Abstract
In hepatitis C virus (HCV) infection, Th responses are implicated in the pathogenesis of liver disease. The dendritic cell (DC) is the most potent activator of CD4 T cells for supporting Th1 differentiation. To clarify the roles of DC of HCV-infected individuals in the development of CD4 T cell responses, we generated peripheral DC with GM-CSF and IL-4 from 24 chronic hepatitis C patients and 14 healthy donors. We then compared their potentials for stimulating allogeneic CD4 T cells, autologous CD4 T cells against influenza A or HCV core Ags, and cytokine production. The DC from the patients (HCV-DC) expressed lower degrees of CD86 than DC from the donors (N-DC), whereas no difference was found in the HLA molecules and other costimulators. HCV-DC stimulated allogeneic T cells less than N-DC; however, influenza A- or core-pulsed HCV-DC retained the potentials for autologous T cell proliferation. In allogeneic DC/T cell cultures, the IFN-γ levels with HCV-DC were lower than those with N-DC, which may be related to the low expressions of IL-12 p35 and p40 transcripts in HCV-DC. The stimulation with LPS disclosed that HCV-DC is less potent in IL-12 p70 production than N-DC. In the autologous cultures, the pulsing of the Ags to HCV-DC increased the IL-12 p40 and IFN-γ production and up-regulated the transcription of both IL-12 subunits. Exogenous IL-2 or IL-12 restored the low allogeneic T cell proliferation with HCV-DC in a dose-dependent manner. Therefore, low expression of CD86 and/or IL-12 is crucially involved in the low allostimulatory capacity of HCV-DC. Low IL-12 and low IFN-γ milieu with HCV-DC on encounters with alloantigens may impede Th1 polarization.
Hepatitis C virus (HCV)3 is one of the major causative agents of chronic non-A, non-B liver disease and often leads to liver cirrhosis or hepatocellular carcinoma (1, 2). In HCV infection, the involvement of immunological reactions against HCV has been implicated in the pathogenesis or progression of liver disease. Recently, some researchers have reported that patients who show strong HCV-specific Th responses eradicate HCV after the initial exposure (3, 4). In contrast, the patients who display only a weak Th response show the tendency to develop to the chronically infected state (3). In chronic HCV infection, the association of stronger Th responses with low-grade liver injury has also been reported (5). These findings demonstrate that the strength of CD4-positive Th responses is related to benign courses of HCV-induced liver disease.
In general, the Th populations in mice and humans could be functionally divided into three major subsets, i.e., Th0, Th1, and Th2 (6, 7). The imbalance or skewness among Th subtypes may be involved in the pathogenesis of a variety of diseases, including viral infection or allergetics. The cytokines present at the initiation of the immune response are the most clearly defined factors determining Th1 and Th2 differentiation from Th0 (6, 7). In this regard, T cell-derived IL-4 and APC-derived IL-12 are the key cytokines (6, 8). In addition to the production of IL-12, professional APCs play important roles in directing the patterns and/or the strength of the immune response by transducing costimulatory signals to T cells. With HCV infection, examinations have been conducted to clarify which Th subset is dominant in relation to the elimination or chronicity of HCV or the degree of liver injury (3, 9, 10). However, most of the reports describing Th subsets were obtained from cytokine analysis with whole PBMC, liver-infiltrating lymphocytes, or established Th clones. Thus, the precise roles of APC in HCV infection in determining the strength of CD4 T cell responses or Th polarization are yet to be investigated.
Dendritic cells (DC) are one of the most potent APCs in vivo (11) and play crucial roles in the enhancement or regulation of cell-mediated immune reactions. Since DC strongly express various costimulatory and/or adhesion molecules (11), they can activate even naive T cells in a primary response. The DC-based approach has been used to establish treatments for several malignant diseases, including B cell lymphoma or melanoma (12, 13). However, APC functions are occasionally suppressed under some tumor-bearing states (14, 15, 16). Thus, the DC function must be assessed in relation to the disease status to apply DC as an immunotherapeutic tool. Recently, procedures have been established for obtaining highly purified DC from mice or humans (17, 18, 19). The in vitro-generated DC from human peripheral blood progenitors are considered to be immature and possess the endocytosis capacity of soluble Ags and present them to T cells (19). In the present study, to assess the DC function in HCV-infected patients and to determine its regulators, we used peripheral blood as a DC source and generation by cytokine stimulation. We next compared the stimulatory capacity for naive or recall Ag-specific CD4 T cells between DC simultaneously generated from HCV-infected patients and from uninfected healthy donors. To clarify the mechanisms for the differences in DC function between patients and donors, DC phenotypes or the cytokines produced in the DC/T cell cultures were examined. We found that DC from HCV-infected individuals exhibited an impaired stimulatory potential against allogeneic CD4 T cells. Exogenous IL-2 or IL-12 restored such defective DC capacity, which offers promise for the modulation of DC functions.
Materials and Methods
Subjects
Twenty-four patients who were positive for both anti-HCV Ab and serum HCV-RNA were enrolled in this study. All patients were negative for HBs Ag and had no apparent history of other types of liver diseases. After informed consent had been obtained from them, liver biopsy was conducted. Eighteen patients were histologically diagnosed as having mild or moderate chronic hepatitis. The remaining six patients did not have complications with cirrhosis or hepatocellular carcinoma according to biochemical and radiological examinations. Serum HCV-RNA was quantified by means of branched DNA probe assay (Chiron HCV-RNA, Emeryville, CA). HCV genotypes were determined by PCR with type-specific primers (20). To assess the DC function of HCV-infected patients, 14 age-matched healthy individuals were assigned as controls. The clinical backgrounds of these subjects are shown in Table I⇓.
Clinical backgrounds of normal volunteers and HCV-infected patientsa
Reagents
NG-monomethyl l-arginine acetate (l-NMMA) was from Research Biochemicals International (Natick, MA). Influenza A Ag (Texas 1/77 strain) was purchased from Chemicon International (Temecula, CA). Recombinant HCV core protein was kindly provided by Mitsubishi Chemical Corporation (Yokohama, Japan), which was expressed by Escherichia coli with a plasmid carrying the HCV genotype 1b cDNA clone spanning from 7 to 353 nucleotides (21).
Generation of DC from PBMC
DC were generated from PBMC according to the methods described by Romani et al. (18). The PBMC were collected from venous blood by Ficoll-Hypaque density-gradient centrifugation. After the PBMC had been suspended in DC culture medium (RPMI 1640 supplemented with 10% FCS, 50 U/ml penicillin, 2 mM l-glutamine, and 50 μM 2-ME), they were placed at 5–10 × 106/well on six-well polystyrene culture plates and stored at 37°C for 2 h. After incubation, nonadherent cells were removed by gently pipetting with warm RPMI 1640. Adherent cells were supplied with DC medium containing 800 U/ml of recombinant human GM-CSF and 500 U/ml of recombinant human IL-4 and cultured for 7 days at 37°C under 5% CO2. Cells were refed fresh medium containing 800 U/ml of GM-CSF and 500 U/ml of IL-4 every 2 days. After 7 days, the cultures developed an adherent monolayer and clusters of DC colonies. The DC yield was defined as the percentage of the obtained DC numbers to the PBMC numbers used as the source. As a control in MLR, adherent cells were harvested on day 1, by washing the wells, and used as the monocyte preparation. The positivity for CD14 in the samples was more than 95%, as confirmed by flow cytometry.
Flow cytometric analysis
On day 7 of culture, DC were harvested, and their surface molecule expression was analyzed using FACS (Becton Dickinson Immunocytometry Systems, San Jose, CA). The efficiencies of cell separation with magnetic bead-tagged Abs were also assessed with FACS on the day of selection. In each step for the staining, 5 × 104 cells were stored with specific Abs for 30 min at 4°C in 50 μl of PBS containing 2% of BSA and 0.1% of sodium azide. For the staining of CD1a, HLA-class I and -class II mouse monoclonal anti-human CD1a (M-T102, IgG2b; PharMingen, San Diego, CA), anti-HLA-ABC (W32/6, IgG2a; Serotec, Oxford, England), or anti-HLA-DR (L234, IgG2a) Abs and subsequently PE-conjugated rat monoclonal anti-mouse IgG κ-chain Ab (X36, IgG1) were used. The phenotypes of DC were examined by staining with the following Abs. The PE-conjugated mouse monoclonal anti-human Abs were anti-CD3 (SK7, IgG1), -CD4 (SK3, IgG1), -CD8 (SK1, IgG1), -CD19 (4G7, IgG1), -CD54 (LB-2, IgG1), and -CD56 (MY31, IgG1). The FITC-labeled mouse monoclonal anti-human Abs were anti-CD11c (3.9, IgG1; Serotec), -CD14 (MφP9, IgG2b), -CD16 (B73.1, IgG1), and -CD40 (EA-5, IgG1; Calbiochem, La Jolla, CA). Biotinylated mouse monoclonal anti-human CD80 (BB1, IgM; PharMingen) and anti-human CD86 (IT2.2, IgG2b; PharMingen) Abs and PE-labeled streptavidin (Life Technologies Biomedical Research Laboratories, Gaithersburg, MD) were used for CD80 (B7-1) and CD86 (B7-2) staining. Isotypic Ab (PE- or FITC-labeled or purified mouse IgM, IgG1, IgG2a, and IgG2b) was substituted for specific Abs to obtain negative controls. All Abs except anti-CD1a, -HLA-ABC, -HLA-DR, -CD11c, -CD40, -CD80, and -CD86 were purchased from Becton Dickinson. For the comparison of HLA and costimulators, the mean fluorescence intensity of stained DC (MFIs) and that of controls (MFIc) were measured using a Consort 30 software program (Becton Dickinson). The degree of surface molecule expression was estimated as the ratio of MFIs to MFIc (MFIs/MFIc) and expressed as the net fluorescence intensity (NFI). All samples were assayed in duplicate.
Mixed lymphocyte reaction
To evaluate the allostimulatory capacity of DC, MLR was performed. To compare the function between DC from HCV-infected patients (HCV-DC) and those from donors (N-DC), allogeneic CD4 T cells were obtained from the same healthy volunteer. The CD4 T cells were prepared from PBMC by removing CD8-positive cells, monocytes, B cells and NK cells with magnetic bead-tagged mouse monoclonal anti-human CD8, CD14, CD19, and CD56 Abs (PerSeptive Diagnostics, Cambridge, MA) according to the manufacturer’s instructions. After the separation, the degrees of positivity of these cells in the samples were all less than 5% and those of CD4-positive cells were more than 95%, respectively (data not shown). After DC had been treated with 50 μg/ml of mitomycin C (MMC) for 45 min at 37°C, they were suspended in DC medium and placed at 1 × 102-2 × 104/well on 96-well flat-bottom culture plates. The CD4-positive cells were mixed with DC or monocytes at 2 × 105/well and cultured for 5 days at 37°C, 5% CO2. From the beginning of some series of the experiments, various cytokines, reagents, or anti-cytokine Abs were added to the culture. Recombinant IL-2 was added at 10, 50, or 100 U/ml, IFN-γ at 500 U/ml, and IL-12 at 10 or 50 ng/ml, respectively. Indomethacin was supplied at 10 μM and l-NMMA was at 5 μM. Neutralizing Abs for IL-10 or FasL were pulsed at 10 μg/ml and anti-TNF-α Ab at 5 μl/well. As controls, rat or hamster IgG or rabbit whole serum was used as substitute for the Abs. During the last 16 h of incubation, pulsing was done with 2.0 μCi/well of [3H]thymidine (ICN Pharmaceuticals, Irvine, CA). Assays were performed in triplicate wells. On day 5, the cells were harvested, and the amounts of [3H]thymidine incorporated to responder cells were counted with a beta counter. The ratio of MLR was determined by the ratios of cpm between HCV-DC and N-DC in the presence of the same reagents at a T cell/DC ratio of 10/1.
Detection of apoptosis
The cells were recovered on day 4 of the MLR culture and stained with FITC-conjugated Annexin V (Trevigen, Gaithersburg, MD) according to the manufacturer’s instructions. The cells not stained served as controls. After the regions of interest had been set separately on DC or T cells, the positivity of Annexin V-bound cells was assessed using FACS (Becton).
Endocytosis assay with dextran
To evaluate the endocytosis potential of DC, 1 mg/ml of FITC-dextran was supplied to 2 × 105 cells of DC and was incubated for 30 min at 37°C. As a control, the DC were given the same doses of FITC-dextran and stored for 30 min at 4°C. After the incubation, the DC were washed three times with ice-cold PBS and subjected to FACS analysis. The degree of uptake of FITC-dextran was assessed by the ratio of MFI between DC stored at 37°C and those at 4°C.
T cell response against recall Ags
DC, 5 × 104, were given influenza A Ag diluted at 1:500 (v/v) or HCV core protein at 10 μg/ml, respectively, and incubated overnight at 37°C. Next, Ag-pulsed DC or unpulsed ones were treated with mitomycin C in the same way as MLR and washed two times with RPMI 1640. Next, DC was placed at 2 × 104/well on 96-well flat-bottom culture plates. Autologous CD4 T cells were prepared from PBMC by removing CD8-, CD14-, CD19-, and CD56-positive cells with magnetic bead-tagged Abs in the same manner as described in the above section. The responder cells were added to DC at 2 × 105/well and incubated for 5 days at 37°C, 5% CO2. During the last 16 h, 2.0 μCi/well of [3H]thymidine (ICN Pharmaceuticals) was pulsed. Assays were performed in triplicate wells. On day 5, the amounts of [3H]thymidine incorporated to T cells were counted with a beta counter. The stimulation index (SI) was defined as the ratio between mean cpm obtained with Ag-pulsed DC and those with unpulsed DC.
RT-PCR of IL-12 mRNA
Detection of mRNA of IL-12 p35 or p40 in DC at the baseline or after the stimulation with recall Ags was done by means of RT-PCR. A sample of 1 × 105 cells was incubated for 48 h in the presence or absence of 1:500 diluted influenza A Ag or 10 μg/ml of HCV core Ag. As used for the internal controls, β-actin mRNA was amplified simultaneously. To discriminate differences in signal intensities of amplified products, the numbers of PCR cycles had been preliminarily optimized.
Total RNA was extracted from 4 × 104 of DC with ISOGEN buffer (Nippon Gene, Toyama, Japan) based on the acid guanidine phenol chloroform method. Next, 2 μg of total RNA was suspended in reaction buffer containing 300 ng of random primers (Life Technologies), 500 μM dinucleotide triphosphate (dNTP) mixture, and 10 mM DTT. The samples were reverse transcribed with 200 units of Superscript II RNase H− reverse transcriptase (Life Technologies) according to the manufacturer’s instructions. Five microliters of cDNA templates was mixed with solution containing 10 pmol each of sense and antisense primers, 0.5 units of Taq polymerase (Takara Shuzo, Kyoto, Japan) and 200 μM of dinucleotide triphosphate. Primers used for PCR were: IL-12 p35, sense, 5′-CTTCACCACTCCCAAAACCTG-3′; antisense, 5′-AGCTCGTCACTCTGTCAATAG-3′; IL-12 p40, sense, 5′-CATTCGCTCCTGCTGCTTCAC-3′; antisense, 5′-TACTCCTTGTTGTCCCCTCTG-3′; β-actin, sense, 5′-TCTACAATGAGCTGCGTGTG-3′; antisense, 5′-GGTGAGGATCTTCATGAGGT-3′. PCR was conducted at 94°C for 30 s, 55°C for 1 min, and 72°C for 1 min and was followed by 10 min extension at 72°C. For amplification, 40 cycles were needed for IL-12 p35, 38 cycles for IL-12 p40, and 30 cycles for β-actin. Amplified PCR products were stained with ethidium bromide, electrophoresed on 1.5% agarose gel, and observed under ultraviolet light. The numbers of base pairs of IL-12 p35, IL-12 p40, and β-actin products were 533, 267, and 314, respectively.
ELISA for cytokines
The cytokines produced from DC or DC/T cell were quantified with ELISA kits. In MLR or recall Ag-specific responses, the supernatants were collected on day 4. To assess the potentials of DC for the production of IL-12 p70, 2 × 105
Statistical analysis
The Wilcoxon test or Mann-Whitney U test was used where appropriate to compare the ages, the NFI of surface molecules, the cpm of [3H]thymidine, the ratio of MLR, the SI of T cell response, and cytokine levels between patients and volunteers. A p value of less than 0.05 was considered statistically significant.
Results
Phenotypic analysis of DC
After 7 days of culture under GM-CSF and IL-4, the cells from patients or volunteers exhibited a DC morphology with veiled process and dendrites. With regard to the phenotypic markers of mononuclear cells, the cells were CD4dim but were negative for CD3, CD8, CD14, CD16, CD19, and CD56 (data not shown). In addition, the cells were strongly positive for CD1a, HLA-ABC, HLA-DR, CD11c, CD40, CD54, CD80 (B7-1), and CD86 (B7-2). The positive ratios of CD1a or CD11c, which are characteristic of DC phenotype, were more than 95%, respectively (data not shown). These results showed that the generated cells from both patients and donors were morphologically and phenotypically compatible with DC. Among the HLA and costimulatory molecules examined, CD86 expression in HCV-positive patients was weaker than that in volunteers (Fig. 1⇓ and Table II⇓). The yields of DC were not different between patients and donors (5 ± 3% vs 5 ± 1%).
Flow cytometric analyses of surface molecules on DC generated from PBMC obtained from a healthy volunteer (A) or an HCV-infected patient (B). Representative results of one of the subjects are shown.
Expressions of surface molecules on DC from normal volunteers or HCV-infected patientsa
Low MLR with DC generated from HCV-infected patients
In MLRs, DC exhibited significantly more stimulatory capacity than freshly prepared monocytes for the proliferation of allogeneic T cells (Fig. 2⇓A). However, the degree of T cell proliferation with HCV-DC was significantly lower than that with N-DC (Fig. 2⇓A). The reduced allogeneic response with HCV-DC was confirmed with the comparison of the cpm at a T cell/DC ratio of 10/1 with large numbers of subjects (p < 0.0001) (Fig. 2⇓B). Since the responders were identical in each series of the comparison, the difference in T cell proliferation mainly depended on the DC difference.
A, [3H]Thymidine incorporation to allogeneic CD4 T cells cultured with DC or monocytes in MLRs. Closed circles are cpms obtained with DC from a normal donor (N-DC), and open circles are those obtained with DC from an HCV-infected patient (HCV-DC), respectively. Closed squares are those with freshly prepared monocytes from donors (N-monocytes). The vertical bars indicate means ± SD. Representative results from the subjects are shown. ∗, p < 0.05 vs the cpm with N-DC; ∗∗, p < 0.01 vs the cpm of [3H]thymidine incorporated into T cells cultured with N-DC. B, The ratio of cpm in allogeneic MLR between with HCV-DC and with N-DC at a T cell/DC ratio of 10/1. #, p < 0.0001 vs N-DC.
Apoptosis is involved in activation-induced cell death, which is one of the mechanisms regulating immune reaction (22). Early in the apoptotic process, phosphatidylserine (PS) becomes exposed on the cell surface by flipping from the inner to the outer leaflet of the cytoplasmic membrane. Thus, early apoptotic cells are bound to Annexin V via phosphatidylserine (23). To investigate whether lower MLR by HCV-DC is due to the induction of apoptosis in DC or CD4 T cells, the cells were stained with FITC-Annexin V. Flow cytometric analysis showed that the positivities of stained DC or T cells were low and that they were not different between patients and volunteers (Fig. 3⇓). These results demonstrated that the lower T cell response caused by HCV-DC is due to the low stimulatory capacity of DC, and not to the susceptibility or inducibility of apoptosis.
Representative results of the staining T cells or DC with Annexin V-FITC recovered on day 4 of MLR. The cells were from samples with N-DC (A) or from those with HCV-DC (B). The shaded curves indicate the fluorescence intensities of cells treated with Annexin V-FITC and the unshaded curves are those without Annexin V-FITC, respectively.
Maintaining of Ag uptake and presentation to T cells in HCV-DC
Dendritic cells actively uptake dextran by way of mannose receptors, which is one of the endocytosis pathways for soluble Ags into DC (24, 25). Flow cytometric analysis revealed that the amount of FITC-dextran taken up by DC did not differ between HCV-infected patients and healthy donors (NFI; 16 ± 3 vs 13 ± 5) (Fig. 4⇓).
Representative results of endocytosis assay of N-DC (A) or HCV-DC (B) with FITC-dextran. The shaded curves indicate the fluorescence intensities of DC incubated with FITC-dextran at 37°C, and the unshaded curves are those incubated with FITC-dextran at 4°C, respectively.
To examine whether or not the low stimulatory potential of HCV-DC is alloantigen specific, we compared the recall Ag-specific T cell response with DC between the groups. For this purpose, we selected two viral Ags, such as influenza A and HCV. Since influenza A virus has been endemic in Japan, both groups of subjects were presumed to have been exposed to influenza A Ag. Among the regions in HCV, we selected the core Ag, which is considered to be the most immunogenic of CD4 T cell responses (26, 27). The SI of T cell response against influenza A Ag did not differ between patients and donors (median 4.86 vs 4.36) (Fig. 5⇓A). By contrast, the SI against the HCV core was significantly higher in patients than that in donors (median 1.8 vs 1.0, p < 0.05) (Fig. 5⇓B). These results indicated that HCV-DC possesses the same potentials as N-DC for the uptake of soluble Ags, presenting them and stimulating T cell proliferation.
SI against influenza A Ag (A) or against HCV core protein (B) in autologous CD4 T cell responses with N-DC or HCV-DC. SI is the ratio of cpm between samples with Ag-pulsed DC and those with unpulsed ones. The horizontal bars indicate the medians of SI. ∗, p < 0.05 vs N-DC.
Low IFN-γ production during MLR with HCV-DC
Cytokines play key roles in determining the strength and/or the phenotypes of the T cell response (6). Thus, we measured cytokine production from DC or CD4 T cells during MLR or T cell response. We selected eight histologically proven, moderate chronic hepatitis patients and eight healthy donors for the comparison. To determine the culture periods for the assessment, we analyzed IL-12 p35, IL-12 p40, IL-2, IL-4, IL-10, and IFN-γ levels in the serially collected supernatants from allogeneic or autologous T cell cultures with N-DC. No significant differences were observed in each cytokine level among samples collected from day 2 to day 4 (data not shown). Therefore, we used the samples collected on day 4 of culture for the analysis.
IL-12 p70 is a biologically active form, which consists of two covalently linked chains of p35 and p40 subunits (28). In MLR, the IL-12 p70 levels were below the threshold in both patients and volunteers. The levels in IL-12 p40, IL-2, IL-4, and IL-10 did not differ between the groups (Table III⇓). By contrast, IFN-γ levels in samples from HCV-DC were significantly lower than those from N-DC (Table III⇓).
Cytokine production during allogeneic MLRa
Next, we compared the cytokine levels during T cell response with influenza A- or HCV core-pulsed DC. Regardless of the stimulation, IL-12 p70 were below the detection limit in both patients and donors. The IL-12 p40 levels were elevated in HCV-DC with either Ag treatment (Table IV⇓). The pulsing influenza Ag to N-DC up-regulated their IL-2 and IFN-γ production, which was also observed after pulsing to HCV-DC. The addition of HCV core to N-DC did not affect the cytokine production. By contrast, the pulsing core to HCV-DC enhanced IL-10 and IFN-γ release (Table IV⇓).
Cytokine production during T cell response with recall Ag-pulsed or -unpulsed DCa
Low expression of IL-12 p35 and p40 transcripts in HCV-DC and low IL-12 p70 production from HCV-DC with LPS stimulation
Since the IL-12 p70 levels were below the threshold of ELISA in this system, we examined the expression of IL-12 p35 and p40 transcripts in DC at the baseline and compared the levels before and after the Ag stimulation. Without the Ags, both IL-12 p35 and p40 expressions in HCV-DC were lower than those in N-DC (Fig. 6⇓). However, those expressions in HCV-DC were up-regulated by pulsing either of the Ags (Fig. 6⇓).
Representative results of RT-PCR analyses for IL-12 p35, IL-12 p40, and β-actin mRNA expression in N-DC or HCV-DC. The amplified products of IL-12 p35, IL-12 p40, and β-actin are 533, 267 and 314 base pairs, respectively. R, DC without pulsing Ags; I, influenza A Ag-pulsed DC; C, HCV-core Ag-pulsed DC.
LPS is one of the most potent stimulants triggering IL-12 production from DC (28). We examined IL-12 p70 release from DC in the same eight patients and eight donors as described in the cytokine studies. With a stimulation of 100 pg/ml of LPS, no significant difference was observed in IL-12 p70 levels between HCV-DC and N-DC (median, 0 pg/ml vs 9 pg/ml, range 0–8 pg/ml vs 0–110 pg/ml). However, with 1 ng/ml of LPS, HCV-DC produced significantly less IL-12 p70 than N-DC (median, 3 pg/ml vs 17 pg/ml, range, 0–39 pg/ml vs 0–525 pg/ml, p < 0.01, Mann-Whitney U test), showing that HCV-DC is less potent in IL-12 p70 productivity.
Reversal by exogenous IL-2 or IL-12 of the low MLR with HCV-DC
We investigated whether the decreased allostimulatory function in HCV-DC could be reversed by exogenous modulations. We compared the changes of MLR ratios with the addition of various reagents. Neutralizing Abs against FasL or TNF-α did not restore the decreased T cell proliferation (Fig. 7⇓A). With regard to the DC-derived inhibitory factors for APC or T cells, IL-10, PGE2, and nitric oxide (NO) are possible candidates (24, 25, 29, 30, 31). PGE2 is released from APC and directly inhibits IFN-γ and IL-12 production from T cells or APC (30). DC-derived NO or IL-10 prevents an excessive degree of allogeneic T cell proliferation (31). Furthermore, IL-10 is also a potent deactivating factor for APC (32). For the reversal of these inhibitors, we added indomethacin, l-NMMA, or anti-IL-10 Ab to the MLR culture. Nevertheless, the treatments did not improve the response to HCV-DC (Fig. 7⇓B).
Changes in the cpm ratios in allogeneic MLR with or without the addition of various reagents to the MLR culture. The ratio was calculated from the cpm in MLR obtained with HCV-DC to those with N-DC at a T cell/DC ratio of 10/1 in the presence of various Abs (A), reagents (B), or cytokines (C). ∗, p < 0.05 vs HCV-DC without reagent, ∗∗, p < 0.01 vs HCV-DC without reagent.
Finally, we supplied IL-2, IL-12, or IFN-γ to the cells. Both IL-2 and IL-12 restored the MLR with HCV-DC in a dose-dependent manner (Fig. 7⇑C). However, the addition of IFN-γ exhibited limited improvement (Fig. 7⇑C). The addition of these cytokines to MLR with N-DC did not increase T cell proliferation (data not shown). In respect of IFN-γ response with HCV-DC after such treatments, 100 U/ml of IL-2 showed a slight increase in IFN-γ levels (data not shown). By contrast, 100 pg/ml of IL-12 significantly increased IFN-γ levels compared with those without it (median, 1057 pg/ml vs 0 pg/ml, range 899-1983 pg/ml vs 0 pg/ml, p < 0.01, Mann-Whitney U test). These results suggested that the low allogeneic response to HCV-DC is mainly due to the decreased production of IL-12.
Discussion
In this study, we compared the functions of DC generated from HCV-infected patients with those from healthy volunteers. We showed that the allogeneic MLR with HCV-DC were lower than those with N-DC, whereas the autologous T cell response against the recall Ags was not. Several explanations are possible for the mechanisms of the reduced allogeneic response with HCV-DC. In general, alloreactivity of T cells is influenced by the number of matched pairs in HLA class II alleles between samples (33). To examine whether lower allogeneic capacity of HCV-DC arises from more HLA class II identities or not, we typed HLA-DRB1 and -DQB1 alleles in some patients and donors and compared their MLR response. Between the patient pairs with one or two HLA-DRB1 or -DQB1 alleles identity and those without, the MLR ratios were not significantly different (data not shown), suggesting that the influence of HLA class II distribution in subjects on the DC-mediated allogenicity may be negligible.
One of the possible explanations is the occurrence of apoptosis during MLR. In mice, CD8α-positive DC has been reported to possess functional FasL and leads CD4 T cells to apoptosis (34, 35). Also, DC often releases NO and then induces apoptosis in an autocrine manner (31). However, in this study, the recovered cells were not apoptotic and no reversal of the low MLR with HCV-DC occurred with anti-FasL, anti-TNF-α Abs, or l-NMMA. Thus, apoptosis is not involved in such a low T cell response. The presence of DC-derived inhibitory factors in T cell proliferation is also unlikely. The failure to restore the MLR with indomethacin, l-NMMA, or anti-IL-10 Ab showed that PGE2, NO, or IL-10 did not contribute to the low response with HCV-DC.
Another possibility is the reduction of immunogenic DC or the increase of tolerogenic DC in the generated HCV-DC population (36). The DC yields for patients were the same as those for healthy donors. Phenotypic markers have been proposed, such as CD11c, which characterizes the DC tolerogenicity (37). However, the phenotypes on DC were similar in patients and donors. Recently, some investigators have shown that exogenous IL-10 prevents maturing DC from achieving immunogenicity (24, 25, 29). To exclude the possibility that HCV-DC progenitors naturally produce IL-10 in vitro and suppress their maturation in an autocrine manner, we examined whether the neutralization of IL-10 during DC generation has an influence on the MLR. We found that anti-IL-10 Ab-treated DC did not differ in the MLR from untreated ones (data not shown). Collectively, the reduced allostimulatory function of HCV-DC does not arise from DC heterogeneity, but might already have been primed in vivo in each DC, possibly due to HCV infection.
With regard to the APC dysfunction in each HCV-DC, the decreased expression of costimulators and/or the decreased cytokine production might be involved. Among the costimulators examined, lower CD86 expression was found on HCV-DC. The importance of the CD86-CD28 system in T cell responses has been well-demonstrated (38). Moreover, it is reported that the T cell anergy induced by the lack of CD80/CD86-CD28 signaling can be reversed with exogenous IL-2 (38). In agreement with those reports, 50 or 100 U/ml of IL-2 improved MLR with HCV-DC in this study. However, even if CD86 expression on HCV-DC is relatively lower than that on N-DC, the levels are still higher than those on freshly prepared monocytes (data not shown). Additionally, IL-2 levels in the MLR samples did not differ between HCV- and N-DC. Therefore, the lower CD86 expression on HCV-DC may not primarily lead to MLR reduction.
The IFN-γ levels in MLR with HCV-DC were extremely low, whereas other cytokines were at nearly the same levels as those with N-DC. In this system, CD4 T cells, but not DC, are the main producers for IFN-γ. Thus, the reduction of IFN-γ-inducing factors, such as IL-12, or the increase of IL-10 may contribute to the low-level IFN-γ release from T cells. In this regard, exogenous IL-12 enhanced IFN-γ production in the MLR with HCV-DC, whereas anti-IL-10 Ab did not (data not shown). Therefore, the low levels of bioactive IL-12 from HCV-DC may lead to low levels of IFN-γ. Furthermore, the successful recovery in T cell proliferation with IL-12, but not with IFN-γ, implied that the impaired IL-12 production may be crucially involved in the low MLR with HCV-DC.
Bioactive IL-12 p70 is a heterodimer composed of p35 and p40 subunits, and those expressions are differently regulated (28). The IL-12 p70 levels in MLR were below the threshold of ELISA with either HCV-DC or N-DC in the present study. The IL-12 p40 levels were higher than p70 levels, but the levels were not different between HCV- and N-DC. At the baseline, the expression of IL-12 p35 and p40 transcripts in HCV-DC were lower than those in N-DC, suggesting that HCV-DC is less potent in IL-12 productivity. The factors directly triggering IL-12 production from DC are known to be bacterial products or signals through the CD40-CD40L pathway (28). In the present study, the HCV-DC released significantly less IL-12 p70 than N-DC with LPS stimulation, which is consistent with the RT-PCR analysis. At present, it is not known why HCV-DC exhibits low expression of IL-12 transcripts. HCV or HCV-derived products may interfere, either directly or indirectly, with the transcription machinery of p35 or p40 genes. Recently, McRae et al. reported that type I IFNs directly down-regulate IL-12 release from DC (39). Hypothetically, HCV may lead to the production of endogenous IFNs and subsequently reduce IL-12 productivity of DC. Direct inoculation of HCV particles or transduction of HCV genome to N-DC, if successful, should provide clues to illuminate this issue.
The impaired IL-12 production from PBMC or APC has been reported for other viral infections, such as with HIV or measles virus (40, 41, 42). In HIV infection, low levels of IL-12 correlated well with the reduction of CD4 T cell count and/or the advancement of the immunodeficiency state (40, 41). In the present study, we found no signs of a generalized immune suppression in the HCV-infected patients, suggesting that a low allogeneic response does not directly reduce the immunocompetency. In this regard, the retention of recall Ag-specific CD4 T cell responses with HCV-DC shown here may contribute to maintain an immune competent state. The IL-12 p40 levels or MLR ratios of patients did not show any correlation with clinical parameters, including alanine aminotransferase levels or HCV-RNA quantity (data not shown). Chouaib et al. reported that, at the sensitizing phase of MLR, endogenous IL-12 plays key roles in the cell proliferation and CTL differentiation (43). Thus, to estimate the implications of low allogeneic response for the pathogenesis of HCV infection, we compared the allogeneic CD8 CTL activity induced with HCV-DC and that with N-DC against relevant DC as targets. In four series of experiments, the lytic activity obtained with HCV-DC was lower than that with N-DC (20 ± 6% vs 44 ± 8%, p < 0.05). These results implied that HCV-DC is less potent than N-DC for the induction of CD4-independent CTL. In general, virus-specific CTL response is essential for the eradication of infected cells. Thus, it is probable that low allostimulatory capacity of HCV-DC is involved, not all but in some part, in the development of HCV persistence.
In T cell response with HCV-DC, T cells specifically proliferated against the recall Ags. The IFN-γ production was up-regulated with the Ag-pulsed HCV-DC compared with those with unpulsed ones. Although the IL-12 p70 levels were below the detection limit, IL-12 p40 levels rose with either of the Ags. The RT-PCR analyses also revealed that the pulsing Ags increased the expressions of both IL-12 p35 and p40 transcripts in HCV-DC. Therefore, the activation of HCV-DC with the recall Ags for the stimulation of IFN-γ production and T cell proliferation may be due to the up-regulation of IL-12.
The low levels of IL-12 and the subsequent low IFN-γ in allogeneic response with HCV-DC could impede Th1 polarization afterward (6, 28). Additionally, in the HCV core-pulsed HCV-DC cultures, IL-10 production was significantly elevated, which also inhibits the Th1 development. Although the IL-4 levels were extremely low under such culture conditions, these data suggested that the core-specific responses with HCV-DC tend to become Th2-like ones, not directly but indirectly, by impeding Th1 skewness. Recent studies reported that the dominance of the Th2-like subset in HCV infection is related to the development of HCV chronicity (3). The present study showed that HCV-DC tends to promote a Th2-like differentiation on encounters with alloantigens or HCV Ags and to provide a favorable environment for HCV persistence.
In summary, the stimulatory potentials of HCV-DC are impaired against alloantigens but not against recall Ags. The possible mechanisms for such impairment are their relatively low expression of CD86 and/or IL-12 at the baseline, and low allogeneic response could be overcome by the addition of IL-2 or IL-12. However, when HCV-DC encounter recall Ags, they are activated to restore the potentials for cytokine production and stimulation of T cell proliferation. Our findings may shed light on the role of DC in the pathogenesis of HCV infection and provide useful information on the modulation of DC function. In the application of DC recovered from HCV-infected patients as an immunopotentiator in vivo, the reduced DC functions demonstrated in this study need to be considered.
Footnotes
-
↵1 This work is supported by a Grant-in-Aid from the Ministry of Education, Science and Culture of Japan.
-
↵2 Address correspondence and reprint requests to Dr. Norio Hayashi, First Department of Medicine, Osaka University Medical School, 2-2 Yamada-oka, Suita, 565-0871, Japan. E-mail address: gastro{at}medone.med.osaka-u.ac.jp
-
↵3 Abbreviations used in this paper: HCV, hepatitis C virus; DC, dendritic cell; HCV-DC, DC obtained from HCV-infected patients; N-DC, DC obtained from normal donors; FasL, Fas ligand; l-NMMA, NG-monomethyl l-arginine acetate; MFI, mean fluorescence intensity; MFIs, MFI of stained DC; MFIc, MFI of controls; NFI, net fluorescence intensity; SI, stimulation index; NO, nitric oxide.
- Received July 23, 1998.
- Accepted February 11, 1999.
- Copyright © 1999 by The American Association of Immunologists
References
In this issue
