Strategies for Targeting Retroviral Integration for Safer Gene Therapy: Advances and Challenges

Abstract

Retroviruses are obligate intracellular parasites that must integrate a copy of the viral genome into the host DNA. The integration reaction is performed by the viral enzyme integrase in complex with the two ends of the viral cDNA genome and yields an integrated provirus. Retroviral vector particles are attractive gene therapy delivery tools due to their stable integration. However, some retroviral integration events may dysregulate host oncogenes leading to cancer in gene therapy patients. Multiple strategies to target retroviral integration, particularly to genetic safe harbors, have been tested with limited success. Attempts to target integration may be limited by the multimerization of integrase or the presence of host co-factors for integration. Several retroviral integration complexes have evolved a mechanism of tethering to chromatin via a host protein. Integration host co-factors bind chromatin, anchoring the complex and allowing integration. The tethering factor allows for both close proximity to the target DNA and specificity of targeting. Each retrovirus appears to have distinct preferences for DNA sequence and chromatin features at the integration site. Tethering factors determine the preference for chromatin features, but do not affect the subtle sequence preference at the integration site. The sequence preference is likely intrinsic to the integrase protein. New developments may uncouple the requirement for a tethering factor and increase the ability to redirect retroviral integration.

Keywords: HIV-1; LEDGF/p75; MLV; gene therapy; retrovirus; targeted integration.

FIGURE 1

Retroviral life cycle and retroviral vector particle transduction. (Left) Retroviruses may enter a target cell by membrane fusion. The capsid core is released to the cytoplasm. Reverse transcription copies the viral genomic RNA (black lines) to a linear double stranded cDNA. Integrase binds the ends of the viral cDNA forming a pre-integration complex (PIC). Lentiviral PICs are able to cross an intact nuclear membrane while all other retroviruses require cellular division to access the host genome. Integrase mediates the stable integration of the vDNA (black) to the host genome (blue) generating the provirus. Host transcription machinery generates viral mRNAs and genomic RNA. Progeny viral particles assemble and are released from the plasma membrane. Following budding from the cell, viral enzyme protease cleaves the polyproteins to generate a mature infectious virus particle. (Right) Retroviral vector particles recapitulate the early steps of the retroviral life cycle. Viral RNA and cDNA depicted in red. However, they do not encode viral proteins. Only the protein of interest is expressed.

FIGURE 2

Timeline of developments in retroviral gene therapy vectors and targeted integration.

FIGURE 3

HIV-1 proviral genome and lentiviral vector genome. (Top) The HIV-1 proviral genome (HIV Genome) has long terminal repeats (LTRs) at each end. These non-coding sequences include the terminal sequences that are bound by integrase. The LTRs also encode sequences necessary for viral gene expression including transcription factor binding sites and a TATA box to initiate RNA Pol II transcription. Every retrovirus includes gag, pol, and env genes. These genes encode the structural, enzymatic, and envelope proteins, respectively. HIV-1 also has six accessory genes. Two of these genes, tat and rev, are spliced. (Bottom) A representative lentiviral vector (WW Vector) for treatment of Wiskott-Aldrich syndrome (WAS) encodes the WAS protein (WASP) gene driven by the human WASP promoter (hWASP) (Aiuti et al., 2013). The post-transcription regulatory element (PRE) mediates export of unspliced mRNA from the nucleus to the cytoplasm for translation (Zufferey et al., 1999). Much of the LTR sequences have been deleted, including transcription factor binding sites, yielding a self-inactivating (SIN) vector.

FIGURE 4

Domains of viral integrases and cofactors. (A) Domains of murine leukemia virus (MLV), prototype foamy virus (PFV) and human immunodeficiency virus (HIV-1) INs. These domains are the N-terminal extension domain (NED), the N-terminal domain (NTD), the catalytic core domain (CCD) and the C-terminal domain (CTD). The numbers along each line represent amino acid residues. (B) Domains of integration host cofactors LEDGF/p75 and Brd4. LEDGF/p75 includes a chromatin binding domain (PWWP) followed by three charged regions (CR1-CR3). Between CR1 and CR2 is a nuclear localization signal (NLS) and two AT-hooks which allow for DNA binding of AT-rich motifs. Near the C-terminus is the integrase binding domain (IBD). Brd4 contains two bromodomains (BD1, BD2) as well as 2 DNA binding motifs, A and B, which in conjunction bind chromatin. There is also a protein interaction extra terminal domain (ET), a serine-rich SEED domain, and a C-terminal motif (CTM).

FIGURE 5

Model of LEDGF/p75 tethering an HIV-1 intasome. A mononucleosome (green) wrapped in DNA (black line) representing a nucleosome. The PWWP domain of LEDGF/p75 (cyan) binds the post translational modification H3K36me3. The integrase binding domain (IBD) of LEDGF/p75 is bound to the HIV-1 intasome (red) which is shown as a tetramer for simplicity. The viral DNA genome is shown in black and the dashed lines represent the viral DNA within the intasome.