Association between virus-specific cytotoxic T-lymphocyte and helper responses in human immunodeficiency virus type 1 infection

Abstract

Cellular immune responses are thought to be an important antiviral host defense, but the relationship between virus-specific T-helper and cytotoxic-T-lymphocyte (CTL) responses has not been defined. To investigate a potential link between these responses, we examined functional human immunodeficiency virus type 1 (HIV-1)-specific memory CTL precursor frequencies and p24-specific proliferative responses in a cohort of infected untreated persons with a wide range of viral loads and CD4 cell counts. Levels of p24-specific proliferative responses positively correlated with levels of Gag-specific CTL precursors and negatively correlated with levels of plasma HIV-1 RNA. These data linking the levels of HIV-specific CTL with virus-specific helper cell function during chronic viral infection provide cellular immunologic parameters to guide therapeutic and prophylactic vaccine development.

FIG. 1

HIV-1-specific CTL precursor frequencies in two subjects with high and low viral loads. For CTLp analysis, serial dilutions of freshly isolated PBMC were placed in 24 replicate wells and stimulated with a CD3-specific MAb (11). For the assessment of proliferative responses, freshly isolated PBMC were placed in six replicate wells in the presence of tetanus toxoid, p24 antigen, or gp160. Viral load measurements were performed on cryopreserved plasma (Amplicor HIV Monitor Test; Roche Molecular Systems, Branchberg, N.J.) according to the manufacturer’s specifications. (A) CTLp in subject CTS-01 (subject 21) (viral load, <400 copies/ml; CD4 count, 900 cells/mm³; duration of infection, 15 years at time of assay) and CTLp in subject 221L (subject 8) (viral load, 200,000 RNA copies/mm³; CD4 count, 800 cells/mm³; duration of infection, 11 years at time of assay). (B) Proliferative responses in subjects CTS-01 (subject 21) and 221L (subject 8). The dotted line represents an SI value of 3.

FIG. 2

Relationship between Gag-specific CTLp and proliferative responses and plasma HIV-1 RNA level. (A) Levels of Gag-specific CTLp do not significantly correlate with low levels of plasma viremia. (B) Increasing p24-specific helper responses are negatively correlated with viral load. Assays were performed on 21 antiretroviral-naive individuals with a wide range of CD4 counts and viral loads. HIV RNA was measured with the Amplicor HIV monitor test according to the manufacturer’s specifications. Individual subjects are numbered.

FIG. 3

Relationship between HIV-1-specific proliferative responses and CTL responses. CTL precursor frequencies directed against Gag, Env, RT, and Nef were plotted against p24 SI values. (A) The presence of a p24-specific proliferative response is strongly associated with increasing levels of Gag-specific CTL precursors. (B to D) Levels of Env, RT, or Nef-specific CTL are not strongly associated with the level of p24-specific help. Individual subjects are numbered, as in Fig. 2.

FIG. 4

CTL precursor frequencies in the presence or absence of HIV-1-specific T-helper responses. (A) Subjects with a p24 SI value of >3 have higher numbers of circulating HIV-1-specific Gag CTL precursors. (B) Subjects with significant helper responses have lower viral RNA levels. Box plots show the mean, upper bounds of the 75th and 90th percentiles and the lower bounds of the 10th and 25th percentiles.