SHP2 promotes NETosis through ERK5 pathway and mediates the development of psoriasis

SHP2 expression by neutrophils is related to develop psoriasis

We collected skin tissue from healthy donor and psoriasis patient, isolated them into single cells and performed single cell RAN-seq to explore the expression of PTPN11. We firstly analyzed the expression levels of 107 genes which encoding human PTP phosphatase family members (Supplementary Figure 1), and showed the top ten in Figure 1a. Results showed that the SHP2 coding gene PTPN11 was highly expressed in skin tissue with psoriasis.

• Download figure
• Open in new tab

Figure 1. SHP2 expression by neutrophils is indispensable in skin tissue of psoriasis.

(a) The violin plot of the top ten expression abundance of 107 PTPs. (b)Neutrophil infiltration was determined by evaluating the intensity of MPO staining of healthy (n=8) and psoriasis patients (n=13) donors, scale bar: 400 μm. (c) Representative immunofluorescence images staining for MPO (green), SHP2 (red) and DAIP (blue) in healthy donors (n=8) and psoriasis patients (n=13) skin section, scale bar: 100 μm.

To determine the function of neutrophils in the pathogenesis of psoriasis, we performed immunohistochemical staining of the skin from patients with psoriasis and healthy donors. Results showed that the infiltration of neutrophils in psoriasis lesions was significantly higher than that in healthy donors (Figure 1c). The results in Figure 1d showed that the expression of SHP2 in neutrophils increased significantly, and had a strong colocalization expression with MPO in the skin lesions of psoriasis patients.

The deficiency of SHP2 in neutrophils improves psoriasis-like phenotype in mice model

To further study the role of SHP2 in psoriasis, we crossed Shp2^flox/flox mice with S100a8cre mice and constructed Cre-driven SHP2 deletion in neutrophils. S100a8^cre-Shp2^flox/flox (SHP2^N KO) mice, and Shp2^flox/flox (SHP2^N WT) mice. The results showed that a series of psoriasis symptoms caused by IMQ on the back of SHP2^N WT mice, such as redness and swelling, scaly coverage, and thickening of the epidermis, were significantly improved in SHP2^N KO mice (Figure 2a, b). The adapted human clinical Psoriasis Area and Severity Index (PASI) score of the IMQ-treated SHP2^N KO mice was also significantly lower than IMQ-treated SHP2^N WT mice (Figure 2c). Accordingly, the increase of epidermal thickness caused by IMQ was also alleviated in SHP2^N KO mice (Figure 2d). In addition, the expression of inflammatory factors TNF-α, IL-1β, IL-6, IL-17a, and CXCL15 were significantly reduced in SHP2^N KO mice skin (Figure 2e). We injected SHP099 (10 mg/kg) which is an effective SHP2 inhibitor into mice intraperitoneally, and the results were also in line with the previous experimental conclusions: after inhibiting SHP2, the symptoms of psoriasis induced by IMQ were improved (Supplementary Figure 2). In summary, our data indicate that the deletion of SHP2 in neutrophils or inhibition can alleviate IMQ-induced psoriasis symptom.

• Download figure
• Open in new tab

Figure 2. SHP2 deficiency in neutrophil alleviates psoriasis-like phenotype in the IMQ-induced murine model.

(a) Phenotypic presentation of back skin sections of sham- or IMQ-treated SHP2N KO and WT mice. (b) The hematoxylin and eosin (H&E) staining of back skin sections of sham- or IMQ-treated SHP2N KO and WT mice, scale bar: 200 μm. Clinical scores (c) or epidermal thickness (d) or of mice dorsal skin. (e) Dorsal skin was infiltered with TNF-α, IL-1β, IL-6, IL-17a and CXCL15 evaluated by ELISA.

NETs are produced in psoriasis patients and can affect SHP2

To explore the relationship between psoriasis and NETs, we collected the skin tissues of psoriasis patients and healthy donors, and detected CitH3 (Figure 3a) and PAD4 (Figure 3b) protein expression, which are 2 key proteins formed in the process of NETosis by immunohistochemistry. The results showed that the expression of CitH3 and PAD4 in skin lesions of psoriasis patients increased significantly compared with the skin of healthy donors. In addition, we used immunofluorescence method to obtain images of NE and MPO in skin sections to observe NETs. The results showed that more NETs were produced in the skin of patients with psoriasis (Figure 3c). Next, we quantified dsDNA which can reflect the generation of NETs from the serum of normal people and psoriasis. Results showed that the amount of dsDNA from psoriasis serum samples are sufficiently larger than that from normal people (Figure 3d). Finally, we analyzed genes’ mRNA expression which relates to NETs formation and PTPN11 in human skin tissue. Data from single cell RNA-seq showed that mRNA expression of ELANE and PADI4 increased with the increase of PTPN11 expression.

• Download figure
• Open in new tab

Figure 3. NETs are produced in psoriasis patients.

Immunohistochemical staining of skin sections from psoriasis patients and healthy donors with anti-PAD4 (a) or anti-CitH3 (b), scale bar: 400 μm. (c) Immunofluorescence staining of skin sections from psoriasis patients and healthy donors with anti-MPO, anti-NE, and DAPI, scale bar: 400 μm. (d) Quantification of dsDNA in serum samples from healthy and psoriasis donors. (e) Expression level of PTPN11, ELANE and PADI4 in human skin.

These data indicate that pathways involved in the formation of NETs are activated in skin tissues of psoriasis patients, and NETs are generated more if there are more SHP2 expression.

NETs can mediate the pathogenesis of psoriasis like symptom in mice

PAD4 protein has been reported to be closely related to the formation of NETs, and it was coding by Padi4 in mice. Next, we constructed Padi4 knockout (Pad4 KO) mice which were unable to form NETs. To confirm the efficiency of PAD4 deletion, we dissociated the skin of PAD4 KO mice and examined the expression of PAD4 using immunohistochemistry (Figure 4a).

• Download figure
• Open in new tab

Figure 4. PAD4 deficiency alleviates psoriasis-like phenotype in the IMQ-induced murine model, and led the growth of SHP2.

(a) The protein level of PAD4 in mice dorsal skin, tested by western blot. (b) Phenotypic presentation of sham- or IMQ-treated PAD4 KO and WT mice, scale bar: 50mm. (c) H&E staining of back skin sections of sham- or IMQ-treated PAD4 KO and WT mice, scale bar: 200 μm. Clinical scores(d) or epidermal thickness(e) of mice dorsal skin. (f) Dorsal skin was infiltered with TNF-α, IL-1β, IL-17a and CXCL15 evaluated by ELISA.

In the following experiment, PAD4 KO or WT mice were randomly divided into 2 groups, which treated with equal amounts of imiquimod or control cream for 3 days, and skin tissue was obtained on day 4. Compared with WT mice, the deletion of Padi4 significantly reversed the psoriasis-like symptoms caused by imiquimod. Scaling was significantly improved in PAD4 KO mice (Figure 4b). In addition, the epidermal hyperplasia was also significantly reduced proven by epidermal thickness and hematoxylin-eosin staining as we can see in Figure 4c, e. What’s more, the PASI score of IMQ treated WT mice continued to increase and peaked on day 4, while the PASI score of PAD4 KO mice continually kept on a low value, and decreased on day 4 (Figure 4d). Furthermore, we tested the expression of psoriasis-related inflammatory cytokines in mice skin. The results showed that Padi4 gene knockout effectively reversed the overexpression of inflammatory factors TNF-α, IL-1β, CXCL15 and IL-17a on the skin of mice caused by IMQ (Figure 4f). These data suggest that reduced level of NETs has an effective fact on ameliorating IMQ-induced psoriatic skin disorders. In conclusion, decreasing NETs can alleviate the development of IMQ-induced psoriasis mice model.

Single-cell RNA-seq reveals the inhibition of SHP2 can decrease neutrophils

We collected the skin tissues of mice which treated with IMQ or SHP099 or none, dissociated them into single cells and performed the single-cell RNA-seq (Figure 6a). Unbiased clustering revealed 5 cell clusters (Figure 6b), and we can see the cell distribution as shown in Figure 6c. According to specific marker genes, we defined these clustered cell populations as fibroblasts, immune cells, keratinocytes, Schwann cells and smooth muscle cells (Figure. 6d). Among them, immune cell subsets were defined as macrophages, T cells, neutrophils and natural killer T cells (Figure 6e). We found that the proportion of neutrophils in immune cells increased significantly after IMQ treatment, and decreased after SHP099 treatment (Figure 6f). Next, we analyzed neutrophil related genes. Results showed that neutrophil infiltration occurred in the skin tissues after IMQ treatment, and SHP099 significantly reversed this phenomenon (Figure 6g, h).

• Download figure
• Open in new tab

Figure 5. Single-cell RNA-seq reveals the inhibition of SHP2 can decrease neutrophils.

(a) Schematic representation of the experimental procedure. (b) Unbiased clustering reveals 5 cellular clusters. (c) Cell atlas of mice skin tissue. (d) Heat map image of functional gene set and cell atlas of each cell population. (e) Unbiased clustering of immune cells. (f) Pie chart of the proportion of neutrophils to immune cells. (g) Heat map image of neutrophil related functional gene sets; (h) Summit plot of neutrophil related functional genes.

• Download figure
• Open in new tab

Figure 6. Inhibition of SHP2 expression in neutrophils can inhibit the formation of NETs.

(a) The protein expression of NE and MPO in the IMQ or SHP099 treated mice. (b) The protein level of CitH3 and PAD4 of mice dorsal skin, tested by western blot. (c) The protein expression of NE and MPO of mice dorsal skin tested by ELISA. (d) Representative immunofluorescence images staining for MPO (green), NE (red) and DAIP (blue) of mice skin section, scale bar: 200 μm. (e) The protein expression of NE and MPO in neutrophils extracted from peripheral blood. (f) Representative immunofluorescence images staining for MPO (green), NE (red) and DAIP (blue) of neutrophils extracted from peripheral blood, scale bar: 100 μm.

Inhibition or deletion of SHP2 can inhibit the formation of NETs

In order to evaluate the effect of SHP2 in NETs formation, we firstly explored NE and MPO expression in IMQ or SHP099 treated mice. Results showed that the expression of protein NE and MPO were decreased by SHP2 inhibition (Figure 6a). Then, we used neutrophils specific SHP2 knockout mice to make some future exploration. ELISA and immunofluorescence were used to test the expression level of NE and MPO of mice dorsal skin. The results showed that IMQ significantly increased the expression of NE and MPO, and SHP2 specific knockout in neutrophils had revered this situation. The same trend was proved in the expression of CitH3 and PAD4 proteins detected by Western blot (Figure 6b, c, d). Next, we extracted neutrophils from peripheral blood of healthy donors, and then processed them with IMQ and/or SHP099. The results showed that IMQ treatment significantly increased the expression of NE and MPO, while SHP099 effectively inhibited the expression of these two proteins in a dose-dependent manner (Figure 6e, f). These results preliminarily confirmed that NETs which are involved in the progression of psoriasis could be down regulated by SHP2 inhibition or deletion.

SHP2 promotes the formation of NETs through the ERK5 pathway

To verify the regulatory relationship between SHP2 and NETs, we isolated neutrophils from human peripheral blood and treated the cells with IMQ or/and SHP099. RNA sequencing showed that: genes encoding proteins involved in the formation of NETs were highly expressed after IMQ treatment, such as MPO, Elane, and PADI4; Genes encoding inflammatory factor, such as IL1B and IL6 were highly expressed too after IMQ treatment, whereas SHP099 reversed the phenomenon. In addition to that, genes which relate to the formation of NETs were enriched[11], such as MAPK family, AKT1, AKT2, GSDMD and so on (Figure 7a). Previous studies have reported that the MAPK pathway is involved in the formation of NETs. Therefore, we used qPCR to explore four major MAPK family classifications: ERK, ERK5, JNK, p38 MAPK, which was encoded by MAPK1, MAPK7, MAPK8, MAPK14 respectively. We found that MAPK7 was significantly overexpressed after IMQ treatment (Figure 7b). Then, we treated neutrophils with IMQ or/and SHP099 for 30 min 60 min or 120 min. We found that SHP099 decreased the high expression of MAPK7 caused by IMQ, and the decrease was more significant with longer treating time (Figure 7c).

• Download figure
• Open in new tab

Figure 7. SHP2 promotes the formation of NETs through the ERK5 pathway.

(a) Heat map image of microarray analysis of RNA from human neutrophils (n=3/group). (b) Expression levels of representative MAPK family genes in neutrophils. (c) Expression levels of MAPK7 in neutrophils in different time limit. (d) Quantification of dsDNA in neutrophils. (e) Representative immunofluorescence images staining for MPO (green), NE (red) and DAIP (blue) of neutrophils, scale bar: 30 μm. (f) The protein expression of MPO in neutrophils tested by ELISA. (g) Quantification of dsDNA in neutrophils.

Next, we treated neutrophils with ERK5 inhibitor: AX-15836 and found that, the formation of NETs was inhibited which similar with the effect of SHP099 (Figure 7d, e). Finally, we treated neutrophils with lovastatin, which has been reported to elevate SHP2 activity in cells[12], and showed that lovastatin was able to induce a high expression of NETs, whereas after inhibition of SHP2 or ERK5, NETs were no longer produced in large amounts. This conclusion was supported by the MPO protein expression measured by ELISA, as well as dsDNA quantitative detection (Figure 7f, g). In conclusion, SHP2 may promote the formation of NETs through the ERK5 pathway.

Animal experiments

All the experimental animals were purchased from GemPharmatech Co. Ltd., and animal were maintained in specific pathogen-free conditions at GemPharmatech Co. Ltd and Experimental Animal Center at Nanjing University. Shp2 flox/flox mice were crossed with S100a8-cre mice on the C57BL/6 background for more than two generations to generate Shp2 flox/flox S100a8-cre mice which SHP2 was conditional knocked out in neutrophils. we also constructed Padi4 knockout on the C57BL/6 background.

Mice were shaved and treated with imiquimod (H20160079, 3M Pharmaceuticals, USA) or petroleum successively. For different experimental needs, 10 mg/kg SHP099, or physiological saline solution was injected intraperitoneally (i.p.). Serum was isolated from the blood of mice which anesthetized with 0.05 mg/kg pelltobarbitalum natricum, and followed by retro-orbital sinus taken. Skin and serum were stored at - 80 □ for future use. All the procedures were carried out in accordance with the Guide for the Care and Use of Laboratory Animals (National Institutes of Health, USA) and ethical regulations of Nanjing university to guarantee animal welfare.

Human specimens

Psoriasis patients and normal healthy donors’ skin tissue were taken to make paraffin sections. Sample acquisitions were approved by the Ethics Committee of West China Hospital. In process, we obeyed the declaration of Helsinki Principles.

PASI score

The changes of mice dorsal skin were observed and calculated according to the PASI scoring standard for the degree of erythema, scaling and infiltration. Then the points of the three items were summed to obtain a total score. The scoring standard refers to the “Severity Severity Scoring Table of Mouse Psoriasis Model” to compare the PAIS scores of the three groups of mice. Each index ranges from (0 to 4) points, 0 points: none; 1 point: mild; 2 Points: moderate; 3 points: obvious; 4 points: extremely obvious.

Neutrophils isolation

Ficoll-Hypaque density gradient centrifugation were used to isolate neutrophils from the peripheral blood of healthy donors. According to the instructions of Human Neutrophil Isolation Kit (LZS11131, Haoyang Biology, China), neutrophil separation solution of different concentration gradients has been already added in advance to the sterile silicified centrifuge tube. Peripheral blood and hydroxyethyl starch were mixed and then gently superimposed on the separation solution. After 20 min’s centrifugation at 800 g, neutrophils were pipetted out, followed by washing, centrifugation, and resuspend. Neutrophils were planted in 9 cm cell culture dishes with RPMI 1640 Medium (01-100-1A, Biological Industries, Israe), 10 % fetal bovine serum (2033119, Biological Industries, Israe) and 1% penicillin-streptomycin (CO222, Beyotime, China). Cells grew in an incubator at 37 □ with 5% CO₂.

Immunohistochemistry

The immunohistochemistry experiment was performed according to the instructions of immunohistochemistry detection kit (KIHC-5, Proteintech Group, USA). Skins were embedded in paraffin and made into 5 μm slides. Slides were treated with 5% sodium citrate buffer (P0081, Beyotime, China) and 2% hydrogen peroxide for antigen retrieval and endogenous peroxidase activity blocking. 3% goat serum was used to block nonspecific binding sites of antibodies. Then the slides were incubated with the primary antibody: rabbit anti-myeloperoxidase antibody (ab208670, Abcam, USA), mouse anti-neutrophil elastase antibody (sc-55549, Santa Cruz Biotechnology, USA) at 4 °C in a humidified chamber overnight. Next day, slides were treated with anti-mouse/rabbit immunohistochemistry detection kit (PK10006, Proteintech, USA). The hematoxylin was used to observe the nucleus. Finally, the slides were imaged by a light microscope (IX61, Olympus, Japan).

Immunofluorescence

Immunofluorescence staining was performed on tissue and cell samples. For tissue samples, slides were deparaffinized, rehydrated, and treated with sodium citrate buffer for antigen retrieval. Slides were incubated with 3% goat serum for 30 min followed by primary antibodies. Cell samples were fixed in poly-L-lysine polylysine (P2100, Solarbio, China), followed by permeabilization and nonspecific binding sites blocking. Samples were treated at 4 °C overnight with the following primary antibodies: rabbit anti-myeloperoxidase antibody (ab208670, Abcam, USA), mouse anti-neutrophil elastase antibody (sc-55549, Santa Cruz Biotechnology, USA), goat anti-shp2 antibady (ab9214, Abcam, USA), rabbit anti-pad4 antibody (ab214810, Abcam, USA), rabbit anti-histone H3 (citrulline R2) antibody (ab174992, Abcam, USA). Next day, Samples were incubated with an Alexa fluorescein-labeled secondary antibody for 2 h. DAPI (C1002, Beyotime, China) was used to stain cell nuclei. Then samples were imaged by an inverted confocal microscope (LSM880 with airyscan, Carl Zeiss, Germany).

Histopathologic assessment

Slides were deparaffinized, rehydrated, and treated with hematoxylin solution and eosin staining according to the instruction of Hematoxylin-eosin staining kit (G1005, Servicebio, China) to evaluate pathological changes in skin tissue. Then slides were dehydrated and sealed with neutral balsam.

Western blotting assay

Protein samples were prepared by tissue lysate and BCA protein assay. 15 μL protein of each sample was loaded and separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto polyvinylidene fluoride (PVDF) membranes (IPVH00010, Merck Millipore, Germany). The membranes were washed with 5% nonfat milk for 2 h to block the non-specific binding sites and then treated with primary antibody: rabbit anti-pad4 antibody (ab214810, Abcam, USA), rabbit anti-histone H3 (citrulline R2) antibody (ab174992, Abcam, USA), mouse anti–β-actin antibody (M20011, abmart, China) overnight at 4 °C. The secondary antibody was incubated with the membranes for 2 hours. Signals were detected and exposed by using LumiGLO^® Reagent (#7003, Cell Signaling Technology, USA).

dsDNA quantification

Quant-it PicoGreen dsDNA Assay kit (P7581, Thermo Fisher Scientific, USA) was used to quantify cfDNA in mice serum. Samples, TE buffer, and Quant-iT PicoGreen reagent were added into a 96-Well Black Opaque Plate. Fluorescence signals were detected by a microplate fluorescence reader (Safire, Tecan, Switzerland) with an excitation wavelength of 480□nm and an emission wavelength of 520□nm.

ELISA

The expression of IL-1β, TNF-α, IL-6, CXCL-15, NE, and MPO was detected by DuoSet ELISA kit (R&D System, USA). Briefly, the capture antibodies were coated in a 96-well enzyme linked immunoassay plate (Costar 3590, Corning Incorporated, USA) at first day in room temperature. After treated with washing buffer, skin lysate and serum were incubated in the plate for 2 h, then the detection antibodies were treated, absorbance at 450 nm was detected by a microplate (Safire, Tecan, Switzerland).

IL-17A.

Quantitative PCR

Trizol was used to extract total RNA (108-95-2, Takara, China). After quality control, we quantified 1 μg RNA, and reverse transcribed it to synthesize single stranded cDNA. We performed quantitative PCR for CFX 100 (bio rad, Hercules, CA) cyclers using the primers listed in Supplementary Table 1. The procedure of the amplification program was: 95 °C for 2.5 min, 95 °C for 15 s, 60 °C for 30 s, for a total of 44 cycles. Dissociation curves were analyzed at the end of amplification. RNA expression levels of GAPDH were used for normalization.

Statistical analysis

All data are expressed as the mean ± SEM and were analyzed by GraphPad Prism 8. 0. Student’s t test and one-way ANOVA test were used for statistical analyses of the data. Differences were considered significant at *P < 0. 05 or **P < 0.01.