Cell Host & Microbe
Previews Lin, L., Ibrahim, A.S., Xu, X., Farber, J.M., Avanesian, V., Baquir, B., Fu, Y., French, S.W., Edwards, J.E., Jr., and Spellberg, B. (2009). PLoS Pathog. 5, e1000703.
Spellberg, B., Ibrahim, A.S., Yeaman, M.R., Lin, L., Fu, Y., Avanesian, V., Bayer, A.S., Filler, S.G., Lipke, P., Otoo, H., and Edwards, J.E., Jr. (2008). Infect. Immun. 76, 4574–4580.
Nanjappa, S.G., and Klein, B.S. (2014). Curr. Opin. Immunol. 28, 27–33.
Torosantucci, A., Bromuro, C., Chiani, P., De Bernardis, F., Berti, F., Galli, C., Norelli, F., Bellucci, C., Polonelli, L., Costantino, P., et al. (2005). J. Exp. Med. 202, 597–606.
Schmidt, C.S., White, C.J., Ibrahim, A.S., Filler, S.G., Fu, Y., Yeaman, M.R., Edwards, J.E., Jr., and Hennessey, J.P., Jr. (2012). Vaccine 30, 7594–7600.
Wu¨thrich, M., Gern, B., Hung, C.Y., Ersland, K., Rocco, N., Pick-Jacobs, J., Galles, K., Filutowicz,
H., Warner, T., Evans, M., et al. (2011a). J. Clin. Invest. 121, 554–568. Wu¨thrich, M., Hung, C.Y., Gern, B.H., PickJacobs, J.C., Galles, K.J., Filutowicz, H.I., Cole, G.T., and Klein, B.S. (2011b). J. Immunol. 187, 1421–1431. Wu¨thrich, M., Brandhorst, T.T., Sullivan, T.D., Filutowicz, H., Sterkel, A., Stewart, D., Li, M., Lerksuthirat, T., LeBert, V., Shen, Z.T., et al. (2015). Cell Host Microbe 17, this issue, 452–465.
The Multidimensional Nature of Antiviral Innate Immunity Reuben S. Harris,1,* Fred W. Perrino,2 and Nadine M. Shaban1 1Department of Biochemistry, Molecular Biology and Biophysics, Institute for Molecular Virology, Center for Genome Engineering, and Masonic Cancer Center, University of Minnesota, Minneapolis, MN 55455, USA 2Department of Biochemistry, Center for Structural Biology, Wake Forest School of Medicine, Winston-Salem, NC 27157, USA *Correspondence: [email protected] http://dx.doi.org/10.1016/j.chom.2015.03.011
Viruses possess elaborate defensive mechanisms to evade host innate immune responses. In this issue of Cell Host & Microbe, Stavrou et al. (2015) reveal how the murine leukemia virus uses a sugar-protein shield to protect from inevitable destruction by cellular innate immune factors including the APOBEC3 DNA mutating enzyme.
Mammalian cells have mechanisms to protect from potentially harmful foreign nucleic acids, including viral RNA and DNA. Deficiencies in these mechanisms can predispose to viral infections and precipitate autoimmune diseases. For instance, defects in a DNA exonuclease (TREX1), a dNTP hydrolase (SAMHD1), a double-stranded RNA editing enzyme (ADAR1), an RNA/DNA hybrid endonuclease (RNaseH2), and a cytosolic RNA receptor (MDA5) all lead to an infectionlike disease called Aicardi-Goutie`res syndrome (AGS) (Crow, 2013). The common denominator is that these are all nucleic acid-catabolizing proteins that participate in pathways for sensing and/or clearing away foreign and self-derived nucleic acids. Presumably, defects in these proteins lead to a build-up of stimulatory nucleic acids that trick immune cells into launching a continual antiviral response characterized by high levels of interferon (IFN) production and upregulation of IFN-responsive genes, some of which encode proteins listed above.
In this issue, Stavrou and colleagues (2015) use murine leukemia virus (MLV) as an innate immune probe to deduce how cells detect viral nucleic acids and implement antiviral programs to limit infection (Stavrou et al., 2015). MLV is a distant retroviral relative of the AIDS viruses HIV-1 and HIV-2. These and other retroviruses use an enzyme called reverse transcriptase (RT) to copy their RNA genome into a double-stranded DNA that inserts into the host cell genome. Reverse transcription is an elegant process that occurs within the relatively safe confines of the viral capsid, a semipermeable lattice that shields the viral nucleic acid from the harsh milieu of the cell. Virus replication is thought to advance almost to completion within the capsid, and viral DNA is only exposed to the cytosol for a short time, if at all, before being inserted into the nuclear genome. Interestingly, most natural antiviral mechanisms target the reverse transcription stage of the retrovirus lifecycle, including SAMHD1, TRIM5a, and
APOBEC3 DNA cytosine deaminases (Harris et al., 2012). It is therefore understandable that many antiretroviral drugs also work by blocking this step of the viral life cycle. Prior work by the same group had shown that the MLV capsid is more fragile in viral variants lacking the amino-terminal end of the capsid precursor polyprotein Gag, which harbors several glycosylation sites (Stavrou et al., 2013). Infection of a murine macrophage cell line with glycoGag mutant viruses results in elevated IFN-b production, which skyrockets in the absence of TREX1, presumably due to the persistence of more immunostimulatory viral DNA replication intermediates. The replication of glyco-Gag mutant viruses is strongly restricted by cellular APOBEC3, as both glyco-Gag mutant and non-mutant viruses are able to replicate robustly and to similarly high levels in APOBEC3 null animals. Moreover, mutations that restore glyco-Gag function frequently occurred in vivo, but only in APOBEC3-expressing animals, implying
Cell Host & Microbe 17, April 8, 2015 ª2015 Elsevier Inc. 423
Cell Host & Microbe
Previews
that the major function of absence of STING, suggestwild-type Gag is shielding ing that this could be the first the virus from APOBEC3defense against retroviral inmediated restriction. Tofections and perhaps other gether with biochemical types of foreign DNA. The postudies of capsid integrity tency of the APOBEC3 dewith and without the glycofense was further evidenced sylation region of Gag, a by the selection of viral mumodel emerged in which tants with fully restored glycosylated Gag fortifies glyco-Gag functionality even capsid stability. in the absence of STING. Figure 1. Multidimensional Antiviral Innate Immunity Cytosolic DNA from a variety of sources, including retrovirus infections, is a The present study leverOverall, a model emerges in substrate for effector mechanisms such as APOBEC3, TREX1, and other ages this unique system to which the innate immune nucleic acid catabolizing proteins, which collectively serve to eliminate the effector proteins APOBEC3 identify innate immune comoffending DNA. If cytosolic DNA persists, it may be sensed by cGAS, and TREX1 are front-line deponents that recognize viral DDX41, and IFI203, and then activation of STING will relay through TBK1 and IRF3 to generate a full-blown interferon response including upregulation fenders that can be further DNA (i.e., sensors and/or of many effector mechanisms. Self DNA can be from a variety of sources, induced if the DNA sensors sensor transducers), rationalincluding dead cells, as well as from within living cells (e.g., retrotransposon are activated by nucleic acids izing that depleting these critDNA not shown). that escape the front line ical components from cells (Figure 1). will block MLV-induced IFNMost aspects of the cytosolic response b production (Stavrou et al., 2015). Four distinct from cGAS (Figure 1). In further clear leads were obtained in a series of support of this model, cells from IFI203 to foreign DNA are conserved in knockdown experiments. Depletion of and STING mutant animals showed mammals. It is therefore likely that many the recently discovered cytosolic DNA re- diminished IFN-b production in response of the observations of Stavrou and colleagues will directly translate to the ceptor, cGAS, suppressed IFN-b produc- to viral infection. One of the most interesting, but debat- human system, and vice versa, as tion, as did knockdown of its downstream signal transducer STING. In this pathway, able, aspects of the work by Stavrou and the field is learning from the ongoing cytosolic DNA is bound by cGAS, trig- colleagues is the concept of hierarchies in identification of new AGS genes (Crow, gering production of the secondary the DNA sensing and clearing mecha- 2013). For instance, given the clear messenger cGAMP and activation of nisms. First, the DNA exonuclease synergy between APOBEC3 and known STING, which in turn activates the kinase TREX1 is a dominant viral DNA clearing AGS proteins in innate immunity, one of TBK1, resulting in phosphorylation and mechanism, and many of the authors’ the human APOBEC3s could easily be nuclear translocation of the transcription observations would not have been involved in the AGS pathway of nucleic factor IRF3 and subsequent induction of possible had they not first weakened this acid defenses. However, evolution is many genes including IFN-b (Barber, defense by depleting TREX1 from cells. a curious beast, and innate immune Second, APOBEC3 appears to matter in proteins often vary in copy number, 2014; Cai et al., 2014) (Figure 1). The other two leads had equally strong both the virus-producing cells and the amino acid composition, and even funceffects on MLV-induced IFN-b production soon-to-be-infected target cells. Many tion between species, and therefore but are harder to integrate in the same prior studies have shown that APOBEC3 some differences between the different pathway. DDX41 is a DExD/H-box heli- packages into assembling viral particles branches of the mammalian phylogenetic case previously implicated in responding in producer cells, yet it exerts its antiviral tree are to be expected. For instance, to cytosolic DNA (Zhang et al., 2011). effects by deaminating viral cDNA and most humans have an armada of seven IFI203 has multiple DNA binding domains blocking reverse transcription in target APOBEC3 proteins, in comparison to and belongs to a larger family of AIM2- cells (likely all within the confines of the the single enzyme produced in mice, like receptor proteins also implicated viral capsid). Several reports, including and partial functional redundancies previously in innate immunity to DNA the predecessor to this study (Stavrou may prevent single gene defects from (Brunette et al., 2012). Interestingly, HIV- et al., 2013), have also suggested that manifesting as AGS. Nevertheless, 1-based particles also triggered IFN-b target-cell APOBEC3 may also interfere as for the DNA sensors, this partial production dependent upon cGAS, with reverse transcription, especially un- redundancy is assuring as it provides IFI203, and STING, but the requirement der conditions in which capsid integrity additional security against potentially for DDX41 was less stringent. The shared may be less than perfect (e.g., glyco- harmful foreign nucleic acids, including requirements suggest the existence of Gag mutants). Thus, APOBEC3 may exert retroviral DNA replication intermediates an underlying general mechanism, but at least two layers of innate immune and many agents of which we are blissthe subtle difference suggests that protection, and this is evidenced by the fully unaware. different retroviruses may trigger different better-than-wild-type infectivity of glycoinnate immune responses. In fact, a series Gag mutant MLV in APOBEC3-deficient ACKNOWLEDGMENTS of co-immunoprecipitation experiments mice in vivo. Moreover, Stavrou and suggested that IFI203, DDX41, and colleagues show that APOBEC3 even R.S.H. laboratory virology studies are supported STING may be part of the same complex, exerts an antiviral effect in the complete by NIAID R01 AI064046 and NIGMS P01 424 Cell Host & Microbe 17, April 8, 2015 ª2015 Elsevier Inc.
Cell Host & Microbe
Previews GM091743, and F.W.P. laboratory studies by NIGMS R01 GM110734.
Cai, X., Chiu, Y.H., and Chen, Z.J. (2014). Mol. Cell 54, 289–296.
REFERENCES
Crow, Y.J. (2013). Handb. Clin. Neurol. 113, 1629– 1635.
Barber, G.N. (2014). Trends Immunol. 35, 88–93. Brunette, R.L., Young, J.M., Whitley, D.G., Brodsky, I.E., Malik, H.S., and Stetson, D.B. (2012). J. Exp. Med. 209, 1969–1983.
Harris, R.S., Hultquist, J.F., and Evans, D.T. (2012). J. Biol. Chem. 287, 40875–40883. Stavrou, S., Nitta, T., Kotla, S., Ha, D., Nagashima, K., Rein, A.R., Fan, H., and Ross, S.R.
(2013). Proc. Natl. Acad. Sci. USA 110, 9078– 9083. Stavrou, S., Blouch, K., Kotla, S., Bass, A., and Ross, S.R. (2015). Cell Host Microbe 17, this issue, 478–488. Zhang, Z., Yuan, B., Bao, M., Lu, N., Kim, T., and Liu, Y.J. (2011). Nat. Immunol. 12, 959–965.
Two Tales (Tails) of SAMHD1 Destruction by Vpx Xiaoyun Ji1 and Yong Xiong1,* 1Department of Molecular Biophysics and Biochemistry, Yale University, New Haven, CT 06520, USA *Correspondence: [email protected] http://dx.doi.org/10.1016/j.chom.2015.03.013
The lentivirus protein Vpx/Vpr recognizes the host restriction factor SAMHD1 at either its N- or C-terminal tail and targets it for destruction by the cellular protein degradation machinery. In this issue of Cell Host & Microbe, Schwefel et al. (2015) report the structural basis of SAMHD1 N-terminal targeting by Vpx.
Hijacking the host protein degradation machinery is a common strategy employed by viruses to target host immune proteins for destruction. Specifically, cullin-RING E3 ubiquitin ligases (CRLs) are frequently exploited by viruses to polyubiquitinate cellular antiviral restriction factors and target them for degradation by the proteasome (Barry and Fru¨h, 2006). This type of E3 ubiquitin ligase machinery contains a cullin scaffolding protein, substrate receptor proteins that specify target proteins, and adaptor proteins that bridge the substrate receptor to the cullin scaffold. Viruses have evolved proteins that mimic substrate receptors in order to recruit specific host restriction factors for degradation. For retroviruses, including HIV-1, the viral accessory proteins Vif, Vpu, Vpr, and/or Vpx act as substrate receptor mimics (Strebel, 2013). These viral proteins mediate the degradation of many extensively studied immune molecules and retroviral restriction factors (Figure 1A): Vif targets the viral genomemutating enzyme APOBEC3G; Vpu targets the viral tethering protein BST-2/ Tetherin and the immune molecule CD4; Vpx or Vpr targets the deoxynucleoside triphosphate (dNTP)-depleting enzyme SAMHD1.
The Vpx/Vpr-mediated degradation of SAMHD1 is the most recently discovered example of viral hijacking of the CRL ubiquitin-proteasome pathway (Hrecka et al., 2011; Laguette et al., 2011). In fact, the knowledge of the interaction between Vpx and the cullin 4A-DDB1DCAF1 ubiquitin ligase predated and directly led to the discovery of SAMHD1 as the targeted restriction factor. SAMHD1 inhibits retroviruses, including HIV-1, in non-dividing cells such as myeloid-lineage and resting CD4+ T cells. The dNTP triphosphohydrolase (dNTPase) activity of the enzyme likely depletes the cellular dNTPs required for viral reverse transcription. Human SAMHD1 is 626 amino acids long and consists of an N-terminal tail (residues 1–35), a protein-interacting sterile alpha motif (SAM) domain (residues 37–109), a catalytic histidine-aspartate (HD) domain (residues 110–600), and a C-terminal tail (residues 601–626). The mechanism by which Vpx/Vpr targets SAMHD1 is among the most intriguing of viral CRL-hijacking events because of the multifaceted targeting modes employed by Vpx/Vpr (Fregoso et al., 2013). Vpx proteins of different simian immunodeficiency virus (SIV) lineages
and HIV-2 recognize either the N or C terminus of SAMHD1 in a species-specific manner. Vpx from HIV-2 and SIVsmm (SIV infecting sooty mangabey monkeys) lineages target the C-terminal tail of host SAMHD1, while Vpx from the SIVmnd-2 (infecting mandrill) lineage targets the N-terminal tail of mandrill SAMHD1. The Vpx proteins between the two lineages have 30%–40% sequence identity and are structurally homologous. Adding to the complexity, in certain SIV lineages this recognition is conferred via the Vpx paralog Vpr. HIV-1 encodes Vpr, but not Vpx, and does not target SAMHD1 even though HIV-1 Vpr interacts with the same host E3 ligase (Hrecka et al., 2007). The target of HIV-1 Vpr is currently unknown. Analysis of the hijacked host E3 ligase components shows that DCAF1, which forms the substrate receptor together with Vpx/Vpr, is virtually identical across species. Additionally, human and monkey SAMHD1 are highly conserved (>93% identical). Perplexing mechanistic and evolutionary questions remain, such as how the structurally homologous Vpx/Vpr proteins use entirely different SAMHD1 motifs (N- or C-terminal) to recruit SAMDH1 to the same E3 ligase and how the species-specific
Cell Host & Microbe 17, April 8, 2015 ª2015 Elsevier Inc. 425
Previews Lin, L., Ibrahim, A.S., Xu, X., Farber, J.M., Avanesian, V., Baquir, B., Fu, Y., French, S.W., Edwards, J.E., Jr., and Spellberg, B. (2009). PLoS Pathog. 5, e1000703.
Spellberg, B., Ibrahim, A.S., Yeaman, M.R., Lin, L., Fu, Y., Avanesian, V., Bayer, A.S., Filler, S.G., Lipke, P., Otoo, H., and Edwards, J.E., Jr. (2008). Infect. Immun. 76, 4574–4580.
Nanjappa, S.G., and Klein, B.S. (2014). Curr. Opin. Immunol. 28, 27–33.
Torosantucci, A., Bromuro, C., Chiani, P., De Bernardis, F., Berti, F., Galli, C., Norelli, F., Bellucci, C., Polonelli, L., Costantino, P., et al. (2005). J. Exp. Med. 202, 597–606.
Schmidt, C.S., White, C.J., Ibrahim, A.S., Filler, S.G., Fu, Y., Yeaman, M.R., Edwards, J.E., Jr., and Hennessey, J.P., Jr. (2012). Vaccine 30, 7594–7600.
Wu¨thrich, M., Gern, B., Hung, C.Y., Ersland, K., Rocco, N., Pick-Jacobs, J., Galles, K., Filutowicz,
H., Warner, T., Evans, M., et al. (2011a). J. Clin. Invest. 121, 554–568. Wu¨thrich, M., Hung, C.Y., Gern, B.H., PickJacobs, J.C., Galles, K.J., Filutowicz, H.I., Cole, G.T., and Klein, B.S. (2011b). J. Immunol. 187, 1421–1431. Wu¨thrich, M., Brandhorst, T.T., Sullivan, T.D., Filutowicz, H., Sterkel, A., Stewart, D., Li, M., Lerksuthirat, T., LeBert, V., Shen, Z.T., et al. (2015). Cell Host Microbe 17, this issue, 452–465.
The Multidimensional Nature of Antiviral Innate Immunity Reuben S. Harris,1,* Fred W. Perrino,2 and Nadine M. Shaban1 1Department of Biochemistry, Molecular Biology and Biophysics, Institute for Molecular Virology, Center for Genome Engineering, and Masonic Cancer Center, University of Minnesota, Minneapolis, MN 55455, USA 2Department of Biochemistry, Center for Structural Biology, Wake Forest School of Medicine, Winston-Salem, NC 27157, USA *Correspondence: [email protected] http://dx.doi.org/10.1016/j.chom.2015.03.011
Viruses possess elaborate defensive mechanisms to evade host innate immune responses. In this issue of Cell Host & Microbe, Stavrou et al. (2015) reveal how the murine leukemia virus uses a sugar-protein shield to protect from inevitable destruction by cellular innate immune factors including the APOBEC3 DNA mutating enzyme.
Mammalian cells have mechanisms to protect from potentially harmful foreign nucleic acids, including viral RNA and DNA. Deficiencies in these mechanisms can predispose to viral infections and precipitate autoimmune diseases. For instance, defects in a DNA exonuclease (TREX1), a dNTP hydrolase (SAMHD1), a double-stranded RNA editing enzyme (ADAR1), an RNA/DNA hybrid endonuclease (RNaseH2), and a cytosolic RNA receptor (MDA5) all lead to an infectionlike disease called Aicardi-Goutie`res syndrome (AGS) (Crow, 2013). The common denominator is that these are all nucleic acid-catabolizing proteins that participate in pathways for sensing and/or clearing away foreign and self-derived nucleic acids. Presumably, defects in these proteins lead to a build-up of stimulatory nucleic acids that trick immune cells into launching a continual antiviral response characterized by high levels of interferon (IFN) production and upregulation of IFN-responsive genes, some of which encode proteins listed above.
In this issue, Stavrou and colleagues (2015) use murine leukemia virus (MLV) as an innate immune probe to deduce how cells detect viral nucleic acids and implement antiviral programs to limit infection (Stavrou et al., 2015). MLV is a distant retroviral relative of the AIDS viruses HIV-1 and HIV-2. These and other retroviruses use an enzyme called reverse transcriptase (RT) to copy their RNA genome into a double-stranded DNA that inserts into the host cell genome. Reverse transcription is an elegant process that occurs within the relatively safe confines of the viral capsid, a semipermeable lattice that shields the viral nucleic acid from the harsh milieu of the cell. Virus replication is thought to advance almost to completion within the capsid, and viral DNA is only exposed to the cytosol for a short time, if at all, before being inserted into the nuclear genome. Interestingly, most natural antiviral mechanisms target the reverse transcription stage of the retrovirus lifecycle, including SAMHD1, TRIM5a, and
APOBEC3 DNA cytosine deaminases (Harris et al., 2012). It is therefore understandable that many antiretroviral drugs also work by blocking this step of the viral life cycle. Prior work by the same group had shown that the MLV capsid is more fragile in viral variants lacking the amino-terminal end of the capsid precursor polyprotein Gag, which harbors several glycosylation sites (Stavrou et al., 2013). Infection of a murine macrophage cell line with glycoGag mutant viruses results in elevated IFN-b production, which skyrockets in the absence of TREX1, presumably due to the persistence of more immunostimulatory viral DNA replication intermediates. The replication of glyco-Gag mutant viruses is strongly restricted by cellular APOBEC3, as both glyco-Gag mutant and non-mutant viruses are able to replicate robustly and to similarly high levels in APOBEC3 null animals. Moreover, mutations that restore glyco-Gag function frequently occurred in vivo, but only in APOBEC3-expressing animals, implying
Cell Host & Microbe 17, April 8, 2015 ª2015 Elsevier Inc. 423
Cell Host & Microbe
Previews
that the major function of absence of STING, suggestwild-type Gag is shielding ing that this could be the first the virus from APOBEC3defense against retroviral inmediated restriction. Tofections and perhaps other gether with biochemical types of foreign DNA. The postudies of capsid integrity tency of the APOBEC3 dewith and without the glycofense was further evidenced sylation region of Gag, a by the selection of viral mumodel emerged in which tants with fully restored glycosylated Gag fortifies glyco-Gag functionality even capsid stability. in the absence of STING. Figure 1. Multidimensional Antiviral Innate Immunity Cytosolic DNA from a variety of sources, including retrovirus infections, is a The present study leverOverall, a model emerges in substrate for effector mechanisms such as APOBEC3, TREX1, and other ages this unique system to which the innate immune nucleic acid catabolizing proteins, which collectively serve to eliminate the effector proteins APOBEC3 identify innate immune comoffending DNA. If cytosolic DNA persists, it may be sensed by cGAS, and TREX1 are front-line deponents that recognize viral DDX41, and IFI203, and then activation of STING will relay through TBK1 and IRF3 to generate a full-blown interferon response including upregulation fenders that can be further DNA (i.e., sensors and/or of many effector mechanisms. Self DNA can be from a variety of sources, induced if the DNA sensors sensor transducers), rationalincluding dead cells, as well as from within living cells (e.g., retrotransposon are activated by nucleic acids izing that depleting these critDNA not shown). that escape the front line ical components from cells (Figure 1). will block MLV-induced IFNMost aspects of the cytosolic response b production (Stavrou et al., 2015). Four distinct from cGAS (Figure 1). In further clear leads were obtained in a series of support of this model, cells from IFI203 to foreign DNA are conserved in knockdown experiments. Depletion of and STING mutant animals showed mammals. It is therefore likely that many the recently discovered cytosolic DNA re- diminished IFN-b production in response of the observations of Stavrou and colleagues will directly translate to the ceptor, cGAS, suppressed IFN-b produc- to viral infection. One of the most interesting, but debat- human system, and vice versa, as tion, as did knockdown of its downstream signal transducer STING. In this pathway, able, aspects of the work by Stavrou and the field is learning from the ongoing cytosolic DNA is bound by cGAS, trig- colleagues is the concept of hierarchies in identification of new AGS genes (Crow, gering production of the secondary the DNA sensing and clearing mecha- 2013). For instance, given the clear messenger cGAMP and activation of nisms. First, the DNA exonuclease synergy between APOBEC3 and known STING, which in turn activates the kinase TREX1 is a dominant viral DNA clearing AGS proteins in innate immunity, one of TBK1, resulting in phosphorylation and mechanism, and many of the authors’ the human APOBEC3s could easily be nuclear translocation of the transcription observations would not have been involved in the AGS pathway of nucleic factor IRF3 and subsequent induction of possible had they not first weakened this acid defenses. However, evolution is many genes including IFN-b (Barber, defense by depleting TREX1 from cells. a curious beast, and innate immune Second, APOBEC3 appears to matter in proteins often vary in copy number, 2014; Cai et al., 2014) (Figure 1). The other two leads had equally strong both the virus-producing cells and the amino acid composition, and even funceffects on MLV-induced IFN-b production soon-to-be-infected target cells. Many tion between species, and therefore but are harder to integrate in the same prior studies have shown that APOBEC3 some differences between the different pathway. DDX41 is a DExD/H-box heli- packages into assembling viral particles branches of the mammalian phylogenetic case previously implicated in responding in producer cells, yet it exerts its antiviral tree are to be expected. For instance, to cytosolic DNA (Zhang et al., 2011). effects by deaminating viral cDNA and most humans have an armada of seven IFI203 has multiple DNA binding domains blocking reverse transcription in target APOBEC3 proteins, in comparison to and belongs to a larger family of AIM2- cells (likely all within the confines of the the single enzyme produced in mice, like receptor proteins also implicated viral capsid). Several reports, including and partial functional redundancies previously in innate immunity to DNA the predecessor to this study (Stavrou may prevent single gene defects from (Brunette et al., 2012). Interestingly, HIV- et al., 2013), have also suggested that manifesting as AGS. Nevertheless, 1-based particles also triggered IFN-b target-cell APOBEC3 may also interfere as for the DNA sensors, this partial production dependent upon cGAS, with reverse transcription, especially un- redundancy is assuring as it provides IFI203, and STING, but the requirement der conditions in which capsid integrity additional security against potentially for DDX41 was less stringent. The shared may be less than perfect (e.g., glyco- harmful foreign nucleic acids, including requirements suggest the existence of Gag mutants). Thus, APOBEC3 may exert retroviral DNA replication intermediates an underlying general mechanism, but at least two layers of innate immune and many agents of which we are blissthe subtle difference suggests that protection, and this is evidenced by the fully unaware. different retroviruses may trigger different better-than-wild-type infectivity of glycoinnate immune responses. In fact, a series Gag mutant MLV in APOBEC3-deficient ACKNOWLEDGMENTS of co-immunoprecipitation experiments mice in vivo. Moreover, Stavrou and suggested that IFI203, DDX41, and colleagues show that APOBEC3 even R.S.H. laboratory virology studies are supported STING may be part of the same complex, exerts an antiviral effect in the complete by NIAID R01 AI064046 and NIGMS P01 424 Cell Host & Microbe 17, April 8, 2015 ª2015 Elsevier Inc.
Cell Host & Microbe
Previews GM091743, and F.W.P. laboratory studies by NIGMS R01 GM110734.
Cai, X., Chiu, Y.H., and Chen, Z.J. (2014). Mol. Cell 54, 289–296.
REFERENCES
Crow, Y.J. (2013). Handb. Clin. Neurol. 113, 1629– 1635.
Barber, G.N. (2014). Trends Immunol. 35, 88–93. Brunette, R.L., Young, J.M., Whitley, D.G., Brodsky, I.E., Malik, H.S., and Stetson, D.B. (2012). J. Exp. Med. 209, 1969–1983.
Harris, R.S., Hultquist, J.F., and Evans, D.T. (2012). J. Biol. Chem. 287, 40875–40883. Stavrou, S., Nitta, T., Kotla, S., Ha, D., Nagashima, K., Rein, A.R., Fan, H., and Ross, S.R.
(2013). Proc. Natl. Acad. Sci. USA 110, 9078– 9083. Stavrou, S., Blouch, K., Kotla, S., Bass, A., and Ross, S.R. (2015). Cell Host Microbe 17, this issue, 478–488. Zhang, Z., Yuan, B., Bao, M., Lu, N., Kim, T., and Liu, Y.J. (2011). Nat. Immunol. 12, 959–965.
Two Tales (Tails) of SAMHD1 Destruction by Vpx Xiaoyun Ji1 and Yong Xiong1,* 1Department of Molecular Biophysics and Biochemistry, Yale University, New Haven, CT 06520, USA *Correspondence: [email protected] http://dx.doi.org/10.1016/j.chom.2015.03.013
The lentivirus protein Vpx/Vpr recognizes the host restriction factor SAMHD1 at either its N- or C-terminal tail and targets it for destruction by the cellular protein degradation machinery. In this issue of Cell Host & Microbe, Schwefel et al. (2015) report the structural basis of SAMHD1 N-terminal targeting by Vpx.
Hijacking the host protein degradation machinery is a common strategy employed by viruses to target host immune proteins for destruction. Specifically, cullin-RING E3 ubiquitin ligases (CRLs) are frequently exploited by viruses to polyubiquitinate cellular antiviral restriction factors and target them for degradation by the proteasome (Barry and Fru¨h, 2006). This type of E3 ubiquitin ligase machinery contains a cullin scaffolding protein, substrate receptor proteins that specify target proteins, and adaptor proteins that bridge the substrate receptor to the cullin scaffold. Viruses have evolved proteins that mimic substrate receptors in order to recruit specific host restriction factors for degradation. For retroviruses, including HIV-1, the viral accessory proteins Vif, Vpu, Vpr, and/or Vpx act as substrate receptor mimics (Strebel, 2013). These viral proteins mediate the degradation of many extensively studied immune molecules and retroviral restriction factors (Figure 1A): Vif targets the viral genomemutating enzyme APOBEC3G; Vpu targets the viral tethering protein BST-2/ Tetherin and the immune molecule CD4; Vpx or Vpr targets the deoxynucleoside triphosphate (dNTP)-depleting enzyme SAMHD1.
The Vpx/Vpr-mediated degradation of SAMHD1 is the most recently discovered example of viral hijacking of the CRL ubiquitin-proteasome pathway (Hrecka et al., 2011; Laguette et al., 2011). In fact, the knowledge of the interaction between Vpx and the cullin 4A-DDB1DCAF1 ubiquitin ligase predated and directly led to the discovery of SAMHD1 as the targeted restriction factor. SAMHD1 inhibits retroviruses, including HIV-1, in non-dividing cells such as myeloid-lineage and resting CD4+ T cells. The dNTP triphosphohydrolase (dNTPase) activity of the enzyme likely depletes the cellular dNTPs required for viral reverse transcription. Human SAMHD1 is 626 amino acids long and consists of an N-terminal tail (residues 1–35), a protein-interacting sterile alpha motif (SAM) domain (residues 37–109), a catalytic histidine-aspartate (HD) domain (residues 110–600), and a C-terminal tail (residues 601–626). The mechanism by which Vpx/Vpr targets SAMHD1 is among the most intriguing of viral CRL-hijacking events because of the multifaceted targeting modes employed by Vpx/Vpr (Fregoso et al., 2013). Vpx proteins of different simian immunodeficiency virus (SIV) lineages
and HIV-2 recognize either the N or C terminus of SAMHD1 in a species-specific manner. Vpx from HIV-2 and SIVsmm (SIV infecting sooty mangabey monkeys) lineages target the C-terminal tail of host SAMHD1, while Vpx from the SIVmnd-2 (infecting mandrill) lineage targets the N-terminal tail of mandrill SAMHD1. The Vpx proteins between the two lineages have 30%–40% sequence identity and are structurally homologous. Adding to the complexity, in certain SIV lineages this recognition is conferred via the Vpx paralog Vpr. HIV-1 encodes Vpr, but not Vpx, and does not target SAMHD1 even though HIV-1 Vpr interacts with the same host E3 ligase (Hrecka et al., 2007). The target of HIV-1 Vpr is currently unknown. Analysis of the hijacked host E3 ligase components shows that DCAF1, which forms the substrate receptor together with Vpx/Vpr, is virtually identical across species. Additionally, human and monkey SAMHD1 are highly conserved (>93% identical). Perplexing mechanistic and evolutionary questions remain, such as how the structurally homologous Vpx/Vpr proteins use entirely different SAMHD1 motifs (N- or C-terminal) to recruit SAMDH1 to the same E3 ligase and how the species-specific
Cell Host & Microbe 17, April 8, 2015 ª2015 Elsevier Inc. 425
