One of the greatest barriers to curing HIV-1 is the ability of the virus to establish a state of latent infection within infected cells. During viral latency, the virus lies dormant in the form of an integrated, replication-competent provirus within the infected cell. In this state, the virus is undetectable by the host immune system and is unaffected by antiretroviral drugs due to extremely weak or null transcription. These viruses, however, can be induced to produce infectious virus later on and propagate the infection as well as reseed the latent reservoir. The factors that lead to the establishment and maintenance of HIV-1 latency are not all known. Current latency reversal methods are unable to effectively purge the latent reservoir in HIV-1-infected patients, which begs the question whether there are populations of latently-infected cells that are not being targeted by present therapeutics. One of our first lines of defense against retroviral infection is the APOBEC3 family of cytidine deaminases. These enzymes restrict retroviral infection mainly through hypermutation of the proviral DNA, leading to inactivation of the virus. However, it has been shown that low levels of mutations do not necessarily inactivate the virus and can sometimes be beneficial for the virus by increasing genetic diversity and viral fitness. Sublethal mutagenesis has been studied on coding regions of the virus, however, their effects on regulatory regions of the virus such as the LTR promoter have been scarcely explored. My project aimed to examine the effects of APOBEC3-induced mutations on the activity of the HIV-1 LTR promoter and investigate whether these mutations could be generating a subset of latently-infected cells. A library of HIV-1 clones with A3G- or A3F-mutated LTRs was generated. We discovered that the 5’ LTR is not mutated during the first round of HIV-1 infection in our system while the 3’ LTR accumulates mutations. A second round of infection is required for the mutations in the 3’ LTR to be copied over to the 5’ LTR and influence promoter activity. The mutations generated a range of effects on the promoter, with some clones being completely inactivated and no longer responsive to Tat or PMA/Ionomycin induction, while other clones were still highly infectious. The clones of interest were the ones in between this spectrum that were weakly-infectious (≤0.25%) prior to stimulation but were able to be induced above a given threshold (≥10-fold). We denoted these clones latency-prone viruses (LPVs) and were able to identify 10 of these clones within our library. We also explored the effects of individual mutations on the promoter and although these clones were highly infectious, we identified 8 positions of interest that led to weaker infectivity and fluorescence when mutated. The mutations in our library were found to hit important transcription factor binding sites and were also identified in a bioinformatics search of HIV-1-infected patient sequences, which confirms the relevance of our in vitro-identified mutations in a physiological setting. This is the first investigation and evidence of the APOBEC3 proteins contributing to the generation of a subset of latent viruses and constitutes an important contribution to the understanding of HIV-1 latency and its reversal. This previously uncharacterized pool of latently-infected cells could explain why current therapies are ineffective as they do not target these molecular modifications but rather focus on reversing epigenetics and reactivation of the virus. These newly-discovered latency-prone viruses would be a useful tool in testing reactivation drugs and establishes the need to develop novel antiretrovirals that will target a broader spectrum of latently-infected cells. This project has effectively illustrated the dual role of this host-encoded restriction factor and provided further insight into HIV-1 latency, the major hurdle towards a cure for HIV-1.

Mammals are constantly challenged by numerous pathogens that pose a threat to their health. Upon infection, retroviruses quickly integrate their genome into that of their host thereby permanently modifying it. Protein members of the APOBEC3 (A3) family exhibit cytidine deaminase activity that specifically acts on single-stranded DNA to deaminate deoxycytidine bases into deoxyuridines. This process is potentially mutagenic because uracil directs the insertion of adenine on the opposite DNA strand. High levels of mutations induced by A3 proteins in the retroviral genome ultimately inactivate progeny viruses. However, under conditions where low levels of A3 proteins are present, sub-lethal mutagenesis can occur and is generally believed to be beneficial for the virus. Powerful and affordable techniques designed to detect rare deamination events generated by these deaminases along the full length of retroviral genomes are therefore essential. Through the course of my studies, I developed such a new tool that I called HyperHRM which was instrumental to my project’s success. In addition to the antiretroviral affects of their catalytic activity, some members of the A3 family have the ability to hinder reverse transcription independently of their enzymatic properties. Yet, the details underlying the deamination-independent restriction by the proteins remain unclear. Through my work, I have advanced our current understanding of this elusive process by defining the essential role for RNA-binding in the inhibition of the early steps of infection by APOBEC3G (A3G). I also demonstrate that the ability to bind RNA is important for the selection of DNA dinucleotides targeted for deamination by A3 enzymes. Based on the premise that the DNA context for deamination may alter viral fitness in various ways, I then investigated the gene inactivation potency of different A3 based on their preferred DNA substrate. My experiments showed that mutations introduced in a 5'CC context by A3G are much more lethal for the virus because of the high frequency of termination codons that are generated. I therefore clearly established that deamination target specificity has a strong influence on the overall restriction potency of A3 proteins and demonstrated that such specificity was linked to the ability of A3 proteins to bind RNA. Finally, in addition to retroviruses, mobile elements such as retrotransposons can also lead to genomic instability if not properly controlled. The A3 protein family has been shown to play a crucial role in the restriction of these elements through a mechanism that is not believed to require the enzymatic activity of the proteins, although the details of the restriction mechanism are not yet understood. Here, I provide molecular insights on the potential mechanism of retrotransposon restriction by showing that the RNA-binding properties of the enzymes are not involved in the restriction of L1 retrotransposition. A complete elucidation of the modes of restriction employed by the A3 could lead to the development of a new generation of antiretroviral drugs. Overall, my research has led to the design of a new research tool to detect and quantify A3-induced mutations in retroviruses, but more importantly, it has enabled a better understanding of how the RNA-binding abilities of A3 proteins play an essential role in the overall restriction potency of retroviruses and retrotransposons.

Les protéines APOBEC3 (A3A-A3H) catalysent la désamination des cytidines (C) présentes sur l'ADN simple brin en thymidine (T). Cette activité cytidines désaminase a initialement été décrite comme impliquée dans la restriction des rétrovirus et de certains virus à ADN par leur capacité à induire de nombreuses mutations C->T, ou hypermutations, sur les génomes viraux. Il apparait néanmoins que leur activité n'est pas restreinte aux génomes viraux et que certaines A3 peuvent induire des mutations sur l'ADN mitochondrial (A3A, C, F, G et H) et nucléaire (A3A et A3B). Ainsi, l'impact somatique des A3 est désormais établi dans la formation de certains cancers, dont la majorité des mutations, portent signatures des APOBEC3. Aux vues de ces observations, nous nous sommes intéressés à la façon dont sont régulées ces enzymes dans le contexte du stress cellulaire viro-induit ou endogène. La première partie de nos travaux a porté sur la protéine A3DE, seul membre de la famille APOBEC3 ne possédant pas d'activité cytidine désaminase. De façon intéressante, il apparait qu'A3DE est surexprimée dans les cirrhoses infectées par le VHB, VHC ou co-infectées par le VHC et le VHB. Nous avons pu mettre en évidence qu'A3DE interagit et module l'activité d'A3F et d'A3G, deux cytidines désaminases exprimées dans le foie et impliquées dans la restriction du VHB. Dans un second temps, nous nous sommes intéressés à la caractérisation du potentiel génotoxique de la protéine A3B. Cette protéine, de par sa localisation strictement nucléaire, constitue la seule A3 à double domaine n'interagissant pas avec A3DE. Contrairement à A3A, A3B est faiblement active sur l’ADN nucléaire et n’induit pas de cassures de l’ADN double brin. Nous avons pu mettre en évidence par mutagénèse les régions de la protéine impliquées dans l’atténuation de la génotoxicité d’A3B par rapport à A3A et que cette atténuation est conservée chez les primates. Enfin, nous avons étudié le rôle et la régulation d’A3A dans le catabolisme. Nous avons mis en évidence que l’ADN mitochondrial cytoplasmique (ADNcymt) active la voie RIG-I/ARN polymérase III ce qui a pour effet d’induire la production d’IFN qui va activer l’expression d’A3A. A3A va ainsi jouer un rôle dans le catabolisme de l’ADNcymt et contribue à l'élimination de cette source de stress cellulaire, mais occasionnant par la même des dommages sur l’ADN nucléaire. Les A3 sont des enzymes fondamentales de la défense immunitaire innée et du catabolisme de l’ADN. Nous montrons qu’A3DE a pour fonction de moduler l’activité d’A3F et d’A3G tandis qu’A3B, possède un phénotype atténué chez tous les primates et s’avère moins génotoxique que’A3A. Cette dernière participe à la dégradation de l’ADN cytoplasmique, limitant ainsi l’inflammation. Néanmoins, A3A peut s’avérer dangereuse pour l’intégrité génomique et contribuer à l’émergence de cancers, notamment en cas d’inflammation chronique. / APOBEC3 proteins (A3A-A3H) catalyse the deamination of cytosine (C) to thymidine (T) on single stranded DNA. This activity, called cytidine deaminase, has initially been described as a mechanism involved in restriction against retroviruses and DNA viruses by massively inducing C->T mutations on viral genome : this phenomenon is called "hypermutations". Nevertheless, this activity is not virus-specific and some A3 can induce mutations on mitochondrial DNA (A3A, C, F, G, H) and nuclear DNA (A3A and A3B). Thus, the impact of those proteins on cancer formation is now established in cancers where mutations mostly show an APOBEC3 signature. In view of those considerations, we decided to study how those enzymes are regulated in the context of a viral cellular stress or an endogenous cellular stress. The first part of our work is focused on A3DE, the only APOBEC3 lacking a cytidine deaminase activity. Interestingly, A3DE is upregulated in cirrhotic livers infected by HBV, HCV or coinfected with HBV & HCV. We show that A3DE inhibits A3F & A3G activity by interacting with those HBV restriction involved A3. Then, we studied the attributes of the genotoxicity potential of A3B. This protein, by his strictly nuclear localization, constitutes the only double domain A3 which is not regulated by A3DE. Unlike A3A, A3B is weakly active on nuclear DNA and does not induce double strand breaks. We determine by directed mutagenesis the clusters of A3B involved in genotoxicity attenuation compared with A3A. We also show that this attenuation is conserved among primates. Finally, we investigated the role and regulation of A3A in the context of DNA catabolism. We proved that mitochondrial cytoplasmic DNA (mtcyDNA) triggers the RIG-I/DNA polymerase III pathway, which induces IFN production leading to A3A expression. So A3A will be involved in mtcyDNA catabolism and contribute to the clearance of this stress signal, but will also induce double strand breaks on nuclear DNA. A3 are major enzymes of the innate immune response and DNA catabolism. We show that A3DE modulates A3F and A3G activity while A3B is attenuated among primates and is less genotoxic than A3A. A3A participates to cytoplasmic DNA catabolism and limits inflammation. Nevertheless, A3A could be dangerous for the genomic integrity and contributes to cancer, especially in cases of chronic inflammation.

Les cellules de la lignée myéloïde jouent un rôle majeur dans la pathogénèse du VIH, servant à la fois de réservoir viral et permettant la transmission du virus aux cellules T. Cependant, ces cellules sont relativement résistantes à l’infection lentivirale par comparaison aux cellules provenant d’autres lignées. Des études provenant de divers laboratoires, dont le notre, ont montré que les étapes précoces de l’infection semblent se dérouler beaucoup moins efficacement dans ces cellules. Nous avons donc cherché à identifier des facteurs spécifiquement exprimés dans ces cellules pouvant être à l’origine de ce blocage. Ainsi, nous avons pu identifier APOBEC3A (A3A), un membre de la famille des APOBEC3s. A3A est spécifiquement exprimée dans les cellules de la lignée myéloïde mais n’était pas connue pour bloquer la réplication du VIH. Nous avons pu montrer que le pool d’A3A présent dans les cellules cibles est capable de cibler les particules entrantes d’une manière dépendante de son activité enzymatique. Sa déplétion, au moyen d’ARNs interférents, permet d’augmenter l’accumulation de l’ADN viral. Nos données suggèrent donc que A3A induit la dégradation des génomes viraux et que son activité antivirale est dirigée plus généralement contre les lentivirus de Primates. Cependant, la protéine Vpx des membres de la famille VIH-2/SIVSM permet de protéger contre l’action de A3A en induisant sa dégradation via le protéasome.Nous avons mis à jour un nouveau rôle de A3A lors de l’infection lentivirale des cellules myéloïdes. Ces données remettent en question le mode de fonctionnement des membres de la famille des APOBECS et ouvrent de nouvelles questions sur leurs modes d’action et leurs régulations dans les cellules primaires. / Myeloid cells are important for HIV pathogenesis both for viral transmission to T cells and as a viral reservoir. Nonetheless, these cells are quite restrictive to HIV infection compared to established cell lines or primary T cells. Studies from our group as well as other laboratories suggest thatthe early steps of infection are particularly inefficient in these cells . We tried to identify cellular factors specifically expressed in myeloid cells that could be responsible for this block. We identified APOBEC3A as one of such factors. A3A is a member of the APOBEC3 family and is the sole specifically expressed in myeloid cells. We showed that A3A blocks HIV incoming viral particles and more generally primates lentiviruses specifically in myeloid cells in a cytidine deaminase dependant manner. Our data suggest that A3A decreases viral DNA accumulation by inducing the degradation of newly synthesized genome most probably after deamination. Among the proteins coded by primate lentivirus, the HIV-2/SIVsm Vpx protein interacts and degrades A3A thus providing partial protection against A3A.Overall, our data reveal a novel role for A3A in the infection of myeloid cells and raises important questions about the regulation of the cell type specific antiviral role of A3A in primary myeloid cells.

"Host resolution of viral infection is dependent upon components of the innate and acquired immune system. The mammalian protein Apobec3 plays an important role as part of the immune system’s innate defenses through its modification of reverse transcribed viral DNA. Recently, Apobec3 was found to directly inhibit HIV-1 and HBV replication through deaminating newly transcribed deoxycytidine residues to deoxyuridine. The ability of mouse and simian Apobec3 variants to inhibit human retroviruses and vice versa highlights the utility of analyzing cross-species homologues. To better understand this editing enzyme, differentially pathogen-susceptible inbred mice were used as an experimental model. The purpose of this project is to examine the effects of murine Apobec3 (muA3) alternative splicing on its DNA-editing characteristics. Three distinct Apobec3 isoforms were isolated from pathogen-susceptible BALB/cByJ (“C”) inbred mice, and two Apobec3 isoforms came from pathogen-resistant C57BL/6ByJ (“Y”) mice. The five muA3 isoforms were cloned, sequenced, and expressed from a constitutive promoter in a haploid Saccharomyces cerevisia strain. MuA3 DNA-editing activity was measured via the CAN1 forward mutation assay. The five isoforms studied in this project were discovered to be strain-specific. One isoform from each mouse strain mutated the yeast CAN1 locus significantly. Additionally, both muA3 isoform mRNAs derived from the pathogen-resistant Y mice were found to persist at a higher level (2.7 -12.4 fold) than any of the C mouse isoforms. This suggests that the absence of exon 5 or some other signal in the Y mice may influence transcript stability. Evidence also suggests that the murine Apobec3 start codon is actually 33bp upstream of its reference start, with implications for previous research performed using muA3. Sequencing analysis of genomic DNA revealed the presence of a 4bp insertion in a region of BALB/cByJ muA3 which may have disrupted an intronic splicing enhancer signal. Furthermore, a novel BALB/cByJ Apobec3 isoform was characterized. This is the first report of strain-specific processing with regard to muA3."

Human APOBEC3 proteins are host-encoded intrinsic restriction factors that can prevent the replication of a broad range of human and animal retroviruses such as HIV, SIV, FIV, MLVs and XMRV. The main pathway of the restriction is believed to occur as a result of the cytidine deaminase activity of these proteins that converts cytidines into uridines in single-stranded DNA retroviral replication intermediates. Uridines in these DNA intermediates disrupt the viral replication cycle and also alter retrovirus infectivity because of the C-to-T transition mutations generated as a result of the deaminase activity on the minus strand DNA. In addition, human APOBEC3 proteins also exhibit a deamination-independent pathway to restrict retroviruses that is not currently well understood. Although the restriction of retroviruses by human APOBEC3 proteins has been intensely studied in vitro, our understanding of how the murine APOBEC3 (mA3) protein restricts retroviruses and/or prevents zoonotic infections in vivo is very limited. In contrast to humans and primates that have 7 APOBEC3 genes, mice have but a single copy. My study of the function and structure of mA3 revealed that it has an inverted functional organization for cytidine deamination in comparison to the human A3G catalytic sites. I have also found that disruption of the integrity of either of these catalytic sites substantially impedes restriction of HIV and MLV. Interestingly, our data shows that mA3 induces a significant decrease in retroviral activity of HIV and MLVs by exploiting both deamination-dependent and -independent pathways. However, the deaminase activity of mA3 is essential to confer long-term restriction of retroviral infection. My observations suggest that mA3 has dual activities, both deamination-dependent and -independent, that work cooperatively to restrict a broad range of human and animal retroviral pathogens. In the context of the intrinsic immune system, APOBEC3 proteins provide a powerful block to the transmission of retroviral pathogens that very few have found ways to evade.

Human APOBEC3 proteins are host-encoded intrinsic restriction factors that can prevent the replication of a broad range of human and animal retroviruses such as HIV, SIV, FIV, MLVs and XMRV. The main pathway of the restriction is believed to occur as a result of the cytidine deaminase activity of these proteins that converts cytidines into uridines in single-stranded DNA retroviral replication intermediates. Uridines in these DNA intermediates disrupt the viral replication cycle and also alter retrovirus infectivity because of the C-to-T transition mutations generated as a result of the deaminase activity on the minus strand DNA. In addition, human APOBEC3 proteins also exhibit a deamination-independent pathway to restrict retroviruses that is not currently well understood. Although the restriction of retroviruses by human APOBEC3 proteins has been intensely studied in vitro, our understanding of how the murine APOBEC3 (mA3) protein restricts retroviruses and/or prevents zoonotic infections in vivo is very limited. In contrast to humans and primates that have 7 APOBEC3 genes, mice have but a single copy. My study of the function and structure of mA3 revealed that it has an inverted functional organization for cytidine deamination in comparison to the human A3G catalytic sites. I have also found that disruption of the integrity of either of these catalytic sites substantially impedes restriction of HIV and MLV. Interestingly, our data shows that mA3 induces a significant decrease in retroviral activity of HIV and MLVs by exploiting both deamination-dependent and -independent pathways. However, the deaminase activity of mA3 is essential to confer long-term restriction of retroviral infection. My observations suggest that mA3 has dual activities, both deamination-dependent and -independent, that work cooperatively to restrict a broad range of human and animal retroviral pathogens. In the context of the intrinsic immune system, APOBEC3 proteins provide a powerful block to the transmission of retroviral pathogens that very few have found ways to evade.

Human APOBEC3 proteins are host-encoded intrinsic restriction factors that can prevent the replication of a broad range of human and animal retroviruses such as HIV, SIV, FIV, MLVs and XMRV. The main pathway of the restriction is believed to occur as a result of the cytidine deaminase activity of these proteins that converts cytidines into uridines in single-stranded DNA retroviral replication intermediates. Uridines in these DNA intermediates disrupt the viral replication cycle and also alter retrovirus infectivity because of the C-to-T transition mutations generated as a result of the deaminase activity on the minus strand DNA. In addition, human APOBEC3 proteins also exhibit a deamination-independent pathway to restrict retroviruses that is not currently well understood. Although the restriction of retroviruses by human APOBEC3 proteins has been intensely studied in vitro, our understanding of how the murine APOBEC3 (mA3) protein restricts retroviruses and/or prevents zoonotic infections in vivo is very limited. In contrast to humans and primates that have 7 APOBEC3 genes, mice have but a single copy. My study of the function and structure of mA3 revealed that it has an inverted functional organization for cytidine deamination in comparison to the human A3G catalytic sites. I have also found that disruption of the integrity of either of these catalytic sites substantially impedes restriction of HIV and MLV. Interestingly, our data shows that mA3 induces a significant decrease in retroviral activity of HIV and MLVs by exploiting both deamination-dependent and -independent pathways. However, the deaminase activity of mA3 is essential to confer long-term restriction of retroviral infection. My observations suggest that mA3 has dual activities, both deamination-dependent and -independent, that work cooperatively to restrict a broad range of human and animal retroviral pathogens. In the context of the intrinsic immune system, APOBEC3 proteins provide a powerful block to the transmission of retroviral pathogens that very few have found ways to evade.

Human APOBEC3 proteins are host-encoded intrinsic restriction factors that can prevent the replication of a broad range of human and animal retroviruses such as HIV, SIV, FIV, MLVs and XMRV. The main pathway of the restriction is believed to occur as a result of the cytidine deaminase activity of these proteins that converts cytidines into uridines in single-stranded DNA retroviral replication intermediates. Uridines in these DNA intermediates disrupt the viral replication cycle and also alter retrovirus infectivity because of the C-to-T transition mutations generated as a result of the deaminase activity on the minus strand DNA. In addition, human APOBEC3 proteins also exhibit a deamination-independent pathway to restrict retroviruses that is not currently well understood. Although the restriction of retroviruses by human APOBEC3 proteins has been intensely studied in vitro, our understanding of how the murine APOBEC3 (mA3) protein restricts retroviruses and/or prevents zoonotic infections in vivo is very limited. In contrast to humans and primates that have 7 APOBEC3 genes, mice have but a single copy. My study of the function and structure of mA3 revealed that it has an inverted functional organization for cytidine deamination in comparison to the human A3G catalytic sites. I have also found that disruption of the integrity of either of these catalytic sites substantially impedes restriction of HIV and MLV. Interestingly, our data shows that mA3 induces a significant decrease in retroviral activity of HIV and MLVs by exploiting both deamination-dependent and -independent pathways. However, the deaminase activity of mA3 is essential to confer long-term restriction of retroviral infection. My observations suggest that mA3 has dual activities, both deamination-dependent and -independent, that work cooperatively to restrict a broad range of human and animal retroviral pathogens. In the context of the intrinsic immune system, APOBEC3 proteins provide a powerful block to the transmission of retroviral pathogens that very few have found ways to evade.

No description available.