Main

Activated monocytes/macrophages, lymphocytes, and neutrophils can permeate the blood-brain barrier(1). These immunocompetent cells, astrocytes, and microglial cells in the CNS produce a variety of cytokines, including IL-1, IL-6, IL-8, tumor necrosis factor-α, interferon-γ, and colony-stimulating factors(1–3). Elevated levels of these cytokines occur in the CSF at the initial stage of bacterial and viral meningitis as shown by us(4, 5) (our unpublished data) and other investigators(1–3). The intracisternal injections of tumor necrosis factor-α and IL-1 induce inflammatory cell extravasation and brain edema, but the administration of their antibodies can reduce the tissue damage(1, 2). These cytokines thus function in the local inflammatory response in meningitis.

Human IL-10 is a cytokine produced by T cells, monocytes/macrophages, and B cells(6, 7). It reduces the expansion of antigen-specific T cells and their production of cytokines such as IL-2, IL-4, IL-5, and interferon-γ through the down-regulation of antigen presentation by accessory cells(8–10). The IL-10 prevents human monocytes from expressing major histocompatibility complex class II antigens and from killing intracellular microorganisms(6–8). It inhibits the synthesis of tumor necrosis factor-α, IL-1, IL-6, IL-8, and IL-10 itself, and also various colony-stimulating factors by human monocytes exposed to LPS(11). The IL-10 decreases the release of proinflammatory cytokines but enhances the production of IL-1 receptor antagonist by LPS-stimulated neutrophils(12, 13). Human IL-10 thus exerts potent immunosuppressive activities.

Recent investigations have shown that the murine brain expresses IL-10 mRNA during the later stage of infection with a protozoan parasiteToxoplasma gondii(14, 15). Intrathecal production of IL-10 has been found in mice having Listeria meningitis(16). IL-10 has been detected in sera of patients with sepsis and in most CSF samples from patients with bacterial meningitis, but found in only 10% of CSF samples from patients with viral meningitis(16, 17). Nevertheless, little is known about which cells produce human IL-10 in the CSF and how the IL-10 levels change during human meningitis. In this study, we show that IL-10 is produced in the CSF in aseptic meningitis, and it may increase relatively late compared with proinflammatory cytokines such as IL-6, IL-8, and G-CSF. IL-10 may play an immunoregulatory role in the CSF of childhood meningitis.

METHODS

Study subjects. We studied 22 patients with aseptic meningitis, and 21 controls without meningitis (Table 1). They ranged in age from 26 d to 15 y and were admitted to our hospital from June 1991 to August 1994. The patients with aseptic meningitis fit the following criteria:1) all had fever, headache, vomiting, and stiff neck; 2) cell counts greater than 35/L in the CSF; 3) sterile CSF found in bacteriologic studies. The pathogens for aseptic meningitis were determined using serum-specific antibody titers and viral cultures as described previously(5), which included mumps virus in six patients, echovirus 30 in four, echovirus 9 in two, and uncertain causes in 10 individuals. The symptomatic stage was designated as the period when any of the meningeal symptoms or signs existed, and the recovery stage as the period after all of the symptoms and signs disappeared. The 1st d of the illness was determined as the day when all of the meningitis symptoms had first occurred. The patients without meningitis (as controls) fit the following criteria:1) cell counts less than 5/L in the CSF and 2) sterile CSF found in bacteriologic studies. The control children consisted of two with epilepsy, three with febrile convulsions, two with relapse-free acute lymphoblastic leukemia (two lumbar punctures were done in each patient), and 14 patients with fever and the meningeal fever-like signs.

Table 1 Clinical characteristics of study subjects
Full size table

Gender and age in both groups were not significantly different. Counts for total white cells, neutrophils, and mononuclear cells, as well as the levels of protein and glucose were significantly different between CSF samples obtained from the meningitis patients with meningeal signs and those from the controls (Table 1).

Sample collection. We obtained institutional approval from the responsible committee and a full informed consent from each of the patient's parents for this study. To minimize the patients' risk throughout this study, lumbar punctures were done only when clinically indicated. We did lumbar punctures again for patients diagnosed finally as aseptic meningitis, only when neutrophils were dominant in the first CSF samples. These patients received a lumbar puncture to decide whether an antibiotic therapy should be continued. Lumbar punctures were done once for each patient to confirm the improvement of the illness during the recovery stage in 14 of 22 patients. The control children who had convulsion or both fever and meningeal signs received lumbar punctures to rule out meningitis. The CSF in leukemia patients was examined to exclude meningeal leukemia. The CSF samples were immediately spun down at 150 × g for 10 min, and the supernatant was stored at-30 °C. Serum specimens were collected at the same time and stored at -30°C.

Cell culture. Mononuclear cells were separate by Ficoll-Hypaque gradient centrifugation from CSF cells(18). The mononuclear cells (1 × 106/mL) were incubated in RPMI 1640 supplemented with 10% FCS (Hyclone Labs, Logan, UT) at 37 °C under air with 5% CO2 for 24 h. The culture supernatant was harvested by centrifugation at 400 × g for 10 min and stored at -30°C.

ELISAs for measuring cytokines. IL-10 concentrations were measured in duplicate by an ELISA (PerSeptive Diagnostics, Cambridge, MA). Briefly, 100 μL of CSF, serum, or standard recombinant human IL-10 were dispensed into 96-well plates precoated with a monoclonal anti-IL-10 antibody. The plate was allowed to settle overnight at 4 °C. IL-10 was then sandwiched with a biotinylated antibody against human IL-10. After that, horseradish peroxidase-conjugated streptavidin, substrate solution, and stop solution were sequentially added. The ELISA system specifically recognized human IL-10, and its measuring range was 10-500 ng/L. The intraassay coefficients of variation were from 2.1 to 8.0%, and the interassay coefficients of variation were from 4.0 to 9.5%.

Concentrations of IL-6 (Toray-Fuji Bionics, Tokyo, Japan) and IL-8(Toray-Fuji Bionics) were quantified using ELISA kits. G-CSF levels were assessed by the chemiluminescent enzyme immunoassay (Chugai Pharmaceutical Co., Ltd., Tokyo)(19). The minimal detectable concentrations were 25 ng/L (IL-6), 12 ng/L (IL-8), and 1 ng/L (G-CSF), respectively.

Reverse-transcribed PCR. Total RNA was extracted by the acidic guanidine isothiocyanate-phenol-chloroform method(20) from 1 × 106 of fresh cells in the CSF. First-strand cDNA was synthesized from approximately 5 μg total RNA in a 100 μL-reaction using a cDNA synthesis kit (Behringer Mannheim) including avian myeloblastosis virus reverse transcriptase. The cDNA (5 μL) was amplified by PCR in a total of 50 μL of Taq polymerase buffer containing 100 μM dNTPs, 1.3 U of Taq polymerase (Behringer Mannheim, Indianapolis, IN), and 250 nM of sense and antisense primers. Human IL-10 primers were designed as 5′-CAGCTCAGCACTGCTCTGTT-3′ (sense, nucleotides 36-55) and 5′-CCTGAGGGTCTTCAGGTTCT-3′ (antisense, nt 371-390)(9). β-Actin primers were made as previously described(21). The reaction consisted of 39 cycles(IL-10) or 21 cycles (β-actin) of denaturation at 94 °C for 1 min, annealing at 55 °C for 1 min and extension at 72 °C for 1 min in a Perkin-Elmer (Norwalk, CT) DNA thermal cycler. The PCR products were electrophoresed on 3% agarose gel and transferred onto nylon filters. The filters were UV-irradiated, prehybridized at 42 °C for 2 h, and hybridized with the following probes at 42 °C for 36-48 h as shown previously(21). An IL-10 oligonucleotide probe, 5′-AACTTGTGAATATAGTCGATGACCTTGTGG-3′ (nucleotides 147-176)(9), was labeled with 32P by the 5′-end-labeling method(21). A β-actin probe(0.7 kb) was labeled with 32P by the random priming method(22). The filters were washed to a stringency of 0.2× SSC with 0.1% SDS at 57 °C and then exposed to a x-ray film for 1-3 d at -80 °C.

Statistical analysis. Results are expressed as means ± SD unless otherwise expressed. Variables were transformed to common logarithms before some statistical analyses. Probability of a significant difference was determined using the Mann-Whitney U test. Relationships between IL-10 levels and other indices were assessed using Pearson's correlation coefficients. Linear regression lines were fitted by the least squares method. Differences were considered significant when the two-tailed p value was below 0.05.

RESULTS

Detection of IL-10 in the CSF. We examined IL-10 concentrations in the CSF of the children either with or without aseptic meningitis. When the patients with aseptic meningitis had the meningeal symptoms, IL-10 in the CSF was detected in 14 of 22 patients (18 of 31 samples). Mean IL-10 levels at the symptomatic stage were 88 ± 146 ng/L (median: 19 ng/L). These values were significantly higher (p = 0.0006) than those of the control children without meningitis; 22 of 23 of these samples (21 of 22 control children) had below the detectable limit (10 ng/L) of IL-10. After the disappearance of meningeal symptoms, 11 of 14 samples (14 patients) were undetectable. Mean IL-10 levels during the recovery stage (11 ± 6 ng/L) were significantly lower than those during the symptomatic stage (p= 0.0226) (Table 2).

Table 2 IL-10 levels in the CSF and serum
Full size table

Figure 1 illustrates the time-dependent changes in IL-10 levels in the CSF. High IL-10 levels were found in 19 of 33 samples before the 6th d of aseptic meningitis; thereafter IL-10 levels in 9 of 12 samples became undetectable. In 14 patients with the illness examined longitudinally, IL-10 levels dropped sequentially in 11 patients, but rose strikingly at the initial stage in three patients.

Figure 1
figure 1

Time-dependent fluctuations of IL-10 levels in the CSF of aseptic meningitis. Only values of increased IL-10 levels during the initial stage of disease in the same individuals are connected. Black circles, symptomatic stage; open circles, stage without meningeal symptoms; dashed line, lower detectable limit for IL-10.

Full size image

Serum IL-10 levels. IL-10 was not detected in any of the serum samples in aseptic meningitis except two samples from the same patient(Table 2). Serum IL-10 levels were significantly lower(p = 0.0011) than corresponding levels in the CSF during the symptomatic stage of the illness.

Relationships between IL-10 levels and clinical indices. Significant correlations were found between IL-10 levels and mononuclear cell counts in the CSF of aseptic meningitis (r = 0.644, p < 0.0001) (Fig. 2). The relationship between the IL-10 levels and neutrophil counts was not significant in the illness (r = 0.235). IL-10 levels in the CSF of the illness were not significantly correlated with clinical features such as either the grade or duration of fever, headache, or vomiting; the concentrations of glucose or protein in the CSF. The correlation between IL-10 levels and only body temperature in aseptic meningitis was significant (r = 0.398, p = 0.0067).

Figure 2
figure 2

Relationships between IL-10 levels and mononuclear cell counts in the CSF of aseptic meningitis. Black circles, symptomatic stage; open circles, stage without meningeal symptoms; solid line, linear regression line; dashed line, lower detectable limit for IL-10.

Full size image

Production of IL-10 by CSF cells. The expression of IL-10 mRNA in fresh CSF cells in meningitis was examined by PCR amplification. The CSF cells were collected on the 1st to 4th d from the onset of the illness.Figure 3A demonstrates the human IL-10-specific hybridization signals: prominent bands were observed in two patients (lanes 3 and 7) with bacterial meningitis and in four (lanes 2, 4, 6, and 9) of seven patients with aseptic meningitis. We also examined the production of IL-10 protein by CSF mononuclear cells obtained on the 1st to 4th d of aseptic meningitis in five patients; these individuals were different from those on which the mRNA studies were done. High levels of IL-10 (152-485 ng/L) were found in the culture supernatant from all of the children(Fig. 3B).

Figure 3
figure 3

Expression of IL-10 mRNA and its protein by cells in the CSF of meningitis. (A) Signals for human IL-10 and β-actin mRNA in fresh CSF cells using reverse transcribed-PCR amplification and hybridization. Lanes 3 and 7, bacterial meningitis; lanes 1, 2, 4-6, 8, and 9, aseptic meningitis. (B) IL-10 protein levels produced by cultures of CSF mononuclear cells in aseptic meningitis.

Full size image

Kinetics of levels of IL-10, IL-6, IL-8, and G-CSF in the CSF.Figure 4 compares the changes in levels of IL-10 and proinflammatory cytokines in the CSF of aseptic meningitis. Mean IL-10 levels reached their peak on the 2nd d (121 ± 171 ng/L, median: 86 ng/L) and the 3rd d (134 ± 227 ng/L, median: 36 ng/L) from the onset of the illness. These values were significantly higher (p = 0.0109 and 0.0100) than those after the 6th d of the disease. In contrast, all mean levels of IL-6, IL-8, and G-CSF were highest on the 1st d of the illness.

Figure 4
figure 4

Kinetics of the levels of IL-10, IL-6, IL-8, and G-CSF in the CSF of aseptic meningitis. Black circles, symptomatic stage; open circles, stage without meningeal symptoms; horizontal bars, mean in each group; dashed lines, lower detectable limit for each cytokine.

Full size image

Correlations in the CSF of aseptic meningitis were significant between levels of IL-6 and IL-8 (r = 0.750, n = 37); between IL-6 and G-CSF (r = 0.695, n = 31); and between IL-8 and G-CSF(r = 0.866, n = 47)(p < 0.0001). However, relationship was not significant between IL-10 levels and concentrations of either IL-6 (r = 0.125, n = 15), IL-8 (r = 0.380, n = 24) or G-CSF (r = 0.184, n = 23).

DISCUSSION

The present study showed that IL-10 is readily detectable in the CSF at the symptomatic stage in 14 of 22 (64%) patients with aseptic meningitis. The IL-10 levels dropped as the meningeal symptoms disappeared. The CSF samples from most of the control children without meningitis did not have detectable levels of IL-10. Our preliminary data showing that high levels of IL-10 in the CSF of bacterial meningitis are consistent with a previous observation by Freiet al.(16). However, they detected IL-10 in the CSF with only 10% of frequency from patients with aseptic meningitis(16); in contrast, we found IL-10 in the CSF in the higher frequency of patients with the illness. This difference may be largely due to our use of a sensitive ELISA, with a lower detectable limit of 10versus 100 ng/L in their bioassay. In any case, our data reveal that IL-10 levels in the CSF are transiently increased in aseptic meningitis.

In further studies, we found that, in contrast to high levels of IL-10 in the CSF, the sera had low levels of IL-10 even during the symptomatic stage of aseptic meningitis, and these IL-10 levels in the sera did not fluctuate throughout the course of the illness. These results suggest that the high levels of IL-10 in the CSF are not likely from IL-10 in the blood that crossed the blood-brain barrier. Actually, our data showed close correlations between IL-10 levels and mononuclear cell counts in the CSF of patients with aseptic meningitis. We further found that IL-10 mRNA is detectable in the CSF cells using reverse-transcribed PCR-assisted amplification, and mononuclear cells in the CSF can release high levels of IL-10 protein in vitro. These results provide direct evidence that IL-10 is produced intrathecally, at least in part, by mononuclear cells.

LPS strongly stimulates IL-10 production by monocyte/macrophage, microglial cells, and astrocytes(11, 23, 24). IL-10 is also secreted by T cells after antigen-specific stimulation in the presence of antigen-presenting cells and by EBV-transformed B cells(7, 8, 10). In bacterial meningitis, LPS and bacterial superantigens such as staphylococcal enterotoxin B(25) can act as stimulators for IL-10 production. In the CSF of aseptic meningitis, elevated levels of a variety of proinflammatory cytokines are detected. However, tumor necrosis factor-α was not detected(1–5), which probably is a major stimulator of IL-10 production by human monocytes(26). Cytokines such as IL-1, IL-6, granulocyte/macrophage-colony-stimulating factor, and interferon-α may not act directly on human monocytes to secrete IL-10(26). Consequently, the major inducer(s) of production of IL-10 in the cytokine network in aseptic meningitis remains under exploration.

Our kinetic study on the cytokines in the CSF suggests that IL-10 levels increase at the later stage, compared with levels of IL-6, IL-8, and G-CSF in aseptic meningitis. This kinetic pattern of IL-10 levels is consistent with recent observations in mice; intracerebral expression of IL-10 mRNA is selectively up-regulated during the later phase of Toxoplasma encephalitis(14, 15) and experimental autoimmune encephalomyelitis(27). The late increase in IL-10 protein has been found in the CSF of murine Listeria meningitis(16). Human IL-10 is also produced relatively latein vitro after activation of monocytes(11) and T cells(28), compared with proinflammatory cytokines. The substantial IL-10 production at the later stage of meningitis seems like a reasonable response for suppression of the inflammatory process maintained by ongoing production of proinflammatory cytokines. Consistent with this hypothesis, administration of IL-10 prevented tissue damage in mice with herpes virus-induced keratitis(29) and lethality in LPS-injected mice(30). In addition to IL-10, transforming growth factor-β, and IL-4 have potent immunosuppressive properties(31, 32). We have found that IL-4 is not detected, and transforming growth factor-β levels do not fluctuate in the CSF throughout meningitis (our unpublished data). Taken together, IL-10 may function as a suppressor of the inflammatory process in human meningitis. However, it remains unclear whether IL-10 functions beneficially or harmfully to generate an appropriate defensive response in the disease. Further studies will be needed to reveal the exact physiologic roles of IL-10 in the illness.