icon star paper   HIV Articles  
Back grey_arrow_rt.gif
 
 
Therapeutic dendritic-cell vaccine for chronic HIV-1 infection
 
 
  Wei Lu1, 2, Luiz Claudio Arraes2, Wylla Tatiana Ferreira2 & Jean-Marie Andrieu1, 2
 
Nature Medicine 28 November 2004
 
1 Institut de Recherche sur les Vaccins et l'Immunothérapie des Cancers et du Sida and the Laboratoire d'Oncologie et Virologie Moléculaire, Faculté de Médecine René Descartes at the Centre Biomédical des Saints-Pères, Université Paris 5, 45 Rue des Saints-Pères, 75006 Paris, France.
 
2 Instituto de Pesquisa en Immunoterapia de Pernambuco and the Laboratorio de Immunologia Keizo Asami, Universidade Federal de Pernambuco, 50670-901 Recife, Brazil.
 
ABSTRACT
 
We present the results of a preliminary investigation of the efficacy of a therapeutic dendritic cell (DC)-based vaccine for HIV-1. We immunized 18 chronically HIV-1-infected and currently untreated individuals showing stable viral loads for at least 6 months with autologous monocyte-derived DCs loaded with autologous aldrithiol-2-inactivated HIV-1. Plasma viral load levels were decreased by 80% (median) over the first 112 d following immunization. Prolonged suppression of viral load of more than 90% was seen in 8 individuals for at least 1 year. The suppression of viral load was positively correlated with HIV-1-specifc interleukin-2 or interferon gamma-expressing CD4+ T cells and with HIV-1 gag--specific perforin-expressing CD8+ effector cells, suggesting that a robust virus-specific CD4+ T-helper type 1 (TH1) response is required for inducing and maintaining virus-specific CD8+ effectors to contain HIV-1 in vivo. The results suggest that inactivated whole virus--pulsed DC vaccines could be a promising strategy for treating people with chronic HIV-1 infection.
 
AUTHOR DISCUSSION
 
This is the first demonstration in humans that a therapeutic vaccine made of autologous monocyte-derived DCs pulsed with autologous inactivated whole HIV-1 is capable of inducing an effective HIV-1-specific T-cell response associated with sustained viral suppression. Taking into account that over the 6 months before immunization, the PVL of the 18 participants remained stable, whereas their mean CD4 cell count decreased by 100 cells/mul, the significant decrease of PVL as well as maintenance of CD4 cell counts observed at 1 year after immunization are particularly promising. Although a nonspecific adjuvant effect of DCs cannot be ruled out, the key observation that enhanced CD4 TH1 as well as perforin-mediated CTL responses elicited by the vaccine correlated with sustained viral suppression strongly supports an immunologic control of chronic HIV-1 infection by the therapeutic vaccine. We should emphasize, however, that the efficacy of such a therapeutic vaccine will not be definitively proven until a randomized trial with an appropriate control arm has been performed.
 
The observation that, at 1 year, the percentage of HIV-1-gag-specific CD8+ T cells expressing perforin was positively correlated with the 1-year PVL decline (Fig. 5f) underscores the major role of perforin-expressing effectors in controlling HIV-1 replication in vivo. This is in keeping with the observation that perforin-expressing HIV-1-specific CD8+ T cells were associated with the partial control of viral replication in some acutely infected patients with structured antiviral treatment interruption as well as in untreated long-term nonprogressors. These findings suggest that intracellular perforin staining, coupled with existing assays for antigen-specific effectors (such as HLA-restricted tetramer binding technique or ICC assay), could be the most sensitive surrogate markers to monitor functional HIV-1-specific CTL activity in vivo. In addition, the significant correlation between viral suppression and durable increase in frequencies of HIV-1-specific CD4+ TH1 cells and CD8+ effectors observed in immunized patients favors the notion that a strong virus-specific CD4+ TH1-cell response is required to enable virus-specific CD8+ effectors to contain HIV-1 replication in vivo. This is in agreement with the findings from several studies showing a correlation between high levels of virus-specific CD4+ T cells and the (spontaneous) control of viral replication in the chronic phase of HIV-1 infection as well as during the course of human hepatitis virus B or C infection. In this regard, virus-specific CD4+ TH1 cells may provide help to promote virus-specific CD8+ T-cell differentiation (such as perforin expression) through a direct mechanism by secreting TH1 cytokines in situ and/or through an indirect mechanism by triggering CD40-mediated DC activation which in turn favors CD8+ T-cell differentiation. Vaccines boosting virus-specific CD4+ T cells could thus be promising strategies for treating people with progressive HIV-1 infection and other chronic viral diseases.
 
Given that stronger 1-year PVL decreases following the DC vaccination were associated with higher baseline CD4 cell counts or HIV-1-specific IL-2-expressing CD4+ T cells (Supplementary Fig. 2 online), which decline progressively along the course of the infection, it is conceivable that an early therapeutic vaccine intervention could increase the probability of achieving sustained viral suppression. This notion is also supported by the dramatic viral suppression that we observed in immunized Chinese macaques with early chronic simian immunodeficiency virus (SIV) infection.
 
Considering the humoral arm of the immune response after vaccination, it is interesting to observe that our inactivated whole HIV-1-loaded DC vaccine did not induce any neutralizing antibodies. This is in keeping with the consensus that there is poor neutralizing antibody response during the course of HIV infection20. We had however observed an increase in the neutralizing antibody response following AT-2-SIV-loaded DC immunization in SIV-infected macaques, This discrepancy might be explained by the fact that in the macaque study the use of in vitro transformed CEMx174 cell lines and cell line--adapted SIVmac251 for detecting neutralizing antibodies could have artificially resulted in serum neutralization titers much higher than those obtained in humans using an assay system involving primary peripheral blood mononuclear cells (PBMC) and primary viral isolates. The stable serum levels of total antibodies specific for HIV-1 in immunized individuals might represent either saturation of anti-HIV-1 antibody-producing capacities or an unresponsiveness resulting from chronic B-cell hyperactivity with impaired antigen-specific humoral immunity as observed in HIV-infected patients.
 
An important question in vaccinology concerns the best antigen source (inactivated whole virus or recombinant proteins or peptides) for generating a protective anti-HIV-1 cellular immunity. Our results suggest that the adoptive transfer of inactivated whole virus--pulsed DCs is an excellent antigenic preparation to expand and activate in vivo virus-specific CTLs with a wide range of potential HLA-restricted epitopes. Up to 5.3% (mean) of HIV-1-specific effector memory (IFN-gamma-expressing) CD8+ T cells were observed 112 d after the immunization whereas only 3.5% (mean) of total CD8+ T cells were stained by gag tetramer at the same time point. Given that only 10% of gag tetramer--staining CD8+ T cells have been reported to express IFN-gamma following gag-specific stimulation, it seems that more than 90% ([5.3% - 0.35%] ÷ 5.3% = 93.4%) of HIV-1-specific memory CD8+ T cells expanded by the inactivated whole virus--pulsed DCs could probably be attributed to HIV-1 antigenic determinants other than gag. In this setting, it has been recognized that AT-2-inactivated SIV/HIV enters the DC through a receptor-mediated mechanism35, 36 eliciting a potent HLA-I-restricted CTL response, whereas recombinant viral proteins enter DCs through nonspecific endocytosis inducing preferentially humoral (antibody) response. The processing of AT-2-inactivated SIV/HIV virion antigens by DCs thus probably differs from that of recombinant proteins in terms of antigenic epitope presentation. Taken together, these differences might explain why an earlier pilot trial with DCs pulsed with a limited number of recombinant HIV-1 antigens or peptides did not have any effect on viral load or CD4 cell count39.
 
Because our vaccine was made of inactivated whole virus-loaded DCs that were matured with cytokines in vitro, a pertinent question was whether the viroimmunologic response following the DC vaccination depends on the levels of DC maturation in the final vaccine preparation. In the present study, the majority of participants' monocyte-derived DCs (51--81%) underwent maturation (CD83 expression) following in vitro cytokine-induced stimulation. This is in agreement with earlier reports showing that monocyte-derived DCs from HIV-1-infected individuals remained functionally intact and capable of priming naive T cells or stimulating memory T cells in vitro. The lack of correlation between the number of mature DCs and viral-load responses allows us to reject the hypothesis that a difference in DC maturation might account for different viral-load responses in immunized individuals. Our final DC preparations contained 2--28% small cells that were mostly T and B cells. Although the potential presentation of the inactivated whole HIV-1 by B cells cannot be ruled out, involvement of B cells is unlikely because currently there is no evidence that HIV can enter human B cells. Moreover, a recent study in the macaque model has shown little involvement of B cells in the presentation of AT-2-inactivated SIV43.
 
Vaccines for treating chronic HIV-1 infection are distinct from prophylactic vaccines against HIV-1. The goal of therapeutic vaccination is the induction of strong and durable cellular responses that can control viral replication established for years in lymphoid tissues. The ultimate goal is to sustainably reduce the viral load of HIV-1-infected individuals to as low a level as possible. This reduction would protect these individuals from disease progression, which would allow them to live without harmful and expensive daily antiretroviral drugs. At the same time, this would minimize their risk of sexually transmitting the virus to healthy people.
 
ARTICLE TEXT
 
Although the natural immune response to human immunodeficiency virus type 1 (HIV-1) is not effective for eradicating the virus, vigorous HIV-1-specific CD4+ TH1 cell responses have been shown to be associated with control of viremia and long-term nonprogression in infected individuals. Early intervention with highly active antiretroviral therapy (HAART) during or shortly after acute infection was also associated with enhanced HIV-1-specific CD4+ TH1-cell responses. In contrast, at a later stage, HAART led to the decline of HIV-1-specific CD4+ TH1-cell and CD8+ cytotoxic T lymphocyte (CTL) responses, suggesting that the functional capacities of HIV-1-capturing antigen presenting cells (APCs), which are required for the induction of the immune response, are progressively lost along the course of the infection. DCs, the most potent APCs, have a pivotal role in the initiation and maintenance of immune responses against viruses and have been found to be impaired in individuals with progressive HIV-1 infection. Recent studies have shown that the adoptive transfer of autologous DCs loaded in vitro with aldrithiol-2 (AT-2)-inactivated HIV-1 induced protective antiviral immunity in hu-PBL-SCID mice. We had previously shown that a therapeutic vaccine made of AT-2-inactivated simian immunodeficiency virus (SIV) strain mac251 (SIVmac251)-loaded DCs led to considerable viral suppression in the absence of any antiviral therapy in Chinese rhesus monkeys that were immunized 2 months after having been infected with SIVmac251.
 
Here, in a preliminary study, we explore the toxicity and the efficacy of a vaccine made of autologous monocyte-derived DCs pulsed with autologous AT-2-inactivated HIV-1 in patients with chronic HIV-1 infection.
 
Results
 
Vaccination and its clinical consequences
We included 18 currently untreated HIV-1-infected individuals in the present study. At day 0, we subcutaneously immunized each of them at the root of both arms and both thighs with a total dose (1 ml) of 3 times 107 quality-controlled AT-2-inactivated HIV-1-loaded viable DCs. We administered two further injections with the same number of AT-2-inactivated HIV-1-loaded viable DCs at 2-week intervals. All patients completed the three doses of the therapeutic vaccine and were followed for 1 year thereafter without antiviral therapy.
 
The only clinical manifestation associated with the vaccine was an increase in the size of peripheral lymph nodes. The mean diameter (± s.e.m) of left and right axillary and inguinal lymph nodes increased from 0.33 ±0.11 cm before the first immunization to 1.17± 0.20 cm (P < 0.01) at day 14 (second immunization), and to 1.61 ± 0.14 cm (P < 0.001) at day 28 (third immunization). The lymph node size remained significantly increased thereafter: 1.50± 0.19 cm at day 112 (P < 0.001), 1.67 ± 0.16 cm at day 224 (P < 0.001), and 1.06 ± 0.17 cm (P < 0.01) at 1 year. No local or systemic side effects developed and no clinical AIDS or milder immunodeficiency-related symptoms (such as weight loss, unexplained fever, chronic diarrhea or oral candidiasis) occurred during the study period.
 
Viral suppression after vaccination
The median plasma viral RNA load (PVL) of the 18 participants, which was stable for at least 6 months before immunization, decreased by 80% (P < 0.01) over the 112 d after the first immunization. It then remained stable until the end of the study. When looking at individual data, we observed that eight individuals' PVLs declined by more than 90%, whereas the response of the other ten individuals was weaker and transient. The median blood HIV-1 cellular viral DNA load of the 18 participants decreased by 50% (P < 0.01) over the first 112 d. This reduced cellular viral DNA load remained statistically significant (P < 0.05) at day 224 but became insignificant at day 357.
 
The participants' blood CD4+ T-cell counts, which decreased (mean 17 cells/mul loss per month) during the 6 months preceding the immunization, increased significantly from day 28 to day 112 but returned progressively to baseline thereafter (Table 1 and Fig. 1f,h and Supplementary Table 4 online). No statistically significant changes were detected in the CD8+ T cell count throughout the study (Table 1 and Fig. 1g,h).
 
HIV-specific immunity after vaccination
Total titers of antibody to HIV-1 remained unchanged following immunization. Neutralizing antibodies specific for autologous HIV-1 isolates were detected at low titers (1/10), the assay detection threshold in two individuals from day 0 to 1 year. They were undetectable in the 16 other participants, except transiently at low titers (1/10) in three patients (one at day 56 and day 112; two others at day 112 only).
 
Using a highly sensitive flow cytometry--based intracellular cytokine (ICC) assay, we observed that HIV-1-specific interleukin-2 (IL-2)-expressing CD4+ T cells increased around threefold (P < 0.01) over 1 year in the 18 immunized participants. HIV-1-specific interferon-gamma (IFN-gamma) expressing CD4+ T cells increased around twofold (P < 0.01), whereas HIV-1-specific IFN-gamma-expressing (memory) CD8+ T cells increased less than twofold (P < 0.01) with a peak increase at day 112. Individual ICC data of the 18 immunized participants can be found in Supplementary Table 6 online. The 1-year PVL change was strongly correlated with the 1-year frequency of HIV-1-specific IL-2- or IFN-gamma-expressing CD4+ T cells. The correlation of the 1-year PVL change with the 1-year frequency of IFN-gamma-expressing CD8+ T cells was much weaker.
 
Article
Published online: 28 November 2004; | doi:10.1038/nm1147
Therapeutic dendritic-cell vaccine for chronic HIV-1 infection
Wei Lu1, 2, Luiz Claudio Arraes2, Wylla Tatiana Ferreira2 & Jean-Marie Andrieu1, 2
 
1 Institut de Recherche sur les Vaccins et l'Immunothérapie des Cancers et du Sida and the Laboratoire d'Oncologie et Virologie Moléculaire, Faculté de Médecine René Descartes at the Centre Biomédical des Saints-Pères, Université Paris 5, 45 Rue des Saints-Pères, 75006 Paris, France.
 
2 Instituto de Pesquisa en Immunoterapia de Pernambuco and the Laboratorio de Immunologia Keizo Asami, Universidade Federal de Pernambuco, 50670-901 Recife, Brazil.
 
Correspondence should be addressed to Jean-Marie Andrieu jean-marie.andrieu@irvics.org or Wei Lu louis.wei-lu@irvics.org
 
We present the results of a preliminary investigation of the efficacy of a therapeutic dendritic cell (DC)-based vaccine for HIV-1. We immunized 18 chronically HIV-1-infected and currently untreated individuals showing stable viral loads for at least 6 months with autologous monocyte-derived DCs loaded with autologous aldrithiol-2-inactivated HIV-1. Plasma viral load levels were decreased by 80% (median) over the first 112 d following immunization. Prolonged suppression of viral load of more than 90% was seen in 8 individuals for at least 1 year. The suppression of viral load was positively correlated with HIV-1-specifc interleukin-2 or interferon-bold gamma-expressing CD4+ T cells and with HIV-1 gag--specific perforin-expressing CD8+ effector cells, suggesting that a robust virus-specific CD4+ T-helper type 1 (TH1) response is required for inducing and maintaining virus-specific CD8+ effectors to contain HIV-1 in vivo. The results suggest that inactivated whole virus--pulsed DC vaccines could be a promising strategy for treating people with chronic HIV-1 infection.
 
Although the natural immune response to human immunodeficiency virus type 1 (HIV-1) is not effective for eradicating the virus, vigorous HIV-1-specific CD4+ TH1 cell responses have been shown to be associated with control of viremia and long-term nonprogression in infected individuals1, 2, 3, 4, 5, 6. Early intervention with highly active antiretroviral therapy (HAART) during or shortly after acute infection was also associated with enhanced HIV-1-specific CD4+ TH1-cell responses7, 8. In contrast, at a later stage, HAART led to the decline of HIV-1-specific CD4+ TH1-cell and CD8+ cytotoxic T lymphocyte (CTL) responses2, 9, 10, suggesting that the functional capacities of HIV-1-capturing antigen presenting cells (APCs), which are required for the induction of the immune response, are progressively lost along the course of the infection. DCs, the most potent APCs, have a pivotal role in the initiation and maintenance of immune responses against viruses11 and have been found to be impaired in individuals with progressive HIV-1 infection12, 13, 14, 15. Recent studies have shown that the adoptive transfer of autologous DCs loaded in vitro with aldrithiol-2 (AT-2)-inactivated HIV-1 induced protective antiviral immunity in hu-PBL-SCID mice16, 17. We had previously shown that a therapeutic vaccine made of AT-2-inactivated simian immunodeficiency virus (SIV) strain mac251 (SIVmac251)-loaded DCs led to considerable viral suppression in the absence of any antiviral therapy in Chinese rhesus monkeys that were immunized 2 months after having been infected with SIVmac251 (ref. 18).
 
Here, in a preliminary study, we explore the toxicity and the efficacy of a vaccine made of autologous monocyte-derived DCs pulsed with autologous AT-2-inactivated HIV-1 in patients with chronic HIV-1 infection.
 
Results
 
Vaccination and its clinical consequences
We included 18 currently untreated HIV-1-infected individuals in the present study (Table 1). At day 0, we subcutaneously immunized each of them at the root of both arms and both thighs with a total dose (1 ml) of 3 times 107 quality-controlled AT-2-inactivated HIV-1-loaded viable DCs (see Supplementary Fig. 1 and Supplementary Tables 1 and 2 online). We administered two further injections with the same number of AT-2-inactivated HIV-1-loaded viable DCs at 2-week intervals. All patients completed the three doses of the therapeutic vaccine and were followed for 1 year thereafter without antiviral therapy.
 
Table 1. Evolution of T-cell counts and plasma viral loads of 18 vaccinated individuals with HIV-1 infection
 
The only clinical manifestation associated with the vaccine was an increase in the size of peripheral lymph nodes. The mean diameter ( s.e.m) of left and right axillary and inguinal lymph nodes increased from 0.33 0.11 cm before the first immunization to 1.17 0.20 cm (P < 0.01) at day 14 (second immunization), and to 1.61 0.14 cm (P < 0.001) at day 28 (third immunization). The lymph node size remained significantly increased thereafter: 1.50 0.19 cm at day 112 (P < 0.001), 1.67 0.16 cm at day 224 (P < 0.001), and 1.06 0.17 cm (P < 0.01) at 1 year. No local or systemic side effects developed and no clinical AIDS or milder immunodeficiency-related symptoms (such as weight loss, unexplained fever, chronic diarrhea or oral candidiasis) occurred during the study period.
 
Viral suppression after vaccination
 
The median plasma viral RNA load (PVL) of the 18 participants, which was stable for at least 6 months before immunization, decreased by 80% (P < 0.01) over the 112 d after the first immunization. It then remained stable until the end of the study (Fig. 1a--c). When looking at individual data (Fig. 1a,c), we observed that eight individuals' PVLs declined by more than 90%, whereas the response of the other ten individuals was weaker and transient (Supplementary Table 3 online). The median blood HIV-1 cellular viral DNA load of the 18 participants decreased by 50% (P < 0.01) over the first 112 d. This reduced cellular viral DNA load remained statistically significant (P < 0.05) at day 224 but became insignificant at day 357 (Fig. 1d,e).
 
Figure 1. Immunologic and virologic evolution of the 18 immunized participants.
 
(a,b) Individual or geometric mean ( s.e.m) (solid diamonds) plasma viral RNA loads (PVL) before and after immunization. Red lines indicate weak and transient responders who showed <90% PVL decrease by 1 year; green lines indicate strong and durable responders who showed >90% PVL decrease by 1 year. (c) Individual change of PVL by 1 year. (d,e) Individual or geometric mean ( s.e.m) (solid squares) of cellular viral DNA loads. (f,g) Individual blood CD4+ or CD8+ T-cell counts. (h) Mean ( s.e.m) CD4+ (solid circles) or CD8+ (open circles) T-cell counts of the 18 participants. *P < 0.05; **P < 0.01.
 
Full Figure and legend (90K)
The participants' blood CD4+ T-cell counts, which decreased (mean 17 cells/mul loss per month) during the 6 months preceding the immunization, increased significantly from day 28 to day 112 but returned progressively to baseline thereafter (Table 1 and Fig. 1f,h and Supplementary Table 4 online). No statistically significant changes were detected in the CD8+ T cell count throughout the study (Table 1 and Fig. 1g,h).
 
HIV-specific immunity after vaccinationTotal titers of antibody to HIV-1 remained unchanged following immunization (Supplementary Table 5 online). Neutralizing antibodies specific for autologous HIV-1 isolates were detected at low titers (1/10), the assay detection threshold in two individuals from day 0 to 1 year. They were undetectable in the 16 other participants, except transiently at low titers (1/10) in three patients (one at day 56 and day 112; two others at day 112 only).
 
Using a highly sensitive flow cytometry--based intracellular cytokine (ICC) assay (representative data from patient 13 in Fig. 2), we observed that HIV-1-specific interleukin-2 (IL-2)-expressing CD4+ T cells increased around threefold (P < 0.01) over 1 year in the 18 immunized participants (Fig. 3a,b). HIV-1-specific interferon-gamma (IFN-gamma) expressing CD4+ T cells increased around twofold (P < 0.01) (Fig. 3c,e), whereas HIV-1-specific IFN-gamma-expressing (memory) CD8+ T cells increased less than twofold (P < 0.01) with a peak increase at day 112 (Fig. 3d,e). Individual ICC data of the 18 immunized participants can be found in Supplementary Table 6 online. The 1-year PVL change was strongly correlated with the 1-year frequency of HIV-1-specific IL-2- (Fig. 3f) or IFN-gamma-expressing CD4+ T cells (Fig. 3g). The correlation of the 1-year PVL change with the 1-year frequency of IFN-gamma-expressing CD8+ T cells was much weaker (Fig. 3h).
 
Figure 2. Intracellular cytokine detection of T cells following stimulation with AT-2-inactivated HIV-1-pulsed DC (patient 13).
 
The data represent the percentage of total T cells secreting IL-2 or IFN-gamma. For the bottom three rows of graphs, the left column and middle columns represent data for CD4+ T cells secreting IL-2 or IFN-gamma, respectively, and the right column represents data for CD8+ T cells secreting IFN-gamma. The monoclonal antibody isotype controls for IL-2 or IFNgamma were less than 0.02% and 0.05% respectively. The background levels of IL-2-expressing CD4+ T cells, IFN-gamma-expressing CD4+ T cells and IFN-gamma-expressing CD8+ T cells in the presence of nonpulsed DCs alone were less than 0.01%, 0.01% and 0.03% respectively.
 
Figure 3. HIV-1-specfic T-cell immunity in the 18 immunized patients.
 
(a,b) Individual or mean ( s.e.m) (solid circles) HIV-1-specific IL-2-expressing CD4+ T cells at baseline and after immunization. (c) Individual HIV-1-specific IFN-gamma-expressing CD4+ T cells. (d) Individual HIV-1-specific IFN-gamma-expressing CD8+ T cells. (e) Mean ( s.e.m) HIV-1-specific IFN-gamma-expressing CD4+ (upward-facing triangles) or CD8+ (downward-facing triangles) T cells. (f,g) Correlation between 1-year changes of PVL from baseline and HIV-1-specific IL-2-expressing or IFN-gamma-expressing CD4+ T cells by 1 year. (h) Correlation between HIV-1-specific IFN-gamma-expressing CD8+ T cells by 1 year and 1-year changes of PVL from baseline. **P < 0.01.
 
Using the combination of antigen-specific human leukocyte antigen (HLA) tetramer binding and four-color flow cytometry in the 10 HLA-A*0201-positive participants, we showed that HIV-1 gag--specific CD8+ T cells increased more than threefold at day 112 (P < 0.05) and remained increased up to 1 year thereafter, whereas the percentage of HIV-1 gag--specific CD8+ T cells expressing perforin (effectors) increased around twofold. By 1 year, there was no correlation between the percentage of gag-specific CD8 cells and the PVL change, whereas a significant linear correlation was observed between the percentage of HIV-1 gag--specific CD8+ T cells expressing perforin and the PVL change. A strong correlation was observed at 1 year between the frequency of HIV-1-gag-specific perforin-expressing CD8+ T cells and the frequency of HIV-1-specific IL-2-expressing CD4+ T cells (Fig. 5g) or the frequency of HIV-specific IFN-gamma-expressing CD4+ T cells. The two individuals who had the highest percentage of virus-specific CD4+ T cells expressing IL-2 by 1 year were the same as those who had the highest percentage of virus-specific CD4+ T cells expressing IFN-gamma. After removing these two individuals from the plot, the perforin-expressing gag-specific CD8+ T cells still correlated significantly with the IL-2-expressing virus-specific CD4+ T cells (r2 = 0.741, P < 0.01) but did not correlate significantly with IFN-gamma-expressing virus-specific CD4+ T cells (r2 = 0.311, P = 0.149).
 
To determine whether any initial biological parameter predicted the 1-year PVL response, we performed a correlation analysis between all initial parameters that we measured and the 1-year PVL change. Baseline PVL showed no correlation with 1-year PVL change. In contrast, the baseline CD4 cell-count correlated positively with 1-year decrease of PVL (r2 = 0.263, P = 0.029) and baseline HIV-1-specific IL-2-expressing CD4+ T cells showed also a (marginal) positive correlation with 1-year decrease of PVL (r2 = 0.169, P = 0.089). On the other hand, HIV-specific IFN-gamma-expressing CD4+ T cells, HIV-specific IFN-gamma-expressing CD8+ T cells, or the major maturation marker (CD83) of DCs did not correlate with 1-year PVL decrease.
 
Methods
 
Participants.
The National Ethical Committee of the Brazilian Ministry of Health approved this phase 1-2 trial. HIV-1-positive volunteers gave written informed consent before being enrolled in this study. The following criteria had to be met 3 months before the first vaccination: age of 18 years, absence of pregnancy, HIV-1 seropositivity for 1 year, no clinical AIDS, absence of antiretroviral therapy for 6 months, hemoglobin 10 g/dl, platelets 100,000 cells/mul, blood CD4+ cells 300 cells/mul, plasma viral RNA 10,000 copies/ml and absence of other chronic diseases. Initially, 20 chronically HIV-1-infected participants were enrolled in this study. Two participants were excluded before being vaccinated: one because his baseline plasma viral load was below 5,000 copies/ml, as determined by repeated viral-load measurements; the other because she had a platelet count below 100,000 cells/mul at day 0. A total of 18 participants were thus included in the present study.
 
Vaccine.
Autologous HIV-1 isolates, which were obtained from CD8-depleted PBMCs by viral culture, were inactivated by AT-2 (Sigma) as described37 under GMP (good manufacturing practices) conditions. The complete inactivation of each of the viral isolates was confirmed by the negative results of a 1-month culture in HIV-negative donor PBMCs. One week before immunization, 1--3 times 1010 PBMCs were collected by a 3-h leukapheresis from each patient and then transferred to a biosafety (level P3) clean room for a standardized 7-d culture. Briefly, we subjected freshly collected PBMCs to plastic adherence at a density of 1 times 106 cells/cm2 in the presence of 0.5% of clinical-use human serum albumin (Laboratoire Fran¸ais du Fractionnement et des Biotechnologies). After a 2-h incubation at 37 °C in 5% CO2, we removed nonadherent cells by rinsing with sterile PBS buffer. Adherent cells were then cultured for 5 d in clinical grade CellGro DC medium (CellGenix) containing 2000 U/ml GM-CSF (Schering-Plouph) and 50 ng/ml IL-4 (CellGenix). At day 5, we exposed 6 times 107 DCs to AT-2-inactivated autologous virus (1 times 109 viral particles/ml) at 37 °C for 2 h and froze the remaining nonpulsed DCs in liquid nitrogen. After two washes to remove free virus, cells were cultured for an additional 2 d in the DC medium supplemented with clinical-grade cytokines IL-1beta (10 ng/ml, CellGenix), IL-6 (100 ng/ml, CellGenix), and TNF-alpha (50 ng/ml, CellGenix). At day 7, we performed quality control with flow cytometry; 3 times 107 quality control--approved viable DCs were resuspended in 1 ml of sterile 0.9% chloride sodium solution and were ready for injection. Each patient received four subcutaneous injections of 0.25 ml in close proximity to left and right axillary and inguinal areas. At weeks 2 and 4, we administered two booster injections with the same number of viable virus-loaded autologous DCs. These DCs were processed from the remaining frozen nonpulsed DCs, which were pulsed with AT-2-inactivated HIV-1 upon thawing just prior to the last 2-d culture.
 
Viral load assays.
We measured plasma HIV-1 RNA using the Monitor kit (Roche Diagnostics). Cell-associated HIV-1 DNA or supernatant HIV-1 RNA was quantified as previously described37, 45.
 
Anti-HIV-1 and neutralizing antibody assays.
We determined serum titers of antibody specific for HIV-1 using a limiting dilution assay with a commercial ELISA kit (GENSCREEN HIV1/2 version 2, Bio-Rad). The titers were determined as reciprocals of the last serum dilutions showing a >=50% positivity by ELISA. We performed neutralization against infection of healthy donor PBMCs by autologous virus in participants' sera as described18. Control cells were exposed to the same infectious dose of autologous virus or to the same virus dose pretreated with the sera taken from five healthy donors.
 
Flow cytometry assays.
Intracellular IL-2- and IFN-gamma-expressing CD4+ T cells or IFN-gamma-expressing CD8+ T cells following ex vivo stimulation with autologous DCs pulsed with AT-2-inactivated autologous virus were detected by previously described ICC assay. The ICC kits and all monoclonal antibodies (including corresponding isotype controls) for the quality control of DCs were purchased from BD Bioscience Europe. We determined cell viability by intracellular DNA staining with propidium iodide (20 mug/ml, Sigma). AT-2-inactivated HIV-1-loaded autologous DCs used as stimulating cells for ICC assay were the same as those used for the booster immunizations. We measured HIV-1-specific CD8+ effector T cells using a combination of HLA-A*0201 HIV gag (amino-acid sequence: SLYNTVATL) tetramer-phycoerythrin (Beckman Coulter) and monoclonal antibodies specific for perforin-FITC, CD8-PerCP-Cy5.5 and CD3-APC (BD Biosciences Europe). The four-color flow cytometry studies were performed on FACSCalibur (BD Biosciences). We determined the specific staining by subtracting the nonspecific staining of isotype controls and determined the gag-tetramer-specific staining by subtracting the background tetramer binding in normal HLA A*0201 donor T cells.
 
Statistical analysis.
Paired data before and at different time points after immunization were compared by the Wilcoxon test.
 
Note: Supplementary information is available on the Nature Medicine website.
 
 
 
 
 
  icon paper stack View Older Articles   Back to Top   www.natap.org