Impaired Thymic Output in Patients with Chronic Hepatitis C Virus Infection

Authors

  • H. J. Hartling,

    1. Viro-Immunology Research Unit, Department of Infectious Diseases, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark
    2. Department of Clinical Immunology, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark
    Search for more papers by this author
  • J. C. Gaardbo,

    1. Viro-Immunology Research Unit, Department of Infectious Diseases, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark
    2. Department of Clinical Immunology, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark
    Search for more papers by this author
  • A. Ronit,

    1. Viro-Immunology Research Unit, Department of Infectious Diseases, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark
    2. Department of Clinical Immunology, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark
    Search for more papers by this author
  • M. Salem,

    1. Viro-Immunology Research Unit, Department of Infectious Diseases, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark
    Search for more papers by this author
  • M. Laye,

    1. Centre of Inflammation and Metabolism, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark
    Search for more papers by this author
  • M. R. Clausen,

    1. Department of Hepatology, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark
    Search for more papers by this author
  • K. Skogstrand,

    1. Department of Clinical Biochemistry and Immunology, Statens Serum Institute, Copenhagen, Denmark
    Search for more papers by this author
  • J. Gerstoft,

    1. Viro-Immunology Research Unit, Department of Infectious Diseases, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark
    Search for more papers by this author
  • H. Ullum,

    1. Department of Clinical Immunology, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark
    Search for more papers by this author
  • S. D. Nielsen

    Corresponding author
    • Viro-Immunology Research Unit, Department of Infectious Diseases, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark
    Search for more papers by this author

Correspondence to: S. D. Nielsen, Viro-Immunology Research Unit, Department of Infectious Diseases, Rigshospitalet, Copenhagen University Hospital, Blegdamsvej 9, 2100 Copenhagen, Denmark.

E-mail: sdn@dadlnet.dk

Abstract

Altered T cell homeostasis in chronic hepatitis C virus (HCV) infection has been demonstrated. However, it is unknown whether fibrosis is associated with more perturbed T cell homeostasis in chronic HCV infection. The aim of this study was to examine and compare T cell subsets including recent thymic emigrants (RTE), naive, memory, senescent, apoptotic and IL-7 receptor α (CD127) expressing CD4+ and CD8+ T cells as well as telomere length and interferon-γ production in HCV-infected patients with (n = 25) and without (n = 26) fibrosis as well as in healthy controls (= 24). Decreased proportions of CD4+ and CD8+ RTE were found in HCV-infected patients, especially in HCV-infected patients with fibrosis (14.3% (9.7–23.0) and 28.8% (16.1–40.5), respectively) compared with healthy controls (24.2% (16.3–32.1), P = 0.004 and 39.1% (31.6–55.0), P = 0.010, respectively). Furthermore, HCV-infected patients with fibrosis presented with a higher proportion of CD4+ T cells expressing CD127 compared with HCV-infected patients without fibrosis [88.4% (84.5–91.0) versus 83.8% (79.9–86.8), P = 0.016]. Thus, impaired thymic output in HCV infection was found, and high proportion of CD127+ T cells may illustrate a compensatory mechanism to preserve T cell counts.

Introduction

The mechanism leading to development of fibrosis in chronic hepatitis C virus (HCV) infection is incompletely understood [1-3]. It is well established, however, that the host immune response plays an important role in development of fibrosis [2-6]. This is supported by an increased risk of fibrosis in older patients with reduced T cell production and in patients with HIV co-infection or other immune dysfunctions [2-5]. Furthermore, during treatment with interferon (IFN)-α and ribavirin, the host cellular immune response is essential for optimal treatment response [7-15]. It is evident that CD8+ T cells have a major role in eliminating HCV-infected hepatocytes, while CD4+ T cells have a role in activating CD8+ T cells and antigen-presenting cells in chronic HCV infection. Furthermore, anti-inflammatory T cell responses in the form of regulatory T cell responses might have an important role in limiting the hepatic inflammation which is a driver of fibrosis [16, 17]. Thus, T cell homeostasis plays an important role for both the natural history of chronic HCV infection and treatment response.

The T cell homeostasis includes production and destruction of T cells. T cells are produced in the thymus where rearrangement of the T cell receptor takes place. Recent thymic emigrants (RTE) are cells with full immunological repertoire produced in the thymus. The fraction of naïve T cells and RTE can be used to assess thymic output [18, 19]. In contrast, exposure to antigens results in a transition from naïve T cells to effector memory cells, and immune activation can lead to T cell senescence, apoptosis and death. Thus, memory T cells, senescent and apoptotic T cells can be used to assess differentiation and destruction of T cells, respectively. Furthermore, interleukin (IL)-7 is essential for T cell homeostasis by promoting T cell survival and proliferation, and the responsiveness is dependent on expression of the IL-7 receptor (IL-7R) [20-22]. Finally, telomere length can be used to assess T cell turnover, as telomere is shortened by each cell division [23].

T cell homeostasis has been extensively examined in HIV-infected and in HIV/HCV co-infected individuals [24-27]. In contrast, only a few studies have assessed thymic output in HCV monoinfected individuals finding lower or similar proportions of naïve CD4+ T cells and CD4+ RTE compared with healthy controls [28-31]. Furthermore, CD8+ RTE are poorly investigated in general and not described in chronic HCV infection [32]. Finally, to our knowledge, it has not been evaluated whether production or destruction of T cells is associated with fibrosis in HCV infection. We hypothesized increasing fibrosis to be associated with higher T cell turnover.

The aim of the present study was to examine the T cell homeostasis using flow cytometric phenotyping and molecular testing to examine telomere length in treatment naive chronic HCV-infected patients with and without fibrosis and in healthy controls in a cross-sectional set-up.

Materials and methods

Study design

A total of 51 patients with chronic HCV infection and 24 healthy individuals were included in a cross-sectional study as described previously [33, 34]. Inclusion criteria for patients were chronic HCV infection with positive anti-HCV and a positive HCV-RNA for more than 6 months. Patients with chronic HCV infection included a group with fibrosis (= 25) and a group without fibrosis (n = 26) as determined by transient elastography using cut-off 8 kPa [33]. The clinical characteristics are presented in Table 1. Of the 25 HCV-infected patients with fibrosis, 12 had a liver stiffness above 12 kPa which were defined as cirrhosis [33]. All patients were Child–Pugh class A and naïve to HCV treatment. All patients were enrolled from Department of Infectious Diseases or Department of Hepatology, Rigshospitalet, University Hospital of Copenhagen. Healthy subjects were recruited among hospital staff. Informed consent was obtained in writing and verbally from all participants. The study was performed in accordance with the ethical guidelines of the 1975 Declaration of Helsinki and approved by the Local Ethical Committee (H-4-2010-012) and the Danish Data Protection Agency.

Table 1. Clinical characteristics of the study population
 HCV mono-infected with fibrosis (n = 25)HCV mono-infected without fibrosis (= 26)Healthy controls (= 24)
  1. NA, not applicable; IQR, interquartile range; IDU, injection drug use; ALT, alanine aminotransferase Comparisons were made between patients with HCV mono-infection with and without fibrosis and healthy controls.

  2. *P < 0.05 by comparison with HCV mono-infected without fibrosis; P < 0.05 by comparison with healthy controls.

Gender (m/f) (m%)16/9 (64)15/11 (58)15/9 (63)
Age, years, median (IQR)54 (48–59)51 (43–57)51(42–56)
Alcohol (>40 g/day), number of individuals (%)3 (12)3 (11.5)0 (0)
Time from HCV diagnosis to study entry (years), median (IQR)5 (3–8) 6 (4–11) NA
Route of HCV infection (IVD/sexual/nosocomial/unknown)11/5/6/314/2/7/3NA
HCV-RNA, IU/ml, median, (IQR)1.1 million (0.12–1.5 million)*2.3 million (1.2–4.4 million)NA
Genotype 1/2/3/410/2/9/115/1/9/0NA
ALT, U/l, median (IQR)104 (60–130)74 (57–129)22 (21–27)
Fibroscan kPa, median (IQR)12.6 (10.3–17.8)‡*5.8 (4.2–6.9)4.3 (3.7–4.9)
IL-28B CC/CT/TT5/9/72/12/10NA
Blood analysis and flow cytometry

Blood collected in ethylenediaminetetraacetic acid (EDTA) tubes was used for flow cytometry. Definition of lymphocyte subsets and gating strategy is shown in Fig. 1. In brief, 100 μl of EDTA blood was incubated in fluorescent dye–conjugated monoclonal antibodies, erythrocytes were lysed with 2 ml of Lysing Solution (Becton Dickinson, BD, Franklin Lakes, NJ, USA), and the samples were washed and resuspended in FACS Flow (BD). Monoclonal antibodies used to determine lymphocyte subsets were CD8 peridinin chlorophyll proteins–cyanine (PerCP-Cy5.5), CD3 fluorescein isothiocyanate (FITC), CD28 phycoerythrin (PE), CCR7-PE, CD31-FITC, CD57-FITC, CD45RA-FITC, CD95-FITC, CD127-FITC, CD27- PE-Cy7 and CD4 allophycocyanin-H7, and appropriate isotype controls all purchased from BD. Acquisition was performed using a FACS Canto, and data were processed using FACS Diva software (BD). For each sample, a minimum of 50,000 cells were acquired. Lymphocyte subsets are given as the proportion (%) of the cell population concerned (CD4+ T cells or CD8+ T cells) (Table 2) as well as absolute cell counts (Table 3).

Table 2. Proportions of lymphocyte subsets
 HCV-infected with fibrosis (n = 25)HCV-infected without fibrosis (n = 26)Comparison between HCV-infected with and without fibrosisHealthy controls (n = 24)Comparison between HCV-infected and healthy controls
  1. RTE, Recent Thymic Emigrants.

  2. Comparisons between groups were analysed by Kruskal Wallis test followed by Mann Whitney U test if significant.

  3. PF+ states the comparison between HCV-infected with fibrosis versus healthy controls. PF− states the comparison between HCV-infected without fibrosis and healthy controls.

  4. Data are given as percentages, median (IQR) unless otherwise stated.

CD4+ T cell subsets as % of CD4+ T cells
RTE (CD45RA+CD31+)14.3 (9.7–23.0)18.0 (10.8–23.2)P = 0.49624.2 (16.4–32.1)

PF+ = 0.002

PF− = 0.032

Naive (CD45RA+CD27+CCR7+)32.7 (26.2–47.8)39.1 (24.7–49.1)P = 0.33947.6 (40.1–59.5)

PF+ = 0.001

PF− = 0.023

CD31Naive (CD31CD45RA+CD27+CCR7+)18.2 (12.5–24.7)20.6 (14.1–28.4)P = 0.67626.4 (18.9–31.3)

PF+ = 0.020

PF− = 0.158

Central memory (CD45RACD27+CCR7+)27.8 (21.2–33.9)27.0 (23.2–34.9)24.1 (17.6–29.0)
Effector memory (CD45RACD27+CCR7)16.6 (8.5–22.4)14.1 (11.0–22.4)12.1 (8.6–17.2)
Late-differentiated (CD45RA+CD27CCR7)2.8 (0.5–8.2)1.9 (0.9–4.7)2.3 (0.7–4.4)
Senescent cells (CD28CD57+)1.3 (0.3–7.4)1.3 (0.8–2.8)2.8 (0.3–4.4)
Apoptotic cells (CD28CD95+)1.5 (0.4–10.3)1.4 (0.8–3.0)3.6 (0.4–4.5)
IL-7R α expressing cells (CD127+)88.4 (84.5–91.0)83.8 (79.9–86.8)P = 0.01682.2 (78.6–87.8)

PF+ = 0.005

PF− = 0.641

CD8+ T cell subsets as % of CD8+ T cells
RTE (CD45RA+CD31+)28.8 (16.1–40.5)37.5 (21.2–46.9)P = 0.05739.1 (31.6–55.0)

PF+ = 0.010

PF− = 0.379

Naive (CD45RA+CD27+CCR7+)22.4 (16.8–34.0)31.6 (20.9–42.1)29.4 (23.2–37.7)
CD31Naive (CD31CD45RA+CD27+CCR7+)1.9 (1.1–3.2)1.8 (1.0–3.5)1.1 (0.7–12.1.9)
Central memory (CD45RACD27+CCR7+)5.7 (3.0–9.0)4.8 (2.3–6.6)P = 0.6052.5 (1.6–3.8)

PF+ = 0.010

PF− = 0.032

Effector memory (CD45RACD27+CCR7)15.5 (8.5–19.7)13.1 (7.5–17.1)9 (7.0–14.3)
Late-differentiated (CD45RA+CD27CCR7)25.4 (11.0–41.1)26.1 (16.6–35.1)30.7 (17.3–42.2)
Senescent cells (CD28CD57+)25.1 (15.7–33.1)24.5 (17.8–35.8)29.6 (19.1–41.3)
Apoptotic cells (CD28CD95+)31.6 (22.5–48.8)36.8 (24.3–45.6)38.4 (30.1–51.1)
IL-7R α expressing cells (CD127+)71.0 (65.8–78.3)72.2 (61.8–77.1)68.1 (60.1–76.9)
Telomere length mean (95% CI) 0.88 (0.62–1.14) 1.15 (0.90–1.39) 1.0 (0.75–1.25)
IFN-γ, pg/ml 175.0 (74.8–311.5) 126.5 (92.0–167.5) P = 0.222 84.0 (44.8–159.3)

pF+ = 0.044

PF− = 0.164

Table 3. Absolute cell counts of T cell subsets
 HCV-infected with fibrosis (= 25)HCV-infected without fibrosis (= 26)Comparison between HCV-infected with and without fibrosisHealthy controls (= 24)Comparison between HCV-infected and healthy controls
  1. RTE, Recent Thymic Emigrants.

  2. Comparisons between groups were analysed by Kruskal Wallis test followed by Mann Whitney U test if significant.

  3. PF+ states the comparison between HCV-infected with fibrosis versus healthy controls. PF− states the comparison between HCV-infected without fibrosis and healthy controls

  4. Data are given as median (IQR).

CD4+ T cell subsets cells/μl
CD4+ T cells983 (828–1125)972 (756–1151)830 (667–956)
RTE (CD45RA+CD31+)141 (90–262)147 (92–198)192 (124–293)
Naive (CD45RA+CD27+CCR7+)301 (224–488)381 (201–455)396 (297–524)
CD31Naive (CD31CD45RA+CD27+CCR7+)155 (112–264)204 (150–285)225 (160–271)
Central memory (CD45RACD27+CCR7+)268 (181–390)247 (179–319)186 (128–239)
Effector memory (CD45RACD27+CCR7)163 (88–225)141 (76–198)99 (79–131)
Late-differentiated (CD45RA+CD27CCR7)20 (6–64)16 (7–40)18 (5–34)
Senescent cells (CD28CD57+)10 (3.4–58)11 (5–26)23 (3–32)
Apoptotic cells (CD28CD95+)9 (3–85)13 (5–29)25 (3–37)
IL-7R α expressing cells (CD127+)873 (680–1028)871 (511–933)= 0.274695 (546–805)

PF+ = 0.011

PF− = 0.294

CD8+ T cell subsets cells/μl
CD8+ T cells513 (405–877)596 (416–783)383 (308–597)
RTE (CD45RA+CD31+)139 (111–211)165 (121–296)172 (121–226)
Naive (CD45RA+CD27+CCR7+)140 (74–181)158 (116–221)125 (100–177)
CD31Naive (CD31CD45RA+CD27+CCR7+)9 (6–17)9 (7–25)6 (3–11)
Central memory (CD45RACD27+CCR7+)30 (12–49)21 (11–37)= 0.42012 (6–26)

PF+ = 0.008

PF− = 0.062

Effector memory (CD45RACD27+CCR7)77 (40–146)61 (36–87)48 (24–78)
Late-differentiated (CD45RA+CD27CCR7)111 (41–219)129 (78–236)127 (49–224)
Senescent cells (CD28CD57+)120 (80–170)130 (83–249)140 (57–225)
Apoptotic cells (CD28CD95+)158 (117–252)196 (106–309)164 (99–291)
IL-7R α expressing cells (CD127+)324 (292–508)378 (297–535)= 0.734247 (180–373)

PF+ = 0.020

PF− = 0.016

Figure 1.

Gating strategy. Representative plots illustrating gating strategies within the (A) CD4+ and the (H) CD8+ T cell compartment of (B, I) recent thymic emigrants (CD45RA+CD31+), (F, M) CD45RA+/- and CD27+/- cells, (G1, N1) naive cells (CD45RA+CD27+CCR7+), (G2, N2) central memory cells (CD45RACD27+CCR7+), (G3, N3) effector memory cells (CD45RACD27+CCR7), (G4, N4) late-differentiated cells (CD45RA+CD27CCR7), (C, J) senescent cells (CD28CD57+), (D, K) apoptotic cells (CD28CD95+) and (E, L) IL-7 receptor α expressing cells (CD127+) including relevant isotype control.

Production of IFN-γ

IFN-γ was measured in whole blood after stimulating with the T cell mitogen phytohaemagglutinin (PHA). In brief, 0.4 ml of whole blood was cultured in 1.6 ml RPMI 1640 and 40 μl PHA (1 μg/μl) and incubated at 37 °C for 24 h. The supernatant was harvested and stored at −80 °C until use. IFN-γ was measured by a bead-based multiplex sandwich immunoassay as described [35]. In short, 50 μl of undiluted sample, calibrators or controls, and 50 μl of a suspension of capture-antibody-conjugated beads were mixed in plate wells (filter plate, MultiScreen MABVN 1.2 μm 96-well, Millipore, France). After 1½ h of incubation, beads were washed twice and subsequently incubated for 1½ h with a mixture (50 μl) of corresponding biotinylated detection antibodies, each diluted 1 : 1000. Fifty microliter of streptavidin–phycoerythrin was added to the wells and incubated for 30 min. Finally, the beads were washed twice and resuspended in 125 μl of buffer and analysed on the Luminex 100™ platform (Luminex Corp., Austin, TX, USA) using BioPlex 5.0 from BioRad Laboratories (Hercules, CA, USA). All samples were measured in duplicates.

Mean telomere length

Peripheral blood mononuclear cells (PBMC) were isolated by density gradient centrifugation and frozen in liquid nitrogen. PBMC were carefully thawed, and DNA was purified using Wizard® SV Genomic DNA Purification System (Promega, Madison, WI, USA) according to manufacturers’ instruction. DNA was used for measurement of mean telomere length by quantitative PCR as described previously [36]. In brief, DNA was quantified using a Nanodrop ND 1000 (Saveen biotech ApS, Aarhus, Denmark). Primers targeting the telomere or the single copy gene 36B4 were used for PCR. Relative telomere-to-36B4 ratios were calculated using the ΔΔCt method (User Bulletin No. 2; ABI PRISM 7700 Sequence Detection System, Applied Biosystems, Carlsbad, CA, USA). Data are given as a percentage relative to the group of healthy controls, that is, 1.

Statistical analyses

Results are given as median and interquartile range (IQR). Differences between groups were analysed by the Kruskal–Wallis test followed by Mann–Whitney U-test if significant. Correlation was calculated by Spearman's test. Two-tailed p values of 0.05 or less were considered significant. All statistical analyses were performed using the Statistical Package for Social Sciences (SPSS version 11.5.0; SPSS Inc., Chicago, IL, USA).

Results

Study population

Hepatitis C virus-infected patients with and without fibrosis and healthy controls were similar regarding gender and age (Table 1). As expected, both groups of HCV-infected patients had higher levels of alanine aminotransferase compared with healthy controls, and the HCV-infected patients with fibrosis had lower levels of HCV-RNA compared with HCV-infected patients without fibrosis (Table 1).

Production and destruction of CD4+ T cells

The proportion of CD4+ RTE was lower in HCV-infected patients with fibrosis and without fibrosis compared with healthy controls (P = 0.002 and P = 0.032, respectively, Table 2). Furthermore, the proportion of naïve CD4+ T cells was lower in both groups of HCV-infected patients compared with healthy controls (P = 0.001 and P = 0.023, respectively, Table 2). No differences in CD4+ RTE and naive CD4+ T cells were found comparing HCV-infected patients with and without fibrosis (Table 2). The proportions of CD31CD45RA+CD27+CCR7+ CD4+ T cells in HCV-infected patients with fibrosis were lower compared with healthy controls (P = 0.020), while no difference was found comparing HCV-infected patients with and without fibrosis (Table 2). Differences in proportion of memory and late-differentiated CD4+ T cells as well as senescent and apoptotic CD4+ T cells were not found between the groups (Table 2). The proportion of CD4+ T cells expressing the IL-7R α (CD127) was higher in HCV-infected patients with fibrosis compared to HCV-infected patients without fibrosis (P = 0.016) and to healthy controls (= 0.005). In contrast, HCV-infected patients without fibrosis did not differ from healthy controls (Table 2). Significant associations between HCV-RNA and the CD4+ T cell subset were not found (data not shown).

Production and destruction of CD8+ T cells

The proportion of CD8+ RTE was lower in HCV-infected patients with fibrosis compared with healthy controls (P = 0.010) and HCV-infected patients without fibrosis (although not reaching statistical significance, = 0.057). In contrast, HCV-infected patients without fibrosis did not differ from healthy controls. The proportions of naïve CD8+ cells and CD31CD45RA+CD27+CCR7+ CD8+ T cells did not differ between patients with HCV infection and healthy controls. However, HCV-infected patients with fibrosis and without fibrosis had a higher proportion of central memory CD8+ T cells compared with healthy controls (= 0.010 and = 0.032, respectively, Table 2). Proportion of senescent cells and apoptotic cells did not differ when comparing the groups (Table 2). Furthermore, expression of CD127 in the CD8+ T cell compartment did not differ between the groups. Finally, significant associations between HCV-RNA and the CD8+ T cell subset were not found (data not shown).

Absolute counts

To further describe the composition of the T cell pool, we assessed the absolute T cell counts (Table 3). In the CD4+ T cell compartment, no differences in absolute counts of CD4+ RTE were found between groups. However, higher absolute counts of CD4+CD127+ T cells were found in HCV-infected patients with fibrosis compared with healthy controls (= 0.011, Table 3). In contrast, no difference between HCV-infected patients with and without fibrosis regarding absolute CD4+CD127+ T cell counts was found.

Similar to the CD4+ T cell compartment, no differences in absolute CD8+ RTE cell counts between the groups were found. The absolute counts of CD8+ T cells expressing CD127 were higher for both groups of HCV-infected patients compared with healthy controls (= 0.020 and = 0.016, respectively, Table 3). Furthermore, the absolute CD8+ central memory T cell counts were higher in HCV-infected patients with and without fibrosis compared with healthy controls (= 0.008 and = 0.062, respectively), although latter did not reach statistical significance.

Telomere length

To assess the proliferative history of the lymphocyte pool, mean telomere lengths were examined. No difference in mean telomere length was found comparing HCV-infected individuals and healthy controls (Table 2).

IFN-γ

Both groups of HCV-infected patients presented with higher PHA-induced IFN-γ production compared with healthy controls, but only HCV-infected patients with fibrosis differed significantly (= 0.044). No difference in PHA-induced IFN-γ production was found between HCV-infected patients with and without fibrosis (= 0.222). Effector memory T cells are known to have an increased capacity for IFN-γ production, but no association between the proportion of effector memory T cells and IFN-γ was found (data not shown). However, inverse correlations between levels of IFN-γ and proportions of naïve CD4+ T cells (r −0.413, = 0.011) and naïve CD8+ T cells (−0.505, = 0.001) were found in HCV-infected patients. In contrast, this was not the case for healthy controls (= 0.357 and = 0.091, respectively).

Discussion

In this study, T cell homeostasis in patients with chronic HCV infection with and without fibrosis was described. Evidence of impaired thymic function was found as proportion of RTE in the CD4+ and the CD8+ T cell compartment as well as naïve CD4+ T cells was significantly lower in HCV-infected patients compared with healthy controls. Especially, HCV-infected patients with fibrosis had impaired thymic output compared with healthy controls. Interestingly, HCV-infected patients with fibrosis presented with high proportion of CD4+ T cells expressing CD127 suggesting a compensatory mechanism for impaired thymic output.

Thymic output assessed by CD4+ RTE in patients with chronic HCV infection has previously been shown to be reduced or unaltered compared with healthy controls [11, 28, 29, 31], but it has not previously been examined if thymic output is associated with fibrosis status in HCV infection. Furthermore, the subset CD8+ RTE has not previously been examined in HCV-infected patients. We found thymic output to be impaired in both CD4+ and CD8+ T cells in individuals with chronic HCV infection, especially in those with fibrosis compared with healthy controls. The impaired thymic output may be explained by thymus exhaustion caused by an increased demand of T cells due to chronic immune activation. Chronic immune activation is a feature of HCV infection as well as HIV infection [3, 33, 37, 38]. In the initial phase of established HIV infection, loss of CD4+ T cells due to immune activation leads to an increased production of RTE until the thymus is unable to meet this demand and the CD4+ T cell count decreases [37-40]. Our data suggest that a similar process may be true for chronic HCV infection, although this process probably is less pronounced in chronic HCV infection compared with HIV infection, as HCV does not infect CD4+ T cells. This is supported by elevated proportion as well as absolute count of central memory CD8+ T cells in HCV-infected patients compared with healthy controls, which is in accordance with increased immune activation and the HCV antigen-specific CD8+ central memory cells described previously [41]. In theory, lower proportion of CD8+ RTE in HCV-infected patients with fibrosis may be a result of a relative increase in central memory cells rather than actual lower thymic output. However, the absolute number of CD8+ central memory is low, and the absolute number of CD8+ RTE is lower, although not significant, in HCV-infected patients compared with healthy controls despite higher total CD8+ T cell count. Thus, it seems that the proportions of CD4+ RTE as well as CD8+ RTE in HCV-infected with fibrosis reflect an impaired thymic output, but due to the cross-sectional design, the present study does not allow causal conclusions.

To ensure that the lower proportions of RTE found in HCV-infected patients were not affected by a lower proportion of CD31 late-differentiated T cells, we compared the proportions of CD31 late-differentiated T cells as well as the proportions of CCR7+ RTE between the groups. The proportion of CD31 CD4+ late-differentiated T cells was low in all groups (4.4%, 6.1% and 5.7% representing HCV-infected patients with fibrosis, without fibrosis and healthy controls, respectively) with no differences between the groups. Similarly, no differences in CD31 CD8+ late-differentiated T cells were found between the groups; however, the proportions of CD31 CD8+ late-differentiated T cells were higher (42.1, 40.2, and 43.4%, respectively). Thus, the differences in RTE presented in this study cannot be explained by lower expression of CD31 on late-differentiated T cells. However, the phenotype of RTE (CD31+CD45RA+) seems to be more specific in the CD4+ T cell compartment compared with the CD8+ T cell compartment which is in accordance with a previous study [32].

IL-7 is a haematopoietic growth factor and essential for lymphopoiesis. Thus, limited production of IL-7 or a dysfunctional response to IL-7 may also explain the impaired thymic output in HCV-infected individuals. Interestingly, high proportion of CD4+ T cells expressing CD127 was found in patients with fibrosis. CD127 is the α-subunit of the receptor for IL-7, and high proportion of CD4+ T cells expressing CD127 might illustrate a compensatory mechanism for the impaired thymic output in HCV-infected patients especially in those patients with fibrosis. In HIV-infected patients, the expression of CD127 on T cells has been found to be inversely associated with the level of IL-7 [37, 42]. This has not been examined in HCV-infected patients, but the liver has shown to be a major source of IL-7 [43]. Thus, impaired thymic output and high proportion of CD4+ T cells expressing CD127 in HCV-infected patients with fibrosis might illustrate a low level of IL-7 caused by a reduced production of IL-7 in the fibrotic liver. The level of IL-7 in HCV infection has been compared with healthy controls with contradictory results, and it has to our knowledge not previously been evaluated whether IL-7 and the stage of fibrosis are associated [31, 44, 45]. Thus, further studies investigating the interplay between thymic output, IL-7 and CD127 and progression of fibrosis are warranted.

Differences in clinical characteristic between the HCV-infected patients with and without fibrosis were limited to the level of HCV-RNA, as the level of HCV-RNA was lower in HCV-infected patients with fibrosis. Previous studies demonstrate no association between current HCV-RNA and liver fibrosis [46-48]. Furthermore, in our cohort, HCV-RNA was not significantly associated with any T cell subsets. An overall tendency for HCV-infected patients with fibrosis to have lower proportion of RTE and naïve cells within both the CD4+ and the CD8+ T cell compartment and shorter mean telomere length compared with HCV-infected without fibrosis was found, suggesting a lower thymic function in these patients. None of these differences were, however, significant. Because thymopoiesis is further impaired during IFN-α treatment, this may explain the association between fibrosis and lower sustained viral response (SVR) after treatment with IFN-α and ribavirin [49]. Furthermore, co-infection with cytomegalovirus (CMV) has been associated with lower SVR [50], and higher prevalence of CMV antibodies was found in HCV-infected patients compared with healthy controls [51]. Whether an underlying CMV infection could influence the proportions of T cell subset including RTE in HCV-infected patients is undetermined, but would be of interest to investigate in futur e studies.

Production of IFN-γ in lymphocytes was higher in HCV-infected patients. IFN-γ is mainly secreted by effector memory T cells [41]. However, no association between IFN-γ and effector memory cells was found (data not shown). Thus, the increased production of IFN-γ might illustrate immune activation and a profile of more active effector memory T cells in patients with HCV infection compared with healthy controls. In contrast, a negative association between production of IFN-γ and proportion of naive cells was found, but whether IFN-γ impairs thymic output is unclear. Furthermore, IFN-γ induces production of the chemokine IP-10. IP-10 has shown to induce trafficking of T lymphocytes to the liver, which is associated with the stage of fibrosis as well as SVR [6, 52]. Thus, the increased production of IFN-γ and the perturbed proportion of T cell subsets in HCV-infected individuals with fibrosis might also be a result of increased trafficking of T cells to the fibrotic liver.

In conclusion, patients with chronic HCV infection seem to have impaired thymic output especially in those with fibrosis compared with healthy controls. Furthermore, expression of CD127 on CD4+ T cells was significantly associated with fibrosis status and further prospective studies are warranted to investigate the interplay between impaired thymic output, expression of CD127 and progression of chronic HCV infection.

Acknowledgment

We gratefully acknowledge the patients and healthy individuals who made this study possible.

Disclosures

The authors have no conflict of interests to disclose.

Financial disclosure

The study was funded by The Danish Council for Independent Research, Lundbeck Foundation, Novo Nordisk Foundation and Augustinus Foundation.

Ancillary