Sequence Notes: Near Full-Length Genome Analysis of the First-Reported HIV-1 Circulating Recombinant Form (CRF)_10CD in Uganda

Abstract

HIV-1 is characterized by remarkable genetic diversity resulting from its high replication rate, error-prone reverse transcriptase enzyme and recombination events. In Uganda, HIV-1 subtype diversity is mostly dominated by subtypes A, D, and A1/D Unique Recombinant Forms (URFs). In this study, deep sequences of HIV from patients with known antiretroviral therapy (ART) status were analyzed to determine the subtypes and to identify drug-resistance mutations circulating in the study population. Of the 187 participant samples processed for next-generation sequencing (NGS), 137 (73%) were successfully classified. The majority of HIV-1 strains were classified as subtype A (75, 55%), D (43, 31%), with other subtypes including C (3, 2%), A1/D (9, 7%) and CRF10_CD (1, <1%). Recombinant analysis of nine complete A1/D HIV genomes identified novel recombination patterns described herein. Furthermore, we report for the first time in Uganda, an HIV-1 CRF10_CD strain from a fisherfolk in a Lake Victoria Island fishing community.

Recent UNAIDS data has shown a decrease in the overall number of new HIV infections in Uganda since 2010, however, infections continue to be elevated in key populations. ¹ Results of the Uganda Population-based HIV Impact Assessment (UPHIA 2020–2021) showed that the prevalence of HIV among adults of age 15 years and older in Uganda was 5.8%, with 80.9% of adults living with HIV being diagnosed, 96.1% on antiretroviral therapy (ART), and 92.2% of ART patients achieving viral load suppression (VLS). ² Key populations in Uganda include people living in the fishing communities (generally referred to as fisherfolk), sex workers, or women at high risk of HIV infection, long distance truckers, men who have sex with men, people in uniformed forces and other high-risk groups.

The HIV subtype distribution in Uganda has historically been dominated by subtype A and D with reports of an increase in the prevalence of Unique Recombinant Forms (URFs). ³ However, no Circulating Recombinant Forms (CRFs) have been reported. A study done in Rakai (Southwest Uganda) identified a potentially new D-A-D CRF. ⁴ Monitoring URFs and CRFs is important for HIV surveillance as recombination may impact on prevention efforts and interfere with the diagnosis and treatment of HIV-1 infection. ⁵ Recombination is key to viral evolution and contributes towards HIV diversity, fitness, drug resistance, immunological escape, and disease progression. ⁵ By definition, an HIV recombinant is a genetic sequence that carries regions from two genetically distinct parental subtypes. ⁶ An important difference between CRFs and URFs is that CRFs have the same recombination breakpoints in the genome while URFs show unique breakpoints with no evidence of onward transmission in the population.

Between 2016–2021, samples were collected from 541 study participants including those from two key populations groups (fisherfolk and sex workers) in the Lake Victoria fishing communities of Kalangala. This study was part of the MRC HIV-1 Molecular Epidemiology Study approved by the Uganda Virus Research Institute Research and Ethics Committee (UVRI-REC) (Federal Wide Assurance (FWA) No. 00001354, Project Identification Code: GC/127/13/06/27) and the Uganda National Council for Science and Technology (UNCST) (FWA No. 00001293, Project Identification Code: HS 1432). All participants were recruited voluntarily and provided written informed consent. The aim of the HIV-1 Molecular Epidemiology Study was to determine HIV-1 subtypes and transmission linkages among both high-risk and general populations in Uganda. The study cohort was 56% female (age 17–69 years, mean 33 years) and 44% male (age 18–60 years, mean 37 years) with 79% of participants being treatment-naïve and 21% on ART. Viral Load Suppression (VLS) defined as HIV RNA < 1,000 copies/mL was achieved in 87% patients on ART. Of 187 samples processed for next-generation sequencing NGS (182 untreated and 5 on ART), 137 (73%) were successfully classified and of these 48, 60, and 28 had genome coverage of 15–50%, 51–89%, and 90–99%, respectively. Viral load (VL) was determined with the RealTime HIV-1 (Abbott Molecular Inc., Des Plaines, IL) assay and specimens with VL > 3.0 log copies/mL were selected for NGS. HIV nucleic acids were extracted and processed using Nextera reagents, XGen HIV custom probe enrichment ⁷ and the NextSeq sequencing platform (Illumina, San Diego, CA). Consensus genomes were generated using BCFtools, ⁸ and preliminary phylogenetic analysis was done using neighbor-joining phylogenetic trees to determine subtypes. In addition, 14 consensus sequences with genome coverage >90% initially classified as pure subtypes (4), and possible recombinants (10) were aligned with reference sequences using MegAlign Pro 17 (DNASTAR, Inc., Madison, WI), and phylogenetic analysis was performed with the PHYLIP software package (J. Felsenstein, University of Washington, Seattle, WA). The Simplot program ⁹ was used to determine recombination breakpoints and confirm the CRF or URF sequences. HIV drug resistance mutations (DRM) were assessed using the Stanford HIV drug resistance database. ¹⁰

The majority of HIV infections were classified as subtype A1 (62, 45%) and D (43, 31%), with other subtypes also present at lower levels, including C (3, 2%), A1/D (9, 7%), and CRF10_CD (1, <1%). Protease-RT and integrase sequences were obtained from 71 treatment-naïve subjects and from three patients on ART. DRM were identified in 21 (30%) of the treatment-naïve subjects and in all 3 (100%) patients on ART with rtK130N as the most frequently observed DRM (23%). The data also indicate a high level of circulating, transmitted DRMs (30%) in treatment-naïve patients. HIV subtyping analysis of HIV near full-length sequences sampled from key populations in Kalangala fishing communities showed that subtypes A1 and D were the major HIV subtypes in circulation. Notably, nine genomes branched basal to A and D clades (Fig. 1A), and recombinant analysis confirmed that these were all A1/D URFs (Fig. 2). Furthermore, we identified for the first time in Uganda a CRF10_CD by phylogenetic (Fig. 1A) and recombinant breakpoint analysis (SimPlot and Bootscan) (Fig. 1B–C). The CRF10_CD study participant was a 33-year-old female who was living in Kalangala Islands. The participant was married and aware of her HIV-positive status but was not linked to care due to HIV-related stigma. She reported having multiple sexual partners and had a VL of 124,965 copies/mL at the time of sample collection. Follow-up molecular surveillance will aim to determine whether the identified CRF has been inwardly transmitted in the study population or whether this was an imported case since Kalangala islands attracts fishermen and other fisherfolk from both Uganda and neighboring Tanzania where CRF10_CD has been reported. ¹¹

FIG. 1.

(A) Neighbor-joining tree shows the phylogenetic relationship of the 14 near-full-length genome nucleotide sequences (shown in red) to HIV group M reference strains (in black). Subtype designation for the reference strains is indicated near the appropriate branch. Bootstrap values supporting monophyletic grouping of strains are provided at the base of the branch. ME-URM-1421495 clustered with other CRF10_CD sequences from Tanzania with high bootstrap support. The nine suspected AD recombinant sequences formed two groups of branches near pure subtype A and D strains. (B) SimPlot and bootscan for reference CRF10_BF061 (upper panel) and ME-URM-1421495 (bottom panel) generated using SimPlot software with a window of 500 bp advancing in 20 bp increments along the gap-stripped alignment with non-recombinant subtype A1, C, D, and SIV reference strains are shown in the legend. Positions in the alignment are shown on the x-axis, percentages of permutated trees (bootstrap values) are shown on the y-axis. ME-URM-1421495 has breakpoints across the genome similar to the reference sequence CRF10_CD. (C) Diagrammatic representation of the ME-URM-1421495 genome map shows the recombination pattern. The numbers above the map indicate the nucleotide positions of the recombination breakpoints in the generated sequence.

FIG. 2.

Intersubtype mosaic patterns of the nine A1/D near-full-length genomes. Breakpoints within genomes are shown based on the nucleotide positions in the gap-stripped alignment. Breakpoints were determined using SimPlot and Bootscan analysis. Subtypes A1, D, and unclassified (U) are color-coded.