Table 1.
Background demographic and clinical characteristics of study participants at baseline.
Table 2.
Subject population virologic, clinical and phylogenetic characteristics.
Fig 1.
HIV-1 blood and CSF populations early during infection can be equilibrated or compartmentalized.
Neighbor-joining phylogenetic trees depicting sequence relationships from subjects with (A) equilibrated and (B) compartmentalized viral populations. env sequences from the CSF (blue circles) and blood plasma (red triangles) are shown. Bootstrap values ≥35 are included (*) at the appropriate nodes. Genetic distance is indicated at the bottom of each figure (0.001, number of nucleotide substitutions per site between env sequences). Compartmentalized CNS populations are designated by an open circle at the appropriate node and a solid blue bracket. Clonal amplification is indicated by a solid black bar.
Fig 2.
Correlates of CSF HIV-1 RNA levels, CNS inflammation, and CNS phylogenetic state.
(A) Pie charts grouping the samples by days p.i. and showing the percentage of samples in each phylogenetic state as a function of time: 0–4 months (‘acute’), 5–12 months (‘early’), and 13–24 (‘established’). States represented include: Not Analyzed by SGA, due to CSF viral load <1,000 copies/ml; Equilibrated (−), CSF WBC <10 cells/μl; Equilibrated (+), CSF WBC ≥10 cells/μl; and Compartmentalized. We used a cut-off of 10 WBC/μl to define a state of substantial CSF pleocytosis because this is two-times the upper limit of normal value of CSF WBC (0 to 5 cells/μl) [22] ensuring that the measured pleocytosis was a robust marker for an inflammatory response. (B) Pie chart showing the percent of samples in each phylogenetic state exhibiting substantial CSF pleocytosis (CSF WBC ≥10 cells/μl). (C) Relationship between plasma and CSF HIV-1 RNA concentrations for not analyzed samples with minimal pleocytosis (NA (−); CSF WBC <10 cells/μl) and equilibrated samples with minimal pleocytosis (Eq (−), CSF WBC <10 cells/μl). Spearman’s rank correlation coefficient and P-value are indicated. Samples with undetectable CSF HIV-1 RNA (<50 copies/ml; limit of detection) were not included in the correlation analysis. (D) Relationship between CSF viral loads for equilibrated with minimal pleocytosis (Eq (−); CSF WBC <10 cells/μl); equilibrated with marked pleocytosis (Eq (+); CSF WBC ≥10 cells/μl); and compartmentalized (Comp) samples. Significant P values (Mann-Whitney Test) comparing relationships between groups are indicated. Samples with evidence of marked pleocytosis (CSF WBC ≥10 cells/μl) are indicated using open circles, while samples with minimal to no pleocytosis are indicated by solid circles.
Fig 3.
An assessment of the viral populations in the two compartments longitudinally.
Examination of longitudinal relationships early during infection for 20 subjects with multiple time points where at least one time point was analyzed for viral populations. This group did not differ significantly from the 17 subjects with longitudinal samples available but not analyzed (due to low CSF viral load) in the average number of samples per person or the time span covered. Within individual subjects, the days post infection for each sample is indicated using circles, which represent the following: open light blue circle, sample not analyzed due to CSF viral load (VL) < 1,000 copies/ml (NA (low CSF VL)); open royal blue circle; sample not analyzed but CSF VL > 1,000 copies/ml (NA (high CSF VL)); closed dark blue circle, equilibrated with minimal CSF pleocytosis (CSF WBC <10 cells/μl) (Equilibrated (−)); closed pink circle, equilibrated with marked CSF pleocytosis (CSF WBC ≥10 cells/μl) (Equilibrated (+)); and closed red circle, compartmentalized. (A) Subjects with all analyzed time points equilibrated with minimal to no pleocytosis and remaining time points not analyzed. (B) Subjects with pleocytosis or compartmentalization in one time point (above dotted line) and subjects with pleocytosis, or compartmentalization in at least two time points (below dotted line). In every subject the compartmentalized (C) viral sequence population changed between sampling time points, indicated by different subscript numbers.
Fig 4.
Compartmentalization can persist and evolve independently within the CSF over time.
Neighbor-joining phylogenetic trees showing how compartmentalization can: (A) be persistent with multiple clonal expansions allowing recombination; (B) consist of sequential transient clonal expansions; and (C) be established with a transmitted variant. env sequences from the CSF are labeled with circles (C, colors designated in figure) and env sequences from the blood plasma are labeled with triangles (P, colors designated in figure). Days p.i. are noted. Bootstrap values ≥ 50 are indicated (*) at the appropriate nodes to highlight the more significant branch points. Genetic distance is indicated at the top of each phylogenetic tree (0.001, number of nucleotide substitutions per site between env sequences.) Compartmentalized populations are indicated at the appropriate node by an open circle and emphasized using a blue bar. BEAST-generated TMRCAs of the entire viral population are noted adjacent to the subject ID, and the TMRCAs of the different compartmentalized linages (subject 9040 and 9021) and transmitted parental lineages (subject 7146) are also noted.
Fig 5.
All viruses required high levels of CD4 for efficient entry, indicative of primarily being selected for replication in CD4+ T cells.
Single-cycle infection of HIV-1 Env-pseudotyped reporter viruses on CD4lowCCR5high 293-Affinofile cells [28]. Receptor expression was measured as follows: CD4low = 972 receptors/cell, CD4high = 72,041 receptors/cell, CCR5high = 15,636 receptors/cell. The data are averaged from triplicate wells for each of 2 to 3 env clones that were generated per indicated amplicon. Amplicons were selected for cloning to represent different portions of the phylogenetic tree. Subjects with no evidence of CNS compartmentalization (i.e. equilibrated) are shown in panel (A); and subjects with CNS compartmentalization at one or more time points are in panel (B). Longitudinal time points are indicated (T1, T2, T3, etc.) and samples with marked pleocytosis are noted (*).
Fig 6.
Four states can define the relationship between virus in the CSF/CNS and blood early during infection.
Blood and CSF/CNS compartments are indicated. Blood plasma viral variants are represented by the red virus particles; and CNS viral variants are represented by the blue virus particles. CD4+ T cells are represented by open circles. Arrows indicate direction of virus movement between compartments. (A) State with a CSF HIV-1 RNA level 1–2% of the viral load in the blood and defined by minimal to no local CNS replication or inflammation, resulting in an equilibrated state between the two compartments (when CSF viral load high enough for analysis). Transparent infected CD4+ T cell represents potential local CNS replication that may be obscured by the import of virus into the CNS from the periphery. (B) State of equilibration between CSF and blood accompanied by high levels of pleocytosis, potentially caused by local CNS replication. Transparent infected CD4+ T cell represents potential local CNS replication likely obscured by the high levels of virus secreted by the infiltrating CD4+ T cells. (C) State of CSF/CNS clonal amplification of identical or nearly identical variants within CD4+ T cells. These clonally amplified populations are characterized by low diversity, signified by all CSF viruses in a single shade of blue. (D) State of genetically complex, compartmentalized viral replication within CD4+ T cells in the CSF/CNS indicative of persistent replication beyond a single clonal amplification event; this complexity is signified by CNS viral variants in multiple shades of blue.