The role of lymphocyte apoptosis in septic shock remains a controversial issue. Using Annexin V and flow cytometry analysis on freshly isolated cells, we evaluated circulating lymphocyte apoptosis in 23 septic shock, 25 sepsis without shock, 7 nonseptic critically ill, and 25 control patients. In patients with sepsis, we compared day 1 lymphocyte apoptosis (i.e., within 3 days of the onset of infection) with that observed 5-7 days after (day 6) according to shock state, mortality, and seventy factors. At day 1, patients in septic shock exhibited higher lymphocyte apoptosis than that present in controls (16.5% +/- 3.5% vs. 3% +/- 0.5%, respectively, P = 0.0001). At day 6, patients with sepsis without shock restored undamaged CD4+ T and CD8+ T lymphocyte counts, whereas patients in septic shock increased only CD4+ T cells. Similarly, survivors restored undamaged lymphocyte count at day 6 (+70%, P < 0.001), whereas nonsurvivors did not. Day 6 undamaged lymphocyte count negatively correlated with day 1 SAPS II, day 6 LOD score, mechanical ventilation, and ICU stay duration. We observed no apoptotic effect of septic shock plasma or septic shock circulating mononuclear cells on target lymphoid cell lines. We found no alteration in any death receptors Fas, TRAIL-R1, TRAIL-R2, or in their ligands on circulating blood cells. Catecholamines and interleukin 10 levels significantly increased in patients with septic shock, but did not correlate with apoptosis levels. We conclude that lymphocyte apoptosis is rapidly increased in blood of patients in septic shock and that lymphocyte apoptosis leads to a profound and persistent lymphopenia associated with poor outcome. These results suggest that lymphocyte apoptosis is one of the main components of human septic shock immune dysfunction and could be related more to microcirculatory disturbance than to circulating factors.

No abstract

After infection, human CMV (HCMV) establishes a latent and persistent infection in immature myeloid progenitors and peripheral blood monocytes. Completion of the HCMV life cycle is possible upon maturation of monocytes to tissue macrophages and under permissive circumstances, e.g., immunosuppression. We investigated the hypothesis that HLA-G molecules could be induced during HCMV reactivation in activated macrophages to favor virus dissemination. In this study, we provide evidence that HLA-G Ags are produced during viral reactivation in macrophages generated after allogeneic stimulation of HCMV latently infected monocytes. While HLA-G surface expression is up-regulated, classical MHC-I molecules are partially down-regulated by HCMV. In vivo, bronchoalveolar macrophages collected from patients suffering from acute HCMV pneumonitis also express HLA-G molecules. The direct correlation between HLA-G Ag induction and HCMV infection was confirmed in U-373 MG astrocytoma cells. Soluble HLA-G expression is stimulated upon HCMV infection, and this modulation depends on the cooperative action of the two immediate-early-1 pp72 and immediate-early-2 pp86 products. Because HLA-G transcription is active in macrophages and U-373 MG astrocytoma cells, it is likely that the modulation of HLA-G protein expression during HCMV replication occurs at a post-transcriptional level. Our data suggest that induction of HLA-G molecules could be an additional mechanism that helps HCMV to subvert host defenses.

Monocyte deactivation has been identified as a major factor of immunosuppression in sepsis and is associated with a loss of surface human leukocyte antigen-DR (HLA-DR) expression on circulating monocytes. Using flow cytometry, quantitative reverse transcription-polymerase chain reaction, we investigated this phenomenon in septic patients. We confirmed the early loss of monocyte HLA-DR expression in all infected patients and demonstrated that this persistent lowered expression at Day 6 correlated with severity scores, secondary infection, and death. This phenomenon occurred at a transcriptional level via a decrease in the class II transactivator A (CIITA) transcription. Furthermore, these abnormalities correlated with the high cortisol levels observed in sepsis and not with those of other putative factors such as catecholamines or interleukin-10. Finally, in vitro studies evidenced that glucocorticoids decrease HLA-DR expression at a transcriptional level via a decrease in CIITA mRNA levels, mainly by down modulating its isoforms I and III. We conclude that in human sepsis, the loss of HLA-DR expression on circulating monocytes is associated with a poor outcome. We suggest that the high endogenous cortisol level observed in septic shock may be a possible new factor involved in the loss of HLA-DR expression on monocytes via its effect on HLA-DR and CIITA transcription.