Ii-TGO mice were generated by intercrossing mice that express an invariant chain promoter-driven reverse transactivator (Ii-rtTA) transgene with mice that express a Tet-regulated ovalbumin fusion protein (TRE-TGO). Upon Dox administration, these mice express an OVA fusion protein that incorporates the transferrin receptor transmembrane domain to facilitate efficient trafficking to endocytic compartments. To compare the pathogenic potential of distinct T cell subsets, OVA-specific DO11 T cells were activated in vitro with OVA-peptide and APCs under Th1 or Th2 skewing conditions, expanded in the presence of IL-2, and then restimulated 2-3 days prior to i.v. injection ( Figure 1A ). In some studies, we used T cells derived from DO11 mice that had been intercrossed with the IL-4 reporter line, 4get (15). We confirmed the functional phenotype of the Th1 and Th2 cells at the time of injection by flow cytometry. Only the Th1 cells expressed IFNγ, and only the Th2 cells expressed IL-4. In addition, the majority of the OTII 4get Th2 cells expressed GFP while few if any of the OTII 4get Th1 cells were GFP+, confirming their commitment to the Th1 lineage ( Figure 1B ). Cytokine concentrations in culture fluids collected after the in vitro restimulation were determined by ELISA, and the results confirmed the expected phenotype; Th1 supernatants contained high levels of IFNγ and the Th2 supernatants contained IL-4 and IL-10 ( Figure 1C ).

In addition, RNA extracted from the restimulated T cells was analyzed by NanoString™ Mouse Immunology code set. Tbx21 (Tbet) was enriched in Th1 cells and Gata3 was enriched in Th2 cells ( Figure 1D ). Additional genes were differentially upregulated in the two populations, including increased S100a8, Il13, Il10rb, and Runx1 in Th2 cells and increased Mif, Itgb1, Ifitm1 and Sell in Th1 cells (heatmap) ( Figure 1E ). The highest upregulated differentially expressed gene (DEG) in Th2 cells was Il1rl1 (volcano plot). Gene set analysis (GSA) revealed Th2 differentiation, lymphocyte trafficking and type I interferon signaling terms were different between Th1 and Th2 cells using a cutoff significance score of 1.3.

We reported previously that sublethally irradiated (400R) IiTGO mice, provided with Dox chow and injected with activated Th2 cells, developed lupus-like skin lesions (5). To determine whether Th1 cells could induce CLE as efficiently as Th2 cells, TLR9^-/- Ii-TGO recipients were sublethally irradiated (400R) and provided with Dox chow 6-18 hrs prior to i.v. injection of Th1 or Th2 DO11 T cells ( Figure 2A ). Skin lesions developed in the Th2 injected mice 3-4 weeks post transfer, but not in the mice injected with Th1 cells, as shown in images of representative mice ( Figure 2B ) and compiled skin score data from 4 experiments ( Figure 2C ). The absence of skin lesions in the Th1-injected mice was not due to the failure of KJ1-26 cells to survive or engraft the recipients, as indicated by splenomegaly ( Figure 2D ) and the initial weight loss in the Th1-injected mice ( Figure 2E ). To further compare the extent of engraftment of the injected DO11 Th1 and Th2 T cells, single cell suspensions obtained from the skin, LN and spleen of the injected mice were analyzed by flow cytometry using the KJ1-26 anti-clonotypic antibody. We found more KJ1-26+ cells in Th2 injected mice in all 3 tissues ( Figure 2F ).

Th1 and Th2 cells are known to express distinct sets of chemokine receptors and Th1 and Th2 cells use different ligands and chemokine receptors to enter the skin (14). In addition to their role in cell migration, chemokine receptors impart functional capacity on T cells. Therefore, we examined DEGs between pre- and post-injection T cells, as well as Th1 and Th2 skewed cells, to better understand factors that might be contributing to T cell function in our model. CXCR6 was significantly higher in Th2 vs. Th1 cells by both NanoString array and flow cytometry ( Figure 2G ). However, there were no significant differences in the expression of other chemokine receptors, based on both Nanostring and flow panels, that have been reported to distinguish Th2 cells from Th1 cells (e.g., CXCR3, CCR4, CCR5, or CCR8) ( Figure 3G ), even though the cells were clearly skewed to the Th2 subset, based on cytokine production and expression of the 4get reporter ( Figure 1 ).

To confirm the presence of the ligands for CXCR6 and CLA in CLE mouse skin, we queried our NanoString dataset which compared RNA isolated from the skin of Th2 injected TLR9^-/- Ii-TGO mice (skin score 3-4) to the skin of TLR9+/+ IiTGO mice (no disease) (5). We found significant increases in CXCL16 (ligand for CXCR6) and SELL (ligand for CLA) but not SELE (another ligand for CLA) in TLR9-/- versus WT skin ( Figure 2H ). Taken together, these data suggest that Th2 skewing promotes skin infiltration by patterning expression of skin-homing molecules, thereby allowing the Th2 cells to follow CXCL16-CXCR6 chemokine and CLA-L-Selectin integrin signals to impart functional capacities in skin.

Next, we compared the injected KJ1-26 Th2 cells to KJ1-26 T cells isolated from lesional skin 4-5 weeks post injection to identify changes in gene expression that developed during the post-injection time frame. Antigen-specific T cells were enriched from the skin using KJ1-26 magnetic beads for positive selection. NanoString analysis of RNA isolated from these cells revealed 150 differentially expressed genes (DEGs) between injected and 4 wk post-injection KJ1-26+ T cells. T cells isolated from lesional skin showed the upregulation of interferon-stimulated genes (ISGs), including IRF5, IFIH1, and MX1 ( Figure 3A ). The data also showed decreased expression of Th2-related genes including IL4RA. These data, combined with the observation that the Th2 upregulated other ISGs led us to further examine the cytokine profile of post-injection DO11 4get T cells.

We tracked 4get expression in both in vitro and ex vivo from T cells, in addition to staining for IFNγ. By 4 weeks post Th2 cell injection a high proportion of KJ1-26+ T cells in the skin and skin-draining LN (sdLN) exhibited a Th1-like phenotype, as defined by Tbet and IFN-γ expression ( Figures 3B, C ). Intriguingly, ~75% of the T cells in the skin were Tbet+/IFNg+ and of these, 25% also were also GFP⁺. Lower GFP detection in combination with Tbet staining may be due to the transcription factor fixation/nuclear permeabilization protocol, which is more harsh than the reagents used for intracellular cytokine staining. Gata3, a master regulator of Th2-associated gene expression, was also reduced in the post-injection T cells expression ( Figure 3D ). Together, these data indicate that the injected KJ1-26+ T cells switch from a Th2 to a Th1-like subset and acquire the capacity to produce IFNγ.

To understand the significance of IFN-γ to development of skin disease, we skewed WT or IFN-γ^-/- DO11 T cells towards a Th2 phenotype and injected them into host mice. IFN-γ^-/- DO11 failed to induce skin lesions in mice ( Figures 3E, F ), despite inducing splenomegaly and an initial drop in body weight ( Figures 3G, H ). These data indicate that the switch towards a Th1 phenotype is required for the development of skin disease in this model.

CLE is cyclical, with relapsing and remitting flares of cutaneous lesions. To model disease flares, mice that had developed skin disease were taken off Dox chow to allow the skin lesions to heal. Dox chow was then readministered to these mice 4 weeks later to reinduce the model autoantigen ( Figure 4A ). Within two weeks, skin disease recurred ( Figure 4B ) without the injection of additional DO11 T cells. Sera collected during the initial clinical presentation and at the time of flare were assayed by protein array to determine whether cytokine titers increased in the mice with flares. We found increases in IL28 (IFNλ) and IFN-γ in flares versus initial clinical presentation ( Figure 4C ). Cytokine titers in the sera of flare control mice, which did not receive the second course of Dox chow, were comparable to those detected during the initial response.

To determine whether these cytokines could be T cell derived, RNA isolated from the cultured Th1 and Th2 cells was compared to T cells isolated from the skin during the initial and flare responses. Gene transcription was assessed with the NanoString probe set as in Figure 3 . We found that T cells isolated from the skin showed minimal expression of IL4, IL5 and IL13, but did express IL28 and IFNγ, both of which trended higher at time of flare ( Figure 4D ) Taken together, these data indicated that the Th1-like phenotype, established during the primary response, persisted in the T cell memory compartment.

All mice were housed in pathogen-free facilities at UMMS, and procedures were approved under protocol #2096 by the UMMS Institutional Animal Care and Use Committee and in accordance with the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals. Mice used for these studies were on the Balb/c background. Age and sex-matched mice were used, and both male and female mice of all strains were tested to avoid gender bias. Replicate experiments were performed two to five times.

Recipient TLR9KO Ii-TGO and WT Ii-TGO controls were generated as previously described (5). BALB/c DO11 mice (C.Cg-Tg [DO11.10]10 Dlo/J; Jackson Laboratory stock no. 003303) or Rag-/- DO11 mice (C.Cg-Rag1tm1Mom Tg(DO11.10)10Dlo/J; stock no. 030666) bred to IL-4/GFP-enhanced transcript (4get) mice (C.129-Il4tm1Lky/J; stock no. 004190) were used as T cell donors.

Magnetic bead-purified DO11 CD4+ T cells (BD IMag magnetic particles) were activated using OVA peptide–pulsed (323-339, In vivogen) irradiated spleen cells (as source of APCs) as described previously (33). Th1 cells were cultured with recombinant mouse IFNγ (10ng/mL) and anti-IL4 antibody (10ug/mL); Th2 cells were cultured with recombinant mouse IL-4 (10ng/mL) and anti-IFNγ (XMG2.1 10ug/mL) and anti-IL12p40 (10ug/mL). All cells (including unskewed Th0) received recombinant mouse IL-2 from J2 supernatant to promote survival and expansion. Cells were split on day 2, fed IL-2 on day 3, and split again on day 4. By day 7, the cells had rested and were re-activated, but not re-skewed, with another batch of OVA-pulsed splenocytes. Cells were harvested on day 10 at the peak of activation post-restim.

Supernatants from restimulated T cells and/or serum from CLE mice were assayed in the RayBiotech Th1/Th2/Th17 mouse Quantibody array per the manufacturer’s protocol. Slides were shipped for scanning array service and data were analyzed by taking the median fluorescence intensity minus the background fluorescence from blank control wells. Data are deposited on GEO Database under accession # GSE186095.

RNA was extracted from polarized cultured T cells, and from post-injection CD3 column-enriched (Miltenyi biotech) T cells from CLE mouse skin using Qiagen RNEasy mini kits. RNA was hybridized for ~18h (BioRad CFX thermocycler) and assayed in the NanoString mouse Immunology panel per the manufacturer’s instructions. Data were analyzed with Rosalind software using NanoString partner analysis. Data are deposited on GEO Database under accession # GSE185355.

10^7 activated and skewed T cells were injected i.v. into sublethally irradiated (4 Gy) age- and sex-matched TLR9WT or TLR9KO Ii-TGO recipient mice. To induce expression of the TGO transgene in the MHCII cells, mice were fed with 200 mg/kg of Dox chow (Bio-Serv). For CLE flares, mice were kept on chow for 4-5 weeks, allowed to heal for 4 weeks, then Dox chow was reintroduced.

Single-cell suspensions obtained from spleen, sdLNs, and skin were analyzed by flow cytometry using fluorochrome-conjugated mAbs listed in Table S1 . Zombie Aqua or Zombie NIR (Biolegend) was used to distinguish live and dead cells. Intracellular staining was carried out on cells incubated with Brefeldin A (Biolegend) in all tissue digestion and FACS staining buffers, approximately for 4 hours. Cells were permeabilized and fixed with transcription factor staining buffer (Invitrogen) or Cytofix/Cytoperm (BD Biosciences) and subsequently incubated with fluorochrome-conjugated mAb to mouse IFN-γ (clone XMG1.2, eBioscience), IL4 (clone 11B11, Biolegend), Tbet (clone 4B10, Biolegend), or GATA3 (clone TWAJ, eBioscience). Flow cytometric analysis was carried out using a Cytek Aurora, and analysis was conducted with FlowJo software 9.7.6 (TreeStar).

Cells were isolated from the skin as described previously (5). Briefly, shaved dorsal skin was harvested, minced, and digested for 45 minutes at 37°C with 2.0 mg/ml collagenase XI from Clostridium histolyticum (Sigma-Aldrich), 0.5 mg/ml hyaluronidase from bovine testes (Sigma-Aldrich), and 0.1 mg/ml DNAse (Sigma-Aldrich). Single cells were washed with 10% cRPMI, filtered through a 100 μm filter, and stained for flow cytometry staining as described above. For samples to be used for ICS, Brefeldin A was added to the digestion buffer and surface stain cocktail. For enrichment of antigen-specific T cells from skin, we used PE-conjugated KJ1-26 antibody and a PE positive selection kit (Miltenyi biotech).

Statistical analyses were performed using Prism software version 7.0 (GraphPad). Experiments are reported as mean ± SEM. Data were analyzed using a 2-tailed Student’s t test for comparison between 2 data sets. Multiple comparisons were analyzed by 1-way ANOVA and 2-way ANOVA, followed by Tukey’s multiple-comparison post hoc test. Differences were considered significant at a P value of less than 0.05.