Impact of IFN-Free and IFN-Based Treatment on Blood Myeloid Dendritic Cell, Monocyte, Slan-DC, and Activated T Lymphocyte Dynamics during HCV Infection

Chronic hepatitis C virus infection leads to the activation of innate immunity, a key component in HCV fibrosis. In the past, the use of IFN-based treatment regimens did not permit an adequate evaluation of the impact of HCV clearance on immune cells, because of their antiviral and immunomodulatory properties. The recent development of direct-acting antiviral (DAA) therapy, which is associated with high rates of sustained virological response, enables a more accurate analysis of the immunological modifications following HCV eradication. We studied the dynamics of blood myeloid dendritic cells, monocytes, slan-DCs, and T lymphocytes during IFN-free and IFN-based regimens in hepatitis C virus infection.


Introduction
Worldwide, an estimated 71 million people are chronically infected with hepatitis C virus (HCV) [1].
Chronic HCV infection is characterized by an aberrant inflammatory response that causes HCV-mediated liver damage, leading to progressive fibrosis, potentially resulting in cirrhosis, liver failure, and hepatocarcinoma (HCC) [2,3].
HCV can counteract both innate and adaptive immune responses by modulating the function of several types of cells of the immune system, including monocytes (Mo), macrophages, dendritic cells (DCs), and T cells [4][5][6][7][8]. As a result of this, the expression profile of circulating pro-and antiinflammatory cytokines and chemokines is altered, leading to chronic infection and persistent inflammation [4][5][6][7][8].
Until August 2011, the commercially available treatment options for HCV infection were limited to interferon-alpha-(IFN-α-) based therapies and ribavirin (RBV) for all geno-types [9][10][11]. IFN-α has antiviral activity and also enhances HCV-specific T cell responses; ribavirin, a nucleoside analogue, has a small direct activity against HCV but reduces hepatic inflammation [12]. It has been speculated that these two drugs act by modulating the immune system for HCV elimination [12].
In recent years, the introduction of several oral IFN-free direct-acting antiviral agents (DAAs) in clinical practice has proved to be a milestone in the management of HCV infection, increasing sustained virological response (SVR) rates up to 95-100% [13]. DAAs are molecules that target specific nonstructural proteins of the virus, with subsequent disruption of viral replication and spread to other cells. Since IFN-free DAA regimens specifically target various steps in the HCV life cycle, they can provide the opportunity to elucidate the relationship between HCV and the innate immune response, without the confounding effect of the IFN-αinduced immune modulation.
2.2. HCV-RNA and HCV Genotype Testing. Plasma HCV-RNA levels were determined by RealTime PCR Roche Cobas TaqMan. HCV genotypes and subtypes 1a and 1b were determined by Abbott RealTime HCV Genotype II.
2.3. Liver Fibrosis Assessment. Liver stiffness (measured in kPa) was determined by transient elastography with the use of a FibroScan machine (Echosens). Advanced fibrosis (F4) was defined as a measure of liver stiffness (LSM) greater than or equal to 14.5 kPa; severe fibrosis (F3) was defined as a LSM value greater than or equal to 10; mild or no fibrosis (F0-F2) was defined as a LSM value less than 10 kPa [15].

Sample
Handling. Venous blood samples were collected from each patient into EDTA-containing tubes (Becton ±Dickinson Systems, San Jose, CA). Patients undergoing Peg-IFN-α/RBV treatment plus telaprevir (TVR) or boceprevir (BOC) had their peripheral blood samples collected at baseline (T0); after 1 (T1), 3 (T2), 6 (T3), and 12 (T end) months after the start of therapy; and within 4-6 months after the end (T post) or interruption of the therapy. Patients on IFN-free DAA regimens had peripheral blood samples collected at baseline (T0), 1 month after the start of T1, and 3 months after the end of the therapy (T post).

Enumeration of DCs and Monocytes.
To enumerate DCs and monocytes, the same staining protocols previously described by our research group have been used [7,16]. The analysis of the results was performed using FlowJo software (FlowJo 9.2, Tree Star, USA). Gating strategies are shown in additional file 1. Moreover, we stained the pDCs of 5 healthy donors in parallel with both the antibody panel used in our study (in which we identified pDC with CD123 and HLA-DR expression) and a new panel, which included BDCA-2, CD123, and HLA-DR, and we found that the two methods were equivalent.
2.6. CD4 and CD8 Activation Markers. To determine activated CD4 and CD8, the same staining protocols previously described by our research group have been used [16].
2.7. Statistical Analysis. All statistical analyses were performed using GraphPad Prism Software version 6.0 (Software MacKiev). Values are given as median and ranges (minimum and maximum values). The nonparametric Mann-Whitney test and the nonparametric Kruskal-Wallis ANOVA with Dunn's posttest were used to compare the differences in values. The nonparametric Wilcoxon matched paired test was applied to perform longitudinal analyses. All differences were considered statistically significant with p values less than 0.05.

Study Population
Characteristics. The study population included 73 patients; 25 were on INF-based treatment, including Peg-IFN-α/RBV plus first-generation protease inhibitors TVR and BOC; and 54 (of which 6 had previously failed with INF-based therapy) were treated with IFN-free DAA regimens with or without RBV. The demographic and clinical features of the patients are listed in Table 1.
In the IFN-based treatment group, 13 patients (52%) were responders and 12 (48%) were non responders, of which 33% NR for virological failure and 67% NR for side effects.
All the 54 patients treated with IFN-free DAA regimens were responders since they reached the SVR12; clinical features of patients are listed in Table 1. Among these patients, six had previously been treated with Peg-IFN-α+RBV and failed to achieve SVR and were retreated with an IFN-free DAA regimen after 1 year.  (Figure 1). No significant differences were found in DCs and Mo counts in the different HCV genotypes (data not shown).  3.2.2. CD4 + and CD8 + T Cell Activation. Regarding activated CD4 + and CD8 + T cells, we stratified patients according to the following: (i) the stage of fibrosis in which no differences were observed in activated T cells, for both CD4 + and CD8 + T cells (data not shown), and (ii) genotypes in which we observed a trend towards an increase in the percentage of CD4 + HLA-DR + CD38 + in patients with other genotypes compared to genotype 1b (2.39% (0.75-4.08) vs. 1.49% (0.17-2.39), respectively, p = 0:37). The median value (range) for genotype 1a was 1.59% (0.54-5.08). Regarding CD8 + T cells, we observed a significantly lower percentage of CD8 + HLA-DR + CD38 + T cells (1.03% (0-3.2)), compared to genotype 1a (2.85% (0.13-8.01)) and other genotypes (2.3% (1.2-8.9)) (p = 0:0059) ( Figure 2). Considering all results, patients with genotype 1b showed the lowest levels of CD4 + and CD8 + T cell activation. Changes in innate immune system cell subsets from the baseline to the end of treatment were longitudinally analyzed in the three different groups, and the results are reported below.

Dynamic Changes in
(2) Responders. In R, classical Mo progressively decreased during therapy at T1 month, T2 month, and T end and slightly increased at T post compared to T0 (baseline) ( Significant differences in mDC values were observed between T0 and T2 month (p = 0:0005) and between T2 month and T3 month (p = 0:00039).
The pDCs significantly decreased up to six months of therapy, with a subsequent progressive increase to values similar to those before the initiation of therapy.
There was a significant decrease in pDCs between T0 and T1 month (7.

CD4 and CD8 Activation Markers
(1) IFN-Based Therapy. At the baseline, the percentage of CD4 and CD8 T cells did not differ between the three groups: R, NR for virological failure, and NR for side effects. For CD4 T cells, the median value for R was 32.3% (8-32.8), for NR for virological failure 30.1% (8.8-31), and for NR for side effects 33.4% (9-35.9) (data not shown). For CD8 T cells, the median  T0  T1  T2  T3  T end T post  T0  T1  T2  T3  T end T post   T0  T1  T2  T3  T end T post  T0  T1  T2  T3  T end T post   T0  T1  T2  T3  T end T post  T0  T1  T2  T3 T end T   Journal of Immunology Research value for R was 14% (7-32), for NR for virological failure 7% (4-22), for NR for side effects 8% (6-18.1) (data not shown).
The percentage of activated CD4 T cells, at baseline, did not differ between the three groups (R, NR for virological failure, and NR for side effects). The median value for R was 1.71% (0.36-8.28), for NR for virological failure 2.22% (0.17-3.66), and for NR for side effects 1.95% (1.4-4.42) (data not shown).

Discussion
The last few years have seen extraordinary advances in the management of patients with chronic HCV infection.
The use of a triple therapy which combined firstgeneration DAAs with pegylated interferon (Peg-IFN) and ribavirin (RBV) significantly improved the chances of achieving sustained virological response (SVR) rates to over 60%. However, treatment with telaprevir and boceprevir was intended for genotype 1 HCV-infected patients and antiviral potency was counteracted by considerable side effects, in particular an increased number of bacterial infections [17,18].
In our study, patients undergoing IFN-based therapy showed a successful rate of 67% in accordance with what is known from the literature [19,20].
In almost half of nonresponder patients, the interruption of therapy was mainly due to neutropenia and infectious diseases (sepsis and pneumonia), in line with previous reports of the association with an increase in treatment-related adverse events compared to the former standard therapy without protease inhibitors. The mechanism of this increased susceptibility to infections seems to be related to the in vitro inhibition of neutrophil elastase activity [21,22].
IFN-free treatments consisting of combinations of second-generation DAAs, with or without ribavirin, showed impressive SVR rates and a safer profile. In our cohort, the success rate was 100% and no patients interrupted the therapy for virological failure or for adverse effects.
In the present study, we evaluated the differences in DCs, Mo, and activated T cells in all HCV-infected patients according to fibrosis stage and genotype. The innate immune system plays a pivotal role in the host-virus interactions during the entire natural course of the disease, and IFNs are the central cytokines responsible for the induction of an antiviral state in cells and for the activation and regulation of the cellular components of innate immunity [23]. IFN-α has potent antiviral activity, and it acts by inducing IFN-stimulated genes (ISGs), which establish a non-virus-specific antiviral state within the cell [24]. Exogenously supplied recombinant IFN-α binds to and activates cellular receptors, leading to the same response as with the endogenous one.
Dendritic cells (DCs) are professional antigen-presenting cells (APCs), playing a key role in the innate immune system in orchestrating the quality and potency of downstream adaptive immune response, through the uptaking and processing of viral antigens, as well as by releasing cytokines to efficiently prime both CD4 + helper T cells and CD8 + cytotoxic T lymphocytes (CTLs) [25].
Two major subsets of DCs can be readily purified from human peripheral blood: plasmacytoid (p)DCs and conventional or myeloid (m)DCs [25]. pDCs and mDCs differ markedly in their ability to capture, process, and present antigens; express costimulatory molecules; and produce cytokines [26].
Slan-DCs are a third subset of DCs that were identified in peripheral blood using the monoclonal antibody M-DC8, which binds to 6-sulfo LacNac (slan), a carbohydrate moiety of the P selectin glycoprotein ligand 1 (PSGL-1) [27]. However, on a two-dimensional flow cytometry dot plot of CD14 and CD16 expression in peripheral blood mononuclear cells (PBMCs), slan-DCs in part overlap with CD14 dim CD16 + monocytes [28], suggesting that they might actually represent a subset of nonclassical monocytes [29].
Regarding their function, blood slan-DCs have been described as potent proinflammatory cells, based on their  Journal of Immunology Research ability to produce large amounts of tumor necrosis factoralpha (TNF-α) and IL-12p70 upon stimulation with tolllike receptor (TLR) ligands [28]. In our study, all HCV-infected patients were analyzed in relation to the stage of fibrosis. Interestingly, an increase in nonclassical Mo, slan-DCs, and pDCs in peripheral blood was found in patients with low fibrosis compared with those with high fibrosis. This could reflect a higher level of these cells in peripheral blood during the early stages of fibrosis, followed by a reduction due to their migration to the liver when fibrosis is progressing. Indeed, peripheral blood monocytes constantly enter the liver to replenish hepatic macrophages, and the number of infiltrating monocytes increases during liver inflammation and modulates fibrogenesis [30]. In fact, during fibrogenesis, there is an infiltration of leukocytes and monocytes/macrophages into the liver; this infiltration is the key for the activation of hepatic stellate cells and subsequent fibrosis; moreover, reducing liver-infiltrating macrophages, it is possible to attenuate fibrosis in some models [30]. During HCV infection, there is a continuous recruitment of monocytes and macrophages into the liver where under local signals became able to secrete a variety of proinflammatory, profibrotic, and anti-inflammatory cytokines and growth factors [29].
Monocytes are attracted to the site of infection and are exposed to viral RNA and protein, which can lead to their activation [31,32]. Nonclassical monocytes (CD14 ++ CD16 + ) preferentially accumulate in the chronically inflamed human liver as a consequence of enhanced recruitment from blood and local differentiation from classical CD14 ++ CD16monocytes [29]. Nonclassical monocytes are major modulators of fibrogenesis in the liver producing profibrogenic cytokines and chemokines. Mascia et al. recently [8] described how monocyte-derived cytokines such as CXCL-10 and sCD14 are increased in all stages of fibrosis, underlying the presence of immune activation from F0 to F4 [8,33]. Slan-DCs seem to contribute to the pathogenesis of chronic inflammatory diseases such as Crohn's disease, rheumatoid arthritis, psoriasis, and HIV since they infiltrate the inflamed ileal mucosa, skin, and synovial tissue [16,28,34]. They are a major source of tumor necrosis factor-α, a pleiotropic cytokine produced even by macrophages/monocytes in response to bacterial LPS, and are significantly expanded in patients with bacterial sepsis [34]. TNF-α has been implicated in the pathogenesis of chronic liver inflammation-activating resident HSCs into fibrogenic myofibroblasts. However, there are currently no data on the possible migration of slan-DCs into the inflamed liver during the progress of fibrosis. The subset of plasmacytoid DCs (pDCs) is considered to be the front line in antiviral immunity, owing to the rapid production of high amounts of type I interferon in response to viruses [35]. Patients with chronic HCV infection have a reduced ability to produce INF after in vitro stimulation of pDCs due to a decrease in their absolute number, an impairment in their function, and an increase in their homing to the liver [32].
Various studies have shown that CD4 + helper T cell-and CD8 + cytotoxic T cell-mediated immune responses determine the outcome of HCV infection. Thus, spontaneous viral clearance of HCV infection is characterized by vigorous and sustained specific CD4 + and CD8 + T cell responses during the acute phase of infection, while in contrast chronic infection is correlated with late, transient, weak, or narrowly focused CD4 + and CD8 + T cell responses [36,37].
We focused our analysis on activated CD4 + and CD8 + cells, and interestingly no differences in activated CD4 + and CD8 + were found in HCV-infected patients stratified for fibrosis stage. Regarding different genotypes, a lower activation of CD8 + cells was observed in patients with genotype 1b compared to other genotypes. Genotype 1b has historically been associated with a lower likelihood of selection of resistant variants and a better response to IFN-based therapy, compared to other genotypes [38]. In this setting, the lower impact on the activation of peripheral CD8 + cells can have an impact on the response to treatment.
To investigate the effect of IFN-based regimens and DAA-containing regimens with or without IFN on dynamic changes in circulating levels of monocytes (Mo) and DCs and activated CD4 and CD8 cells, we performed a longitudinal study.
During the course of INF-based regimens, we observed a decrease in pDCs after 1 month of therapy, which persisted up to the end of therapy, with subsequent increase after the interruption of treatment, although without a restoration to pretreatment values. mDCs tended to remain stable over the treatment course, although a decrease in number was observed at month 3, followed by a recovery at month 6.
No differences in both pDCs and mDCs were found in patients who failed due to virological failure or side effects.
Regarding monocytes, a decrease in classical monocytes, nonclassical monocytes, and slan-DCs was observed after 1 month of therapy that persisted after 6 months. No differences were observed in non responders.
These variations in the levels of circulating innate immune cells could be a consequence of the IFN-induced depression of bone marrow activity, as suggested by the well-known reduction in granulocytes during treatment [39], and could explain the increased risk of bacterial infections observed in our study population.
On the other hand, DAA therapy lacks the immunomodulatory effects of IFN, possibly explaining the absence of differences in DCs and Mo counts during interferon-free treatments.
To address the dynamics of immune activation in HCV infection, we analyzed CD38 and HLA-DR expression on the CD4 + and CD8 + T cells during the two different treatments.
Both responders and non responders for side effects had increased levels of activation of CD4 + cells over the course of IFN-based therapy, which remained persistently elevated even after the end of treatment in non responders for side effects, while no alterations were observed in non responders for virological failure. On the other hand, the percentage of activated CD8+ cells was significantly increased only in responders during the course of IFN-based regimens.
Our results are in contrast with previously published data on the decrease in immune activation in the T cell compartment after IFN-based therapy [40]. However, Radkowski et al. reported increased levels of immune activation 9 Journal of Immunology Research associated with persistence of HCV-RNA in PBMCs after successful treatment of chronic HCV infection, which could be linked to an increased risk of developing immunemediated extrahepatic complications of HCV infection in some patients [41].
On the contrary, during the course of DAA-based regimens, we observed a decrease in activated CD8 + T cells after the end of therapy. This highlights the importance of immune activation in the pathogenesis of chronic HCV infection, considering the reduced immune activation following complete viral clearance as a consequence of successful treatment with DAAs.
Finally, the observation of a higher number of classical and nonclassical monocytes, together with increased levels of pDCs and slan-DCs in responders, could suggest a possible role of the evaluation of these cell subsets in predicting treatment response during the course of IFN-based regimens. In fact, those cell subsets are involved in the immune response to HCV and have been found in peripheral blood and in the liver of patients with chronic HCV, suggesting a primary role in the immune response to the virus [31,34,35]. Thus, higher levels of these subpopulations of cells of the innate immune system might enhance the response against the virus and facilitate its eradication.
There are several limitations in our study. Firstly, we did not evaluate the functional status of circulating monocytes and DCs, since the analysis was limited to numerical changes in peripheral blood samples. Moreover, HCVinfected patients were not treated with the same DAA regimen, and this may have had an impact on the results, due to the possibility of differential effects of the various DAAs on immune function.
Taken together, these results could suggest that increased circulating levels of specific cell subsets within the innate immune compartment could promote antiviral response during the course of IFN-based regimens and could help in predicting the outcome of treatment in some patients.

Data Availability
All data used to support the findings of this study are included within the article and in the additional materials.

Conflicts of Interest
The authors declare that there is no conflict of interest regarding the publication of this article.