New Mechanism for APOBEC3G? AUTOIMMUNITY TLR Ligands Are Not Sufficient to Break Cross-Tolerance to Self-Antigens

RESEARCH HIGHLIGHTS

VIRAL IMMUNITY IN BRIEF

New mechanism for APOBEC3G? AUTOIMMUNITY TLR ligands are not sufficient to break cross-tolerance to self-antigens. Hamilton-Williams, E. E. et al. J. Immunol. 174, 1159–1163 (2005). The presence of autoreactive T cells in individuals who do not have autoimmune disease shows that this is not sufficient for the induction of pathology. This paper looked at the potential role of Toll-like receptor (TLR) ligation in providing non- specific inflammatory signals that are required to turn potential autoreactivity into overt disease. Transgenic mice expressing ovalbumin in pancreatic β-cells under control of the rat insulin promoter (RIP-mOVA mice) do not develop diabetes when injected with OVA-specific CD8+ T (OT-I) cells in the absence of specific CD4+ T-cell help, owing to T-cell deletion by cross-tolerizing dendritic cells (DCs). When these mice were also injected with ligands for TLR2, TLR3, TLR4 or TLR9, which could activate the DCs for cross-priming, the Members of the APOBEC protein did not significantly affect antiviral OT-I cells were still deleted unless OVA-specific CD4+ family all contain consensus cytidine- activity against susceptible strains T (OT-II) cells were added. Therefore, TLR ligation alone deaminase motifs, some of which have of HIV-1. Comparison of the differ- is not sufficient to break tolerance in this model, which is been shown to catalyse the deami- ent APOBEC3G mutants showed supported by the fact that infection with TLR-ligand-bearing nation of cytidine to uridine in DNA that antiviral activity depends on at pathogens does not commonly result in autoimmunity. This is or RNA. Replication over the mutated least one of the domains being in contrast to a recent study in Nature Medicine by Lang et al. uridine (which is recognized as thymi- intact, but it doesn’t matter which (11, 138–145 (2005); covered as a highlight in the February dine by the polymerase machinery) one. It therefore seems that there issue of Nature Reviews Immunology), which showed that TLR then results in a guanosine (G) to might be an alternative mechanism, ligation could convert the presence of autoreactive T cells into adenosine (A) mutation in the pairing independent of DNA deamination, for autoimmune disease in another transgenic mouse model of strand. APOBEC3G has potent anti- the antiviral activity of APOBEC3G. diabetes, through upregulation of MHC class I expression on retroviral activity, and consistent with Consistent with this, HIV-1 cDNAs the target organ by interferon-α produced in response to TLR its proposed role as a cytidine deami- recovered from cells infected with stimulation. The difference between these studies could relate nase, it induces high levels of deleteri- APOBEC3G-susceptible virions pro- to the requirement for CD4+ T-cell help to prevent deletion of ous G to A mutations in HIV-1 cDNA. duced in the presence of C-terminal- CD8+ T cells in the former model but not in the latter. Given that APOBEC3G has two con- mutated APOBEC3G showed no sensus cytidine-deaminase motifs, evidence of G to A hypermuta- REGULATORY TCELLS one at each terminus, Sheehy and col- tion, despite the presence of a strong leagues set out to find which of these antiretroviral effect. Contact-mediated suppression by CD4+CD25+ is important for the antiviral effect. The authors experimentally ruled regulatory cells involves a granzyme B-dependent, First, they tested the ability of each out a dominant-negative effect of perforin-independent mechanism. domain to mutate DNA in an in vitro C-terminal mutations on any DNA- assay in which mutation of the RNA mutating activity of the N-terminus of Gondek, D. C. et al. J. Immunol. 174, 1783–1786 (2005). polymerase B gene of Escherichia coli APOBEC3G. They also showed that The in vitro suppressive function of CD4+CD25+ regulatory by APOBEC3G is measured by the the failure of C-terminal APOBEC3G T (TReg) cells is contact dependent, and cross-linking of frequency of acquired resistance to mutants to edit viral cDNA did not glucocorticoid-induced tumour-necrosis factor (TNF)- rifampicin. Mutations in the amino result from a lack of virion packaging. receptor-related protein (GITR) on TReg cells abrogates this. (N)-terminal domain of APOBEC3G Therefore, these results call into Using DNA-microarray analysis, Gondek et al. showed that did not affect the frequency of question the previously assumed stimulation of TReg cells solely through the T-cell receptor rifampicin resistance, whereas muta- absolute correlation between cytidine- increased levels of granzyme B cDNA but that concomitant tions in the carboxyl (C)-terminus deaminase activity and antiretroviral GITR signalling markedly diminished such upregulation. markedly decreased the frequency of function of APOBEC3G, which might Functionally, TReg cells isolated from granzyme-B-deficient resistant bacterial colonies. Surpri- have implications for the function mice were impaired in their ability to suppress in vitro singly, this indicates that only the of other members of the APOBEC CD4+CD25– T-cell proliferation. However, the suppressive

C-terminal cytidine-deaminase motif family. function of TReg cells was perforin independent, so the of APOBEC3G has DNA-mutating Kirsty Minton mechanism by which granzyme B affects TReg-cell-mediated activity. References and links suppression remains unclear. Interestingly, TReg cells induced However, although mutations at ORIGINAL RESEARCH PAPER Newman, E. N. C. CD4+CD25– T-cell apoptosis, so further studies will aim to et al. Antiviral function of APOBEC3G can be the C-terminus inhibited the muta- dissociated from cytidine deaminase activity. define the molecular link between TReg-cell production of tional activity of APOBEC3G, they Curr. Biol. 15, 166–170 (2005). granzyme B and CD4+CD25– T-cell apoptosis.