miR-26a-5p inhibits the proliferation of psoriasis-like keratinocytes in vitro and in vivo by dual interference with the CDC6/CCNE1 axis

Screening of DEGs and DEmiRs

To ensure the accuracy of the interaction between miRNAs and their targets, we selected the GSE142582 series, in which the miRNA profile and mRNA profile of each group came from the same sample. The human skin samples included in the GSE142582 series were obtained from psoriatic patients and healthy volunteers who received no drug treatment within three months of recruitment. Disease severity was assessed by the psoriatic area and severity index (PASI). Five-millimeter-thick perforated biopsies of psoriatic plaque tissue (PS) and five control skin samples were obtained from surgically discarded samples from healthy people (NNs) (provided by the Shanghai Tenth People’s Hospital) [33]. DEGs and differentially expressed miRNAs (DEmiRs) in the psoriasis lesion group (PS) and the normal skin group (NN) were analyzed using DESeq2 [34]. The screening criteria were set as FDR <0.1 and |log2-fold change (FC)| >1. Gene set enrichment analysis (GSEA) of the DEGs was performed via the WEB-based Gene Set Analysis Toolkit (http://www.webgestalt.org/) [35]. A p-value < 0.05 was considered to indicate statistical significance in screening the enrichment results. Genes enriched in psoriasis-related pathways were selected as candidate target genes for DEmiRs.

Target gene prediction and miRNA-mRNA regulatory network construction

The DEmiR-associated targets of the dataset GSE142582 were predicted using TargetScan [36], miRTarBase [37], miRDB [38, 39], and RNAhybrid [40]. Only the targets included in all of these databases were selected for further analysis. The miRNA-target genes were subsequently overlapped with the DEGs, and the negative interaction pairs between the DEmiRs and DEGs were used to construct the miRNA–mRNA network via Cytoscape [41] software (version 3.9.1). RNAhybrid was further used to calculate the minimum free energies (MFEs) between miRNAs and target genes, whereas miRNA:mRNA hybridization was considered plausible when the MFE scores were in the relatively high-energy interval (−30 kcal/mol < MFE ≤ −14 kcal/mol) [42]. Hybridization was performed by using RNAhybrid in domain mode, and the short sequence was hybridized to the best fitting region of the long sequence. This tool is primarily meant to predict miRNA targets. The reporter assay validity for identifying microRNA-target interactions was performed with miRTarBase [43].

Ethical statement

The animal experiments in this study were approved by the Animal Protection and Research Ethics Committee of Jilin University, and all procedures were performed strictly in accordance with the guidelines for the Care and Use of Laboratory Animals. Specific pathogen-free BALB/c mice were purchased from Liaoning Changsheng Biotechnology Co., Ltd., (Benxi, Liaoning, China), raised at the Experimental Animal Center of Jilin University, and fed a standard maintenance diet. All animal operations were performed under anesthesia, and every effort was made to minimize pain. The human tissue used for cell isolation was provided by the First Affiliated Hospital of Jilin University (2020-708), and informed consent and confidentiality were obtained from the donors.

HaCaT cell culture

The human immortalized keratinocyte (HaCaT) cell line was purchased from American Type Culture Collection (ATCC-PCS-200-011). The cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM; Gibco, Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 10% (v/v) fetal bovine serum (FBS; Gibco) and 1% penicillin/streptomycin (Boster, Wuhan, China). The cells were cultivated in a humidified incubator at 37°C with 5% CO₂. The medium was changed every two days.

Isolation of human primary epidermal keratinocytes (HEKs)

Human skin tissue was obtained from the First Affiliated Hospital of Jilin University. After several washes with physiological saline containing antibiotics, the tissue was cut into small squares with a side length of 5 mm, soaked in 0.25% trypsin and incubated overnight at 4°C. After overnight incubation, the tissue fragments were transferred to phosphate-buffered saline (PBS) buffer supplemented with antibiotics, after which the epidermal layer was carefully removed using ophthalmic forceps. The peeled epidermal layer was cut into smaller pieces and centrifuged at 2000 rpm for 10 minutes. The supernatant was discarded after centrifugation, the epidermal fragments were resuspended in K-SFM (Cat. 17005042; Thermo Fisher Scientific) supplemented with 1% penicillin/streptomycin, and the mixture was placed in a 37°C incubator under 5% carbon dioxide.

Induction of cells with TNF-α

Human recombinant tumor necrosis factor was purchased from PeproTech, Inc., (Cat. 300-1A, Rocky Hill, NJ, USA). The powder was initially reconstituted in water to 1.0 mg/mL, and for extended storage, the TNF-α solution was further diluted to the proper concentration using PBS containing 5% BSA and stored at −80°C. When used to induce cells, the TNF-α solution was further diluted with cell culture medium to the corresponding concentration, after which the cells were cultured for a certain duration.

microRNA transfection

The microRNA and scramble-miRNA mimics were synthesized by GenScript (Nanjing, Jiangsu, China) and were dissolved in water to a concentration of 1 μg/mL for storage. The sequences are shown in Table 1. For starvation, the medium was changed to Opti-modified Eagle’s medium (Opti-MEM, Gibco, Thermo Fisher Scientific) 12 hours before transfection. Lipofectamine 3000 transfection reagent (Cat. L3000008, Invitrogen, Carlsbad, CA) was used for miRNA transfection, and the manufacturer’s protocol was followed. In brief, 5 μl of transfection reagent and 5 μg of miRNA were added to 125 μl of Opti-MEM, after which the diluted transfection reagent and miRNA mixture were mixed to a total volume of 250 μl. After resting for 15 min at room temperature, the mixtures were added to the cells in the six-well plate drop by drop. The same dose of transfection reagent without miRNAs was added to the control group. The basal medium was resuspended 8 h after transfection.

Cell proliferation and viability assay

CCK-8 (Cat. AR1199, Beyotime, Shanghai, China) reagent was used to monitor cell viability. Cells (5,000 cells/well) were seeded into 96-well plates. The CCK-8 reagent was diluted 1:10 with DMEM, 100 μl was added to each well, and then, every group of cells was treated and incubated for 2 hours at 37°C. Optical density (OD) values were monitored using a microplate reader at 450 nm (Infinite M200 Pro NanoQuant, Tecan, Switzerland) every 12 hours. Cell viability curves were plotted with the OD values as the ordinate and points in time as the abscissa.

Flow cytometric analysis

The cells were seeded into 6-well plates, cultured to 40–50% (5~6 × 10⁵) confluence, and then transfected with miRNAs for 72 h. The cells were harvested, washed with cold phosphate-buffered saline (PBS), and fixed in 70% ethanol at 4°C overnight. RNA was removed by using RNase A at 37°C for 30 min. Finally, the cells were stained with propidium iodide (PI; Cell Cycle and Apoptosis Analysis Kit, Cat. C1052; Beyotime, Shanghai, China) for 30 min at room temperature and analyzed on a FACSAria III flow cytometer (BD Biosciences). The cell cycle peak was generated using Modfit software (ver. 5.0).

5-Ethynyl-2′-deoxyuridine (EdU) staining

EdU staining was performed using a BeyoClick™ EdU Cell Proliferation Kit with Alexa Fluor 488 (Cat. C0071S; Beyotime, Shanghai, China) according to the manufacturer’s protocol. In brief, cells were plated in six-well plates after the corresponding treatments, incubated with 10 μM EdU for 2 hours at 37°C and then fixed with 4% paraformaldehyde for 15 min, permeabilized with 0.5% Triton X-100 for 20 min and stained with Click Additive Solution and Hoechst 33342 in the dark. Images were captured using an EVOS f1 fluorescence microscope (Thermo Fisher Scientific). The percentage of EdU-positive cells was analyzed using Image-Pro Plus 6.0 software (Media Cybernetics, Silver Spring, USA).

IMQ-induced psoriatic mouse model

For establishment of IMQ-induced psoriasis-like skin inflammation, 20 six-week-old BALB/c female mice were randomly divided into four groups (one Vaseline group and three IMQ groups) after a one-week stabilization period. On the day before induction, the dorsal skin of the mice was shaved and then treated with 5% imiquimod cream (62.5 mg) (Sichuan Mingxin Pharmaceutical Co., Ltd., Chengdu, Sichuan, China), and the same dose of Vaseline was applied to the Vaseline group once daily for 1 to 4 days. The day before IMQ plastering was defined as Day 0, and the psoriatic mouse model was generated via continuous smearing for 4 days. The erythema score, scaling score, and skin thickness severity index were calculated from 1–5, accurate to one decimal place, and the three parameters were summed to obtain a cumulative score [44].

Injection of miRNAs into a psoriatic mouse model

On the 4th day after the psoriasis model was established, the psoriatic mice were randomly divided into three groups: the IMQ group, the scr-miRNA (scrambled-microRNA) (GenScript, Nanjing, China) group, and the miR-26a group. The scr-miRNA group and the miR-26a (10 μg) group were subcutaneously injected with compounds from the Entranster™ In Vivo Kit (Cat. 18668-11-1; Engreen Biosystem Co., Ltd., Beijing, China) and scr-miRNA or miR-26a mimics, respectively. Moreover, the negative control group and IMQ group were injected with the abovementioned transfection reagent in combination with saline. Daily skin changes were recorded with photos by a Canon 700D Single lens Reflex camera.

Histological analysis

The skin and spleen tissues from each group of mice were collected after euthanasia, soaked in 4% paraformaldehyde at 4°C overnight, dehydrated in increasing concentrations of ethanol for various durations (70%—6 h, 80%—1 h, 96%—1 h, 100%—3 h), and clarified in xylene, followed by embedding in paraffin. Sections 5 μm in thickness were cut, and after dewaxing, H&E staining was performed using an H&E staining kit (Cat. AR1180, Boster, Wuhan, Hubei, China). The stained sections were imaged with an Olympus CX43 microscope, and the layer thicknesses were measured by the tool within the software.

Immunohistochemistry

Immunohistochemistry was performed by the standard procedure. In brief, the tissue slices were dewaxed in xylene for 10 minutes, after which the xylene was removed with PBS. The slices were placed in heated boiling EDTA antigen retrieval solution and boiled for 20 minutes. The slices were then cooled with cold water to room temperature, after which the antigen retrieval solution was removed with PBS. Normal goat serum solution (Cat. AR0009, Boster, Wuhan, Hubei, China) was added to the slices, which were incubated in a wet box for 30 minutes, after which the unbound serum protein was washed off with PBS. The primary antibody was diluted 100-fold and added to the slices. The samples were incubated in a wet box for 1–2 hours, after which the unbound primary antibody was washed off with PBS. The 100-fold diluted secondary antibody was added to the slices. The samples were incubated in a wet box for 1–2 hours. Finally, the unbound secondary antibody was washed off with PBS, the slices were dried, and the slides were observed under a microscope after they were dried. The antibodies used for the immunohistochemistry assays are shown in Table 2.

Total RNA and miRNA extraction and qRT-PCR

For qRT-PCR analyses, total RNA was extracted from cells (72 hours after miRNA transfection) and tissues with TRIzol reagent (DP419, Tiangen, Beijing, China), and complementary DNA (cDNA) was synthesized using a FastKing RT Kit (with gDNase) (KR116, Tiangen Biotech, China) according to the manufacturer’s recommendations. A miRcute miRNA extraction kit (DP501, Tiangen, Beijing, China) was used for miRNA extraction, and a miRcute Plus miRNA First-Strand cDNA kit (KR211, Tiangen, Beijing, China) was used for miRNA RT-PCR. Real-time qPCR assays were performed using SuperReal PreMix Plus (SYBR Green) (FP205, Tiangen, Biotech, China), a miRcute Plus miRNA qPCR kit (FP411, Tiangen, Beijing, China), and an IQ5 Multicolor Real-Time PCR detection system (Bio-Rad, Hercules, California, USA). The expression levels of the treated samples were normalized to the levels of the GAPDH and U6 genes, and the relative gene expression levels were calculated via the 2^−ΔΔCT formula [45].

Primer design and validation

The primer pairs used in the RT-qPCR assay were designed using AlleleID software version 7.0 (Premier Biosoft, http://www.premierbiosoft.com/). All the sequences were tested for their potential secondary structure and dimerization ability using the OligoAnalyzer 3.1 program (Integrated DNA Technologies, https://sg.idtdna.com/pages). The specificity of the primers was identified using the Primer-blast program (https://blast.ncbi.nlm.nih.gov/Blast.cgi). The validated sequences are shown in Tables 3, 4.

Protein isolation and western blotting

Total protein was extracted from tissues and cells using radioimmunoprecipitation assay lysis buffer (Cell lysis buffer for Western and IP; P0013; Beyotime, Shanghai, China) supplemented with phenylmethanesulfonyl fluoride (PMSF; ST506 Beyotime, Shanghai, China). The protein concentration was measured using an enhanced BCA protein assay kit (Cat. P0010, Beyotime, Shanghai, China). Total protein extracts were separated on 10% gels via SDS-PAGE and then transferred to 0.45 nm polyvinylidene fluoride membranes (Millipore, USA). The proteins were probed with specific antibodies after the blot was blocked with 5% nonfat milk (Cat. AR0104; Boster, Wuhan, China). The antibodies used are listed in Table 2.

Dual-luciferase reporter assay

The target sites of miR-26a-5p on CDC6 and the CCNE1 3′UTR were predicted by mirPath v.3 from DIANA TOOLS (https://dianalab.e-ce.uth.gr/html/mirpathv3/index.php?r=mirpath). Dual-luciferase reporter plasmids containing a 200 bp fragment containing the target site of miR-26a-5p were synthesized by Sangon Biotech (Shanghai, China). HEK293 and NIH3T3 cells were transfected with 5 μg of Psi-CHECK2-wt CDC6-3′UTR (or wtCCNE1-3′UTR), 10 μmol of miR-26a and Psi-CHECK2-wt CDC6-3′UTR (or wtCCNE1-3′UTR); miR-26a and Psi-CHECK2-mut CDC6-3′UTR (or wtCCNE1-3′UTR); or Scr-miR-26a and Psi-CHECK2-wt CDC6-3′UTR (or wtCCNE1-3′UTR). Forty-eight hours after transfection, a Dual-Lumi™ II Luciferase Assay Kit (Cat. No. RG089S; Beyotime, Shanghai, China) and a microplate reader (Infinite M200 Pro NanoQuant, Tecan, Switzerland) were used to determine the fluorescence intensities of firefly luciferase and Renilla luciferase, the specific values of which were used as the relative luciferase activity.

Statistical analysis

Statistical analyses were performed using GraphPad Prism for Windows software (GraphPad Software, version 8.0.1; San Diego, CA, USA; https://www.graphpad.com/). The values are expressed as the means ± SDs. Comparisons between two groups were performed using Student’s t test. ^*p < 0.05, ^**p < 0.01, ^***p < 0.001, and ^****p < 0.0001 are the significance thresholds, and ns means not significant.

Availability of data and material

All data in the study are available.

In silico identification of the miR-26a-5p/CCNE1/CDC6 axis in psoriasis

We performed a gene set enrichment analysis (GSEA) of 2985 DEGs (1392 upregulated and 1593 downregulated DEGs) obtained from the GSE142582 dataset with the KEGG signaling pathways and found that the IL-17 signaling pathway and cell cycle were associated with the development of psoriasis among the pathways significantly enriched with upregulated genes (Figure 1A). We next selected the top 10 downregulated DEmiRs (let-7a-5p, let-7b-5p, let-7c-5p, let-7d-5p, let-7g-3p, miR-423-5p, miR-30a-5p, miR-26a-5p, miR-22-3p and miR-195-5p) from the 232 DEmiRs (99 upregulated and 133 downregulated DEmiRs) obtained from the GSE142582 dataset to analyze the targeting relationship between the upregulated genes in the two pathways, resulting in the generation of a miRNA-mRNA network based on 24 targets involved in the cell cycle and only 8 targets belonging to the IL-17 signaling pathway, as shown in Figure 1B. miR-26a-5p was first chosen as a candidate therapeutic miRNA. To identify reliable targets, we performed a comprehensive analysis considering the expression correlation, minimum free energy (MFE), and upregulated fold change values of the 7 target genes (CCNE1, CCNE2, CDC6, CHEK1, MAD2L1, PTTG1 and TTK). We found that CDC6, CCNE1 and PTTG1 exhibited the strongest negative correlation with miR-26a-5p (r < −0.8). However, PTTG1 had the lowest MFE (−13.5) in humans harboring miR-26a-5p, which suggested that PTTG1 may not be a primary target of miR-26a-5p in humans (Figure 1C). We therefore transfected miR-26a-5p mimics into TNF-α-induced HaCaT cells and human epidermal keratinocytes (HEKs). qPCR revealed that PTTG1 was not downregulated as was CDC6 or CCNE1 by miR-26a-5p, causing more than 50% downregulation of both HaCaT and HEK cell lines, suggesting that CDC6 and CCNE1 are the most reliable effective targets of miR-26a-5p (Figure 1D).

Figure 1. In silico identification of the miR-26a-5p/CCNE1/CDC6 axis in psoriasis (A) GSEA bar chart showing the top 10 enhanced and top 10 inhibited KEGG pathways for DEGs; the color scale represents the FDR, and the length of the bar represents the normalized enrichment score (NES) (positively correlated, blue >0; negatively correlated, orange <0). (B) The interaction network of 10 downregulated miRNAs and target genes involved in the cell cycle and the IL17 pathway. Let-7a-5p, let-7b-5p, let-7c-5p, let-7d-5p, and let-7g-3p are uniformly represented by let-7 because they all share the same target genes. (C) Multifactor correlation estimation between miR-26a-5p and target genes. (D) Target gene expression in HaCaT and HEKs cells. Orange columns: cells induced with TNF-α without miRNA transfection; blue columns: cells induced with TNF-α and transfected with miR-26a-5p. The data are presented as the mean ± SD (n = 3). ^*p < 0.05; ^**p < 0.01; ^***p < 0.001; ^****p < 0.0001.

Verification of cross-species conservation of the miR-26a-5/CDC6/CCNE1 axis in humans and mice

To ensure consistency between our in vivo and in vitro experiments, we first performed multiple sequence alignment of the published miR-26 family sequences. The results showed that only miR-26 showed cross-species conservation, and all 29 miR-26 sequences were the same except for those of ssa (Salmo salar) and tch (Tupaia chinensis) (Figure 2A). Fortunately, the relationships of 7 targets (CCNE1, CCNE2, CDC6, CHEK1, MAD2L1, PTTG1 and TTK) of miR-26a-5p were confirmed in mice with target prediction analysis, and CDC6 was the gene that most highly bound to miR-26a-5p in both species (Figure 2B). To confirm our hypothesis, we performed dual luciferase analysis to verify that CDC6 and CCNE1 are direct targets of miR-26a-5p in both humans and mice. ELISAs also revealed that the fluorescence intensity of cell extracts from 3′UTR- and miR-26a-cotransfected cells was significantly lower than that of the cells from the mutant sequence-cotransfected and scr-miR-transfected groups by more than 30% (p < 0.001). These findings demonstrated that miR-26a-5p can disrupt the expression of CDC6 and CCNE1 in both human and mouse cells, confirming their direct interaction and regulatory relationship (Figure 2C, 2D).

Figure 2. Verification of cross-species conservation of the miR-26a-5/CDC6/CCNE1 axis. (A) Multiple sequence alignment of orthologous miRNAs in the miR-26 family. (B) Comparison of the minimum free energy of miR-26a-5p at the 3′UTR binding sites of target genes between humans (blue columns) and mice (orange columns). (C, D) Dual-luciferase reporter assay of target genes of miR-26a-5p and statistical analysis of the relative fluorescence intensity of each transfection group. The data are presented as the mean ± SD (n = 3). ^*p < 0.05. ^***p < 0.001.

In vitro validation of the role of miR-26a-5p in inhibiting keratinocyte proliferation

To investigate the inhibitory effect of miR-26a-5p on the proliferation of keratinocytes under pathological and normal physiological conditions, we selected HaCaT cells and HEKs as the target cells for miRNA transfection, and miR-26a-5p was measured in the transfected cells by qRT-PCR (Figure 3A). The viability assays (CCK-8) of both miR-26a-5p mimic-transfected HaCaT cells and HEKs showed significantly decreased proliferation compared to that of the negative control cells (Figure 3B). Furthermore, according to the EdU incorporation assays, the number of EdU-positive cells was significantly reduced following the above transfection of miR-26a-5p mimics in both cell types. Interestingly, we found that the number of HaCaT cells was significantly lower than that of HEKs after transfection with the miR-26a-5p mimics, indicating that miR-26a-5p seems to have a stronger "killing" effect on HaCaT cells (Figure 4). We further observed that miR-26a-5p induced cell cycle arrest at the G1/S phase in HaCaT cells and HEKs by a flow cytometry assay. As shown in Figure 3C, 3D, following treatment with miR-26a-5p, a greater percentage of cells remained in the G0/G1 phase of the cell cycle (HaCaT: 34% to 59%; HEKs: 69% to 83%), and this change was accompanied by a reduced percentage of cells in the S phase (HaCaT: 49% to 34%; HEKs: 23% to 11%). In addition, we found that miR-26a-5p significantly reduced the expression of CCNE1 and CDC6 at the mRNA level (Figure 3E). The western blot results further confirmed that miR-26a-5p, as predicted by the luciferase reporter gene assay, was able to interfere with CDC6 and CCNE1 expression in vitro (Figure 3F) Finally, to ensure the safety of miR-26a-5p for future clinical treatment, researchers must evaluate the effect of the dosage of miR-26a-5p on the proliferation of normal cells. Therefore, we performed a proliferation inhibition assay on uninduced HaCaT and HEK cells, and the CCK-8 results showed that 2, 5 and 8 μg of miR-26a-5p did not affect the viability of HaCaT or HEK cells that were not induced by TNF-a (Supplementary Figure 1).

Figure 3. In vitro validation of the ability of miR-26a-5p to inhibit keratinocyte proliferation (A) miR-26a-5p expression in transfected HaCaT cells and HEKs compared with that in untreated control cells. (B) Proliferative ability (growth curve) of cells in each group. Dark blue line: untreated control; red line: cells induced with TNF-α; gray line: cells induced with TNF-α and transfected with miR-26a-5p. (C, D) Proportional changes in the cell cycle according to flow cytometry assays between TNF-α-induced cells transfected with or without miR-26a-5p. (E, F) mRNA and protein expression of miR-26a targets in HaCaT cells and HEKs. Group 1: Nontreated cells; Group 2: Cells with transfection reagent added; Group 3: Cells transfected with miR-26a. The data are presented as the mean ± SD (n = 5). ^*p < 0.05; ^**p < 0.01; ^***p < 0.001; ^****p < 0.0001.

Figure 4. Intuitive observation of the cell proliferation rate of HaCaT keratinocytes and HEKs by EdU staining. After 48 hours, the cells were transfected with 50 nM miRNA mimics. The percentage of EdU incorporation was analyzed using Image-Pro Plus 6.0 software. Scale bars: 200 μm.

In vivo evaluation of the therapeutic effects of miR-26a-5p in a psoriasis mouse model

To further assess the efficacy of miR-26a-5p in the treatment of psoriasis in vivo, we topically applied imiquimod cream containing 5% IMQ to the shaved back skin of the mice for eight consecutive days to induce psoriasis. The mice in the Vaseline group and IMQ group were injected with saline (0.9%) once a day, while those in the miRNA treatment group were injected intradermally with miR-26a-5p and scr-miRNA from 4 to 8 days (Figure 5A). Compared with those in the IMQ-treated wild-type (WT) group and scr-miRNA-injected group, the erythema score, scaling score, and skin thickness severity index were markedly lower in the miR-26a-5p treatment group (Figure 5B, 5C). Then, we isolated the epidermis and dermis from miRNA-injected skin samples, and the qRT-qPCR results showed that the content of miR-26a-5p in the epidermis was much greater than that in the dermis (Figure 5D). Then, the dorsal skin samples of each group were stained with hematoxylin and eosin (H&E), and the thickness of the epidermis was measured. Quantitative analysis revealed that the thickness of the epidermis was greatest in the IMQ and SCR-miR groups and was significantly greater than that in the miR (miR-26a-5p) and Vaseline (control) groups. As observed above, the thickness of the epidermis of the mice treated with miR-26a-5p was very similar to that of the control group (Figure 6A). Moreover, spleen sections from the miR-26a-5p group showed no white pulp depletion (Figure 6B). The white pulp and red pulp boundaries were also clearer in the Vaseline group and the miR-26a-5p group than in the IMQ and scr-miR groups, indicating that miR-26a-5p not only inhibited the proliferation of epidermal keratinocytes in psoriasis-like skin but also reduced the inflammatory reaction in the spleen. Changes in the spleen circumference were also observed, suggesting that the spleen was enlarged due to the inflammatory reaction. The above results confirmed that the in vivo therapeutic effects of miR-26a-5p in the IMQ mouse model are consistent with those in patients with psoriasis. Furthermore, we found that normal mouse dorsal skin also exhibited tolerance to 10 μg of miR-26a-5p (Supplementary Figure 2). Finally, we confirmed that miR-26a-5p triggered cascade downregulation of biomarkers from the IL-23/IL-17A axis associated with psoriasis development while suppressing CDC6 and CCNE1 expression (Figure 6C, 6D). These findings suggested that miR-26a-5p-induced cell cycle arrest can in turn weaken the activity of driving factors such as IL-17A.

Figure 5. In vivo study design and pathological observation of mouse skin tissue. (A) Schematic diagram of the whole treatment process for the mice in each group. In brief, the hair on the backs of all the mice was shaved off on Day 0. From Day 1 to Day 8, IMQ was administered to the backs of the mice in Groups 2, 3 and 4. Moreover, the backs of the mice in Group 1 were daubed with Vaseline. From Day 4 to Day 8, the mice in Groups 1 and 2 were injected with saline once every day, and the mice in Groups 3 and 4 were injected with scr-miRNA and the miR-26a-5p mimic, respectively. Tissue samples were collected on Day 9. (B) Representative macroscopic views of the dorsal skin of mice in each group. (C) Statistical analysis of the skin erythema score, scale score, and skin thickness severity index of each group of mice. (D) miR-26a expression in the dermis and epidermis of injected mice and in the normal skin tissue of mice in the Vaseline group. The data are presented as the mean ± SD (n = 5). ^***p < 0.001.

Figure 6. Histology and immunohistology of skin tissue and spleen sections. (A) H&E staining of paraffin sections of the dorsal skin and comparison of epidermal layer thickness. Scale bar: 200 μm. (B) Spleen sections and circumferential comparisons of mice from each group. Scale bar: 2 mm. (C, D) CCNE1 and CDC6 antibody staining of skin tissue sections for immunohistology analysis. Scale bar: 100 μm. The data are presented as the mean ± SD (n = 5). ^*p < 0.05; ^**p < 0.01.

Based on these experimental results, 4 days after miR-26a-5p was injected subcutaneously into IMQ-induced mice, PTTG1 was not downregulated at the mRNA level among the 7 target genes (Figure 7A). CDC6 and CCNE1, which act as initial DNA replication switches, were downregulated at the protein level (Figure 7B). In addition, the core genes in the IL-23/IL-17A axis were downregulated by miR-26a-5p interference (Figure 7C).

Figure 7. Expression of proteins and genes related to psoriasis. (A) mRNA expression of CDC6, CCNE1, and representative genes of the IL-23/IL-17A/STAT3 axis was detected in the IMQ-induced mouse model using qRT-PCR. (B, C) The protein expression of CDC6, CCNE1, and representative genes of the IL-23/IL-17A/STAT3 axis was examined in the IMQ-induced mouse model using Western blotting. The data are presented as the mean ± SD (n = 5). ^*p < 0.05. ^**p < 0.01. ^***p < 0.001. ^****p < 0.0001.