Hello Readers! My apologies for the unexpected hiatus as preliminary exams and the end of the semester have occupied the bulk of my time recently. I thought I would make the most of the situation and post the written portion that I’ve recently completed as it is an interesting subject I was unaware of until recently. Studies in this area may lead to future treatments for retroviral infections such as human immunodeficiency virus 1 (HIV-1), the infectious agent responsible for acquired immune deficiency syndrome (AIDS) by showing exactly how the host protein APOBEC3G exerts an antiviral effect against this virus in the cell.

The current pandemic of human immunodeficiency virus 1 (HIV-1) has led to an intensive search for effective antiretroviral therapies. Among the recently discovered mediators of host cellular immunity to HIV is APOBEC3G (apolipoprotein B mRNA editing catalytic polypeptide-like 3 G), originally known as CEM15¹. APOBEC3G (A3G) is a member of the APOBEC/AID family of deaminases and produced in only a subset of human cells; notably among them monocytes, CD4⁺ T cells,T-lymphocytes, and immature dendritic cells^2–5. Prior to the discovery of A3G it was known that HIV-1 was capable of infecting a number of human cell types but that HIV-1 without the Vif (viral infectivity factor) protein (HIV-1 ΔVif) was only capable of producing infectious virus in a subset of these cell lines termed permissive^6,7. Through the utilization of heterokaryon analysis with permissive and non-permissive cell lines it was possible to assay for the production of infectious HIV-1 ΔVif and it was observed that there was a dominant restriction factor of HIV-1 ΔVif in non-permissive cells⁶. Once A3G was discovered it was shown to be the responsible agent in the inhibition the production of infectious HIV-1 ΔVif from non-permissive cells when incorporated into the HIV virion¹, and that the HIV-1 Vif protein is responsible for the suppression of A3G and evasion of this host antiviral effect⁸. The discovery of A3G as a potent retroviral agent has wide-ranging implications in human innate immunity to retroviruses and for potential therapies. As such, functional analysis of the multitude of A3G’s antiviral effects, its interaction with Vif as well as host factors, and the potential to generate therapies that target these interactions have undergone extensive further investigation.

Cytidine Deamination and A3G: Soon after the discovery of A3G, it was shown that A3G functions as an antiviral defense mechanism though the lethal editing via cytidine deamination of nascent HIV-1 reverse transcripts⁹. This editing function results in a high level of guanosine to adenosine (G-to-A) mutations in the plus-strand HIV-1 DNA. These mutations preferentially occur in a 5’-GGG to 5’-GGA manner in the positive sense strand due to deamination of cytosine to uracil in the negative sense strand¹⁰. An approach using point mutation and x-ray crystallography have shown that the composition of charged and hydrophobic residues in the active site groove are all important the substrate binding and deaminase activity¹¹. Current opinion is that these editing mutations occur during A3G association with the single-stranded negative sense DNA template produced during reverse transcription. It has been shown that A3G preferentially targets single stranded DNA over both double stranded DNA and an RNA-DNA hybrid, which would account for the editing of the nascent HIV-1 transcript during reverse transcription^12,13. The end result of these mutations are often the disruption of promoters, introduction of new stop codons, abnormal splicing events, and aberrant protein production from events such as the mutation of the tryptophan codon TGG to the TAG, TAA, or TGA stop codons¹⁴. However, the introduction of these stop codons may not be intentional lethal editing by A3G but merely a byproduct of the particular target sequence preference for deamination. This conclusion is supported by work showing that even with different target sequence preferences, mutant forms of A3G still retained deaminase function and an antiviral phenotype¹⁵.

The sequences to be deaminated in the nascent transcript are targeted by a recognition loop of residues 314-322 in the C-terminal domain of A3G^15,16. When this loop was substituted with the corresponding loops from either APOBEC3F or the related AID (activation-induced deaminase) the targeted sequence in the negative strand shifted to mirror that of the donor indicating that this region is key in the determination of target sites for deamination¹⁵. Interestingly, despite the differences in target sequence preference all loop-graft variants of A3G demonstrated antiviral activity and generated extensive G-to-A mutations in viral genomes, which indicates that the specific targeting of A3G to the nascent transcript sequence is independent of the actual deaminase activity. Furthermore, due to differences in targeting preference it was expected that the mutants used in Kohli et al would generate two fold less stop codons and therefore be two-fold reduced in the GFP reporter expression assay; however experimental results confirmed that these mutants generated a matching four-fold magnitude reduction in GFP expression compared to wild type A3G. This conclusion supports the theory that the antiviral effect of cytidine deamination is due to broad non-specific editing and not the specific introduction of stop codons.

There is also evidence in the fossilized in our own genomes that A3G has been responsible for the hypermutation and inactivation of human endogenous retroviruses (HERVs) during the course of human evolution¹⁷. Here Lee et al reconstituted HERV-K(HML-2) in a functional form as a consensus sequence from many duplicated HERVs in the human genome and demonstrated that A3G is the protein responsible for the high levels of hypermutation found in two of these proviruses (HERV-KI and HERV-K60). This was done using sequencing to compare the functional, reconstituted HERV-K_CON to mutations in each of the endogenous proviruses and their flanking regions in the genome in order to determine whether the mutations resulted from specific activity against each HERV or as a result of random mutagenesis. This level of analysis provided evidence of high levels of G-to-A mutations in HERV-KI and HERV-60, and by putting these mutations in the context of the flanking nucleotides it became apparent that these mutations were occurring at the preferred deamination target sites of A3G that had been demonstrated in previous work^9,14. This finding is valuable in that it shows the capability of A3G to inactivate a retrovirus by cytidine deamination in the context of an actual human infection.

Antiviral activity independent of deaminase functions of A3G: While deamination activity is inarguably a function of the CD2 domain of A3G^11,18 there is also a significant body of evidence that supports an additional antiviral function independent of deaminase activity. Support for this concept began with the discovery that some APOBEC3 proteins retain antiviral activity even when activity of the deaminase domain has been abolished through point mutation^12,19,20. This theory has gained traction though studies of multiple deaminase-deficient APOBEC3 proteins (A3A, A3G, and A3F) and their antiviral effects^20–22.

Recent studies involving the targeted mutation of the CD1 and CD2 domains of A3G has elucidated how these two distinct domains contribute to the antiviral effect. Highly purified and catalytically active A3G was generated using a baculovirus expression system and used to investigate the relative contributions of each zinc finger motif in regards to nucleic acid binding, deamination, and antiviral activity. By using this approach with wild type and zinc finger mutants they were able to demonstrate that the CD1 zinc finger is primarily responsible for encapsidation, while the CD2 domain was exclusively associated with deamination activity¹². The study by Li et al shows that when the CD2 domain responsible for deaminase activity is replaced with the CD1 domain the resulting deaminase-deficient CD1-CD1 mutant protein still led to reduction in late DNA synthesis, indicating that A3G can inhibit viral DNA synthesis via a non-deamination mechanism unrelated to the function of the CD2 deaminase domain¹⁸.

There is additional information suggesting that A3G targets the integration pathway of HIV-1 as well. This was initially demonstrated by Mbisa et al where RT-PCR assays were utilized to examine defects in plus-strand integration and DNA transfer of HIV-1 cDNAs produced in the presence of A3G²³. With this approach they demonstrated that cDNAs produced in the presence of A3G showed abnormal cleavage of the tRNA primer on the 3’ end of the viral DNA which resulted in the formation of  a 6-bp extension at the viral U5 end of the 3’-LTR resulting in an inappropriate substrate for integration into the host genome²⁴. However, out of line with known functions, they showed that it was necessary to have a fully functioning CD2 domain in order to achieve this effect, indicating that the CD2 domain may have an additional antiviral effect beyond deamination.

Further research has shown that A3G is capable of directly interfering with HIV-1 integrase (IN) in addition to generating unsuitable integration substrates²⁵. Here Luo et al demonstrated that A3G was capable of inhibiting the accumulation of HIV plus-strand viral DNA at all time points studied and that at 12 hours post infection these transcripts were reduced by 60% in modified 293 cells expressing A3G. This inhibition was tracked to a direct physical interaction between the first linker region at residues 104 to 156 in the N-terminal domain of A3G and the C-terminal domain of integrase. These results support the prior works of Mbisa et al that demonstrated that A3G is capable of inhibiting HIV-1 DNA integration through defects in the plus-strand transfer and DNA integration pathway^23,24. Interestingly, these results together suggest that the region responsible for this specific antiviral effect is located outside of both CD1 and CD2 domains indicating that further structural studies are necessary to understand the nature of this interaction and how it exerts an antiviral effect against HIV in vivo.

An additional proposed mechanism for the non-deamination antiviral effect of A3G is the direct interference of A3G with HIV-1 reverse transcriptase (RT). Early work in this area suggested that the nucleic acid binding properties of both A3G and HIV-1 RT contribute to the inhibition of reverse transcription by competing for access to template DNA²⁶. Iwatani et al demonstrated that catalytically inactive A3G was capable of significantly inhibiting the negative ssDNA synthesis of reverse transcription as read out by production D-18 primed (-) ssDNA, indicating that A3G can interfere with reverse transcription²⁶. This hypothesis is further bolstered by the work of Wang et al using a cell-based co-immunopreciptation assay coupled with deletion analysis to demonstrate that A3G directly binds HIV-1 RT between residues 65-132 in the N-terminal region of A3G²⁷. Overexpression of a peptide fragment derived from residues 65-132 of A3G resulted in an attenuation of the anti-HIV effect of A3G on reverse transcription. It must be kept in mind then that while this peptide is capable of binding RT competitively with intact A3G, the loss of antiviral effect in competition assays indicates that regions outside of this fragment on A3G are responsible for reverse transcription inhibition in vivo. This mechanism was shown to be functional in the absence of HIV-1 IN, indicating that this may be a wholly independent antiviral mechanism unrelated to the interference with IN proposed by Luo et al²⁵.

Antiviral activity of A3G not limited to HIV-1: Significantly, it was shown by Mangeat et al that A3G is not only effective against HIV-1, but has a broad range of antiviral effects against diverse retroviruses including simian immunodeficiency virus (SIV), equine infectious anemia virus (EIAV), and murine leukemia virus (MLV)⁹. Furthermore, when Vif was introduced via plasmid expression it knocked down the antiviral effect of A3G in all viruses studied, indicating that A3G targets a broad range of retroviruses (EIAV, MLV) and it not limited to HIV and SIV. This has implications for the development of drugs against emerging retroviruses by making use of this broad antiretroviral effect.

Studies with the related protein APOBEC3A (A3A)  have shown effective deaminase-independent inhibition of not retroviruses, but parvoviruses²¹. A3A has been shown to have no effect against lentiviruses in vivo, but here it was observed to inhibit the production of recombinant adeno-associated virus (rAAV) as well as the autonomous parvovirus minute virus of mice (MVM). It is known that A3A has only one cytidine deaminase domain and that this is highly homologous to the CD1 domain of A3G as determined through sequence alignment²⁸. When specific residues conferring anti-rAAV activity were discovered in the VS1 domain of A3A and substituted into the CD2 of A3G it was shown that the modified A3G demonstrated a gain-of-function activity against rAAV, indicating that this VS1 region contributes to the antiviral effect against parvoviruses. These studies further demonstrated that the antiviral effects of some APOBEC3 family members are independent of their deaminase activity and that, in this case at least, the activity can be traced to specific regions in the APOBEC3 protein structure. Additionally, this VS1 mutation is located in the CD2 domain, further indicating that antiviral effects are not restricted to deaminase activity of the CD2 domain of A3G.

Furthermore, other studies have shown that A3G has an antiviral effect against the hepatitis B virus in the absence of deamination^29,30. A3G downregulates the production of both human and duck HBV virions independently of the extensive genomic editing seen in HIV-1 models, indicating that A3G has an antiviral effect against HBV independent of deaminase activity³⁰. In another study HBV was passaged through the semi-permissive HepG2 cell line and when mutant HepG2 cells for deaminase deficient A3G were utilized there was no difference in reduction of the levels of HBV replicative intermediates between the two cell line conditions, indicating that the reduction in viral synthesis was the driving force against HBV, not deamination²⁹. Due to the non-essential nature of deaminase activity in relation to suppression of the production of replication-competent HBV nucleocapsids it has been postulated that the deaminase function is not a part of human innate immunity against HBV, which leaves the role of non-deaminase functions responsible for innate defense³⁰. As such, this function of A3G against HBV warrants further investigation to see if this phenotype holds against other related viruses and if it is the result of a broad antiviral effect targeting steps in viral DNA synthesis.

Mechanism of Vif against A3G: It was initially shown that A3G is targeted for proteosomal degradation by a complex containing HIV-1 Vif-Cul5-SCL^8,31.  Targeting for degradation by the proteasome is achieved by polyubiquitination and has was confirmed using immunoprecipitation of Myc-tagged ubiquitin and HA-tagged A3G³². It has been determined that HIV-1 Vif interacts with A3G at the N-terminal linker sequence outside of the two CD motifs⁸.  It has been recently demonstrated that Vif interacts with the T-cell differentiation factor CFB-β in order to mediate binding with CUL5. This observation was achieved through the targeted knockdown of CFB-β via siRNA which resulted in a gain of antiviral activity against HIV-1 with Vif³². Furthermore, reciprocal knockdown of CUL5 did not result in disruption of the Vif-CBF-β interaction, indicating that CBF-β is uniquely involved in mediating the interaction of Vif and CUL5. This interaction between Vif, A3G, and CBF-β has been shown to be a key determinant in whether A3G can be incorporated into virions and exert an antiviral effect^33,34. As such, this interaction between CBF-β, A3G, and Vif is worth further study as a target for antiretroviral therapy.

Interaction of A3G and Vif and their contribution to drug resistance in HIV-1: Focus on the non-catalytic antiviral function of A3G is advantageous for the development of novel antiretrovirals in that the exploitation of this feature may lead to use of A3G function independent of cytidine deamination. It has been previously demonstrated that suboptimal suppression of A3G by Vif results in enhanced mutation rates and development of resistance to antivirals to HIV-1^14,34,35. The use of naturally occurring point mutants in HIV-1 Vif with suboptimal activity against A3G was shown to induce the appearance of lamivudine (3TC) resistance in viral populations before the virus was cultured in the presence of 3TC¹⁴. Significantly, sequencing of the quasispecies confirmed that these mutations were the result of A3G-directed cytidine deamination events and that they occurred in greater than 40% of proviruses coding for the defective Vif mutants. Sadler et al showed that A3G can induce a sub-lethal level of mutations which actually contributes to a greater diversity of the viral quasipecies³⁵. It was further shown that 3TC resistance is more rapidly fixed in viral populations generated in the presence of both A3G and 3TC and suggested that editing results in greater genetic variation in the viral population, acting as a reservoir that can shape the overall population³⁶. This was shown elegantly by using CEM-SS cells engineered to express A3G. When compared to non-A3G expressing CEM-SS cells, virus produced in the presence of 3TC was 10, 10,000, and 100,000 fold increased at 9, 13, and 16 days post infection, respectively; indicating that HIV-1 coding for partially active Vif exposed to A3G acquired the 3TC resistant phenotype more rapidly as a result of cytidine deamination events³⁶.

Future Directions: Due to the observable antiviral function of A3G it has become a subject of intensive research. While the primary mechanism against HIV-1 appears to be thoroughly rooted in deaminase functions^9,10,12,17, additional research has revealed that A3G retains its antiviral activity in the absence of deamination^{15,19–22,29,30,37}. Furthermore, A3G  has been shown to have antiviral effects against a diverse array of viruses, from related retrovirueses such as SIV, EIAV, and MLV⁹ to the more distantly related Hepadnaviridae^29,30, and AAV²¹. In fact, the antiviral effects of A3G may apply to many viruses that use a single-stranded DNA intermediate during the course of infection which can be targeted by A3G. This has wide-ranging implications in antiviral therapy if this broad mechanism can be appropriately understood and utilized.

The A3G protein has been shown to be a determining factor in HIV-1 infectivity, and HIV-1 has countered this antiviral effect with Vif⁷. This is an example of a molecular arms race where despite the rapid rate of evolution of the A3G protein in humans³⁸, HIV is much more capable of evolving rapidly to evade this innate host defense under selective pressure due to the very nature of its replicative fidelity^14,35,36. Therefore, it would be advantageous to design therapies to block the function of Vif and avert the degradation of A3G, effectively restoring antiviral activity against HIV-1³⁹. In fact, prior work has shown that the proteasome inhibitor MG132 is capable of blocking the Vif-directed polyubiquitination and degradation of A3G and results in A3G packaging in the virion⁴⁰. This work shows that disruption of the proteosomal degradation pathway may hold future treatments for HIV, as protease inhibitors are an accepted method of treating Hepatitis C virus⁴¹.

Additional research into host factors involved in this Vif-A3G interaction has shown that the T-cell differentiation factor CBF-β is critical for Vif recruitment of the complete E3 ubiquitin ligase complex that targets A3G for degradation³². Excitingly, Zhou et al were able to generate a stable complex containing Vif-Cul5-CFBβ-EloB/C which will be invaluable for future studies. Further studies using crystallographic analysis will reveal the specific structural and electrostatic determinants of these interactions, providing valuable information for the development of drug therapies that have high affinity to these regions and effectively block the action of Vif and CBF-β in the formation of the Vif-Culd5-CBFβ-EloB/C complex that targets A3G for degradation and exclusion from the virion. As such, disruption of this complex to prevent the polyubiqutination of A3G and to free A3G for incorporation into HIV virions is an attractive target for rational drug design. Furthermore, these kind of structural studies would be highly useful for understanding and validating the observed non-deaminase antiviral effect of A3G including its interaction with integrase and reverse transcriptase. Understanding of these interactions would provide valuable tools for the rational design of antivirals to be added to HAART regiments that inhibit the functions of integrase and reverse transcriptase to further reduce the ability of HIV to escape the treatment regimen through mutation and increase favorable patient outcomes.

Citations:

1.        Sheehy, A.M., Gaddis, N.C., Choi, J.D. & Malim, M.H. Isolation of a human gene that inhibits HIV-1 infection and is suppressed by the viral Vif protein. Nature 418, 646-50 (2002).

2.        Peng, G. et al. Myeloid differentiation and susceptibility to HIV-1 are linked to APOBEC3 expression. Blood 110, 393-400 (2007).

3.        Vetter, M.L., Johnson, M.E., Antons, A.K., Unutmaz, D. & D’Aquila, R.T. Differences in APOBEC3G expression in CD4+ T helper lymphocyte subtypes modulate HIV-1 infectivity. PLoS pathogens 5, e1000292 (2009).

4.        Chiu, Y.-L. et al. Cellular APOBEC3G restricts HIV-1 infection in resting CD4+ T cells. Nature 435, 108-14 (2005).

5.        Stopak, K.S., Chiu, Y.-L., Kropp, J., Grant, R.M. & Greene, W.C. Distinct patterns of cytokine regulation of APOBEC3G expression and activity in primary lymphocytes, macrophages, and dendritic cells. The Journal of biological chemistry 282, 3539-46 (2007).

6.        Simon, J.H., Gaddis, N.C., Fouchier, R.A. & Malim, M.H. Evidence for a newly discovered cellular anti-HIV-1 phenotype. Nature medicine 4, 1397-400 (1998).

7.        Strebel, K. et al. The HIV “A” (sor) gene product is essential for virus infectivity. Nature 328, 728-30

8.        Marin, M., Rose, K.M., Kozak, S.L. & Kabat, D. HIV-1 Vif protein binds the editing enzyme APOBEC3G and induces its degradation. Nature medicine 9, 1398-403 (2003).

9.        Mangeat, B. et al. Broad antiretroviral defence by human APOBEC3G through lethal editing of nascent reverse transcripts. Nature 424, 99-103 (2003).

10.      Harris, R.S. et al. DNA deamination mediates innate immunity to retroviral infection. Cell 113, 803-9 (2003).

11.      Holden, L.G. et al. Crystal structure of the anti-viral APOBEC3G catalytic domain and functional implications. Nature 456, 121-4 (2008).

12.      Iwatani, Y., Takeuchi, H., Strebel, K. & Levin, J.G. Biochemical activities of highly purified, catalytically active human APOBEC3G: correlation with antiviral effect. Journal of virology 80, 5992-6002 (2006).

13.      Yu, Q. et al. Single-strand specificity of APOBEC3G accounts for minus-strand deamination of the HIV genome. Nature structural & molecular biology 11, 435-42 (2004).

14.      Mulder, L.C.F., Harari, A. & Simon, V. Cytidine deamination induced HIV-1 drug resistance. Proceedings of the National Academy of Sciences of the United States of America 105, 5501-6 (2008).

15.      Kohli, R.M. et al. Local sequence targeting in the AID/APOBEC family differentially impacts retroviral restriction and antibody diversification. The Journal of biological chemistry 285, 40956-64 (2010).

16.      Kohli, R.M. et al. A portable hot spot recognition loop transfers sequence preferences from APOBEC family members to activation-induced cytidine deaminase. The Journal of biological chemistry 284, 22898-904 (2009).

17.      Lee, Y.N., Malim, M.H. & Bieniasz, P.D. Hypermutation of an ancient human retrovirus by APOBEC3G. Journal of virology 82, 8762-70 (2008).

18.      Li, X. et al. Functional analysis of the two cytidine deaminase domains in APOBEC3G. Virology 414, 130-6 (2011).

19.      Newman, E.N.C. et al. Antiviral function of APOBEC3G can be dissociated from cytidine deaminase activity. Current biology : CB 15, 166-70 (2005).

20.      Bishop, K.N., Holmes, R.K. & Malim, M.H. Antiviral potency of APOBEC proteins does not correlate with cytidine deamination. Journal of virology 80, 8450-8 (2006).

21.      Narvaiza, I. et al. Deaminase-independent inhibition of parvoviruses by the APOBEC3A cytidine deaminase. PLoS pathogens 5, e1000439 (2009).

22.      Holmes, R.K., Koning, F.A., Bishop, K.N. & Malim, M.H. APOBEC3F can inhibit the accumulation of HIV-1 reverse transcription products in the absence of hypermutation. Comparisons with APOBEC3G. The Journal of biological chemistry 282, 2587-95 (2007).

23.      Mbisa, J.L. et al. Human immunodeficiency virus type 1 cDNAs produced in the presence of APOBEC3G exhibit defects in plus-strand DNA transfer and integration. Journal of virology 81, 7099-110 (2007).

24.      Mbisa, J.L., Bu, W. & Pathak, V.K. APOBEC3F and APOBEC3G inhibit HIV-1 DNA integration by different mechanisms. Journal of virology 84, 5250-9 (2010).

25.      Luo, K. et al. Cytidine deaminases APOBEC3G and APOBEC3F interact with human immunodeficiency virus type 1 integrase and inhibit proviral DNA formation. Journal of virology 81, 7238-48 (2007).

26.      Iwatani, Y. et al. Deaminase-independent inhibition of HIV-1 reverse transcription by APOBEC3G. Nucleic acids research 35, 7096-108 (2007).

27.      Wang, X. et al. The cellular antiviral protein APOBEC3G interacts with HIV-1 reverse transcriptase and inhibits its function during viral replication. Journal of virology 86, 3777-3786 (2012).

28.      Jarmuz, A. et al. An anthropoid-specific locus of orphan C to U RNA-editing enzymes on chromosome 22. Genomics 79, 285-96 (2002).

29.      Noguchi, C. et al. Dual effect of APOBEC3G on Hepatitis B virus. The Journal of general virology 88, 432-40 (2007).

30.      Rösler, C. et al. APOBEC-mediated interference with hepadnavirus production. Hepatology (Baltimore, Md.) 42, 301-9 (2005).

31.      Yu, X. et al. Induction of APOBEC3G ubiquitination and degradation by an HIV-1 Vif-Cul5-SCF complex. Science (New York, N.Y.) 302, 1056-60 (2003).

32.      Zhang, W., Du, J., Evans, S.L., Yu, Y. & Yu, X.-F. T-cell differentiation factor CBF-β regulates HIV-1 Vif-mediated evasion of host restriction. Nature 481, 376-9 (2012).

33.      Zhou, X., Evans, S.L., Han, X., Liu, Y. & Yu, X.-F. Characterization of the Interaction of Full-Length HIV-1 Vif Protein with its Key Regulator CBFβ and CRL5 E3 Ubiquitin Ligase Components. PloS one 7, e33495 (2012).

34.      Jäger, S. et al. Vif hijacks CBF-β to degrade APOBEC3G and promote HIV-1 infection. Nature 481, 371-5 (2012).

35.      Sadler, H.A., Stenglein, M.D., Harris, R.S. & Mansky, L.M. APOBEC3G contributes to HIV-1 variation through sublethal mutagenesis. Journal of virology 84, 7396-404 (2010).

36.      Kim, E.-Y. et al. Human APOBEC3G-mediated editing can promote HIV-1 sequence diversification and accelerate adaptation to selective pressure. Journal of virology 84, 10402-5 (2010).

37.      Holmes, R.K., Malim, M.H. & Bishop, K.N. APOBEC-mediated viral restriction: not simply editing? Trends in biochemical sciences 32, 118-28 (2007).

38.      Zhang, J. & Webb, D.M. Rapid evolution of primate antiviral enzyme APOBEC3G. Human molecular genetics 13, 1785-91 (2004).

39.      Huthoff, H. & Malim, M.H. Identification of amino acid residues in APOBEC3G required for regulation by human immunodeficiency virus type 1 Vif and Virion encapsidation. Journal of virology 81, 3807-15 (2007).

40.      Liu, B., Yu, X., Luo, K., Yu, Y. & Yu, X.-F. Influence of primate lentiviral Vif and proteasome inhibitors on human immunodeficiency virus type 1 virion packaging of APOBEC3G. Journal of virology 78, 2072-81 (2004).

41.      McHutchison, J.G. et al. Telaprevir for previously treated chronic HCV infection. The New England journal of medicine 362, 1292-303 (2010).

 Edit: Clarified rAAV data as the location of the VS1 loop in the A3G substitution mutant was in the CD2 domain of the C-terminal fragment,  not the CD1 domain.