Failure to restore the Vgamma2-Jgamma1.2 repertoire in HIV-infected men receiving highly active antiretroviral therapy (HAART)

Abstract

Gammadelta (gammadelta) T cells expressing the Vgamma2-Jgamma1.2Vdelta2 (Vgamma9-JPVdelta2, alternate nomenclature) T cell receptor (TCR) constitute the major peripheral blood population of gammadelta T cells in adult humans and are specifically depleted during human immunodeficiency virus (HIV) disease. Vgamma2-Jgamma1.2Vdelta2 T cells provide a convenient model for assessing the impact of antiretroviral therapy on cell populations that are not susceptible to direct infection because they do not express CD4 and depletion occurs by indirect mechanisms. We obtained longitudinal PBMC samples from 16 HIV-infected individuals who enrolled in the Multicenter AIDS Cohort Study (MACS) and were starting highly active antiretroviral therapy (HAART). Vgamma2-Jgamma1.2Vdelta2 T cells were depleted in these individuals as a result of HIV infection. Despite evidence for clinical benefits of HAART, the Vgamma2-Jgamma1.2Vdelta2 T cell repertoire did not recover after HAART initiation irrespective of treatment duration. These studies highlight important defects among cell subsets lost due to indirect effects of HIV.

Figure 1. HAART effects on CD4+ or Vδ2+ cells and on the Vγ2 repertoire

For 13 HIV-infected donors (6 suppressors and 7 non-suppressors), we assessed the blood CD4 (upper row) and Vδ2 (middle row) frequency by flow cytometry and estimated the Vγ2-Jγ1.2 frequency (lower row) based on Vγ2 spectratype analysis (Vγ2_990–996). Vγ2_990–996 represents the sum relative frequency of Vγ2 chains 990, 993 and 996 nucleotides in length based on spectratyping. Data is plotted for each time point within the 2 year study interval (left panels), or by comparing (right panels) averaged values at initial and final time points for the 13 HIV-infected donors or 9 HIV-negative controls (over a comparable two year interval). Open bars in the right panels show data for HIV-negative or for the last visit before HAART initiation (Pre-HAART) for HIV+ individuals. Suppressor indicates patients with complete viral suppression after HAART. Non-Suppressor means there was viremia despite treatment. Solid bars show data for late time points that are 2–2.5 years after the early value. HIV-positive indicates the sum of 13 patients, whereas suppressors (n=6) and non-suppressors (n=7) are also shown separately. Error bars indicate standard error of the mean. A significant difference was noted for CD4 lymphocyte frequency in the Suppressor population (upper right panel) with a value of p = 0.008. All other differences were not significant.

Figure 2. Selective Depletion of Longer Vγ2 Chains in HIV-positive Donors Despite HAART

We compared the Vγ2 chain length distribution for a group of HIV-negative adults (white) with Vγ2 spectratypes obtained from HAART-treated HIV-positive adults determined for an early time point (gray) and at an interval approximately 2 years later (black). The averaged relative frequency for each chain length among HIV-negative or HIV-positive donors is plotted (± standard error of the mean). Asterisks (*) indicate statistical significance when comparing seronegative controls with both early and late time points from seropositive donors (Student’s t-test, p<0.05). No individual Vγ2 chain lengths reach statistical significance when comparing only early and late time point relative frequencies for HIV-positive donors.

Figure 3. Sustained Vγ2-Jγ1.2 depletion in HIV-infected individuals receiving HAART

The Vγ2-Jγ1.2 frequency was determined at two or three time points (minimum 6 months between intervals for HIV-positive donors and up to 7 years for uninfected controls) by analysis of Vγ2 sequences from 7 representative HIV-positive donors from the HIV-positive cohort (upper) or 9 healthy, HIV-negative donors (lower). The HIV+ donors MPD01 and MPD11 are separated by a dashed line to indicate that we did not have pre-HAART samples for these individuals and instead substituted data from the first visit after HAART initiation. Each data point represents the analysis of on average 40 Vγ2 CDR3 sequences (range: 19–114) at one of the time points (range: 0.5–2 years for HIV-positive donors, or 0.5–7 years for uninfected controls). The Vγ2 chain sequences with in-frame stop codons were excluded.

Figure 4. Targeted Instability within the Vγ2-Jγ1.2 Repertoire of HIV-infected donors

Vγ2 sequences were obtained from 7 HIV-infected donors at 6 month intervals for 2 years (study visits 1–5, left column) and 9 HIV-negative volunteers in 1999, 2004 or 2006 (right column). The HIV+ donors MPD01 and MPD11 are separated by a dashed line to indicate that we did not have pre-HAART samples for these individuals and instead substituted data from the first visit after HAART initiation. PBMC were purified from the blood of HIV-positive or negative donors and Vγ2 chains were amplified by RT-PCR. For each donor, Vγ2 (upper row) or Vγ2-Jγ1.2 (lower row) nucleotypes were compared between/among time points and those present at more than one time point were considered conserved. The number of sequences that encoded conserved Vγ2 or Vγ2-Jγ1.2 nucleotypes was divided by the total number of sequences encoding Vγ2 or Vγ2-Jγ1.2 nucleotypes, respectively. The conserved repertoire proportion was calculated for each time point and plotted for all donors. Each time point represents an average of 40 individual sequences (range: 19–114) but is a single determination and does not have a standard error. Truncated Vγ2 chains encoding in-frame stop codons were excluded from the analysis. Dotted lines indicate the approximate limit of detection for specific Vγ2 sequences (~10%). Part of these data for HIV-negative donors were reported earlier (Hebbeler et al. 2006).

Figure 5. Peripheral Blood Vδ2 Frequencies are not Related to Changes in CD4 cell Count

Flow cytometry was performed on specimens taken from 10 donors (5 suppressors and 5 non-suppressors) at the pre-HAART visit and at the peak of CD4 cell count post-HAART. We compared the Vδ2+ cell frequencies with the CD4 cell counts at these time points (panel A). There was no significant relationship between these values. Similarly, we compared the change in Vγ2 repertoire (on the basis of spectratyping to define the proportion of chains with lengths between 990–996 nucleotides) for 15 donors (6 suppressors, 7 non-suppressors and 2 persons where we used the initial visit after HAART as the baseline). Again, we did not see a significant relationship between Vγ2 chain repertoire and CD4 cell count.