MINIREVIEW How Viruses Escape from Cytotoxic T Lymphocytes

VIROLOGY 234, 179–185 (1997) ARTICLE NO. VY978674

MINIREVIEW

How Viruses Escape from Cytotoxic T Lymphocytes: Molecular Parameters and Players

Michael B. A. Oldstone

Division of Virology, Department of Neuropharmacology, The Scripps Research Institute, 10550 N. Torrey Pines Road, La Jolla, California 92037 Received May 27, 1997; accepted June 9, 1997

Viruses that persist in infected hosts must evolve successful strategies to avoid recognition by the immune system. The primary player in antiviral immune surveillance is the CD8/ cytotoxic T lymphocyte (CTL), and the battle drawn between the CTLs and viruses is the focus of this review. In this struggle, viruses can follow multiple distinct pathways. For example, DNA viruses often adopt the strategy of encoding proteins that interfere with the immune response along routes of antigen presentation. Such interference prevents the viral peptide from binding to the major histocompatibility complex (MHC) class I glycoprotein; therefore, no virus–MHC complex forms for recognition by antiviral CTLs. RNA viruses, having fewer genes, generate swarms of quasispecies that can contain mutated viral proteins. When such mutants occur in viral peptides presented to the MHC protein or the residue recognized by the CTL receptor, CTL recognition and activation fail. If, instead, the mutation occurs in the viral peptide flanking sequence, the infected cell may not process the viral peptide from the cytosol to the endoplasmic reticulum. Viruses can also directly or indirectly attack dendritic cells and CD4/ or CD8/ T lymphocytes, other routes that interfere with immune functions. Dendritic cells are the primary professional antigen-presenting cells and are critical for the activation of CTL responses. CD4/ T lymphocytes provide help for long-term CD8/ CTL activity and are necessary for its maintenance. Consequently, interference with either dendritic or CD4/ cell types constitutes yet another way that viruses can disable the immune response and persistently infect their host. ᭧ 1997 Academic Press

The specific immune response generated by a host play of viral structural proteins on cell surfaces occurs against infecting viruses can be segregated into two dis- late in the infectious cycle, and large amounts of antibody tinct pathways: humoral (antibody) and cellular (T lym- (ú5 1 106 molecules) and complement are required to phocytes). Both routes play important roles in protection lyse infected cells (Sissons et al., 1979). In the absence from or prevention of viral infections, although clinical of complement, antibodies can bind to viral proteins on and basic research results point to the T lymphocyte as the cell’s surface and alter viral transcription inside the the major player in most viral infections. With the excep- cell (Fujinami and Oldstone, 1979; Levine and Griffin, tion of several enteroviral infections, most viral infections 1992; Levine et al., 1991). In contrast, T cells need recog- are controlled even by humans born with genetic defi- nize only a portion (peptide) of viral protein expressed ciencies in antibody production and with little or no im- on the cell surface complex with the host major histo- munoglobulin (Good, 1991; Whitton and Oldstone, 1996). compatibility (MHC) molecule (reviewed in Whitton and Children with Bruton’s sex-linked a-gammaglobulinemia Oldstone, 1996). Viral proteins (peptides) recognized in- not only recover in the absence of detectable antibody clude those made immediately after infection. Recent evi- but resist subsequent exposure to most viruses, indicat- dence indicates that a single peptide complexed to the ing that they also have and retain immune memory. How- appropriate MHC glycoprotein on a target cell can elicit ever, in those with impaired T cell responses, such as a CTL response (Sykulev et al., 1996). The fact that T individuals with DiGeorge’s syndrome (congenital cells offer defense against a virus during the intracellular athymic aplasia), acquired immunodeficiency syndrome phase of its life cycle, coupled with these cells’ extraordi- (AIDS), or leukemia, or those receiving immunosuppres- nary sensitivity and ability to recognize immediate-early sive therapy, the frequency and severity of viral infections nonstructural components of the virus, which allows re- are enhanced. In most instances, there is also some moval of the factories that make new viruses prior to impairment of antibody responses, yet adoptive transfer assembly, serves to make T cells a most formidable op- of hyperimmune immunoglobulin can moderate, but often ponent of viruses. In addition to lysing virally infected is unable to clear, the viral infection. Antibodies, most cells, CTLs also inhibit viral gene expression by a noncy- often through their hypervariable regions, recognize viral totoxic mechanism involving release of soluble factors antigens as intact proteins that are circulating in the such as interferon-g and tumor necrosis factor (TNF) blood and other fluids or that are attached to cell mem- (Guidotti et al., 1994). branes (Whitton and Oldstone, 1996). However, the dis- Class I MHC-restricted CTLs play a pivotal role in

0042-6822/97 $25.00 179 Copyright ᭧ 1997 by Academic Press All rights of reproduction in any form reserved.

AID VY 8674 / 6a3e$$$381 07-17-97 10:33:42 vira AP: VY 180 MICHAEL B. A. OLDSTONE aborting and preventing both acute and persisting viral infections. The presentation of viral peptide (antigens) by MHC class I molecules utilizes primarily a cytosolic pathway, and MHC class I molecules are found on nearly all nucleated cells of the body except neurons (Joly et al., 1991). The absence of MHC molecules in neurons is one of the cardinal reasons why such cells, when infected by viruses, are not lysed by CTL. The selective advantage for the host is that these essential, nonre- placeable cells are not destroyed, but viruses also enjoy the advantage of dwelling in a cell where they avoid immunologic surveillance. Hence, it is not surprising that many DNA and RNA viruses reside in and persistently infect neurons. The intracellular processing of viral peptides so they can combine with MHC molecules occurs in the endoplasmic reticulum (ER) and is a critical step in the antigen-presenting pathway (Figs. 1 and 2). Neurons FIG. 1. Antigen presentation by the MHC class I and class II path- escape antigen processing by failing to transcribe the ways (adapted from Whitton and Oldstone, 1996). For the class I pathway, the thick line represents a viral protein newly synthesized in the heavy chain of the MHC molecule (Joly et al., 1991) and cell’s cytoplasm. The protein is degraded to shorter peptides in the not transcribing proteins that are important in trans- proteasome. A change in amino acid residues flanking the CTL epitope porting viral peptides to the ER (Joly and Oldstone, 1992). (peptide) sequence can abort cleavage at this step. Peptides are trans- In cells other than neurons, a basic and active strategy located into the ER by the TAP transport mechanism. Shown schemati- used by viruses to avoid immune recognition, which cally are two peptides (l and ᭺) that are both transported to the ER. One of these (᭺) is selected on the basis of its binding affinity by the would destroy the cell they wish to take over for produc- MHC molecule, followed by assembly of the MHC class I heavy chain tion of their own progeny, is to neutralize or abort host peptide and MHC class I light chain (b2M). The MHC–peptide complex molecules important in the antigen presentation path- passes into the Golgi and trans-Golgi to reach the cell’s surface, where way. Alternatively, viruses can mutate their own amino it is available to circulating CD8/ CTLs. For the class II pathway, the acids required for binding to MHC molecules in the ER viral protein outside the cell is taken up by endocytosis and internalized to reach an acid vesicle where degradation to peptide fragments takes or to T cell receptors (TCR). Figure 1 depicts antigen place. Meanwhile, the class II molecule is synthesized within the ER processing via both class I and class II pathways. and transported through the Golgi network where the class II molecule- Within the cell’s cytosol, newly synthesized viral pro- containing vesicle fuses with the endosome containing the fragmented teins are degraded into peptides by the ubiquitin-pro- protein. The class II molecule does not bind to peptides transported teasome system (Rechsteiner et al., 1993; Lowe et al., into the ER because another protein, the invariant chain (not shown), occupies the binding site. Later, in the endosome, the invariant chain 1995; Driscoll and Goldberg, 1990). Structurally, the is degraded or lost, allowing the MHC class II molecule to bind, again proteasome particle consists of a 20S catalytic core of by selective affinity, to the peptide. Finally, the MHC–peptide complex four donut-shaped heptameric stacks of homologous is transported to the cell’s surface where it is sought by CD4/ T cells. subunits. Inside the core’s cavity are catalytic sites. Additional subunits assemble with the 20S core to form a 26S complex that degrades intact proteins in an ATP- CTL epitopes have been found within signal se- dependent manner. The protein’s peptide fragments quences (Hudrisier et al., 1997). then pass into the ER through the action of transport- Two fundamentally different tacks can be employed associated proteins (TAP) (Townsend and Trowsdale, by viruses to interfere with antigen presentation. First, 1993). Figure 1 shows at least two viral peptides (a the DNA viruses associated with persistent infections black circle and a white circle) that have been trans- are likely to encode gene products that interfere with ported into the ER. One of these (the white circle) is any of the multiple steps along the antigen-presenting selected on the basis of binding affinity to the a and pathway (reviewed Fruh et al., 1997). DNA or RNA vi- b chains of the class I molecule’s heavy chain. The ruses that generate greater numbers of CD4/ than conformation of the complex formed by this binding is CD8/ CTLs likely signal a defect in antigen presenta- altered by its interaction with the MHC light chain (b2 tion because the MHC class I pathway evolved to han- microglobulin), allowing the completed complex to dle cytosolic infections. The MHC class II pathway and pass into and through the Golgi to the cell surface. CD4/ T cells are initiated through an extracellular Peptides degraded from whole proteins can also reach route (Fig. 1). CD4/ CTLs were found several years ago the MHC class I heavy chain, but in a TAP-independent in humans with herpes simplex virus (HSV) infection manner, when they originate from signal sequences (Yasukawa and Zarling, 1984), although in most other and are translated on ribosomes bound to the rough viral infections studied, including HSV infection of mice ER (Henderson et al., 1992). Although not common, (Vasilakos and Michael, 1993; Niemialtowski et al.,

AID VY 8674 / 6a3e$$$382 07-17-97 10:33:42 vira AP: VY VIRAL ESCAPE FROM CYTOTOXIC T LYMPHOCYTES 181

1994), and in a recent study of HSV infection of humans (Posavad et al., 1996), CD8/ CTLs are generated. The reason for the generation of CD4/ anti-HSV CTLs was unclear until others showed that HSV ICP-47 efficiently blocked antigen presentation in human cells but did so poorly in murine cells (York et al., 1994; Fruh et al., 1995; Hill et al., 1995; Ahn et al., 1996a). The HSV protein ICP-47 inhibits binding of its peptides to TAP, a molecule necessary for translocating the peptide from the cytosol to the ER. Currently, three proteins of human cytomegalovirus (HCMV) that inhibit MHC class I antigen presentation have been identified. One is the US11 protein, which exports the MHC heavy chain from the ER into the cytosol (Jones et al., 1995; Wiertz et al., 1996); another is US3, which retains MHC molecules in the ER, preventing them from traveling to the plasma membrane (Ahn et al., 1996b). A third protein, US6, inhibits peptide translocation by TAP (Ahn et al., 1997). The m152 protein of murine cytomegalovirus (MCMV) prevents MHC trafficking out of the ER (Tahle et al., 1995), whereas MCMV p34 binds to MHC class I molecules on the plasma membrane and prevents their interaction with the TCR (Hill et al., 1996). Recently Slater FIG. 2. Schematic profile of the effect of a single amino acid mutation and Campbell (1997) noted that the MCMV-induced (1, ᭡) in the sequence (l) flanking the 9-amino-acid peptide (᭺) that defect in MHC class I synthesis and transport varied is processed to bind to the MHC class I molecule. The MHC-bound in a cell-specific manner, perhaps dependent upon the peptide is presented at the cell’s surface for recognition by the T cell / state of differentiation or transcriptional activity of the receptor (TCR) of the CD8 CTL. In row 1, the polypeptide is degraded appropriately in the proteasome and translocated by TAP molecules infected cell. The adenovirus E3/19K glycoprotein was to the ER, where it binds tightly to the MHC molecule. The peptide– the first viral component shown to inhibit MHC class I MHC complex then binds correctly to the TCR (motif A). In row 2, a antigen presentation (Burgert and Kvist, 1985; Anders- mutation in the flanking residue prevents recognition and processing son et al., 1985). This viral complex resides in the ER, by the proteasome. In rows 3 and 4, mutations in the CTL epitope and the 6 amino acid residues terminating the 15- peptide occur either at the residue facing the TCR (1) or at the anchor residue (᭡) that binds to the MHC. In the first instance (motif B), al- membered cytoplasmic tail of its gp19 component are though binding of the viral CTL epitope to the MHC and transport of necessary and sufficient for retention in the ER (Nils- the complex to the cell’s surface are not impeded, binding to the TCR son et al., 1989; Jackson et al., 1990). When these last does not occur. In the latter instance (motif C), the viral CTL epitope 6 amino acids (DEKKMP) are transplanted onto the does not bind within the MHC heavy chain groove, b2microglobulin cytoplasmic tails of other membrane proteins that nor- does not lock into place, and the MHC–peptide complex does not form in the ER. This figure was modified from the Journal of Experimental mally travel out of the ER, their ability to leave the ER Medicine, 1994, Vol. 180, pp. 779–782, by copyright permission of the is eliminated. This inhibition of MHC class I antigen Rockefeller University Press. presentation by the adenovirus E3 complex has been confirmed both in vitro and in vivo (Efrat et al., 1995; von Herrath et al., 1997). However, to date, none of the the immunodominant peptide would influence its pro- HSV, HCMV, or MCMV proteins has been shown to cessing and block presentation to MHC molecules in interfere with MHC class I antigen presentation in vivo, the ER (Eisenlohr et al., 1992). Accordingly, effective so their biological significance has not yet been estab- processing of the Friend/Moloney/Rauscher (FMR) lished. There is no reason why a specific gene prod- peptide RSPWFTTL encoded by MV-p15E led to the uct(s) of RNA viruses cannot also interfere with antigen defective processing of the FMR peptide in tumor cells, processing (Howcroft et al., 1993), and some do. while the KSPWFTTL peptide encoded by AKV-MCF Second, RNA viruses have attracted attention be- was processed correctly (Ossendorp et al., 1996). Sub- cause of their potentially high frequency of mutation sequent experiments showed that proteasome-medi- in the viral peptides that interfere with processing from ated digestion due to the R r K change led to degrada- the cytosol to the ER and because of their ability to bind tion of the peptide and its failure to present TAP for to MHC class I molecules or to TCR (Fig. 2). Recently, a translocation into the ER. single amino acid change in a viral peptide was shown Both in vitro evidence and in vivo evidence indicate to influence antigen processing. It had been antici- that viruses can escape recognition of virus-specific CTL pated that such a mutation in a viral sequence flanking by generating mutants in the viral peptide that binds to

AID VY 8674 / 6a3e$$$382 07-17-97 10:33:42 vira AP: VY 182 MICHAEL B. A. OLDSTONE

MHC molecules or is recognized by the TCR (Aebischer recent serial observations of CTL escape variants in et al., 1991; Lewicki et al., 1995a; Koup, 1994; Pircher et HIV infection (Borrow et al., 1997). That is, within 15 al., 1990; Lewicki et al., 1995b; Moskophidis and Zinker- days after initiation of a primary HIV infection, a single nagel, 1995; Borrow et al., 1997; Goulder et al., 1997; immunodominant epitope in gp160, amino acids 29– Harrer et al., 1996; Klenerman et al., 1994; Bertoletti et 39, engaged in a monospecific MHC class I restricted al., 1994; Couillin et al., 1995). The MHC-bound peptide CTL response that rapidly eliminated the infecting vi- sequence is linear, resulting from proteolytic fragmenta- ruses. At that time, no CTL responses to the other HIV tion of viral protein synthesized within the cell. Several proteins occurred, so the kinetics for generation of a rules concerning the structure of these peptides and their CTL escape variant under CTL pressure in vivo could binding to MHC as well as the peptide–MHC complex be accurately measured. By 30 to 72 days after onset binding the TCR are now established. Studies of endoge- of the acute retroviral symptoms, and under CTL pres- nously processed viral peptides indicate that they vary sure in which approximately 1 of every 20 CD8/ CTL in length from 8 to 11 amino acids and display MHC lymphocytes in the blood was directed against that allele-specific motifs. Mutational and crystallographic single gp160 aa 29–39 epitope, this HIV peptide mu- studies of MHC molecules complexed with viral peptide tated, allowing the virus to escape CTL recognition. show that the molecule’s flexible conformations allow The kinetics of CTL escape was comparable to that these 8- to 11-amino-acid peptides to bind within the observed for patients receiving anti-retroviral drug MHC groove once their anchoring residues are fixed therapy and clearly indicated for the first time the bio- (Bjorkman et al., 1987; Garcia et al., 1996). Analysis of logical significance of generating such CTL responses residues flanking the anchoring residue(s) indicates the in vivo in HIV infection for the control of viral load critical importance of minor pockets of MHC-binding (Borrow et al., 1997). In the HIV-infected patient, a new clefts in peptide selectivity, leading to the concept that CTL epitope located near the COO0 end of gp160 and these structural factors are likely responsible for the pref- CTL responses to GAG, POL, and NEF now appeared erential selection of specific peptides so often observed for the first time (Borrow et al., 1997). The newly gener- in interactions with MHC molecules (Garcia et al., 1996; ated CTL responses continued to limit the viral load, Hudrisier et al., 1996; reviewed in Whitton and Oldstone, which dropped from 200,000–300,000 to approxi- 1996). mately 1000 RNA molecules/ml. Similar results oc- Studies with lymphocytic choriomeningitis virus curred in experiments with LCMV (Lewicki et al., (LCMV) have documented CTL escape variants with mu- 1995b) in that following mutation additional CTL epi- tations in all the known immunodominant epitopes se- topes were generated and the viral load was con- lected under CTL pressure (Lewicki et al., 1995a). By this trolled. means, viral mutants are generated that escape recogni- Goulder et al. (1997) also noted CTL escape variants tion by CTL because of a single amino acid mutation in two HIV-infected patients studied during the late in any one of the known three immunodominant CTL stages of disease. These and other mutations in the CTL epitopes. These mutations did not occur at the anchor epitope may also inhibit CTL activation by an ‘‘antagonist residue but at residues that interact with receptors for activity.’’ Described in both HIV and hepatitis B virus in- the CTL. However, as long as one of these three immuno- fections (Klenerman et al., 1994; Bertoletti et al., 1994), dominant epitopes does not mutate, viruses are cleared such mutated sequences can bind and interact with the efficiently and effectively with the usual kinetics just as TCR (in contrast to the B motif shown in Fig. 2) but are though all three were present (Lewicki et al., 1995b). unable to deliver a full stimulatory signal so that anergy Importantly, when all three immunodominant epitopes occurs. The molecular basis of this failed response is mutate in the same virion, the host can still mount nonim- not clear. munodominant lower affinity CTLs against the virus and If CTL escape variants are unlikely to lead to viral still control infection. However, in this case, more time persistence, except when a uniquely restricted MHC is required to purge virus infection (Lewicki et al., 1995b; allele(s) is present (de Campos-Lima et al., 1993) or Moskophidis and Zinkernagel, 1995). Thus, under CTL when only a monotypic CTL response is made (Pircher pressure, escape variants can arise, but the host has et al., 1990; Borrow et al., 1997), what accounts for the other options including a hierarchical control of CTL epi- failure of CTL to control the infection? Since expres- topes that function even when immunodominant epi- sion of a single CTL epitope is apparently sufficient topes mutate. The inference drawn is that, unless a par- for control of viral infection (Lewicki et al., 1995b), the ticular CTL response is restricted to one epitope (Pircher existence of a virus that evades immune surveillance et al., 1990), even mutations in CTL epitopes (viral pep- because a few or several of its CTL epitopes have tide) identifiable by biochemical and in vitro assays are mutated is unlikely. Although one to three immuno- probably not of major importance in persistent viral infec- dominant CTL epitopes have been found for a single tions. viral protein, the numbers of viral proteins encoded in This hypothesis receives further reinforcement from the genome, the presence of two major MHC HLA-

AID VY 8674 / 6a3e$$$382 07-17-97 10:33:42 vira AP: VY VIRAL ESCAPE FROM CYTOTOXIC T LYMPHOCYTES 183 restricting haplotypes per individual, and the hierar- effectively escape the host’s immune surveillance. De- chical control and subsequent release of nonimmuno- fining the parameters and manipulating the players dominant CTL epitopes when the immunodominant involved promise to be important strategies in the epitopes have mutated all indicate that it would likely treatment of persistent viral infections. be rare for a virus to totally escape CTL recognition by this means. Yet, viruses escape CTL immune sur- ACKNOWLEDGMENTS veillance. How does this occur? There are at least This is Publication 10722-NP from the Department of Neuropharma- three main mechanisms. First, viruses that infect the cology, The Scripps Research Institute, La Jolla, California. This work thymus of a newborn negatively select and thereby was supported in part by USPHS Grants AI09484 and AG04342. remove T lymphocytes that ordinarily would recognize that virus (reviewed in Borrow and Oldstone, 1997; REFERENCES Jamieson and Ahmed, 1988). When this occurs, few if any low-affinity T lymphocytes would be available for Aebischer, T., Moskophidis, D., Rohrer, U. H., Zinkernagel, R. H., and Hengartner, R. (1991). In vitro selection of lymphocytic choriomeningi- positive selection and generation of effector CTL. Sec- tis virus escape mutants by cytotoxic T lymphocytes. Proc. Natl. Acad. ond, viruses that infect or interfere with dendritic cells Sci. USA 88, 11047–11051. (Borrow et al., 1995), the professional antigen-present- Ahn, K., Meyer, T. H., Uebel, S., Sempe, P., Djaballah, H., Yang, Y., ing cells of the body, could prevent the generation of Peterson, P. A., Fruh, K., and Tampe, R. (1996a). Molecular mecha- CTLs so none could be activated to seek virus-infected nism and species specificity of TAP inhibition by herpes simplex virus ICP47. EMBO J. 15, 3247–3255. cells throughout the body. Since dendritic cells are Ahn, K., Angulo, A., Ghazal, P., Peterson, P. A., Yang, Y., and Fruh, K. repopulated from precursors of macrophages in the (1996b). Human cytomegalovirus inhibits antigen presentation by a bone marrow, any virus that also infects these cells sequential multistep process. Proc. Natl. Acad. Sci. USA 93, 10990– in the bone marrow could prevent the differentiation 10994. of dendritic cells to replace those destroyed by the Ahn, K., Gruhler, A., Galocha, B., Jones, T. R., Wiertz, E. J. H. J., Ploegh, H. L., Peterson, P. A., Yang, Y., and Fruh, K. (1997). The ER-luminal virus or the antiviral immune response. Third, a virus domain of the HCMV glycoprotein US6 inhibits peptide translocation / could lower or remove the supply of CD4 T lympho- by TAP. Immunity. [in press]. cytes. Results from experimental models clearly dem- Andersson, M., Paabo, S., Nilsson, T., and Peterson, P. A. (1985). Im- onstrate that to maintain effective CTLs in vivo for paired intracellular transport of class I MHC antigens as a possible more than a few weeks, helper CD4/ cells are re- means for adenoviruses to evade immune surveillance. Cell 43, 215– 222. quired (Matloubian et al., 1994; Battegay et al., 1994; Battegay, M., Moskophidis, D., Rahemtulla, A., Hengartner, H., Mak, Tishon et al., 1995; Thomsen et al., 1996). When such T. W., and Zinkernagel, R. M. (1994). Enhanced establishment of a cells are not present, CD8/ CTLs lose their antiviral virus carrier state in adult CD4/ T-cell-deficient mice. J. Virol. 68, protective function in vivo. Although all the factors 4700. made by CD4/ T cells necessary to maintain CD8 ac- Bertoletti, A., Sette, A., Chisari, F. V., Penna, A., Levrero, M., De Carli, M., Fiaccadori, F., and Ferrari, C. (1994). Natural variants of cytotoxic tivity are not known, one has been identified: inter- epitopes are T-cell receptor antagonists for antiviral cytotoxic T cells. feron-g (Tishon et al., 1995). Recently, Thomsen et Nature 369, 407–410. al. (1996) noted that removal of B lymphocytes also Bjorkman, P. J., Saper, M. A., Samraoui, B., Bennett, W. S., Strominger, interfered with maintaining clearance of virus and al- J. L., and Wiley, D. C. (1987). Structure of the human class I histocom- lowed virus to persist. However, it is not clear whether patibility antigen HLA-A2. Nature 329, 506–512. Borrow, P., Lewicki, H., Wei, X., Horwitz, M. S., Peffer, N., Meyers, H., the deficiency noted when B lymphocytes were re- Nelson, J. A., Gairin, J. E., Hahn, B. H., Oldstone, M. B. A., and Shaw, moved was due to a role for antibody, the interaction G. M. (1997). Antiviral pressure exerted by HIV-1-specific cytotoxic T of B cells with CD4/ T helper cells, or a failure in lymphocytes (CTLs) during primary infection demonstrated by rapid antigen presentation by B cells. In summary, all these selection of CTL escape virus. Nature Med. 3, 205–211. options, i.e., infection or disruption of dendritic cell Borrow, P., and Oldstone, M. B. A. (1997). Lymphocytic choriomeningitis virus. In ‘‘Viral Pathogenesis’’ (N. Nathanson et al., Eds.), pp. 593– function as well as infection of T lymphocytes with 627. Lippincott–Raven, Philadelphia. either early-life abrogation of their potential function Borrow, P., Evans, C. F., and Oldstone, M. B. A. (1995). Virus-induced or removal of CD4/ lymphocytes as helpers for the immunosuppression: Immune system-mediated destruction of virus- antiviral CD8/ CTL response, allow viruses to persist. infected dendritic cells results in generalized immune suppression. Such a loss of CD4/ T cell help has been postulated J. Virol. 69, 1059–1070. Burgert, H.-G., and Kvist, S. (1985). An adenovirus type 2 glycoprotein to occur in HIV infection (Meyaard et al., 1994; Schnitt- blocks cell surface expression of human histocompatibility class I man et al., 1990). These options, coupled with the ef- antigens. Cell 41, 987–997. fect of viruses on antigen processing, mutation, and Burrows, J. M., Burrows, S. R., Poulsen, L. M., Sculley, T. B., Moss, D. J., selection for CTL escape variants, and surely other and Khanna, R. (1996). Unusually high frequency of Epstein–Barr pathways yet to be uncovered, all point to a multifac- virus genetic variants in Papua New Guinea that can escape cytotoxic T-cell recognition: Implications for virus evolution. J. Virol. 70, toral and complex chess game played between the 2490–2496. host and the virus for a winning immune response Couillin, I., Connan, F., Culmann-Penciolelli, B., Gomard, E., Guillet, that overcomes infection or a winning virus that can J. G., and Choppin, J. (1995). HLA-dependent variants in human immu-

AID VY 8674 / 6a3e$$$383 07-17-97 10:33:42 vira AP: VY 184 MICHAEL B. A. OLDSTONE

nodeficiency virus NEF protein alter peptide/HLA binding. Eur. J. elements at peptide non-anchor residues. J. Biol. Chem. 271, 17829– Immunol. 25, 728–732. 17836. de Campos-Lima, P. O., Gavioli, R., Zhang, Q. J., Wallace, L. E., Dolcetti, Jackson, M. R., Nilsson, T., and Peterson, P. A. (1990). Identification of R., Rowe, M., Rickinson, A. B., and Masucci, M. G. (1993). HLA-A11 a consensus motif for retention of transmembrane proteins in the epitope loss isolates of Epstein–Barr virus from a highly A11/ popu- endoplasmic reticulum. EMBO J. 9, 3153–3162. lation. Science 260, 98–100. Jamieson, B. D., and Ahmed, R. (1988). T-cell tolerance: Exposure to Driscoll, J., and Goldberg, A. L. (1990). The proteasome (multicatalytic virus in utero does not cause a permanent deletion of specific T protease) is a component of the 1500-kDa proteolytic complex which cells. Proc. Natl. Acad. Sci. USA 85, 2265–2268. degrades ubiquitin-conjugated proteins. J. Biol. Chem. 265, 4789– Joly, E., and Oldstone, M. B. A. (1992). Neuronal cells are deficient in 4792. loading peptides onto MHC class I molecules. Nature 8, 1185. Efrat, S., Fejer, G., Brownlee, M., and Horwitz, M. (1995). Prolonged Joly, E., Mucke, L., and Oldstone, M. B. A. (1991). Viral persistence in survival of pancreatic islet allografts mediated by adenovirus immu- neurons explained by lack of major histocompatibility complex class noregulatory transgenes. Proc. Natl. Acad. Sci. USA 92, 6947– I expression. Science 253, 1283–1285. 6951. Jones, T. R., Hanson, L. K., Sun, L., Slater, J. S., Stenberg, R. M., and Eisenlohr, L. C., Yewdell, J. W., and Bennink, J. R. (1992). Flanking se- Campbell, A. E. (1995). Multiple independent loci within the human quences influence the presentation of an endogenously synthesized cytomegalovirus unique short region down-regulate expression of peptide to cytotoxic T lymphocytes. J. Exp. Med. 175, 481–487. major histocompatibility complex class I heavy chains. J. Virol. 69, Fruh, K., Ahn, K., and Peterson, P. A. (1997). Inhibition of MHC class I 4830–4841. antigen presentation by viral proteins. J. Mol. Med. [in press]. Klenerman, P., Rowland-Jones, S., McAdam, S., Edwards, J., Daenke, Fruh, K., Ahn, K., Djaballah, H., Sempe, P., van Endert, P. M., Tampe, S., Lalloo, D., Koppe, B., Rosenberg, W., Boyd, D., Edwards, A., Gian- R., Peterson, P. A., and Yang, Y. (1995). A viral inhibitor of peptide grande, P., Phillips, R. E., and McMichael, A. J. (1994). Cytotoxic T- transporters for antigen presentation. Nature 375, 415–418. cell activity antagonized by naturally occurring HIV-1 gag variants. Fujinami, R. S., and Oldstone, M. B. A. (1979). Antiviral antibody reacting Nature 369, 403–407. on the plasma membrane alters measles virus expression inside the Koup, R. A. (1994). Virus escape from CTL recognition. J. Exp. Med. cell. Nature 279, 529–530. 180, 779–782. Garcia, K. C., Degano, M., Stanfield, R. L., Brunmark, A., Jackson, M. R., Levine, B., Hardwick, J. M., Trapp, B. D., Crawford, T. O., Bollinger, R. C., Peterson, P. A., Teyton, L., and Wilson, I. A. (1996). An ab T cell and Griffin, D. E. (1991). Antibody-mediated clearance of alphavirus ˚ receptor structure at 2.5 A and its orientation in the TCR–MHC com- infection from neurons. Science 254, 856–860. plex. Science 274, 209–219. Levine, B., and Griffin, D. E. (1992). Persistence of viral RNA in mouse Good, R. A. (1991). Experiments of nature in the development of modern brains after recovery from acute alphavirus encephalitis. J. Virol. 66, immunology. Immunol. Today 12, 283–286. 6429–6435. Goulder, P. J. R., Phillips, R. E., Colbert, R. A., McAdam, S., Ogg, G., Lewicki, H., Tishon, A., Borrow, P., Evans, C. F., Gairin, J. E., Hahn, K. M., Nowak, M. A., Giangrande, P., Luzzi, G., Morgan, B., Edwards, A., Jewell, D. A., Wilson, I. A., and Oldstone, M. B. A. (1995a). CTL escape McMichael, A. J., and Rowland-Jones, S. (1997). Late escape from an viral variants. I. Generation and molecular characterization. Virology immunodominant cytotoxic T-lymphocyte response associated with 210, 29–40. progression to AIDS. Nature Med. 3, 312–317. Lewicki, H. A., von Herrath, M. G., Evans, C. F., Whitton, J. L., and Old- Guidotti, L. G., Ando, K., Hobbs, M. V., Ishikawa, T., Runkel, L., Schreiber, stone, M. B. A. (1995b). CTL escape viral variants. II. Biologic activity R. D., and Chisari, F. V. (1994). Cytotoxic T lymphocytes inhibit hepati- in vivo. Virology 211, 433–450. tis B virus gene expression by a noncytolytic mechanism in transgenic mice. Proc. Natl. Acad. Sci. USA 91, 3764–3768. Lowe, J., Stock, D., Jap, B., Zwickl, P., Baumeister, W., and Huber, R. Harrer, T., Harrer, E., Kalams, S. A., Barbosa, P., Trocha, A., Johnson, (1995). Crystal structure of the 20 S proteasome from the archaeon ˚ R. P., Elbeik, T., Feinberg, M. B., Buchbinder, S. P., and Walker, B. D. T. acidophilum at 3.4 A resolution. Science 268, 533–539. / (1996). Cytotoxic T lymphocytes in asymptomatic long-term nonpro- Matloubian, M., Concepcion, R. J., and Ahmed, R. (1994). CD4 T cells / gressing HIV-1 infection. Breadth and specificity of the response and are required to sustain CD8 cytotoxic T-cell responses during relation to in vivo viral quasispecies in a person with prolonged chronic viral infection. J. Virol. 68, 8056–8063. infection and low viral load. J. Immunol. 156, 2616–2623. Meyaard, L., Otto, S. A., Hooibrink, B., and Miedema, F. (1994). Quantita- / Henderson, R. A., Michel, H., Sakaguchi, K., et al. (1992). HLA2.1-associ- tive analysis of CD4 T cell function in the course of human immuno- ated peptides from a mutant cell line: A second pathway of antigen deficiency virus infection. J. Clin. Invest. 94, 1947. presentation. Science 255, 1264–1266. Moskophidis, D., and Zinkernagel, R. M. (1995). Immunobiology of cyto- Hill, A., Jugovic, P., York, I., Rus, G., Bennink, J., Yewdell, J., Ploegh, H., toxic T-cell escape mutants of lymphocytic choriomeningitis virus. J. and Johnson, D. (1995). Herpes simplex virus turns off the TAP to Virol. 69, 2187–2193. evade host immunity. Nature 375, 411–415. Niemialtowski, M. G., Godfrey, V. L., and Rouse, B. T. (1994). Quantita- / / Hill, A., Kleijnen, M., Schust, D., Koszinowski, U., and Ploegh, H. (1996). tive studies on CD4 and CD8 cytotoxic T lymphocyte responses Murine cytomegalovirus (MCMV) encodes a 34 kDa glycoprotein against herpes simplex virus type 1 in normal and b2M-deficient which binds to class I MHC molecules at the surface of infected mice. Immunobiology 190, 183–194. cells. In ‘‘Cell Biology of Virus Entry, Replication and Pathogenesis,’’ Nilsson, T., Jackson, M., and Peterson, P. A. (1989). Short cytoplasmic [Keystone Symposium, Santa Fe, NM, Abstract 410]. sequences serve as retention signals for transmembrane proteins Howcroft, T., Strebel, K., Martin, M., and Singer, D. (1993). Repression in the endoplasmic reticulum. Cell 58, 707–718. of MHC class I gene promoter activity by two-exon TAT of HIV. Ossendorp, F., Eggers, M., Neisig, A., Ruppert, T., Groettrup, M., Sijts, Science 260, 1320. A., Mengede, E., Kloetzel, P.-M., Neefjes, J., Koszinowski, U., and Hudrisier, D., Oldstone, M. B. A., and Gairin, J. E. (1997). The signal Melief, C. (1996). A single residue exchange within a viral CTL epi- sequence of lymphocytic choriomeningitis virus contains an immuno- tope alters proteasome-mediated degradation resulting in lack of dominant cytotoxic T cell epitope that is restricted by both H-2Db antigen presentation. Immunity 5, 115–124. and H-2Kb molecules. Virology. [in press]. Pircher, H., Moskophidis, D., Rohrer, U., Burki, K., Hengartner, H., and Hudrisier, D., Mazarguil, H., Laval, F., Oldstone, M. B. A., and Gairin, Zinkernagel, R. M. (1990). Viral escape by selection of cytotoxic T J. E. (1996). Binding of viral antigens to major histocompatibility com- cell-resistant virus variants in vivo. Nature 346, 629–633. plex class I H-2Db molecules is controlled by dominant negative Posavad, C. M., Koelle, D. M., and Corey, L. (1996). High frequency of

AID VY 8674 / 6a3e$$$383 07-17-97 10:33:42 vira AP: VY VIRAL ESCAPE FROM CYTOTOXIC T LYMPHOCYTES 185

CD8/ cytotoxic T-lymphocyte precursors specific for herpes simplex cytic choriomeningitis virus-infected MHC class II-deficient mice and viruses in persons with genital herpes. J. Virol. 70, 8165–8168. B cell-deficient mice. J. Immunol. 157, 3074. Rechsteiner, M., Hoffman, L., and Dubiel, W. (1993). The multicatalytic Tishon, A., Lewicki, H., Rall, G., von Herrath, M., and Oldstone, M. B. A. and 26 S proteases. J. Biol. Chem. 268, 6065–6068. (1995). An essential role for type 1 interferon-g in terminating persis- Schnittman, S. M., Lane, H. C., Greenhouse, J., Justement, J. S., Baseler, tent viral infection. Virology 212, 244–250. M., and Fauci, A. S. (1990). Preferential infection of CD4/ memory T Townsend, A., and Trowsdale, J. (1993). The transporters associated cells by human immunodeficiency virus type 1: Evidence for a role with antigen presentation. Semin. Cell Biol. 4, 53–61. in the selective T-cell functional defects observed in infected individ- Vasilakos, J. P., and Michael, J. G. (1993). Herpes simplex virus class uals. Proc. Natl. Acad. Sci. USA 87, 6058. I-restricted peptide induces cytotoxic T lymphocytes in vivo indepen- / Sissons, J. G. P., Schreiber, R. D., Perrin, L. H., Cooper, N. R., Muller- dent of CD4 T cells. J. Immunol. 150, 2346–2355. Eberhard, H. J., and Oldstone, M. B. A. (1979). Lysis of measles virus- von Herrath, M. G., Efrat, S., Oldstone, M. B. A., and Horwitz, M. (1997). infected cells by the purified cytolytic alternative complement path- Expression of adenoviral E3 transgenes in b-cells prevents autoim- mune diabetes. Proc. Natl. Acad. Sci. USA, in press. way and antibody. J. Exp. Med. 150, 445–454. Whitton, J. L., and Oldstone, M. B. A. (1996). Immune response to vi- Slater, J. S., and Campbell, A. E. (1997). Down-regulation of MHC class ruses. In ‘‘Fields Virology’’ (B. N. Fields et al., Eds.), pp. 345–374. I synthesis by murine cytomegalovirus occurs in immortalized but Lippincott-Raven, Philadelphia. not primary fibroblasts. Virology 229, 221–227. Wiertz, E. J. H. J., Jones, T. R., Sun, L., Bogyo, M., Geuze, H. J., and Sykulev, Y., Joo, M., Vturina, I., Tsomides, T. J., and Eisen, H. N. (1996). Ploegh, H. L. (1996). The human cytomegalovirus US11 gene product Evidence that a single peptide–MHC complex on a target cell can dislocates MHC class I heavy chains from the ER to the cytosol. Cell elicit a cytolytic T cell response. Immunity 4, 565–571. 84, 769–779. Tahle, R., Szepan, U., Hengel, H., Geginat, G., Lucin, P., and Koszinow- Yasukawa, M., and Zarling, J. M. (1984). Human cytotoxic T cell clones ski, U. H. (1995). Identification of the mouse cytomegalovirus geno- directed against herpes simplex virus infected cells: Lysis restricted mic region affecting major histocompatibility complex class I mole- by HLA-11 MB and DR antigens. J. Immunol. 133, 422–427. cule transport. J. Virol. 69, 6098–6105. York, I. A., Roop, C., Andrews, D. W., Riddell, S. R., Graham, F. L., and Thomsen, A. R., Johansen, J., Marker, O., and Christensen, J. P. (1996). Johnson, D. C. (1994). A cytosolic herpes simplex protein inhibits Exhaustion of CTL memory and recrudescence of viremia in lympho- antigen presentation to CD8/ T lymphocytes. Cell 77, 525–535.

AID VY 8674 / 6a3e$$$383 07-17-97 10:33:42 vira AP: VY