An optimal viral peptide recognized by CD8+ T cells binds very tightly to the restricting class I major histocompatibility complex protein on intact cells but not to the purified class I protein

Amino-terminal alteration of the HLA-A*0201-restricted human immunodeficiency virus pol peptide increases complex stability and in vitro immunogenicity

Initial studies suggested that major histocompatibility complex class I-restricted viral epitopes could be predicted by the presence of particular residues termed anchors. However, recent studies showed that nonanchor positions of the epitopes are also significant for class I binding and recognition by cytotoxic T lymphocytes (CTLs). We investigated if changing nonanchor amino acids could increase class I affinity, complex stability, and T-cell recognition of a natural viral epitope. This concept was tested by using the HLA-A 0201-restricted human immunodeficiency virus type 1 epitope from reverse transcriptase (pol). Position 1 (P1) amino acid substitutions were emphasized because P1 alterations may not alter the T-cell receptor interaction. The peptide with the P1 substitution of tyrosine for isoleucine (I1Y) showed a binding affinity for HLA-A 0201 similar to that of the wild-type pol peptide in a cell lysate assembly assay. Surprisingly, I1Y significantly increased the HLA-A 0201-peptide complex stability at the cell surface. I1Y sensitized HLA-A 0201-expressing target cells for wild-type pol-specific CTL lysis as well as wild-type pol. Peripheral blood lymphocytes from three HLA-A2 HIV-seropositive individuals were stimulated in vitro with I1Y and wild-type pol. I1Y stimulated a higher wild-type pol-specific CTL response than wild-type pol in all three donors. Thus, I1Y may be an "improved" epitope for use as a CTL-based human immunodeficiency virus vaccine component. The design of improved epitopes has important ramifications for prophylaxis and therapeutic vaccine development.

Identification of a novel peptide derived from the melanocyte-specific gp100 antigen as the dominant epitope recognized by an HLA-A2.1-restricted anti-melanoma CTL line

Cytotoxic T lymphocytes (CTL) reactive with human melanoma tumor cells occasionally display cross-reactivity with normal melanocytes. Previously, we identified the melanocyte lineage-specific antigen gp 100 that is expressed by both melanoma cells and normal melanocytes, as a target antigen for tumor-infiltrating lymphocytes derived from a melanoma patient (TIL 1200). Here, we demonstrate that the oligoclonal HLA-A2.1-restricted TIL 1200 line is reactive with 2 distinct peptides derived from the gp 100 protein. Apart from the peptide corresponding to gp 100 amino acids 457-466, we identified the gp 100 peptide 154-162 as a second epitope recognized by TIL 1200. A 100-fold lower concentration of this novel gp 100 peptide was required for target-cell sensitization compared to peptide 457-466, indicating that the 154-162 peptide is the dominant gp 100 epitope for TIL 1200. Together with the recently described gp 100 280-288 epitope, 3 distinct CTL epitopes have now been identified in gp 100, all presented in the context of HLA-A2.1. Therefore, gp 100 is an attractive target antigen in the development of immuno-therapeutic protocols against melanoma.

Persistent high frequency of human immunodeficiency virus-specific cytotoxic T cells in peripheral blood of infected donors

Human immunodeficiency virus (HIV)-specific cytotoxic T lymphocytes (CTLs) are thought to play a major role in the immune response to HIV infection. The HIV-specific CTL response is much stronger than previously documented in an infectious disease, yet estimates of CTL frequency derived from limiting-dilution analysis (LDA) are relatively low and comparable to other viral infections. Here we show that individual CTL clones specific for peptides from HIV gag and pol gene products are present at high levels in the peripheral blood of three infected patients and that individual CTL clones may represent between 0.2% and 1% of T cells. Previous LDA in one donor had shown a frequency of CTL precursors of 1/8000, suggesting that LDA may underestimate CTL effector frequency. In some donors individual CTL clones persisted in vivo for at least 5 years. In contrast, in one patient there was a switch in CTL usage suggesting that different populations of CTLs can be recruited during infection. These data imply strong stimulation of CTLs, potentially leading some clones to exhaustion.

Induction and characterization of cytotoxic T-lymphocytes recognizing a mutated p21ras peptide presented by HLA-A*0201

The ras oncogene is frequently found to be activated in human cancer through point mutations at codons 12, 13 or 61. We explored whether these altered p21ras protein sequences contain peptide sequences that can activate naive CD8+ cytotoxic T lymphocytes (CTL). Several wild-type and mutated p21ras peptides were identified that carry a binding motif for human leukocyte antigen (HLA)-A*0201. Two peptides were found to bind strongly to this allele. CD8+ CTL bulk cultures specifically reacting with one of these peptides could be induced, using processing-defective T2 cells loaded with peptide CLLDILDTAGL as stimulators. The peptide is derived from p21ras, position 51-61, and carries a 61 Gln-->Leu mutation. In contrast, a 9-mer peptide CLLDILDTA corresponding to amino acid sequence 51-59 of wild-type p21ras did not yield reactive CTL cultures. T-cell clones with low affinity for the 11-mer peptide were isolated from CLLDILDTAGL-reactive bulk cultures. These T cells did not lyse melanoma cells transfected with 61-Leu N-ras, although lysis was found when these transfectants were pulsed with the 11-mer peptide. Possibly, T cells of higher affinity may be required to demonstrate processed peptide on the cell surface. The combined experiments suggest that a peptide derived from mutated p21ras can be recognized by HLA class I-restricted CTL, whereas an analogous wild-type p21ras peptide may not be immunogenic.

Peptide binding to MHC class I molecules: implications for antigenic peptide prediction

The human mayor histocompatibility complex class I molecule HLA-A2 preferentially binds peptides that contain Leu at P2 and Val or Leu at the C terminus. The other amino acids in the peptide also contribute to binding positively or negatively. It is possible to estimate the binding stability of HLA-A2 complexes containing particular peptides by applying coefficients, deduced from a large amount of binding data, that quantify the relative contribution of each amino acid at each position. In this review, we describe the molecular basis for these coefficients and demonstrate that estimates of binding stability based on the coefficients are generally concordant with experimental measurements of binding affinities. Peptides that contained cysteine were predicted less well, possibly because of complications resulting from peptide dimerization and oxidation. Apparently, peptide binding affinity is largely controlled by the rate of dissociation of the HLA/peptide/beta 2-microglobulin complex, whereas the rate of formation of the complex has less impact on peptide affinity. Although peptides that bind tightly to HLA-A2, including many antigenic peptides bind much more weakly. Therefore, a full understanding of why certain peptides are immunodominant will require further research.

Characterization of two Epstein-Barr virus epitopes restricted by HLA-B7

We have identified two epitopes for Epstein-Barr virus specific cytotoxic T lymphocytes (CTL) restricted by the common allele HLA-B7. They are EBNA 3C 881-9 (QPRAPIRPI) and EBNA 3A 379-387 (RPPIFIRRL). The epitopes conform well to the recently described motif for HLA-B7-binding peptides (Huckzo et al., J. Immunol. 1993. 151:2572). Titration of the peptides in CTL assays and detergent lysate binding assays revealed that extending the peptides at either the N or C terminus did not reduce their affinity for HLA-B7. This behavior contrasted with HLA-B51, which binds peptides with a similar motif to B7, and has identical amino acid residues at sites expected to form the "F" pocket of the peptide-binding groove. HLA-B51 also bound the peptide EBNA 3C 881-9, but was unable to bind peptides extended at the C terminus.

Naturally processed viral peptides recognized by cytotoxic T lymphocytes on cells chronically infected by human immunodeficiency virus type 1

We have established long-term cultures of several cell lines stably and uniformly expressing human immunodeficiency virus type 1 (HIV-1) in order to (a) identify naturally processed HIV-1 peptides recognized by cytotoxic T lymphocytes (CTL) from HIV-1-seropositive individuals and (b) consider the hypothesis that naturally occurring epitope densities on HIV-infected cells may limit their lysis by CTL. Each of two A2-restricted CD8+ CTL specific for HIV-1 gag or reverse transcriptase (RT) recognized a single naturally processed HIV-1 peptide in trifluoroacetic acid (TFA) extracts of infected cells: gag 77-85 (SLYNTVATL) or RT 476-484 (ILKEPVHGV). Both processed peptides match the synthetic peptides that are optimally active in cytotoxicity assays and have the consensus motif described for A2-associated peptides. Their abundances were approximately 400 and approximately 12 molecules per infected Jurkat-A2 cell, respectively. Other synthetic HIV-1 peptides active at subnanomolar concentrations were not present in infected cells. Except for the antigen processing mutant line T2, HIV-infected HLA-A2+ cell lines were specifically lysed by both A2-restricted CTL, although infected Jurkat-A2 cells were lysed more poorly by RT-specific CTL than by gag-specific CTL, suggesting that low cell surface density of a natural peptide may limit the effectiveness of some HIV-specific CTL despite their vigorous activity against synthetic peptide-treated target cells.

Human immunodeficiency virus-1 reverse transcriptase immunodominant CD4+ T cell epitopes: a peptide-based multiparametric assessment in the mouse

We previously identified an immunodominant CD4+ T cell determinant in the carboxy-terminal region of HIV-1 reverse transcriptase (RT528-543). The present study aimed at enumerating all the potential sites of HIV-1 RT recognized by Th cells in the BALB/c (H-2d) mouse model. To achieve this we used a panel of 62 overlapping 15-mer synthetic peptides covering the whole RT sequence to assay the following parameters: (i) immunogenicity in naive BALB/c mice injected either with peptides pools or individual peptides; (ii) antigenicity, as detected by their ability to restimulate in vitro T cells from BALB/c mice primed with native RT; (iii) MHC class II (Ad)-binding capacity as measured by the inhibition of the antigen-specific, Ad-restricted presentation of unfolded apamin (4-Acm) by fixed antigen-presenting cells to Ad/4-Acm-specific, interleukin-2-producing T hybridoma cells; and (iv) the presence of typical or degenerate consensus Ad-binding motifs. The results in this study permitted identification of three novel immunodominant RT mouse CD4+ T cell sites (RT276-290, RT375-389 and RT411-425) located in regions of limited polymorphism among RT from several HIV isolates. Some of these RT segments were found to be in the vicinity of B cell or H-2Kk- or HLA-A2-restricted cytotoxic T lymphocyte epitopes. Finally, the approach used in this study was found to be very efficient for enumerating most T cell recognition sites in a complex protein, a result that would have not been achieved by a single parameter-based analysis.

HIV-1 proteins in infected cells determine the presentation of viral peptides by HLA class I and class II molecules and the nature of the cellular and humoral antiviral immune responses--a review

The goals of molecular virology and immunology during the second half of the 20th century have been to provide the conceptual approaches and the tools for the development of safe and efficient virus vaccines for the human population. The success of the vaccination approach to prevent virus epidemics was attributed to the ability of inactivated and live virus vaccines to induce a humoral immune response and to produce antiviral neutralizing antibodies in the vaccinees. The successful development of antiviral vaccines and their application to most of the human population led to a marked decrease in virus epidemics around the globe. Despite this remarkable achievement, the developing epidemics of HIV-caused AIDS (accompanied by activation of latent herpesviruses in AIDS patients), epidemics of Dengue fever, and infections with respiratory syncytial virus may indicate that conventional approaches to the development of virus vaccines that induce antiviral humoral responses may not suffice. This may indicate that virus vaccines that induce a cellular immune response, leading to the destruction of virus-infected cells by CD8+ cytotoxic T cells (CTLs), may be needed. Antiviral CD8+ CTLs are induced by viral peptides presented within the peptide binding grooves of HLA class I molecules present on the surface of infected cells. Studies in the last decade provided an insight into the presentation of viral peptides by HLA class I molecules to CD8+ T cells. These studies are here reviewed, together with a review of the molecular events of virus replication, to obtain an overview of how viral peptides associate with the HLA class I molecules. A similar review is provided on the molecular pathway by which viral proteins, used as subunit vaccines or inactivated virus particles, are taken up by endosomes in the endosome pathway and are processed by proteolytic enzymes into peptides that interact with HLA class II molecules during their transport to the plasma membrane of antigen-presenting cells. Such peptides are identified by T-cell receptors present on the plasma membrane of CD4+ T helper cells. The need to develop viral synthetic peptides that will have the correct amino acid motifs for binding to HLA class I A, B, and C haplotypes is reviewed. The development of HIV vaccines that will stimulate, in an uninfected individual, the humoral (antibody) and cellular (CTL) immune defenses against HIV and HIV-infected cells, respectively, and may lead to protection from primary HIV infection are discussed.(ABSTRACT TRUNCATED AT 400 WORDS)

Induction of a major histocompatibility complex class I-restricted cytotoxic T-lymphocyte response to a highly conserved region of human immunodeficiency virus type 1 (HIV-1) gp120 in seronegative humans immunized with a candidate HIV-1 vaccine

Efforts to induce broadly reactive immunity against human immunodeficiency virus type 1 (HIV-1) have been impaired by the extent of sequence variation exhibited by this lentivirus. Cytotoxic T lymphocytes (CTL) specific for other viruses such as influenza virus have been shown to mediate immunity against divergent viral strains, a property that is related to the ability of CTL to recognize processed antigen derived from conserved viral proteins. A recent candidate HIV-1 vaccine regimen has been described in which subjects receive a primary immunization with a recombinant vaccinia virus expressing gp160 and then a booster immunization with recombinant gp160. Volunteers immunized with this regimen have exhibited augmented humoral responses and have also developed CD4+ and CD8+ CTL specific for gp160. In this report, we have identified the epitopes recognized by CD4+ and CD8+ CTL obtained from two vaccines. An immunodominant CD8+ CTL response was HLA-A3.1 restricted and recognized a 10-amino-acid epitope (gp120/38-47) in a highly conserved region of gp120. CTL specific for the epitope gp120/38-47 were able to lyse targets sensitized with peptides corresponding to all known natural sequence variants in this region. In addition, other HLA class I-restricted CTL epitopes were identified in relatively conserved regions of gp120 and gp41, and CD4+ CTL were shown to recognize two different regions of gp120. Thus, in these two volunteers, immunization with a single strain of HIV-1 induced CD4+ and CD8+ CTL that are specific for multiple conserved regions of HIV-1 and would be expected to recognize a broad range of viral isolates.

Two tyrosinase nonapeptides recognized on HLA-A2 melanomas by autologous cytolytic T lymphocytes

A number of cytolytic T lymphocyte (CTL) clones derived from several melanoma patients have been found to recognize a majority of melanomas from HLA-A2 patients. We have reported previously that two such CTL clones recognize a product of the tyrosinase gene that is presented by HLA-A2. Here we show that one of these CTL clones recognizes a peptide encoded by the first nine amino acids of the putative signal sequence of tyrosinase. The other CTL clone recognizes a different tyrosinase peptide corresponding to amino acids 368-376. Both peptides contain consensus motifs of HLA-A2 binding peptides.

A simple assay for detection of peptides promoting the assembly of HLA class I molecules

Synthetic peptides derived from influenza virus and human immunodeficiency virus were tested for their ability to promote the assembly of HLA-A2 and HLA-B51 molecules in T2 cell lysates. Specific assembly was detected by an enzyme-linked immunosorbent assay. The most significant HLA-A2 assembly was obtained in the presence of peptides known to be targets for HLA-A2-restricted cytotoxic T lymphocytes (influenza matrix M.58-66 and HIV Pol 476-484). Three of a batch of Nef peptides corresponding to epitopic regions for cytotoxic T lymphocytes, caused significant assembly of HLA-A2 (Nef 83-91, 137-145 and 144-153), but only at high concentrations (100 microM). As these peptides bound relatively weakly, it is unlikely that they are good candidates for HLA-A2-restricted CTL epitopes. Peptides matrix M.60-68, Nef 186-194, and Plasmodium falciparum sh.77-85 produced the most significant assembly of HLA-B51. These peptides have a dominant hydrophobic anchor residue (V, L. I) at position 9 that could occupy pocket "F". Our results also suggest that another hydrophobic residue (V, L) at position 3 or 4 may anchor to hydrophobic pocket "D" of HLA-B51. Proline at position 2 greatly increases HLA-B51 anchoring.

A quantitative assay to measure the interaction between immunogenic peptides and purified class I major histocompatibility complex molecules

A direct and sensitive biochemical assay to measure the interaction in solution between peptides and affinity-purified major histocompatibility complex (MHC) class I molecules has been generated. Specific binding reflecting the known class I restriction of cytotoxic T cell responses was obtained. Adding an excess of beta 2-microglobulin (beta 2m) significantly increased the rate of peptide association, but it did not affect the rate of dissociation. Binding was complicated by a rapid and apparently irreversible loss of functional MHC class I at 37 degrees C which might limit the life span of empty MHC class I thereby preventing the inadvertent exchange of peptides at the target cell surface. All class I molecules tested bound peptides of the canonical octa- to nona-meric length. However, one class I molecule, Kk, also bound peptides, which were much longer suggesting that the preference of class I molecules for short epitopes is not absolute and may be caused by factors other than the peptide-MHC class I binding event itself.

A ubiquitous protein is the source of naturally occurring peptides that are recognized by a CD8+ T-cell clone

We previously isolated from mouse spleen an octapeptide (LSPFPFDL) that in association with the class I major histocompatibility complex protein Ld is recognized by the antigen-specific receptor of an alloreactive CD8+ T-cell clone (2C). Guided by an assay dependent upon the same 2C T-cell receptor, we have now isolated from the same source another naturally occurring peptide. The second peptide (VAITRIEQLSPFPFDL) includes the entire octapeptide sequence and preliminary evidence suggests that it may be a natural precursor of the octapeptide. On finding extensive sequence homology between the 16-mer and part of human 2-oxoglutarate dehydrogenase, we determined the cDNA sequence of mouse 2-oxoglutarate dehydrogenase and found that the deduced amino acid sequence matches precisely the two naturally occurring peptides, indicating their origin by cellular processing of this ubiquitous self protein.

Comparison of the P2 specificity pocket in three human histocompatibility antigens: HLA-A*6801, HLA-A*0201, and HLA-B*2705

Coordinates from x-ray structures of HLA-A*6801, HLA-A*0201, and HLA-B*2705 were analyzed to examine the basis for their selectivity in peptide binding. The pocket that binds the side chain of the peptide's second amino acid residue (P2 residue) shows a preference for Val, Leu, and Arg in these three HLA subtypes, respectively. The Arg-specific pocket of HLA-B*2705 differs markedly from those of HLA-A*0201 and HLA-A*6801, as a result of numerous differences in the side chains that form the pocket's surface. The cause of the specificity differences between HLA-A*0201 and HLA-A*6801 is more subtle and depends both on a change in conformation of pocket residue Val-67 and on a sequence difference at residue 9. The Val-67 conformational change appears to be caused by a shift in the position of the alpha 1-domain alpha-helix relative to the beta-sheet in the cleft and may, in fact, depend on amino acid differences remote from the P2 pocket. Analysis of the stereochemistry of the P2 side chain interacting with its binding pocket permits an estimate to be made of its contribution to the free-energy change of peptide binding.

A soluble, single-chain Kd molecule produced by yeast selects a peptide repertoire indistinguishable from that of cell-surface-associated Kd

Peptide binding to a soluble, single-chain Kd protein produced by the yeast strain Kluyveromyces lactis, and to Kd molecules on Kd-expressing cells (P815) was studied using radiolabeled Kd-restricted peptides. The stability of the peptide-Kd complexes formed was monitored in the absence and presence of unlabeled competitor peptides. Radioiodination of the Tyr anchor residue in position 2 of the peptide interferes with binding. A Kd-biased peptide library and a modified antigenic peptide in which a second Tyr was added in positions 6 and 8, respectively, were therefore used to assay binding. Recombinant and cell-associated Kd molecules are very similar in the following respects: the ease with which the proteins can be loaded with labeled peptide; the spectrum of peptides selected from a peptide library; the stability of the labeled peptide-Kd complex formed; and the ability to partially dissociate the class I-peptide complex with exogenous, unlabeled peptides. These results imply that measurements of peptide binding to soluble Kd molecules are a reliable indicator of the peptide-binding properties of Kd proteins on living cells. The large quantities of soluble recombinant Kd protein currently available represent an invaluable tool not only for dissecting the molecular mechanisms of antigen presentation but also for vaccinations and the design of T cell-specific toxins.

The effect of anchor residue modifications on the stability of major histocompatibility complex class I-peptide interactions

Anchor residues in peptides determine the specificity of binding to major histocompatibility complex class I molecules through interactions of their side chains with pockets in the peptide-binding groove. We have compared the kinetics of association of a Sendai virus nucleoprotein-derived peptide (FAPGNYPAL, termed SV9) with H-2Kb class I molecules, and the same peptide iodinated on the anchor residue tyrosine (125I-SV9). Even though the association rates were too rapid for direct measurements, competition studies indicated that they were similar for SV9 and 125I-SV9. To measure the binding of non-radioactive SV9 directly, SV9 was tritiated (3H-SV9). 3H-SV9 remained stably associated with H-2Kb molecules, whereas 125I-SV9 dissociated in a temperature-dependent fashion. Thus, modifications on anchor residues do not necessarily have to affect the specificity and association kinetics of peptide binding to class I molecules but can affect the stability of the resulting class I-peptide interaction. The dissociation of peptides with modified and, more generally, suboptimal anchor residue side chains may explain the presence of empty class I molecules and free class I heavy chains at the cell surface.

In vivo priming of cytotoxic T lymphocyte responses in relation to in vitro up-regulation of major histocompatibility complex class I molecules by short synthetic peptides

Cytotoxic T lymphocytes (CTL) recognize target antigens as short peptides presented by major histocompatibility complex class I molecules (MHC-I). Externally added peptides can sensitize target cells by binding directly to MHC-I without any need for internal processing. Those which are similar in length to endogenously processed peptides are more potent in this respect than slightly longer peptides. Peptide MHC-I interactions can also be reflected as up-regulation of MHC-I in vitro on certain cells. We have compared the capacity of Db, Kb- and Ld-binding peptides, which are slightly different in length, to up-regulate MHC-I in vitro with their immunogenicity in vivo, in relation to generation of CTL responses. A clear correlation between these two different functions was found. We have also modified a 9-mer Db-binding peptide by adding cystein to the amino terminus and lysine to the amino- or carboxy terminus and studied the effects on MHC-I up-regulation and in vivo immunogenicity. Cystein and lysine contain reactive groups which are likely to influence the binding of modified peptides into the antigen-binding groove of Db. These small modifications of the optimal 9-mer peptide strongly influenced their functions but still there was a correlation between MHC-I up-regulation and CTL responses. Up-regulation of MHC-I in vitro may reflect a capacity of peptides to accumulate on the surface of particular antigen-presenting cells in vivo.

Antivirally protective cytotoxic T cell memory to lymphocytic choriomeningitis virus is governed by persisting antigen

The basis of antiviral protection by memory cytotoxic T lymphocytes (CTL) was investigated in vivo and in vitro using lymphocytic choriomeningitis virus (LCMV) and recombinant vaccinia viruses expressing the LCMV-glycoprotein (vacc-GP) or -nucleoprotein (vacc-NP). The widely replicating LCMV with a tendency to persist induced solid long-term protective memory. The poorly replicating vaccinia recombinant viruses revealed in the vaccinated host that the antiviral capacity of the secondary immune T cell response and the protection against lethal LCM was dependent upon the immunizing antigen and its dose. Protection against lethal choriomeningitis is less sensitive to assess memory because it depends upon high levels of CTL precursors (p) and/or on an activated state of memory CTL. In contrast, antiviral protection measured as the capacity of the primed host to reduce virus titers after challenge infection correlated with elevated CTLp frequencies after immunization with live LCMV or recombinant vaccinia virus-expressing the major LCMV epitope. CTLp frequencies were constantly increased up to 70 d for LCMV immune mice, but rapidly decreased a few weeks after immunization with low dose vaccinia recombinant virus. For example, mice primed with 2 x 10(6) plaque-forming units (PFU) of vacc-NP, or 2 x 10(2) PFU, or 2 x 10(6) PFU of vacc-GP were antivirally protected on day 7 but not after day 30 when CTLp could not be measured any longer in vitro. However, greater priming doses of vacc-NP (10(4) or 2 x 10(6) PFU) as well as LCMV (2 x 10(2) PFU) induced elevated levels of CTLp and antiviral protection for 60 d or longer. Adoptive transfer experiments of immune spleen cells into syngeneic recipients without addition of antigen demonstrated that maintenance of the antiviral protective capacity of the transferred cells depended on the presence of viral antigen. Thus, antiviral protection by memory CTL may be rather short-lived since it is based on activated T cells continuously stimulated by persisting antigen. This is best achieved by high immunizing antigen doses yielded either by widely replicating viruses or high doses of poorly replicating recombinant vaccines.

Extensive peptide ligand exchange by surface class I major histocompatibility complex molecules independent of exogenous beta 2-microglobulin

Certain class I major histocompatibility complex molecules expressed on live cells have been shown to bind exogenous peptide ligands. However, it remains controversial whether this binding occurs by peptide exchange or to empty surface class I molecules. In this report we compare the surface binding and dissociation of two virus-derived ligands of the Ld class I molecule of the mouse. The peptide ligands were previously identified in immune responses to cytomegalovirus or lymphochoriomeningitis virus as immunodominant, optimally sized, and Ld restricted. Ligand dissociation was monitored on live cells indirectly by measuring the surface turnover of Ld-peptide complexes or directly by using labeled peptides. The cytomegalovirus-derived and lymphochoriomeningitis virus-derived peptides appeared to dissociate relatively rapidly; however, the cytomegalovirus-derived peptide had a more rapid off-rate than the lymphochoriomeningitis-derived peptide. Furthermore, these rates of dissociation appear to span that seen with endogenous Ld-associated peptides expressed by cells at 37 degrees C. Exploiting the extraordinary accessibility of the surface Ld ligand binding site we developed an assay to quantitate peptide ligand exchange. Cells were precoated with saturating amounts of unlabeled peptide by overnight incubation and were then tested for secondary binding of labeled peptides in a 4-h assay. Our results unequivocally demonstrate the potential for surface class I molecules to undergo peptide exchange. Furthermore, peptide exchange was found to be largely independent of exogenous beta 2-microglobulin. This result implies that beta 2-microglobulin association and not beta 2-microglobulin exchange is the critical factor in peptide exchange by surface class I molecules. Because of the exquisite ability of T cells to discriminate different amounts of ligand bound to class I, the binding of exogenous peptides could play a critical role in normal or aberrant immune responses.