Several years ago it was shown that very high doses of an entry inhibitor-based microbicide could protect macaques from infection with SHIV162-p3. A paper in J Virology now raises some important caveats about this approach, including the first reported example of a microbicide selecting for drug-resistant virus. The microbicide in question is a CCR5 inhibitor called PSC-RANTES. The challenge virus used in the studies is called SHIV162-p3, a lab-created SIV/HIV hybrid in which an HIV envelope that uses CCR5 to gain entry into CD4 T cells is inserted into an SIV genome instead of the SIV envelope.
The new paper, authored by Dawn Dudley and colleagues from Case Western Reserve University and the Tulane Primate Research Center, is based on an analysis of macaques given lower doses of PSC-RANTES that were not protective against SHIV162-p3 infection. In most cases, viruses in these animals showed mutations that were deemed unrelated to PSC-RANTES because similar mutations were seen in controls that were challenged with the same virus in the absence of the microbicide. But the virus in one macaque contained two mutations, K315R in gp120 and N640D in gp41, which were very rare in the challenge virus stock (after infection, compared to their frequency in the challenge virus, these mutations were increased at least 25-fold and 75-fold). Importantly, PSC-RANTES failed to completely inhibit this mutant SHIV, even at high concentrations that inhibited the parent SHIV162-p3 and all other R5-using HIV isolates tested.
Dudley and colleagues acknowledge that this apparent example of resistance was only seen in 1/25 macaques studied, and there is an outlying possibility that the mutations represent a random adaptation of the virus to better replicate in macaque cells. However, they argue that the weight of evidence supports their conclusion that the mutations arose from drug selection pressure, and that this possibility is particularly concerning given that SHIV162-p3 is a very homogenous virus stock from which the emergence of resistance would have been considered unlikely.
Discussing the implications for microbicides generally, they note that the positive results reported to date with entry inhibitor-based microbicides (PSC-RANTES, BMS-378806, CMPD167, and C52L) have all involved administering relatively high doses and the SHIV162-p3 challenge virus (which “is exquisitely sensitive to many entry inhibitors”). As a result, they believe that “the barrier for selecting resistance to an anti-HIV microbicide (such as PSC-RANTES) has been set very high and possibly beyond physiological relevance.” To address these concerns, the authors recommend the development of macaque models that use multiple R5-SHIVs as a challenge in order to better duplicate the diversity of viruses to which individuals are typically exposed. Because the resistance described in their study was associated with a dose of PSC-RANTES lower than that which conferred full protection, they also caution that waning microbicide levels after application could conceivably provide the conditions necessary for the emergence of drug resistance.
JVI Accepts, published online ahead of print on 11 March
2009
J. Virol. doi:10.1128/JVI.00055-09
Selection of SHIV resistant to a vaginal microbicide in macaques
Dawn M. Dudley, Jennifer L. Wentzel, Matthew S. Lalonde, Ronald S. Veazey, and Eric J. Arts
Department of Molecular Biology and Microbiology, Division of Infectious Diseases, Department of Medicine, Department of Biochemistry, Case Western Reserve University, Cleveland, OH, 44106; Department of Pathology, Tulane National Primate Research Center, Tulane University Health Sciences Center, Covington, LA
Abstract
PSC-RANTES binds to CCR5, inhibits HIV-1 entry, and has been shown to protect rhesus macaques as a vaginal microbicide from SHIVSF162-p3 infection in a dose-dependent manner. In this study, env gene sequences from SHIVSF162-p3-infected rhesus macaques treated with PSC-RANTES were analyzed for possible drug escape variants. Two specific mutations located in the V3 region of gp120 (K315R) and C-helical domain of gp41 (N640D) were identified in a macaque (m584) pre-treated with a 100 µM dose of PSC-RANTES. These two env mutations were found throughout infection (through week 77) but were only found at low frequencies in the inoculating SHIVSF162-p3 stock and in the other SHIVSF162-p3-infected macaques. HIV-1 env genes from the m584 macaque and from inoculating SHIVSF162-p3 were cloned into an HIV-1 backbone. Increases in IC50 values to PSC-RANTES with envm584 was modest (7-fold) and most pronounced in cells expressing rhesus macaque as compared to human CCR5. Nonetheless, virus harboring the envm584, unlike inoculating envp3, could replicate even at the highest tissue culture PSC-RANTES concentrations (100 nM). Dual virus competitions revealed a dramatic increase in fitness of chimeric virus containing the envm584 (K315R/N640D) over that containing the envp3, but again, only in rhesus CCR5-expressing cells. This study is the first to describe the immediate selection and infection of a drug-resistant SHIV variant in the face of a protective vaginal microbicide, PSC-RANTES. This rhesus CCR5-specific/PSC-RANTES-resistance selection is particularly alarming given the relative homogeneity of the SHIVSF162-p3 stock as compared to the potential exposure to a heterogeneous HIV-1 population in human transmission.
Comments