Zachary Peters1, Lindsay Mendyka2, Sicong Shan1, William Rigby3, Christopher Burns4, Randolph Noelle5 and Sladjana Skopelja-Gardner4, 1Dartmouth College, Hanover, NH, 2Dartmouth Hitchcock, Lebanon, NH, 3Dartmouth-Hitchcock, Norwich, VT, 4Dartmouth Hitchcock Medical Center, Lebanon, NH, 5Geisel School of Medicine at Dartmouth, Hanover, NH
Background/Purpose: Persistent production of type I interferons (IFN-Is) is one of the hallmarks of lupus skin disease that is exacerbated by ultraviolet (UV) light. We previously showed IFN-I induction by UV occurs in a cGAS-STING-dependent manner, but the IFN-I signature returns to baseline levels in healthy skin, unlike in lupus. Here, we propose that the immune checkpoint VISTA is a regulator of skin IFN-I production in keratinocytes of therapeutic relevance to lupus.
Methods: Skin biopsies from B6, B6.Vsir-/- (VISTA-deficient), B6.Vsir-/-Sting-/-, KRT14creVsirfl/fl (Vsir-/- in keratinocytes), and cre-Vsirfl/fl female mice (3 mo) were collected prior to, 3 and 24h after UVB (500mJ/cm2). Gene expression was quantified by RNA-seq (Rosalind). Skin infiltrating cells were quantified by flow cytometry. Human keratinocytes were isolated from healthy skin and culturedin vitro. Cells were treated with agonistic anti-VISTA (803) or isotype IgG2a (20ug/ml) prior to UVB (50mJ/cm2), in the presence or absence of IFNa (100U). Expression of IFN-Is and IFN-I stimulated genes (ISGs) were quantified by qPCR (4hr after UV) and IFN-I score derived.
Results: At baseline, B6.Vsir-/- skin had increased IFN-k mRNA, IFN-b protein, as well as total STING protein levels (Fig. 1A-C). RNA-seq revealed increased expression of Sting and upstream cytosolic DNA sensors in VISTA-deficient mouse skin before and after UV. B6.Vsir-/- mice exhibited a 5-fold higher skin IFN-I score after UV compared to B6 mice (Fig.1D). Higher baseline and UV-induced IFN-I response in VISTA-deficient skin was suppressed in the absence of STING (Vsir-/-Sting-/- mice). Moreover, fewer skin-infiltrating neutrophils and inflammatory monocytes were recruited to Vsir-/-Sting-/- vs. STING-sufficient Vsir-/- skin post UV. RNAseq also revealed decreased expression of DNA repair genes like Ogg1, Xpd and Pole in B6.Vsir-/- skin, which are essential for repair of UV-induced DNA damage. VISTA deficient mouse keratinocytes produced 2-fold higher IFN-k at baseline ex vivo and accumulated higher levels of oxidized DNA (8-OHdG) compared with B6 cells (Fig. 2B-C).Mice with a conditional VISTA deletion only in keratinocytes exhibited a 10-fold higher baseline IFN-I signature and increased IFN-b protein levels compared to controls (Fig. 2D-E). Expression of VISTA on human keratinocytes decreased 4 hr after UV, but increased 5-fold 24 hr after UV(Fig. 3A).Pre-treatment of human keratinocytes with an agonistic anti-VISTA antibody (803) suppressed UV-induced IFN-k, Stat1 phosphorylation and IFN-I score, even in keratinocytes with a pre-existing IFN-I signature (Fig. 3C-D).
Conclusion: These studies identify VISTA as a suppressor of IFN-I production in keratinocytes and demonstrate STING-dependent IFN-I production in the absence of VISTA. As oxidized DNA is resistant to degradation in the cytosol where it can serve as a potent trigger of cGAS-STING, VISTA promotion of DNA repair may indirectly suppress STING signals. Antibody-mediated activation of VISTA in keratinocytes demonstrates its potential as a target to suppress IFN-I production in the context of photosensitivity and lupus.
Figure 1. Elevated skin IFN-I production in the absence of VISTA. A. IFN-β immunoblot of B6 and B6.Vsir-/- skin lysates; levels normalized to β-actin (ImageJ). B. IFN-k gene expression in B6 and Vsir-/- whole skin measured by qPCR. C. STING immunoblot of B6 and B6.vsir-/- skin lysates. D. Skin IFN score in B6 and Vsir-/- mice, before and after UVB exposure (1x500mJ/cm2). Differences in gene expression determined relative to baseline B6 and all statistical significance determined by two-way ANOVA (A,D) or Student’s T test(B,C) (n=5, *p<0.05, **p<0.01, ***p<0.005, ****p<0.00001). IFN-I score calculated as sum normalized expression of 6 ISGs relative to baseline B6 (ISGs: IRF7, IFIT1, IFIT3, ISG15, IFI27La, Mx1) and normalized to 18s.
Figure 2. VISTA regulates IFN-I production in keratinocytes. A. Mouse epidermis stained for VISTA (red), Keratin-14 (green), and DAPI (blue). B. Baseline IFN-κ protein in B6 and Vsir-/- keratinocyte lysates after 6 days in culture, quantified relative to B-actin. C. Mouse keratinocytes stained for 8-hydroxyguanosine (yellow) and DAPI (blue). D. IFN-b immunoblot of skin lysates from B6, B6.Vsir-/-, Vsirfl/fl and KRT14Cre:Vsirfl/fl mice. E. Baseline skin IFN-I score of Vsirfl/fl and KRT14Cre:Vsirfl/fl mice. Significance determined by two-way ANOVA (A). (n=3, * p<0.05, p<0.01). Differences in gene expression relative to Vsirfl/fl control and statistical significance determined by T test (n=4).
Figure 3. Anti-VISTA agonistic IgG suppresses UV-induced IFN-I signature in human keratinocytes. A. VISTA surface expression on primary human keratinocytes measured by flow cytometry, without UV, 4 and 24hr after UVB (50mJ/cm2). B. IFN-k gene expression in human keratinocytes pre-treated with either anti-VISTA INX803 IgG2 or isotype control IgG2 antibody. C. IFN-I scores of human keratinocytes pre-treated with 803 or isotype control IgG2 before and after UV exposure were derived as the sum of normalized expression levels of 9 interferon-stimulated genes (CXCL10, BST2, ifitm3, ifi27, isg15, ifit1, mx1, Ly6E, irf7). D. Immunoblot of phosphorylated STAT1 and IFN-k in human keratinocyte lysates from cells before and after UVB exposure, in the absence or presence of IFN⍺ (100U, 24hr prior to UV exposure) and pre-treated with anti-VISTA 803 or isotype control IgG2 antibody. Pixel intensity calculated relative to b-actin (ImageJ) and fold change of P-STAT1 and IFN-k in the presence of IFNa calculated relative to baseline NoUV. Statistical significance was determined by two-way ANOVA(A,C) and Student’s t-test (B,D); *p<0.05, **p<0.01, ns= not significant.
Z. Peters: None; L. Mendyka: None; S. Shan: None; W. Rigby: None; C. Burns: None; R. Noelle: None; S. Skopelja-Gardner: None.
.jpg)
.jpg)
.jpg)
