Science Spotlight

Fight fire with fire, viral delivery of anti-HIV gene therapy

Scheme for AAV mediated gene delivery. Both cellular penetration and gene expression limit efficiency of gene therapy.
Adeno-associated virus (AAV) vectors carrying DNA encoding GFP were used to test the gene delivery capabilities of this virus. T cells were exposed to AAV, and first through PCR researchers measured rate of AAV delivery of GFP. Second, successful expression of the gene product was determined by flow cytometry. Both steps limited this technology and were used to generate a model of viral delivery pharmacodynamics.
Figure provided by Dr. Pavitra Roychoudhury

Research over the last three decades has yielded major advances in treating and preventing HIV infections; however, a cure has remained a distant goal. To clear viral infections from a body the host immune system must destroy circulating virus, and purge viral components from infected cells - often by killing those cells. If infected cells are not removed they remain a viral reservoir causing chronic infection such as with HIV. Natural clearance of viral reservoirs is often accomplished by degrading the viral genome, but HIV incorporates its genome into host chromosomes making this impossible without cell death. One key component for an HIV cure is genetic engineering.

As gene editing technologies mature the cure seems less distant. Currently, at least four different enzymatic mechanisms have been engineered to target and destroy single protein coding sequences in the human genome. All of these approaches require the delivery of large enzymes to the infected cells in patients. To accomplish this, ironically, viruses like HIV have been engineered to deliver DNA encoding gene therapies to infected cells. In a recent publication in the Journal of Antimicrobial Chemotherapy researchers in the Schiffer and Jerome Labs (Vaccine and Infectious Disease and Clinical Research Divisions) explore the efficiency of delivering gene therapy using adeno-associated virus (AAV) vectors.

HIV infects cells expressing the CD4 cell surface marker (primarily T cells) making these the target for a gene-editing cure, which limits the delivery mechanism as Dr. Daniel Stone explained, "We chose AAV primarily because it is able to readily infect our target cell population of CD4+ T cells. In contrast some viral vectors, such as MLV based retroviral vectors, cannot infect non-dividing cells so would be of limited use for a HIV therapy targeting latently infected quiescent T cells."

For successful therapy, AAV must enter the T cell, and then express the gene-editing enzyme it carries in its engineered genome. These features are cell type specific and largely determined by the serotype of AAV and the promoter region. Researchers compared the ability of eight different AAV serotypes and eight unique promoter sequences to deliver and express GFP rather than a gene-editing enzyme. Three of the serotypes and four promoters performed best, however successful GFP expression was inefficient. At doses where a single T cell was exposed to 100,000 AAV particles that cell only expressed two GFP genes successfully due to restrictions in both AAV delivery and GFP expression. Quantitative PCR analysis revealed that the first bottleneck was AAV delivery rates, ~1 in 220 viral particles were taken up by T cells, but AAV delivery did not guarantee GFP expression as 1 in 15 to 1 in 1500 of AAVs taken up by the cell resulted in GFP expression.

Using these experimental results the researchers generated a model to determine the treatment required for an HIV cure. The model is designed to predict how weekly infusions with a gene therapy would reduce a baseline infection of 10 million T cells. This is dependent on the potency of the gene-editing enzyme and how well it is expressed in T cells. For example, this group extrapolated from previous findings that 4-9 copies of a gene-editing enzyme (zinc finger nuclease) targeting the HIV pol gene must be expressed per cell to affect 50% HIV genomes inserted in the T cell population – this is the potency. For expression, this study demonstrated the best AAV/promoter combo produced an average expression of two to three gene copies per cell. Using these data the model shows it would require 6 months to deplete the infected cell population 10,000 fold and a complete cure would be achieved after a year. This lengthy treatment would also require up to 4 x 1014 viral particles for each transfusion.

While laboratory use of these gene-editing technologies is becoming commonplace, this work emphasizes how vast the translational space remains. Both potency and delivery of these therapies would benefit from improvements. In another delivery approach for gene therapy T cells are isolated from patients first and then genetically modified ex vivo before being infused back into patients. Dr. Stone explains that this may have a role in the HIV cure, but cannot be the sole approach. "It is very likely that ex vivo T cell transduction would enable higher levels of gene delivery to T cells than in vivo gene delivery, however, HIV infected T cells reside in many tissues beyond the blood, so a therapy that only transduces T cells isolated from blood would not reach many of the latently infected target cells. In theory in vivo therapeutic delivery could enable transduction of HIV infected cell types in hard to reach places, such as macrophages within the brain."

These findings are directing researchers towards new important questions, yet a model is always limited by the input data. The next steps in this work are to model viral delivery using in vivo data from murine studies, "Next, we plan to model endonuclease treatment efficacy in a humanized mouse model of HIV infection. We are also interested in modeling the pharmacokinetics of endonuclease-based approaches in vitro and in vivo to look at the time course from absorption to clearance." said Dr. Pavitra Roychoudhury.


Roychoudhury P, De Silva Feelixge HS, Pietz HL, Stone D, Jerome KR, Schiffer JT. (2016). Pharmacodynamics of anti-HIV gene therapy using viral vectors and targeted endonucleases. J Antimicrob Chemother. 2016. Epub ahead of print.

Funding for this research was provided by the National Institutes of Health, and the University of Washington Center for AIDS Research.