Profound and persistent decrease of circulating dendritic cells is associated with ICU-acquired infection in patients with septic shock.

PURPOSE: Septic shock induces a decrease in dendritic cells (DCs) that may contribute to sepsis-induced immunosuppression. We analyzed the time course of circulating DCs in patients with septic shock and its relation to susceptibility to intensive care unit (ICU)-acquired infections. METHODS: We enrolled adult patients with septic shock (n = 43), non-septic shock (n = 29), and with sepsis without organ dysfunction (n = 16). Healthy controls (n = 16) served as reference. Blood samples were drawn on the day of shock (day 1), then after 3 and 7 days. Myeloid (mDC) and plasmacytoid (pDC) DCs were counted by flow cytometry. Cell surface HLA-DR expression was analyzed in both DC subsets. RESULTS: At day 1, median mDC and pDC counts were dramatically lower in septic shock patients as compared to healthy controls (respectively, 835 mDCs and 178 pDCs/ml vs. 19,342 mDCs and 6,169 pDCs/ml; P < 0.0001) but also to non-septic shock and sepsis patients (P < 0.0001). HLA-DR expression was decreased in both mDCs and pDCS within the septic shock group as compared to healthy controls. DC depletion was sustained for at least 7 days in septic shock patients. Among them, 10/43 developed ICU-acquired infections after a median of 9 [7.5-11] days. At day 7, mDC counts increased in patients devoid of secondary infections, whereas they remained low in those who subsequently developed ICU-acquired infections. CONCLUSION: Septic shock is associated with profound and sustained depletion of circulating DCs. The persistence of low mDC counts is associated with the development of ICU-acquired infections, suggesting that DC depletion is a functional feature of sepsis-induced immunosuppression.

Circulating dendritic cells and interferon-alpha production in patients with tuberculosis: correlation with clinical outcome and treatment response.

Dendritic cells (DC) have been characterized recently as having an important role in the initiation and control of immunological response to Mycobacterium tuberculosis infection. Blood DC have been subdivided into myeloid (mDC) and plasmacytoid (pDC) subsets, on the basis of differences in phenotype markers and function. Little is known about the enumeration and functional evaluation of circulating DC in patients with tuberculosis and their correlation with clinical outcome during the course of anti-tuberculous treatment. We assessed circulating mDC and pDC counts measured by a newly developed single-platform flow cytometric assay based on TruCOUNT, as well as the production of interferon (IFN)-alpha after in vitro stimulation by herpes simplex virus (HSV-1) in 24 patients with active tuberculosis (TB) and 37 healthy donors. Absolute numbers of both DC subsets were decreased significantly in patients with active TB compared to controls. Similarly, the production of IFN-alpha was highly impaired. In 13 patients these parameters were assessed longitudinally, before and after the specific anti-microbial treatment. Most interestingly, in all nine patients with successful anti-tuberculous therapy there was a significant and marked increase of pDC counts and IFN-alpha production. In contrast, no significant longitudinal variations in DC counts and IFN-alpha production were observed in four patients with lack of response to specific treatment. In conclusion, active TB is associated with a defect in blood DC numbers and IFN-alpha production that is restored after bacterial clearance and clinical improvement, as a result of effective anti-tuberculous treatment.

Dendritic cells in innate immune responses against HIV.

Dendritic cells (DCs) were recently found to be innate immunity effectors against tumoral cells and viruses. (i) In response to most viruses, including HIV, plasmacytoid DCs are responsible for most of the type I IFN secretion, which is strongly anti-viral and induces TH1 type responses. Myeloid DCs secrete IL-12, which is also important for TH1-type and cytotoxic responses. In HIV patient blood, both DC population numbers decrease as early as the primary stage. Plasmacytoid DC numbers correlate with type I IFN secretion, which is a prognosis predictor, particularly under treatment. IL-12 secretion is also defective. Immunotherapies to replace the defective cytokines or to restore a potentially defective DC-T lymphocyte feed-back might help patients restore their immune responses under antiviral therapy. (ii) After measles and other viral infections, or incubation with dsRNA, DCs become cytotoxic and consequently exhibit natural killer function, through upregulation of type I IFN secretion which enhances TRAIL expression. In HIV infection, this mechanism was not demonstrated yet, but it might a) be responsible for the massive apoptosis of uninfected lymphocytes, and b) increase specific immunity through cross-presentation of antigens from infected cells killed by DCs. (iii) DCs direct expansion and effector functions of NK cells in the absence of adaptive-type cytokines and modulate NKT cell IFN-gamma production. Reciprocally, NK activation triggers DC maturation. HIV-1 Tat inhibits NK cell cytotoxicity directly and probably through inhibition of IL-12 secretion by DC. Therefore, understanding the functions of DCs in innate immune responses and in pathogenesis will help obtain better HIV replication control.

Reduced blood CD123+ (lymphoid) and CD11c+ (myeloid) dendritic cell numbers in primary HIV-1 infection.

Successful immunologic control of HIV infection is achieved only in rare individuals. Dendritic cells (DCs) are required for specific antigen presentation to naive T lymphocytes and for antiviral, type I interferon secretion. Two major blood DC populations are found: CD11c+ (myeloid) DCs, which secrete IL-12, and CD123+ (IL-3-receptor+) DCs (lymphoid), which secrete type I interferons in response to viral stimuli. The authors have previously found a decreased proportion of blood CD11c+ DCs in chronic HIV+ patients. In this study, 26 to 57 days after infection and before treatment, CD123+ and CD11c+ DC numbers were dramatically reduced in 13 HIV+ patients compared with 13 controls (P =.0002 and P =.001, respectively). After 6 to 12 months of highly active antiretroviral therapy, DC subpopulation average numbers remained low, but CD123+ DC numbers increased again in 5 of 13 patients. A strong correlation was found between this increase and CD4 T-cell count increase (P =.0009) and plasma viral load decrease (P =.009). Reduced DC numbers may participate in the functional impairment of HIV-specific CD4+ T cells and be responsible for the low type I interferon responsiveness already known in HIV infection. The restoration of DC numbers may be predictive of immune restoration and may be a goal for immunotherapy to enhance viral control in a larger proportion of patients.

Downregulation of major histocompatibility class I on human dendritic cells by HIV Nef impairs antigen presentation to HIV-specific CD8+ T lymphocytes.

The HIV early regulatory Nef protein downregulates surface expression of major histocompatibility class I (MHC I) molecules on various immortalized cell lines and on T lymphocytes. MHC I-restricted presentation induces CD8+ T cell responses, which have a major role in limiting HIV infection. Induction of primary immune responses requires dendritic cells, which are major candidates as the first cells that can internalize the virus and present it to T cells in mucosal contamination. To test the effect of Nef on MHC I-restricted antigen presentation by dendritic cells, we used recombinant vaccinia viruses. Flow cytometric analysis of double labeling for a vaccinia protein and MHC I showed that HIV-1 Lai Nef indeed downregulated MHC I surface expression on dendritic cells. MHC I-restricted presentation to a Nef-specific CD8+ cell clone from an infected patient was decreased in an interferon gamma ELISpot assay. Presentation of a reverse transcriptase epitopic peptide on sorted Nef-infected cells was decreased in a peptide concentration-dependent way, confirming the role of MHC I downregulation in the impairment of the CD8+ cell-specific response. Therefore, Nef downregulates MHC I surface expression on human dendritic cells, impairing presentation to HIV-specific CD8+ cells. This action of Nef probably induces a deleterious delay in the early CD8+ responses during the first days of infection and at the onset of new viral mutants.

Lipopeptide presentation pathway in dendritic cells.

Lipopeptides are currently being evaluated as candidate vaccines in human volunteers. They elicit cytotoxic responses from CD8(+) T lymphocytes, whereas peptides without a lipidic moiety usually do not. The exact processing and presentation pathways leading to association with MHC class I molecules has not yet been defined. This is of particular interest in dendritic cells, which are required for primary T cell stimulation. We have tracked lipopeptides derived from an HLA-A2.1-restricted HIV-1 Reverse Transcriptase epitope, by N-terminal addition of an N-epsilon-palmitoyl-lysine. Entry of the lipopeptides into human monocyte-derived dendritic cells (MDC) was mediated by endocytosis, as assessed by colocalization using analogs labelled with rhodamine, and by confocal microscopy. This internalization in DC induced functional stimulation of CD8(+) T lymphocytes specific for the epitopes, quantified by Interferon-gamma ELISPOT assays. The peptide alone was not visualized inside the DC and was only presented through direct surface association to HLA-A*0201. Therefore, lipopeptides provide a model system to define precisely the cross-presentation pathways that lead exogenous proteins to associate with class I MHC molecules within dendritic cells. Using this approach, cross-presentation pathways can be better defined and vaccine lipopeptides can be further optimized for MHC class I association in human dendritic cells.

Investigation of human spleen dendritic cell phenotype and distribution reveals evidence of in vivo activation in a subset of organ donors.

Although the mouse spleen dendritic cell (DC) is perhaps the most intensively studied DC type, little has been published concerning its human equivalent. In this report, rare event flow cytometry and in situ immunofluorescence were used to study the surface phenotype and distribution of HLA-DR(+) CD3(-)14(-)16(-)19(-) human spleen DC. Spleens from organ donors with different clinical histories were used. Most (81% +/- 9%; n = 14) spleen DCs expressed high levels of the integrin CD11c. CD11c(+) DCs were distributed in 3 distinct regions-the peri-arteriolar T-cell zones, the B-cell zones, and the marginal zone, where they formed a ring of cells surrounding the white pulp, just inside a ring of CD14(+) red pulp macrophages, apparently more regularly organized than the previously described marginating DC population in the mouse spleen. The T-cell zones contained CD86(+) DCs, among which a subpopulation expressed CD83. These mature/activated CD86(+) DCs represented a minority (12% +/- 8%) of total spleen DCs in most organ donors: most spleen DCs are immature. In 3 of 18 (17%) donors, however, most (54%-81%) of spleen DCs were CD86(+), suggesting that in vivo DC activation had occurred. In one donor, a radical shift in DC distribution from the marginal zone to the T-cell zones was also observed. This activation of spleen DCs in vivo was reminiscent of the effects of experimental microbial product injection in mice, and it seemed to correlate with bacterial infection or multiple trauma. (Blood. 2001;97:3470-3477)

HIV-specific effector cytotoxic T lymphocytes and HIV-producing cells colocalize in white pulps and germinal centers from infected patients.

Human immunodeficiency virus (HIV) infection is characterized by the massive infiltration of secondary lymphoid organs with activated CD8(+) T lymphocytes. While converging data indicated that these cells were HIV-specific cytotoxic T lymphocytes (CTLs) responsible for HIV spread limitation, direct evidence was lacking. Here, the presence of HIV-specific effector CTLs was demonstrated directly ex vivo in 15 of 24 microdissected splenic white pulps from an untreated patient and in 1 of 24 tonsil germinal centers from a second patient with incomplete viral suppression following bitherapy. These patients had plasma HIV RNA loads of 5900 and 820 copies per milliliter. The frequencies of HIV-1 DNA(+) cells in their lymphoid organs were more than 1 in 50 and 1 in 175, respectively. Spliced viral messenger RNA (a marker for ongoing viral replication) was present in most immunocompetent structures tested. Conversely, CTL activity was not found in spleens from 2 patients under highly active antiretroviral therapy, with undetectable plasma viral load. These patients had much lower spleen DNA(+) cell frequencies (1 in 2700 and 1 in 3800) and no white pulps containing spliced RNA. CTL effector activity as well as spliced viral messenger RNA were both concentrated in the white pulps and germinal centers. This colocalization indicates that viral replication in immunocompetent structures of secondary lymphoid organs triggers anti-HIV effector CTLs to these particular locations, providing clues to target therapeutic intervention.

Endocytosis of an HIV-derived lipopeptide into human dendritic cells followed by class I-restricted CD8(+) T lymphocyte activation.

CD8(+) T lymphocytes, which are major immune effectors, require primary stimulation by dendritic cells (DC) presenting MHC class I molecule-bound epitopes. Sensitization to exogenous protein epitopes that are not synthesized in DC, such as cross-priming, is obtained through pathways leading to their association with MHC class I. To follow class I-restricted pathways in human DC, we have tracked a lipopeptide derived from the conserved HLA-A*0201-restricted HIV-1 reverse transcriptase 476-484 epitope, by N-terminal addition of an Nepsilon-palmytoyl-lysine. Indeed, lipopeptides elicit cytotoxic responses from CD8(+) T lymphocytes, whereas peptides without a lipid moiety do not. The lipopeptide and its parent peptide were labeled unequivocally by rhodamine to study their entry into immature monocyte-derived human DC by confocal microscopy. The lipid moiety induced endocytosis of the lipopeptide, assessed by rapid entry into vesicles, colocalization with Dextran-FITC and dependence on energy. Internalization occurred even when actin filaments were depolymerized by Cytochalasin B. This internalization induced functional stimulation of specific CD8(+) T lymphocytes in IFN-gamma ELISPOT assays. The peptide alone was not visualized inside the DC and was presented through direct surface association to HLA-A*0201. Therefore, lipopeptides are a unique opportunity to define precisely the pathways that lead exogenous proteins to associate with MHC class I molecules in DC. The results will also be useful to design lipopeptide vaccines.

Extension of HLA-A*0201-restricted minimal epitope by N epsilon-palmitoyl-lysine increases the life span of functional presentation to cytotoxic T cells.

The delineation of the minimal requirements for efficient delivery of functional cytotoxic epitopes into APC could be a step toward the definition of "minimal length" lipopeptides for the modulation of CTL activity. Several analogues of the HLA-A*0201-restricted HIV-1 polymerase (pol476-484) minimal cytotoxic epitope were obtained by modifying P0, P1, or P10 positions by a single N epsilon-palmitoyl-lysine residue. The use of fluorescent derivatives confirmed the cell-permeating activities and suggested that a P0- and a P1-modified lipopeptide possessing ionizable extremities fulfills the structural requirements for MHC loading. The expressions of HLA-peptide complexes at the surface of TAP-deficient cells incubated with the parent epitope or lipopeptide derivatives were compared, in terms of intensity and stability. Both lipopeptides induced a considerably prolonged expression of conformationally correct complexes, which were dependent on the integrity of the exocytosis pathway, suggesting a dynamic mechanism of formation or reloading of the complexes from an intracellular pool. The agonistic activities of the different HLA-peptide complexes were evaluated using two independent T cell lines from HIV-infected donors. We report that a lipodecapeptide obtained by N-terminal addition of a N epsilon-palmitoyl-lysine to the pol476-484 epitope was able to increase the life span of functional presentation to cytotoxic T cells specific for the parent peptide.

Depletion in blood CD11c-positive dendritic cells from HIV-infected patients.

OBJECTIVES: To quantify blood dendritic cells from HIV-positive patients and to study the expression of functional molecules, in relation to HIV viral load, CD4 cell counts and antiretroviral treatment. DESIGN AND METHODS: Three-colour flow cytometry analysis was used to quantify blood dendritic cells without previous isolation from whole blood and to study the expression of functional molecules (MHC class II, CD11c, CD83, CD86) by dendritic cells from 30 HIV-positive patients, 15 of whom were treated with combined antiretroviral therapy (viral loads from undetectable to 5.4 log copies/ml, CD4 cell counts 1-1895 cells/mm3) and 11 non-infected controls. RESULTS: The median proportion of blood dendritic cells from HIV-positive patients was significantly decreased when the plasma viral load was above 200 copies/ml: 0.2% (0.1-1.1, n = 19) compared with 0.4% (0.2-0.8, n = 11) in patients with undetectable viral load whether they were treated or not, and to 0.4% (0.2-1.3, n = 11) in controls (P = 0.02). A major decrease of the CD11c positive dendritic cells was observed in all HIV-positive samples, with only 18% (mean; range: 0.3-80%, median 4.2%) compared with 44% (11-70%, median 42%) of control dendritic cells (P = 0.0006). In contrast, the proportion of dendritic cells expressing CD86, was slightly higher in HIV-positive patients than in controls (P = 0.03). CONCLUSIONS: The decreased proportion of blood dendritic cells correlated with virus replication and the lack of dendritic cells expressing CD11c are the first evidence of strong dendritic cell alterations in HIV-positive patients. Although the proportion of blood dendritic cells are in the normal range in treated HIV-positive patients with undetectable viral load, the CD11c alterations persist indicating that antiretroviral therapy might only partly correct the alterations of the circulating dendritic cells.

Calcium responses elicited in human T cells and dendritic cells by cell-cell interaction and soluble ligands.

The interactions between a human CD4+ T cell clone and monocyte-derived human dendritic cells (DC) were analyzed with an imaging system. The first question addressed was the relationship between the formation of a contact zone and the triggering of a Ca2+ response in the T cells, in the presence or absence of antigen. Interaction of T cells with DC pulsed with the antigen led to the formation of a stable contact zone, followed by the appearance in the T cells of large and sustained Ca2+ oscillations. In the absence of antigen, contact zones formed normally and, surprisingly, Ca2+ responses were also observed, characterized by rare and small transients. Antigen-independent Ca2+ responses were not MHC restricted. The possible influence of Ca2+ responses in the DC on the efficiency of antigen presentation was then Investigated. In DC, Ca2+ responses can be elicited by a variety of stimuli: cell adhesion, platelet-activating factor, UTP and chemotactic molecules (formyl-Met-Leu-Pro, RANTES, MIP-1beta and SDF-1alpha). Importantly, Ca2+ responses were also induced in approximately 30% of DC as a result of their interaction with T cells. However, the efficiency of antigen presentation (as judged by the percentage of T cells presenting a Ca2+ response) was independent of the Ca2+ level in DC. Thus, imaging the interactions between human T cells and DC led us to observe two novel phenomena: DC-induced but antigen-independent Ca2+ responses in T cells and T cell-induced Ca2+ responses in DC.

Monocyte-derived dendritic cells have a phenotype comparable to that of dermal dendritic cells and display ultrastructural granules distinct from Birbeck granules.

Most monocyte-derived dendritic cells (DC) display CD1a, like Langerhans cells (LC) and some dermal DC, but their relationship with these skin DC remains unclear. To address this issue, we studied the expression of different antigens characteristic of skin DC and of monocyte/macrophages in CD1a+ and CD1a- monocyte-derived DC. Their phenotype indicated that they may be related to dermal DC rather than to LC, i.e., they were all CD11b-positive, and 72% were Factor XIIIa-positive, but they did not express E-cadherin nor VLA-6. It is interesting that CD1a+ and CD1a-cells showed intracytoplasmic granules that were different from LC Birbeck granules. These phenotypical and ultrastructural features are comparable to those of CD14-derived DC obtained from cord blood precursors [C. Caux et al. J. Exp. Med. 184, 695-706]. These results show a close relationship between these two in vitro models, which are both related to dermal DC.

The importance of pairwise interactions between peptide residues in the delineation of TCR specificity.

A minimal, nonamer epitope (TEMEKEGKI) from the reverse transcriptase protein of HIV-1, restricted by H-2Kk, was identified and the function of individual residues determined. Besides classical anchor residues at positions 2 and 9, methionine at position 3 was identified as an important MHC anchor and improved binding of a different (malarial) nonamer epitope to H-2Kk, albeit while also abolishing CTL recognition. Lysine at position 5 was replaceable by alanine for CTL raised against wild-type peptide but abolished recognition for CTL raised against the variant 5ALA peptide, indicating a unidirectional cross-reactivity. Interestingly, one CTL line raised against the 5ALA substituted peptide was permissive for a double substitution at positions 5 and 6, in which lysine was permissive at position 5 only if the adjacent glutamic acid was replaced by alanine. Extensive analysis revealed three distinct patterns of responses with peptides doubly substituted in this region: recognition of both single substitutions but not the double substitution, recognition of only one single substitution but also the double substitution, or recognition of both single substitutions and the double substitution. A second complementary substitution can therefore restore function lost through a first substitution. Thus, no residue acts independently of its neighbors, and pairs of substitutions may give results not predictable from the effects of each taken singly. This finding may have bearing on viral infections (such as HIV), in which the accumulation of two mutations in the epitope may lead to the reengagement of memory CTL previously silenced by the initial mutation.

Cytotoxic T-cell responses to HIV-1 reverse transcriptase, integrase and protease.

OBJECTIVES: To determine immunodominant regions and new epitopes for cytotoxic T cells (CTL) directed against the HIV-1 pol products reverse transcriptase (RT), integrase and protease in a large cohort of patients at different stages of disease. DESIGN AND METHODS: Cross-sectional analysis of 98 patients from the French IMMUNOCO cohort (CD4 counts: 125-1050 x 10(6) cells/l), monitored for CTL recognition of HIV-1 pol products using recombinant vaccinia virus constructs and synthetic peptides. RESULTS: Memory CTL responses against HIV-1 pol products were detected in 78% of all patients whatever the stage of disease. RT was more immunogenic (81%, 30 out of 37 patients) than integrase and protease (51% and 24%, respectively). CTL recognition of RT was more frequent against Pol amino acids 310-460 (61%, 11 out of 18 patients) than against the other three portions (Pol 168-310, Pol 450-600, Pol 590-728) in patients with CD4 counts > 400 x 10(6)/l, whereas in patients at advanced stages no prominent differences were observed. Two new clusters of antigenic regions were found in the NH2 segment: three epitopes between amino-acids Pol 200 and 217 and four epitopes between amino-acids Pol 346 and 387, using five different HLA-restricting elements. A new cluster of three conserved epitopes was found in the COOH segment of RT. CONCLUSIONS: This study shows that memory CTL responses against HIV-1 RT, integrase and protease are detectable in most patients at different stages of disease. The capacity of CTL to recognize simultaneously clusters of epitopes may become important for the immune control to reinforce antiretroviral drug efficiency.

Low CD83, but normal MHC class II and costimulatory molecule expression, on spleen dendritic cells from HIV+ patients.

Dendritic cells (DCs), which are the most potent antigen-presenting cells for T lymphocytes, are targets for HIV in vitro and in vivo. Antigen presentation by DCs has been suggested to be impaired during HIV infection; however, the extent to which DCs from HIV+ individuals are altered, particularly in lymphoid organs where T cell stimulation takes place, is not clear. To address this question, the levels of expression of functionally important molecules by spleen DCs from HIV+ patients (n = 6), and HIV- organ donors (n = 5) were compared. By rare event analysis of flow cytometry data, spleen DCs from HIV+ patients were not depleted, representing 0.6 +/- 0.4% of spleen mononuclear cells compared with 0.8 +/- 0.5% in HIV- spleens. Fresh HIV+ spleen DCs were MHC II+ and weakly CD86+CD40+, but negative for CD83 and CD80, and hence had a normal phenotype, showing no signs of in vivo activation. After 24 hr of culture, they upregulated the expression of MHC II, CD40, CD80, and CD86 to levels just as high as those on DCs from organ transplant donors. However, cultured DCs from HIV+ spleens showed lower expression of CD83, compared with DCs from HIV- spleens. The biological significance of this observation will be appreciated further when the function of this molecule is better known. These results suggest that putative defects in antigen presentation by DCs from HIV+ patients are not related to the surface expression of MHC II, CD40, CD80, or CD86.

Dynamics of HIV variants and specific cytotoxic T-cell recognition in nonprogressors and progressors.

Infection with the human immunodeficiency virus (HIV) results in a disease characterized by a rapid viral replication, immunodeficiency and chronic immune activation. The vigorous polyspecific cytotoxic T-cell (CTL) response directed against multiple HIV epitopes reduces HIV-infected cell numbers, although unable to eradicate the virus. The plasticity of the specific CTL repertoire ensures adaptation to the high rate of viral variation that can be found in CTL epitopes of several HIV-1 proteins. However, viral persistence occurs despite continuous CTL recognition and although functional importance of conserved sites in the different HIV proteins may impose constraints to viral variation. In the reverse transcriptase (RT) which is a major target for antiretroviral therapy, the impact of the continuous pressure of drug therapy is more obvious than that of the CTLs. Shifts in immunodominant RT regions seem to allow the maintenance of the HIV-1 RT CTL recognition with disease progression and antiretroviral therapy. In respect to new highly active drug combinations, understanding the capacity of virus-specific CTLs to control residual viral variants seems very important and may allow development of efficient immunotherapies to prevent drug-induced viral resistance.