In recent years there has been an intense debate about the nature of HIV pathogenesis. It has been argued that CD4 depletion is primarily a result of infection and subsequent cell death caused by HIV virions that homeostatic mechanisms are eventually unable to correct. 1,2 Others have emphasized that while HIV does infect some CD4 lymphocytes, its primary mechanism of pathogenesis is by causing immune activation of both CD4 and CD8 cells, resulting in apoptosis of this activated population. 3–5 Evidence in support of the latter theory has been provided by measuring the proliferation of cells in vivo using radioactive labels such as bromodeoxyuridine (BrdU) 5 or deuterated glucose. 6–8 These studies demonstrated reductions in not only CD4 but also CD8 cell proliferation in HIV-infected subjects when the HIV viral load was reduced by drug therapy. Further support for the importance of immune activation in HIV pathogenesis has been provided by the study of simian immunodeficiency (SIV) in different simian species. Sooty mangabeys and African green monkeys, which are the natural hosts of SIV, tolerate high SIV viral loads yet maintain relatively normal CD4 counts and live normal lifespans. 9,10 By contrast, macaques infected with SIV undergo a course of infection similar to that in humans, with high viral loads resulting in CD4 cell depletion, illness, and death. BrdU labeling studies have demonstrated that the sooty mangabeys have low rates of both CD4 and CD8 cell turnover when compared with the macaques, 11 suggesting that they have developed mechanisms to avoid immune-activated cell death.
Studies in patients infected with HIV have demonstrated that immune activation, determined by the expression of CD38 on CD8+ T lymphocytes, is associated with disease progression 12–16 and that effective antiretroviral therapy results in a decline in the expression of this marker in parallel with the fall in HIV viral load. 17 These findings suggest that HIV drives immune activation, although some authors have stressed the role of additional infections that might contribute to HIV-induced activation. 18,19
One serious drawback of previous studies has been the exclusive examination of blood T lymphocytes. We included the lung in our study for two reasons: First, previous studies have demonstrated that the lung is a site of HIV replication 20–23 and therefore the investigation of this organ in addition to the blood may provide a more closely associated picture between viral replicative events and immune activation than the examination of blood alone. Second, the role of additional respiratory pathogens in stimulating immune activation could be assessed by comparison with HIV-positive subjects in whom no bronchoalveolar lavage (BAL) pathogen was identified.
In this study we examined both the activation and proliferation of CD8+ memory T cells in both lung and blood. Activation was determined by CD38 expression on CD8+ lymphocytes using a simple, reproducible gating strategy. Proliferation was estimated in both the CD38bright and CD38dim CD8+ populations using Ki67, a nuclear marker associated with dividing cells. 24,25
HIV-infected patients undergoing bronchoscopy for suspected respiratory disease were invited to take part in the study, which was approved by the hospital ethics committee. Thirty-five HIV-positive patients were tested and a control group of 5 healthy individuals was also examined. The HIV cohort mostly consisted of patients with advanced disease, with only 5 patients on antiretroviral therapy. The median CD4 count of the HIV-positive patients was 75 cells/μL (interquartile range [IQR]: 11–265) and median HIV viral load was 164,000 copies/mL (IQR: 49,300–413,000) at the time of investigation. The demographic characteristics, CD4 counts, HIV viral loads, and BAL diagnoses of the study population are depicted in Table 1.
Determination of HIV Viral Load
HIV viral loads were determined by the automated Amplicor polymerase chain reaction (PCR) (Cobas Amplicor; Roche Diagnostics, Basel, Switzerland). The minimal level of detection was 50 HIV copies/mL and the upper limit 750,000 copies/mL. Viral load was not determined in BAL for the following reason. The procedure of BAL involves the instillation of large volumes of saline (typically 180–200 mL) with the bronchoscope wedged into a subsegmental bronchus. The return of both fluid and cells is highly variable and dependent on operator technique, patient tolerability, and the pathologic state of the lungs. For these reasons, BAL HIV viral loads are unreliable and unstandardized and were not performed in this study.
Standard Investigations for Respiratory Pathogens in BAL
All BAL samples from HIV-positive individuals were investigated for the presence of respiratory viruses, including influenza, parainfluenza, adenovirus, and respiratory syncytial virus. Cytomegalovirus direct antigen fluorescent foci test and PCR were performed in those with CD4 counts < 100 cells/μL. Culture of BAL for bacteria, fungi, and mycobacteria was undertaken and in addition, a stained specimen was examined by a cytopathologist for the presence of Pneumocystis jirovecii, acid fast bacilli, and fungal pathogens.
Bronchoscopy and Sample Preparation
Fiberoptic bronchoscopy was performed according to established guidelines. 26 BAL was performed from an area of radiologically abnormal lung; otherwise standard right middle lobe lavage was performed. The samples were kept on ice and then divided and aliquots sent to relevant laboratories for pathologic investigations. The remaining BAL was kept for immunologic analysis. The BAL samples were filtered using a 100-μm filter (Celltrics, Partec, GmBH, Münster, Germany) and washed in phosphate-buffered saline. At the time of bronchoscopy, 5 mL of ethylenediamine tetra-acetic acid peripheral blood was also taken.
Flow Cytometry and Gating Strategies
A new diagnostic protocol to determine the proportions of lymphocytes, granulocytes, and alveolar macrophages in BAL was performed as previously described. 27 In addition, BAL CD4/CD8 ratios were analyzed by staining specimens with a panel of CD3, CD4, and CD8 monoclonal antibodies in a similar manner. These procedures were carried out using a volumetric flow cytometer (Cytoron absolute, Raritan, NJ) that enabled the absolute counts of CD4 and CD8 lymphocytes to be determined.
An aliquot of BAL containing 1 × 105 CD8+ lymphocytes and a sample of blood containing the same number of cells were then fixed and permeabilized as previously described. 28 No lysis step was included as adequate decanting during the fixation and permeabilization stage removed nearly all red cells. Following this procedure, the separate samples were stained at 4°C for 30 minutes, followed by a wash step. The following monoclonal antibodies were used in pretitrated optimal concentrations in a single 4-color panel: Ki67 FITC (Immunotech, Marseilles, France), CD38 PE (Caltag Medsystems, Towcester, UK), CD8 PECy7 (Caltag Medsystems), and CD45RA APC (clone sn130, Southern Biotech, Birmingham, AL). Stained blood and BAL specimens were run on a FACSCalibur flow cytometer (Becton Dickinson, San Jose, CA). A total of 20,000 CD8+ lymphocytes were acquired and the list mode data were analyzed using Winlist 4.0 software (Verity, Inc., Topsham, VA). Primary immunologic gating of CD8+ lymphocytes was performed and these events were confirmed to lie within a lymphoid scatter gate. These CD8+ cells were further investigated in terms of their naive/memory phenotype by their CD45RA isoform expression (Fig. 1).
Since CD45RA− memory CD8+ lymphocytes formed most of BAL lymphocytes with only a small percentage of CD45RA+ naive/revertant cells, we investigated the degree of activation in the memory pool in both blood and BAL compartments using CD38 expression. Decisions regarding placement of the gate to differentiate between CD38+ and CD38− CD8 cells were determined from staining of control blood. The majority of CD8+ T cells in healthy individuals are CD38dim with only a few activated cells, although natural killer cells express this marker brightly. Therefore, we elected to draw a gate that would differentiate between these predominantly CD38dim CD8+ lymphocytes and the CD38bright natural killer cells. Using this method on blood from healthy controls that had been fixed and permeabilized, the gate was drawn at log 1,3 mean fluorescence intensity CD38 expression (Fig. 2). The proportions of CD38bright and CD38dim CD8+ cells in BAL and blood from the HIV-positive study population and the normal controls were then determined by the same cut-off.
Lastly, the proportions of memory CD8+ lymphocytes that expressed the marker of cell cycling Ki67 were determined in the CD38bright and CD38dim subpopulations (Fig. 1).
Median values and IQRs were expressed in the text. Nonparametric analysis by the Mann-Whitney method was used to compare the data sets.
Diagnoses in the HIV-Positive Patients With Respiratory Disease and BAL Lymphocyte Percentages
In 19 HIV-positive patients a respiratory pathogen was identified in BAL. The diagnoses were 9 culture-confirmed cases of tuberculosis (TB), 7 Pneumocystis carinii pneumonias (PCP), 1 bacterial pneumonia, and 1 cytomegalovirus (CMV) infection. In addition, 1 patient had multiple infections with TB, PCP, and CMV simultaneously. In 19 HIV-positive patients, no pathogens were identified in BAL. Three subjects in this group were excluded because flow cytometry demonstrated a marked BAL neutrophilia suggestive of a bacterial lung infection, despite failure to culture an organism. Sixteen HIV-positive patients were therefore included in this group. The BAL lymphocyte percentages were highly variable (Table 1). When compared with the control subjects, BAL from the HIV subjects without respiratory disease contained a lymphocytosis (median 26.1 vs. 10.2%, P = 0.06). This BAL lymphocytosis was more marked in the patients with respiratory disease (median 38.7%, P = 0.02 compared with control values).
CD45 Isoform Expression of CD8+ T Lymphocytes in BAL and Blood in HIV-Positive Patients and Control Subjects
BAL CD8+ T lymphocytes from HIV-positive patients were overwhelmingly of a memory phenotype with a median of 97.5% CD45RA− (IQR: 96.1–97.9). There was no significant difference in the proportion of CD45RA− phenotype between those patients with respiratory pathogens isolated in BAL (median 97.6%) and those in whom no pathogens were isolated (median 97.1%). In the control patients, slightly fewer (median 91.8%, IQR: 86.4–97.2%) of BAL lymphocytes were CD45RA−. Therefore, BAL in both the HIV-infected patients and the control subjects contained predominantly memory CD8+ T lymphocytes.
Peripheral blood from the HIV-positive patients contained a median value of 56.3% (IQR: 48.5–73.1%) of CD45RA−CD8+ T lymphocytes. The corresponding value in blood from the control patients was 36.9% (IQR: 32.1–42.0%).
CD38 Expression in CD45RA− CD8+ Lymphocytes From BAL and Blood of HIV-Positive Patients and Control Subjects
The proportions of CD38bright CD8+ T lymphocytes were examined in both blood and BAL. The results were stratified according to the blood HIV viral load values between undetectable to 100,000 copies/mL (n = 14) and >100,000 copies/mL (n = 21) and these were compared with the CD38 activation status in the control subjects (Fig. 3). The lower limit of detection of the assay was 50 copies/mL, with an upper limit of detection of 750,000 copies/mL. Four patients had unspecified HIV viral loads above this level and 3 had values < 50 copies/mL.
Higher percentages of CD38bright CD8+ T lymphocytes in both blood and BAL were associated with higher blood HIV viral loads (Fig. 3). For the patients in the lower category of viral load, the median percentage of CD38bright CD8+ lymphocytes was 37.8% (IQR: 29.2–63.6%) in BAL and 30.3% in blood (IQR: 29.2–54.8%). In the higher HIV viral load category, the CD8+ lymphocytes were more activated in both BAL and blood compartments, with median values of 69.7% (IQR: 44.6–83.8%) and 75.9% (IQR: 65.2–87.9%), respectively. When the control patients were examined, the percentages of activated CD8 lymphocytes were much lower than in the HIV-positive patients in both BAL (median 4.5%, IQR: 3.2–5.7%) and blood (median 12.3%, IQR: 4.2–14%). There was a statistically significant difference in CD8 activation measured by CD38 expression between the control values and those in the lower viral load category in both lung (P = 0.005) and blood (P = 0.0.18). A significantly higher CD38 expression was also determined when the lower and higher HIV viral load categories were compared in both BAL (P = 0.006) and blood (P = 0.003).
CD38 Expression in CD45RA− CD8+ Lymphocytes From BAL of HIV-Positive Patients With and Without Respiratory Pathogens
Since higher viral loads were associated with lower CD4 counts and increased rates of pathogens detected in the lung, we repeated our analysis comparing the proportion of activated CD8+ T lymphocytes for the same HIV viral load categories in BAL from HIV-positive patients with respiratory disease and those in whom no pathogens were identified in BAL (Fig. 4). The aim for this analysis was to assess the relative contributions of respiratory pathogens and of HIV viral load in inducing CD8 cell activation.
In the patients in whom BAL revealed no pathogens, the median values for CD38bright CD8+ T lymphocytes in BAL were 37.7 and 68.2% for the low and high viral load groups, respectively. In the HIV-positive patients with respiratory pathogens, the corresponding values were 49.9 and 81.0%, respectively. Despite a trend toward higher CD38 expression in the subjects with respiratory co-infections, the differences were not statistically significant either for the low HIV viral load groups (P = 0.62) or for the higher viral load groups (P = 0.50). These data suggest that HIV viral load is the most significant factor in stimulating CD8 lymphocyte activation.
Expression of Ki67 in Activated and Unactivated CD8+ Lymphocytes in Lung and Blood
We next investigated the relationship between CD38 activation and CD8 lymphocyte proliferation by comparing the proportions of Ki67+ CD8+ cells in the CD38bright and CD38dim populations in both BAL and blood (Fig. 5). Activated, CD38bright CD8+ T lymphocytes were associated with higher percentages of Ki67+ cells in BAL (median 2.37, IQR: 1.65–3.98%) than in the CD38dim CD8+ lymphocytes (median 1.10%, IQR: 0.38–1.535%). Increased percentages of Ki67+ CD8 lymphocytes in the activated cells were also documented in blood (median 1.48%, IQR: 0.77–3.03%) when compared with the unactivated cells (median 0.04%, IQR: 0–0.29%). These difference between the Ki67+ populations in the activated and unactivated CD8+ lymphocytes were highly significant for both compartments (P ≤ 0.0001). In the control subjects; high proportions of Ki67+ CD8+ cells were noted among the rarer CD38bright population in BAL (median 7.68, IQR: 1.96–11.5%), while in the predominant CD38dim component, Ki67 expression was very low (median 0.12%, IQR: 0–0.54%). Ki67+ CD8 cells in blood from the controls were very low both in the CD38bright (median 0.24%) and the CD38dim (median 0.01%) populations.
In this study we have investigated the relationship between blood HIV viral load and the features of activation and cell cycling of memory CD8+ T lymphocytes in both blood and lung. We have demonstrated as expected that activated CD8+ T lymphocytes, identified by their particularly high CD38 expression in both lung and blood compartments, have higher rates of cell cycling when compared with the CD38dim CD8+ cells. Moreover, this high level of CD8 activation and cycling detected at both sites was correlated with the blood HIV viral load.
Early studies noted that CD38 expression on CD8 lymphocytes was associated with accelerated HIV disease progression. 12–14,29 However, this finding was most clearly documented in the CD45RO+ memory CD8+ pool. 14 The demonstration that activated memory, but not naive CD8+ lymphocytes, is correlated with CD4 decline is not surprising since it is well known that it is the memory population that responds to antigen by proliferation and a change in various phenotypic and functional markers. 30–32
While previous studies have demonstrated increased lymphocyte proliferation in HIV-infected humans and SIV-infected animals using Ki67 expression or radiolabeling, this is the first investigation to study the correlation between immune activation and cell proliferation in both the blood and a relevant tissue compartment, the lung. The decay rates of lymphocytes radiolabeled with BrdU or deuterated glucose provide evidence of increased proliferation or death within the total lymphocyte pool, although the interpretation of the results of some studies has been questioned in the light of immunologic models of lymphocyte activation. 4 In this study we have demonstrated that activated CD38bright memory CD8+ cells have significantly higher rates of cell cycling as measured by Ki67 expression in both blood and lung. This finding, in conjunction with the relationship between HIV viral load and CD8 activation, is direct evidence for HIV in promoting increased proliferation. A caveat to this assertion is that some investigators have questioned the validity of Ki67 as a marker of the fraction of cells proliferating. 16 These authors have demonstrated that nearly half of the Ki67+ CD4+ lymphocytes also expressed CTLA-4, a marker for activated cells arrested at the G1 stage of the cell cycle. 33 However, Ki67 expression in CD8+ lymphocytes may be a more reliable measure of the fraction of cells proliferating since only 10% of Ki67+ lymphocytes also co-expressed CTLA-4. 16
We have not measured the HIV viral load in the lung since the process of BAL introduces a highly variable dilution factor rendering quantitative HIV viral loads difficult to interpret. Nevertheless, our data support the findings of previous investigators that the lung is a site of HIV replication. 20–23 It has also been demonstrated that purified CD8+ lymphocytes from the lung of HIV-infected subjects have high rates of spontaneous apoptosis. 34
Since our cohort included HIV-infected patients both with and without respiratory disease, we were able to further investigate the relationship between HIV and CD8 lymphocyte activation and to consider whether co-infections with respiratory pathogens could play an important role in this process. Although we demonstrated no significant difference between the CD38bright CD8+ lung lymphocytes in the patients with respiratory pathogens when compared with those without pathogens in the 2 HIV viral load categories, there was a trend for slightly higher rates of BAL CD8 activation in the patients co-infected with respiratory pathogens (Fig. 4). We therefore conclude that HIV plays a primary role in driving cellular activation and that respiratory co-pathogens provide a more minor contribution.
However, one limitation of this study was that the patients in whom no respiratory pathogens were found in BAL may possibly have had occult infections, despite a rigorous search using culture, PCR, and microscopy, in addition to the exclusion of patients with a BAL neutrophilia who were judged likely to have a bacterial co-infection. A further caveat is that we cannot conclude that co-infections may not exacerbate immune activation, since in our cohort most patients had either TB or PCP and we would need to examine a larger sample with a wider range of respiratory pathogens. Nevertheless, it is interesting to note that the patient in whom CMV was the only respiratory pathogen found did not have especially activated BAL CD8+ lymphocytes.
In summary, we have developed a simple, reliable gating strategy for determining the CD38 activation status of CD8+ lymphocytes. Using such a system we have shown that CD8 cell activation is related to the blood HIV viral load in both blood and an important tissue site, the lung. The primary role of HIV in stimulating this immune activation is strengthened by the demonstration that respiratory co-infections did not significantly increase the CD8 activation state of lung CD8+ lymphocytes when compared with those with matched HIV viral loads without respiratory pathogens. Lastly, we have also documented that more activated, CD38bright CD8 cells expressed the cell cycling marker Ki67 than the CD38dim cells. Taken together, these findings are consistent with the model that HIV drives cell activation and proliferation.
1. Ho DD, Neumann AU, Perelson AS, et al. Rapid turnover of plasma virions and CD4 lymphocytes in HIV-1 infection. Nature. 1995; 373:123–126.
2. Mohri H, Bonhoeffer S, Monard S, et al. Rapid turnover of T lymphocytes in SIV-infected rhesus macaques. Science. 1998; 279:1223–1227.
3. Hazenberg MD, Stuart JW, Otto SA, et al. T-cell division in human immunodeficiency virus (HIV)-1 infection is mainly due to immune activation: a longitudinal analysis in patients before and during highly active antiretroviral therapy (HAART). Blood. 2000; 95:249–255.
4. Grossman Z, Meier-Schellersheim M, Sousa AE, et al. CD4+ T-cell depletion in HIV infection: are we closer to understanding the cause? Nat Med. 2002; 8:319–323.
5. Kovacs JA, Lempicki RA, Sidorov IA, et al. Identification of dynamically distinct subpopulations of T lymphocytes that are differentially affected by HIV. J Exp Med. 2001; 194:1731–1741.
6. Deeks SG, Hoh R, Grant RM, et al. CD4+ T cell kinetics and activation in human immunodeficiency virus-infected patients who remain viremic despite long-term treatment with protease inhibitor-based therapy. J Infect Dis. 2002; 185:315–323.
7. Mohri H, Perelson AS, Tung K, et al. Increased turnover of T lymphocytes in HIV-1 infection and its reduction by antiretroviral therapy. J Exp Med. 2001; 194:1277–1287.
8. Ribeiro RM, Mohri H, Ho DD, et al. In vivo dynamics of T cell activation, proliferation, and death in HIV-1 infection: why are CD4+ but not CD8+ T cells depleted? Proc Natl Acad Sci U S A. 2002; 99:15572–15577.
9. Kaur A, Grant RM, Means RE, et al. Diverse host responses and outcomes following simian immunodeficiency virus SIVmac239 infection in sooty mangabeys and rhesus macaques. J Virol. 1998; 72:9597–9611.
10. Broussard SR, Staprans SI, White R, et al. Simian immunodeficiency virus replicates to high levels in naturally African green monkeys without inducing immunologic or neurologic disease. J Virol. 2001; 75:2262–2275.
11. Kaur A, Barabasz A, Rosenzweig M, et al. Dynamics of T-lymphocyte turnover in sooty mangabeys, a non-pathogenic host of simian immunodeficiency virus infection. Paper presented at: 9th
Conference on Retroviruses and Opportunistic Infections; 2002; Seattle.
12. Levacher M, Hulstaert F, Tallet S, et al. The significance of activation markers on CD8 lymphocytes in human immunodeficiency syndrome: staging and prognostic value. Clin Exp Immunol. 1992; 90:376–382.
13. Giorgi JV, Liu Z, Hultin LE, et al. Elevated levels of CD38
+ CD8+ T cells in HIV infection add to the prognostic value of low CD4+ T cell levels: results of 6 years of follow-up. The Los Angeles Center, Multicenter AIDS Cohort Study. J Acquir Immune Defic Syndr. 1993; 6:904–912.
14. Bofill M, Mocroft A, Lipman M, et al. Increased numbers of primed activated CD8+CD38
+CD45RO+ T cells predict the decline of CD4+ T cells in HIV-1 infected patients. AIDS. 1996; 10:827–834.
15. Liu Z, Cumberland WG, Hultin LE, et al. Elevated CD38
antigen expression on CD8+ T cells is a stronger marker for the risk of chronic HIV disease progression to AIDS and death in the Multicenter AIDS Cohort Study than CD4+ cell count, soluble immune activation markers, or combinations of HLA-DR and CD38
expression. J Acquir Immune Defic Syndr Hum Retrovirol. 1997; 16:83–92.
16. Leng Q, Borkow G, Weisman Z, et al. Immune activation correlates better than HIV plasma viral load with CD4 T-cell decline during HIV infection. J Acquir Immune Defic Syndr. 2001; 27:389–397.
17. Tilling R, Kinloch S, Goh LE, et al. Parallel decline of CD8+/CD38
++ T cells and viraemia in response to quadruple highly active antiretroviral therapy in primary HIV infection. AIDS. 2002; 16:589–596.
18. Bentwich Z, Kalinkovich A, Weisman Z. Immune activation is a dominant factor in the pathogenesis of African AIDS. Immunol Today. 1995; 16:187–191.
19. Cohen Stuart JW, Hazebergh MD, Hamann D, et al. The dominant source of CD4+ and CD8+ T-cell activation in HIV infection is antigenic stimulation. J Acquir Immune Defic Syndr. 2000; 25:203–211.
20. Chayt KJ, Harper ME, Marselle LM, et al. Detection of HTLV-III RNA in lungs of patients with AIDS and pulmonary involvement. JAMA. 1986; 256:2356–2359.
21. Linnemann Jr, CC Baughman RP, Frame PT, et al. Recovery of human immunodeficiency virus and detection of p24 antigen in bronchoalveolar lavage
fluid from adult patients with AIDS. Chest. 1989; 96:64–67.
22. Twigg HL, Soliman DM, Day RB, et al. Lymphocytic alveolitis, bronchoalveolar lavage
viral load, and outcome in human immunodeficiency virus infection. Am J Respir Crit Care Med. 1999; 159:1439–1444.
23. Semenzato G, Agostini C, Chieco-Bianchi L, et al. HIV load in highly purified CD8+ T cells retrieved from pulmonary and blood compartments. J Leukoc Biol. 1998; 64:298–301.
24. Gerdes J, Lemke H, Baisch H, et al. Cell cycle analysis of a cell proliferation-associated human nuclear antigen defined by the monoclonal antibody Ki-67. J Immunol. 1984; 133:1710–1715.
25. Schwarting R, Gerdes J, Niehus J, et al. Determination of the growth fraction in cell suspensions by flow cytometry using the monoclonal antibody Ki-67. J Immunol Methods. 1986; 90:65–70.
26. British Thoracic Society guidelines on diagnostic flexible bronchoscopy. Thorax. 2001; 56:i1-21.
27. Barry SM, Condez A, Johnson MA, et al. Determination of bronchoalveolar lavage
leukocyte populations by flow cytometry in patients investigated for respiratory disease. Cytometry. 2002; 50:291–297.
28. Kappelmayer J, Gratama JW, Karaszi E, et al. Flow cytometric detection of intracellular myeloperoxidase, CD3 and CD79a: interaction between monoclonal antibody clones, fluorochromes and sample preparation protocols. J Immunol Methods. 2000; 242:53–65.
29. Liu Z, Hultin LE, Cumberland WG, et al. Elevated relative fluorescence intensity of CD38
antigen expression on CD8+ T cells is a marker of poor prognosis in HIV infection: results of 6 years of follow-up. Cytometry. 1996; 26:1–7.
30. Akbar AN, Terry L, Timms A, et al. Loss of CD45R and gain of UCHL1 reactivity is a feature of primed T cells. J Immunol. 1988; 140:2171–2178.
31. Hamann D, Baars PA, Rep MH, et al. Phenotypic and functional separation of memory and effector human CD8+ T cells. J Exp Med. 1997; 186:1407–1418.
32. Sallusto F, Lenig D, Forster R, et al. Two subsets of memory T lymphocytes with distinct homing potentials and effector functions. Nature. 1999; 401:708–712.
33. Brunner MC, Chambers CA, Chan FK, et al. CTLA-4-mediated inhibition of early events of T cell proliferation. J Immunol. 1999; 162:5813–5820.
34. Agostini C, Siviero M, Facco M, et al. Antiapoptotic effects of IL-15 on pulmonary Tc1 cells of patients with human immunodeficiency virus infection. Am J Respir Crit Care Med. 2001; 163:484–489.