No evidence for abnormal immune activation in peripheral blood T cells in patients with hepatitis C virus (HCV) infection with or without cryoglobulinaemia
Professor Patrice Cacoub Md Department of Internal Medicine, Hôpital La Pitié-Salpêtrière, 83 Boulevard de l'Hôpital, 75651 Cedex 13 Paris, France.
The aim of this study was to investigate the peripheral blood lymphocyte (PBL) phenotypes and T cell repertoire in patients with HCV infection, with or without mixed cryoglobulinaemia (MC). The patients were: Group 1, 23 patients with HCV infection and MC; Group 2, 14 patients with HCV infection but without MC; Group 3, 10 patients with symptomatic essential MC. Twenty healthy blood donors were used as controls. Blood lymphocyte counts were determined, and flow cytometry was used to measure proportions of B cells (CD19+), natural killer (NK) cells (CD16+ CD56+), T cells (CD3+), CD4+ T cell subsets (memory CD4+ CD45RO+; naive CD4+ CD45RO−; Th0/Th2 CD4+ CD7−; activated CD4+ CD25+), and CD8+ T cell subsets (immunoregulatory CD8+ CD57+; cytotoxic CD8+ S6F1+, activated CD8+ CD25+). Bias in the usage of T cell receptor (TCR) Vβ chains was studied in a subgroup of 10 representative patients of Group 1 using a polymerase chain reaction (PCR) analysis of the Vβ segments with a series of 20 oligonucleotides specific for the Vβ families. The three groups were comparable for blood lymphocyte counts, and we observed no abnormal repartition of the following PBL subsets: T cells (CD3+), CD4+ and CD8+ subpopulations, B cells (CD19+), and the NK cells (CD16+ 56+). In none of the groups could we observe lymphocyte ex vivo activation as assessed by the normal expression of the activation cell markers: CD25 on CD4+ or CD8+ T cells, or CD5 on B cells. The repartition of naive and memory (CD45RO−/RO+) CD4+ T cells was normal and we did not observe any amplification of the CD4+ CD7− T cell subset differentiated in vivo in Th0/Th2 cells. There was no significant amplification of cytotoxic (SF6+) and immunoregulatory (CD57+) CD8+ T cells in HCV patients with or without MC. Finally, the usage of Vβ families in the TCR repertoire was normal in the patients tested. In patients with chronic HCV infection, with or without MC, we did not find any significant expansion or abnormal activation of T, B and NK cell subsets, dysbalance of the naive/memory subsets, or expansion of the Th0/Th2 subpopulation. These findings differ from the profound immune alterations that are observed in other chronic infections such as HIV or Epstein–Barr virus. Although this study was restricted to the peripheral blood, it suggests that in chronic HCV infection, MC is not the consequence of a chronic activation or dysregulation of the peripheral blood immune cells.
The syndrome of mixed cryoglobulinaemia (MC) is characterized by the clinical triad of purpura, arthralgia and asthenia associated with cryoglobulins composed of different immunoglobulins, of which the IgM rheumatoid factor component is monoclonal in type II MC and polyclonal in type III MC [ 1]. The clinical manifestations of MC result from an immune complex-type vasculitis and the deposition of immunoglobulins and complement in the lesions [ 2]. The antigens that trigger the production of antibodies found in immune complexes are usually unknown. In view of the frequent biologic and histological liver abnormalities in MC [ 3], the hepatotropic antigens that can trigger production of MC have been sought for many years. An important role for HCV in MC pathogenesis has been demonstrated by studies that found anti-HCV antibodies in about 60–80% of patients with ‘essential’ MC [ 4–9]. A high prevalence of HCV RNA was found in the sera and/or cryoprecipitates of these patients, lending further support to the hypothesis that HCV has an important role in the pathogenesis of MC. With quantitative polymerase chain reaction (PCR), HCV RNA was found to be 20–100 times more concentrated in cryoprecipitates than in supernatants [ 8, 10]. However, the precise mechanisms involved in the pathogenesis of MC in patients with chronic HCV infection remain unknown.
HCV is known to infect lymphocytes [ 11], and a chronic viral replication inside the immune system might induce immunological disorders such as MC or autoimmune manifestations as a consequence of permanent immune cell activation or dysregulation. Other viruses that infect the immune system with a high replication level, either chronically, such as HIV, or acutely, such as Epstein–Barr virus (EBV) or cytomegalovirus (CMV), induce major dysregulation and activation of the lymphoid subsets that are detectable in periphery. In some disorders, chronic antigenic stimulation can bias T helper cell differentiation toward the Th2 subset, whose production of IL-4, IL-5, and IL-10 stimulates B cell proliferation and antibody production [ 12]. Chronic viral replication usually activates CD8+ T cells, prompting their differentiation into cytotoxic effector cells, and the expansion of CD8+ CD57+ cells, which down-regulate lymphocyte activity. It is also possible that a biased usage of T cell receptor (TCR) variable regions might occur during such chronic viral infection and participate in the pathogenesis of the autoimmune disorders associated with HCV [ 13–15].
We previously described in healthy donors a minor subset of CD4+ T cells that preferentially produce IL-4 and IL-10 but low amounts of IL-2, and were defined as Th0/Th2 cells [ 16]. As this T cell subset is expanded in clinical situations associated with chronic antigenic stimulation such as HIV infection [ 17] and transplantation [ 18], we have now looked for a possible relation between Th0/Th2 cells and the development of MC, with or without HCV infection. In addition, we have examined the role of immune cell activation and dysregulation in MC associated with HCV, by determining blood lymphocyte phenotypes in HCV patients with or without MC, with special interest in T cell activation status, and the expansion of memory (CD45RO+) CD4+ T cells.
PATIENTS AND METHODS
We studied different groups of patients. Group 1 included 23 patients (14 males, nine females, aged 54 years (range 36–68 years)) with chronic HCV infection and MC. Chronic HCV infection was defined by the presence of alanine aminotransferase levels more than twice the upper limit of normal range, anti-HCV antibodies detected by third generation test (ELISA, recombinant-based immunoblot assay (RIBA)), liver biopsy findings compatible with chronic hepatitis C, and no other cause of liver dysfunction (e.g. chronic hepatitis B, autoimmune hepatitis, primary biliary cirrhosis). Patients were considered to have a significant cryoglobulin if they had a minimum serum cryoglobulin level of 0.05 g/l at two determinations. Group 2 included 14 patients (eight males, six females, aged 51 years (range 32–65 years)) with chronic HCV infection but without MC. No patient with HCV infection (Groups 1 and 2) received interferon-alpha (IFN-α) therapy at the time of the study. Group 3 included 10 patients (seven males, three females, aged 57 years (range 34–72 ;years)) with symptomatic ‘essential MC’, without evidence of HCV infection or any other autoimmune, infectious or malignant haematological disorder. In addition, 20 healthy blood donors (aged 26 years, range 23–38 years) were used as a control group (Group 4). All the subjects of Group 4 had normal serum transaminase value (ALT), no anti-HCV antibodies and no cryoglobulins. This study was approved by the local ethical committee, and all subjects gave their informed consent.
Blood was drawn and kept at 37°C until clotted. All sera were stored at −80°C until assayed for anti-HCV antibodies, according to the manufacturers' instructions, using third-generation ELISA (Ortho Diagnostic Systems, Raritan, NJ) and RIBA (Chiron, Emeryville, CA).
Detection of HCV RNA and HCV genotyping
Sera used to detect HCV RNA were stored at −80°C and were previously unthawed, to avoid false negativity due to RNA destruction by RNases and false positivity due to contamination. Serum RNA was extracted, reverse transcribed to make cDNA, and amplified by PCR as previously described [ 19]. To assess the specificity of the PCR products, Southern blotting was performed under stringent conditions using a radiolabelled oligonucleotide probe. All experiments were performed in parallel with (i) serum-free lysis buffer to detect possible contamination prior to amplification (extraction and cDNA step), (ii) the reaction mixture without DNA (PCR step), and (iii) negative sera. Each sample was tested in at least two different series. HCV genotyping was done using a second generation Line Probe Assay (LiPA; Innogenetics, Brussels, Belgium), which employs biotinylated universal primers specific for the 5′ non-coding region of HCV RNA. Amplification products were then hybridized to genotype-specific probes. This identifies HCV types and subtypes 1a, 2a, 2b, 3a, 4 or 5 [ 20].
Analysis of liver biopsy specimen
For all patients with chronic HCV infection (groups 1 and 2), liver biopsy specimens were evaluated according to a previously validated scoring system [ 21], which includes 27 semiquantitatively scored items. Our analysis focused upon the inflammatory activity and the severity of fibrosis. The necroinflammatory activity was graded on a scale of 0–3 (A0–A3 = 0, no activity; 1, mild; 2, moderate; or 3, severe), taking into account the severity of portal and lobular necroinflammatory lesions. Fibrosis was graded on a scale of 0–4 (F0–F4 = 0, no fibrosis; 1, stellate enlargement of portal tract but without septa formation; 2, stellate enlargement of portal tract with rare septa formation; 3, numerous septa without cirrhosis; or 4, cirrhosis).
Detection and characterization of cryoglobulins
Cryoglobulins were isolated from the patients' sera, purified and characterized by immunoblotting at 37°C as previously described [ 22]. Using this method, only 5/131 (3.8%) healthy blood donors had MC, all of them with a cryoglobulin level < 0.03 g/l [ 10]. In the present study, patients were considered to have a significant cryoglobulin if they had a minimum serum cryoglobulin level of 0.05 g/l at two determinations. After immunochemical analysis [ 22], cryoglobulins were classified as type I, II or III, according to Brouet et al. [ 23].
Flow cytometry analysis
A panel of MoAbs was used for this study. All MoAbs were directly labelled with either FITC or PE. Antibodies to CD25 (IOT14-FITC), CD45RO (IOL2-FITC), CD71 (IOA71-FITC), and in a dual colour kit CD4 (IOT4a-FITC) and CD8 (IOT8a-PE), were purchased from Immunotech (Marseille-Luminy, France). Antibodies to CD4 (Leu-3a–FITC-PE), CD8 (Leu-2a–PE), CD45RO (Leu-45RO–PE), HLA-DR (FITC) and in dual colour kits CD3 (Leu-4–FITC) plus CD16+ CD56 (Leu-11 + Leu-19–PE), and CD8 (Leu-2a–FITC) plus CD3 (Leu-4–PE), were obtained from Becton Dickinson (Grenoble, France). Antibodies to CD29 (4B4–PE) were obtained from Coulter Clone (Margency, France). A dual-colour kit CD19 (HD37–PE)/CD5 (DK23–FITC) was purchased from Dako (Glostrup, Denmark). Before staining, peripheral blood lymphocytes (PBL) were isolated on Ficoll–Hypaque gradient (Pharmacia, Uppsala, Sweden). Two-colour stainings were performed in a one-step procedure by incubating cells with MoAbs for 20 min at 4°C. Cells were then washed twice and analysed on a FACScan flow cytometer (Becton Dickinson, Sunnyvale, CA). At least 5000 cells were analysed on the flow cytometer after single gating on lymphoid cells for all MoAb combinations. The percentages of lymphocyte subpopulations were all expressed as percentages of total lymphocyte counts.
Vβ chain of the TCR
To study the hypothesis of a selective in vivo deletion of TCR Vβ chains, we performed in 10 patients of Group 1 a PCR analysis of the Vβ segments with a series of oligonucleotides specific for the Vβ families (1–20). We used a standard PCR analysis of the first 20 Vβ segments of the whole Vβ family usage, as previously described in detail [ 24]. Briefly, RNA extracted from total PBL was submitted to reverse-transcriptase amplification. The 20 oligonucleotides specific for the Vβ segments were used together with the Vβ primer and the Cβ primer as internal control to amplify the C-α fragment. We used for PCR amplification 40 cycles of 1 min at 92°C, 1 min at 52°C and 2 min at 72°C. Of each PCR product, 10% was subjected to gel electrophoresis through 2% agarose gels in presence of ethidium bromide and visualized by exposure to UV light.
All data were expressed as the mean ± s.d. Statistical analysis used χ2 or Fisher's exact test for comparisons of percentages. Mean quantitative values were compared using the unpaired Student's t-test. Non-parametric analyses were done with Mann–Whitney U-test. Significance was assessed at P = 0.05. All calculated P values are two-tailed.
In patients of Group 1 ( Table 1), MC was of type II (n = 7) or type III (n = 16), with a level ranging from 0.05 to 0.76 g/l. In type II MC, the monoclonal components were IgM κ (n = 5), IgG κ (n = 2), IgM λ (n = 1), IgG λ (n = 1), where two patients had two monoclonal components (IgM κ + IgM λ, and IgM κ + IgG κ). In patients of Group 3 ( Table 2), MC was of type II (n = 7) or type III (n = 3), with a level ranging from 0.04 to 0.35 g/l. The monoclonal components were IgM λ (n = 3), IgM κ (n = 2), IgG κ (n = 1), IgG λ (n = 1) and IgA λ (n = 1), where one patient had two monoclonal components (IgM κ + IgM λ). All patients of Group 3 had symptomatic essential MC ( Table 2).
Table 1. .
Virological, immunological and liver histological data of patients with HCV infection and mixed cryoglobulinaemia (Group 1) * HCV genotyping was done using the Line Probe Assay (LiPA) method, as described in Patients and Methods.† Type II and III are mixed cryoglobulins, composed of different immunoglobulins, with a monoclonal component in type II and only polyclonal immunoglobulins in type III, according to Brouet [ 20
] (see Materials and Methods).‡ Metavir score: liver biopsy specimens were evaluated according to a previously validated scoring system, graded on a scale of 0–3 for the inflammatory activity (A0–A3), and 0–4 for the severity of fibrosis (F0–F4).
Table 2. .
Clinical and immunological data of patients with essential mixed cryoglobulinaemia (Group 3) * Types II and III are mixed cryoglobulins, composed of different immunoglobulins, with a monoclonal component in type II and only polyclonal immunoglobulins in type III, according to Brouet [ 20
] (see Patients and Methods).
HCV PCR was positive in most patients tested in Group 1 (13/16 = 81%) ( Table 1), and Group 2 (8/10 = 80%) ( Table 3). The distribution of HCV genotypes was heterogeneous but similar in both groups ( Tables 1 and 3). For Group 1, HCV genotype 1b was found in 6/14 (43%) patients, 2a in 3/14 (22%), 4 in 2/14 (14%), 1a in 1/14 (7%), 1c in 1/14 (7%), and 3 in 1/14 (7%). For Group 2, HCV genotype 1b was found in 3/6 (50%) patients, 3 in 2/6 (33%), and 1a in 1/6 (17%). Liver biopsy specimen analysis often found signs of chronic active hepatitis, but there was no significant difference in the distribution of necroinflammatory lesions or fibrosis between the groups. In patients of Group 1, the necroinflammatory activity scoring was A0 in 2/20 (10%) patients, A1 in 14/20 (70%) and A3 in 4/20 (20%). In this group, the fibrosis score was F0 in 3/20 (15%) patients, F1 in 6/20 (30%), F2 in 8/20 (40%), F3 in 2/20 (10%) and F4 in 1/20 (5%). In patients of Group 2, the necroinflammatory activity scoring was A0 in 3/13 (23%) patients, A1 in 6/13 (46%), A2 in 2/13 (15%) and A3 in 1/13 (8%). In Group 2, the fibrosis score was F1 in 7/13 (54%) patients, F2 in 3/13 (24%), F3 in 2/13 (15%) and F4 in 1/13 (8%).
Table 3. .
Virological and liver histological data of patients with HCV infection without mixed cryoglobulinaemia (Group 2) * HCV genotyping was done using the Line Probe Assay (LiPA) method, as described in Patients and Methods.† Metavir score: live biopsy specimens were evaluated according to a previously validated scoring system, graded on scale of 0–3 for the inflammatory activity (A0–A3), and 0–4 for the severity of fibrosis (F0–F4).
Lymphoid subset analysis
Blood lymphocytes counts were not significantly different between the groups and they were comparable to those of healthy controls: Group 1 1920 ± 420/μl (range 1200–2700); Group 2 2020 ± 550/μl (800–4800); Group 3 1790 ± 340/μl (1100–2700); Group 4 2060 ± 350/μl (1400–2500) ( Table 4).
Table 4. .
Lymphocyte subsets in cryoglobulinaemic patients with or without HCV infection * Group 1, patients with HCV infection and mixed cryoglobulinaemia. Group 2, patients with HCV infection, without cryoglobulinaemia. Group 3, patients with symptomatic essential mixed cryoglobulinaemia. Healthy blood donors were used as controls.† T cells (CD3+
). Natural killer (NK) cells (CD16+
). B cells (CD19+
). T CD4+
helper (memory CD4+
; Th0/Th2 CD4+
; activated CD4+
). T CD8+
; cytotoxic CD8+
, activated CD8+
When lymphocyte subsets of each group were studied, no differences were seen in the percentages or absolute counts of CD3+, CD4+ or CD8+ T cells, and the CD4/CD8 ratios were all normal. When looking for immune activation, no significant increase of the IL-2 receptor α-chain (CD25) was observed on CD4+ or CD8+ cells. Similarly, each of the naive (CD45RO−)/memory (CD45RO+) CD4 subpopulations represented ≈ 50% of the CD4 lymphocytes in all groups tested. We did not observe any significant expansion of the CD4+ CD7− subset, previously shown to be differentiated in vivo in Th0/Th2 cells [ 16–18].
The percentages of immunoregulatory cells expressing CD57, or of cytotoxic cells defined by S6F1 expression, were greater in the CD8+ subsets from patients with HCV (Groups 1 and 2) than in the essential MC group or normal controls. These differences were not statistically significant, however. The proportions of B cells (CD19+) were also normal in all groups, with no enhancement of the CD5+ subset. Natural killer (NK) (CD16+ 56+) lymphocytes also did not significantly differ between the groups.
Vβ chain of the TCR
We analysed TCR Vβ usage of the first 20 Vβ segments of the whole Vβ family in 10 representative patients of Group 1. The profile obtained after amplification from total peripheral blood mononuclear cells was comparable in each case to data obtained in normal controls (data not shown). These results indicate that no bias in Vβ usage was detectable in the peripheral blood TCR in the context of HCV–MC.
Many studies have demonstrated that HCV infection is responsible for the production of the majority of previously named ‘essential’ MC [ 6–11]. However, the mechanism(s) involved in cryoglobulin production during chronic HCV infection remain unknown. From a theoretical point of view, there may be viral, genetic, immunological or environmental factors.
Viral factors have been studied, but in all studies except one [ 25] there was no evidence for a relation between MC and a particular HCV genotype or the level of HCV viraemia. In this study, we looked for viral factors but this search was negative. When we compared HCV patients with or without MC, we did not find significant differences in the frequency of PCR HCV positivity (80%) or the distribution of HCV genotype, with a majority of type 1b, as expected in Europe. Moreover, liver histological lesions were similar between these groups.
The immune response of the host to this chronic viral infection has also been previously analysed. Analysis of the humoral immune response showed a marked increase of IgG1 serum levels, but not of the other IgG subclasses. Serum immunoglobulin levels were not significantly different according to the presence or absence of MC [ 26]. In this study, the distribution of type II/III MC in HCV patients was 1/3–2/3, as previously reported [ 19, 27], and the monoclonal components in the type II MC were mainly IgM κ, with no significant difference seen between HCV+ and HCV−groups.
The present study was designed to analyse the cellular immune disorders induced by chronic HCV infection. We did not find any significant T cell expansion, restriction of Vβ chain of the TCR, dysregulation between naive or memory CD4 lymphocyte subsets, expansion of Th0/Th2, or expansion of activated T cells or B cells. Since only blood lymphocytes were studied, we cannot exclude that such abnormalities might occur in infected tissues. Since HCV is able to infect lymphocytes [ 11, 28], chronic replication in lymphocytes may have induced significant immune alterations both in lymphoid organs and in the periphery. It is possible that a bias in the CD4 T cell response would be seen in a study of virus-specific T cell populations resident in the liver. Intrahepatic CD4 T cells specific for the HCV NS4 antigen may in fact differ from their counterparts in the peripheral blood [ 29]. T cell clones derived from peripheral blood and liver appeared to be qualitatively different in their ability to provide B cell help for certain immunoglobulin subclasses. Moreover, the epitope specificity may be different in these two tissues, as the TCR repertoire of the clones seems also distinct [ 29]. No abnormal usage in the Vβ segment of the TCR was observed when analysing the PBL from 10 HCV patients with MC. This finding supports the lack of massive activation of the PBL in HCV with or without MC, and may be in accordance with results of Davies et al. [ 30], who investigated the TCR repertoire in autoimmune thyroiditis. A bias in Vβ family usage was observed in the thyroid-infiltrating lymphocytes only, but not in the PBL [ 30]. However, with the method used in the present study (PCR), we cannot rule out small changes in percentages of lymphocytes expressing different TCR. Such minor changes might have been detected using flow cytometry analysis, but this method was not available for the analysis of the Vβ segment at that time.
CD4 and CD8 T cell responses are unable to eliminate persistent HCV infection. There is, however, evidence that these cellular immune responses do provide some control of ongoing virus replication during the chronic phase of infection [ 31]. HCV-specific CD4 cell lines have been found to produce cytokines (IL-2, IFN-γ, IL-4, IL-5, tumour necrosis factor-alpha (TNF-α), TNF-β) associated with a Th0 phenotype [ 29]. Looking for a peripheral amplification of Th0/Th2 cells, we investigated the CD4+ CD7− subset, which can be expanded in the context of chronic virus replication, such as HIV infection [ 16, 17], chronic antigen stimulation, such as transplantation [ 18], or with B cell lymphoproliferation. However, we failed to observe any consistent expansion of this particular subset in the peripheral blood. Similarly, we did not find any increase in serum levels of IL-6 which might have a role in the B cell dysregulation observed with MC (data not shown). Thus, although cellular immune responses probably do help control ongoing virus replication, we did not find evidence for the role of PBL B or T subpopulations in the production of MC.
In patients with chronic HCV infection with or without MC, we did not find any significant expansion or abnormal activation of T, B and NK cell subsets, imbalance of the naive/memory subsets, or expansion of the Th0/Th2 subpopulation. These findings differ from the profound immune alterations that are observed in other chronic viral infections such as HIV or EBV. Although this study was restricted to the peripheral blood, it suggests that in chronic HCV infection, MC is not the consequence of a chronic activation or dysregulation of the studied subsets of peripheral blood immune cells.
The members of the Multivirc group of La Pitié-Salpêtrière Hospital are: Anatomopathologie: Frédéric Charlotte, Yves Le Charpentier; Anaesthésie-Réanimation: Bruno Riou; Biochimie: Jacques Delattre, Françoise Imbert-Bismuth, Annie Piton; Centre de Transfusion Sanguine: Jean-Jacques Fournel, Loan Nguyen; Chirurgie Cardiaque: Richard Dorent, Iradj Gandjbakhch; Endocrinologie: Luc Foubert, Gérard Turpin; Hépato-gastro-entérologie: Brigitte Bernard, Cécile Blot, Philippe Nguyen, Philippe Mathurin, Joseph Moussali, Pierre Opolon, Michèle Perrin, Thierry Poynard, Rodolphe Sobetski; Immunochimie: Pascale Ghillani, Lucille Musset; Immunologie cellulaire: Brigitte Autran, Patrice Debré; Maladies infectieuses: François Bricaire, Marc Gentilini; Médecine Interne: Patrice Cacoub, Olivier Chosidow, Camille Francès, Pierre Godeau, Serge Herson, Sophie Pelletier, Jean-Charles Piette; Médecine Nucléaire: André Aurengo, Thierry Delbot, Laurence Leenhardt; Neurologie: Pascal Brunet, Olivier Gout, Jean-Marc Leger, Olivier Lyon-Caen; Oncologie Médicale: Eric Antoine, Christian Borel, David Khayat; Pharmacie: Anne-Marie Fievet, Frédérique Plassart, Alain Thuillier; Rhumatologie: Pierre Bourgeois, Nathalie Wrona; Virologie: Françoise Lunel, Jean-Marie Huraux, Vincent Thibaut; Transplantation rénale: Olivier Bitker
This study was presented in part at the 59th Annual Scientific Meeting of the American College of Rheumatology, San Francisco, October 1995.