Accepted Manuscript Dendritic Cell Maturation in HCV Infection: Altered Regulation of MHC Class I Antigen Processing-Presenting Machinery Patrizia Leone, Mariangela Di Tacchio, Simona Berardi, Teresa Santantonio, Massimo Fasano, Soldano Ferrone, Angelo Vacca, Franco Dammacco, Vito Racanelli PII: DOI: Reference:

S0168-8278(14)00256-6 http://dx.doi.org/10.1016/j.jhep.2014.04.007 JHEPAT 5109

To appear in:

Journal of Hepatology

Received Date: Revised Date: Accepted Date:

30 August 2013 4 March 2014 6 April 2014

Please cite this article as: Leone, P., Di Tacchio, M., Berardi, S., Santantonio, T., Fasano, M., Ferrone, S., Vacca, A., Dammacco, F., Racanelli, V., Dendritic Cell Maturation in HCV Infection: Altered Regulation of MHC Class I Antigen Processing-Presenting Machinery, Journal of Hepatology (2014), doi: http://dx.doi.org/10.1016/j.jhep. 2014.04.007

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1

DENDRITIC CELL MATURATION IN HCV INFECTION: ALTERED REGULATION OF MHC CLASS I ANTIGEN PROCESSING-PRESENTING MACHINERY Patrizia Leone, Mariangela Di Tacchio, Simona Berardi, *Teresa Santantonio, *Massimo Fasano, §Soldano Ferrone, Angelo Vacca, Franco Dammacco, and Vito Racanelli

Department of Internal Medicine and Clinical Oncology, University of Bari Medical School, Bari, Italy; §

*

Department of Infectious Diseases, University of Foggia, Foggia, Italy;

Department of Surgery, Massachusetts General Hospital, Harvard Medical School, Boston,

MA, USA

Word count: 5476; number of figures: 4; number of tables: 2 Abbreviations mDC, myeloid dendritic cells; pDC, plasmacytoid dendritic cells; HCV, hepatitis C virus; TAP, transporter-associated proteins; PRR, pattern recognition receptors; TLR, Toll-like receptors; polyI:C, polyinosinic:polycytidylic acid; IL, interleukin; IFN, interferon; APM, antigen processing-presenting machinery; β2m, β2-microglobulin; PB, peripheral blood; PBMC, peripheral blood mononuclear cells; MESF, molecular equivalents of soluble fluorochrome; PI, persistently infected; SR, spontaneous resolvers; HD, healthy donors Conflict of interest The authors have no conflicting financial interests Financial support Italian Association for Cancer Research (AIRC), Milan, Italy

Corresponding author Vito Racanelli, MD, Department of Internal Medicine and Clinical Oncology, University of Bari Medical School, Policlinico, 11 Piazza G. Cesare, 70124 Bari, Italy; phone: +39 080 5478050; fax: +39 080 5478045; e-mail: [email protected]

2

ABSTRACT Background & aims: Modulation of dendritic cell (DC) function has been theorized as one of the mechanisms used by hepatitis C virus (HCV) to evade the host immune response and cause persistent infection. Methods: We used a range of cell and molecular biology techniques to study DC subsets from uninfected and HCV-infected individuals. Results: We found that patients with persistent HCV infection have lower numbers of circulating myeloid DC and plasmacytoid DC than do healthy controls or patients who spontaneously recovered from HCV infection. Nonetheless, DC from patients with persistent HCV infection display normal phagocytic activity, typical expression of the class I and II HLA and co-stimulatory molecules, and conventional cytokine production when stimulated to mature in vitro. In contrast, they do not display the strong switch from immunoproteasome to standard proteasome subunit expression and the upregulation of the transporter-associated proteins following stimulation, which were instead observed in DC from uninfected individuals. This different modulation of components of the HLA class I antigen processing-presenting machinery results in a differential ability to present a CD8+ T cell epitope whose generation is dependent on the LMP7 immunoproteasome subunit. Conclusions: Overall, these findings establish that under conditions of persistent HCV antigenemia, HLA class I antigen processing and presentation are distinctively regulated during DC maturation.

3

INTRODUCTION Dendritic cells (DC) play a crucial role in immune responses to infectious pathogens, including viruses [1]. In humans, DC circulating in the blood characteristically express high levels of HLA class II molecules and are proficient in antigen uptake and processing. However, they express low levels of HLA class I and co-stimulatory molecules (e.g. CD80, CD86) and lack common lineage markers such as CD3, CD14, CD16, CD19, CD20 and CD56. These lineage-negative (Lin-) cells are subdivided into CD11c+ myeloid DC (mDC) and CD11c-CD123+ plasmacytoid DC (pDC) [2]. Blood mDC and pDC are immature cells that are derived from bone marrow precursors and are migrating to their target sites. mDC home preferentially to peripheral tissues, the site of entry of most viruses. In contrast, pDC localize preferentially in the T cell areas of lymphoid organs such as thymus, bone marrow, spleen, tonsils and lymph nodes [3]. At these sites, mDC and pDC interact with pathogens and mature. The interaction of mDC and pDC with pathogens occurs through pattern recognition receptors (PRR) [1], of which the Toll-like receptors (TLR) are the most studied. TLR form a family of molecules, expressed on the cell surface and in endosomes, that recognize typical pathogen molecules.

For

instance,

mDC

express

TLR-3

that

recognizes

and

binds

polyinosinic:polycytidylic acid (polyI:C). pDC express TLR-7 and TLR-9 that bind singlestranded viral RNA and hypomethylated CpG motifs in bacterial DNA, respectively. The TLR-mediated maturation process comprises a series of transformations that reduce the antigen-capturing capacity, increase HLA class I and II and co-stimulatory molecule expression, develop an exceptional efficiency in presenting antigens to T cells, and augment the secretion of cytokines modulating T cell activation. In particular, mDC are major

4

producers of interleukin (IL)-12, while pDC are specialized in producing type I interferon (IFN) [4]. In patients with persistent hepatitis C virus (HCV) infection, virus-specific T cells often have defective effector function and overall low numbers [5, 6]. One hypothesis to explain the failure of these patients to generate or maintain an effective T cell response is that the virus impairs DC function. However, in spite of extensive studies, there is little consensus: some investigators reported reduced numbers and impaired functions of one or both DC subsets during persistent HCV infection, while others found that these cells are normal (reviewed in [7]). These inconsistencies may be due to differences in the DC purification and maturation protocols or in the measures used to evaluate DC activity. Nonetheless, the model of DC impairment by HCV clashes with the fact that persistently HCV-infected patients are not overtly immunocompromised and hence are not abnormally susceptible to other viral diseases [8]. Moreover, it is unlikely that HCV impairs DC in their ability to generate a T cell response against itself but not against other viruses. One important area of DC activity that has not been previously investigated regards the expression of the HLA class I antigen processing-presenting machinery (APM) components. These include the standard proteasome subunits delta (β1), MB1 (β5) and zeta (β2); the IFNinducible proteasome (immunoproteasome) subunits LMP2 (β1i), LMP7 (β5i) and LMP10 (β2i); the peptide transporters TAP1 and TAP2; and the endoplasmic reticulum chaperones calnexin, calreticulin, ERp57 and tapasin. Acting in concert, these molecules are responsible for the generation of antigenic peptides and for their translocation from the cytosol to endoplasmic reticulum and loading onto β2-microglobulin (β2m)-associated HLA class I heavy chains [9]. The generation of antigenic peptides, in particular, changes according to the types of catalytic subunits incorporated into proteasome particles. Indeed, the standard proteasome and the immunoproteasome have different cleavage specificities and produce

5

different sets of peptides [10]. Defects in the function or expression of APM components may affect the production of CD8 + T cell epitopes in terms of efficiency and repertoire. In order to clarify the effects of HCV infection on blood DC function, we examined the phenotype and function of mDC and pDC from persistently HCV-infected patients compared to the corresponding cells from uninfected persons, both healthy controls and patients who spontaneously recovered from HCV infection.

6

PATIENTS AND METHODS Three groups of individuals were investigated: 15 patients who spontaneously recovered from HCV infection, 15 patients persistently infected with HCV, and 10 healthy blood donors. The first group included individuals who had spontaneously cleared HCV RNA more than 3 years earlier, while the second group was composed of individuals with chronic hepatitis who were antiviral therapy-naive at the time of sampling. All subjects were negative for the hepatitis B surface antigen and for antibodies to human immunodeficiency virus. The study protocol was approved by the University of Bari Medical School Ethics Committee and conformed to the good clinical practice guidelines of the Italian Ministry of Health and the ethical guidelines of the Declaration of Helsinki, as revised and amended in 2004. Informed consent was obtained from each individual. Reagents and procedures for specimen collection; virological and clinical tests; cell isolation, culture, staining and analysis; ELISA, real-time PCR, and statistical analyses are provided in the supplementary methods file.

7

RESULTS To search for possible DC defects under conditions of persistent HCV antigenemia, we explored a variety of features and functions of blood DC from 15 patients with persistent HCV infection (persistently infected, PI), 15 patients who spontaneously recovered from HCV infection (spontaneous resolvers, SR), and 10 healthy control individuals (healthy donors, HD) (Table 1). First, we determined the frequency of mDC and pDC in peripheral blood of the three groups. To this aim, freshly drawn whole blood samples were immunostained and analyzed by flow cytometry. Frequencies were determined as the proportion of live cells that were Lin-, HLADR+ (HLA class II), and either CD11c+ (mDC) or CD123 + (pDC) (Figure 1A). Overall, there was considerable variability, within and among the groups. The median percentage of mDC was greater in HD (1.56%; range 0.31%-3.40%) than in PI (0.70%; range 0.07%-1.90%) with intermediate values in SR (1.10%; range 0.30%-2.90%). However, only the difference between HD and PI was significant (P = 0.0472) (Figure 1B). A similar pattern was observed for the absolute number of mDC, which was significantly different (P = 0.0414) between HD with the highest value (19.00/µL; range 8.00-30.00) and PI with the lowest value (9.50/µL; range 1.00-25.00) (Figure 1C). The median percentage of pDC was comparable among the three groups although it tended to progressively decrease moving from HD (0.75%; range 0.23%-1.09%) to SR (0.50%; range 0.20%-1.60%) and to PI (0.30%; range 0.01%-1.14%) (Figure 1D). The absolute number of pDC was similar in HD (6.00/µL; range 2.00-17.00) and SR (6.00/µL; range 2.00-11.00) but significantly lower in PI (3.50/µL; range 1.00-7.00) (Figure 1E). Altogether, these results indicate that the frequency of mDC and pDC is reduced in peripheral blood of PI patients. Moreover, a progressive decrease in the absolute numbers

8

of mDC and pDC paralleled the increase in the histological grade of chronic hepatitis in PI (Figure 1F-G). We next examined the expression of HLA and co-stimulatory molecules on the surface of cells exposed or not to a maturation stimulus, i.e. poly(I:C) for mDC and the CpG motifcontaining oligonucleotide ODN2216 for pDC. To this aim, freshly prepared samples of peripheral blood mononuclear cells (PBMC) were cultured in the absence or presence of the TLR ligand for 20 hours prior to immunostaining with mAb specific for HLA-ABC (class I), HLA-DR (class II), CD80 or CD86 (together with mAb for the common lineage markers and DC markers, as in Figure 1). Flow cytometric analysis of expression levels, expressed in molecular equivalents of soluble fluorochrome (MESF) units, revealed a significant upregulation of HLA-ABC and HLA-DR (Figure 1H) as well as CD80 and CD86 (Figure 1I) in mDC and pDC within each group. Expression levels were similar when compared among the three groups in either the unstimulated or stimulated condition. Stimulation with a control oligonucleotide lacking the CpG motif resulted in no substantial changes in expression levels of HLA-ABC, HLA-DR, CD80 or CD86 (data not shown). Overall these results demonstrate that the ability of mDC and pDC from PI to phenotypically mature is equivalent to that of HD and SR. We also compared the ability of mDC and pDC from the three groups to phagocytize foreign particles. To this end, mDC and pDC were immunomagnetically isolated from PBMC samples and pooled for each patient (to have enough cells for the experiments). Pooled DC were labeled with the membrane dye PKH67 (green fluorescence) and incubated at 37°C with necrotic HepG2 cells (labeled with the red dye PKH26) with no stimulant or with poly(I:C) and ODN2216 combined. After 20 hours, the uptake of apoptotic cells by DC was evaluated by flow cytometry as the percentage of doubly stained cells. The DC showed strong and comparable phagocytic activities with no significant differences among the three groups in the

9

unstimulated condition. Combined poly(I:C) and ODN2216 stimulation strongly reduced (but did not abolish) their phagocytic activities to similar extents in all groups (Figure 2A). Phagocytosis was negligible in control samples incubated on ice (data not shown). Pooled DC were also incubated with opsonized fluorescent latex beads, immunostained for surface markers and analyzed by flow cytometry. No inter- and intra-group differences were found between mDC and pDC regarding the degree of phagocytosis (Figure 2B-D). These results indicate that DC from PI are fully able to undergo maturation, with a decline in their phagocytic capacity that parallels the increased expression of HLA and co-stimulatory accessory molecules. To investigate cytokine production, immunomagnetically purified mDC and pDC were stimulated with poly(I:C) and ODN2216, and supernatant levels of IL-12p70 and IFN-α, respectively, were measured by ELISA. Poly(I:C) induced IL-12 secretion in comparable amounts among the three groups (Figure 2E). In the same way, ODN2216-induced IFN-α was readily observed in all three groups (Figure 2F). Similar results were observed for stimulated PBMC, assayed by flow cytometry (Figure 2G, H). Stimulation with a control oligonucleotide lacking the CpG motif did not increase IFN-α production (data not shown). Overall these findings suggest that the cytokine production capacity of mDC and pDC from PI is equivalent to that of HD and SR. We then examined the levels of APM proteins in mDC and pDC from the three groups. Freshly prepared PBMC were cultured in the absence or presence of poly(I:C) and ODN2216, intracellularly immunostained and analyzed by flow cytometry. The standard proteasome subunits delta, MB1 and zeta and the immunoproteasome subunit LMP2 were significantly upregulated in mDC from HD and SR, whereas their levels did not change in response to stimulation in PI (Figure 3A). Similar results were observed for pDC (Figure 3B). The immunoproteasome subunits LMP7 and LMP10 were, instead, significantly downregulated in

10

both mDC and pDC from all three groups (with the exception of LMP7 in pDC from PI) (Figure 3A-C). The transporters TAP1 and TAP2 were upregulated in both mDC and pDC from HD and SR whereas in PI they did not change in mDC and were downregulated in pDC. The chaperones tapasin, calnexin, calreticulin and ERp57 were significantly upregulated in both mDC and pDC from all three groups (Figure 3A-B). Expression levels of the APM components were then used to calculate ratios between the immunoproteasome and standard proteasome subunits; this was done for each patient and averaged for each group (Table 2). The three ratios, namely LMP2/delta, LMP7/MB1, and LMP10/zeta, all decreased upon stimulation, in both mDC and pDC from all three groups. These results suggest that DC maturation involves a replacement of immunoproteasome subunits by standard proteasome subunits. The differences in mean values between unstimulated and stimulated conditions, however, were substantially lower in PI than in SR and HD. To determine whether variations in expression of the 12 APM components reflected differences in transcriptional activity, we quantified the relevant mRNA in poly(I:C)stimulated mDC and ODN2216-stimulated pDC and expressed the data as fold change relative to the mean levels for unstimulated mDC and pDC. In mDC, delta, MB1, zeta, tapasin, calnexin, calreticulin, and ERp57 mRNA levels were upregulated while LMP7 and LMP10 mRNA levels were downregulated in all three groups (Figure 3D). Similar results were observed for pDC (Figure 3E). Differences between groups emerged for LMP2, TAP1 and TAP2, which were upregulated in both mDC and pDC from HD and SR but downregulated in mDC and pDC from PI. No significant differences among the groups were found in the extent of change when this had the same direction (Figure 3D-E). These results demonstrate that variations in the protein levels of APM components are regulated at the transcriptional level, with the exception of LMP2 in mDC and pDC from PI. Overall these

11

findings indicate that, following stimulation, mDC and pDC from PI do not undergo the strong switch from immunoproteasome to standard proteasome subunit expression and do not display the TAP upregulation which was seen in DC from uninfected individuals. To evaluate whether the differences in APM expression correlated with different in vitro abilities to present antigens via the HLA class I pathway, we used two approaches. In a first set of experiments, immunomagnetically purified mDC and pDC were cultured with immunomagnetically purified, 5(6)-carboxyfluorescein diacetate succinimidyl ester (CFSE)labeled allogeneic CD8 + T cells in the absence or presence of poly(I:C) and ODN2216, respectively. The percentage of CD8 + T cells induced to proliferate was measured by flow cytometry after 7 days. The CD8 + T cell responses to unstimulated mDC and pDC were small and comparable among HD, SR, and PI. The CD8 + T cell responses to poly(I:C)-treated mDC and ODN2216-treated pDC were more vigorous, but remained similar among groups (Figure 4A-B). Control cultures without DC did not exhibit CD8+ T cell proliferation (data not shown). In a second set of experiments, immunomagnetically purified and pooled mDC and pDC from selected HLA-A*0201 (A2)-positive patients (4 HD, 4 SR, 4 PI) were cultured with immunomagnetically purified autologous CD8 + T cells and the Influenza Matrix protein in the absence or presence of poly(I:C) and ODN2216 combined. The Influenza Matrix protein was chosen because it is a recall antigen and the presentation of its immunodominant A2-restricted Matrix 58-66 epitope is LMP7-dependent [11]. The percentage of Matrix 58-66 epitope-specific CD8 + T cells was detected by pentamer staining and flow cytometry after 7 days. DC from HD and SR were less able to expand Matrix 58-66 epitope-specific CD8+ T cells than DC from PI in cultures with poly(I:C) or ODN2216 (Figure 4C, E). Of note, the opposite was true when the experiment was repeated using the synthetic Matrix 58-66 peptide instead of the whole protein (Figure 4D-E). Nonspecific pentamer binding was excluded by performing the same experiment in A2-negative control patients (data not shown).

12

Finally, to confirm the different ability of stimulated DC from PI to present the processed HLA-class I-restricted Matrix 58-66 epitope, we used a short-term readout such as IFN-γ release. Specifically, we generated a Matrix 58-66 epitope-specific CD8 + T cell line from the PBMC of the A2-positive patients. We then immunomagnetically purified CD8 + T cells from these cell lines and incubated them with autologous immunomagnetically purified and pooled Matrix protein-loaded mDC and pDC in the presence of poly(I:C) and ODN2216. IFN-γ production was measured by flow cytometry of intracellularly stained cells after 6 hours of incubation and by ELISA in supernatants after 48 hours. Both assays showed higher CD8 + T cell stimulatory activity of DC from PI than from HD and SR (Figure 4F-G). On the whole, these data suggest that the altered regulation of HLA class I APM components in DC from PI changes the processing ability of these cells, which can have a substantial impact on CD8 + T cell activation.

13

DISCUSSION This study shows that persistently HCV-infected patients (PI) have functionally competent blood DC, although at lower frequency than uninfected persons. Moreover, in these cells, the HLA class I APM is regulated differently in PI compared to uninfected individuals. The different APM profile of PI translates into a different stimulatory activity of CD8+ T cells. Our study adds to the growing evidence that reduced circulating levels of mDC and pDC are a distinctive mark of HCV infection [12-16]. These findings may be explained by an increased migration of mDC and pDC to the liver (i.e. the site of chronic inflammation) or, alternatively, an increased propensity of these cells to undergo apoptosis, as has been demonstrated for pDC in response to type I IFN [17]. Our study, however, reveals that the loss of circulating DC associates with the histological grade of liver inflammation. Therefore, the decline in DC is more likely due to their compartmentalization into the liver where they have been histologically described [15, 16, 18-21]. This possibility is supported by the fact that pDC sense HCV-infected hepatocytes and produce high amounts of type I IFN in response [22]. In line with previous reports [14, 23-26], our study also documents that DC from PI undergo normal phenotypic and functional maturation in response to appropriate stimuli and are able to promote allogeneic CD8 + T cell proliferation. These results support the clinical evidence that persistently HCV-infected patients are not immunocompromised [8], although we cannot exclude that liver-resident DC or DC selectively involved in promoting HCV-specific immune responses may be defective. However, since current knowledge about intrahepatic DC is largely incomplete, we do not know whether blood DC remain normally functional once they have migrated to the infected liver or if the tolerogenic liver microenvironment [27] causes cell dysfunction.

14

The novel aspect of our study is that pertaining to the assessment of APM during mDC and pDC maturation. In uninfected persons (HD and SR) this process consists of two events: first, a dramatic change in the type of molecules involved in the processing of antigenic proteins, e.g. a replacement of immunoproteasome subunits with standard proteasome subunits; and second, a strong increase in molecules needed for transport (TAP) of protein fragments and assembly (calnexin, calreticulin, tapasin, ERp57) of the HLA class I–peptide complex. In persistently HCV-infected patients (PI), conversely, mDC and pDC maturation is not characterized by the strong switch from immunoproteasome to standard proteasome subunit expression and by the upregulation of TAP1 and TAP2. This different expression of APM components has functional consequences in that it entails a differential ability to present an immunodominant HLA class I-restricted peptide, whose generation from the native protein is strictly dependent on the presence of the LMP7 immunoproteasome subunit. Although the exact explanation for these findings is unknown, one hypothesis is that, under continuing antigenic pressure, DC maximize their capacity to capture and process antigens into immunodominant epitopes at the expense of peptide translocation and HLA class I-peptide complex formation. In particular, DC try to generate more restricted sets of epitopes, thus focusing the immune response toward T cell epitopes exclusively or preferentially generated in infected and inflamed tissues. Whether the proteasome activator PA28, a protein that regulates the incorporation of immunosubunits into proteasomes [28], plays a role in the mismatch between the changes in mRNA and protein levels of LMP2 upon maturation also remains to be established. A final comment must be made on the effect of TLR stimulation on DC maturation and activation of T cells. In our study, poly(I:C) and ODN2216 promoted changes in the expression pattern of APM components in DC, highlighting differences in epitope-specific CD8 + T cell responses. This finding, along with recent work showing that TLR ligands

15

enhance HLA class I presentation of exogenous antigens by DC [29], breaks the classical

view that the TLR ligand-mediated adjuvant effect on T cell responses only involves

stimulation through ‘second signals’ such as costimulatory molecules (collectively termed

signal 2) and cytokines (collectively termed signal 3). Instead, this finding supports the

concept of a ‘direct’ effect of TLR on the so-called signal 1, i.e. the antigen-HLA class I

complex. This possibility, and the other observations emerging from this research, deserve

further investigation, given that this study only considered the response to a single

immunoproteasome-dependent epitope, due to the small number of well-known CD8+ T cell

epitopes presented by human cells that depend on specific proteasome subunits and derive

from recall antigens to which both uninfected and HCV-infected individuals are immune. In conclusion, this study establishes that, under conditions of persistent HCV antigenemia, HLA class I APM is distinctively regulated during DC maturation. The implications of our study may go beyond HCV infection by providing insight into the general behavior of DC in both health and disease. This knowledge may help design novel immunotherapies to enhance the immune response to persisting pathogens.

16

ACKNOWLEDGEMENTS This study was supported by research funding from Associazione Italiana per la Ricerca sul Cancro (AIRC), Milan, Italy.

17

REFERENCES [1] Ueno H, Klechevsky E, Morita R, Aspord C, Cao T, Matsui T, et al. Dendritic cell subsets in health and disease. Immunol Rev 2007;219:118-142. [2] Ziegler-Heitbrock L, Ancuta P, Crowe S, Dalod M, Grau V, Hart DN, et al. Nomenclature of monocytes and dendritic cells in blood. Blood 2010;116:e74-e80. [3] Merad M, Sathe P, Helft J, Miller J, Mortha A. The dendritic cell lineage: ontogeny and function of dendritic cells and their subsets in the steady state and the inflamed setting. Annu Rev Immunol 2013;31:563-604. [4] Ito T, Kanzler H, Duramad O, Cao W, Liu YJ. Specialization, kinetics, and repertoire of type 1 interferon responses by human plasmacytoid predendritic cells. Blood 2006;107:2423-2431. [5] Racanelli V, Rehermann B. Hepatitis C virus infection: when silence is deception. Trends Immunol 2003;24:456-464. [6] Racanelli V, Leone P, Grakoui A. A spatial view of the CD8+ T-cell response: the case of HCV. Rev Med Virol 2011;21:347-357. [7] Ryan EJ, O'Farrelly C. The affect of chronic hepatitis C infection on dendritic cell function: a summary of the experimental evidence. J Viral Hepat 2011;18:601-607. [8] Jou JH, Muir AJ. In the clinic. Hepatitis C. Ann Intern Med 2012;157:ITC6. [9] Sant A, Yewdell J. Antigen processing and recognition. Curr Opin Immunol 2003;15:66-68.

18

[10] Guillaume B, Chapiro J, Stroobant V, Colau D, Van HB, Parvizi G, et al. Two abundant proteasome subtypes that uniquely process some antigens presented by HLA class I molecules. Proc Natl Acad Sci U S A 2010;107:18599-18604. [11] Gileadi U, Moins-Teisserenc HT, Correa I, Booth BL, Jr., Dunbar PR, Sewell AK, et al. Generation of an immunodominant CTL epitope is affected by proteasome subunit composition and stability of the antigenic protein. J Immunol 1999;163:6045-6052. [12] Anthony DD, Yonkers NL, Post AB, Asaad R, Heinzel FP, Lederman MM, et al. Selective impairments in dendritic cell-associated function distinguish hepatitis C virus and HIV infection. J Immunol 2004;172:4907-4916. [13] Averill L, Lee WM, Karandikar NJ. Differential dysfunction in dendritic cell subsets during chronic HCV infection. Clin Immunol 2007;123:40-49. [14] Pachiadakis I, Chokshi S, Cooksley H, Farmakiotis D, Sarrazin C, Zeuzem S, et al. Early viraemia clearance during antiviral therapy of chronic hepatitis C improves dendritic cell functions. Clin Immunol 2009;131:415-425. [15] Ulsenheimer A, Gerlach JT, Jung MC, Gruener N, Wachtler M, Backmund M, et al. Plasmacytoid dendritic cells in acute and chronic hepatitis C virus infection. Hepatology 2005;41:643-651. [16] Wertheimer AM, Bakke A, Rosen HR. Direct enumeration and functional assessment of circulating dendritic cells in patients with liver disease. Hepatology 2004;40:335-345. [17] Swiecki M, Wang Y, Vermi W, Gilfillan S, Schreiber RD, Colonna M. Type I interferon negatively controls plasmacytoid dendritic cell numbers in vivo. J Exp Med 2011;208:2367-2374.

19

[18] Galle MB, DeFranco RM, Kerjaschki D, Romanelli RG, Montalto P, Gentilini P, et al. Ordered array of dendritic cells and CD8+ lymphocytes in portal infiltrates in chronic hepatitis C. Histopathology 2001;39:373-381. [19] Kunitani H, Shimizu Y, Murata H, Higuchi K, Watanabe A. Phenotypic analysis of circulating and intrahepatic dendritic cell subsets in patients with chronic liver diseases. J Hepatol 2002;36:734-741. [20] Lai WK, Curbishley SM, Goddard S, Alabraba E, Shaw J, Youster J, et al. Hepatitis C is associated with perturbation of intrahepatic myeloid and plasmacytoid dendritic cell function. J Hepatol 2007;47:338-347. [21] Velazquez VM, Hon H, Ibegbu C, Knechtle SJ, Kirk AD, Grakoui A. Hepatic enrichment and activation of myeloid dendritic cells during chronic hepatitis C virus infection. Hepatology 2012;56:2071-2081. [22] Takahashi K, Asabe S, Wieland S, Garaigorta U, Gastaminza P, Isogawa M, et al. Plasmacytoid dendritic cells sense hepatitis C virus-infected cells, produce interferon, and inhibit infection. Proc Natl Acad Sci U S A 2010;107:7431-7436. [23] Barnes E, Salio M, Cerundolo V, Francesco L, Pardoll D, Klenerman P, et al. Monocyte derived dendritic cells retain their functional capacity in patients following infection with hepatitis C virus. J Viral Hepat 2008;15:219-228. [24] Dolganiuc A, Paek E, Kodys K, Thomas J, Szabo G. Myeloid dendritic cells of patients with chronic HCV infection induce proliferation of regulatory T lymphocytes. Gastroenterology 2008;135:2119-2127. [25] Longman RS, Talal AH, Jacobson IM, Rice CM, Albert ML. Normal functional capacity in circulating myeloid and plasmacytoid dendritic cells in patients with chronic hepatitis C. J Infect Dis 2005;192:497-503.

20

[26] Piccioli D, Tavarini S, Nuti S, Colombatto P, Brunetto M, Bonino F, et al. Comparable functions of plasmacytoid and monocyte-derived dendritic cells in chronic hepatitis C patients and healthy donors. J Hepatol 2005;42:61-67. [27] Racanelli V, Rehermann B. The liver as an immunological organ. Hepatology 2006;43:S54-S62. [28] Koch F, Trockenbacher B, Kampgen E, Grauer O, Stossel H, Livingstone AM, et al. Antigen processing in populations of mature murine dendritic cells is caused by subsets of incompletely matured cells. J Immunol 1995;155:93-100. [29] Crespo MI, Zacca ER, Nunez NG, Ranocchia RP, Maccioni M, Maletto BA, et al. TLR7 triggering with polyuridylic acid promotes cross-presentation in CD8alpha+ conventional dendritic cells by enhancing antigen preservation and MHC class I antigen permanence on the dendritic cell surface. J Immunol 2013;190:948-960.

21

Table 1. Demographic and clinical parameters of the study population Patient group PI (n=15)

SR (n=15)

HD (n=10)

5

8

6

10

7

4

52.4 (9.5)

53.7 (10.1)

46.5 (8.3)

1b

1b

NA

1124 (404)

0

NA

Yes

Yes

No

1

2

NA

NA

2

8

NA

NA

3

4

NA

NA

1

NA

NA

70.3 (25.5)

20.9 (8.0)

22.9 (3.0)

Gender, n

Male Female

Age, years

a

HCV genotype

HCV RNA, IU/ml a Anti-HCV antibodies Hepatitis histological grade, n

4 ALT, U/l

a

Abbreviations: ALT, alanine aminotransferase; NA, not applicable. a Values are mean (SD)

22

Table 2. Immunoproteasome/standard proteasome subunit MESF ratios

LMP2/delta

LMP7/MB1

LMP10/zeta

ex vivo

2.30 ± 1.37

2.12 ± 1.76

9.11 ± 1.60

stimulated

1.70 ± 1.05

0.92 ± 1.08

0.56 ± 0.80

difference

0.60

1.20

8.55

ex vivo

1.95 ± 1.47

2.52 ± 1.58

8.34 ± 4.35

stimulated

1.66 ± 1.53

1.02 ± 0.66

1.48 ± 1.42

difference

0.29

1.50

6.86

ex vivo

1.68 ± 0.69

2.19 ± 1.23

7.92 ± 1.94

stimulated

1.58 ± 1.10

1.54 ± 1.17

4.69 ± 2.40

difference

0.10

0.65

3.23

ex vivo

1.91 ± 1.45

2.99 ± 1.88

6.47 ± 2.46

stimulated

1.37 ± 0.87

1.07 ± 0.74

2.82 ± 1.73

difference

0.54

1.92

3.65

ex vivo

1.85 ± 0.97

2.70 ± 1.62

5.28 ± 1.54

stimulated

1.26 ± 0.25

1.19 ± 0.74

1.39 ± 0.71

difference

0.59

1.51

3.89

ex vivo

1.47 ± 1.10

1.98 ± 1.53

4.88 ± 2.51

stimulated

1.32 ± 0.76

1.60 ± 1.08

3.70 ± 2.84

difference

0.15

0.38

1.18

mDC

HD

SR

PI

pDC

HD

SR

PI

mDC and pDC were stimulated with poly(I:C) and ODN2216, respectively.

23

LEGEND TO FIGURES Figure 1. Frequency and phenotype of circulating DC in healthy donors (HD), spontaneous HCV resolvers (SR), and persistently HCV-infected patients (PI). (A) Strategy of flow cytometric analysis of immunostained whole blood cells: gating was done sequentially on live cells, on lineage-negative cells, and then on either CD11c+HLA-DRhi cells (mDC) or CD123 +HLA-DR+ cells (pDC). (B-E) Percentages and absolute numbers of mDC and pDC, expressed as median, interquartile range (box), and range (whiskers). Kruskal-Wallis analysis of variance and Mann-Whitney test for comparisons of groups. (F-G) Association between mDC and pDC frequency and histological grade in PI. (H-I) Changes in expression levels of HLA and co-stimulatory molecules in mDC and pDC upon stimulation with poly(I:C) and ODN2216, respectively. Bar graphs show median and standard deviation of expression levels in gated cells reported in units of molecular equivalents of soluble fluorochrome (MESF) in HD (white), SR (light gray) and PI (dark gray). Wilcoxon signed rank test. Overlay plots are representative and show isotype control (light line) and the expression of the indicated molecule in the absence (dotted line) or presence (dark line) of stimulation in each group. Figure 2. Phagocytosis and cytokine production by DC. (A-D) Changes in phagocytic activity of mDC and pDC upon stimulation with poly(I:C) and ODN2216, determined by flow cytometry. (A) Representative plots and percentages of pooled DC containing HepG2 fragments from HD (white), SR (light gray), and PI (dark gray). Mann-Whitney test. (B-C) Representative plots and percentages of mDC (B) and pDC (C) containing fluorescent latex beads from HD (white), SR (light gray), and PI (dark gray). Mann-Whitney test. (D) Representative micrograph of a DC (from a PI patient) that has engulfed fluorescent latex beads. (E-H) Cytokine production by mDC and pDC in response to no stimulation or stimulation with poly(I:C) and ODN2216. (E-F) Levels of IL-12 and IFN-α in mDC and pDC

24

culture supernatants, respectively, assessed by ELISA in HD (white), SR (light gray) and PI (dark gray). Mann-Whitney test. (G-H) Flow cytometric analysis of IL-12 (G) and IFN-α (H) production by mDC and pDC, respectively. Representative plots are gated on mDC and pDC according to Figure 1. IL-12 and IFN-α expression levels in positive cells are reported in MESF units in HD (white), SR (light gray) and PI (dark gray). The MESF rate was calculated as the MESF of the stimulated sample divided by that of the unstimulated sample for the same subject. Figure 3. APM component regulation in DC. (A-B) Changes in protein expression levels upon stimulation with poly(I:C) and ODN2216 in mDC and pDC, respectively, assessed by flow cytometry in HD (white), SR (light gray) and PI (dark gray). Graphs show expression levels in positive cells reported in MESF units. Wilcoxon signed rank test for within-subject comparisons. (C) Representative flow cytometry plots from one SR and one PI patient showing LMP7 expression in pDC gated according to Figure 1. Histogram overlays show the different shift in fluorescence. (D-E) Relative levels of mRNA in mDC and pDC stimulated with poly(I:C) and ODN2216, respectively. Transcript levels were determined by real-time PCR and the 2-∆∆Ct method, were normalized to those of TATA-box binding protein for the same samples, and then were expressed as fold change relative to the average value for unstimulated mDC or pDC. Values are mean and SD for 10 HD (white), 15 SR (light gray) and 15 PI (dark gray). Kruskal-Wallis test. Figure 4. CD8+ T cell activation by DC. (A-B) Mixed lymphocyte reactions. Top, Percentages of allogeneic CD8+ T cells undergoing one or more divisions upon incubation with untreated and poly(I:C)-treated mDC (A) and ODN2216-treated pDC (B) from HD (white), SR (light gray), and PI (dark gray), determined by flow cytometry. Bottom, Representative CFSE profiles from one patient of each group. Sequential peaks of lower

25

fluorescence intensity identify subsequent generations of proliferating daughter cells. (C-G) CD8 + T cell epitope presentation assays. (C-D) Percentages of autologous CD8+ T cells binding the Influenza Matrix 58-66 pentamer upon incubation with Matrix protein-loaded (C) and Matrix 58-66 peptide-loaded (D) DC in absence and presence of poly(I:C) and ODN2216, determined by flow cytometry in HD (white), SR (light gray), and PI (dark gray). Percentages refer to the overall population of live cells. Kruskal-Wallis test. (E) Representative plots from one PI patient showing the different sizes of pentamer-binding populations. (F-G) IFN-γ production by in vitro expanded autologous Influenza Matrix 58-66 peptide-specific CD8 + T cells upon incubation with Matrix protein-loaded DC in presence of poly(I:C) and ODN2216, determined by flow cytometry (F) and ELISA (G) in HD (white), SR (light gray) and PI (dark gray). Percentages refer to the overall population of live cells. Kruskal-Wallis test.

Figure 1

HLA-DR H

mDC

A SSC

HLA-DR

Lin-

CD11c

Lineage (CD3,14,16,19,20,56)

HLA-DR

pDC FSC

CD123

1

HD SR

PI

HD SR

2.0

20

1.5

15

1.0 0.5 0.0

PI

HD SR

PI

5 0

HD SR

20

5.0

10 0

PI

2.5 0.0

0 1 2 3 4 5

Histological grade

0 1 2 3 4 5

Histological g grade g

0

p=0.0017 p=0.0046 p=0.0020 0

200

400

600

ODN2216 -

p=0.0010 0

p=0.0010 400

600

HLA-DR MESF U x

104

p=0.0020 20

40

HLA-DR

CD86

60

104

p