Gene Expression Profiling Suggests that Complement Activation Is Important for Blister Formation in Bullous Pemphigoid.

Bullous pemphigoid (BP) is the most common autoimmune bullous disease that typically presents in elderly patients with severe pruritus, with or without tense blisters on erythematous skin (Schmidt and Zillikens, 2013Schmidt E. Zillikens D. Pemphigoid diseases.Lancet. 2013; 381: 320-332Abstract Full Text Full Text PDF PubMed Scopus (706) Google Scholar). One in five patients lack blisters, a disease variant termed non-BP (NBP) (Di Zenzo et al., 2012Di Zenzo G. Della Torre R. Zambruno G. Borradori L. Bullous pemphigoid: from the clinic to the bench.Clin Dermatol. 2012; 30: 3-16Abstract Full Text Full Text PDF PubMed Scopus (114) Google Scholar; Meijer et al., 2019Meijer J.M. Diercks G.F.H. de Lang E.W.G. Pas H.H. Jonkman M.F. Assessment of diagnostic strategy for early recognition of bullous and nonbullous variants of pemphigoid.JAMA Dermatol. 2019; 155: 158-165Crossref PubMed Scopus (44) Google Scholar). In both phenotypes, autoantibodies are directed against hemidesmosomal antigens BP180 and BP230. Studies focusing on the bullous phenotype of BP support the hypothesis that autoantibodies against BP180 mediate blister formation by activation of the complement system through classical and alternative pathways, attracting inflammatory cells toward the skin (Dainichi et al., 2017Dainichi T. Chow Z. Kabashima K. IgG4, complement, and the mechanisms of blister formation in pemphigus and bullous pemphigoid.J Dermatol Sci. 2017; 88: 265-270Abstract Full Text Full Text PDF PubMed Scopus (23) Google Scholar; Nelson et al., 2006Nelson K.C. Zhao M. Schroeder P.R. Li N. Wetsel R.A. Diaz L.A. et al.Role of different pathways of the complement cascade in experimental bullous pemphigoid.J Clin Invest. 2006; 116: 2892-2900Crossref PubMed Scopus (89) Google Scholar). Complement-independent mechanisms of blistering were also described, involving the depletion of the BP180 molecule from the hemidesmosome by autoantibody-induced pinocytosis (Iwata and Ujiie, 2017Iwata H. Ujiie H. Complement-independent blistering mechanisms in bullous pemphigoid.Exp Dermatol. 2017; 26: 1235-1239Crossref PubMed Scopus (23) Google Scholar). Previous studies on the pathogenesis of BP assessed the mRNA expression levels of cytokines and chemokines in BP but not in NBP and only in single-cell types (Furudate et al., 2014Furudate S. Fujimura T. Kambayashi Y. Kakizaki A. Aiba S. Comparison of CD163+ CD206+ M2 macrophages in the lesional skin of bullous pemphigoid and pemphigus vulgaris: the possible pathogenesis of bullous pemphigoid.Dermatology. 2014; 229: 369-378Crossref PubMed Scopus (32) Google Scholar; Messingham et al., 2014Messingham K.N. Holahan H.M. Frydman A.S. Fullenkamp C. Srikantha R. Fairley J.A. Human eosinophils express the high affinity IgE receptor, FcεRI, in bullous pemphigoid.PLoS One. 2014; 9: e107725Crossref PubMed Scopus (66) Google Scholar). To date, it is unknown why patients with NBP do not develop blisters, whereas the autoantibody profile can be similar to that of BP. To contribute to the molecular understanding of both BP phenotypes, we assess the transcriptomics of inflamed lesional skin of 12 patients with BP and 12 patients with NBP and 7 control patients with pruritus of unknown origin (Table 1). Patients with NBP clinically suffered from pruritus but lacked bullae now and in the past, whereas patients with BP suffered from pruritus combined with bullae on the skin. Diagnostic inclusion criteria consisted of the two-of-three rule, meaning that patients had to meet at least two of the following three criteria: (i) pruritus and/or cutaneous blisters, (ii) positive linear IgG or complement component C3c staining by direct immunofluorescence microscopy, and (iii) positive IgG staining on the epidermal side of salt-split skin by indirect immunofluorescence microscopy (Meijer et al., 2019Meijer J.M. Diercks G.F.H. de Lang E.W.G. Pas H.H. Jonkman M.F. Assessment of diagnostic strategy for early recognition of bullous and nonbullous variants of pemphigoid.JAMA Dermatol. 2019; 155: 158-165Crossref PubMed Scopus (44) Google Scholar). Control samples were retrospectively selected. Inclusion criteria were pruritus of uncertain origin and negative direct immunofluorescence microscopy and immunoserology test results. All biopsies were taken from lesional erythematous inflamed skin. Quantification of 180 gene transcripts associated with innate and adaptive immune responses was performed using the NanoString nCounter Myeloid Innate Immunity Panel (NanoString Technologies, Seattle, WA). The Advanced Analysis tool (version 2.0.134) embedded within the nSolver Analysis Software 4.0 (NanoString Technologies) was used for data analysis (Supplementary Materials and Methods). The study was approved by the local Ethics Committee of the University of Groningen (Groningen, The Netherlands). Individual patient consent was not required for this study because none of the patients objected to the use of leftover human tissue for research purposes, a choice that is standardly given to each patient of the University Medical Center Groningen. Heatmaps were created with unsupervised clustering.Table 1Patient Characteristics and Gene Expression ResultsPatient CharacteristicsGene Expression1Whether gene expression was scored low or high was determined on the basis of unsupervised clustering and visualization of the data by heatmaps (see Figure 1 and Supplementary Figures S1 and S2).NBP Versus BPAgeSexDIF IgG; C3cIIF SSS Roof StainingAntigen Recognition2Antigen recognition is based on results of immunoblot and/or ELISA.Complement ActivationTh1 ResponseTh2 ResponseNBP 167Female2+; negIgG 1+BP180 onlyLowLowLowNBP 241Male+/2+; negIgG +/2+BP180 onlyLowLowLowNBP 395Male+/2+; negnegBP180 onlyLowLowLowNBP 4101Female2+; negIgG1+ IgA-/+BP180 + BP230LowHighHighNBP 588Female2+; negIgG2+BP180 + BP230LowLowHighNBP 678Femaledub; 3+IgA1+, IgG3+BP180 + BP230LowHighHighNBP 782Femaledub;negIgG 1+BP180 + BP230LowLowLowNBP 879Maleneg; negIgG+/2+BP230 onlyLowLowLowNBP 979Maleneg; negIgG 2+BP230 onlyLowLowLowNBP 1091Femaleneg; negIgG3+, IgA1+BP230 onlyLowHighHighNBP 1192Femaleneg; negIgG 2+BP230 onlyLowLowHighNBP 1286Femaleneg; negIgG 1+BP230 onlyLowHighLowBP 176Male3+; 2+IgG 3+BP180 + BP230LowHighHighBP 253Male3+; 2+IgG 3+BP180 onlyLowHighHighBP 387Male2+; 3+IgG 2+BP180 + BP230HighHighHighBP 471Female2+; +IgA1+, IgG3+BP180 + BP230HighHighHighBP 580Male+; 3+IgG 3+BP180 onlyHighHighHighBP 687Female+; 2+IgG 1+BP180 onlyLowHighHighBP 773Female+; +/-IgG 3+BP180 + BP230HighHighHighBP 880Female+; negIgG 3+BP230 onlyHighHighHighBP 989Femaleneg; 2+IgG 2+BP180 + BP230LowHighHighBP 1061Maleneg; negIgG 3+BP180 + BP230HighHighHighBP 1172Female3+; +IgG 2+BP180 onlyHighHighHighBP 1290Female2+; 2+IgG 2+BP230 onlyLowHighHighAbbreviations: BP, bullous pemphigoid; C3c, complement component C3c; DIF, direct immunofluorescence microscopy; NBP, nonbullous pemphigoid; IIF SSS, indirect immunofluorescence on salt-split skin; neg, negative; Th, T helper.1 Whether gene expression was scored low or high was determined on the basis of unsupervised clustering and visualization of the data by heatmaps (see Figure 1 and Supplementary Figures S1 and S2).2 Antigen recognition is based on results of immunoblot and/or ELISA. Open table in a new tab Abbreviations: BP, bullous pemphigoid; C3c, complement component C3c; DIF, direct immunofluorescence microscopy; NBP, nonbullous pemphigoid; IIF SSS, indirect immunofluorescence on salt-split skin; neg, negative; Th, T helper. Interestingly, genes related to complement activation were highly expressed in 7 of 12 (58%) BP biopsies, in 1 of 7 (14%) control biopsies, but not in NBP biopsies (0 of 12) (Figure 1). Moreover, all BP biopsies showed a strong dual T helper (Th) 1 and Th2 response (Supplementary Figures S1 and S2). Four NBP biopsies showed dual Th1- and Th2-related gene expression. Five control pruritus samples showed dual high expression for Th1- and Th2-related genes. One NBP and one control sample showed high Th2 gene expression only. No other notable clustering of BP and NBP samples was observed in the heatmaps of genes related to angiogenesis, antigen presentation, cell cycle and apoptosis, cell migration and adhesion, chemokine signaling, cytokine signaling, differentiation and maintenance of myeloid cells, Fc receptor signaling, GF signaling, IFN signaling, lymphocyte activation, pathogen response, T-cell activation and checkpoint signaling, and toll-like receptor signaling. Our finding showed that the majority of BP samples showed high expression of genes related to complement activation, whereas none of the NBP samples did emphasize the importance of complement activation in the blistering mechanism. In line, previous studies reported significantly fewer complement component C3c deposits along the basement membrane zone in NBP skin than in BP skin using direct immunofluorescence microscopy (Meijer et al., 2019Meijer J.M. Diercks G.F.H. de Lang E.W.G. Pas H.H. Jonkman M.F. Assessment of diagnostic strategy for early recognition of bullous and nonbullous variants of pemphigoid.JAMA Dermatol. 2019; 155: 158-165Crossref PubMed Scopus (44) Google Scholar; Romeijn et al., 2017Romeijn T.R. Jonkman M.F. Knoppers C. Pas H.H. Diercks G.F.H. Complement in bullous pemphigoid: results from a large observational study.Br J Dermatol. 2017; 176: 517-519Crossref PubMed Scopus (26) Google Scholar). It is hypothesized that complement activation induced by anti-BP180 IgG autoantibody binding to BP180 may lead to migration of mast cells, eosinophils, and neutrophils toward the skin (Heimbach et al., 2011Heimbach L. Li Z. Berkowitz P. Zhao M. Li N. Rubenstein D.S. et al.The C5a receptor on mast cells is critical for the autoimmune skin-blistering disease bullous pemphigoid.J Biol Chem. 2011; 286: 15003-15009Abstract Full Text Full Text PDF PubMed Scopus (60) Google Scholar; Nelson et al., 2006Nelson K.C. Zhao M. Schroeder P.R. Li N. Wetsel R.A. Diaz L.A. et al.Role of different pathways of the complement cascade in experimental bullous pemphigoid.J Clin Invest. 2006; 116: 2892-2900Crossref PubMed Scopus (89) Google Scholar). On activation, immune cells release cytotoxic substances and proteases that degrade extracellular matrix proteins and therefore cause a subepidermal split (Lo Schiavo et al., 2013Lo Schiavo A. Ruocco E. Brancaccio G. Caccavale S. Ruocco V. Wolf R. Bullous pemphigoid: etiology, pathogenesis, and inducing factors: facts and controversies.Clin Dermatol. 2013; 31: 391-399Abstract Full Text Full Text PDF PubMed Scopus (140) Google Scholar). Several mice studies showed that complement activation through both the classical pathway and alternative pathway might be essential for blister development (Heimbach et al., 2011Heimbach L. Li Z. Berkowitz P. Zhao M. Li N. Rubenstein D.S. et al.The C5a receptor on mast cells is critical for the autoimmune skin-blistering disease bullous pemphigoid.J Biol Chem. 2011; 286: 15003-15009Abstract Full Text Full Text PDF PubMed Scopus (60) Google Scholar; Liu et al., 1995Liu Z. Giudice G.J. Swartz S.J. Fairley J.A. Till G.O. Troy J.L. et al.The role of complement in experimental bullous pemphigoid.J Clin Invest. 1995; 95: 1539-1544Crossref PubMed Scopus (265) Google Scholar; Mihai et al., 2018Mihai S. Hirose M. Wang Y. Thurman J.M. Holers V.M. Morgan B.P. et al.Specific inhibition of complement activation significantly ameliorates autoimmune blistering disease in mice.Front Immunol. 2018; 9: 535Crossref PubMed Scopus (24) Google Scholar; Nelson et al., 2006Nelson K.C. Zhao M. Schroeder P.R. Li N. Wetsel R.A. Diaz L.A. et al.Role of different pathways of the complement cascade in experimental bullous pemphigoid.J Clin Invest. 2006; 116: 2892-2900Crossref PubMed Scopus (89) Google Scholar). Mice deficient in important complement factors (C5aR, factor B, or C4) were injected with pathogenic anti-BP180 IgG and clearly showed less or no blistering compared with wild-type mice. However, our study also reports five BP samples without a high expression of complement-related genes; still, the patients expressed a blistering phenotype. Studies have suggested complement-independent blistering through autoantibody-induced internalization of the complete BP180 protein, therefore weakening the hemidesmosomal adhesion strength (Iwata and Ujiie, 2017Iwata H. Ujiie H. Complement-independent blistering mechanisms in bullous pemphigoid.Exp Dermatol. 2017; 26: 1235-1239Crossref PubMed Scopus (23) Google Scholar; Natsuga et al., 2012Natsuga K. Nishie W. Shinkuma S. Ujiie H. Nishimura M. Sawamura D. et al.Antibodies to pathogenic epitopes on Type XVII collagen cause skin fragility in a complement-dependent and -independent manner.J Immunol. 2012; 188: 5792-5799Crossref PubMed Scopus (62) Google Scholar). Our observations underscore the complexity of blister formation in bullous BP and the complement system with its various activation routes. This explorative pilot study is not without limitations. First, our study has a limited sample size and retrospective character. Second, in contrast to an RNA-sequencing approach, the current analyses were performed with a predesigned panel containing innate and adaptive immune response genes. For instance, IL31and IL-31RA, which can be expressed in pruritic disease and dermatomyositis (Kim et al., 2018Kim Y.J. Lee H.J. Ryu J.S. Kim Y.H. Jeon S. Oh J.Y. et al.Prospective clinical trial of corneal reconstruction with biomaterial-free cultured oral mucosal epithelial cell sheets.Cornea. 2018; 37: 76-83Crossref PubMed Scopus (25) Google Scholar), were not part of the panel; as such, the importance of these and other untested genes in the pathophysiology of blister formation remains unknown. In contrast, the NanoString analysis can be performed on RNA extracted from formalin-fixed paraffin-embedded tissue samples and enables the proper histological assessment of the tissue of interest. Finally, these explorative gene expression profiles do require further validation on a protein level. On the basis of the data of this explorative study, it may be carefully concluded that genes related to complement activation showed a higher expression in BP skin than in NBP skin, suggesting an important role in blister formation. Further studies for validation of the genetic data are needed and should assess specific gene and protein expression in a larger patient cohort to provide greater insight into the mechanism of pruritus and blistering in BP. Datasets related to this article can be found at NanoString dataset, Mendeley Data, V1, doi: 10.17632/kgcthpzfpt.1, an open-source online data repository hosted at Mendeley Data (Lamberts, 2022Lamberts A. Nanostring dataset. University of Groningen, 2022https://doi.org/10.17632/kgcthpzfpt.1Crossref Google Scholar). Aniek Lamberts: http://orcid.org/0000-0001-6357-5264 Nika Kotnik: http://orcid.org/0000-0002-2083-5860 Joost M. Meijer: http://orcid.org/0000-0001-7654-3528 Leon C. van Kempen: http://orcid.org/0000-0003-0646-0705 Gilles F. H. Diercks: http://orcid.org/0000-0001-8053-216X Barbara Horváth: http://orcid.org/0000-0001-8559-3674 AL reports receiving financial support granted by the International Pemphigus and Pemphigoid Foundation for conducting this study. LCVK reports receiving grants and nonfinancial support from Amgen, Bayer, Biocartis, Invitae, Merck, NanoString, and Roche and is a consultant with advisory boards for AstraZeneca, Bayer, Janssen-Cilag, and Merck. BH reports fees from Janssen-Cilag (advisory boards, educational grants, consultations, investigator initiative studies), AbbVie (advisory boards, educational grants, consultations, investigator initiative studies), Novartis Pharma (advisory boards, consultations, investigator initiative studies), UCB Pharma (advisory boards, consultations), Leo Pharma (consultations), Solenne B.V. (investigator initiative studies), Celgene (consultations, investigator initiative studies), Akari therapeutics (consultations, investigator initiative studies), Philips (consultation), Roche (consultation), Regeneron (consultation), Sanofi (consultation), and Argenx (advisory boards, consultations), which fees were paid to the institution. The remaining authors state no conflict of interest. Financial support for this work was granted by the International Pemphigus and Pemphigoid Foundation. Conceptualization: AL, NK, LCVK, GFHD, BH; Data Curation: AL, NK, JM, LCVK, GFHD, BH; Formal Analysis: AL, LCVK; Funding Acquisition: AL, JM, GFHD, BH; Methodology: AL, LCVK, GFHD, BH; Software: LCVK; Supervision: JM, GFHD, BH; Visualization: AL, NK, LCVK; Writing – Original Draft Preparation: AL; Writing – Review and Editing: NK, JM, LCVK, GFHD, BH Patients diagnosed with bullous pemphigoid (BP) (n = 14) and non-BP (n = 12) at the outpatient clinic of our dermatology department were retrospectively selected. Diagnostic inclusion criteria consisted of the two-of-three rule, meaning that patients had to meet at least two of the following three criteria: (i) pruritus and/or cutaneous blisters, (ii) positive linear IgG or complement component C3c staining by direct immunofluorescence microscopy, and (iii) positive IgG staining on the epidermal side of salt-split skin by indirect immunofluorescence microscopy (Meijer et al., 2019Meijer J.M. Diercks G.F.H. de Lang E.W.G. Pas H.H. Jonkman M.F. Assessment of diagnostic strategy for early recognition of bullous and nonbullous variants of pemphigoid.JAMA Dermatol. 2019; 155: 158-165Crossref PubMed Scopus (47) Google Scholar). In addition, sufficient formalin-fixed and paraffin-embedded tissue had to be available for histopathological assessment and RNA extraction. Patient characteristics were assessed by reviewing patient charts. Patients using systemic immunosuppressive drugs at the time of the biopsy were excluded. The use of topical corticosteroids was avoided but was allowed at the lowest class if a more suitable biopsy was not available. All biopsies were taken from lesional erythematous inflamed skin. Skin samples of patients with pruritic skin lesions (n = 10) were included as control samples and were retrospectively selected. Inclusion criteria were pruritus of uncertain origin and negative direct immunofluorescence microscopy and immunoserology test results. Histopathology skin samples must show aspecific inflammation without an exact diagnosis by the pathologist. The study was approved by the local Ethics Committee of the University of Groningen (Groningen, The Netherlands). The NanoString nCounter (Tsang et al., 2017Tsang H.F. Xue V.W. Koh S.P. Chiu Y.M. Ng L.P. Wong S.C. NanoString, a novel digital color-coded barcode technology: current and future applications in molecular diagnostics.Expert Rev Mol Diagn. 2017; 17: 95-103Crossref PubMed Scopus (90) Google Scholar) Myeloid Innate Immunity Panel (NanoString Technologies, Seattle, WA) was used to quantify the expression of 180 genes associated with innate and adaptive immune responses. RNA was isolated from four 5-μm thick formalin-fixed and paraffin-embedded skin sections of 14 patients with confirmed BP, 12 with confirmed non-BP, and 10 with pruritus using the RNeasy mini kit (Qiagen, Hilden, Germany) according to the supplier's instructions. RNA (100 ng as measured by Qubit [Thermo Fisher Scientific, Waltham, MA]) was hybridized with the NanoString reporter and captured probes overnight at 65 °C. Each probe set (reporter and capture probe) is designed to bind a unique mRNA target and has a unique color-coded molecular tag. The RNA−probe complexes were loaded on a nCounter cartridge and washed and read on a SPRINT platform. The Advanced Analysis (version 2.0.134) tool embedded within the nSolver Analysis Software 4.0 (NanoString Technologies) was used for data analysis and comprised a correction for hybridization differences between samples and selection of reference genes using the build-in GeNorm algorithm. When a sample was flagged because of low overall counts and/or high normalization factors, this indicated low RNA quality and excluded the sample from further analyses. Two BP samples and three pruritus control samples were excluded from the study owing to low quality of the transcript data. Gene expression levels were successfully measured in 12 BP, 12 non-BP biopsies, and 7 pruritus control biopsies. Heatmaps were created with unsupervised clustering. Samples were labeled as BP, non-BP, or control.Supplementary Figure S2Heatmap with unsupervised clustering of expressed genes involved in a T helper 2 response. Highly expressed genes are depicted in orange, whereas genes with lower expression are in blue. The samples are labeled BP, NBP, or control. BP, bullous pemphigoid; NBP, nonbullous pemphigoid.View Large Image Figure ViewerDownload Hi-res image Download (PPT)