Computational design of antiviral RNA interference strategies that resist human immunodeficiency virus escape

Abstract

Recently developed antiviral strategies based upon RNA interference (RNAi), which harnesses an innate cellular system for the targeted down-regulation of gene expression, appear highly promising and offer alternative approaches to conventional highly active antiretroviral therapy or efforts to develop an AIDS vaccine. However, RNAi is faced with several challenges that must be overcome to fully realize its promise. Specifically, it degrades target RNA in a highly sequence-specific manner and is thus susceptible to viral mutational escape, and there are also challenges in delivery systems to induce RNAi. To aid in the development of anti-human immunodeficiency virus (anti-HIV) RNAi therapies, we have developed a novel stochastic computational model that simulates in molecular-level detail the propagation of an HIV infection in cells expressing RNAi. The model provides quantitative predictions on how targeting multiple locations in the HIV genome, while keeping the overall RNAi strength constant, significantly improves efficacy. Furthermore, it demonstrates that delivery systems must be highly efficient to preclude leaving reservoirs of unprotected cells where the virus can propagate, mutate, and eventually overwhelm the entire system. It also predicts how therapeutic success depends upon a relationship between RNAi strength and delivery efficiency and uniformity. Finally, targeting an essential viral element, in this case the HIV TAR region, can be highly successful if the RNAi target sequence is correctly selected. In addition to providing specific predictions for how to optimize a clinical therapy, this system may also serve as a future tool for investigating more fundamental questions of viral evolution.

FIG. 1.

Antiviral efficacy of the combination RNAi approach. (A) Evolution of neutral viral genomes simulated under conditions of combinatorial RNAi inhibition. In each series, the overall RNAi cleavage probability on a per-locus basis (P_cleavage) was held constant, while an increasing number of evenly spaced viral sequences was targeted for degradation. In this and all other figures, antiviral efficacy refers to the percentage of trials resulting in an extinction of the viral population. Each data point represents the result of at least 200 independent simulation trials. Symbols for different values of P_cleavage are as follows: 0.9 (black squares); 0.8 (grey triangles); 0.75 (grey circles); 0.7 (white squares); 0.6 (black circles). (B) Distribution of viral escape times for two values of P_cleavage: 0.8 (black) and 0.7 (white). Time until viral escape (in number of simulation cycles) was normalized to the maximum value depicted. Frequency was normalized to the total number of escape events.

FIG. 2.

Nonuniform RNAi: bipartite cell population and neutral viral genomes. (A) Evolution of neutral viral genomes in a cell population consisting of a mixture of cells expressing either uniform levels of siRNA (corresponding to P_cleavage) or no siRNA. In each series, P_cleavage was held constant while the fraction of nonexpressing cells was increased. (B) Data points from panel A plotted against the mean P_cleavage of the population, calculated by multiplying the value of P_cleavage in siRNA-expressing cells by the fraction of cells that expressed siRNA. Symbols for different values of P_cleavage are as follows: 0.95 (white diamonds); 0.9 (black squares); 0.8 (grey triangles); 0.75 (grey circles).

FIG. 3.

Nonuniform RNAi: binomial distribution of siRNA expression and neutral viral genomes. (A) Depiction of binomial distribution, centered about a P_cleavage of 0.8. Shaded region is ±1 standard deviation (σ). (B) Evolution of neutral viral genomes on a cell population with binomially distributed siRNA expression. In each series, mean P_cleavage was held constant while the standard deviation of the distribution (σ) was increased. The right-most data point in each series represents the widest possible distribution for that particular value of mean P_cleavage. Symbols for different values of mean P_cleavage were as follows: 0.9 (black squares); 0.85 (grey diamonds); 0.8 (grey triangles); 0.75 (grey circles); 0.5 (black diamonds).

FIG. 4.

Choice of RNAi target sequence in TAR, as shown in Tat/TAR functional viral genomes reproduced on cell lines that uniformly expressed siRNA against a single region of the TAR RNA. In each series, different 21-bp target sequences were used for a fixed value of P_cleavage, and antiviral efficacy was recorded; the location of the target sequence is reported as the center base of the target. Symbols for different values of mean P_cleavage were as follows: 0.9 (black squares); 0.85 (grey diamonds); 0.8 (grey triangles); 0.78 (grey circles); 0.77 (white squares); 0.76 (black circles).

FIG. 5.

Nonuniform RNAi: bipartite cell population and Tat/TAR functional viral genomes. (A to C) Similar to Fig. 2, Tat/TAR functional genomes replicated on a mixed population of cells expressing either uniform levels of siRNA (corresponding to P_cleavage) or no siRNA were analyzed. In each series, P_cleavage was held constant while the fraction of nonexpressing cells was increased. A highly conserved sequence (bases 18 to 38) (A), an intermediately conserved sequence (bases 31 to 51) (B), and a nonconserved sequence (bases 47 to 67) (C) were used. (D to F) Data points from panels A to C, respectively, plotted against the mean P_cleavage of the population, which was calculated by multiplying the value of P_cleavage in siRNA-expressing cells by the fraction of cells that expressed siRNA. Symbols for different values of mean P_cleavage are as follows: 0.9 (black squares); 0.85 (grey diamonds); 0.8 (grey triangles).

FIG. 6.

Nonuniform RNAi: binomial distribution of siRNA expression and Tat/TAR functional viral genomes. Similar to Fig. 3, Tat/TAR functional genomes replicated on a cell population with binomially distributed siRNA expression. In each series, mean P_cleavage was held constant while the standard deviation of the distribution (σ) was increased. A highly conserved sequence (bases 18 to 38) (A), an intermediately conserved sequence (bases 31 to 51) (B), and a nonconserved sequence (bases 47 to 67) (C) were used. The right-most data point in each series represents the widest possible distribution for that particular value of mean P_cleavage. Symbols for different values of mean P_cleavage are as follows: 0.9 (black squares); 0.85 (grey diamonds); 0.8 (grey triangles); 0.79 (white squares); 0.78 (grey circles); 0.77 (black circles); 0.76 (grey squares); 0.75 (black triangles); 0.74 (white circles).