Activated peripheral CD8 lymphocytes express CD4 in vivo and are targets for infection by human immunodeficiency virus type 1

Abstract

There is increasing evidence that CD8 lymphocytes may represent targets for infection by human immunodeficiency virus type 1 (HIV-1) in vivo whose destruction may contribute to the loss of immune function underlying AIDS. HIV-1 may infect thymic precursor cells destined to become CD4 and CD8 lymphocytes and contribute to the numerical decline in both subsets on disease progression. There is also evidence for the induction of CD4 expression and susceptibility to infection by HIV-1 of CD8 lymphocytes activated in vitro. To investigate the relationship between CD8 activation and infection by HIV-1 in vivo, activated subsets of CD8 lymphocytes in peripheral blood mononuclear cells (PBMCs) of HIV-seropositive individuals were investigated for CD4 expression and HIV infection. Activated CD8 lymphocytes were identified by expression of CD69, CD71, and the human leukocyte antigen (HLA) class II, the beta-chain of CD8, and the RO isoform of CD45. CD4(+) and CD4(-) CD8 lymphocytes, CD4 lymphocytes, other T cells, and non-T cells were purified using paramagnetic beads, and proviral sequences were quantified by PCR using primers from the long terminal repeat region. Frequencies of activated CD8 lymphocytes were higher in HIV-infected study subjects than in seronegative controls, and they frequently coexpressed CD4 (mean frequencies on CD69(+), CD71(+), and HLA class II(+) cells of 23, 37, and 8%, respectively, compared with 1 to 2% for nonactivated CD8 lymphocytes). The level of CD4 expression of the double-positive population approached that of mature CD4 lymphocytes. That CD4 expression renders CD8 cell susceptible to infection was indicated by their high frequency of infection in vivo; infected CD4(+) CD8 lymphocytes accounted for between 3 and 72% of the total proviral load in PBMCs from five of the eight study subjects investigated, despite these cells representing a small component of the PBMC population (<3%). Combined, these findings provide evidence that antigenic stimulation of CD8 lymphocytes in vivo induces CD4 expression that renders them susceptible to HIV infection and destruction. The specific targeting of responding CD8 lymphocytes may provide a functional explanation for the previously observed impairment of cytotoxic T-lymphocyte (CTL) function disproportionate to their numerical decline in AIDS and for the deletion of specific clones of CTLs responding to HIV antigens.

FIG. 1

Binding sites of the pan-LTR and C-LTR primers used for detection of proviral LTR sequences. Numbers in parentheses indicate the 5′ base positions in the HXB2 genome. Abbreviations: PBS, primer binding site; OS, outer, sense; IS, inner, sense; IAS, inner, antisense; OAS, outer, antisense.

FIG. 2

Expression of the activation markers CD69 (A), CD71 (B), and HLA class II (C) on CD4 and CD8 lymphocytes from HIV-seronegative (S1 to S5) and HIV-seropositive (p4 to p17) individuals. HIV-seropositive individuals were ranked from low CD4 count (p4) to high CD4 count (p17); mean values for expression of each activation marker on CD4 and CD8 lymphocytes in the seronegative and seropositive groups are indicated above the histograms. A minimum of 7,500 lymphocytes from each individual were collected and analyzed.

FIG. 3

Comparison of frequencies of CD4 expression on activated and nonactivated CD8 lymphocytes identified by CD69 (A), CD71 (B), and HLA class II (C) from HIV-seronegative (S1 to S5) and HIV-seropositive (p4 to p17) individuals. Mean values for CD4 expression on each activated and nonactivated lymphocyte subset are indicated above the histograms. A minimum of 7,500 lymphocytes from each individual were collected and analyzed.

FIG. 4

Flow cytometric analysis of CD4 expression on activated and nonactivated CD8 lymphocytes. (A) Forward and side scatter plots of immunomagnetically isolated T lymphocytes (purity of T cell isolation: CD3⁺ αβ-TCR⁺, 94.69%; CD3⁺ γδ-TCR⁺, 2.18%). (B) Dual-parameter plot of CD8 and CD71 expression from the R1 gate highlighting the activated CD8 lymphocytes (CD71⁺, R2 gate) and nonactivated CD8 lymphocytes (CD71⁻, R3 gate). (C) Histogram of CD4 expression from CD71⁺ CD8 lymphocytes (R3 gate). (D) Histogram of CD4 expression from CD8⁻ lymphocytes from panel B. (E) Histogram of CD4 expression from CD71⁻ CD8 lymphocytes (R2 gate). The frequency of CD4-expressing cells is shown in the histograms; the GMFI of CD4 expression in CD4⁺ populations is indicated below this frequency. Open histograms refer to the binding of an isotype-matched control antibody. A minimum of 25,000 CD8β⁺ lymphocytes from each individual were collected and analyzed.

FIG. 5

Analysis of the levels of expression of CD45RA (naive) and CD45RO (memory) cell markers on CD8 β-chain-positive lymphocytes expressing CD4 molecules. A minimum of 15,000 CD8β⁺ lymphocytes from each individual were collected and analyzed.

FIG. 6

Correlation between proviral loads in different lymphocyte subsets measured by the C-LTR primers (x axis) and the pan-LTR primers (y axis); the regression line and Spearman's correlation coefficient are shown. Symbols: ○, CD4 lymphocyte; ●, CD4⁺ CD8 lymphocyte; □, other T cell; ⧫, non-T cell.

FIG. 7

Contribution to overall proviral load of CD4 lymphocytes, CD4⁺ and CD4⁻ CD8 lymphocytes, other T cells, and non-T cells (expressed as copies per 10⁶ PBMCs) in peripheral blood samples from eight HIV-seropositive individuals. HIV-seropositive individuals were ranked from low CD4 count (p2) to high CD4 count (p20). VL, circulating viral load (RNA copies per milliliter of plasma).