Regulation of HIV-1 replication by CpG dinucleotides, ZAP and KHNYN

Student thesis: Doctoral ThesisDoctor of Philosophy


The functional relevance of RNA sequence biases in the human immunodeficiency virus type 1 (HIV-1) genome remains poorly understood and is necessary for a complete understanding of the viral lifecycle. CpG dinucleotides are suppressed in many vertebrate RNA viruses, including HIV-1, and introducing CpGs into RNA virus genomes inhibits their replication. This could be a result of altered RNA structure, splice sites, or RNA binding protein interactions. The zinc finger antiviral protein (ZAP) binds viral RNA with increased CpG abundance and is necessary for CpGs to inhibit viral replication in some contexts. However, it remains unknown how the number and position of CpG dinucleotides in viral genomes affect restriction by ZAP and whether ZAP-mediated restriction is the only mechanism driving CpG dinucleotide suppression in viral genomes.

To gain a better understanding of how CpG dinucleotides affect HIV-1 replication, we synonymously increased their abundance in multiple regions of the viral genome and analysed the effect on RNA expression, protein abundance, and infectious virus production. We found that the antiviral effect of CpGs was not correlated with their abundance and that CpGs inserted into some regions of the viral genome sensitize the virus to ZAP antiviral activity more efficiently than insertions into other regions. Increasing CpG dinucleotide abundance in the 5’ region of the viral gene env sensitized the virus to endogenous levels of ZAP more efficiently than increased CpG content in either gag or pol, with this sensitivity modulated by interferon treatment or ZAP overexpression. In addition, we demonstrated that, in some contexts, CpGs can inhibit HIV-1 replication through ZAP-independent mechanisms. One of these is the activation of a cryptic splice site at the expense of a canonical splice site when synonymous mutations are introduced into the 5’ region of gag. Overall, we demonstrate that the location and sequence context of CpG dinucleotides in the viral genome determine the magnitude and mechanism of their antiviral activity.

ZAP has no known enzymatic activity and must interact with other cellular proteins to inhibit viral replication. We conducted a yeast-2-hybrid screen for ZAP-interacting proteins and identified KHNYN, a protein with no previously known function. KHNYN localizes to the cytoplasm, and co-immunoprecipitation assays confirmed that it has an RNase-resistant interaction with ZAP. KHNYN overexpression selectively inhibits HIV-1 containing increased CpG content in env by decreasing genomic RNA abundance, Gag expression, and infectious virus production. KHNYN requires ZAP and the ZAP cofactor Tripartite motif-containing protein 25 (TRIM25) for its antiviral activity. Crucially, depletion of KHNYN eliminated the deleterious effect of CpG dinucleotides in the 5’ region of env on HIV-1 RNA abundance and infectious virus production and enhanced murine leukemia virus (MLV) viral protein expression, phenocopying ZAP knockout cells.

While KHNYN protein structure has not been experimentally determined, it is predicted to contain a putative RNA binding extended di-K Homology (KH)-like domain and an NYN endonuclease domain. Deleting the extended di-KH-like domain reduced KHNYN antiviral function, and specific point mutations previously shown to inhibit NYN domain nuclease activity abolished its ability to inhibit infectious virus production. This suggests that CpG containing viral RNA is cleaved by the NYN endonuclease domain in KHNYN and that the extended di-KH-like domain is necessary for KHNYN antiviral activity. Our work identifies KHNYN as a novel ZAP cofactor essential for the innate immune restriction of CpG containing viral RNA.

In sum, we have used HIV-1 as a model system to conduct a detailed analysis of the mechanisms underlying CpG dinucleotide inhibition of viral replication and identified KHNYN as a novel ZAP cofactor required for the degradation of CpG-containing retroviral RNA. Beyond allowing for a better understanding of viral biology, elucidating the mechanism and cellular factors involved in the inhibition of viral replication by CpG dinucleotides could allow for the creation of viruses useful for human health, such as the development of novel, live attenuated viral vaccines via synonymous genome modification. 
Date of Award1 Jan 2022
Original languageEnglish
Awarding Institution
  • King's College London
SupervisorChad Swanson (Supervisor) & Charlotte Odendall (Supervisor)

Cite this