Deep Gene Sequence Cluster Analyses of Multi-Virus-Infected Mucosal Tissue Reveal Enhanced Transmission of Acute HIV-1

Abstract

Exposure of the genital mucosa to a genetically diverse viral swarm from the donor HIV-1 can result in breakthrough and systemic infection by a single transmitted/founder (TF) virus in the recipient. The highly diverse HIV-1 envelope (Env) in this inoculating viral swarm may have a critical role in transmission and subsequent immune response. Thus, chronic (Env_chronic) and acute (Env_acute) Env chimeric HIV-1 were tested using multivirus competition assays in human mucosal penile and cervical tissues. Viral competition analysis revealed that Env_chronic viruses resided and replicated mainly in the tissue, while Env_acute viruses penetrated the human tissue and established infection of CD4⁺ T cells more efficiently. Analysis of the replication fitness, as tested in peripheral blood mononuclear cells (PBMCs), showed similar replication fitness of Env_acute and Env_chronic viruses, which did not correlate with transmission fitness in penile tissue. Further, we observed that chimeric Env viruses with higher replication in genital mucosal tissue (chronic Env viruses) had higher binding affinity to C-type lectins. Data presented herein suggest that the inoculating HIV-1 may be sequestered in the genital mucosal tissue (represented by chronic Env HIV-1) but that a single HIV-1 clone (e.g., acute Env HIV-1) can escape this trapped replication for systemic infection.IMPORTANCE During heterosexual HIV-1 transmission, a genetic bottleneck occurs in the newly infected individual as the virus passes from the mucosa, leading to systemic infection with a single transmitted HIV-1 clone in the recipient. This bottleneck in the recipient has just been described (K. Klein et al., PLoS Pathog 14:e1006754, https://doi.org/10.1371/journal.ppat.1006754), and the mechanisms involved in this selection process have not been elucidated. However, understanding mucosal restriction is of the utmost importance for understanding dynamics of infections and for designing focused vaccines. Using our human penile and cervical mucosal tissue models for mixed HIV infections, we provide evidence that HIV-1 from acute/early infection, compared to that from chronic infection, can more efficiently traverse the mucosal epithelium and be transmitted to T cells, suggesting higher transmission fitness. This study focused on the role of the HIV-1 envelope in transmission and provides strong evidence that HIV transmission may involve breaking the mucosal lectin trap.

Keywords: HIV; fitness; transmission.

FIG 1

Schematic of transmission fitness assay and analyses of multivirus infections. (a) Human explant tissue was exposed to a mixture of Env_acute and Env_chronic viruses for 3 h, washed, and cultured overnight. Cells migrating out of the tissue were collected and cocultured with PM1 CD4⁺ T cells. After10 days, both tissue and MC+PM1 cocultures were lysed, and DNA was extracted for PCR amplification of the C2-V3 envelope region. After ligation of barcoded oligonucleotides, purification, and quantification, equivalent amounts of products were subject to NGS. Data analysis was performed using SeekDeep, a custom pipeline sorting sequences based on barcodes and aligning sequences within barcode bins. (b) Results of multivirus infections are presented as percent sequence reads of each virus detected in the assay in tissue or MC+PM1 cocultures. Gray bars indicate the percentage of sequences from that barcoded PCR product (representing this multivirus competition) mapped to one of the seven viruses in the competitions. It is important to note that these analyses involve counting HIV isolates based on phylogenetic clustering, where low-frequency mutations do not impact the analyses. This was verified as shown in panel c. The relative amount of each virus in the competition was determined using maximum-likelihood trees of the HIV-1 C2-V3 env DNA amplified from tissue or MC+PM1 cocultures. Env sequence in the maximum likelihood tree of panel c in the bold and larger colored text and the black text represent the C3-V3 gp120 sequences of all the chimeric Env HIV-1 strains that were employed in the various multi-virus competitions used in this study (see below). For the representative multi-virus competition results shown in this figure, the Env chimeric viruses used in this competition are the branch points on the tree labeled with the colored text. The small text labels (text not important) show the individual branches/sequences defined as NGS reads "counted" from the multivirus competitions by SeekDeep. The reference sequences for the viruses in this competition are shown in italic, bold, large text, namely B1-1054.TC4.1499 (red), B2-6244 (yellow), B3-PRB926 (dark blue), B4-WEAUd15.410.5017 (light blue), B17-REJO.D12.1972 (purple), BB19-700010040.C9.4520 (brown), and B20-63358.p3.4013 (orange on the tree but pink for rest of figure). A few recombinant B19/B4 sequences were identified in this competition and were identified as a mixed brown/light blue text label. Of all HIV sequences in the maximum-likelihood tree (c) from the matched-input viruses (B1, B2, B3, B4, B19, and B20), only B1, B4, and B19 replicated in the tissue (see replicate 1 of the tissue in panel b). Less than 1%, or 4 sequences, were recombinants of B4 and B19 with different breakpoints.

FIG 2

Characterization of the migratory cell phenotype in penile and cervical tissue. The percentages of CD1c-, CD83-, CD207-, and CD209-positive cells of all viable mDCs isolated from penile (n = 5) and cervical (n = 3) tissue were assessed by a multicolor flow panel on an LSRIIFortessa flow cytometer. Box-and-whisker plots show median, minimum/maximum, and 25th/75th percentiles of the percentage of viable cells stratified by CD1c, CD83, CD207, and CD209 expression. Each symbol represents an individual penile or cervical tissue donor.

FIG 3

Higher transmission fitness of Env_acute versus Env_chronic viruses in multivirus competition assays in human tissue. (a) Penile tissues were dissected into 3-mm³ explants and exposed to eight different mixtures of a panel of Env_acute viruses and the same three Env_chronic viruses. (b) Following overnight culture, migrating cells from the tissue were collected and cocultured with PM1 T cells (MC+PM1). After 10 days in culture, DNA from tissue and MC+PM1 cells was isolated, PCR amplified, and analyzed by NGS. The cumulative percent replication in penile tissue (a) and in MC+PM1 cocultures (b) of Env_chronic and Env_acute viruses for tissue A and tissue B in each of the eight multivirus panels is shown. (c) The combined average replication of all Env_chronic and all Env_acute viruses of the eight virus competitions in penile tissues and Env_chronic (green bars) and Env_acute (red bars) viruses in MC+PM1 cocultures was determined. Each competition was done in triplicates in tissues from two donors. (d) Average replication of Env_chronic (bars with green outlines) and Env_acute (bars with red outlines) viruses of one competition set in cervical tissue and the corresponding MC+PM1 cocultures were evaluated by NGS. Statistical significance was determined using the Holm-Sidak method (a and b) (*, P < 0.05) and a two-tailed Mann-Whitney test (c and d) (***, P < 0.001; ****, P < 0.0001).

FIG 4

Multivirus competition assays to assess the transmission fitness across penile and cervical tissue. Penile and cervical explants were exposed to mixtures of three Env_chronic viruses (I10, K44, and Q0) and three to five Env_acute viruses (B1, B2, B3, B4, B7, B8, B9, B14, B17, B19, and B20) for 3 h. After overnight culture, migrating cells from the tissue were collected and cocultured with CD4⁺ PM1 T cells (MC+PM1). Following 10 days in culture, both tissue and MC+PM1 cocultures were lysed and PCR amplified, and replication of viruses was analyzed by NGS. (a to p) Percent replication of eight different mixtures of Env_chronic and Env_acute viruses in tissues from donor A and donor B and Env_chronic and Env_acute viruses in the corresponding MC+PM1 cocultures. Each graph shows the percentage of sequence reads for donor A and donor B from the NGS analysis mapped to the viruses present in the competitions. (q and r) Side-by-side comparison of Env_chronic and Env_acute virus replication in penile tissue and Env_chronic and Env_acute virus replication in cervical tissue (q) and comparison of transmission fitness of Env_chronic and Env_acute viruses in MC+PM1 derived from penile tissue with transmission fitness of Env_chronic and Env_acute viruses in MC+PM1 from cervical tissue (r) of one competition set (I10, K44, and Q0 and B1, B3, B4, B19, and B20).

FIG 5

Replication fitness does not predict transmission fitness. Multivirus infection assays in penile tissue were performed, and the average replication of each Env_chronic and each Env_acute virus in tissue (a) and in MC+PM1 cocultures (b) was determined by sequencing. (c) Each of the Env_acute viruses was competed against the Env_chronic controls in PHA- and IL-2-activated PBMCs to assess the pathogenic fitness (21). The average pathogenic fitness versus the three Env_chronic viruses is shown. (d) The percent replication of Env_acute viruses in competitions in tissue and in MC + PM1 cocultures was plotted against the average percent replication in PBMC competitions. ND, not done.

FIG 6

Binding profile of Env_acute and Env_chronic viruses to C-type lectins. Binding to DC-SIGN and langerin was determined using an acoustic biosensor. Binding affinity (dissociation constant [K_d], in nanomolar units) was determined by binding the lectins of interest to the chip and allowing viruses to flow over the surface. Average replication in tissue of Env_acute (red squares) and Env_chronic (green squares) viruses was correlated with the binding affinity to DC-SIGN (a), langerin (b), and PHA (c). Average transmission fitness in MC+PM1 cocultures of Env_acute (red squares) and Env_chronic (green squares) viruses was plotted against the binding affinity to DC-SIGN (d), langerin (e), and PHA (f). PHA was tested as the nonspecific control. Each symbol represents one virus. Statistical analysis was performed using a Spearman rank correlation analysis.

FIG 7

Presence of MBL in different mucosal tissue types. (a) Fluorescent deconvolution microscopy images of ectocervix, endocervix, inner foreskin, and outer foreskin stained for MBL (red) and DAPI (blue). (b) Autofluorescence (green) of the corresponding tissues. Magnification, ×40. Bar, 40 μm. Dashed lines indicate the luminal surface. L, lumen; E, epithelium; S, stroma.

FIG 8

Mannan alters the transmission of HIV-1 across mucosal tissue. (a) Cervical and (b) penile tissues were incubated in medium alone or medium containing mannan for 12 h before exposure to a mix of one Env_acute and one Env_chronic virus. Migratory cells were harvested after 24 h and cocultured with PM1 T cells. After 10 days in culture, the total HIV-1 copy number in tissue and MC+PM1 cells with and without mannan was assessed using a gag-specific qPCR. Each assay was performed with tissue from three donors. The values are means and standard errors of the means.