A decade of viral mutations and associated drug resistance in a population of HIV-1+ Puerto Ricans: 2002-2011

Abstract

Puerto Rico has one of the highest rates of HIV/AIDS seen for any US state or territory, and antiretroviral therapy has been a mainstay of efforts to mitigate the HIV/AIDS public health burden on the island. We studied the evolutionary dynamics of HIV-1 mutation and antiretroviral drug resistance in Puerto Rico by monitoring the population frequency of resistance-associated mutations from 2002 to 2011. Whole blood samples from 4,475 patients were analyzed using the TRUGENE HIV-1 Genotyping Kit and OpenGene DNA Sequencing System in the Immunoretrovirus Research Laboratory at Universidad Central del Caribe. Results show that 64.0% of female and 62.9% of male patients had HIV-1 mutations that confer resistance to at least one antiretroviral medication. L63P and M184V were the dominant mutations observed for the protease (PRO) and reverse transcriptase (RT) encoding genes, respectively. Specific resistance mutations, along with their associated drug resistance profiles, can be seen to form temporal clusters that reveal a steadily changing landscape of resistance trends over time. Both women and men showed resistance mutations for an average of 4.8 drugs over the 10-year period, further underscoring the strong selective pressure exerted by antiretrovirals along with the rapid adaptive response of HIV. Nevertheless, both female and male patients showed a precipitous decrease for overall drug resistance, and for PRO mutations in particular, over the entire course of the study, with the most rapid decrease in frequency seen after 2006. The reduced HIV-1 mutation and drug resistance trends that we observed are consistent with previous reports from multi-year studies conducted around the world. Reduced resistance can be attributed to the use of more efficacious antiretroviral drug therapy, including the introduction of multi-drug combination therapies, which limited the ability of the virus to mount rapid adaptive responses to antiretroviral selection pressure.

Fig 1. Cohort sample analysis and HIV-1 genotyping workflow used for the study.

Fig 2. Observed levels of antiretroviral resistance-associated mutations in Puerto Rico from 2002 to 2011.

(A) Drug resistance and mutation frequencies recorded by gender. (B) Percentages of individuals seen for the most commonly observed antiretroviral drug resistance and the most frequent mutations in the protease (PRO) and reverse transcriptase (RT) genes.

Fig 3. Trends in HIV-1 drug resistance levels in Puerto Rico from 2002 to 2011.

The names of the antiretroviral drugs are color-coded with respect to their class of action as shown in the key in the middle of the figure. (A) Normalized drug resistance frequencies, color-coded as shown in the key, are shown for the drugs analyzed here over the 2002–2011 period. White boxes indicate an absence of resistance mutations for a drug in any given year. Normalized drug resistance fraction profiles are grouped using hierarchical clustering, revealing two clusters with distinct peak resistance periods. (B) The maximum observed resistance fractions are shown for all drugs alongside (C) box-plot distributions of the maximum observed resistance fractions.

Fig 4. Trends in HIV-1 protease mutations in Puerto Rico from 2002 to 2011.

(A) Normalized protease (PRO) mutation frequencies, color-coded as shown in the key, are shown for the 2002–2011 period. White boxes indicate an absence of PRO resistance mutations in any given year. PRO mutation frequency profiles are grouped using hierarchical clustering, revealing five distinct clusters. Mutation names are shown adjacent to the clustering plot. (B) Box-plot distributions of the normalized PRO mutation frequencies. Box-plots are color-coded using a rainbow scheme to visually distinguish the five PRO mutation frequency clusters.

Fig 5. Trends in HIV-1 reverse transcriptase mutations in Puerto Rico from 2002 to 2011.

(A) Normalized reverse transcriptase (RT) mutation frequencies, color-coded as shown in the key, are shown for the 2002–2011 period. White boxes indicate an absence of RT resistance mutations in any given year. RT mutation frequency profiles are grouped using hierarchical clustering, revealing five distinct clusters. Mutation names are shown adjacent to the clustering plot. (B) Box-plot distributions of the normalized RT mutation frequencies. Box-plots are color-coded using a rainbow scheme to visually distinguish the five RT mutation frequency clusters.

Fig 6. Average number of drug resistance or mutations per individual during the 2002 to 2011 period.

Values were computed for all individuals (black-solid), females (blue-dotted), and males (red-dashed).