Furman for helpful comments on the manuscript. This work has benefited from valuable discussions with Robert Huber, Wolfgang Baumeister, Robert Tampe, Michelle Groll, and Alfred Goldberg. The study was supported by grants from the US-Israel Binational Science Foundation, and from the Israel Cancer

References (50)

  • P Paz et al.

    Discrete proteolytic intermediates in the MHC class I antigen processing pathway and MHC I-dependent peptide trimming in the ER

    Immunity

    (1999)
  • W.R Pearson

    Empirical statistical estimates for sequence similarity searches

    J. Mol. Biol.

    (1998)
  • S.J Powis et al.

    The rat cim effectTAP allele-dependent changes in a class I MHC anchor motif and evidence against C-terminal trimming of peptides in the ER

    Immunity

    (1996)
  • S Uebel et al.

    Specificity of the proteasome and the TAP transporter

    Curr. Opin. Immunol.

    (1999)
  • T Wenzel et al.

    Existence of a molecular ruler in proteasomes suggested by analysis of degradation products

    FEBS Letters

    (1994)
  • L.C Anton et al.

    Dissociation of proteasomal degradation of biosynthesized viral proteins from generation of MHC class I-associated antigenic peptides

    J. Immunol.

    (1998)
  • A.M Benham et al.

    Allelic differences in the relationship between proteasome activity and MHC class I peptide loading

    J. Immunol.

    (1998)
  • C.C Bergmann et al.

    Differential effects of flanking residues on presentation of epitopes from chimeric peptides

    J. Virol.

    (1994)
  • C.C Bergmann et al.

    Flanking residues alter antigenicity and immunogenicity of multi-unit CTL epitopes

    J. Immunol.

    (1996)
  • A Craiu et al.

    Two distinct proteolytic processes in the generation of a major histocompatibility complex class I-presented peptide

    Proc. Natl Acad. Sci. USA

    (1997)
  • S Daniel et al.

    Relationship between peptide selectivities of human transporters associated with antigen processing and HLA class I molecules

    J. Immunol.

    (1998)
  • B Ehring et al.

    Effects of major-histocompatibility-complex-encoded subunits on the peptidase and proteolytic activities of human 20 S proteasomes. Cleavage of proteins and antigenic peptides

    Eur. J. Biochem.

    (1996)
  • L.C Eisenlohr et al.

    Expression of a membrane protease enhances presentation of endogenous antigens to MHC class I-restricted T lymphocytes

    Cell

    (1992)
  • T Elliott et al.

    Processing of major histocompatibility class I-restricted antigens in the endoplasmic reticulum

    J. Exp. Med.

    (1995)
  • M Gaczynska et al.

    Peptidase activities of proteasomes are differentially regulated by the major histocompatibility complex-encoded genes for LMP2 and LMP7

    Proc. Natl Acad. Sci. USA

    (1994)
  • Cited by (51)

    • Immunoinformatics approach to designing a multi-epitope vaccine against Saint Louis Encephalitis Virus

      2021, Informatics in Medicine Unlocked
      Citation Excerpt :

      All nucleated cells contain MHC-I molecules which facilitate this discrimination process by CTLs by identifying small peptide fragments of a foreign protein that exhibit antigenic properties termed as epitopes. Therefore, the prediction of epitopes that can be detected by MHC-I molecules is crucial in order to better label pathogenic proteins [80,81]. A total of 55 epitope sequences were selected from the predictions made by NetCTL 1.2 server for MHC-I epitopes, which were further subjected to screening for antigenicity, toxicity, allergenicity and conservancy analysis using Vaxijen 2.0, AllerTOP v2.0 and ToxinPred respectively.

    • Major histocompatibility complex linked databases and prediction tools for designing vaccines

      2016, Human Immunology
      Citation Excerpt :

      However, despite these difficulties in constructing a consensus method, a few consensus prediction tools are available as web services; these include IEDB-MHC II consensus and NetMHCcons, both of which demonstrate increased epitope prediction accuracy (Table 4). The proteasome usually generates precursors of MHC binding ligands with an extension at the N-termini [46]. Therefore, prediction of proteasomal cleavage sites is an alternative way to predict CTL epitopes.

    • A melanoma multiepitope polypeptide induces specific CD8+ T-cell response

      2007, Cellular Immunology
      Citation Excerpt :

      Collected data of proteasome cleavage products (e.g. [9,10]) were incorporated into predictive algorithms [11–13]. In particular, it has been shown based on an analysis of about 300 naturally processed peptides, that both the C-terminal residue and its flanking residue play a role in determination of proteasome cleavage specificity [10]. Furthermore, it was demonstrated that changing the C-terminal flanking residue could enhance immunogenicity of a multiepitope peptide [14].

    • Immunological validation of the EpitOptimizer program for streamlined design of heteroclitic epitopes

      2007, Vaccine
      Citation Excerpt :

      The program will be further improved by adding new features, including matrices for additional MHC-I alleles of mouse, human and macaque and incorporation of matrices for MHC-II alleles. To facilitate peptide selection, a proteasome algorithm may be useful in conjunction with EpitOptimizer [41,42]. It is important to acknowledge that predictive algorithms are not 100% accurate [8,23,43].

    View all citing articles on Scopus
    1

    Edited by W. Baumeister

    View full text