- Email: [email protected]
Human cytomegalovirus: Host immune modulation by the viral US3 gene
The International Journal of Biochemistry & Cell Biology 41 (2009) 503–506
Contents lists available at ScienceDirect
The International Journal of Biochemistry & Cell Biology journal homepage: www.elsevier.com/locate/biocel
Pathogens in focus
Human cytomegalovirus: Host immune modulation by the viral US3 gene Ziqi Liu a , Michael Winkler b , Bonita Biegalke c,∗ a
Program in Molecular and Cellular Biology, Ohio University, 228 Irvine Hall, Athens, OH 45701, United States Institute for Virology, University Clinic Ulm, Albert-Einstein-Allee 11, 89081 Ulm, Germany c Ohio University College of Osteopathic Medicine, Department of Biomedical Sciences and Program in Molecular and Cellular Biology, Athens, OH 45701, United States b
a r t i c l e
i n f o
Article history: Received 13 April 2007 Received in revised form 24 September 2008 Accepted 10 October 2008 Available online 18 October 2008 Keywords: Cytomegalovirus Immune evasion US3 MHC class I heavy chains Tapasin
a b s t r a c t Human cytomegalovirus (HCMV) is a common infection, opportunistically causing disease in people with immune system deficits. HCMV expresses several proteins that contribute to avoidance of the host immune response. The US3 gene is one of the first immune evasion genes expressed following infection. Expression of the US3 gene is highly regulated, with the gene encoding autoregulatory proteins. The largest of the US3 proteins, a 22 kDa resident endoplasmic reticulum protein, binds to MHC class I heavy chain complexes and components of the peptide loading complex, delaying the maturation of the MHC class I complexes and presentation of viral antigen on the surface of infected cells. A smaller US3 protein, a 17 kDa US3 protein, competes with the 22 kDa for protein interactions, counteracting, in part, the effects of the larger protein. The US3 amino acid sequence is highly conserved among clinical isolates and laboratory strains, suggesting an important role for this gene in natural infections in the human host. © 2008 Elsevier Ltd. All rights reserved.
1. Introduction Human cytomegalovirus (HCMV) infects 50–80% of adults world-wide (Britt and Alford, 1996). HCMV is acquired by exposure to body fluids containing infectious virus. The incidence of infection increases with age, with infection typically acquired either during early childhood or at the onset of sexual activity. Initial infection and replication is followed by the establishment of a lifelong asymptomatic infection. Associated disease is rarely seen in healthy, immunocompetent adults despite the frequency of infection. Rather, HCMV is an opportunistic pathogen, causing clinically significant disease in immune-immature or immune-compromised individuals. Immunosuppressed individuals such as transplant recipients, immunocompromised individuals including people with HIV infection, and immuno-immature individuals or neonates are all prone to develop HCMV-associated disease (Britt and Alford, 1996). HCMV is the most common viral infection in transplant recipients, where infection is associated with a variety of clinical presentations, including a viral syndrome (fever, malaise, thrombocytopenia), invasive infection of a particular tissue (e.g. pneumonitis, retinitis, or gastroenteritis) and an increase in graft rejection as well as graft-versus-host disease (American Society of
∗ Corresponding author. Tel.: +1 740 593 2377. E-mail address: [email protected] (B. Biegalke). 1357-2725/$ – see front matter © 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.biocel.2008.10.012
Transplant Surgeons, 2004). In people with HIV infection, HCMVassociated disease is one of the correlates defining the development of AIDS, although the frequency of HCMV-associated disease in HIV-infected individuals has declined with the advent of HAART therapy (Griffiths, 2006). In addition to causing clinical disease in immuno-incompetent adults, neonates can also develop clinically significant disease following HCMV infection. Approximately 1% of newborn infants are infected with HCMV; of the infected infants ∼5–10% present initially with symptomatic disease including hepatosplenomegaly, thrombocytopenic purpura, and microcephaly. Of the remaining infected infants, approximately 10–20% will develop later sequelae, most commonly sensorineuronal deafness (Boppana et al., 2005). HCMV-associated disease in immuno-incompetent populations results from the disruption of the balance between the host and the virus, with the outcome of viral replication and the associated development of symptoms influenced not only by viral factors, but also by the host immune response. 2. Overview of viral pathogenesis The genome of HCMV is large (∼235 kb) and encodes approximately 200 genes (Murphy and Shenk, 2008). Of the viral genes, only ∼40–50 (20–25%) are essential for lytic viral replication in cell culture, suggesting that the remainder of the viral genes are involved in aspects of viral replication in the host. Included in the non-essential viral genes are genes that encode chemokines or
504
Z. Liu et al. / The International Journal of Biochemistry & Cell Biology 41 (2009) 503–506
chemokine receptors, anti-apoptotic genes, genes that contribute to the infection of particular cell types, and genes that modulate host immune-cell recognition of virally infected cells (Mocarski, 2002). In addition to the contributions of the nonessential viral genes to pathogenicity, viral genes essential for replication alter normal cell processes, such as regulation of the cell cycle, to insure the generation of progeny viral particles (Castillo and Kowalik, 2004). Viral-specific cytotoxic T cells are generated in response to infection and play a major role in limiting viral replication. Viral antigens are presented on the surface of infected cells as a tripartite complex of major histocompatibility complex (MHC) class I heavy chains, 2 microglobulin and antigenic peptide. The generation of the tripartite complex occurs in the endoplasmic reticulum (ER) where MHC class I heavy chains associate with 2 microglobulin. The antigenic peptide is then added to the complex with the assistance of ER chaperones and the transporter for antigen presentation (TAP) (Heemels and Ploegh, 1995). HCMV encodes several proteins that modulate or alter the host immune response to viral replication (Powers et al., 2008). In particular, one gene cluster, the US2–US11 gene cluster, encodes several immunomodulatory proteins. The US2–US11 genes appear to have arisen through gene duplication. The proteins encoded by the US2 and US11 genes bind to MHC class I complexes, resulting in transport of the complexes from the ER to the cytosol where they are degraded. The US6 protein blocks the TAP, which then inhibits the maturation of the MHC class I complex. The normal trafficking of the MHC class I complexes from the ER is delayed by the US10 protein. US3 proteins retain MHC class I complexes in the ER, also delaying their maturation (Ahn et al., 1996; Jones et al., 1996). US2 and US3 proteins also interfere with MHC class II antigen presentation (Powers et al., 2008). Natural killer cells (NK) are activated by and destroy cells that lack sufficient amounts of MHC class I complexes on the cell surface. In addition to encoding proteins that interfere with cytotoxic T cells, HCMV also encodes several proteins that modulate the NK cell response to virally infected cells (Powers et al., 2008).
3. US3 expression Expression of the US3 gene results in retention of MHC class I heavy chains in the endoplasmic reticulum, interfering with the presentation of antigen on the surface of US3-expressing cells. Transcriptional regulatory elements controlling US3 expression include silencer and enhancer elements as well as a transcriptional repressive element (Weston, 1988; Biegalke, 1995; Thrower et al., 1996; Chan et al., 1996). A complex network of proteins interacts with the regulatory elements to control US3 transcription. The US3 gene is an immediate early gene, with transcripts detectable by one hour post-infection (Weston, 1988; Liu et al., 2002). Following infection, abundant levels of US3 transcripts are seen early in infection, with transcripts peaking at 3 h post-infection, followed by a dramatic decline in transcript levels by 5 h post-infection. The enhancer element is presumed to be important for the initial burst of transcriptional activity, while the binding of the early viral UL34 gene product to the transcriptional repressive element (crs, Fig. 1A) is responsible for the decline in US3 expression at 5 h post-infection (LaPierre and Biegalke, 2001; Biegalke et al., 2004). The primary US3 transcript is subject to alternative splicing, generating three predominant transcripts; a full length unspliced RNA encoding a 22 kDa protein, a singly spliced RNA encoding a 17 kDa protein and a doubly spliced RNA predicted to encode a 3.5 kDa protein (Fig. 1A; Tenney et al., 1993; Liu et al., 2002). Temporally, the unspliced RNA appears earliest following infection, followed by the singly and doubly spliced transcripts (Liu et al., 2002).
Fig. 1. (A) Diagram of the US3 gene and transcripts. The regulatory region is indicated: S, silencer; E, enhancer; T, TATA box; crs, cis-respressive signal; ORF, open reading frame. Diagrams of the three major spliced transcripts are depicted with shared amino acids indicated by the black rectangles. The sizes of each of the proteins encoded by the transcripts are listed. US, unspliced; SS, singly spliced; DS, doubly spliced. (B) Diagram of the 22 kDa US3 open reading frame. The signal peptide, luminal and transmembrane domains, and the cytoplasmic tail are indicated. The -strands in the luminal domain are indicated by the black rectangles; the strands associate in the following manner: a–b, c–c , d–e, e–f and strands a and g are held together with a disulfide bond linking the two conserved C residues indicated by the arrows. *Proline substitution found in some clinical isolates. The numbers indicate the amino acids comprising the protein domains.
4. US3 function The 22 kDa US3 protein encoded by the unspliced transcript is an resident ER protein comprised of a short signal peptide, a ER luminal domain, and a 20 amino acid transmembrane-spanning region followed by a short carboxyl-terminal tail (Fig. 1B; Tenney et al., 1993; Ahn et al., 1996; Jones et al., 1996). The structure of the luminal domain of the 22 kDa US3 protein was determined by NMR analysis and compared to the structure of the related US2 protein (Misaghi et al., 2004). Similar to the US2 protein, the US3 protein contains an Ig-like fold comprised of seven -strands, with a conserved disulfide bond linking strands G and A (Fig. 1B; see Misaghi et al., 2004, for a model of the luminal domain of US3). The 22 kDa US3 protein oligomerizes in vitro and in vitro, and binds to MHC class I molecules in vivo as an oligomer. The ER luminal portion of the 22 kDa US3 protein was defined as containing the intracellular localization signal (Lee et al., 2000) when assayed by sensitivity to endoglycosidase H digestion or by immunofluorescent antibody staining. However, expression of 22 kDa US3 truncations as EGFP-fusion proteins suggests that the carboxyl-terminal tail and/or the transmembrane region contribute to ER localization (Zhao and Biegalke, 2003). The 22 kDa US3 protein physically interacts with MHC class I heavy chains (Ahn et al., 1996; Jones et al., 1996) and is sufficient to retain MHC heavy chains in the ER (Zhao and Biegalke, 2003). Deletion of the carboxyl terminal tail from the 22 kDa US3 protein reduces interactions with MHC class I heavy chains while deletion of the transmembrane region eliminates the interaction (Zhao and Biegalke, 2003; Gruhler et al., 2002). Similarly, mutation of the cysteine at position 44 (Fig. 1B, first arrow) which links -strands a and g also abolishes heavy chain–US3 interactions. In addition to interacting with MHC class I heavy chains, the 22 kDa US3 protein binds to TAP, tapasin, and protein disulfide isomerase (PDI), members of the MHC class I peptide loading complex (Park et al., 2004, 2006). Tapasin is a MHC class I-dedicated chaperone involved in loading peptide onto the MHC class I complex in an allele-specific manner (Fig. 2B). PDI functions to regulate the oxidation state of MHC class I molecules, facilitating the selection of
Z. Liu et al. / The International Journal of Biochemistry & Cell Biology 41 (2009) 503–506
505
Fig. 2. (A) Effect of US3 proteins on the processing of MHC class I heavy chains. Parental cells (U373), cells constitutively expressing the US3 gene from a clinical isolate (Oe), or an isolate (TB40E) with a glutamic acid to lysine substitution at position 63 (E63 to K63 ) were pulsed (P) with radioactive amino acids and or chased for 1 h with non-radioactive amino acids (C). MHC class I complexes were immunoprecipitated using antibody W6/32 and either left untreated or treated with endoglycosidase H (PE or CE). Proteins resistant (R) or sensitive (S) to endoglycosidase H are indicated. Diagram of MHC class I complex assembly. The newly synthesized heavy chain is inserted into the ER upon translation, and then (1) with the assistance of chaperones, associates with 2-microglobulin. The TAP complex, including tapasin and protein disulfide isomerase, processes and loads antigenic peptides onto the MHC class I complex (2). The 22 kDa US3 protein interferes with the peptide loading process (indicated by the dotted circle), preventing the normal processing and maturation of the oligosaccharides occurs as the MHC class I complex travels from the ER (endoglycosidase H sensitive) to the Golgi (endoglycosidase H resistant; 3). The complex is then transported to the cell surface, resulting in the presentation of the antigen in the context of the MHC class I complex on the cytoplasmic membrane (4).
optimal peptides for MHC class I complex formation. The interaction of the 22 kDa US3 protein and tapasin interferes with peptide loading and MHC complex assembly, resulting in a delay in maturation. Expression of the 22 kDa US3 protein results in a decrease in the amount of PDI present in cells, and an increase in the levels of reduced class I molecules. Thus, the 22 kDa US3 protein inhibits peptide loading onto MHC class I molecules by interacting with both tapasin and PDI. Carboxyl-terminal tail and transmembrane deletions of the 22 kDa US3 protein have a decreased ability to prevent the maturation of MHC class I heavy chains, suggesting that the transmembrane region is required to inhibit peptide loading (Liu et al., 2002; Park et al., 2004). The singly-spliced US3 transcript encodes a 17 kDa protein that lacks the transmembrane-spanning region but has 134 amino acids in common with the full length US3 protein (Fig. 1A; Tenney et al., 1993). Despite having 134 amino acids in common with the 22 kDa protein, the 17 kDa US3 protein, when expressed as a EGFPfusion protein, localizes to the ER, to cytoplasmic vesicles and to the Golgi, indicative of processing through the secretory pathway. The 17 kDa US3 protein does not bind to MHC class I heavy chains, as demonstrated by the absence of coprecipitation (Zhao and Biegalke, 2003). Cells expressing the 17 kDa US3 protein process MHC class I heavy chains as efficiently as the control cells, demonstrating that the 17 kDa protein does not affect the processing of the MHC class I complex (Liu et al., 2002).
In contrast to cells expressing individual US3 proteins, in cells expressing both the 22 and the 17-kDa US3 proteins, more of the MHC class I heavy chains mature normally and reach the cell surface (Liu et al., 2002; Shin et al., 2006). The variation seen in levels of MHC class I processing results from the expression of the 17 kDa US3 protein. The 17 kDa US3 protein, like the 22 kDa protein, also binds to tapasin, with the two proteins competing for tapasin. The 22 kDa US3 protein blocks the interaction of tapasin with TAP; the tapasin-TAP interaction is restored by the addition of the 17 kDa protein. In the presence of the 17 kDa protein, tapasin interactions with the 22 kDa protein are decreased. The competition between the 17 kDa and the 22 kDa US3 proteins for tapasin binding suggest that the US3 gene encodes an autoregulator of its immune evasion function. The larger US3 protein is synthesized first during infection, following by synthesis of the smaller US3 proteins. Multiple mechanisms, including autoregulation, are used by the virus to ensure that the US3 gene functions are precisely regulated. Biochemical evidence for the generation of the predicted 3.5 kDa protein encoded by the doubly-spliced transcript has not yet been obtained partly because of the size of the predicted protein. Unlike the two larger US3 proteins discussed above, the 3.5 kDa US3 protein, when expressed as an EGFPfusion protein localized exclusively to the Golgi (Zhao and Biegalke, 2003).
506
Z. Liu et al. / The International Journal of Biochemistry & Cell Biology 41 (2009) 503–506
5. Clinical correlations The contribution of the US3 gene to immune-evasion has led to the sequence analysis of the US3 genes from clinical isolates as well as a variety of laboratory strains. While there is sequence variation in the DNA encoding the US3 proteins, an analysis of the amino acid sequence data for the US3 open reading frame reveals a relatively small number of amino acid changes, most of which are conservative. Two of the amino acid changes seen in some clinical isolates are not conservative and are worthy of mention. The first amino acid variation is at position 63 (located in the loop between -strands b and c; Fig. 1B); the consensus US3 amino acid sequence has a glutamic acid residue in this position. Mutation of Glu63 prevented retention of US3 in the ER (Lee et al., 2003); however, in ∼10% of the US3 genes sequenced, Glu63 has been replaced by a lysine residue. Analysis of an isolate with the Glu to Lys63 substitution did not alter the ability of the US3 gene to retain MHC class I complexes in the ER (Fig. 2A). The other interesting amino acid substitution is at position 107, where a proline is inserted in place of either a serine or a threonine in ∼20% of the US3 genes sequenced (see asterisk (*), Fig. 1B). This substitution occurs in the loop between -strands E and F; the impact of the proline on the structure and function of the US3 proteins is unknown. We have analyzed the ability of US3 genes from clinical isolates to inhibit MHC class I heavy chain processing. Cell lines expressing the US3 genes from either laboratory strains or from clinical isolates were established; the resulting cell lines transcribed RNAs corresponding to both the unspliced and singly spliced US3 transcripts (data not shown). The effect of the US3 proteins on MHC class I molecule processing was examined by pulsing cells with radiolabeled amino acids, followed by a chase period. MHC class I complexes were immunoprecipitated using the monoclonal antibody W6/32; the relative amount of mature MHC class I complexes was determined by subjecting the proteins to endoglycosidase H treatment. As seen in Fig. 2A, expression of US3 proteins resulted in a marked increase in the amount of immature MHC class I complexes (endoglycosidase H sensitive), either following the initial labeling period of following the chase period (compare parental U373 cells with the TB40E and Oe isolates). In conclusion, the US3 proteins encoded by clinical isolates alter antigen presentation through the MHC class I complexes, similar to the laboratory stains of virus. The US3 gene is conserved in laboratory and clinical isolates; homologs of US3 are present in rhesus and chimpanzee cytomegaloviruses (Pande et al., 2005); and a functional homolog is present in murine cytomegalovirus. The conservation and the large number of immune modulatory genes present in the cytomegaloviruses demonstrates the significance of these viral genes for establishment of infection in the host. However, in spite of the utilization of multiple immune evasion tactics, immunecompetent individuals mount a robust immune response to the virus, generating cytotoxic T cells to several immunodominant epitopes and to minor epitopes, including the US3 gene. The elicitation of a substantial immune response to the virus following infection suggests that a vaccine able to elicit the generation of a cytotoxic T response is feasible. Deletion of the US2–US11 genes from the viral genome may contribute to the development of a safe, protective vaccine. HCMV is an intriguing virus; its existence as an opportunistic pathogen is useful in understanding the interplay between viral proteins and normal cellular functions. The US3 gene is an excellent example of the levels of complexity used by the virus to initially undergo lytic replication when possible, while establishing and maintaining a latent infection in the host, thereby ensuring a reser-
voir of virus capable of reactivating and replicating when conditions permit. Acknowledgements Research in B. Biegalke’s laboratory is supported by the American Heart Association and by the National Institutes of Health. References Ahn KS, Angulo A, Ghazal P, Peterson PA, Yang Y, Früh K. Human cytomegalovirus inhibits antigen presentation by a sequential multistep process. Proc Natl Acad Sci USA 1996;93:10990–5. American Society of Transplant Surgeons. Cytomegalovirus. Am J Transplant 2004;4:51–8. Biegalke BJ. Regulation of human cytomegalovirus US3 gene transcription by a cisrepressive sequence. J Virol 1995;69:5362–7. Biegalke BJ, Lester E, Rana R. Characterization of the human cytomegalovirus UL34 gene. J Virol 2004;78:9579–83. Boppana SB, Fowler KB, Pass RF, Rivera LB, Bradford RD, Lakeman FD, et al. Congenital cytomegaloviurs infection: association between virus burden in infancy and hearing loss. J Pediatr 2005;146:817–23. Britt WJ, Alford CA. Cytomegalovirus. In: Fields BN, Knipe DM, Howley PM, editors. Fields Virology. Philadelphia: Lippincott-Raven; 1996. p. 2493–524. Castillo JP, Kowalik TF. HCMV infection: modulating the cell cycle and cell death. Int Rev Immunol 2004;23:113–39. Chan YJ, Tseng WP, Hayward GS. Two distinct upstream regulatory domains containing multicopy cellular transcription factor binding sites provide basal repression and inducible enhancer characteristics to the immediate-early IES (US3) promoter from human cytomegalovirus. J Virol 1996;70:5312–28. Griffiths PD. CMV as a cofactor enhancing progression of AIDS. J Clin Virol 2006;35:489–92. Gruhler A, Peterson PA, Fruh K. Human cytomegalovirus immediate early glycoprotein US3 retains MHC class I molecules by transient association. Traffic 2002;1:318–25. Heemels M-T, Ploegh H. Generation, translocation and presentation of MHC class I-restricted peptides. Annu Rev Biochem 1995;64:463–91. Jones TR, Wiertz EJHJ, Sun L, Fish KN, Nelson JA, Ploegh HL. Human cytomegalovirus US3 impairs transport and maturation of major histocompatibility complex class I heavy chains. Proc Natl Acad Sci USA 1996;93:11327–33. LaPierre L, Biegalke BJ. Identification of a novel transcriptional repressor encoded by human cytomegalovirus. J Virol 2001;75:6062–9. Lee S, Yoon J, Park B, Jun Y, Jin M, Sung H, et al. Structural and functional dissection of human cytomegalovirus US3 in binding major histocompatibility complex class I molecules. J Virol 2000;74:11262–9. Lee S, Park B, Ahn K. Determinant for endoplasmic reticulum retention in the luminal domain of the human cytomegalovirus US3 glycoprotein. J Virol 2003;77:2147–56. Liu W, Zhao Y, Biegalke BJ. Analysis of human cytomegalovirus US3 gene products. Virology 2002;301:32–42. Misaghi S, Sun Z-YJ, Stern P, Gaudet R, Wagner G, Ploegh H. Structural and functional analysis of human cytomegalovirus US3 protein. J Virol 2004;78:413–23. Mocarski ES. Immunomodulation by cytomegaloviruses: manipulative strategies beyond evasion. Trends Microbiol 2002;10:332–9. Murphy E, Shenk T. Human cytomegalovirus genome. Curr Topics Microbiol Immun 2008;325:2–19. Pande NT, Powers C, Ahn K, Früh K. Rhesus cytomegalovirus contains functional homologues of US2, US3, US6, and US11. J Virol 2005;79:5786–98. Park B, Kim Y, Shin J, Lee S, Cho K, Früh K, et al. Human cytomegalovirus inhabits tapasin-dependent peptide loading and optimization of the MHC class I peptide cargo for immune evasion. Immunity 2004;20:71–85. Park B, Sungwook L, Kin E, Cho K, Riddell SR, Cho S, et al. Redox regulation facilitates optimal peptide selection by MHC class I during antigen processing. Cell 2006;127:369–82. Powers C, DeFelippis V, Malouli D, Fruh K. Cytomegalovirus immune evasion. Curr Topics Microbiol Immun 2008;325:333–59. Shin J, Park B, Kim Y, Biegalke BJ, Kang S, Ahn K. A short isoform of human cytomegalovirus US3 functions as a dominant negative inhibitor of the fulllength form. J Virol 2006;80:5397–404. Tenney DJ, Santomenna LD, Goudie KB, Colberg-Poley AM. The human cytomegalovirus US3 immediate-early protein lacking the putative transmembrane domain regulates gene expression. Nucleic Acids Res 1993;21:2931– 7. Thrower AR, Bullock GC, Bissell JE, Stinski MF. Regulation of a human cytomegalovirus immediate-early gene (US3) by a silencer-enhancer combination. J Virol 1996;70:91–100. Weston K. An enhancer element in the short unique region of human cytomegalovirus regulates the production of a group of abundant immediate early transcripts. Virology 1988;162:406–16. Zhao Y, Biegalke BJ. Functional analysis of the human cytomegalovirus immune evasion protein, pUS322 kDa. Virology 2003;315:353–61.
Contents lists available at ScienceDirect
The International Journal of Biochemistry & Cell Biology journal homepage: www.elsevier.com/locate/biocel
Pathogens in focus
Human cytomegalovirus: Host immune modulation by the viral US3 gene Ziqi Liu a , Michael Winkler b , Bonita Biegalke c,∗ a
Program in Molecular and Cellular Biology, Ohio University, 228 Irvine Hall, Athens, OH 45701, United States Institute for Virology, University Clinic Ulm, Albert-Einstein-Allee 11, 89081 Ulm, Germany c Ohio University College of Osteopathic Medicine, Department of Biomedical Sciences and Program in Molecular and Cellular Biology, Athens, OH 45701, United States b
a r t i c l e
i n f o
Article history: Received 13 April 2007 Received in revised form 24 September 2008 Accepted 10 October 2008 Available online 18 October 2008 Keywords: Cytomegalovirus Immune evasion US3 MHC class I heavy chains Tapasin
a b s t r a c t Human cytomegalovirus (HCMV) is a common infection, opportunistically causing disease in people with immune system deficits. HCMV expresses several proteins that contribute to avoidance of the host immune response. The US3 gene is one of the first immune evasion genes expressed following infection. Expression of the US3 gene is highly regulated, with the gene encoding autoregulatory proteins. The largest of the US3 proteins, a 22 kDa resident endoplasmic reticulum protein, binds to MHC class I heavy chain complexes and components of the peptide loading complex, delaying the maturation of the MHC class I complexes and presentation of viral antigen on the surface of infected cells. A smaller US3 protein, a 17 kDa US3 protein, competes with the 22 kDa for protein interactions, counteracting, in part, the effects of the larger protein. The US3 amino acid sequence is highly conserved among clinical isolates and laboratory strains, suggesting an important role for this gene in natural infections in the human host. © 2008 Elsevier Ltd. All rights reserved.
1. Introduction Human cytomegalovirus (HCMV) infects 50–80% of adults world-wide (Britt and Alford, 1996). HCMV is acquired by exposure to body fluids containing infectious virus. The incidence of infection increases with age, with infection typically acquired either during early childhood or at the onset of sexual activity. Initial infection and replication is followed by the establishment of a lifelong asymptomatic infection. Associated disease is rarely seen in healthy, immunocompetent adults despite the frequency of infection. Rather, HCMV is an opportunistic pathogen, causing clinically significant disease in immune-immature or immune-compromised individuals. Immunosuppressed individuals such as transplant recipients, immunocompromised individuals including people with HIV infection, and immuno-immature individuals or neonates are all prone to develop HCMV-associated disease (Britt and Alford, 1996). HCMV is the most common viral infection in transplant recipients, where infection is associated with a variety of clinical presentations, including a viral syndrome (fever, malaise, thrombocytopenia), invasive infection of a particular tissue (e.g. pneumonitis, retinitis, or gastroenteritis) and an increase in graft rejection as well as graft-versus-host disease (American Society of
∗ Corresponding author. Tel.: +1 740 593 2377. E-mail address: [email protected] (B. Biegalke). 1357-2725/$ – see front matter © 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.biocel.2008.10.012
Transplant Surgeons, 2004). In people with HIV infection, HCMVassociated disease is one of the correlates defining the development of AIDS, although the frequency of HCMV-associated disease in HIV-infected individuals has declined with the advent of HAART therapy (Griffiths, 2006). In addition to causing clinical disease in immuno-incompetent adults, neonates can also develop clinically significant disease following HCMV infection. Approximately 1% of newborn infants are infected with HCMV; of the infected infants ∼5–10% present initially with symptomatic disease including hepatosplenomegaly, thrombocytopenic purpura, and microcephaly. Of the remaining infected infants, approximately 10–20% will develop later sequelae, most commonly sensorineuronal deafness (Boppana et al., 2005). HCMV-associated disease in immuno-incompetent populations results from the disruption of the balance between the host and the virus, with the outcome of viral replication and the associated development of symptoms influenced not only by viral factors, but also by the host immune response. 2. Overview of viral pathogenesis The genome of HCMV is large (∼235 kb) and encodes approximately 200 genes (Murphy and Shenk, 2008). Of the viral genes, only ∼40–50 (20–25%) are essential for lytic viral replication in cell culture, suggesting that the remainder of the viral genes are involved in aspects of viral replication in the host. Included in the non-essential viral genes are genes that encode chemokines or
504
Z. Liu et al. / The International Journal of Biochemistry & Cell Biology 41 (2009) 503–506
chemokine receptors, anti-apoptotic genes, genes that contribute to the infection of particular cell types, and genes that modulate host immune-cell recognition of virally infected cells (Mocarski, 2002). In addition to the contributions of the nonessential viral genes to pathogenicity, viral genes essential for replication alter normal cell processes, such as regulation of the cell cycle, to insure the generation of progeny viral particles (Castillo and Kowalik, 2004). Viral-specific cytotoxic T cells are generated in response to infection and play a major role in limiting viral replication. Viral antigens are presented on the surface of infected cells as a tripartite complex of major histocompatibility complex (MHC) class I heavy chains, 2 microglobulin and antigenic peptide. The generation of the tripartite complex occurs in the endoplasmic reticulum (ER) where MHC class I heavy chains associate with 2 microglobulin. The antigenic peptide is then added to the complex with the assistance of ER chaperones and the transporter for antigen presentation (TAP) (Heemels and Ploegh, 1995). HCMV encodes several proteins that modulate or alter the host immune response to viral replication (Powers et al., 2008). In particular, one gene cluster, the US2–US11 gene cluster, encodes several immunomodulatory proteins. The US2–US11 genes appear to have arisen through gene duplication. The proteins encoded by the US2 and US11 genes bind to MHC class I complexes, resulting in transport of the complexes from the ER to the cytosol where they are degraded. The US6 protein blocks the TAP, which then inhibits the maturation of the MHC class I complex. The normal trafficking of the MHC class I complexes from the ER is delayed by the US10 protein. US3 proteins retain MHC class I complexes in the ER, also delaying their maturation (Ahn et al., 1996; Jones et al., 1996). US2 and US3 proteins also interfere with MHC class II antigen presentation (Powers et al., 2008). Natural killer cells (NK) are activated by and destroy cells that lack sufficient amounts of MHC class I complexes on the cell surface. In addition to encoding proteins that interfere with cytotoxic T cells, HCMV also encodes several proteins that modulate the NK cell response to virally infected cells (Powers et al., 2008).
3. US3 expression Expression of the US3 gene results in retention of MHC class I heavy chains in the endoplasmic reticulum, interfering with the presentation of antigen on the surface of US3-expressing cells. Transcriptional regulatory elements controlling US3 expression include silencer and enhancer elements as well as a transcriptional repressive element (Weston, 1988; Biegalke, 1995; Thrower et al., 1996; Chan et al., 1996). A complex network of proteins interacts with the regulatory elements to control US3 transcription. The US3 gene is an immediate early gene, with transcripts detectable by one hour post-infection (Weston, 1988; Liu et al., 2002). Following infection, abundant levels of US3 transcripts are seen early in infection, with transcripts peaking at 3 h post-infection, followed by a dramatic decline in transcript levels by 5 h post-infection. The enhancer element is presumed to be important for the initial burst of transcriptional activity, while the binding of the early viral UL34 gene product to the transcriptional repressive element (crs, Fig. 1A) is responsible for the decline in US3 expression at 5 h post-infection (LaPierre and Biegalke, 2001; Biegalke et al., 2004). The primary US3 transcript is subject to alternative splicing, generating three predominant transcripts; a full length unspliced RNA encoding a 22 kDa protein, a singly spliced RNA encoding a 17 kDa protein and a doubly spliced RNA predicted to encode a 3.5 kDa protein (Fig. 1A; Tenney et al., 1993; Liu et al., 2002). Temporally, the unspliced RNA appears earliest following infection, followed by the singly and doubly spliced transcripts (Liu et al., 2002).
Fig. 1. (A) Diagram of the US3 gene and transcripts. The regulatory region is indicated: S, silencer; E, enhancer; T, TATA box; crs, cis-respressive signal; ORF, open reading frame. Diagrams of the three major spliced transcripts are depicted with shared amino acids indicated by the black rectangles. The sizes of each of the proteins encoded by the transcripts are listed. US, unspliced; SS, singly spliced; DS, doubly spliced. (B) Diagram of the 22 kDa US3 open reading frame. The signal peptide, luminal and transmembrane domains, and the cytoplasmic tail are indicated. The -strands in the luminal domain are indicated by the black rectangles; the strands associate in the following manner: a–b, c–c , d–e, e–f and strands a and g are held together with a disulfide bond linking the two conserved C residues indicated by the arrows. *Proline substitution found in some clinical isolates. The numbers indicate the amino acids comprising the protein domains.
4. US3 function The 22 kDa US3 protein encoded by the unspliced transcript is an resident ER protein comprised of a short signal peptide, a ER luminal domain, and a 20 amino acid transmembrane-spanning region followed by a short carboxyl-terminal tail (Fig. 1B; Tenney et al., 1993; Ahn et al., 1996; Jones et al., 1996). The structure of the luminal domain of the 22 kDa US3 protein was determined by NMR analysis and compared to the structure of the related US2 protein (Misaghi et al., 2004). Similar to the US2 protein, the US3 protein contains an Ig-like fold comprised of seven -strands, with a conserved disulfide bond linking strands G and A (Fig. 1B; see Misaghi et al., 2004, for a model of the luminal domain of US3). The 22 kDa US3 protein oligomerizes in vitro and in vitro, and binds to MHC class I molecules in vivo as an oligomer. The ER luminal portion of the 22 kDa US3 protein was defined as containing the intracellular localization signal (Lee et al., 2000) when assayed by sensitivity to endoglycosidase H digestion or by immunofluorescent antibody staining. However, expression of 22 kDa US3 truncations as EGFP-fusion proteins suggests that the carboxyl-terminal tail and/or the transmembrane region contribute to ER localization (Zhao and Biegalke, 2003). The 22 kDa US3 protein physically interacts with MHC class I heavy chains (Ahn et al., 1996; Jones et al., 1996) and is sufficient to retain MHC heavy chains in the ER (Zhao and Biegalke, 2003). Deletion of the carboxyl terminal tail from the 22 kDa US3 protein reduces interactions with MHC class I heavy chains while deletion of the transmembrane region eliminates the interaction (Zhao and Biegalke, 2003; Gruhler et al., 2002). Similarly, mutation of the cysteine at position 44 (Fig. 1B, first arrow) which links -strands a and g also abolishes heavy chain–US3 interactions. In addition to interacting with MHC class I heavy chains, the 22 kDa US3 protein binds to TAP, tapasin, and protein disulfide isomerase (PDI), members of the MHC class I peptide loading complex (Park et al., 2004, 2006). Tapasin is a MHC class I-dedicated chaperone involved in loading peptide onto the MHC class I complex in an allele-specific manner (Fig. 2B). PDI functions to regulate the oxidation state of MHC class I molecules, facilitating the selection of
Z. Liu et al. / The International Journal of Biochemistry & Cell Biology 41 (2009) 503–506
505
Fig. 2. (A) Effect of US3 proteins on the processing of MHC class I heavy chains. Parental cells (U373), cells constitutively expressing the US3 gene from a clinical isolate (Oe), or an isolate (TB40E) with a glutamic acid to lysine substitution at position 63 (E63 to K63 ) were pulsed (P) with radioactive amino acids and or chased for 1 h with non-radioactive amino acids (C). MHC class I complexes were immunoprecipitated using antibody W6/32 and either left untreated or treated with endoglycosidase H (PE or CE). Proteins resistant (R) or sensitive (S) to endoglycosidase H are indicated. Diagram of MHC class I complex assembly. The newly synthesized heavy chain is inserted into the ER upon translation, and then (1) with the assistance of chaperones, associates with 2-microglobulin. The TAP complex, including tapasin and protein disulfide isomerase, processes and loads antigenic peptides onto the MHC class I complex (2). The 22 kDa US3 protein interferes with the peptide loading process (indicated by the dotted circle), preventing the normal processing and maturation of the oligosaccharides occurs as the MHC class I complex travels from the ER (endoglycosidase H sensitive) to the Golgi (endoglycosidase H resistant; 3). The complex is then transported to the cell surface, resulting in the presentation of the antigen in the context of the MHC class I complex on the cytoplasmic membrane (4).
optimal peptides for MHC class I complex formation. The interaction of the 22 kDa US3 protein and tapasin interferes with peptide loading and MHC complex assembly, resulting in a delay in maturation. Expression of the 22 kDa US3 protein results in a decrease in the amount of PDI present in cells, and an increase in the levels of reduced class I molecules. Thus, the 22 kDa US3 protein inhibits peptide loading onto MHC class I molecules by interacting with both tapasin and PDI. Carboxyl-terminal tail and transmembrane deletions of the 22 kDa US3 protein have a decreased ability to prevent the maturation of MHC class I heavy chains, suggesting that the transmembrane region is required to inhibit peptide loading (Liu et al., 2002; Park et al., 2004). The singly-spliced US3 transcript encodes a 17 kDa protein that lacks the transmembrane-spanning region but has 134 amino acids in common with the full length US3 protein (Fig. 1A; Tenney et al., 1993). Despite having 134 amino acids in common with the 22 kDa protein, the 17 kDa US3 protein, when expressed as a EGFPfusion protein, localizes to the ER, to cytoplasmic vesicles and to the Golgi, indicative of processing through the secretory pathway. The 17 kDa US3 protein does not bind to MHC class I heavy chains, as demonstrated by the absence of coprecipitation (Zhao and Biegalke, 2003). Cells expressing the 17 kDa US3 protein process MHC class I heavy chains as efficiently as the control cells, demonstrating that the 17 kDa protein does not affect the processing of the MHC class I complex (Liu et al., 2002).
In contrast to cells expressing individual US3 proteins, in cells expressing both the 22 and the 17-kDa US3 proteins, more of the MHC class I heavy chains mature normally and reach the cell surface (Liu et al., 2002; Shin et al., 2006). The variation seen in levels of MHC class I processing results from the expression of the 17 kDa US3 protein. The 17 kDa US3 protein, like the 22 kDa protein, also binds to tapasin, with the two proteins competing for tapasin. The 22 kDa US3 protein blocks the interaction of tapasin with TAP; the tapasin-TAP interaction is restored by the addition of the 17 kDa protein. In the presence of the 17 kDa protein, tapasin interactions with the 22 kDa protein are decreased. The competition between the 17 kDa and the 22 kDa US3 proteins for tapasin binding suggest that the US3 gene encodes an autoregulator of its immune evasion function. The larger US3 protein is synthesized first during infection, following by synthesis of the smaller US3 proteins. Multiple mechanisms, including autoregulation, are used by the virus to ensure that the US3 gene functions are precisely regulated. Biochemical evidence for the generation of the predicted 3.5 kDa protein encoded by the doubly-spliced transcript has not yet been obtained partly because of the size of the predicted protein. Unlike the two larger US3 proteins discussed above, the 3.5 kDa US3 protein, when expressed as an EGFPfusion protein localized exclusively to the Golgi (Zhao and Biegalke, 2003).
506
Z. Liu et al. / The International Journal of Biochemistry & Cell Biology 41 (2009) 503–506
5. Clinical correlations The contribution of the US3 gene to immune-evasion has led to the sequence analysis of the US3 genes from clinical isolates as well as a variety of laboratory strains. While there is sequence variation in the DNA encoding the US3 proteins, an analysis of the amino acid sequence data for the US3 open reading frame reveals a relatively small number of amino acid changes, most of which are conservative. Two of the amino acid changes seen in some clinical isolates are not conservative and are worthy of mention. The first amino acid variation is at position 63 (located in the loop between -strands b and c; Fig. 1B); the consensus US3 amino acid sequence has a glutamic acid residue in this position. Mutation of Glu63 prevented retention of US3 in the ER (Lee et al., 2003); however, in ∼10% of the US3 genes sequenced, Glu63 has been replaced by a lysine residue. Analysis of an isolate with the Glu to Lys63 substitution did not alter the ability of the US3 gene to retain MHC class I complexes in the ER (Fig. 2A). The other interesting amino acid substitution is at position 107, where a proline is inserted in place of either a serine or a threonine in ∼20% of the US3 genes sequenced (see asterisk (*), Fig. 1B). This substitution occurs in the loop between -strands E and F; the impact of the proline on the structure and function of the US3 proteins is unknown. We have analyzed the ability of US3 genes from clinical isolates to inhibit MHC class I heavy chain processing. Cell lines expressing the US3 genes from either laboratory strains or from clinical isolates were established; the resulting cell lines transcribed RNAs corresponding to both the unspliced and singly spliced US3 transcripts (data not shown). The effect of the US3 proteins on MHC class I molecule processing was examined by pulsing cells with radiolabeled amino acids, followed by a chase period. MHC class I complexes were immunoprecipitated using the monoclonal antibody W6/32; the relative amount of mature MHC class I complexes was determined by subjecting the proteins to endoglycosidase H treatment. As seen in Fig. 2A, expression of US3 proteins resulted in a marked increase in the amount of immature MHC class I complexes (endoglycosidase H sensitive), either following the initial labeling period of following the chase period (compare parental U373 cells with the TB40E and Oe isolates). In conclusion, the US3 proteins encoded by clinical isolates alter antigen presentation through the MHC class I complexes, similar to the laboratory stains of virus. The US3 gene is conserved in laboratory and clinical isolates; homologs of US3 are present in rhesus and chimpanzee cytomegaloviruses (Pande et al., 2005); and a functional homolog is present in murine cytomegalovirus. The conservation and the large number of immune modulatory genes present in the cytomegaloviruses demonstrates the significance of these viral genes for establishment of infection in the host. However, in spite of the utilization of multiple immune evasion tactics, immunecompetent individuals mount a robust immune response to the virus, generating cytotoxic T cells to several immunodominant epitopes and to minor epitopes, including the US3 gene. The elicitation of a substantial immune response to the virus following infection suggests that a vaccine able to elicit the generation of a cytotoxic T response is feasible. Deletion of the US2–US11 genes from the viral genome may contribute to the development of a safe, protective vaccine. HCMV is an intriguing virus; its existence as an opportunistic pathogen is useful in understanding the interplay between viral proteins and normal cellular functions. The US3 gene is an excellent example of the levels of complexity used by the virus to initially undergo lytic replication when possible, while establishing and maintaining a latent infection in the host, thereby ensuring a reser-
voir of virus capable of reactivating and replicating when conditions permit. Acknowledgements Research in B. Biegalke’s laboratory is supported by the American Heart Association and by the National Institutes of Health. References Ahn KS, Angulo A, Ghazal P, Peterson PA, Yang Y, Früh K. Human cytomegalovirus inhibits antigen presentation by a sequential multistep process. Proc Natl Acad Sci USA 1996;93:10990–5. American Society of Transplant Surgeons. Cytomegalovirus. Am J Transplant 2004;4:51–8. Biegalke BJ. Regulation of human cytomegalovirus US3 gene transcription by a cisrepressive sequence. J Virol 1995;69:5362–7. Biegalke BJ, Lester E, Rana R. Characterization of the human cytomegalovirus UL34 gene. J Virol 2004;78:9579–83. Boppana SB, Fowler KB, Pass RF, Rivera LB, Bradford RD, Lakeman FD, et al. Congenital cytomegaloviurs infection: association between virus burden in infancy and hearing loss. J Pediatr 2005;146:817–23. Britt WJ, Alford CA. Cytomegalovirus. In: Fields BN, Knipe DM, Howley PM, editors. Fields Virology. Philadelphia: Lippincott-Raven; 1996. p. 2493–524. Castillo JP, Kowalik TF. HCMV infection: modulating the cell cycle and cell death. Int Rev Immunol 2004;23:113–39. Chan YJ, Tseng WP, Hayward GS. Two distinct upstream regulatory domains containing multicopy cellular transcription factor binding sites provide basal repression and inducible enhancer characteristics to the immediate-early IES (US3) promoter from human cytomegalovirus. J Virol 1996;70:5312–28. Griffiths PD. CMV as a cofactor enhancing progression of AIDS. J Clin Virol 2006;35:489–92. Gruhler A, Peterson PA, Fruh K. Human cytomegalovirus immediate early glycoprotein US3 retains MHC class I molecules by transient association. Traffic 2002;1:318–25. Heemels M-T, Ploegh H. Generation, translocation and presentation of MHC class I-restricted peptides. Annu Rev Biochem 1995;64:463–91. Jones TR, Wiertz EJHJ, Sun L, Fish KN, Nelson JA, Ploegh HL. Human cytomegalovirus US3 impairs transport and maturation of major histocompatibility complex class I heavy chains. Proc Natl Acad Sci USA 1996;93:11327–33. LaPierre L, Biegalke BJ. Identification of a novel transcriptional repressor encoded by human cytomegalovirus. J Virol 2001;75:6062–9. Lee S, Yoon J, Park B, Jun Y, Jin M, Sung H, et al. Structural and functional dissection of human cytomegalovirus US3 in binding major histocompatibility complex class I molecules. J Virol 2000;74:11262–9. Lee S, Park B, Ahn K. Determinant for endoplasmic reticulum retention in the luminal domain of the human cytomegalovirus US3 glycoprotein. J Virol 2003;77:2147–56. Liu W, Zhao Y, Biegalke BJ. Analysis of human cytomegalovirus US3 gene products. Virology 2002;301:32–42. Misaghi S, Sun Z-YJ, Stern P, Gaudet R, Wagner G, Ploegh H. Structural and functional analysis of human cytomegalovirus US3 protein. J Virol 2004;78:413–23. Mocarski ES. Immunomodulation by cytomegaloviruses: manipulative strategies beyond evasion. Trends Microbiol 2002;10:332–9. Murphy E, Shenk T. Human cytomegalovirus genome. Curr Topics Microbiol Immun 2008;325:2–19. Pande NT, Powers C, Ahn K, Früh K. Rhesus cytomegalovirus contains functional homologues of US2, US3, US6, and US11. J Virol 2005;79:5786–98. Park B, Kim Y, Shin J, Lee S, Cho K, Früh K, et al. Human cytomegalovirus inhabits tapasin-dependent peptide loading and optimization of the MHC class I peptide cargo for immune evasion. Immunity 2004;20:71–85. Park B, Sungwook L, Kin E, Cho K, Riddell SR, Cho S, et al. Redox regulation facilitates optimal peptide selection by MHC class I during antigen processing. Cell 2006;127:369–82. Powers C, DeFelippis V, Malouli D, Fruh K. Cytomegalovirus immune evasion. Curr Topics Microbiol Immun 2008;325:333–59. Shin J, Park B, Kim Y, Biegalke BJ, Kang S, Ahn K. A short isoform of human cytomegalovirus US3 functions as a dominant negative inhibitor of the fulllength form. J Virol 2006;80:5397–404. Tenney DJ, Santomenna LD, Goudie KB, Colberg-Poley AM. The human cytomegalovirus US3 immediate-early protein lacking the putative transmembrane domain regulates gene expression. Nucleic Acids Res 1993;21:2931– 7. Thrower AR, Bullock GC, Bissell JE, Stinski MF. Regulation of a human cytomegalovirus immediate-early gene (US3) by a silencer-enhancer combination. J Virol 1996;70:91–100. Weston K. An enhancer element in the short unique region of human cytomegalovirus regulates the production of a group of abundant immediate early transcripts. Virology 1988;162:406–16. Zhao Y, Biegalke BJ. Functional analysis of the human cytomegalovirus immune evasion protein, pUS322 kDa. Virology 2003;315:353–61.