Immune evasion by adenovirus E3 proteins: exploitation of intracellular trafficking pathways.

Adenoviruses (Ads) are nonenveloped viruses which replicate and assemble in the nucleus. Therefore, viral membrane proteins are not directly required for their multiplication. Yet, all human Ads encode integral membrane proteins in the early transcription unit 3 (E3). Previous studies on subgenus C Ads demonstrated that most E3 proteins exhibit immunomodulatory functions. In this review we focus on the E3 membrane proteins, which appear to be primarily devoted to remove critical recognition structures for the host immune system from the cell surface. The molecular mechanism for removal depends on the E3 protein involved: E3/19K prevents expression of newly synthesized MHC molecules by inhibition of ER export, whereas E3/10.4-14.5K down-regulate apoptosis receptors by rerouting them into lysosomes. The viral proteins mediating these processes contain typical transport motifs, such as KKXX, YXXphi, or LL. E3/49K, another recently discovered E3 protein, may require such motifs to reach a processing compartment essential for its presumed immunomodulatory activity. Thus, E3 membrane proteins exploit the intracellular trafficking machinery for immune evasion. Conspicuously, many E3 membrane proteins from Ads other than subgenus C also contain putative transport motifs. Close inspection of surrounding amino acids suggests that many of these are likely to be functional. Therefore, Ads might harbor more E3 proteins that exploit intracellular trafficking pathways as a means to manipulate immunologically important key molecules. Differential expression of such functions by Ads of different subgenera may contribute to their differential pathogenesis. Thus, an unexpected link emerges between viral manipulation of intracellular transport pathways and immune evasion.

Subversion of host defense mechanisms by adenoviruses.

Adenoviruses (Ads) cause acute and persistent infections. Alike the much more complex herpesviruses, Ads encode numerous immunomodulatory functions. About a third of the viral genome is devoted to counteract both the innate and the adaptive antiviral immune response. Immediately upon infection, E1A blocks interferon-induced gene expression and the VA-RNA inhibits interferon-induced PKR activity. At the same time, E1A reprograms the cell for DNA synthesis and induces the intrinsic cellular apoptosis program that is interrupted by E1B/19K and E1B/55K proteins, the latter inhibits p53-mediated apoptosis. Most other viral stealth functions are encoded by a separate transcription units, E3. Several E3 products prevent death receptor-mediated apoptosis. E3/14.7K seems to interfere with the cytolytic and pro-inflammatory activities of TNF while E3/10.4K and 14.5K proteins remove Fas and TRAIL receptors from the cell surface by inducing their degradation in lysosomes. These and other functions that may afect granule-mediated cell death might drastically limit lysis by NK cells and cytotoxic T cells (CTL). Moreover, Ads interfere with recognition of infected cell by CTL. The paradigmatic E3/19K protein subverts antigen presentation by MHC class I molecules by inhibiting their transport to the cell surface. In concert, these viral countermeasures ensure prolonged survival in the infected host and, as a consequence, facilitate transmission. Elucidating the molecular mechanisms of Ad-mediated immune evasion has stimulated corresponding research on other viruses. This knowledge will also be instrumental for designing better vectors for gene therapy and vaccination, and may lead to a more rational treatment of life-threatening Ad infections, e.g. in transplantation patients.

Immunomodulatory functions encoded by the E3 transcription unit of adenoviruses.

Persistent viruses have evolved multiple strategies to escape the host immune system. One important prerequisite for efficient viral reproduction in the face of an ongoing immune response is prevention of premature lysis of infected cells. A number of viruses achieve this goal by interfering with antigen presentation and recognition of infected cells by cytotoxic T cells (CTL). Another viral strategy aims to block apoptosis triggered by host defense mechanisms. Both types of strategies seem to be realized by human adenoviruses (Ads). The early transcription unit E3 of Ads encodes proteins that inhibit antigen presentation by MHC class I molecules as well as apoptosis induced by tumor necrosis factor alpha (TNF-alpha) and Fas ligand (FasL). Here, we will describe the organization of the E3 regions of different Ad subgroups and compare the structure and functions of the known immunomodulatory E3 proteins.

The luminal part of the murine cytomegalovirus glycoprotein gp40 catalyzes the retention of MHC class I molecules.

Murine cytomegalovirus (MCMV) interferes with the MHC class I pathway of antigen presentation. The type I transmembrane glycoprotein gp40, encoded by the gene m152, retains major histocompatibility complex (MHC) class I complexes in the endoplasmic reticulum (ER)-Golgi intermediate compartment (ERGIC)/cis-Golgi. These MHC class I complexes are stable, show an extended half-life and do not exchange beta(2)-microglobulin, whereas gp40 reaches an endosomal/lysosomal compartment where it subsequently is degraded. The analysis of regions within the viral protein that are essential for MHC class I retention revealed that a secreted form of gp40, lacking the cytoplasmic tail and the transmembrane region, still has the capacity to retain MHC class I complexes. Continuous expression of gp40 is not required for MHC class I retention. Our data indicate that the retention of MHC class I complexes in the ERGIC/cis-Golgi is triggered by gp40 and does not require the further presence of the viral protein.

The amyloid precursor-like protein 2 associates with the major histocompatibility complex class I molecule K(d).

Amyloid precursor-like protein 2 (APLP2) is a member of a protein family related to the amyloid precursor protein, which is implicated in Alzheimer's disease. Little is known about the physiological function of this protein family. The adenovirus E3/19K protein binds to major histocompatibility complex (MHC) class I antigens in the endoplasmic reticulum, thereby preventing their transport to the cell surface. In cells coexpressing E3/19K and the MHC K(d) molecule, K(d) is associated with E3/19K and two cellular protein species with masses of 100 and 110 kDa, termed p100/110. Interestingly, p100/110 are released from the complex upon the addition of K(d)-binding peptides, suggesting a role for these proteins in peptide transfer to MHC molecules. Here we demonstrate by microsequencing, reactivity with APLP2-specific antibodies, and comparison of biochemical parameters that p100/110 is identical to human APLP2. We further show that the APLP2/K(d) association does not require the physical presence of E3/19K. Thus, APLP2 exhibits an intrinsic affinity for the MHC K(d) molecule. Similar to the binding of MHC molecules to the transporter associated with antigen processing, complex formation between APLP2 and K(d) strictly depends upon the presence of beta(2)-microglobulin. Conditions that prolong the residency of K(d) in the endoplasmic reticulum lead to a profound increase of the association and a drastic reduction of APLP2 transport. Therefore, this unexpected interplay between these unrelated molecules may have implications for both MHC antigen and APLP2 function.

A cytomegalovirus glycoprotein re-routes MHC class I complexes to lysosomes for degradation.

Mouse cytomegalovirus (MCMV) early gene expression interferes with the major histocompatibility complex class I (MHC class I) pathway of antigen presentation. Here we identify a 48 kDa type I transmembrane glycoprotein encoded by the MCMV early gene m06, which tightly binds to properly folded beta2-microglobulin (beta2m)-associated MHC class I molecules in the endoplasmic reticulum (ER). This association is mediated by the lumenal/transmembrane part of the protein. gp48-MHC class I complexes are transported out of the ER, pass the Golgi, but instead of being expressed on the cell surface, they are redirected to the endocytic route and rapidly degraded in a Lamp-1(+) compartment. As a result, m06-expressing cells are impaired in presenting antigenic peptides to CD8(+) T cells. The cytoplasmic tail of gp48 contains two di-leucine motifs. Mutation of the membrane-proximal di-leucine motif of gp48 restored surface expression of MHC class I, while mutation of the distal one had no effect. The results establish a novel viral mechanism for downregulation of MHC class I molecules by directly binding surface-destined MHC complexes and exploiting the cellular di-leucine sorting machinery for lysosomal degradation.

The adenovirus E3/10.4K-14.5K proteins down-modulate the apoptosis receptor Fas/Apo-1 by inducing its internalization.

Adenoviruses (Ads) have evolved multiple mechanisms to evade the host immune response. Several of the immunomodulatory Ad proteins are encoded in early transcription unit 3 (E3). The E3/19K protein interferes with antigen presentation and T cell recognition, whereas the E3/10.4K, 14.5K, and 14.7K proteins can protect cells from tumor necrosis factor alpha-mediated lysis. Here, we describe an additional activity of E3 proteins. Transfectants expressing all E3 proteins of Ad2 exhibit a profound reduction of the apoptosis receptor CD95 (Fas, APO-1) on the cell surface. In contrast, cells expressing only the E3A region have normal Fas levels. Thus, one of the E3B proteins (10.4K, 14.5K, or 14.7K) seems to be responsible for this effect. To identify the E3B products involved, each individual E3B ORF was selectively disrupted. Examination of stable cell lines containing the mutated E3 regions showed that Fas expression is restored when either the 10.4K or the 14.5K ORF is disrupted, whereas mutation of the 14.7K ORF does not rescue Fas expression. Loss of Fas on the cell surface is accompanied by a similar decrease of total Fas levels. However, in the presence of lysosomotropic agents Fas accumulates in endosomal/lysosomal vesicles, indicating that 10.4K-14.5K induce internalization and degradation of Fas. Down-regulation of Fas but not CD40 is also observed during infection and as a consequence, Ad-infected cells are protected from Fas-mediated apoptosis. Thus, the Fas system is implicated in Ad pathogenesis.

Tumor necrosis factor alpha induces the adenovirus early 3 promoter by activation of NF-kappaB.

The early transcription unit 3 (E3) of human adenoviruses encodes proteins which appear to subvert host defense mechanisms. For example, the E3/19K protein inhibits the transport of major histocompatibility complex (MHC) class I molecules to the cell surface and thereby prevents cell lysis by cytotoxic T cells. Tumor necrosis factor alpha (TNF) stimulates expression of MHC molecules on the cell surface of normal cells but not of E3(+) cells, rather, a further reduction of MHC expression is evident. This was attributed to the increased expression of E3/19K upon TNF treatment, an effect also observed for other E3 proteins. We investigated the mechanism of the TNF-mediated up-regulation of E3 products. We show that TNF stimulates expression of a luciferase reporter gene driven by the E3 promoter. Mutation of individual transcription factor binding sites within the E3 promoter reveals the importance of the NF-kappaB binding site kappa2 for TNF inducibility. Electrophoretic mobility shift assays using antibodies directed against various members of the NF-kappaB family demonstrate that stimulation by TNF is mediated by the p50-p65 NF-kappaB complex. TNF inducibility does not depend on coexpression of E1A and can be observed during infection. Interestingly, the E3 promoter seems to be the only early promoter responsive to TNF and the only adenovirus promoter containing an NF-kappaB site. The implications of this regulatory mechanism for the adenovirus life cycle and its pathogenesis are discussed.

Early region 3 of adenovirus type 19 (subgroup D) encodes an HLA-binding protein distinct from that of subgroups B and C.

Early region 3 (E3) of human adenoviruses (Ads) codes for proteins that appear to control viral interactions with the host. For example, the most abundant E3 protein, E3/19K, inhibits the transport of newly synthesized class I major histocompatibility molecules to the cell surface, thereby interfering with antigen presentation. So far, the E3 regions of Ad subgroups A, B, C, and F have been characterized. We have cloned the E3A region of Ad type 19a (Ad19a), which belongs to the largest subgroup, D, and causes epidemic keratoconjunctivitis in humans. The sequence reveals five open reading frames (ORFs) with the potential to encode the Ad19 equivalent of pVIII, as well as proteins 12.2K, 16.2K, and 18.6K. The last ORF predicts a novel 49K protein which has no counterpart in other subgroups. Both the sequence and the overall organization of the E3 region from Ad19a shows a closer relationship to group B than to group C Ads. The 18.6K ORF represents the Ad19 homolog of the Ad2 E3/19K protein. By using 293 cells stably transfected with the Adl9a E3A region, we showed by immunoprecipitation, pulse-chase experiments, and fluorescence-activated cell sorter analysis that the Ad19 E3/19K protein binds to and prevents the transport of major histocompatibility complex molecules to the cell surface. The similar but distinct functional activity of the Ad19 E3/19K protein, combined with the new sequence which differs from those of subgroup B and C proteins, allows a more precise definition of amino acids essential for HLA binding.

Activation of transcription factor NF-kappaB by the adenovirus E3/19K protein requires its ER retention.

We have recently shown that the accumulation of diverse viral and cellular membrane proteins in the ER activates the higher eukaryotic transcription factor NF-kappaB. This defined a novel ER-nuclear signal transduction pathway, which is distinct from the previously described unfolded protein response (UPR). The well characterized UPR pathway is activated by the presence of un- or malfolded proteins in the ER. In contrast, the ER stress signal which activates the NF-kappaB pathway is not known. Here we used the adenovirus early region protein E3/19K as a model to investigate the nature of the NF-kappaB-activating signal emitted by the ER. E3/19K resides in the endoplasmic reticulum where it binds to MHC class I molecules, thereby preventing their transport to the cell surface. It is maintained in the ER by a retention signal sequence in its carboxy terminus, which causes the protein to be continuously retrieved to the ER from post-ER compartments. Mutation of this sequence allows E3/19K to reach the cell surface. We show here that expression of E3/19K potently activates a functional NF-kappaB transcription factor. The activated NF-kappaB complexes contained p50/p65 and p50/c-rel heterodimers. E3/19K interaction with MHC class I was not important for NF-kappaB activation since mutant proteins which no longer bind MHC molecules remained fully capable of inducing NF-kappaB. However, activation of both NF-kappaB DNA binding and kappaB-dependent transactivation relied on E3/19K ER retention: mutants, which were expressed on the cell surface, could no longer activate the transcription factor. This identifies the NF-kappaB-activating signal as the accumulation of proteins in the ER membrane, a condition we have termed "ER overload." We show that ER overload-mediated NF-kappaB activation but not TNF-stimulated NF-kappaB induction can be inhibited by the intracellular Ca2+ chelator TMB-8. Moreover, treatment of cells with two inhibitors of the ER-resident Ca(2+) -dependent ATPase, thapsigargin and cyclopiazonic acid, which causes a rapid release of Ca2+ from the ER, strongly activated NF-kappaB. We therefore propose that ER overload activates NF-kappaB by causing Ca2+ release from the ER. Because NF-kappaB plays a key role in mounting an immune response, ER overload caused by viral proteins may constitute a simple antiviral response with broad specificity.

Tumor necrosis factor alpha increases expression of adenovirus E3 proteins.

Human adenovirus can cause persistent infections in man. Implicated in this phenomenon is the early transcription unit 3 (E3) of the virus which encodes proteins that are primarily devoted to counteract the lytic attack by the host immune system: Several E3 proteins (14.7K, 10.4K and 14.5K) protect infected cells from the lytic activity of tumor necrosis factor alpha (TNF) while the most abundant E3 protein, E3/19K, inhibits lysis by cytotoxic T cells. E3/19K interacts with class I histocompatibility (MHC) antigens in the rough endoplasmic reticulum, thereby preventing transport of MHC molecules to the cell surface and, consequently, MHC-restricted T cell recognition. In addition, the 10.4K and 14.5K proteins downregulate cell surface expression of the epidermal growth factor receptor. Interestingly, adenovirus-mediated pneumonia in mice is accompanied by induction of TNF, a cytokine known to enhance MHC expression. We previously showed that TNF is unable to restore MHC class I expression in E3/19K transfected cells but rather leads to a further reduction of MHC antigens. This effect correlated with an increased production of E3/19K mRNA and protein. We now find in addition an upregulation of other E3 proteins in transfected as well as in infected cells. This coordinated upregulation of E3 proteins indicates that TNF stimulates the E3 promoter, probably by activating the transcription factor NF-kappa B. Thus, a novel interaction between the immune system and adenovirus is described in which the virus takes advantage of an immune mediator to promote expression of several immunosubversive proteins supporting its escape from immunosurveillance.

Conserved cysteine residues within the E3/19K protein of adenovirus type 2 are essential for binding to major histocompatibility complex antigens.

The E3/19K protein of human adenovirus type 2 is a resident transmembrane glycoprotein of the endoplasmic reticulum. Its capacity to associate with class I histocompatibility (MHC) antigens abrogates cell surface expression and the antigen presentation function of MHC antigens. At present, it is unclear exactly which structure of the E3/19K protein mediates binding to MHC molecules. Apart from a stretch of approximately 20 conserved amino acids in front of the transmembrane segment, E3/19K molecules from different adenovirus subgroups (B and C) share little homology. Remarkably, the majority of cysteines are conserved. In this report, we examined the importance of cysteine residues for the structure and function of E3/19K. We show that E3/19K contains intramolecular disulfide bonds. By using site-directed mutagenesis, individual cysteines were replaced by serines and mutant proteins were stably expressed in 293 cells. On the basis of the differential binding of monoclonal antibody Tw1.3 and cyanogen bromide cleavage experiments, a structural model of E3/19K is proposed, in which Cys-11 and Cys-28 as well as Cys-22 and Cys-83 are linked by disulfide bonds. Both disulfide bonds (all four cysteines) are absolutely critical for the interaction with human MHC antigens. This was demonstrated by three criteria: loss of E3/19K coprecipitation, lack of transport inhibition, and normal cell surface expression of MHC molecules. Mutation of the three other cysteines had no effect. This indicates that a conformational determinant based on two disulfide bonds is crucial for the function of the E3/19K molecule, namely, to bind and to inhibit transport of MHC antigens.

Identification of amino acids within the MHC molecule important for the interaction with the adenovirus protein E3/19K.

The E3/19K protein of human adenovirus type 2 binds to class I MHC Ags thereby interfering with their cell surface expression and Ag presentation function. Currently, it is unclear exactly which structure of MHC molecules is recognized by the E3/19K protein. We have previously demonstrated that the murine H-2Kd Ag is able to associate with E3/19K, whereas the allelic H-2Kk molecule is not. By using exon shuffling between Kd and Kk molecules, the alpha 1 and alpha 2 domains of MHC class I molecules were identified as essential structures for binding the viral protein. In this report, we have examined the contribution of individual amino acids within the alpha 2 domain of MHC for binding E3/19K. First, we show that within this domain the alpha-helical part is most important for the interaction with E3/19K. By using site-directed mutagenesis, Kd-specific amino acids were introduced into the alpha-helix of the alpha 2 domain of Kk. By using the expression of mutagenized proteins in E3/19K+ cells, we have identified Tyr 156 and Leu 180 as being essential for the association with the E3/19K protein. In addition, Kd residue Glu 163 seems to contribute to the complex formation. Furthermore, analysis of a panel of Kd/Dd recombinants indicates that a similar region in the Dd molecule, namely, the C-terminal half of the alpha 2 domain, affects binding to E3/19K. Combining these results with Ab binding data, we present two alternative models of how the adenovirus protein may bind to the alpha 1 and alpha 2 domains.

Down-regulation of HLA antigens by the adenovirus type 2 E3/19K protein in a T-lymphoma cell line.

Adenoviruses of subgroup C can establish persistent infections in human beings. The exact site of persistence has not been established, but lymphoid tissues are certainly one reservoir. Experimental evidence suggests that early transcription unit 3 (E3) of the virus is involved in this phenomenon. In particular, the most abundant protein of this region, the E3/19K protein, seems to fulfill an important role in viral escape from the immune response. We previously demonstrated that in nonlymphoid cells E3/19K interferes with the antigen presentation function of class I major histocompatibility complex (MHC) antigens by inhibiting their transport to the cell surface. However, the function of the E3 products in lymphoid cells was not investigated. To examine this, the T-lymphoma cell line Jurkat was transfected with a DNA fragment comprising the entire E3 region of adenovirus type 2. We show here that E3/19K is expressed in the absence of the viral transactivator E1A with a rate of biosynthesis similar to that in nonlymphoid 293 cells. Furthermore, inhibition of transport and down-regulation of MHC antigens was comparable in both cell lines. In contrast, various T-cell molecules containing immunoglobulin-like domains showed a normal expression pattern in the transfected cells. A detailed analysis of the interaction between E3/19K and the MHC class I antigens of Jurkat (HLA-A3 and HLA-B35) revealed a differential sensitivity for down-regulation by E3/19K. The data demonstrate that E3/19K exerts its function also in lymphoid cells without affecting other lymphoid cell surface molecules. The implications for persistence of adenovirus in lymphoid cells in vivo are discussed.

Novel proteins associated with MHC class I antigens in cells expressing the adenovirus protein E3/19K.

Assembly of histocompatibility class I antigens (MHC) with beta 2-microglobulin (beta 2m) and peptide takes place in the rough endoplasmic reticulum (ER). At present, it is unclear why peptides generated in the cytosol or ER by proteolysis are not further degraded. Also, it is an open question whether assembly and/or peptide binding is self-instructive or is promoted by additional molecules, for example, chaperones. We previously demonstrated that the adenovirus glycoprotein E3/19K binds to human and some mouse MHC molecules in the ER, interfering with their transport to the cell surface. Here we show that immunoprecipitates from human cells that express transfected E3/19K and murine MHC Kd molecules not only contain MHC heavy chain, beta 2m and E3/19K but also two additional proteins with apparent molecular weights of 100 kDa and 110 kDa. Biochemical characterization of these proteins, designated p100 and p110, demonstrates that they are transmembrane glycoproteins with a similar if not identical protein backbone. Consistent with a role as chaperones, we find that glucose starvation induces complex formation between p100/110 and MHC-E3/19K. Most interestingly, p100 and p110 are displaced from the complex by addition of Kd-specific peptides. Therefore, p100 and p110 might be chaperones that promote correct folding of MHC antigens and/or peptide binding to MHC.

Tumor necrosis factor alpha stimulates expression of adenovirus early region 3 proteins: implications for viral persistence.

Human adenovirus (Ad) can cause persistent infections in humans. Early region 3 (E3) of the virus appears to be implicated in this phenomenon. This transcription unit encodes proteins that interfere in various ways with host cell functions, including (i) cell-surface expression of histocompatibility class I antigens (HLA), (ii) cell-surface expression of the epidermal growth factor receptor (EGF-R), and (iii) the biological activity of tumor necrosis factor alpha (TNF-alpha). We transfected the human cell line 293 with the entire E3 region of Ad2 and investigated the influence of the cytokines TNF-alpha and interferon gamma (IFN-gamma) on cell-surface expression of HLA class I and the EGF-R. Whereas IFN-gamma treatment induced expression of HLA to some extent but not that of the EGF-R, TNF-alpha treatment augmented the reduction of these cell-surface molecules. Subsequent studies on the mechanism of this effect showed a TNF-alpha-dependent upregulation of E3 protein (E3/19K) and mRNA. The significance of this phenomenon was confirmed in infection experiments. A dramatic increase in the amount of E3/19K, even after short induction with low doses of TNF-alpha could be demonstrated. The study provides evidence for an interaction between the immune system and Ad in which the virus takes advantage of an immune mediator to escape immunosurveillance of the host.

E3/19K from adenovirus 2 is an immunosubversive protein that binds to a structural motif regulating the intracellular transport of major histocompatibility complex class I proteins.

We have previously expressed in transgenic mice a chimeric H-2Kd/Kk protein called C31, which contains the extracellular alpha 1 domain of Kd, whereas the rest of the molecule is of Kk origin. This molecule functions as a restriction element for alloreactive and influenza A-specific cytotoxic T lymphocytes (CTL) but is only weakly expressed at the cell surface of splenocytes. Here, we show that the low cell surface expression is the result of slow intracellular transport and processing of the C31 protein. A set of hybrid molecules between Kd and Kk were used to localize the regions in major histocompatibility complex (MHC) molecules that are important for their intracellular transport and to further localize the structures responsible for binding to the adenovirus 2 E3/19K protein. This protein appears to be an important mediator of adenovirus persistence. It acts by binding to the immaturely glycosylated forms of MHC class I proteins in the endoplasmic reticulum (ER), preventing their passage to the cell surface and thereby reducing the recognition of infected cells by virus-specific T cells. We find the surprising result that intracellular transport and E3/19K binding are controlled primarily by the first half of the second domain of Kd, thus localizing these phenomena to the five polymorphic residues in this region of the Kd protein. This result implies that the E3/19K protein may act by inhibiting peptide binding or by disrupting the oligomerization of MHC class I molecules required for transport out of the ER. Alternatively, the E3/19K protein may inhibit the function of a positively acting transport molecule necessary for cell surface expression of MHC class I molecules.

A role for clonal inactivation in T cell tolerance to Mls-1a.

Clonal deletion plays a major part in the maintenance of natural self-tolerance in both normal and transgenic mice. Self antigens that are expressed in the thymus result in the physical elimination of autoreactive thymocytes at a particular stage in their development. For example, the majority V beta 6- and V beta 8.1-bearing T cells that recognize the minor lymphocyte-stimulating antigen, Mls-1a (ref. 10) , are clonally deleted in the thymuses of normal mice and transgenic mice expressing Mls-1a (refs 2, 3, 9). In contrast, a very different mechanism of tolerance involving the functional inactivation, but not elimination, of autoreactive cells, termed clonal inactivation or clonal anergy, has been implicated in some experimentally manipulated systems of tolerance. To test further the mechanisms involved in self-tolerance, we have generated transgenic mice expressing a V beta 8.1 beta chain on greater than 95% of peripheral T cells and have tested tolerance to Mls-1a in these mice. Surprisingly, a significant fraction of the CD4+ peripheral cells that survived deletion were non-responsive in vitro to any stimulus tested. Naturally occurring tolerance to a self antigen expressed in the thymus can thus be mediated by clonal anergy, as well as by clonal deletion.