Activation of myeloid dendritic cells, effector cells and regulatory T cells in lichen planus

The current study enrolled patients with cutaneous LP (n = 18; 3 males, 15 females) from the Dermatological Outpatient Clinic of the Hospital das Clínicas de São Paulo (HC-FMUSP) as well as healthy individuals as controls (n = 22; 5 males, 17 females). The majority cohort of patients with LP had the cutaneous form of LP and association with oral form was observed in 3 patients. Patients with other forms of LP, such as lichen planopilaris or drug-induced LP, as well as patients with LP associated with autoimmune diseases and other skin diseases were not included. The majority of the patients had at least two affected limbs, with papules involving up to 30 % of the trunk. None of the patients took any type of topical or systemic corticosteroids, retinoids or immunosuppressants for 1 month prior to blood collection. A serological screening for hepatitis C was performed. The median age was 43.9 ± 13.5 years (range: 22–65 years) for the patients with LP and 40.5 ± 14.3 years (range: 20–71 years) for the healthy individuals. All the subjects provided written informed consent under the approval of the São Paulo University Institutional Use Committee (CAPPesq no. 0709/11).

To analyse Treg populations, PBMCs were isolated from heparinized venous blood by Ficoll-Hypaque gradient centrifugation (GE Healthcare Bio-Sciences AB, Uppsala, Sweden) and diluted in RPMI medium supplemented with 10 % AB human serum (Sigma, St. Louis, MO, USA). The PBMCs were incubated with antibodies against the following proteins at 4 °C for 30 min: CD3 (S4.1, Qdot, Invitrogen, Carlsbad, CA, USA), CD4 (RPA T4, Horizon V500), CD8 (RPA T8, Alexa Fluor 700), CD45RA (HI 100, APC-H7), CCR7 (3D12, Alexa Fluor 647), and CD127 (HIL-7R-M21, RPE-Cy7). All antibodies were from BD Biosciences (San Jose, CA, USA). The cells were then fixed, permeabilized and incubated with an anti-Foxp3 antibody (PCH101, e-Bioscience, San Diego, CA, USA) at 4 °C for 30 min. The Tregs were characterized as CD3⁺CD4⁺CD25⁺CD127^low/−Foxp3⁺ or CD8⁺CD25⁺CD127^low/−Foxp3⁺ and as thymus-derived Tregs (tTregs; CD45RA⁺Foxp3^low) or peripherally derived Tregs (pTregs; CD45RA⁻Foxp3^high). A total of 400,000 events were acquired using a flow cytometer (LSR Fortessa, BD Biosciences, USA) with FACS-Diva software (BD Bioscience). The data were analysed using FlowJo software, version 9.4.11 (Tree Star, Inc., Ashland, OR, USA).

Cultures of PBMCs (2.0 × 10⁶ cells/500 µL) were incubated in 48-well plates (Costar, Cambridge, MA, USA) in RPMI 1640 medium with ligands for TLR4 (lipopolysaccharide, 2 µg/mL), TLR7 (imiquimod, 2.5 µg/mL), TLR7/TLR8 (CL097, 5 µg/mL), or TLR9 (oligodeoxynucleotide CpG, 4 µM/mL) for 16 h or SEB (enterotoxin B of Staphylococcus aureus, 1 μg/mL) as a positive control for 6 h at 37 °C under 5 % CO₂. The stimulus concentrations and culture times were previously determined using a dose–response curve (Cardoso EC, 2013). All the ligands were obtained from Invivogen (San Diego, CA, USA). Brefeldin A (10 µg/mL, Sigma) was added to the cells 4 h after the beginning of the culture. mDCs were characterized as Lin⁻CD11c⁺HLA-DR⁺, and pDCs were characterized as Lin⁻CD123⁺HLA-DR⁺. After the Brefeldin A incubation, the cells were washed and incubated with human IgG for 10 min followed by staining with LIVE/DEAD (Invitrogen), fixation with Cytofix/Cytoperm solution (BD Biosciences) for 20 min and a wash in Perm/Wash solution. The cells were stained with antibodies against the following proteins: Lineage 1 cocktail (SK-7, Lin1, a mix of CD3, CD14, CD16, CD19, CD20 and CD56) (FITC), HLA-DR (G46-6, Horizon V500), CD123 (7G3, PERCP Cy5.5), CD11c (B-ly6, Alexa fluor 700), IFN-α (7N4-1, Horizon V500) and TNF-α (6401.111, PE). For monofunctional and polyfunctional T cell analyses, PBMCs were stained with LIVE/DEAD (Invitrogen, viability marker), fixed and permeabilized with Cytofix/Cytoperm solution (BD Biosciences) and then stained with antibodies against CD3 (SP34-2, BV605), CD4 (RPA-34, Horizon V500), CD8 (RPA-T8, PERCP Cy5.5), TNF (MAb11, PECy7), IL-10 (JES3-19F1, APC), CD38 (HIT2, Alexa Fluor 700) and IFN-γ (B27, Horizon V450) from BD Biosciences and antibodies against IL-22 (22URTI, PE) and IL-17 (N49-653, Alexa Fluor 488) from eBioscience. A total of 500,000 events were acquired with a flow cytometer (LSRFortessa, BD) and analysed with FlowJo software. The fluorescence minus one (FMO) protocol was used for all analyses to determine the gates. Boolean gate arrays were then created using FlowJo software. This analysis determined the frequency of each cytokine based upon all possible combinations of cytokines. To analyse the polychromatic flow cytometry data and to generate graphical representations of the T cells, the SPICE Program was used (Version 2.9, Vaccine Research Center, NIAID, NIH).

The Mann–Whitney U test was used to compare variables between the patients with LP and the healthy controls. P ≤ 0.05 was considered significant.

Previously, we verified that mononuclear cells from patients with LP have altered cytokine secretion upon activation of TLRs with synthetic ligands. This finding mainly applies to intracellular TLR7/TLR8 and TLR9 signalling [3]. One cytokine that is abundantly produced in PBMCs upon TLR activation is TNF [3]. Based on these studies, we analysed the frequency of TNF-α⁺ mDCs within populations of PBMCs [20]. To evaluate the responsiveness of DCs (mDCs and pDCs), populations of PBMCs were assessed using flow cytometry to determine the frequencies of TNF-α⁺ mDCs and IFN-α⁺ pDCs after stimulation with ligands for TLR4 (LPS), TLR7 (imiquimod) and TLR7/8 (CL097), and TLR9 (CpG) or with SEB as a positive control. The gating strategy used to evaluate the pDCs and mDCs is depicted in Additional file 1: Figure S1.

As shown in Fig. 1, an increased frequency of CD11c⁺HLA-DR⁺TNF-α⁺ mDCs was found in the LP samples after stimulation with LPS/TLR4 and CL097/TLR7-TLR8. No difference was observed following TLR7 activation, indicating that activation of TLR8 using the CL097 compound in mDCs potently induced TNF-α secretion. To induce type I interferon secretion by CD123⁺HLA-DR⁺ pDCs, we used TLR7 and TLR9 agonists, which resulted in similar frequencies of IFN-α⁺ pDCs in the LP and HC groups (Fig. 1b).

To verify that T cells were present in the patients with LP, we evaluated populations of tTregs and pTregs, derived from thymic or peripheral T cells, respectively [8]. To accomplish this goal, we evaluated the frequencies of CD4⁺ and CD8⁺Foxp3⁺ Tregs in peripheral blood samples collected from the LP and HC groups. The gating strategy used to identify Tregs is shown in Additional file 2: Figure S2.

The patients with LP exhibited increased frequencies of CD4⁺CD25⁺Foxp3⁺CD127^low/− T cells and CD8⁺ Tregs (Fig. 2a). Among the different subtypes of Tregs analysed, we verified that similar frequencies of CD4⁺ and CD8⁺ tTregs were present in both groups (Fig. 2b), while the patients with LP had a higher frequency of CD4⁺ pTregs than the healthy controls (Fig. 2c).

Previously, we verified that TLR activation produces an adjuvant effect by inducing cytokine production in PBMCs [3]. In the current study, we evaluated how stimulation of the innate immune system through TLR activation in T cells influences LP. Additionally, we stimulated PBMCs with TLR agonists such as LPS/TLR4, CL097/TLR7-8, CpG/TLR9 and SEB and then assessed intracellular levels of IFN-γ, IL-10, IL-22 and TNF in CD4⁺ and CD8⁺ T cells. The gating strategy used to quantify cytokines secreted from T cells is depicted in Additional file 3: Figure S3 (a, b).

We assessed the frequencies of CD4⁺ T cells that secreted IFN-γ and IL-10 (Fig. 3) and CD8⁺ T cells that secreted IFN-γ, IL-22 and TNF (Additional file 4: Figure S4) because these cytokines are specifically modified after TLR activation. Relative to the HC group, in the patients with LP, CD4⁺IFN-γ⁺ T cell frequency increased after stimulation with a TLR4 agonist and decreased after TLR9 activation, but it was not altered by SEB stimulation (Fig. 3a). Moreover, the patients with LP had a higher frequency of unstimulated CD4⁺IL-10⁺ T cells, whereas this population decreased following TLR4 or SEB stimulation.

Unstimulated CD8⁺ T cells exhibited increased production of proinflammatory cytokines, such as IFN-γ and IL-22 (Additional file 4: Figure S4). After stimulation with CL097/TLR7-8 or SEB, the LP group displayed a high frequency of IL-22⁺ cells (Additional file 4: Figure S4b). Upon TLR9 stimulation, the frequency of CD8⁺ T cells secreting TNF or IFN-γ (Additional file 4: Figure S4c) decreased in the LP group.

We also evaluated the ratios between Th17 cells and Tregs and Th1 cells and Tregs using baseline data (Additional file 5: Figure S5). These ratios were similar between the LP and HC groups.

To evaluate the frequencies of Th22 and Tc22 cells in peripheral blood, we analysed production of the cytokines IL-17, IFN-γ and IL-22 in T cells via flow cytometry. Th22 and Tc22 cells are IL-17⁻IFN-γ⁻IL22⁺, and the gating strategy used to identify these cells is shown in Fig. 4a. Under no-stimulation conditions, the LP group exhibited increased frequencies of Th22 and Tc22 cells compared to the HC group; this trend remained for Th22 cells following stimulation with SEB (Fig. 4b).

Next, we evaluated whether stimulation of PBMCs with TLRs could induce the production of polyfunctional T cells in patients with LP. To accomplish this, we assessed levels of the cytokines IL-17, IL-10, IFN-γ, TNF and IL-22 in T cells via flow cytometry using a Boolean gating strategy, which is described in Additional file 3: Figure S3c. As shown in Fig. 5, the patients with LP had fewer CD4⁺ T cells that simultaneously secreted the 5 evaluated cytokines following treatment with CL097 and CpG. However, TLR4 or TLR7/8 and TLR9 activation induced the production of some of these cytokines in T cells from patients with LP, although these cells did not appear to produce IL-10. Moreover, in the patients with LP, we observed an increased frequency of CD4⁺ T cells simultaneously secreting 3 of the evaluated cytokines compared to the HC following treatment with CL097 or CpG (pie chart, yellow slice). Notably, the most prominent combinations of 3 cytokines in this context included IL-22, IL-17, IFN-γ, or TNF, but not IL-10.

The production of polyfunctional CD8⁺ T cells was high in the LP group under no-stimulation conditions (Fig. 6). Following stimulation with SEB or CL097, the frequencies of CD8⁺ T cells that simultaneously secreted 2, 3, or 4 of the above-referenced cytokines increased in the samples collected from the patients with LP. Although no differences were noted between the groups when the cytokines were combined, a different response was observed upon activation with CL097 (pie chart, Fig. 6).