ABSTRACT
Streptococcus pneumoniae infections are associated with considerable morbidity and mortality throughout the world. The immunopathology is characterized by an intense inflammatory reaction, including a strong acute-phase response and increased numbers of neutrophils in the circulation. However, little is known regarding the T-cell response during in vivo infections in humans. The purpose of this study was to test the hypothesis that activated T cells producing type 1 cytokines were engaged in the host response to pneumococcal infections. The phenotype and function of T cells were studied in 22 patients at admission to a department of infectious diseases and after antibiotic treatment for 1 week compared with an age-matched, healthy control group. Pneumococcal infections induced lymphopenia in the circulation due to the disappearance of activated T lymphocytes with a type 1 cytokine profile. In contrast, the numbers of naive T cells and interleukin-4-producing T cells did not change. Activated type 1 cytokine-producing cells reappeared in the circulation in relation to the treatment and clinical improvement. The underlying mechanisms during infection may include sequestration in the peripheral tissues and/or apoptosis. In fact, increased activation-induced apoptosis in the remaining peripheral lymphocytes and elevated levels of soluble Fas ligand were detected at admission to the hospital. In conclusion, these data suggest that activated T lymphocytes with a type 1 cytokine profile are highly engaged in the in vivo immune response to S. pneumoniae.
Streptococcus pneumoniae is the most common pathogen causing community-acquired pneumonia and otitis media, and it may lead to septicemia and meningitis, with a high risk of mortality despite sufficient antibiotic treatment. The highest rates of invasive infections are seen in infants, elderly individuals (20), and immunocompromised patients (26). Pneumococcal infections together with influenza are thus the fifth leading cause of death among persons over 65 years in the Western world (4). The effectiveness of the available polyvalent polysaccharide pneumococcal vaccine is questionable (19). Accordingly, throughout the world, S. pneumoniae infections are associated with considerable morbidity and mortality.
The immunopathology of pneumococcal infections is characterized by an intense inflammatory reaction, including a strong acute-phase response and leucocytosis due to increased numbers of neutrophiles (25). The fact that human immunodeficiency virus (HIV) patients (26) and splenectomized patients (9) are at especially high risk for developing severe, invasive pneumococcal infections suggests that T cells may be key players in the immune response. However, to our knowledge, the T-cell response in humans during in vivo infections has not previously been described. It has recently been shown that T cells are important for help with the induction of a humoral immune response against S. pneumoniae (24). Furthermore, patients with interleukin-12 (IL-12) deficiency develop more severe pneumococcal infections, suggesting that the lack of IL-12 for stimulation of a type 1 cytokine response characterized by IL-2 and gamma interferon (IFN-γ) production has a crucial role in the outcome of the infection (10). In accordance with this, it has been suggested that IFN-γ produced by the NK or T cell is of importance in the host defense against S. pneumonia (23). In vitro experiments have shown that Streptococcus pyogenes strains induce production of IFN-γ (type 1 response), but only weakly induce production of IL-4 (16, 18). Furthermore, IFN-γ rather than IL-4 is the dominant cytokine elicited in tonsils in vivo (1). Accordingly, a type 1 cytokine response may be important for the clearance of streptococcal infections.
The balance between T-cell proliferation and T-cell death may also be crucial for the host response. Soluble FasL (sFasL) is produced by activated lymphocytes and may induce apoptosis (2). However, sFasL can also protect cells from apoptosis by inhibiting Fas/FasL ligation (5), and may be one explanation of why spontaneous apoptosis among pulmonary neutrophiles during community-acquired pneumonia is decreased (8). Fas/FasL seems to be a system of importance in maintaining homeostasis among T cells during various infections (12) and is further involved in the tissue response in S. pneumoniae infections in mice (17).
Studies of other acute infections as well as chronic infections have shown a disappearance of certain lymphocyte subsets from the peripheral blood (3, 7, 22) concomitant with increased susceptibility to activation induced cell death in the remaining peripheral lymphocytes (14, 21). It is unclear if this lymphopenia is due to sequestering of cells in lymphoid tissue or is a consequence of the increased apoptosis. Chemotherapy of bacterial, viral, and parasitic infections induces a subsequent increase in the percentages and absolute numbers of T cells in the peripheral blood (3, 6, 7, 11), suggesting a release of T cells concomitant with the clearance of the infectious agents. Accordingly, T cells isolated during treatment seem to offer a convenient opportunity to examine functional aspects of different subsets that have been engaged in the in vivo immune response to the infection.
The purpose of the present study was to test the hypothesis that activated T cells producing type 1 cytokines were involved in the host response to pneumococcal infections in vivo in humans. Furthermore, we hypothesized that these cells were prone to undergo apoptosis when released to the peripheral blood during treatment. Thus, we investigated the percentages and absolute numbers of naïve, memory or activated, and cytokine-producing T cells in 22 hospitalized patients with pneumococcal infections during the first week of chemotherapy and compared them with those of an age-matched healthy control group. Furthermore, we measured the frequencies of apoptosis among peripheral lymphocytes and levels of sFasL in serum.
MATERIALS AND METHODS
Patients.Twenty-two consecutive patients (10 men, 12 women) with infections caused by S. pneumoniae in a department of infectious diseases were included. The median age was 68 years (range, 37 to 91 years). None of the patients was HIV positive or had cancer. Eleven patients had bacteremia, including 6 patients with meningitis, but none had multiple organ failure. The remaining patients had noninvasive pneumonia. A microbiologic diagnosis of S. pneumoniae infection was obtained in 19 patients. The remaining three patients were included because they had radiologically verified lobar pneumonia, a high level of C-reactive protein (CRP) at admission, and an immediate clinical response to treatment with penicillin G. The median length of hospitalization was 13 days (range, 2 to 38 days). Blood samples were collected during the first 24 h after admission (day 0) and after 7 days. Twenty-two healthy, sex- and age-matched subjects served as a control group.
Isolation of mononuclear cells.Peripheral blood mononuclear cells (PBMC) were isolated by Lymphoprep (Nyegaard, Oslo, Norway) density centrifugation. The cells were frozen and stored in liquid nitrogen. Before use, the cells were rapidly thawed and washed. The viability of the cells was ascertained by trypan blue exclusion. Previous studies have shown that freezing does not significantly change the expression of the surface markers used, intracellular cytokine expression, or phytohemagglutinin (PHA)-induced apoptosis (13). Blood samples were collected during the same time period and treated in the exact same way for patients and controls. For each donor, samples from days 0 and 7 were analyzed on the same day together with samples from the healthy age- and sex-matched control in all of the cellular immune assays in order to eliminate the influence of day-to-day variation.
Clinical chemistry tests.Standard laboratory procedures were used for clinical chemistry tests. The concentration of leukocyte subsets was determined with a cell counter (Technic H.1; Miles, Tarrytown, N.Y.).
Cultivation of PBMC.PBMC were resuspended in RPMI 1640, supplemented with 15% heat-inactivated pooled human serum, 20 IU of penicillin per ml, and 20 μg of streptomycin per ml (Gibco, Paisley, United Kingdom) and seeded into 24-well plates (Nunc, Roskilde, Denmark). Each well contained 106 PBMC in 1 ml of medium. The cells were cultured at 37°C in a humidified atmosphere with 5% CO2. For intracellular cytokine detection, the cells were incubated with 1.5 μM monensin (Sigma, St. Louis, Mo.), 1 μM ionomycin, and 50 ng of phorbol myristate acetate per ml for 4 h. For detection of apoptosis, the cells were incubated with PHA in a final concentration of 40 μg/ml for 24 h.
Flow cytometry.For surface staining, three-color flow cytometry was applied by using a monoclonal antibody (MAb) conjugated to either fluorescein isothiocyanate (FITC), phycoerythrin, or Cy5 and specific for the determinants listed: CD3 (UCHT1; DAKO, Glostrup, Denmark) CD4 (MT310; DAKO) CD8 (DK25; DAKO), CD19 (HD37; DAKO), CD45R0 (F0800; DAKO), CD45RA (4KB5;DAKO), CD62L (FMC46, DAKO), CD56 (MOC-1; DAKO), and CD95 (M019751; Pharmingen, San Diego, Calif.). All samples were analyzed on a flow cytometer (Epics XL-MCL; Coulter, Hialeah, Fla.). Data were analyzed with Winlist software (Verity, Topsham, Maine). The absolute number of the lymphocyte subset was found by multiplying the lymphocyte counts by the proportion of lymphocytes.
Intracellular cytokine staining.The method used for intracellular cytokine staining was based on other studies (15). In brief, following surface staining, cells were washed twice in staining buffer, fixed in staining buffer containing 2% formaldehyde (Sigma), and washed twice in a freshly made saponin buffer (phosphate-buffered saline [PBS], bovine serum albumin [BSA], NaN3 containing 0.1% [wt/vol] saponin Σ), and incubated with anticytokine antibody (IL-2, tumor necrosis factor alpha [TNF-α], IFN-γ, or IL-4 from Pharmingen) for 30 min in the dark (40°C). Following intracellular labeling, the cells were washed twice in saponin buffer and twice in staining buffer, resuspended in the same buffer, and analyzed.
Analysis of apoptosis by annexin V and 7AAD staining.PBMC were labeled with annexin V (M046701; Pharmingen) and 7-aminoactinomycin D (7AAD) (Sigma) in binding buffer (Pharmingen) according to the manufacturer's directions and analyzed by flow cytometry.
Levels of sFas ligand in plasma.Levels of sFasL in plasma were measured by enzyme-linked immunosorbent assay (ELISA) (Naka-ku, Nagoya, Japan).
Statistical analysis.Statistical calculations were performed with Sigma Stat software. Data are presented as medians and quartiles, because initial analysis revealed that the lymphocyte data were not normally distributed. Parameters obtained on day 1 were compared to data obtained on day 7 by Wilcoxon's signed-rank test for paired data. Furthermore, data from patients on day 1 or day 7 were compared to those from controls by Mann-Whitney rank-sum test for independent groups. P values of <0.05 were considered significant. Preliminary data analysis revealed that there were no significant differences between patients with pneumonia and patients with bacteremia or meningitis according to the parameters investigated. Consequently, patient data were pooled in the final analyses.
RESULTS
Patients are characterized by low numbers of lymphocytes.Patients had elevated levels of CRP, increased numbers of neutrophils, and low hemoglobin levels at admission to the hospital compared to the control group (Table 1). These parameters were partly normalized after the first week of treatment, although they were still significantly different from those of the controls on day 7. Thrombocytes were within the range of the controls on day 1, but elevated on day 7, whereas the monocyte count was constantly increased on days 1 and 7. In contrast, the number of lymphocytes was decreased at admission, but increased during the first week of hospitalization, although it did not reach the levels in the controls on day 7. Accordingly, the acute phase of the infection was associated with lymphopenia and a subsequent release of lymphocytes to the peripheral blood in response to chemotherapy.
Clinical data from patients with pneumococcal infections at days 1 and 7 after the start of treatment and in a control groupa
Activated or memory T lymphocytes, but not naive cells, disappear from circulation.In order to further characterize the distribution of lymphocytes during infection, we looked at changes in percentages of different subsets. On day 1, percentages of CD8+ cells were severely decreased, whereas those of CD4+ cells were unaltered. After 1 week, the percentage of CD4+ cells was elevated compared to that at day 1 and to that of controls, whereas the percentage of CD8+ cells was increased compared to that at day 1, but did not reach the ranges observed in controls (Table 2). When the absolute numbers in the blood were calculated, it was revealed that the numbers of both CD4+ and CD8+ cells were decreased on day 1. On day 7, the number of CD4+ cells was normalized: whereas the number of CD8+ cells was elevated compared with that at day 1, it was still decreased compared with the number in controls (Fig. 1A and B). No significant differences in percentages of naive cells (CD62L+ CD45RA+) and activated or memory cells (CD45R0+ and CD95+) among CD4+ and CD8+ cells were observed (Table 2). However, calculations of the absolute numbers revealed that the infection-induced decrease in counts of CD4+ and CD8+ cells in the circulation was due to a disappearance of activated cells, whereas the number of naive cells was unaffected (Fig. 1A and B). The percentage of NK cells was decreased at admission, but not after 1 week when compared to that of the control, whereas the percentage of B cells was unaffected (Table 2).
Absolute numbers of CD4+ subsets (A), CD8+ subsets (B), CD4+ cytokine-producing cells (C), and CD8+ cytokine-producing cells (D). Results (medians and quartiles) from patients with pneumococcal infections at day 1 (black bars) and day 7 (white bars) after the start of treatment and in a control group (gray bars) are shown. P < 0.05 for significant difference from day 1 (∗) or from day 7 (#).
Percentages of lymphocyte subsets in patients with pneumococcal infections at days 1 and 7 after the start of treatment and in a control group
T lymphocytes with a type 1 cytokine profile reappear in the circulation in relation to treatment.The percentage of IL-2-producing CD4+ cells was lower at day 1 when compared to patients at day 7, and lower percentages of IFN-γ-producing CD4+ cells were observed in patients on days 1 and 7 when compared to those in the control group (Fig. 2A). Within the CD8+ subset, more IFN-γ-producing cells were found at day 7 than at day 1 (Fig. 2B). When the absolute number of T lymphocytes with potential cytokine production was calculated, it became clear that CD4+ cells and CD8+ cells producing IL-2, TNF-α, and IFN-γ disappeared from the circulation in relation to the acute phase of the infection, but had partly returned on day 7. The percentage or absolute number of CD4+ T lymphocytes expressing IL-4 was not affected during the infection (Fig. 1C and D).
Percentage of cytokine-producing CD4+ (A) or CD8+ (B) cells from patients with pneumococcal infections at day 1 (black bars) and day 7 (white bars) after the start of treatment and in a control group (gray bars). Medians and quartiles are shown. P < 0.05 for significant difference from day 1 (∗) or from day 7 (#).
Increased apoptosis among lymphocytes and elevated levels of sFasL in serum.Annexin V-positive, 7AAD-negative cells following PHA stimulation were identified as markers of cells primed for apoptosis, whereas 7AAD-positive cells marked dead cells (Fig. 3). The percentages of annexin V-positive, 7AAD-negative cells were significantly elevated at day 1 both compared to the levels at day 7 and in the control group. There was no difference between the levels in day 7 patients and the controls. Furthermore, there was no difference in 7AAD-positive cells in relation to the disease (data not shown). Finally, levels of sFasL in serum were significantly higher in patients at admission compared to those in controls, whereas the difference between day 1 and day 7 did not reach significance (Fig. 4).
PHA-induced apoptosis (Annexin V+7AAD−) of lymphocytes from patients with pneumococcal infections (black bar) and 7 days after initiation of chemotherapy (white bar), as well as age-matched, healthy control donors (gray bars). Medians and quartiles are shown. P < 0.05 for significant difference from day 1 (∗).
Levels of sFasL in serum obtained from patients with pneumococcal infections at admission (black bar) or 7 days after admission (white bar) and from age-matched, healthy control donors (gray bar). Medians and quartiles are shown. P < 0.05 for significant difference from day 1 (∗).
DISCUSSION
The major findings in this study were that pneumococcal infections induced lymphopenia due to disappearance of activated T lymphocytes from the peripheral blood. The CD8+ cell compartment was affected more strongly than CD4+ cells, because the percentage of CD8+ cells decreased, whereas the percentage of CD4+ cells was unaltered on admission to the hospital. One week of antibiotic treatment restored the number of activated CD4+ cells in the peripheral blood, whereas the number of activated CD8+ cells had not recovered completely. The compartment of naive T cells was unaffected by the infection. Accordingly, the infection-induced lymphopenia might be due to a sequestration of T lymphocytes in peripheral lymphoid organs with a subsequent release to the circulation when the infection was cleared. Furthermore, infection-induced apoptosis may contribute. Thus, increased levels of FasL in serum and increased PHA-induced apoptosis among PBMC were detected at day 0. However, it is not likely that apoptosis is responsible for the low CD4 count, because the absolute number was completely restored after 1 week, and furthermore, we did not detect any changes in the percentages of naive and activated cells. This indicates that the recovery was due to trafficking rather than an inflow of de novo-synthesized cells. In contrast, both sequestration and apoptosis may contribute to the low CD8+ cell count. In order to elucidate the function of the redistributed T cells, their cytokine profile was investigated. This analysis showed that IL-2-, TNF-α-, and IFN-γ-producing T cells disappeared from the circulation during the acute phase of the infection, but reappeared partly after treatment for 1 week. Furthermore, the percentages of IL-2-producing CD4+ cells and IFN-γ-producing CD8+ cells were increased on day 7 compared to day 1, and the percentage of IFN-γ-producing CD4+ cells was decreased on day 1 compared to that in controls, suggesting that these subsets are especially affected. In contrast, CD4+ lymphocytes with potential IL-4 production were not affected during the infection. Accordingly, these results show that pneumococcal infections are associated with trafficking of activated CD4+ and CD8+ T lymphocytes with a type 1 cytokine profile.
In accordance with the present study, investigations of other acute infections have shown a disease-induced decrease of lymphocytes in the peripheral blood, which may be due to a distribution to the sites of inflammation (3, 7, 22). Chemotherapy of bacterial (7), viral (3), and parasitic (6, 11) infections induces an increase in frequencies and absolute numbers of T cells in the circulation, and it has been hypothesized that this increase reflect a redistribution of T cells sequestered during the disease because the short time period makes it unlikely that these cells are newly synthesized. In the present context, this hypothesis suggests that CD4+ cells and CD8+ cells with potential type 1 cytokine production are engaged in the in vivo immune response against pneumococcal infections. This interpretation of the data is consistent with previous in vitro studies, demonstrating that pneumococcal antigens induce T-lymphocyte production of IFN-γ (18) and the report of severe clinical symptoms in relation to IL-12 deficiency (1). Furthermore, in accordance with our data, other studies have indicated that IL-4 plays a less significant role during infections caused by S. pneumoniae. The number of CD4+ T cells producing TNF-α in this study was low, although there was a significant increase in these cells at day 7, and the control group demonstrated higher numbers than patients did, indicating that this subset was also engaged in the immune response. Thus, in addition to monocytes, macrophages, and other antigen-presenting cells, CD4+ T cells may be an important source of TNF-α during the infection.
Studies of other infections have also demonstrated that lymphopenia in the acute phase is accompanied by increased susceptibility to activation-induced cell death of the remaining peripheral lymphocytes (14, 21). Accordingly, this may represent a general phenomenon in the host response. It is unlikely that apoptosis is the only mechanism responsible for the lymphopenia in the present study, although apoptosis may contribute, because the lymphocyte count was already partly restored after 1 week. We expected that activation-induced apoptosis would peak on day 7 when stimulated cells returned to the circulation. In contrast, the peak in apoptosis already occurred at admission to the hospital together with increased levels of sFasL in plasma. Thus, apoptosis was probably related to the infection rather than an altered distribution of T-lymphocyte subsets, because the percentages of naive and activated cells were unaltered among T lymphocytes. We suggest that the infection induces a higher susceptibility to apoptosis among bystander lymphocytes as a consequence of increased levels of proteins, such as sFasL. This, concomitant with a gradual release of exhausted cells from areas with inflammation, leads to the high level of preapoptotic cells observed in the patients. However, further studies are needed to examine these phenomena more precisely. For instance, it would have been very interesting to evaluate activation-induced apoptosis concomitant with spontaneous apoptosis and Fas-induced apoptosis at different time points. Such detailed analyses were not possible, because we chose to use frozen cells in order to perform cellular assays on the same day of samples from days 0 and 7 for each donor together with cells from the healthy control. Thus, based on our laboratory experience, the day-to-day variation in cellular assays is much larger than changes induced by freezing. We have previously examined how freezing of polymorphonuclear cells (PMNC) from healthy donors does not affect the cellular assays used in the present study (15), but we cannot exclude that the freezing procedure affected cells from patients during the acute phase of infection, and this could especially affect apoptosis assays. On the other hand, an increased frailty of cells during the freezing process due to preactivation in patients also carries its own information about differences in function during acute-phase infection.
In conclusion, pneumococcal infections induced lymphopenia due to the disappearance of activated T lymphocytes with a type 1 cytokine profile. CD4+ cells reappeared in the circulation in relation to the treatment, whereas the number of CD8+ cells was only partly restored after 1 week. Accordingly, the underlying mechanisms might include sequestration in the peripheral tissues with subsequent release to the circulation when the bacteria are cleared with regard to the traffic of CD4+ cells, whereas increased apoptosis might also be involved in the decreased number of CD8+ cells. Furthermore, these data suggest that activated T lymphocytes with a type cytokine 1 profile are engaged in the in vivo immune response to S. pneumoniae.
ACKNOWLEDGMENTS
We acknowledge the excellent technical assistance of Hanne Villumsen and Gitte Grauert.
FOOTNOTES
- Received 7 February 2002.
- Returned for modification 29 April 2002.
- Accepted 29 May 2002.
Editor: E. I. Tuomanen
- American Society for Microbiology