Research ArticleSKIN

Neutrophil extracellular traps, B cells, and type I interferons contribute to immune dysregulation in hidradenitis suppurativa

See allHide authors and affiliations

Science Translational Medicine  04 Sep 2019:
Vol. 11, Issue 508, eaav5908
DOI: 10.1126/scitranslmed.aav5908

Implicating the immune system in hidradenitis suppurativa

The cause of the painful skin condition hidradenitis suppurativa (acne inversa) is unknown. To better understand pathogenesis of this debilitating disease, Byrd et al. examined blood and lesional skin of patients and healthy controls. They observed prominent neutrophil extracellular traps (NETs), plasma cells, autoantibodies, and a type I IFN signature in patient samples. Inflammation and netting neutrophils may promote B cell activation and autoantibody secretion, and all of these mechanisms may contribute to tissue damage. These results suggest that multiple arms of the immune system are dysregulated in hidradenitis suppurativa and could be potentially targeted in new therapies.


Hidradenitis suppurativa (HS), also known as acne inversa, is an incapacitating skin disorder of unknown etiology manifested as abscess-like nodules and boils resulting in fistulas and tissue scarring as it progresses. Given that neutrophils are the predominant leukocyte infiltrate in HS lesions, the role of neutrophil extracellular traps (NETs) in the induction of local and systemic immune dysregulation in this disease was examined. Immunofluorescence microscopy was performed in HS lesions and detected the prominent presence of NETs. NET complexes correlated with disease severity, as measured by Hurley staging. Neutrophils from the peripheral blood of patients with HS peripheral also displayed enhanced spontaneous NET formation when compared to healthy control neutrophils. Sera from patients recognized antigens present in NETs and harbored increased antibodies reactive to citrullinated peptides. B cell dysregulation, as evidenced by elevated plasma cells and IgG, was observed in the circulation and skin from patients with HS. Peptidylarginine deiminases (PADs) 1 to 4, enzymes involved in citrullination, were differentially expressed in HS skin, when compared to controls, in association with enhanced tissue citrullination. NETs in HS skin coexisted with plasmacytoid dendritic cells, in association with a type I interferon (IFN) gene signature. Enhanced NET formation and immune responses to neutrophil and NET-related antigens may promote immune dysregulation and contribute to inflammation. This, along with evidence of up-regulation of the type I IFN pathway in HS skin, suggests that the innate immune system may play important pathogenic roles in this disease.


Hidradenitis suppurativa (HS), also known as acne inversa, is a debilitating inflammatory disorder of the pilosebaceous unit and apocrine gland-rich anatomical regions. HS is characterized by recurrent painful abscess-like nodules with malodorous purulent drainage, sinus tracts, and scarring in the axillae, groin, anogenital region, and other intertriginous areas. Conflicting reports estimate the prevalence of HS as 0.053 to 4% (1), with a 3:1 female/male ratio, and women of childbearing age bearing the burden of this disease (2). Although it is also more prevalent among African Americans (3, 4), very few studies have focused on this population.

HS is associated with stigmatization and substantial morbidity, with quality of life scores worse than psoriasis and most other dermatologic conditions (5). Patients with HS are more likely to suffer from depression, anxiety, and sexual dysfunction, and the impact of the disease on quality of life is worse in women (5, 6). Common current treatment options include intralesional corticosteroids, antibiotics, biologics [tumor necrosis factor (TNF) inhibitors], and surgical resection.

The etiology of HS is unknown, with obesity and smoking among risk factors (7). Neutrophils are prominent in affected lesional skin (8). This is linked to recent associations between dermal interleukin-17 (IL-17) and T helper 17 (TH17) enhanced responses in HS that may drive neutrophil migration into affected areas and promote tissue damage (8, 9). However, the role of neutrophils in the pathogenesis of HS remains unclear. Among the several functions of neutrophils, these cells have the ability to form neutrophil extracellular traps (NETs) after exposure to certain microbes or sterile stimuli. NETs are web-like structures composed of nucleic acids and granule or cytosolic proteins that are extruded to the extracellular space after exposure to various danger signals (10, 11). During NET formation, peptidylarginine deiminase 4 (PAD4) is activated, promoting histone citrullination. Histone H3 and H4 citrullination has been detected in tissues where NETs are present (10). Dysregulation of NET formation has been implicated in various autoinflammatory and autoimmune conditions, as NETs can activate several aberrant innate and adaptive immune responses, including the induction of type I interferons (IFNs) and the NLRP3 inflammasome, leading to amplification of inflammation (1214).

NETs are considered a putative source of modified autoantigens targeted by the immune system in individuals with systemic lupus erythematosus (SLE) (15, 16), rheumatoid arthritis (RA) (10, 11, 17), and anti-neutrophil cytoplasmic antibody vasculitis (AAV) (18). Both enhanced NET formation and impaired NET clearance have been associated to the development of autoantibodies in various autoimmune conditions (12, 19, 20). It has been a matter of debate in the recent literature whether HS displays features of autoimmunity, although this condition has been associated with several chronic inflammatory diseases (2124).

Infiltration of neutrophils in affected lesions during active disease has been reported in HS (8), but their role in the initiation and/or perpetuation of this condition has not been systematically explored. Here, we analyze mostly biospecimens from African American subjects with HS and report substantial systemic and lesional innate and adaptive immune dysregulation in this condition, in association with disease progression.


Netting neutrophils are present in HS lesions and peripheral blood

Histopathologic analysis of affected skin lesions from HS patients by hematoxylin and eosin (H&E) staining demonstrated infiltration of neutrophils when compared to skin from healthy donors. Multilobulated neutrophils were found primarily around the hair follicles (Fig. 1A), and this was corroborated by the presence of myeloperoxidase (MPO)–positive cells in the same areas in HS dermis but not in healthy control dermis (Fig. 1A, lower panel). To assess whether NETs are present in HS skin, we stained the dermis with antibodies recognizing MPO and citrullinated histone 4 (cit-H4). HS lesional skin was characterized by the presence of NET structures that were not observed in controls (Fig. 1B). To corroborate the presence of NETs in HS skin, we analyzed homogenized tissue lysates to detect cit-H3/DNA complexes by enzyme-linked immunosorbent assay (ELISA). Complexes of cit-H3/DNA were significantly (P = 0.029) increased in HS skin when compared to healthy controls (Fig. 1C); this was validated by Western blot to detect cit-H3 (Fig. 1D). There was a significant positive correlation between amount of NETs disgorged within lesions and HS severity index (Hurley stage; r2 = 0.7537, P < 0.0001 with 95% confidence) (Fig. 1E). To assess whether enhanced NET formation is confined to the skin, HS peripheral blood neutrophils were analyzed ex vivo. These neutrophils displayed enhanced NET formation in the absence of exogenous in vitro stimulation when compared to control neutrophils (Fig. 1F). This difference is also in the absence of clinically assessed infectious etiologies. Serum NET complexes were not found increased in HS (fig. S1, A and B); however, HS serum failed to properly degrade healthy control NETs in vitro (fig. S1C), suggesting that NET clearance mechanisms may be impaired in the circulation of these patients. Together, these results indicate that NET formation is enhanced in HS, both systemically and in lesional skin, in association with disease severity and that NET degradation mechanisms may be impaired in this disease.

Fig. 1 NETs are present in HS lesions and correlate with disease severity.

(A) Biopsies from HS patients stained with H&E (upper panel, neutrophils indicated with arrowheads) and immunofluorescence staining against myeloperoxidase (MPO; green) and Hoechst (DNA, blue) (lower panel). Epi, epidermis; Derm, dermis; HF, hair follicle. Scale bars, 50 μm. (B) Representative confocal images of lesions from HS patients stained for citrullinated histone H4 (cit-H4, red) and Hoechst (DNA, blue). Arrows show colocalization of cit-H4 and DNA. Scale bars, 100 μm. (C) Homogenized lysates from resected tissues from various Hurley stages were tested for cit-H3 and DNA complexes. Results are the mean ± SEM. *P < 0.05, Mann-Whitney U test. (D) Cit-H3 protein abundance was evaluated in tissue from HS patients and assessed by Western blot. (E) Correlation of cit-H3 and DNA complexes with HS severity as determined by the Hurley stage. One-way ANOVA Brown-Forsythe test analysis was used. (F) NETosis of peripheral blood neutrophils from HS patients and control neutrophils. MPO is red, and DNA is blue. Scale bars, 50 μm.

Autoantibodies to NET components and to citrullinated proteins develop in HS

Given that HS neutrophils display enhanced NET formation, we hypothesized that individuals affected by this condition would display autoantibodies to NET components. Purified, ionophore-generated NETs from healthy donors were probed with control or HS sera. Western blot analysis showed differential recognition of peptides by HS serum immunoglobulin G (IgG) when compared to control IgG (Fig. 2A), suggesting that HS serum recognizes autoantigens present in NETs and the putative presence of autoantibodies in HS. Total IgG was significantly increased in HS sera (P = 0.0011) when compared to sera from healthy volunteers (Fig. 2B). HS skin displayed IgG deposition in the epidermis, as well as infiltration by IgG-positive plasma cells (Fig. 2C), suggesting dysregulation in antibody responses. Flow cytometry analysis of HS peripheral blood mononuclear cells (PBMCs) demonstrated significantly increased frequency of plasma cells (P = 0.0424) and increased B cells when compared to controls. In addition, activated CD86+ plasma B cells (P = 0.0095) were significantly elevated in HS, supporting dysregulation of the B cell compartment in this disease (Fig. 2, D to I). TNFα is increased in HS serum, and anti-TNFα therapy has been reported effective in some HS patients (25). Anti-TNFα treatment can promote autoimmunity and autoantibody development (26). However, the cohort of HS patients we studied was not prescribed or had discontinued use of any anti-TNFα agents at least 2 to 4 weeks before sample collection. Furthermore, there was no correlation between use of anti-TNFα treatment and B cells, activated B cells, plasma cells, IgG, or autoantibody quantities (table S1).

Fig. 2 Antibodies recognizing NET antigens and citrullinated peptides are present in HS serum.

(A) NETs were resolved in a SDS–polyacrylamide gel electrophoresis gel and transferred onto a nitrocellulose membrane. Membrane was probed and incubated with serum from control or HS patients. Autoantibodies were detected using anti-IgG secondary antibody. Detection of MPO and cit-H3 were used as loading controls for NETs. (B) ELISA of total IgG in HS, RA, and control (Ctrl) samples. (C) Representative images of control and HS tissue stained for IgG immunohistochemistry within the epidermis and IgG-positive plasma cells (arrowheads); nuclei and DNA were stained with hematoxylin (blue). Scale bars, 50 μm. Peripheral blood from control and HS patients was analyzed for the frequency and activation status of (D and E) plasma cells, (F and G) memory B cells, and (H and I) CD19+ B cells by flow cytometry. MFI, mean fluorescence intensity. Results are the mean ± SEM. *P < 0.05, **P < 0.01, Mann-Whitney U test.

Because NETs contain citrullinated proteins that can serve as autoantigens in RA (10, 17), we next tested whether HS patients develop autoantibodies to citrullinated proteins. Clinical testing to detect antibodies to citrullinated peptides [anti-cyclic citrullinated peptide (anti-CCP) test] was performed in HS sera. Although anti-CCP positivity, particularly at high titers, is considered highly specific for RA (2729), HS sera also showed significantly higher titers of anti-CCP antibodies (P = 0.0127) when compared to healthy control sera, albeit lower than the titers found in RA (Fig. 3A). To further corroborate this finding, we used a bead-based assay against citrullinated antigens that has been previously described (30). An antigen bead assay assessed the presence of autoantibodies to native and citrullinated proteins in HS serum. HS sera displayed significantly elevated antibodies recognizing citrullinated vimentin 1–16 (P = 0.0078), citrullinated fibrinogenA 27–43 (P = 0.0329), citrullinated filaggrin 48–65 (P = 0.0326), and citrullinated enolase-1A 5–21 (P = 0.0006) (Fig. 3, B to E) when compared to control sera. HS sera also showed more antibodies against citrullinated H2A, H2B, tenascin, ApoA1 231–248, and ApoE 277–296 (Fig. 3F). Unsupervised clustering of autoantibodies to citrullinated proteins demonstrated distinct enrichment in association with specific Hurley stages. Autoantibodies against citrullinated fibronectin and citrullinated filaggrin were more prevalent in Hurley stage 1 and decreased in Hurley stages 2 and 3. In contrast, autoantibodies against citrullinated ApoE, citrullinated vimentin, and citrullinated fibrinogenA 211–230 were more prevalent in Hurley stage 2 compared to stages 1 and 3. Hurley stage 3 was characterized primarily by autoantibodies against H2B and tenascin C1 (Fig. 3G). Together, HS is characterized by increased plasma cells and IgG and by the development of autoantibodies against citrullinated cellular and matrix proteins, some of them previously reported to be present in neutrophils and NETs.

Fig. 3 Autoantibodies against citrullinated antigens are present in HS patients.

(A) Control (Ctrl) (n = 7), HS (n = 21), and RA (n = 10) sera were tested for the presence of autoantibodies against cyclic citrullinated peptides (anti-CCP) using a commercially available kit. Control sera were used as negative control, and RA sera were used as positive control. Results are the mean ± SEM. *P < 0.05, ***P < 0.001, Mann-Whitney U test. (B) Control (n = 9), HS (n = 16), and RA (n = 9) sera were analyzed for the presence of antibodies recognizing citrullinated vimentin, (C) citrullinated fibrinogen, (D) citrullinated filaggrin, and (E) citrullinated enolase using a bead-based assay. Results are the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, Mann-Whitney U test. (F) Heat map of the elevated antibodies recognizing citrullinated peptides present in HS sera (n = 16) when compared to control sera (n = 10). RA sera (n = 9) were used as control. (G) Heat map of unsupervised clustering of citrullinated autoantibodies stratified by Hurley stage (stage I, n = 4; stage II, n = 5; stage III, n = 8).

Enzymes involved in citrullination are elevated in HS skin

PADs are enzymes involved in the conversion of arginine to citrulline. There are five PAD isoforms, and two of them have been well studied in myeloid cells, PAD2 and PAD4 (10, 31). As histone citrullination was increased in HS skin and both of these PADs can citrullinate histones, we hypothesized that PADs can increase protein citrullination in HS skin. In addition to myeloid-relevant PADs, PAD1 and PAD3 have been reported to be expressed in normal skin (32) and no difference in PADI1 or PADI3 gene expression was detected in HS skin when compared to controls (fig. S2A). In contrast, PADI2 gene (P = 0.0158) was significantly higher in comparison to control skin (Fig. 4A). Likewise, protein expression was higher in HS skin when compared to control skin (Fig. 4B). Although we were not able to detect PADI4 gene expression in HS skin (fig. S2A), PAD4 protein was higher in HS when compared to healthy skin (Fig. 4B). We then assessed the activity of the PAD enzymes in HS skin by using a probe that specifically recognizes citrullinated proteins. Confirming the observations of enhanced histone citrullination, this citrulline probe assay demonstrated that HS skin contains a number of increased citrullinated proteins when compared to control skin (Fig. 4C). The pattern of citrullination found in HS skin was comparable to the pattern generated with recombinant PADs 1 to 4, suggesting that PADs are involved in the generation of citrullinated proteins in HS skin. Of note, immunofluorescence of HS skin showed differential expression of PADs compared to control skin (Fig. 4, D to F). PAD1 was localized throughout the epidermis in control skin, whereas this expression was more basal in HS lesions. Furthermore, compared to control skin, PAD2 appeared to be more localized to the lower stratum corneum and granular layer and PAD3 appeared to be more localized to the lower stratum spinosum and basal layer in HS. Together, these results suggest that PADs are active in HS lesional skin and may contribute to enhanced citrullination of autoantigens detected in HS lesions, potentially promoting the formation of autoantibodies targeting citrullinated peptides.

Fig. 4 PAD expression and activity are increased in HS lesions.

(A) PADI2 qPCR analysis of lesions from HS patients and control tissue. *P < 0.05, Mann-Whitney U test. (B) PAD2 and PAD4 protein analysis in tissue from HS patients as assessed by Western blot. Tubulin was used as a loading control. Densitometry of PAD2/Tub and PAD4/Tub (lower panel) (C) PAD activity was measured using Rh-PG probe against citrulline in HS and control tissue. Control tissue with recombinant PADs 1 to 4 was used as a positive control for substrates for citrullination pattern detection. Representative confocal images of NETs (cit-H4, red) and (D) PAD1 (green), (E) PAD2 (green), and (F) PAD3 (green) in HS and control tissues. Arrows indicate coexistence of NETs and PADs. Hoechst was used to counterstain nuclei in blue. Scale bars, 100 μm.

PADs are increased in the lungs of smokers (33), and smoking is a major risk factor for HS. The majority of the HS patients analyzed in our cohort were current or former smokers; therefore, we cannot exclude that tobacco use could have contributed to increased PAD expression in these patients. However, we did not detect differential PAD expression based on smoking status of patients with HS (β = −3.415; P = 0.6628, SD = 7.7) (fig. S2B). Likewise, because obesity is a known risk factor for HS, we assessed whether body mass index (BMI) correlated with enhanced NET formation and disease severity. Although the majority of the patients are overweight or obese, there was no direct correlation between elevated BMI and disease severity (β = −0.068; P = 0.353, SD = 0.073) (fig. S2C).

Type I IFN pathway is activated in HS lesional skin

IFNs can amplify innate and adaptive immune responses, and previous studies have demonstrated that NETs can promote induction of type I IFN signatures in skin, kidney, and peripheral blood in various autoimmune and chronic inflammatory conditions (3438). We tested whether HS patients display a type I IFN gene signature in skin by homogenizing HS lesional skin and isolating RNA and proteins for analysis. qPCR analysis showed significantly increased expression of type I IFN–regulated genes IFI44L (P = 0.0026), MX1 (P = 0.0044), CXCL10 (P = 0.0219), RSAD2 (P = 0.0428), and IFI27 (P = 0.0086) (Fig. 5A). To corroborate these findings, we performed Western blot analysis of homogenized skin lysates to determine the presence of activated key transcription factors involved in the type I IFN pathway, phosphorylated IRF3 (p-IRF3), and p-IRF7 (39, 40). As shown in Fig. 5B, enhanced expression of p-IRF3 and p-IRF7 was detected in HS skin, further supporting activation of the type I IFN pathway in HS lesions.

Fig. 5 An enhanced type I IFN signature is detected in HS lesions.

RNA was isolated from homogenized HS lesional tissue or control tissue. (A) qPCR analysis of HS lesions (n = 16) and control tissue (n = 5) for type I IFN genes IFI44L, MX1, CXCL10, RSAD2, and IFI27. (B) Protein analysis of key IFN regulators, IRF3 and IRF7, in HS tissue and control tissue. β-Actin was used as a loading control. (C) NETs and pDCs were visualized in HS tissue by staining cit-H4 (red) and CD303/BDCA-2 (green), respectively, showing coexistence of NETs and CD303-positive cells (arrows). Nuclei were stained blue. Scale bars, 100 μm. (D) PBMCs and (E) neutrophils from HS patients (n = 4) and control (n = 6) were analyzed for type I IFN–regulated genes. Results are the mean ± SEM. *P < 0.05, **P < 0.01, Mann-Whitney U test.

Western blot analysis also showed enhanced IFN-α protein expression in HS when compared to control tissues. As it is known that NETs can activate plasmacytoid dendritic cells (pDCs) to produce IFN-α (13, 38, 41), we tested whether pDCs and NETs may interact in HS skin. Confocal analysis demonstrated that NETs coexist with pDCs in HS skin (Fig. 5C), suggesting that NETs priming pDCs may be involved in the up-regulation of type I IFN responses detected in HS skin. To further address whether the type I IFN signature is a systemic or localized feature of HS, we performed a reporter assay to quantify IFN activity in the serum by incubating HeLa cells with HS and control sera for 6 hours. qPCR showed no differences between control and HS induction of IFN-inducible genes in the cell line, suggesting no enhanced type I IFNs in the circulation of HS patients (fig. S3). To further corroborate this finding, PBMCs and neutrophils from HS patients and controls were analyzed for a type I IFN gene signature. Although no differences were observed between HS and control PBMCs (Fig. 5D), CXCL10 (P = 0.0317) and IFI27 (P = 0.0190) genes were significantly elevated in the peripheral blood HS neutrophils when compared to controls (Fig. 5E). These results indicate that the type I IFN signature present in HS patients is primarily restricted to lesional skin and may also involve peripheral blood neutrophils.


Severe dermal inflammation with substantial neutrophil infiltration is characteristic of HS lesions (8). However, how neutrophils contribute to HS pathogenesis and their role in activating other immune cell types and amplifying inflammation in this disease had not been systematically investigated. We now report that HS neutrophils are primed to form NETs and that these structures associate with disease severity and progression and with the activation of a type I IFN response in the skin. Furthermore, we describe that HS sera contain autoantibodies targeting NET components and citrullinated proteins and increased circulating plasma cells.

A major histologic finding of HS lesions is the infiltration of various inflammatory cells (42). Infiltrating macrophages in HS lesions release TNFα, IL-12, and IL-23, which can lead to the production of IL-17 by T helper cells (9). Besides TH17 cells, innate cells may contribute to IL-17 synthesis in HS (8, 43). In turn, increases in IL-17 in HS lesional skin may promote neutrophil migration to the skin. It is possible that some of the cytokines increased in HS lesions promote the formation of NETs. Previous studies in other inflammatory conditions have proposed a role for NETs as autoantigens and amplifiers of inflammation and innate and adaptive immune responses (10). We now provide evidence that enhanced NET formation in HS externalizes autoantigens that are recognized by HS serum antibodies. Some of the antibodies recognizing citrullinated peptides that were detected in the serum of HS patients are considered important in the pathogenesis of RA (10, 17). In addition, the enhanced NET response exhibited by HS circulating and skin neutrophils may contribute to enhanced inflammation and activation of other immune cells, including skin pDCs, thereby promoting an enhanced type I IFN response in HS lesions.

Inadequate NET clearance has been proposed to play a role in enhancing the half-life of NET autoantigens in autoimmune diseases such as SLE (12, 19, 20) and appears to be a mechanism present in HS, at least in vitro. NET complexes in circulation were not enhanced in HS and no enhanced systemic type I IFN response was detected, except for increased CXCL10 and IFI27 in the HS neutrophils. In contrast, enhanced NET formation in HS skin may promote localized immune dysregulation and enhanced skin type I IFN responses. The discrepancy between the lack of enhanced NET complexes in circulation with evidence of enhanced in vitro NET formation by HS neutrophils, along with impaired in vitro NET degradation by HS serum, remains to be determined. One possibility is that the degree of NET formation in the circulation of HS patients remains low enough that, even with impairments in the degradation of these complexes, no significant elevations are observed in serum. This is in contrast to HS skin tissue where the degree of NET formation, along with the enhanced proinflammatory milieu in the skin, may enhance the half-life of these structures and promote immune dysregulation. It is also possible that some of the neutrophils isolated from the peripheral blood of HS patients are more prone to form NETs in vitro and show enhanced type I IFN responses because they were previously exposed to the HS lesional skin milieu and then undergone reverse transmigration (44) back to the circulation and on their way into the bone marrow.

B cells and plasma cells are present in chronic HS lesions (42, 45). IL-10 is up-regulated in HS (46), and this cytokine can influence B cell differentiation to plasma cells and antibody production (47). It is possible that enhanced generation of autoantigens in HS lesional skin and peripherally, through increased NET generation, could play a role in activating B cells, generating plasma cells, and promoting an aberrant adaptive immune response. This is also supported by the identification of elevated IgG in HS patients, deposition of IgG in HS lesions, and increased plasma cells that we identified in HS. The role of the autoantibodies detected in HS in disease pathogenesis, as well as dependence on T cells, remains to be determined. Autoantibodies to ApoA1 and ApoE were detected in the serum of HS patients. Emerging data demonstrate the potential role of anti–apolipoprotein A1 in atherothrombosis (48), which could have implications for the described increased risk for vascular disease in HS (49).

One of the known hallmarks of RA pathogenesis is the preferential presentation of citrullinated peptides to the adaptive immune system and the development of autoantibody responses to citrullinated peptides, including autoantibodies that recognize antigens present in NETs (17, 50, 51). Of interest, autoantibodies recognizing citrullinated H2A, H2Bcit, α-enolase, vimentin, and clusterin, considered specific to RA (17), were also detected in HS subjects. Furthermore, PADs 1 to 4 were present in HS skin, with PAD2 and PAD4 proteins being more abundant in HS skin. Of interest, there was differential skin distribution of PAD 1 to 3 proteins when compared to controls. As PAD1 and PAD3 can citrullinate extracellular antigens, we cannot exclude their contributions to the enhanced citrullination of proteins in HS skin, in addition to the putative role of myeloid PAD2 and PAD4. Considering the complexity of chronic inflammation and the infiltration of multiple cellular types in HS lesions, it is possible that other cells besides neutrophils also contribute to the heightened citrullination found in HS skin through the activation of PAD2 and perhaps other PADs. PAD2 is expressed by different inflammatory cells, including macrophages (52), which are increased in HS lesions (9, 42, 53).

An increase in type I IFN responses was detected in HS lesional skin but not in PBMCs, although isolated neutrophils showed an increase in CXCL10 and IFI27, suggesting a possible role for IFN-γ. We currently cannot exclude the contributions of other IFNs to the HS signature. We hypothesize that enhanced NET formation may contribute to pDC activation in HS skin and this is partially supported by observing that netting neutrophils and pDCs are present in same areas in HS skin. NETs have been shown to activate pDCs to produce abundant type I IFN in SLE in a DNA- and TLR7/9 (Toll-like receptor 7/9)–dependent manner (37). Whether other pathways driven by the presence of skin microbes or by the production of type I IFNs by other skin cell types play important roles in immune dysregulation in HS is unknown.

These data were generated from the analysis of mostly African American patients, a cohort largely understudied in HS. It is not known whether there are pathogenic differences across ethnicities in HS to confirm whether these observations are recapitulated in these other groups. Other limitations in this study include the relatively small sample size, the need to further understand in future experiments the role that neutrophil/type I IFN dysregulation plays in the clinical manifestations of the disease, and how these abnormalities in immune responses regulate other innate and adaptive immune cell types in HS.

Overall, we have characterized a putative axis of immune dysregulation characterized by enhanced NET formation and type I IFN responses associated to autoantibody generation in HS. This could potentially dictate the development of new therapeutics for this devastating condition.


Study design

The study investigated the role of neutrophils and NETs in promoting immune dysregulation in HS. HS lesions from patients who underwent surgery were collected or tissue from healthy control patients after abdominoplasties and mammoplasties, as explained in the next section. Pilot studies were done to test our hypotheses as reported herein; subsequently, the number of patients was increased to corroborate the findings and for statistical analysis. Other biospecimens (blood, serum, etc.) were obtained from consented patients as needed. Patients who were diagnosed clinically and correlated with histopathologic characteristics of HS (42) were included. Patients younger than 16 years of age were excluded. None of the patients were on immunosuppressants, and immunosuppressive therapy had been stopped at least 2 to 4 weeks before tissue collection. No data were excluded. There were at least three biological replicates for all experiments. The number of patients included is indicated in the figure legends and was determined, in part, by the number of samples available at specific time points. Symbols denoted in the figures represent one patient or biological replicate. Investigators were blinded for the Hurley staging of the patients during experiments. Once the data were obtained, the Hurley stage (disease severity) was known for interpretation of the data. Primary data are reported in data file S1.

Collection of tissue samples

The study was approved by the Johns Hopkins University Institutional Review Board, and informed consent was obtained. Normal (non-HS) and lesional HS skin were obtained from surgical resections during cosmetic surgery and from Johns Hopkins’ tissue bank obtained from patients undergoing surgical removal of HS skin. Patients with Hurley stages 2 and 3 were considered to have chronic HS, i.e., sinus tracts and fibrotic tissue. Visual assessment of the patient at the time of surgery as well as clinical and histopathological diagnosis were verified. Each sample was collected and stored in liquid nitrogen for transport. Samples were then embedded in optimal cutting temperature (OCT) media and placed at 20°C. Once solidified, 5-8 μm cryostat sections were placed on Superfrost Plus slides (Thermo Fisher Scientific) and placed at 20°C until immunofluorescent staining was done as described below. Some skin sections were fixed in 10% formalin and embedded in paraffin. H&E staining was performed for histological characterization.


Rabbit polyclonal antibodies used were anti–IFN-α (1:500, Abcam), anti–p-IRF7 Ser471/472 (1:500, Cell Signaling), anti–β-actin (1:1000, Abcam), anti-PAD1 (1:100, Proteintech), anti-PADI2 [1:500 dilution (Western blot); 1:200 (immunofluorescence), Proteintech], anti-PAD3 (1:200, arigo Biolaboratories), anti-MPO (1:1000, Dako), and anti-H3 (1:500, Abcam). Monoclonal rabbit anti–p-IRF3 Ser386 (1:500, Abcam) and mouse anti-PADI4 clone 6A2 (1:500 dilution, Lifespan Biosciences) were also used. CD303 (BDCA-2) fluorescein isothiocyanate (FITC)–coupled (1:100) was from Miltenyi. The secondary antibodies purchased from LI-COR and used at a 1:10,000 dilution were donkey anti-mouse 800, donkey anti-mouse 680, donkey anti-rabbit 800, and goat anti-rabbit 680. Anti-CCP analysis was done by the Clinical Immunology Laboratory in Johns Hopkins Hospital, following established chemiluminescent immunoassay protocols (QUANTA Flash CCP3; Inova Diagnostics).

NET isolation

NETs were isolated, as previously described (16). Briefly, HS neutrophils were plated in 24-well tissue culture plates in RPMI without phenol red. Cells were incubated for 3 hours at 37°C. Supernatants were harvested, and NETs were digested with micrococcal nuclease (10 U/ml; Thermo Fisher Scientific) for 15 min at 37°C. Supernatants were collected and centrifuged at 5000 rpm for 5 min at 4°C. NET supernatants were transferred to a fresh tube and stored at −80°C until used.

Detection of NETs, PADs, and IgG in skin

To detect NETs in skin, frozen sections were blocked with 10% bovine serum albumin (BSA) for 1 hour at room temperature. Where indicated, anti-PAD1 (1:100, Proteintech), anti-PAD2 (1:200, Proteintech), anti-PAD3 (1:200, arigo Biolaboratories), anti-MPO (1:1000, Dako), and anti-human histone H4 (citrulline 3) rabbit polyclonal antibodies (1:1000) (EMD Millipore) diluted in 5% BSA were added to sections and placed at 4°C overnight. Each slide was washed three times with phosphate-buffered saline (PBS), and Alexa Fluor 555 donkey anti-rabbit IgG (Invitrogen) secondary antibody (1:400) was added and incubated for 1 hour at room temperature. Nuclei were counterstained with (1:1000) Hoechst for 10 min at room temperature followed by five more washes with PBS. Tissue sections were sealed with a coverslip containing ProLong Gold solution (Thermo Fisher Scientific). Images were acquired on a Zeiss LSM780 confocal laser-scanner microscope. IgG immunohistochemistry was done by the Johns Hopkins Medical Institutions (JHMI) Immunopathology Core.

qPCR analysis in skin and PBMCs

Tissue sections (3 to 5 mm) were homogenized using liquid nitrogen, mortar, and pestle. One-half of the homogenized tissue was resuspended in 500 μl of TRI Reagent (Sigma-Aldrich). After 30 min, solution was centrifuged for 10 min at 14,000 rpm, and supernatant was transferred to a fresh Eppendorf tube. RNA was isolated using the Direct-zol RNA MiniPrep Kit (Zymo Research) according to the manufacturer’s instructions. RNA from HS and control PBMCs was isolated through the same method. Total RNA (500 ng) was reverse-transcribed using iScript RT single-strand complementary DNA (cDNA) (Bio-Rad). qPCR was performed using the TaqMan Gene Expression Master Mix (Thermo Fisher Scientific), human GAPDH primers (Hs99999905_m1), and sequence-specific primers for IFI44 (Hs00197427_m1), IFI44L (Hs00915292_m1), IFI27 (Hs01086373_g1), RSAD2 (Hs00369813_m1), CXCL10 (Hs00171042_m1), HERC5 (Hs00180943_m1), MX1 (Hs00895608_m1), IFNA1 (Hs04189288_g1), IFNB1 (Hs01077958_s1), PADI1 (Hs00203458_m1), PADI2 (Hs01042505_m1), PADI3 (Hs01042496_m1), and PADI4 (Hs01057483_m1). Fold difference was calculated using the ΔCt equation.

Serum type I IFN activity assay

HeLa cells were cultured in a humidified incubator with 5% CO2 at 37°C in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum, 2 mM glutamine, penicillin (100 U/ml), and streptomycin (100 μg/ml). The cells were plated in a 24-well plate and treated with 50% HS or control serum for 24 hours. Cells were washed twice in PBS and suspended in TRI Reagent (Sigma-Aldrich). RNA was extracted, cDNA was prepared, and qPCR was performed as described above.

Cit-H3/DNA complex ELISA

Cit-H3/DNA NET complexes were quantified by ELISA. A 96-well plate was coated with rabbit polyclonal anti–cit-H3 (Abcam) at 1:400 in PBS overnight at 4°C. Wells were washed three times in washing buffer (0.5% Tween in PBS) and blocked in blocking buffer (1% BSA in PBS) at room temperature for 1 hour. After washing twice, 5 μg of total protein isolated from HS or control skin was added to the wells in blocking buffer and incubated overnight at 4°C. The wells were washed three times and incubated with mouse monoclonal anti–double-stranded DNA antibody (EMD Millipore) at 1:100 in blocking buffer. After washing three times, goat anti-mouse–conjugated horseradish peroxidase antibody (Bio-Rad) was added to the wells in blocking buffer at 1:10,000. Wells were washed five times followed by the addition of TMB substrate (Sigma-Aldrich) and stop solution (Sigma-Aldrich). The absorbance was measured at 450 nm, and values were calculated as an optical density index. Assay was performed in duplicate.

Citrullinated autoantibody bead-based assay

As previously described (54), serum anti-citrullinated protein antibodies targeting 40 RA-associated autoantigens were measured using a bead-based immunoassay on a Bio-Plex platform. Serum was diluted to a 1:30 ratio in a proprietary sample dilution buffer (Bio-Rad), mixed with the antigens conjugated to spectrally distinct fluorescent microspheres (Bio-Rad), and then incubated with an anti-human phycoerythrin (PE)–labeled antibody (Jackson ImmunoResearch). The resulting fluorescence intensities were analyzed on a Luminex 200 platform (Luminex Corporation).

Serum-induced degradation of NETs

This was performed as previously described (55). Briefly, control neutrophils were resuspended in unsupplemented RPMI (1 × 106 cells/ml) and stimulated with phorbol 12-myristate 13-acetate (PMA) (500 ng/ml) for 4 hours at 37°C to induce NETs. After stimulation, PMA-treated neutrophils without serum were fixed with 4% paraformaldehyde (PFA) and stored at 4°C. Neutrophils were treated with 1% serum from either healthy controls, HS subjects, or RA subjects for 16 hours at 37°C. Neutrophils were fixed with 4% PFA and stored at 4°C before immunofluorescence was done to visualize NETs.

Western blot

Tissue sections (3 to 5 mm) were flash-frozen with liquid nitrogen. Frozen skin was ground with pestle and mortar. One-half of the homogenized tissue was then resuspended in 500 μl of lysis buffer [50 mM tris-HCl (pH 7.4), 300 mM NaCl, 0.5% (w/v) Triton X-100, 5 mM EDTA] enriched with a cocktail of protease inhibitors (Roche) (16) and placed at 4°C on a rotator for approximately 1 hour. The supernatants were transferred to a fresh Eppendorf tube. Total protein was quantified using bicinchoninic acid (BCA) protein assay (Thermo Fisher Scientific). Equal amounts of total protein were resolved in a 4 to 12% gradient bis-tris gel (Invitrogen), transferred onto a nitrocellulose membrane, and blocked with 10% BSA for 30 min at room temperature. After overnight incubation with primary antibodies, membranes were washed three times with PBS-Tween (PBS-T) and incubated with secondary antibody coupled to IRDye 800CW. Membranes were developed using a Li-COR Odyssey Clx scanner.

Immunophenotyping of B cells by flow cytometry

After Ficoll centrifugation, HS and matched healthy control PBMCs were obtained and counted, and 2 × 106 cells were resuspended in fluorescence-activated cell sorting (FACS) buffer (PBS containing 0.5% BSA) and spun at 1500 rpm for 5 min. Cells were resuspended in Fc receptor blocking solution (Human TruStain FcX, 1:100 in FACS buffer) for 10 min on ice. Cells were then spun at 1500 rpm for 5 min and resuspended in a cocktail of fluorescently labeled antibodies and stained for 30 min on ice. Single stains of UltraComp eBeads (Invitrogen) were used as compensation controls. Cells were washed twice with FACS buffer, spun at 1500 rpm for 5 min, and then fixed with 2% PFA (Chem Cruz) for 15 min on ice. Fluorescently labeled antibodies were obtained from BioLegend unless otherwise specified. The following antibodies were obtained for this study: CD19–Pacific Blue, CD27–allophycocyanin (APC)–Cy7, CD138-FITC, CD80-PE, CD69-AF700, CD86-APC, CD44-PE, CD25-AF700, and CD30-APC. Antibody concentrations were chosen on the basis of titration and guided by developer recommendations. Analysis of cells was performed using a FACSCanto II flow cytometer and FlowJo software (v10.3).

PAD activity using Rh-PG probe

Citrulline probe was used as previously described (16). Briefly, 10 μg of total protein from control and HS skin was incubated in the presence of 10 mM CaCl2 (Quality Biological) and 1 mM dithiothreitol (Sigma-Aldrich) overnight at 37°C. Samples were incubated with 100 μM citrulline–Rh-PG probe (Cayman) in 20% trichloroacetic acid for 30 min at 37°C. Human recombinant PAD1 (Cayman Chemical Company), PAD2 (Sigma-Aldrich), PAD3 (Cayman Chemical Company), and PAD4 (Sigma-Aldrich) were used as positive controls. Samples were washed with cold acetone twice and separated in a 4 to 12% gradient bis-tris gel (Invitrogen). Citrullinated proteins were visualized in an Azure c600 Imaging system (Azure Biosystems).

Statistical analysis

Data were analyzed using GraphPad Prism software. For samples with non-Gaussian distribution, we used Mann-Whitney U test. Where indicated, one-way analysis of variance (ANOVA) Brown-Forsythe test analysis was used. Results are presented as the mean ± SEM.


Fig. S1. Analysis of neutrophil stimulation.

Fig. S2. Analysis of PAD expression and BMI in HS.

Fig. S3. Type I IFN serum activity is not increased in HS serum.

Table S1. Correlation analysis of anti-TNF treatment with B cells and autoantibodies.

Data file S1. Primary data.


Acknowledgments: We thank A. Chien, L. Garza, F. Succaria, D. William, J. Zarif, R. Miller, R. Barnes, N. Viloria, the JHMI Reference Histology and Immunopathology Cores, and the Clinical Immunology Laboratory in Johns Hopkins Hospital. Funding: This study was supported, in part, by the Intramural Research Program, NIAMS/NIH, ZIA AR041199. A.S.B. was a participant in the Johns Hopkins Ethnic Skin Fellowship, which was supported, in part, by a grant from Valeant Pharmaceuticals. The company was not involved in the conception of the research idea, conduct of the experiments, and the writing of this manuscript. This study was also supported, in part, by the Johns Hopkins School of Medicine Physician Scientist Training Program (PSTP) and the Danby Hidradenitis Suppurativa Foundation Research Grant. Author contributions: C.C.-R., A.S.B., L.J.O., P.M.C., A.Z.R., and C.C. performed the experiments; C.C.-R., A.S.B., L.J.O., and P.M.C. analyzed the data and performed statistical analyses; C.C.-R., A.S.B., P.M.C., C.C., A.Z.R., M.L.K., J.S.R., S.K., L.S.M., W.H.R., G.A.O., and M.J.K. provided scientific input and were involved in overall design and/or manuscript preparation. A.S.B., C.C.-R., and M.J.K. drafted the manuscript; J.A.C., S.M.M., J.M.S., O.A., and K.P.B. surgically resected the analyzed skin sections and provided clinical and scientific input. Competing interests: All unrelated to the work reported and had no impact or conflicts related to this manuscript: A.S.B. and G.A.O. were formerly investigators for Eli Lilly. J.M.S. is a consultant/speaker for allergen and co-founder of LifeSprout. L.S.M. is a paid consultant for AstraZeneca, Janssen Research & Development, and Armirall and has received stock from Noveome Biotherapeutics, which are developing therapeutics against infections, inflammatory, and autoimmune diseases. L.S.M. serves on the scientific advisory board for Integrated Biotherapeutics, which is developing vaccines against pathogens. A.Z.R. has NDAs with Pliant Therapeutics and xMD Diagnostics and is on the Advisory Board for Escala Therapeutics. L.S.M. has received grants from AstraZeneca, MedImmune, Pfizer, Regeneron Pharmaceuticals, Boehringer Ingelheim, and Moderna Therapeutics, which are developing therapeutics against infections, inflammatory, and autoimmune diseases. A.S.B. received honorarium for speaking at the 77th Annual Society for Investigative Dermatology meeting—AbbVie Sponsored Symposium. All other authors declare that they have no potential conflicts of interest. Data and materials availability: All data associated with this study are in the paper or the Supplementary Materials. HS biospecimens are available from Johns Hopkins University School of Medicine under a material transfer agreement with the University.
View Abstract

Stay Connected to Science Translational Medicine

Navigate This Article