ORIGINAL ARTICLE
Peripheral and Local Human Papillomavirus 16–Specific CD8 þ T-Cell Expansions Characterize Erosive Oral Lichen Planus Manuelle Viguier1,2,13, Herve´ Bachelez2,3,13, Be´atrice Poirier1, Je´re´my Kagan4,5,6,7, Maxime Battistella1,8, Franc¸ois Aubin9, Antoine Touze´10, Maryvonnick Carmagnat11, Camille France`s12, Marie-Lise Gougeon1,4 and Nicolas Fazilleau4,5,6,7,14 Erosive oral lichen planus (OLP) is a chronic, disabling mucocutaneous dysimmune rare disease characterized by mucosal inflammatory erosive lesions with pathological evidence for a marked CD8 þ cytotoxic T-lymphocyte (CTL) infiltration. However, the specificity of lesional CTL in OLP has never been analyzed. To investigate the molecular mechanisms underlying dysregulation of T-cell immune responses in patients with OLP, we studied the diversity and antigen specificity of the TCR expressed by CD8 þ T cells using dextramer staining, spectratyping, and TCR sequencing in 10 OLP patients undergoing extracorporeal photochemotherapy. Expansions of TCRVb3bearing CD8 þ T cells were found in peripheral blood and in lesional tissues of OLP patients. Spectratyping and sequencing studies identified specific clonotypes in each patient. These expansions were enriched with human papillomavirus 16 (HPV16)–specific CD8 þ T cells in HLA-A*0201 þ patients as shown by their immune recognition of the E711-20 immunodominant epitope. Under treatment with extracorporeal photochemotherapy, clonotypic CD8 þ T-cell expansions decreased in parallel with clinical remission. Altogether, these data establish a link between HPV infection and OLP pathogenesis by identifying a massive clonal expansion of CD8 þ T cells with increased frequency of HPV 16–specific CD8 þ T cells in OLP patients. Journal of Investigative Dermatology advance online publication, 23 October 2014; doi:10.1038/jid.2014.397
INTRODUCTION Lichen planus is a chronic inflammatory disease of unknown etiology involving the skin and/or mucous membranes, 1
Antiviral Immunity Biotherapy and Vaccine Unit, Institut Pasteur, Paris, France; 2Sorbonne Paris Cite´ Universite´ Paris Diderot, Paris, France; Universite´ Paris Descartes, Imagine Institute, Necker Hospital, Paris, France; 4 Centre de Physiopathologie de Toulouse Purpan, Toulouse, France; 5INSERM, U1043, Toulouse, France; 6CNRS, UMR5282, Toulouse, France; 7Universite´ Toulouse III Paul-Sabatier, Toulouse, France; 8Universite´ Paris Diderot, UMRS728, INSERM, Paris, France; 9Service de Dermatologie, Centre Hospitalier Universitaire, SFR FED 4234, EA 3181, Universite´ de Franche Comte´, Besanc¸on, France; 10UMR 1282 INRA-Universite´ Franc¸ois Rabelais, Tours, France; 11Laboratoire d’Immunologie, Hoˆpital Saint-Louis, Paris, France and 12 Service de Dermatologie-Allergologie, Hoˆpital Tenon, Universite´ Paris VI, AP-HP, Paris, France 3
13
The first two authors contributed equally to this work.
14
The last two authors contributed equally to this work.
Correspondence: Herve´ Bachelez, Sorbonne Paris Cite´ Universite´ Paris Diderot, Paris 75000, France or Marie-Lise Gougeon, Antiviral Immunity Biotherapy and Vaccine Unit, Institut Pasteur, Paris 75724, France or Nicolas Fazilleau, Centre de Physiopathologie de Toulouse Purpan, Toulouse 31300, France E-mail: [email protected] or [email protected] or [email protected] Abbreviations: CDR3, complementary determining region 3; CTL, cytotoxic T lymphocyte; ECP, extracorporeal photochemotherapy; HPV, human papillomavirus; OLP, oral lichen planus; PBMC, peripheral blood mononuclear cell Received 24 March 2014; revised 4 July 2014; accepted 16 August 2014; accepted article preview online 10 September 2014
& 2014 The Society for Investigative Dermatology
leading to a variety of clinical presentations (Le Cleach and Chosidow, 2012). Erosive oral lichen planus (OLP) is the most prevalent subtype, affecting 0.1 to 4% of the general population, and is characterized by a severely disabling course in its erosive form that often requires immunomodulatory systemic therapies such as oral steroids, immunosuppressants, and more recently extracorporeal photochemotherapy (ECP) (Becherel et al., 1998; Edwards and Kelsch, 2002; Guyot et al., 2007). Pathologically, OLP is characterized by a predominant infiltration of immune cells, mainly CD8 þ cytotoxic T lymphocytes (CTLs), associated with apoptosis of epithelial cells and disruption of the basement membrane zone (Zhou et al., 2002). Similarly, the presence in OLP lesions of CTL in the vicinity of damaged epithelial cells supports the hypothesis that a dysregulated CTL response of unknown antigen specificity has a key role in the pathogenesis of the disease (Santoro et al., 2004). However, the nature of these antigens remains controversial. Indeed, although autoreactivity of lesional skin–derived CTL against epithelial cells has been suggested in OLP patients (Sugerman et al., 2000; Gorsky and Epstein, 2011; Syrjanen et al., 2011), other studies identified the presence of DNA from several human papillomavirus (HPV) subtypes in mucosal lesions of OLP, raising the issue of the self versus non-self nature of the target antigens in OLP (Jontell et al., 1990; Young and Min, 1991; Miller et al., 1993; Lodi et al., 2005; Yildirim et al., 2011). So far, the lack of any study using accurate tools to address the specificity of lesional CTL in OLP www.jidonline.org
1
M Viguier et al.
HPV-Specific CD8 þ T Cells in Lichen Planus
the peripheral blood CD8 þ T lymphocytes (data not shown, 17.27±4.51%). The quantitative PCR products were further used for the amplification of dye-labeled oligonucleotides specific for Cb and allowed to assess the complementary determining region 3 (CDR3) length distribution. For each TCRVb gene segment tested and for the TCRVb3 gene segment from CD4 þ and CD8 þ T cells from healthy donors or from CD4 þ T cells isolated from OLP patients (Figure 1b), CDR3b length distribution profiles displayed Gaussian-like curves, the hallmark of a polyclonal T-cell repertoire. In contrast, CDR3b distribution profiles of the TCRVb3 gene segment were altered with single expansions in blood CD8 þ T cells from all OLP patients, as observed in OLP no. 5 for a CDR3b size of 11 amino acids (Figure 1b). Therefore, we determined the Vb-Jb rearrangement corresponding to the observed clonal expansion and further cloned and sequenced it. As shown in Figure 1c, one single CDR3b sequence was found in each patient. Furthermore, we revealed that similar TCRVb3 þ clonal expansions were also present in tissue lesions of three
precludes any conclusion about mechanisms underlying this dysregulation. In this study, using dextramer, spectratyping, and sequencing in 10 patients with severe erosive OLP, we investigated the antigen specificity of the CD8 þ T-cell expansions. RESULTS AND DISCUSSION TCRVb3 þ clonal expansions alter CD8 þ T-cell repertoires of OLP patients at flare
To understand the molecular events that govern the physiopathology of OLP, we investigated the T-cell repertoire of 10 patients with OLP. At flare, the usage of each TCRVb gene segment, as estimated by quantitative real-time reverse-transcriptase–PCR in CD4 þ and CD8 þ peripheral blood T cells, revealed striking predominance of the TCRVb3 gene segment in CD8 þ T cells (26.49±4.81%), whereas it was used at a much lower level in the circulating CD4 þ counterparts (11.49±2.24%) or in both isolated CD4 þ and CD8 þ T cells from healthy donors (Figure 1a). Results from quantitative real-time reverse-transcriptase–PCR were confirmed by flow cytometry, showing predominance of TCRVb3 expression in
Vβ3 usage in PBMC
**
Vβ3-Cβ
PBMC
*
CD4
CD8
HD 1 25
11
11
11
11
OLP 5 CD4
CD8
CD4
HD OLP
Jβ
1 2 3 4 5 6 7 8 9 10
1.5 2.7 2.7 2.7 2.2 2.1 1.2 2.7 2.7 2.2
OLP CDR3β (nucleotidic sequence)
CDR3β (AA)
TGTGCCAGCAGTTTTGTCATTGGGGGGGTTAGCAATCAGCCCCAG TGTGCCAGCAGTGAACCGGGACAGGGTTTTGGTGAGCAG TGTGCCAGCAGTATAGGGCTCCACTACGAGCAG TGTGCCAGCAGTTTAGGTGTGCACTACGAGCAG TGTGCCAGCAGTCAAGGAGTCTTCCTGGGGGCCGGGGAGCTG TGTGCCAGCAGTTTATTGGAGGGACTAGCGGGAGTGGATGAGCAG TGTGCCAGCAGTTATATAGGCCTACAGAGTAAAGCTAACTATGGCTAC TGTGCCAGCAGTGAACCGGGACAGGGTTTTGGTGAGCAG TGTGCCAGCAGTTTCCAACCAGGGGGGCAG TGTGCCAGCAGTTTAACCACCGGGGAGCTG
In situ
5 SEPGQGFGEQ In Vβ3+ (%)
Vβ3-Jβ2.7
CD8
2.5
OLP 2 þ
10
5 Vβ3-Jβ2.7
0
10
SFVIGGVSNPQ SEPGQGFGEQ SIGLHYEQ SLGVHYEQ SQGVFLGAGEL SLLEGLAGVDEQ SYIGLQSKANYGY SEPGQGFGEQ SFQPGGQ SLTTGEL
Vβ3-Jβ2.2 2.5
0
8 OLP 4 þ
SQGVFLGAGEL In Vβ3+ (%)
0
SLGVHYEQ In Vβ3+ (%)
Frequency (%)
50
5
0
11 OLP 5
Figure 1. TCRVb3 clonal expansions skewed peripheral and lesional CD8 T-cell repertoires of oral lichen planus (OLP) patients at flare. (a) Reversetranscriptase–PCR (RT–PCR) amplifications of the TCRVb3 usage in CD4 þ and CD8 þ T cells from the blood of healthy donors (HDs) or OLP patients. PBMC, peripheral blood mononuclear cell. (b) CDR3b length distributions for CD4 þ and CD8 þ Vb3-Cb rearrangements from a representative HD and a representative OLP patient. The peak of the 11th codon is marked on the abscissa axis. (c) Table of CDR3 clonotypes identified in each OLP patient (nucleotidic and amino-acid (AA) sequences). (d) CDR3b length distributions (left) and clonotype distribution estimated by sequencing (right) in lichen lesions in three representative OLP patients (OLP nos. 2, 4, and 5). *Po0.05, **Po0.01.
2
Journal of Investigative Dermatology (2014), Volume 00
M Viguier et al.
HPV-Specific CD8 þ T Cells in Lichen Planus
OLP patients (Figure 1d). Overall, neither CDR3b length nor consensus sequence was shared across all OLP patients. However, all detected CD8 þ clonal expansions were using the TCRVb3 gene segment, even in patients not sharing common class I HLA alleles (Supplementary Table S1 online). Interestingly, 5 out of the 10 clonotypic TCRs were composed of the Vb-Jb2.7 rearrangement, and one clonotypic sequence was found in two different patients (OLP nos. 2 and 8; Figure 1c). Of note, flow cytometric characterization of TCRVb3 þ CD8 þ T cells in peripheral blood mononuclear cells (PBMCs) confirmed their activated status as shown by surface expression of HLA-DR (data not shown, 20.99± 4.52%, n ¼ 7), CD38 (24.85±6.18%, n ¼ 4), and perforin content (15.2±4.54%, n ¼ 7). Altogether, these data show that clonotypic TCRVb3 þ CD8 þ T-cell expansions are detected in situ and in the blood of OLP patients at flare. These sets of data extend previous observations reported in two studies, one reporting the increased presence of TCRVb3and TCRVa2-positive cells in the oral mucosal T-cell infiltrate of 7 OLP patients (Simark-Mattsson et al., 1994), and the other performed in 12 OLP patients showing that CD8 þ T cells had skewed TCR repertoire toward the usage of a restricted set of TCRVb segments including TCRVb3 (Gotoh et al., 2008). Tracking of CD8 þ TCRVb3 þ clonal expansions following disease activity in OLP patients
We next reasoned that the CD8 þ clonal expansions could actively participate in the physiopathology of OLP and tested
this hypothesis by performing a longitudinal study of the TCR repertoire in all OLP patients. PBMCs were isolated at different time points corresponding to distinct clinical status: flare (before starting ECP), partial remission, complete remission following ECP, and relapse after ECP withdrawal. For all collected samples, TCRVb usage and CDR3b length distribution were performed as well as sequencing of Vb-Jb rearrangements corresponding to the clonal expansions at flare (Figure 1c). This allowed a follow-up of peaks corresponding to TCRVb3 þ clonal expansions in the T-cell compartment (Figure 2 and Supplementary Figure S1 online). For three patients (OLP nos. 5, 6 (Figure 2) and 8 (Supplementary Figure S1 online)), we observed a disappearance of the peaks corresponding to the TCRVb3 þ clonal expansions that strictly correlated with a clinical improvement (either partial remission or complete remission). However, for the seven remaining OLP patients (as an example OLP no. 4, Figure 2), CDR3 distribution profiles were all disturbed irrespective of the clinical status of the patients. Nevertheless, we found a strong correlation between the clonotype usage and the clinical status of OLP patients for 8 out of 10 patients. A pronounced decrease or disappearance of the frequency of TCRVb3 þ clonotype was observed in OLP patients 1, 4, 5, 6, and 8, whereas a slight decrease was found in OLP cases 2, 3, and 7 (Figure 2 and Supplementary Figure S1 online). Moreover, the clonotype identified at initial flare was also found at high frequency at the time of clinical relapse (OLP nos. 4, and 5 (Figure 2) and nos. 9 and 10 (Supplementary Figure S1 online)). Overall, these data support that clinical remission 80 SLGVHYEQ In Vβ3+ (%)
Vβ3-Jβ2.7
OLP 4 8 PR 1
8 PR 2
8 Relapse
Vβ3-Jβ2.2
OLP 5 11 PR 1
11 Relapse
Vβ3-Jβ2.1
OLP 6
12 Flare
12 CR
11 PR 2
12 PR
Flare
PR 1
PR 2 Relapse
Flare
0 PR 1 Relapse PR 2
100 75 50 25 0
SLLEGLAGVDEQ In Vβ3+ (%)
11 Flare
0
SQGVFLGAGEL In Vβ3+ (%)
8 Flare
40
15 10 5 0
Flare
0 CR
0 PR
Figure 2. Decrease or disappearance of TCRVb3 þ clonal expansions over the clinical course of the disease in oral lichen planus (OLP) patients. CDR3b length distributions (left column) for different CD8 þ Vb3-Jb rearrangements over the clinical course of the disease following extracorporeal photochemotherapy (ECP) and clonotype distribution estimated by sequencing (right column) for three different representative OLP patients (OLP nos. 4, 5, and 6). CR, complete remission; PR, partial remission.
www.jidonline.org
3
M Viguier et al.
HPV-Specific CD8 þ T Cells in Lichen Planus
obtained with ECP in OLP is associated in most cases with a decrease or disappearance of TCRVb3 clonotypic CD8 þ T-cell expansions, suggesting the pathogenic relevance of these clonotypic expansions. CD8 þ T cells expressing clonal TCRVb3 þ in OLP patients are specific for HPV16
Lesional CTLs from OLP patients are suspected to recognize an antigen associated with major histocompatibility complex class I molecule on lesional keratinocytes (Sugerman et al., 2000). Furthermore, several studies have shown the presence of several HPV subtypes in mucosal OLP lesions (Jontell et al., 1990; Young and Min, 1991; Miller et al., 1993; Lodi et al., 2005; Yildirim et al., 2011). Detection of current or past HPV infection could be performed in 6 out of our 10 patients either on tissue biopsies or by serology. Three out of six OLP tested patients were found HPV positive and, among them, HPV16 DNA and/or HPV16-specific IgG were found in 2 out of the 3 HPV-positive patients (OLP nos. 2 and 5). To test whether TCRVb3 þ clonal expansions might exhibit HPV specificity, we stained PBMCs from the five HLA-A*0201 þ OLP patients of the cohort with peptide-containing major histocompatibility complex class I dextramer. For all these patients, a distinct population of CD8 þ T cells expressing a TCRVb3 specific for the E711-20 immunodominant peptide of HPV16 was detected, as shown for OLP no. 2 at flare or OLP no. 4 under ECP (Figure 3a; mean percentage of CD8 þ Dext-HPV þ in TCRVb3 þ CD3 þ cells: 2.62±1.7%, n ¼ 5). As expected, almost no TCRVb3 þ T cells were stained with HLA-A2/HIVp17 Gag dextramer in these HIV-negative patients (Figure 3a; CD8 þ Dext-HIV þ : 0.64±0.3%, n ¼ 5). We then sorted out HPV-specific CD8 þ TCRVb3 þ cells as well as their CD8 þ TCRVb3 þ Dext-HPV- counterparts from OLP nos. 3, 4, and 7. Spectratyping analysis and CDR3b sequencing were performed, revealing a striking enrichment of Dext-HPV þ populations with the clonotype in contrast with Dext-HPV- subset that showed no enrichment (Figure 3b). In the last series of experiments, we performed in situ immunostaining on OLP lesion from OLP no. 2. The presence of cells expressing CD3, CD8, and the cytotoxic marker TIA1 (T-cell-restricted intracellular antigen 1) were detected (Figure 4a), some of these bearing TCR specific for HLA-A2/HPV E711-20 peptide, whereas staining with the irrelevant Dext-HIV yielded negative results in the same patient (Figure 4b). Overall, these data demonstrate that a notable proportion of clonal CD8 þ blood T cells of OLP patients are HPV16 specific, also infiltrating mucosal and/or skin lesions, and decreasing or disappearing following clinical remission, suggesting their pathogenic role in OLP. These results are reminiscent of results from previous in vitro studies showing the capacity of this E711-20 HPV16 epitope to favor the stimulation and expansion of CTL clones expressing TCRVb3, without evidence of HLA class I restriction (Nilges et al., 2003; Schreurs et al., 2003). These sets of data also raise the HPV infection status in OLP patients. Similarly, a large set of published HPV detection studies support a pathogenic contribution of some HPV subtypes in OLP (Gorsky and Epstein, 2011; Syrjanen et al., 2011). Notably, detection of 4
Journal of Investigative Dermatology (2014), Volume 00
viral DNA mainly derived from HPV6/11/16/31/33 has been reported in mucosal oral and/or genital lesions from OLP patients in several studies, providing a rationale for their role as putative antigenic stimuli (Lodi et al., 2005; Mattila et al., 2012). This latter hypothesis has been reinforced by a recent exhaustive systematic review of HPV detection in 39 selected cross-sectional studies, showing that HPV DNA is significantly more frequently detected in OLP lesions compared with controls (Syrjanen et al., 2011). Indeed, the positive detection of previous or current infection with HPV16 in 2/6 of our patients, including 1 patient who showed clear expansion of HPV16 E711-20–specific cells in blood and lesional mucosa, is a strong support for the contribution of HPV as an antigenic stimulus of CTL expansion that characterizes severe erosive OLP. Finally, the presence of HPV16 E711-20–specific CD8 þ T cells in the vicinity of dying epithelial cells in a studied patient correlates with their potential involvement in tissue damage of OLP, even though this latter observation does not rule out the deleterious role of other subtypes than HPV16 that were not addressed in our patients. As some of our patients did not show either the presence of HPV16 DNA or the serological evidence of previous HPV16 infection, it remains to be determined whether other HPV subtypes or even alternative viral agents would be able to trigger similar CTL expansions in OLP (Lodi et al., 2005). Nevertheless, HPV serology is a poor marker of current infection and HPV DNA detection can be rendered difficult, as most HPV infections are transient (Ho et al., 1998; Vaccarella et al., 2010). As most adult individuals are infected at some point with one or several HPV subtypes, the present results also raise the mechanisms underlying dysimmunity in OLP patients. Supporting the autoimmune hypothesis and its possible link with HPV is the previously established presence among lesional T lymphocytes of two OLP patients of CTL clones reacting toward autologuous keratinocytes immortalized with HPV16 E6 and E7 (Sugerman et al., 2000). It is worth noticing that a causal role for HPV infection in OLP does not necessarily require constant tissue viral replication, as an initial viral stimulation might be sufficient to trigger the founder immune response, with following autoimmune T-cell expansions related to molecular mimicry, unsequestration of masked self epitopes, or both (Selmi et al., 2012). As such, it might explain why we were able to detect antibodies to HPV16 in the sera of only two patients of our cohort. In line with the former of these hypotheses, molecular mimicry between HPV16 E7 protein and human self has been revealed by computer-based analyses, providing a rational basis for further investigations of OLP lesional CTL regarding their immunoreactivity toward HPV versus self-candidate antigens (Natale et al., 2000). However, alternative mechanisms such as alterations of the TCRVb3 þ /HLA-peptide synapse by some viral products cannot be ruled out. Finally, it also remains to be determined whether HPV and/or other viral stimuli are involved in other forms of lichen planus. Our data also suggest that prophylactic HPV16 vaccine in addition to preventing cervical dysplasia and cancer may also protect against OLP.
M Viguier et al.
HPV-Specific CD8 þ T Cells in Lichen Planus
OLP 2
OLP 4
CD3+ Vβ3+
HIV-1 SLYNTVATL
HPV YMLDLQPETT
HPV YMLDLQPETT
0.81%
2.86%
2.34%
Dextramer HLA-A*0201
0.73%
CD3+ Vβ3+
HIV-1 SLYNTVATL
CD8 80
Isolated T cells HPV+
HPV– OLP 7
SYIGLQSKANYGY In Vβ3+ (%)
Unsorted T cells
60 40 20 0
13
13
13
0 Unsorted HPV– HPV+ T cells Isolated T cells
Isolated T cells HPV–
HPV+
OLP 3
8
SIGLHYEQ In Vβ3+ (%)
50 Unsorted T cells
25
0
8
8
Unsorted HPV– HPV+ T cells Isolated T cells
Isolated T cells HPV–
HPV+
OLP 4
SLGVHYEQ In Vβ3+ (%)
100 Unsorted T cells
75 50 25 0
8
8
8
Unsorted HPV– HPV+ T cells Isolated T cells
Figure 3. A subpopulation of peripheral CD8 þ T cells from oral lichen planus (OLP) patients is human papillomavirus 16 (HPV16) specific and is enriched in TCRVb3 þ clonal expansions. (a) Flow cytometric detection using dextramer of HIV- and HPV-specific CD8 þ T cells in CD3 þ Vb3 þ cells from the blood of two representative HLA-A*0201 þ OLP patients (OLP nos. 2 and 4). (b) CDR3b length distributions for Vb3-Cb rearrangements on unsorted T cells or isolated T cells with HPV-Dextramer (left column) and clonotype distribution estimated by CDR3 sequencing (right column) for three different representative HLA-A*0201 þ OLP patients (OLP nos. 7, 3, and 4).
MATERIALS AND METHODS Patients and healthy donors Ten patients with OLP treated with ECP were enrolled. They consisted of 8 women and 2 men (mean age: 56.5 years (range: 25–78)) (Supplementary Table S1 online). The characteristics of the eight patients have been previously reported (Guyot et al., 2007). Previous treatment with glucocorticoids and immunosuppressive agents was stopped 4 weeks before ECP initiation. All patients reached remission under ECP, either complete remission or partial remission (defined by regression of lesions of at least 50%) (Guyot et al., 2007). Clinical
statuses were determined at the time of blood collection. The Institutional Review Board (Comite´ de Protection des Personnes Ile de France IV, 2009/10NI) approved the study that was qualified as noninterventional. The patients were fully informed and had no objection to participate in the study. Written patient consent was not required because French law considers that it is not mandatory for noninterventional biological collection for immunological and molecular research, with the exception of genetic studies on germinal DNA. Healthy donors were blood donors from the Etablissement Franc¸ais du Sang. www.jidonline.org
5
M Viguier et al.
HPV-Specific CD8 þ T Cells in Lichen Planus
CD3
CD8
Dextramer HLA-A*0201 HIV-1 SLYNTVATL
TIA1
Dextramer HLA-A*0201 HPV YMLDLQPETT
Figure 4. Human papillomavirus 16 (HPV16)–specific CD8 þ T cells infiltrate lichen lesions. (a) The in situ immunostainings of CD3, CD8 (original magnification 200, scale bars ¼ 100 mm), and TIA-1 (T-cell-restricted intracellular antigen 1) (original magnification 400, scale bar ¼ 50 mm) on a biopsy taken from mucosal lesions in oral lichen planus (OLP) patient no. 2. (b) The in situ immunostainings using either dextramer HIV-1 (left) or dextramer HPV (right) ( 630, scale bars ¼ 25 mm; inset original magnification 1,000, scale bars ¼ 10 mm).
Preparation of PBMCs Blood was collected at baseline before any ECP treatment (‘‘flare’’) and at several time points. PBMCs were prepared by Ficoll-Hypaque density gradient centrifugation and cryopreserved in liquid nitrogen in 8% dimethylsulfoxide, 42% fetal calf serum, and 50% RPMI-1640 medium (Invitrogen, Carlsbad, CA).
(Invitrogen Life Technologies, Carlsbad, CA). Quantitative PCR amplifications were performed with TaqMan Universal PCR Master Mix (Applied Biosystems, Carlsbad, CA), TCRVb gene segment– specific oligonucleotides, TCRCb-specific antisense primer, and a fluorescent probe specific for the TCRCb gene segment (Fazilleau et al., 2007).
Antibodies and flow cytometric analysis
Cloning and sequencing of CDR3b rearrangements
Anti-CD3 FITC (Clone BW264/56, Miltenyi Biotec, Bergisch Gladbech, Germany), anti-CD8-APC (SK1, BD Pharmingen, Franklin Lakes, NJ), Dextramer-HPV E711-20-PE (Dext-HPV PE; HLA-A*0201; YMLDLQPETT, Immudex, Copenhagen, Denmark), Dextramer-HIV-1 P17 Gag 77-85-PE (Dext-HIV-PE; HLA-A*0201, Bayport, MN; SLYNTVATL, Immudex), and anti-TCR-Vb3-FITC (JOVI-3, Ancell, Bayport, MN) were used. Cells were incubated with dextramer for 10 minutes at room temperature, and then antibodies were added for 30 minutes at 4 1C. Stained cells were washed with phosphatebuffered saline/0.5% BSA/0.1% NaN3 and analyzed using a FACSCalibur flow cytometer (Becton Dickinson,Franklin Lakes, NJ).
Vb-Jb PCR products were cloned in pCR 4Blunt TOPO vector (Invitrogen Life Technologies). Sequencing reactions were performed using the ABI PRISM Big Dye Terminator Reaction Kit (Applied Biosystems). Reaction products were analyzed on an ABI 3130XL 16 capillaries (Applied Biosystems).
Purification of T-lymphocyte subsets Dead cells were removed using the MACS Dead Cell Removal kit (Miltenyi Biotec). T cells were isolated using anti-CD8 or anti-CD4 antibody-coated immunomagnetic beads and stabilized in PrepProtect buffer (Miltenyi Biotec). Cell purities exceeded 85%. CD3 þ CD8 þ TCRVb3 þ Dext-HPV þ or Dext-HPV- populations were isolated using a MoFlo Astrios cytometer cell sorter (Beckman Coulter, Villepinte, France).
TCR repertoire analysis RNA was extracted using RNAeasy kit (Qiagen, Venlo, The Netherlands) either from sorted PBMCs or from total skin or mucosal OLP lesions and reverse-transcribed into complementary DNA using Superscript II 6
Journal of Investigative Dermatology (2014), Volume 00
Detection of HPV subtypes Tissue samples were obtained either atraumatically using a cytobrush (DNAPAP Cervical Sampler; Digene, Qiagen) secondarily placed in Cervical Specimen Transport Medium (Digene, Qiagen) or by tissue biopsy secondarily frozen. DNA extraction was performed using the QIAGEN BioRobot EZ1 with the EZ1 DNA Tissue Kit (Qiagen). Human papillomavirus detection was performed using two highly sensitive and specific assays combining a specific multiplex PCR with DNA microarray primer extension as described previously (Schmitt et al., 2006; Ruer et al., 2009). Multiplex PCR reactions were performed for the amplification of a-HPV, b-HPV, g-HPV, and mHPV types, respectively. The assays were combined in one single chip to detect 27 HPV types from genus a, 25 cutaneous HPV types from genus b, 16 cutaneous HPV types from genus g, and one from genus m. Following PCR amplification, 10 ml of each reaction mixture was analyzed using the Luminex technology (Luminex, Austin, TX). The detection limit was determined using the cutoff of 10 median reporter fluorescence intensity and varied between 100 and 800 pg of PCR products (Schmitt et al., 2006). For each sample, three multiplex
M Viguier et al.
HPV-Specific CD8 þ T Cells in Lichen Planus
PCR reactions were performed. Search for serum IgG antibodies against HPV16 major capsid protein L1 was performed using ELISA (Combita et al., 2002). Briefly, serum samples were tested in triplicate in a 1:20 dilution, and virus like particle-bound antibodies were detected using anti-human IgG conjugated to peroxidase diluted at 1:20,000 (Southern Biotech, Birmingham, AL). Cutoff value for positivity based on sample from young children (mean reactivity plus 3 SD) was 0.2. Previously tested samples were retested on the assay plate to confirm ensibility and consistency of the assay (Ferguson et al., 2006).
In situ immunostaining Acetone-fixed cryosections were incubated with dextramer for 75 minutes at room temperature and mounted with VECTASHIELD Mouting Medium with 4’,6-diamidino-2-phenylindole (Vector Laboratories, Burlingame, CA). Slides were studied using an Axiovert 200M microscope with MRm camera (Zeiss, Oberbochen, Germany).
Statistical analysis Unpaired t-tests were calculated using Prism software (Irvine, CA). CONFLICT OF INTEREST The authors state no conflict of interest, except AT who was a speaker at GSK symposia and for whom GSK and Sanofi Pasteur provided meeting and travel fees.
ACKNOWLEDGMENTS We are indebted to Laure Buisson for her help in sequencing analyses (UDEAR, CNRS, France), Pierre-Henri Comme`re for cell sorting (Flow Cytometry Facility, Pasteur Institute), Nathalie Parquet (Saint-Louis Hospital, Paris, France) and Claire Rabian (Saint-Louis Hospital, Paris, France) for their help in cell collection and cell preparation, and to Tarik Gheit (International Agency for Research in Cancer, Lyon, France) for HPV genotyping. This work was supported by Institut Pasteur, the Socie´te´ Franc¸aise de Dermatologie, AVENIR INSERM (NF), Association pour la Recherche sur le Cancer (ARC AO 2009, NF), la Ligue Nationale Contre le Cancer (NF), Conseil Re´gional Midi-Pyre´ne´es (NF), Institut National contre le Cancer (PLBIO10-195 and INCA-6530, NF), Collery Grant from Acade´mie Nationale de Me´decine (MB), and International Re-integration Grant Marie Curie. SUPPLEMENTARY MATERIAL Supplementary material is linked to the online version of the paper at http:// www.nature.com/jid
Gotoh A, Hamada Y, Shiobara N et al. (2008) Skew in T cell receptor usage with polyclonal expansion in lesions of oral lichen planus without hepatitis C virus infection. Clin Exp Immunol 154:192–201 Guyot AD, Farhi D, Ingen-Housz-Oro S et al. (2007) Treatment of refractory erosive oral lichen planus with extracorporeal photochemotherapy: 12 cases. Br J Dermatol 156:553–6 Ho GY, Bierman R, Beardsley L et al. (1998) Natural history of cervicovaginal papillomavirus infection in young women. N Engl J Med 338:423–8 Jontell M, Watts S, Wallstrom M et al. (1990) Human papilloma virus in erosive oral lichen planus. J Oral Pathol Med 19:273–7 Le Cleach L, Chosidow O (2012) Clinical practice. Lichen planus. N Engl J Med 366:723–32 Lodi G, Scully C, Carrozzo M et al. (2005) Current controversies in oral lichen planus: report of an international consensus meeting. Part 1. Viral infections and etiopathogenesis. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 100:40–51 Mattila R, Rautava J, Syrjanen S (2012) Human papillomavirus in oral atrophic lichen planus lesions. Oral Oncol 48:980–4 Miller CS, White DK, Royse DD (1993) In situ hybridization analysis of human papillomavirus in orofacial lesions using a consensus biotinylated probe. Am J Dermatopathol 15:256–9 Natale C, Giannini T, Lucchese A et al. (2000) Computer-assisted analysis of molecular mimicry between human papillomavirus 16 E7 oncoprotein and human protein sequences. Immunol Cell Biol 78:580–5 Nilges K, Hohn H, Pilch H et al. (2003) Human papillomavirus type 16 E7 peptide-directed CD8 þ T cells from patients with cervical cancer are cross-reactive with the coronavirus NS2 protein. J Virol 77:5464–74 Ruer JB, Pepin L, Gheit T et al. (2009) Detection of alpha- and beta-human papillomavirus (HPV) in cutaneous melanoma: a matched and controlled study using specific multiplex PCR combined with DNA microarray primer extension. Exp Dermatol 18:857–62 Santoro A, Majorana A, Bardellini E et al. (2004) Cytotoxic molecule expression and epithelial cell apoptosis in oral and cutaneous lichen planus. Am J Clin Pathol 121:758–64 Schmitt M, Bravo IG, Snijders PJ et al. (2006) Bead-based multiplex genotyping of human papillomaviruses. J Clin Microbiol 44:504–12 Schreurs MW, Scholten KB, Kueter EW et al. (2003) In vitro generation and life span extension of human papillomavirus type 16-specific, healthy donorderived CTL clones. J Immunol 171:2912–21 Selmi C, Leung PS, Sherr DH et al. (2012) Mechanisms of environmental influence on human autoimmunity: a National Institute of Environmental Health Sciences expert panel workshop. J Autoimmun 39:272–84
REFERENCES
Simark-Mattsson C, Bergenholtz G, Jontell M et al. (1994) T cell receptor V-gene usage in oral lichen planus; increased frequency of T cell receptors expressing V alpha 2 and V beta 3. Clin Exp Immunol 98:503–7
Becherel PA, Bussel A, Chosidow O et al. (1998) Extracorporeal photochemotherapy for chronic erosive lichen planus. Lancet 351:805
Sugerman PB, Satterwhite K, Bigby M (2000) Autocytotoxic T-cell clones in lichen planus. Br J Dermatol 142:449–56
Combita AL, Bravo MM, Touze A et al. (2002) Serologic response to human oncogenic papillomavirus types 16, 18, 31, 33, 39, 58 and 59 virus-like particles in colombian women with invasive cervical cancer. Int J Cancer 97:796–803
Syrjanen S, Lodi G, von Bultzingslowen I et al. (2011) Human papillomaviruses in oral carcinoma and oral potentially malignant disorders: a systematic review. Oral Dis 17 (Suppl 1):58–72
Edwards PC, Kelsch R (2002) Oral lichen planus: clinical presentation and management. J Can Dent Assoc 68:494–9 Fazilleau N, Bachelez H, Gougeon ML et al. (2007) Cutting edge: size and diversity of CD4 þ CD25high Foxp3 þ regulatory T cell repertoire in humans: evidence for similarities and partial overlapping with CD4 þ CD25- T cells. J Immunol 179:3412–6
Vaccarella S, Franceschi S, Clifford GM et al. (2010) Seroprevalence of antibodies against human papillomavirus (HPV) types 16 and 18 in four continents: the International Agency for Research on Cancer HPV Prevalence Surveys. Cancer Epidemiol Biomarkers Prev 19:2379–88 Yildirim B, Senguven B, Demir C (2011) Prevalence of herpes simplex, Epstein Barr and human papilloma viruses in oral lichen planus. Med Oral Patol Oral Cir Bucal 16:e170–4
Ferguson M, Heath A, Johnes S et al. (2006) Results of the first WHO international collaborative study on the standardization of the detection of antibodies to human papillomaviruses. Int J Cancer 118: 1508–14
Young SK, Min KW (1991) In situ DNA hybridization analysis of oral papillomas, leukoplakias, and carcinomas for human papillomavirus. Oral Surg Oral Med Oral Pathol 71:726–9
Gorsky M, Epstein JB (2011) Oral lichen planus: malignant transformation and human papilloma virus: a review of potential clinical implications. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 111:461–4
Zhou XJ, Sugerman PB, Savage NW et al. (2002) Intra-epithelial CD8 þ T cells and basement membrane disruption in oral lichen planus. J Oral Pathol Med 31:23–7
www.jidonline.org
7
Peripheral and Local Human Papillomavirus 16–Specific CD8 þ T-Cell Expansions Characterize Erosive Oral Lichen Planus Manuelle Viguier1,2,13, Herve´ Bachelez2,3,13, Be´atrice Poirier1, Je´re´my Kagan4,5,6,7, Maxime Battistella1,8, Franc¸ois Aubin9, Antoine Touze´10, Maryvonnick Carmagnat11, Camille France`s12, Marie-Lise Gougeon1,4 and Nicolas Fazilleau4,5,6,7,14 Erosive oral lichen planus (OLP) is a chronic, disabling mucocutaneous dysimmune rare disease characterized by mucosal inflammatory erosive lesions with pathological evidence for a marked CD8 þ cytotoxic T-lymphocyte (CTL) infiltration. However, the specificity of lesional CTL in OLP has never been analyzed. To investigate the molecular mechanisms underlying dysregulation of T-cell immune responses in patients with OLP, we studied the diversity and antigen specificity of the TCR expressed by CD8 þ T cells using dextramer staining, spectratyping, and TCR sequencing in 10 OLP patients undergoing extracorporeal photochemotherapy. Expansions of TCRVb3bearing CD8 þ T cells were found in peripheral blood and in lesional tissues of OLP patients. Spectratyping and sequencing studies identified specific clonotypes in each patient. These expansions were enriched with human papillomavirus 16 (HPV16)–specific CD8 þ T cells in HLA-A*0201 þ patients as shown by their immune recognition of the E711-20 immunodominant epitope. Under treatment with extracorporeal photochemotherapy, clonotypic CD8 þ T-cell expansions decreased in parallel with clinical remission. Altogether, these data establish a link between HPV infection and OLP pathogenesis by identifying a massive clonal expansion of CD8 þ T cells with increased frequency of HPV 16–specific CD8 þ T cells in OLP patients. Journal of Investigative Dermatology advance online publication, 23 October 2014; doi:10.1038/jid.2014.397
INTRODUCTION Lichen planus is a chronic inflammatory disease of unknown etiology involving the skin and/or mucous membranes, 1
Antiviral Immunity Biotherapy and Vaccine Unit, Institut Pasteur, Paris, France; 2Sorbonne Paris Cite´ Universite´ Paris Diderot, Paris, France; Universite´ Paris Descartes, Imagine Institute, Necker Hospital, Paris, France; 4 Centre de Physiopathologie de Toulouse Purpan, Toulouse, France; 5INSERM, U1043, Toulouse, France; 6CNRS, UMR5282, Toulouse, France; 7Universite´ Toulouse III Paul-Sabatier, Toulouse, France; 8Universite´ Paris Diderot, UMRS728, INSERM, Paris, France; 9Service de Dermatologie, Centre Hospitalier Universitaire, SFR FED 4234, EA 3181, Universite´ de Franche Comte´, Besanc¸on, France; 10UMR 1282 INRA-Universite´ Franc¸ois Rabelais, Tours, France; 11Laboratoire d’Immunologie, Hoˆpital Saint-Louis, Paris, France and 12 Service de Dermatologie-Allergologie, Hoˆpital Tenon, Universite´ Paris VI, AP-HP, Paris, France 3
13
The first two authors contributed equally to this work.
14
The last two authors contributed equally to this work.
Correspondence: Herve´ Bachelez, Sorbonne Paris Cite´ Universite´ Paris Diderot, Paris 75000, France or Marie-Lise Gougeon, Antiviral Immunity Biotherapy and Vaccine Unit, Institut Pasteur, Paris 75724, France or Nicolas Fazilleau, Centre de Physiopathologie de Toulouse Purpan, Toulouse 31300, France E-mail: [email protected] or [email protected] or [email protected] Abbreviations: CDR3, complementary determining region 3; CTL, cytotoxic T lymphocyte; ECP, extracorporeal photochemotherapy; HPV, human papillomavirus; OLP, oral lichen planus; PBMC, peripheral blood mononuclear cell Received 24 March 2014; revised 4 July 2014; accepted 16 August 2014; accepted article preview online 10 September 2014
& 2014 The Society for Investigative Dermatology
leading to a variety of clinical presentations (Le Cleach and Chosidow, 2012). Erosive oral lichen planus (OLP) is the most prevalent subtype, affecting 0.1 to 4% of the general population, and is characterized by a severely disabling course in its erosive form that often requires immunomodulatory systemic therapies such as oral steroids, immunosuppressants, and more recently extracorporeal photochemotherapy (ECP) (Becherel et al., 1998; Edwards and Kelsch, 2002; Guyot et al., 2007). Pathologically, OLP is characterized by a predominant infiltration of immune cells, mainly CD8 þ cytotoxic T lymphocytes (CTLs), associated with apoptosis of epithelial cells and disruption of the basement membrane zone (Zhou et al., 2002). Similarly, the presence in OLP lesions of CTL in the vicinity of damaged epithelial cells supports the hypothesis that a dysregulated CTL response of unknown antigen specificity has a key role in the pathogenesis of the disease (Santoro et al., 2004). However, the nature of these antigens remains controversial. Indeed, although autoreactivity of lesional skin–derived CTL against epithelial cells has been suggested in OLP patients (Sugerman et al., 2000; Gorsky and Epstein, 2011; Syrjanen et al., 2011), other studies identified the presence of DNA from several human papillomavirus (HPV) subtypes in mucosal lesions of OLP, raising the issue of the self versus non-self nature of the target antigens in OLP (Jontell et al., 1990; Young and Min, 1991; Miller et al., 1993; Lodi et al., 2005; Yildirim et al., 2011). So far, the lack of any study using accurate tools to address the specificity of lesional CTL in OLP www.jidonline.org
1
M Viguier et al.
HPV-Specific CD8 þ T Cells in Lichen Planus
the peripheral blood CD8 þ T lymphocytes (data not shown, 17.27±4.51%). The quantitative PCR products were further used for the amplification of dye-labeled oligonucleotides specific for Cb and allowed to assess the complementary determining region 3 (CDR3) length distribution. For each TCRVb gene segment tested and for the TCRVb3 gene segment from CD4 þ and CD8 þ T cells from healthy donors or from CD4 þ T cells isolated from OLP patients (Figure 1b), CDR3b length distribution profiles displayed Gaussian-like curves, the hallmark of a polyclonal T-cell repertoire. In contrast, CDR3b distribution profiles of the TCRVb3 gene segment were altered with single expansions in blood CD8 þ T cells from all OLP patients, as observed in OLP no. 5 for a CDR3b size of 11 amino acids (Figure 1b). Therefore, we determined the Vb-Jb rearrangement corresponding to the observed clonal expansion and further cloned and sequenced it. As shown in Figure 1c, one single CDR3b sequence was found in each patient. Furthermore, we revealed that similar TCRVb3 þ clonal expansions were also present in tissue lesions of three
precludes any conclusion about mechanisms underlying this dysregulation. In this study, using dextramer, spectratyping, and sequencing in 10 patients with severe erosive OLP, we investigated the antigen specificity of the CD8 þ T-cell expansions. RESULTS AND DISCUSSION TCRVb3 þ clonal expansions alter CD8 þ T-cell repertoires of OLP patients at flare
To understand the molecular events that govern the physiopathology of OLP, we investigated the T-cell repertoire of 10 patients with OLP. At flare, the usage of each TCRVb gene segment, as estimated by quantitative real-time reverse-transcriptase–PCR in CD4 þ and CD8 þ peripheral blood T cells, revealed striking predominance of the TCRVb3 gene segment in CD8 þ T cells (26.49±4.81%), whereas it was used at a much lower level in the circulating CD4 þ counterparts (11.49±2.24%) or in both isolated CD4 þ and CD8 þ T cells from healthy donors (Figure 1a). Results from quantitative real-time reverse-transcriptase–PCR were confirmed by flow cytometry, showing predominance of TCRVb3 expression in
Vβ3 usage in PBMC
**
Vβ3-Cβ
PBMC
*
CD4
CD8
HD 1 25
11
11
11
11
OLP 5 CD4
CD8
CD4
HD OLP
Jβ
1 2 3 4 5 6 7 8 9 10
1.5 2.7 2.7 2.7 2.2 2.1 1.2 2.7 2.7 2.2
OLP CDR3β (nucleotidic sequence)
CDR3β (AA)
TGTGCCAGCAGTTTTGTCATTGGGGGGGTTAGCAATCAGCCCCAG TGTGCCAGCAGTGAACCGGGACAGGGTTTTGGTGAGCAG TGTGCCAGCAGTATAGGGCTCCACTACGAGCAG TGTGCCAGCAGTTTAGGTGTGCACTACGAGCAG TGTGCCAGCAGTCAAGGAGTCTTCCTGGGGGCCGGGGAGCTG TGTGCCAGCAGTTTATTGGAGGGACTAGCGGGAGTGGATGAGCAG TGTGCCAGCAGTTATATAGGCCTACAGAGTAAAGCTAACTATGGCTAC TGTGCCAGCAGTGAACCGGGACAGGGTTTTGGTGAGCAG TGTGCCAGCAGTTTCCAACCAGGGGGGCAG TGTGCCAGCAGTTTAACCACCGGGGAGCTG
In situ
5 SEPGQGFGEQ In Vβ3+ (%)
Vβ3-Jβ2.7
CD8
2.5
OLP 2 þ
10
5 Vβ3-Jβ2.7
0
10
SFVIGGVSNPQ SEPGQGFGEQ SIGLHYEQ SLGVHYEQ SQGVFLGAGEL SLLEGLAGVDEQ SYIGLQSKANYGY SEPGQGFGEQ SFQPGGQ SLTTGEL
Vβ3-Jβ2.2 2.5
0
8 OLP 4 þ
SQGVFLGAGEL In Vβ3+ (%)
0
SLGVHYEQ In Vβ3+ (%)
Frequency (%)
50
5
0
11 OLP 5
Figure 1. TCRVb3 clonal expansions skewed peripheral and lesional CD8 T-cell repertoires of oral lichen planus (OLP) patients at flare. (a) Reversetranscriptase–PCR (RT–PCR) amplifications of the TCRVb3 usage in CD4 þ and CD8 þ T cells from the blood of healthy donors (HDs) or OLP patients. PBMC, peripheral blood mononuclear cell. (b) CDR3b length distributions for CD4 þ and CD8 þ Vb3-Cb rearrangements from a representative HD and a representative OLP patient. The peak of the 11th codon is marked on the abscissa axis. (c) Table of CDR3 clonotypes identified in each OLP patient (nucleotidic and amino-acid (AA) sequences). (d) CDR3b length distributions (left) and clonotype distribution estimated by sequencing (right) in lichen lesions in three representative OLP patients (OLP nos. 2, 4, and 5). *Po0.05, **Po0.01.
2
Journal of Investigative Dermatology (2014), Volume 00
M Viguier et al.
HPV-Specific CD8 þ T Cells in Lichen Planus
OLP patients (Figure 1d). Overall, neither CDR3b length nor consensus sequence was shared across all OLP patients. However, all detected CD8 þ clonal expansions were using the TCRVb3 gene segment, even in patients not sharing common class I HLA alleles (Supplementary Table S1 online). Interestingly, 5 out of the 10 clonotypic TCRs were composed of the Vb-Jb2.7 rearrangement, and one clonotypic sequence was found in two different patients (OLP nos. 2 and 8; Figure 1c). Of note, flow cytometric characterization of TCRVb3 þ CD8 þ T cells in peripheral blood mononuclear cells (PBMCs) confirmed their activated status as shown by surface expression of HLA-DR (data not shown, 20.99± 4.52%, n ¼ 7), CD38 (24.85±6.18%, n ¼ 4), and perforin content (15.2±4.54%, n ¼ 7). Altogether, these data show that clonotypic TCRVb3 þ CD8 þ T-cell expansions are detected in situ and in the blood of OLP patients at flare. These sets of data extend previous observations reported in two studies, one reporting the increased presence of TCRVb3and TCRVa2-positive cells in the oral mucosal T-cell infiltrate of 7 OLP patients (Simark-Mattsson et al., 1994), and the other performed in 12 OLP patients showing that CD8 þ T cells had skewed TCR repertoire toward the usage of a restricted set of TCRVb segments including TCRVb3 (Gotoh et al., 2008). Tracking of CD8 þ TCRVb3 þ clonal expansions following disease activity in OLP patients
We next reasoned that the CD8 þ clonal expansions could actively participate in the physiopathology of OLP and tested
this hypothesis by performing a longitudinal study of the TCR repertoire in all OLP patients. PBMCs were isolated at different time points corresponding to distinct clinical status: flare (before starting ECP), partial remission, complete remission following ECP, and relapse after ECP withdrawal. For all collected samples, TCRVb usage and CDR3b length distribution were performed as well as sequencing of Vb-Jb rearrangements corresponding to the clonal expansions at flare (Figure 1c). This allowed a follow-up of peaks corresponding to TCRVb3 þ clonal expansions in the T-cell compartment (Figure 2 and Supplementary Figure S1 online). For three patients (OLP nos. 5, 6 (Figure 2) and 8 (Supplementary Figure S1 online)), we observed a disappearance of the peaks corresponding to the TCRVb3 þ clonal expansions that strictly correlated with a clinical improvement (either partial remission or complete remission). However, for the seven remaining OLP patients (as an example OLP no. 4, Figure 2), CDR3 distribution profiles were all disturbed irrespective of the clinical status of the patients. Nevertheless, we found a strong correlation between the clonotype usage and the clinical status of OLP patients for 8 out of 10 patients. A pronounced decrease or disappearance of the frequency of TCRVb3 þ clonotype was observed in OLP patients 1, 4, 5, 6, and 8, whereas a slight decrease was found in OLP cases 2, 3, and 7 (Figure 2 and Supplementary Figure S1 online). Moreover, the clonotype identified at initial flare was also found at high frequency at the time of clinical relapse (OLP nos. 4, and 5 (Figure 2) and nos. 9 and 10 (Supplementary Figure S1 online)). Overall, these data support that clinical remission 80 SLGVHYEQ In Vβ3+ (%)
Vβ3-Jβ2.7
OLP 4 8 PR 1
8 PR 2
8 Relapse
Vβ3-Jβ2.2
OLP 5 11 PR 1
11 Relapse
Vβ3-Jβ2.1
OLP 6
12 Flare
12 CR
11 PR 2
12 PR
Flare
PR 1
PR 2 Relapse
Flare
0 PR 1 Relapse PR 2
100 75 50 25 0
SLLEGLAGVDEQ In Vβ3+ (%)
11 Flare
0
SQGVFLGAGEL In Vβ3+ (%)
8 Flare
40
15 10 5 0
Flare
0 CR
0 PR
Figure 2. Decrease or disappearance of TCRVb3 þ clonal expansions over the clinical course of the disease in oral lichen planus (OLP) patients. CDR3b length distributions (left column) for different CD8 þ Vb3-Jb rearrangements over the clinical course of the disease following extracorporeal photochemotherapy (ECP) and clonotype distribution estimated by sequencing (right column) for three different representative OLP patients (OLP nos. 4, 5, and 6). CR, complete remission; PR, partial remission.
www.jidonline.org
3
M Viguier et al.
HPV-Specific CD8 þ T Cells in Lichen Planus
obtained with ECP in OLP is associated in most cases with a decrease or disappearance of TCRVb3 clonotypic CD8 þ T-cell expansions, suggesting the pathogenic relevance of these clonotypic expansions. CD8 þ T cells expressing clonal TCRVb3 þ in OLP patients are specific for HPV16
Lesional CTLs from OLP patients are suspected to recognize an antigen associated with major histocompatibility complex class I molecule on lesional keratinocytes (Sugerman et al., 2000). Furthermore, several studies have shown the presence of several HPV subtypes in mucosal OLP lesions (Jontell et al., 1990; Young and Min, 1991; Miller et al., 1993; Lodi et al., 2005; Yildirim et al., 2011). Detection of current or past HPV infection could be performed in 6 out of our 10 patients either on tissue biopsies or by serology. Three out of six OLP tested patients were found HPV positive and, among them, HPV16 DNA and/or HPV16-specific IgG were found in 2 out of the 3 HPV-positive patients (OLP nos. 2 and 5). To test whether TCRVb3 þ clonal expansions might exhibit HPV specificity, we stained PBMCs from the five HLA-A*0201 þ OLP patients of the cohort with peptide-containing major histocompatibility complex class I dextramer. For all these patients, a distinct population of CD8 þ T cells expressing a TCRVb3 specific for the E711-20 immunodominant peptide of HPV16 was detected, as shown for OLP no. 2 at flare or OLP no. 4 under ECP (Figure 3a; mean percentage of CD8 þ Dext-HPV þ in TCRVb3 þ CD3 þ cells: 2.62±1.7%, n ¼ 5). As expected, almost no TCRVb3 þ T cells were stained with HLA-A2/HIVp17 Gag dextramer in these HIV-negative patients (Figure 3a; CD8 þ Dext-HIV þ : 0.64±0.3%, n ¼ 5). We then sorted out HPV-specific CD8 þ TCRVb3 þ cells as well as their CD8 þ TCRVb3 þ Dext-HPV- counterparts from OLP nos. 3, 4, and 7. Spectratyping analysis and CDR3b sequencing were performed, revealing a striking enrichment of Dext-HPV þ populations with the clonotype in contrast with Dext-HPV- subset that showed no enrichment (Figure 3b). In the last series of experiments, we performed in situ immunostaining on OLP lesion from OLP no. 2. The presence of cells expressing CD3, CD8, and the cytotoxic marker TIA1 (T-cell-restricted intracellular antigen 1) were detected (Figure 4a), some of these bearing TCR specific for HLA-A2/HPV E711-20 peptide, whereas staining with the irrelevant Dext-HIV yielded negative results in the same patient (Figure 4b). Overall, these data demonstrate that a notable proportion of clonal CD8 þ blood T cells of OLP patients are HPV16 specific, also infiltrating mucosal and/or skin lesions, and decreasing or disappearing following clinical remission, suggesting their pathogenic role in OLP. These results are reminiscent of results from previous in vitro studies showing the capacity of this E711-20 HPV16 epitope to favor the stimulation and expansion of CTL clones expressing TCRVb3, without evidence of HLA class I restriction (Nilges et al., 2003; Schreurs et al., 2003). These sets of data also raise the HPV infection status in OLP patients. Similarly, a large set of published HPV detection studies support a pathogenic contribution of some HPV subtypes in OLP (Gorsky and Epstein, 2011; Syrjanen et al., 2011). Notably, detection of 4
Journal of Investigative Dermatology (2014), Volume 00
viral DNA mainly derived from HPV6/11/16/31/33 has been reported in mucosal oral and/or genital lesions from OLP patients in several studies, providing a rationale for their role as putative antigenic stimuli (Lodi et al., 2005; Mattila et al., 2012). This latter hypothesis has been reinforced by a recent exhaustive systematic review of HPV detection in 39 selected cross-sectional studies, showing that HPV DNA is significantly more frequently detected in OLP lesions compared with controls (Syrjanen et al., 2011). Indeed, the positive detection of previous or current infection with HPV16 in 2/6 of our patients, including 1 patient who showed clear expansion of HPV16 E711-20–specific cells in blood and lesional mucosa, is a strong support for the contribution of HPV as an antigenic stimulus of CTL expansion that characterizes severe erosive OLP. Finally, the presence of HPV16 E711-20–specific CD8 þ T cells in the vicinity of dying epithelial cells in a studied patient correlates with their potential involvement in tissue damage of OLP, even though this latter observation does not rule out the deleterious role of other subtypes than HPV16 that were not addressed in our patients. As some of our patients did not show either the presence of HPV16 DNA or the serological evidence of previous HPV16 infection, it remains to be determined whether other HPV subtypes or even alternative viral agents would be able to trigger similar CTL expansions in OLP (Lodi et al., 2005). Nevertheless, HPV serology is a poor marker of current infection and HPV DNA detection can be rendered difficult, as most HPV infections are transient (Ho et al., 1998; Vaccarella et al., 2010). As most adult individuals are infected at some point with one or several HPV subtypes, the present results also raise the mechanisms underlying dysimmunity in OLP patients. Supporting the autoimmune hypothesis and its possible link with HPV is the previously established presence among lesional T lymphocytes of two OLP patients of CTL clones reacting toward autologuous keratinocytes immortalized with HPV16 E6 and E7 (Sugerman et al., 2000). It is worth noticing that a causal role for HPV infection in OLP does not necessarily require constant tissue viral replication, as an initial viral stimulation might be sufficient to trigger the founder immune response, with following autoimmune T-cell expansions related to molecular mimicry, unsequestration of masked self epitopes, or both (Selmi et al., 2012). As such, it might explain why we were able to detect antibodies to HPV16 in the sera of only two patients of our cohort. In line with the former of these hypotheses, molecular mimicry between HPV16 E7 protein and human self has been revealed by computer-based analyses, providing a rational basis for further investigations of OLP lesional CTL regarding their immunoreactivity toward HPV versus self-candidate antigens (Natale et al., 2000). However, alternative mechanisms such as alterations of the TCRVb3 þ /HLA-peptide synapse by some viral products cannot be ruled out. Finally, it also remains to be determined whether HPV and/or other viral stimuli are involved in other forms of lichen planus. Our data also suggest that prophylactic HPV16 vaccine in addition to preventing cervical dysplasia and cancer may also protect against OLP.
M Viguier et al.
HPV-Specific CD8 þ T Cells in Lichen Planus
OLP 2
OLP 4
CD3+ Vβ3+
HIV-1 SLYNTVATL
HPV YMLDLQPETT
HPV YMLDLQPETT
0.81%
2.86%
2.34%
Dextramer HLA-A*0201
0.73%
CD3+ Vβ3+
HIV-1 SLYNTVATL
CD8 80
Isolated T cells HPV+
HPV– OLP 7
SYIGLQSKANYGY In Vβ3+ (%)
Unsorted T cells
60 40 20 0
13
13
13
0 Unsorted HPV– HPV+ T cells Isolated T cells
Isolated T cells HPV–
HPV+
OLP 3
8
SIGLHYEQ In Vβ3+ (%)
50 Unsorted T cells
25
0
8
8
Unsorted HPV– HPV+ T cells Isolated T cells
Isolated T cells HPV–
HPV+
OLP 4
SLGVHYEQ In Vβ3+ (%)
100 Unsorted T cells
75 50 25 0
8
8
8
Unsorted HPV– HPV+ T cells Isolated T cells
Figure 3. A subpopulation of peripheral CD8 þ T cells from oral lichen planus (OLP) patients is human papillomavirus 16 (HPV16) specific and is enriched in TCRVb3 þ clonal expansions. (a) Flow cytometric detection using dextramer of HIV- and HPV-specific CD8 þ T cells in CD3 þ Vb3 þ cells from the blood of two representative HLA-A*0201 þ OLP patients (OLP nos. 2 and 4). (b) CDR3b length distributions for Vb3-Cb rearrangements on unsorted T cells or isolated T cells with HPV-Dextramer (left column) and clonotype distribution estimated by CDR3 sequencing (right column) for three different representative HLA-A*0201 þ OLP patients (OLP nos. 7, 3, and 4).
MATERIALS AND METHODS Patients and healthy donors Ten patients with OLP treated with ECP were enrolled. They consisted of 8 women and 2 men (mean age: 56.5 years (range: 25–78)) (Supplementary Table S1 online). The characteristics of the eight patients have been previously reported (Guyot et al., 2007). Previous treatment with glucocorticoids and immunosuppressive agents was stopped 4 weeks before ECP initiation. All patients reached remission under ECP, either complete remission or partial remission (defined by regression of lesions of at least 50%) (Guyot et al., 2007). Clinical
statuses were determined at the time of blood collection. The Institutional Review Board (Comite´ de Protection des Personnes Ile de France IV, 2009/10NI) approved the study that was qualified as noninterventional. The patients were fully informed and had no objection to participate in the study. Written patient consent was not required because French law considers that it is not mandatory for noninterventional biological collection for immunological and molecular research, with the exception of genetic studies on germinal DNA. Healthy donors were blood donors from the Etablissement Franc¸ais du Sang. www.jidonline.org
5
M Viguier et al.
HPV-Specific CD8 þ T Cells in Lichen Planus
CD3
CD8
Dextramer HLA-A*0201 HIV-1 SLYNTVATL
TIA1
Dextramer HLA-A*0201 HPV YMLDLQPETT
Figure 4. Human papillomavirus 16 (HPV16)–specific CD8 þ T cells infiltrate lichen lesions. (a) The in situ immunostainings of CD3, CD8 (original magnification 200, scale bars ¼ 100 mm), and TIA-1 (T-cell-restricted intracellular antigen 1) (original magnification 400, scale bar ¼ 50 mm) on a biopsy taken from mucosal lesions in oral lichen planus (OLP) patient no. 2. (b) The in situ immunostainings using either dextramer HIV-1 (left) or dextramer HPV (right) ( 630, scale bars ¼ 25 mm; inset original magnification 1,000, scale bars ¼ 10 mm).
Preparation of PBMCs Blood was collected at baseline before any ECP treatment (‘‘flare’’) and at several time points. PBMCs were prepared by Ficoll-Hypaque density gradient centrifugation and cryopreserved in liquid nitrogen in 8% dimethylsulfoxide, 42% fetal calf serum, and 50% RPMI-1640 medium (Invitrogen, Carlsbad, CA).
(Invitrogen Life Technologies, Carlsbad, CA). Quantitative PCR amplifications were performed with TaqMan Universal PCR Master Mix (Applied Biosystems, Carlsbad, CA), TCRVb gene segment– specific oligonucleotides, TCRCb-specific antisense primer, and a fluorescent probe specific for the TCRCb gene segment (Fazilleau et al., 2007).
Antibodies and flow cytometric analysis
Cloning and sequencing of CDR3b rearrangements
Anti-CD3 FITC (Clone BW264/56, Miltenyi Biotec, Bergisch Gladbech, Germany), anti-CD8-APC (SK1, BD Pharmingen, Franklin Lakes, NJ), Dextramer-HPV E711-20-PE (Dext-HPV PE; HLA-A*0201; YMLDLQPETT, Immudex, Copenhagen, Denmark), Dextramer-HIV-1 P17 Gag 77-85-PE (Dext-HIV-PE; HLA-A*0201, Bayport, MN; SLYNTVATL, Immudex), and anti-TCR-Vb3-FITC (JOVI-3, Ancell, Bayport, MN) were used. Cells were incubated with dextramer for 10 minutes at room temperature, and then antibodies were added for 30 minutes at 4 1C. Stained cells were washed with phosphatebuffered saline/0.5% BSA/0.1% NaN3 and analyzed using a FACSCalibur flow cytometer (Becton Dickinson,Franklin Lakes, NJ).
Vb-Jb PCR products were cloned in pCR 4Blunt TOPO vector (Invitrogen Life Technologies). Sequencing reactions were performed using the ABI PRISM Big Dye Terminator Reaction Kit (Applied Biosystems). Reaction products were analyzed on an ABI 3130XL 16 capillaries (Applied Biosystems).
Purification of T-lymphocyte subsets Dead cells were removed using the MACS Dead Cell Removal kit (Miltenyi Biotec). T cells were isolated using anti-CD8 or anti-CD4 antibody-coated immunomagnetic beads and stabilized in PrepProtect buffer (Miltenyi Biotec). Cell purities exceeded 85%. CD3 þ CD8 þ TCRVb3 þ Dext-HPV þ or Dext-HPV- populations were isolated using a MoFlo Astrios cytometer cell sorter (Beckman Coulter, Villepinte, France).
TCR repertoire analysis RNA was extracted using RNAeasy kit (Qiagen, Venlo, The Netherlands) either from sorted PBMCs or from total skin or mucosal OLP lesions and reverse-transcribed into complementary DNA using Superscript II 6
Journal of Investigative Dermatology (2014), Volume 00
Detection of HPV subtypes Tissue samples were obtained either atraumatically using a cytobrush (DNAPAP Cervical Sampler; Digene, Qiagen) secondarily placed in Cervical Specimen Transport Medium (Digene, Qiagen) or by tissue biopsy secondarily frozen. DNA extraction was performed using the QIAGEN BioRobot EZ1 with the EZ1 DNA Tissue Kit (Qiagen). Human papillomavirus detection was performed using two highly sensitive and specific assays combining a specific multiplex PCR with DNA microarray primer extension as described previously (Schmitt et al., 2006; Ruer et al., 2009). Multiplex PCR reactions were performed for the amplification of a-HPV, b-HPV, g-HPV, and mHPV types, respectively. The assays were combined in one single chip to detect 27 HPV types from genus a, 25 cutaneous HPV types from genus b, 16 cutaneous HPV types from genus g, and one from genus m. Following PCR amplification, 10 ml of each reaction mixture was analyzed using the Luminex technology (Luminex, Austin, TX). The detection limit was determined using the cutoff of 10 median reporter fluorescence intensity and varied between 100 and 800 pg of PCR products (Schmitt et al., 2006). For each sample, three multiplex
M Viguier et al.
HPV-Specific CD8 þ T Cells in Lichen Planus
PCR reactions were performed. Search for serum IgG antibodies against HPV16 major capsid protein L1 was performed using ELISA (Combita et al., 2002). Briefly, serum samples were tested in triplicate in a 1:20 dilution, and virus like particle-bound antibodies were detected using anti-human IgG conjugated to peroxidase diluted at 1:20,000 (Southern Biotech, Birmingham, AL). Cutoff value for positivity based on sample from young children (mean reactivity plus 3 SD) was 0.2. Previously tested samples were retested on the assay plate to confirm ensibility and consistency of the assay (Ferguson et al., 2006).
In situ immunostaining Acetone-fixed cryosections were incubated with dextramer for 75 minutes at room temperature and mounted with VECTASHIELD Mouting Medium with 4’,6-diamidino-2-phenylindole (Vector Laboratories, Burlingame, CA). Slides were studied using an Axiovert 200M microscope with MRm camera (Zeiss, Oberbochen, Germany).
Statistical analysis Unpaired t-tests were calculated using Prism software (Irvine, CA). CONFLICT OF INTEREST The authors state no conflict of interest, except AT who was a speaker at GSK symposia and for whom GSK and Sanofi Pasteur provided meeting and travel fees.
ACKNOWLEDGMENTS We are indebted to Laure Buisson for her help in sequencing analyses (UDEAR, CNRS, France), Pierre-Henri Comme`re for cell sorting (Flow Cytometry Facility, Pasteur Institute), Nathalie Parquet (Saint-Louis Hospital, Paris, France) and Claire Rabian (Saint-Louis Hospital, Paris, France) for their help in cell collection and cell preparation, and to Tarik Gheit (International Agency for Research in Cancer, Lyon, France) for HPV genotyping. This work was supported by Institut Pasteur, the Socie´te´ Franc¸aise de Dermatologie, AVENIR INSERM (NF), Association pour la Recherche sur le Cancer (ARC AO 2009, NF), la Ligue Nationale Contre le Cancer (NF), Conseil Re´gional Midi-Pyre´ne´es (NF), Institut National contre le Cancer (PLBIO10-195 and INCA-6530, NF), Collery Grant from Acade´mie Nationale de Me´decine (MB), and International Re-integration Grant Marie Curie. SUPPLEMENTARY MATERIAL Supplementary material is linked to the online version of the paper at http:// www.nature.com/jid
Gotoh A, Hamada Y, Shiobara N et al. (2008) Skew in T cell receptor usage with polyclonal expansion in lesions of oral lichen planus without hepatitis C virus infection. Clin Exp Immunol 154:192–201 Guyot AD, Farhi D, Ingen-Housz-Oro S et al. (2007) Treatment of refractory erosive oral lichen planus with extracorporeal photochemotherapy: 12 cases. Br J Dermatol 156:553–6 Ho GY, Bierman R, Beardsley L et al. (1998) Natural history of cervicovaginal papillomavirus infection in young women. N Engl J Med 338:423–8 Jontell M, Watts S, Wallstrom M et al. (1990) Human papilloma virus in erosive oral lichen planus. J Oral Pathol Med 19:273–7 Le Cleach L, Chosidow O (2012) Clinical practice. Lichen planus. N Engl J Med 366:723–32 Lodi G, Scully C, Carrozzo M et al. (2005) Current controversies in oral lichen planus: report of an international consensus meeting. Part 1. Viral infections and etiopathogenesis. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 100:40–51 Mattila R, Rautava J, Syrjanen S (2012) Human papillomavirus in oral atrophic lichen planus lesions. Oral Oncol 48:980–4 Miller CS, White DK, Royse DD (1993) In situ hybridization analysis of human papillomavirus in orofacial lesions using a consensus biotinylated probe. Am J Dermatopathol 15:256–9 Natale C, Giannini T, Lucchese A et al. (2000) Computer-assisted analysis of molecular mimicry between human papillomavirus 16 E7 oncoprotein and human protein sequences. Immunol Cell Biol 78:580–5 Nilges K, Hohn H, Pilch H et al. (2003) Human papillomavirus type 16 E7 peptide-directed CD8 þ T cells from patients with cervical cancer are cross-reactive with the coronavirus NS2 protein. J Virol 77:5464–74 Ruer JB, Pepin L, Gheit T et al. (2009) Detection of alpha- and beta-human papillomavirus (HPV) in cutaneous melanoma: a matched and controlled study using specific multiplex PCR combined with DNA microarray primer extension. Exp Dermatol 18:857–62 Santoro A, Majorana A, Bardellini E et al. (2004) Cytotoxic molecule expression and epithelial cell apoptosis in oral and cutaneous lichen planus. Am J Clin Pathol 121:758–64 Schmitt M, Bravo IG, Snijders PJ et al. (2006) Bead-based multiplex genotyping of human papillomaviruses. J Clin Microbiol 44:504–12 Schreurs MW, Scholten KB, Kueter EW et al. (2003) In vitro generation and life span extension of human papillomavirus type 16-specific, healthy donorderived CTL clones. J Immunol 171:2912–21 Selmi C, Leung PS, Sherr DH et al. (2012) Mechanisms of environmental influence on human autoimmunity: a National Institute of Environmental Health Sciences expert panel workshop. J Autoimmun 39:272–84
REFERENCES
Simark-Mattsson C, Bergenholtz G, Jontell M et al. (1994) T cell receptor V-gene usage in oral lichen planus; increased frequency of T cell receptors expressing V alpha 2 and V beta 3. Clin Exp Immunol 98:503–7
Becherel PA, Bussel A, Chosidow O et al. (1998) Extracorporeal photochemotherapy for chronic erosive lichen planus. Lancet 351:805
Sugerman PB, Satterwhite K, Bigby M (2000) Autocytotoxic T-cell clones in lichen planus. Br J Dermatol 142:449–56
Combita AL, Bravo MM, Touze A et al. (2002) Serologic response to human oncogenic papillomavirus types 16, 18, 31, 33, 39, 58 and 59 virus-like particles in colombian women with invasive cervical cancer. Int J Cancer 97:796–803
Syrjanen S, Lodi G, von Bultzingslowen I et al. (2011) Human papillomaviruses in oral carcinoma and oral potentially malignant disorders: a systematic review. Oral Dis 17 (Suppl 1):58–72
Edwards PC, Kelsch R (2002) Oral lichen planus: clinical presentation and management. J Can Dent Assoc 68:494–9 Fazilleau N, Bachelez H, Gougeon ML et al. (2007) Cutting edge: size and diversity of CD4 þ CD25high Foxp3 þ regulatory T cell repertoire in humans: evidence for similarities and partial overlapping with CD4 þ CD25- T cells. J Immunol 179:3412–6
Vaccarella S, Franceschi S, Clifford GM et al. (2010) Seroprevalence of antibodies against human papillomavirus (HPV) types 16 and 18 in four continents: the International Agency for Research on Cancer HPV Prevalence Surveys. Cancer Epidemiol Biomarkers Prev 19:2379–88 Yildirim B, Senguven B, Demir C (2011) Prevalence of herpes simplex, Epstein Barr and human papilloma viruses in oral lichen planus. Med Oral Patol Oral Cir Bucal 16:e170–4
Ferguson M, Heath A, Johnes S et al. (2006) Results of the first WHO international collaborative study on the standardization of the detection of antibodies to human papillomaviruses. Int J Cancer 118: 1508–14
Young SK, Min KW (1991) In situ DNA hybridization analysis of oral papillomas, leukoplakias, and carcinomas for human papillomavirus. Oral Surg Oral Med Oral Pathol 71:726–9
Gorsky M, Epstein JB (2011) Oral lichen planus: malignant transformation and human papilloma virus: a review of potential clinical implications. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 111:461–4
Zhou XJ, Sugerman PB, Savage NW et al. (2002) Intra-epithelial CD8 þ T cells and basement membrane disruption in oral lichen planus. J Oral Pathol Med 31:23–7
www.jidonline.org
7
