Усовершенствованная последовательность транскриптомов и искусственный интеллект/машинное обучение (AI/ML) в имитируемом псориазом воспалении кератиноцитов человека.

1. Introduction

Psoriasis is a chronic autoimmune disease that is not contagious and often coexists with other medical conditions [ 1 ]. It is characterized by longlasting symptoms and a tendency to recur, leading to sleep disturbances and concentration difficulties [ 2 ]. The clinical manifestations of psoriasis can range from small patches to widespread involvement of the entire body, and can even trigger suicidal thoughts in individuals with mental health issues [ 3 ]. Psoriasis is also associated with an increased risk of developing other serious illnesses such as arthritis, metabolic syndrome, cardiovascular disease, crohn’s disease, and respiratory complications [ 4 – 6 ]. The etiology of psoriasis is believed to be multifactorial, involving genetic factors, skin injury, climate, immunodeficiency, certain infections, and the use of specific medications [ 7 – 9 ]. The development of psoriasis is a complex process, and evidence suggests that both innate and adaptive immune systems play a role in the disease [ 10 , 11 ]. Various genetic and environmental factors stimulate immune cells to release inflammatory cytokines, such as TNF-α, IFN-γ, IL-6, IL-12, and IL-23, which in turn cause abnormal growth and development of skin cells called keratinocytes [ 11 – 15 ]. Keratinocytes are activated by Th1 and Th17 cells, resulting in the production of antimicrobial peptides, inflammatory cytokines, and chemokines, which contribute to the characteristic inflammation observed in the skin of patients [ 16 – 18 ]. The involvement of Th17 and Tγδ cells that produce IL-17A is highlighted by variations in genes encoding the IL-23 receptor and IL-12, which have been shown to be crucial for psoriasis development. This disease is characterized by the activation of the IL-12/IL-23/IL-17 cytokine axis [ 11 , 17 , 19 – 21 ].

Imiquimod (IMQ; Aldara®) is an immune modulator that is effective in limiting viral infections and inhibiting tumor growth [ 22 – 25 ]. It exerts its primary mode of action through Toll-like receptors (TLR), specifically TLR-7 and TLR-8, which activate NF-κB [ 26 ]. NF-κB activation induces cell death in tumor cells and disrupts adenosine receptor signaling pathways [ 27 – 32 ]. Researchers have developed an IMQ-induced mouse model to enhance our understanding of psoriasis, which is widely used in basic research [ 33 – 37 ]. However, the relationship between IMQ-induced psoriasis in mice and human psoriasis in terms of genetics and clinical characteristics is not fully understood [ 38 ]. Although previous studies have used the HaCaT keratinocyte model to investigate the effects of antipsoriatic agents, there is still limited knowledge about the mechanism of action of IMQ-induced inflammation, proliferation, differentiation, and signal transduction in psoriasis-like keratinocytes [ 39 – 41 ].

In our previous research, we conducted a genetic analysis using genome-wide association studies (GWAS) to investigate the association between single nucleotide polymorphisms (SNPs), human leukocyte antigen (HLA) genotypes, and psoriasis. The results of our study showed that 153 genes significantly correlated with psoriasis phenotypes. In the Taiwanese population with psoriasis, we found a significant link between HLA-A 02:07 and HLA-C 06:02. The development and progression of psoriasis are influenced by pro-inflammatory cytokines such as IL-12, IL-23, and TNF [ 42 ]. To further investigate psoriasis-related mechanisms, in this study we established an in vitro model that mimics psoriasis-like proliferation and inflammation induced by IMQ. Subsequently, we performed a gene-level analysis and examined the signaling pathways associated with psoriasis.

2. Methods

2.1. Chemicals and reagents

The reagents used in this study were obtained from Thermo Fisher Scientific (Waltham, MA, USA): Dulbecco’s Modified Eagle medium (DMEM), fetal bovine serum (FBS), trypsin-EDTA, L-glutamine, penicillin G, and streptomycin. The compounds 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), dimethyl sulfoxide (DMSO), and imiquimod (IMQ) were acquired from Sigma–Aldrich (Merck KGaA, Darmstadt, Germany). Fetal bovine serum (FBS) was obtained from Invitrogen (Carlsbad, CA, USA).

2.2. Cell culture

The HaCaT cell line, derived from human normal keratinocytes, was obtained from Elabscience Biotechnology Co., Ltd (Elabscience®; Catalog No.: CL-0090). The cells were cultured in DMEM supplemented with 10% heat-inactivated FBS, penicillin (100 U/ml), and streptomycin (100 μg/mL) and maintained at 37 °C in a 5%CO 2 atmosphere [ 43 , 44 ].

2.3. Cell viability assay

HaCaT cells were cultivated in 96-well plates at a density of 1 × 10 4 cells/100 μL. They were then exposed to different concentrations of IMQ for 24 h. The concentrations used were 0, 25, 50, and 100 μM. After a 20-h incubation period, a solution of MTT (0.5 mg/mL) was added to each well and incubated at 37 °C for 4 h. The growth media was then extracted, and the formazan crystals were dissolved using 100 μL of DMSO. Absorbance was measured using the SpectraMax iD3 multimode microplate reader (Molecular Devices Ltd., San Jose, CA, USA) at a wavelength of 570/620 nm. Cell viability was calculated as a percentage relative to the control group [ 45 , 46 ].

2.4. Cell morphology

HaCaT cells were cultured in 24-well plates at a concentration of 2.5 × 10 5 cells/mL. They were then exposed to various concentrations (0, 25, 50, and 100 μM) of IMQ for 24 h. The cell images were analyzed using a phase-contrast microscope with a 200× magnification (Leica Microsystems GmbH, Wetzlar, Germany) [ 44 , 47 ].

2.5. Whole transcriptome sequencing

Total RNA was isolated using TRIzol ® Reagent (Invitrogen, USA) following the manufacturer’s instructions provided in the manual. RNA was measured using an ND-1000 spectrophotometer (Nanodrop Technology, USA) at a wavelength of 260 nm and assessed for quality using a Bioanalyzer 2100 (Agilent Technology, USA) with an RNA 6000 LabChip kit (Agilent Technology, USA). Library preparation and sequencing processes followed the standard protocol provided by Illumina (Illumina, USA). Library building was performed using the SureSelect XT HS2 mRNA Library Preparation kit from Agilent, USA, and library size selection was carried out using AMPure XP beads from Beckman Coulter, USA. Transcriptome sequencing was performed using the Illumina NovaSeq 6000 system, which utilizes the sequencing-by-synthesis (SBS) method. Welgene Biotech’s process, which utilizes Illumina’s basecalling application bcl2fastq (v2.20), was used to create sequencing data in the form of FASTQ reads. To assess the accuracy of the sequence base obtained fromthe sequencingplatforms,we used the Phred quality score (Q score). We employed the bcl2fastq conversion software v2.20, the official basecalling application developed by Illumina, to convert BCL files from all Illumina sequencing platforms into FASTQ reads. For adaptor cutting and sequence quality trimming, we utilized Trimmomatic (v0.36) with a sliding windowtechnique. Next,we aligned the next-generation sequencing reads to the genomes using HISAT2 software version 2.0.1. Ultimately, the pipeline is applied to the human reference genome (GRCh38.p14) obtained from the UCSC Genome Browser, which may be accessed at the following URL: http://asia.ensembl.org/Homo_sapiens/Info/Index?db=core . Differential expression was studied using StringTie (version 2.1.4) and DEseq (version 1.39.0) with genome bias detection and correction. Genes exhibiting a lowlevel of expression (less than0.3 TPMvalue) in either the treatment or control samples, or in both, were eliminated. Genes that exhibited a p-value of less than 0.05 and a fold change greater than 2.0, were deemed to be significantly differentially expressed. Each stage underwent rigorous quality control and meticulous monitoring [ 48 – 50 ].

2.6. Bio-informatics network analysis

The outcome of the volcano plotwas generatedusing the EnhancedVolcano package,which can be accessed at the following URL: https://bioconductor.org/packages/release/bioc/html/EnhancedVolcano.html . The gene set enrichment analysis (GSEA) was executed using the clusterProfiler package. The package can be found at https://www.bioconductor.org/packages/release/bioc/html/clusterProfiler.html . The Gene Set Enrichment Analysis (GSEA) analysis displayed p-values of categories that were less than 0.5. To investigate a molecular network, the Ingenuity Pathway Analysis (IPA) program (version: 107193442; Qiagen Sciences, Inc.) was utilized. The program received data collection consisting of gene symbols and Log2 ratios (IMQ/control). Each gene was associated with a relevant entity in the IPA Knowledge database. Networks and bio-informatics pathways were created using the IPA algorithm [ 42 , 51 ].

2.7. Quantitative real time-PCR (QPCR) analysis

HaCaT cells were cultivated in a 75-T flask at a density of 2 × 10 6 cells/mL. They were then exposed to 0 and 100 μM IMQ, commercially known as Aldara, for 18 h at 37 °C. The EZ2 RNA/miRNA Tissue/Cells Kit (QIAGEN) was used to extract the total RNA. The quantity of RNA samples was assessed using NanoDrop One, a device manufactured by Thermo Fisher Scientific. The RNA samples were reverse-transcribed for 120 min at 37 °C using the High Capacity cDNA Reverse Transcription Kit, following the standard protocol provided by the manufacturer (Applied Biosystems). Quantitative PCR was conducted under the following conditions: a 10-min incubation at 95 °C, followed by 40 cycles of a 15-s incubation at 95 °C and a 1-min incubation at 60 °C. The reaction mixture consisted of 2X Power SYBR Green PCR Master Mix (Applied Biosystems) and 200 nM forward and reverse primers. The assay was performed three times on an Applied Biosystems 7900 Real-Time PCR system, and the fold changes in expression were determined using the comparative CT method. Human glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as an endogenous control. The primers used for TNF-α, IL-6, and GAPDH are listed in Table 1 [ 52 ].

2.8. Artificial intelligence/machine learning (AI/ML) disease pathways prediction analysis

The QIAGEN artificial intelligence/machine learning (AI/ML) database of IPA software (version 107193442; Qiagen Sciences, Inc.) utilizes artificial intelligence/machine learning (AI/ML) disease pathway prediction analysis. It generates over 1500 diseases, phenotypes, and functional pathways. These pathways display primary genes or molecules that influence a specific disease and its related phenotypes. IPA software was used to forecast potential diseases and phenotypes based on genes acquired by RNA sequencing [ 53 , 54 ].

2.9. Statistical analysis

Data are presented as mean ± standard deviation. To analyze the variations between groups and conduct multiple comparisons, a one-way analysis of variance was performed, followed by Dunnett’s test. The study utilized SPSS software version 26.0, which was developed in Chicago, IL, USA. The statistical significance level was set at P < 0.001.

3. Results

3.1. Effect of IMQ on cell viability and morphology of HaCaT cells

The study design is illustrated in Fig. 1 . To determine the effect of IMQ on the survival of HaCaT keratinocytes, cell viability was evaluated using the MTT assay. The data in Fig. 2A shows that the application of IMQ at concentrations of 25, 50, and 100 μM resulted in cell viability rates of 101.78 ± 2.04%, 115.05 ± 6.60%, and 131.17 ± 3.78% respectively in HaCaT cells. As demonstrated in Fig. 2B , treatment with IMQ at concentrations of 25, 50, and 100 μM did not alter cell shape. Our finding indicated that IMQ increased cell viability in a concentration-dependent manner.

3.2. Whole transcriptome sequencing and bioinformatics network analysis

A comprehensive analysis was performed on the transcriptome sequencing profile to investigate the mechanism of action (MOA) and network of IMQ in HaCaT keratinocytes. The highest dose of IMQ (100 μM) was used for whole transcriptome sequencing. The rawdata fromwhole transcriptome sequencing of HaCaT cells treated with IMQ are provided in Supplementary Table S1 ( https://www.biomedicinej.com/cgi/editor.cgi?article=1468&window=additional_files&context=biomedicinej ).

Fig. 3A displays volcano plots showing 513 differentially expressed genes. Among these genes, 382 were up-regulated (indicated in red) and 131 were down-regulated (indicated in blue) in IMQ-treated HaCaT keratinocytes. Gene numbers were analyzed along with pathways ( Fig. 3B ), leading to the prioritization of processes, such as the immune system, cellular immune response, cellular stress and injury, disease-specific pathways, pathogen-influenced signaling, cytokine signaling, intracellular and second messenger signaling, humoral immune response, proliferation and development, cellular growth, ingenuity toxicity list pathways, cancer, and apoptosis.

Fig. 3C provides a graphical overview of the signaling pathways in IMQ-treated HaCaT cells compared with the control. The 25 most important signaling pathways included IFN-γ signaling, interferon alpha/beta signaling, macrophage classical activation signaling pathway, pathogen-induced cytokine storm signaling pathway, role of hypercytokinemia/hyperchemokinemia in the pathogenesis of influenza, IL-4 and IL-13 signaling, interferon signaling, protein ubiquitination pathway, antigen presentation pathway, and unfolded protein response. Other pathways included the CGAS-STING signaling pathway, glucocorticoid receptor signaling, Th1 pathway, hepatic fibrosis, glycosylation signaling pathway, coronavirus pathogenesis pathway, complement system and IL-10 signaling, tumor microenvironment pathway, and immunogenic cell death signaling pathway. The pathways involved also include the signaling of the acute phase response, function of PKR in the activation of interferon production, initiation of antiviral defense mechanisms, pyroptosis signaling pathway, macrophage alternative activation signaling pathway, cachexia signaling pathway, multiple sclerosis signaling pathway, and Th1 and Th2 activation pathways.

Gene Set Enrichment Analysis (GSEA) is a bioinformatics analytical method that determines if a pre-defined set of genes exhibits statistically significant and consistent differences between two phenotypes [ 55 ]. In this study, GSEA analysis was performed using Gene Ontology categories. The results showed significant enrichment in the response to interferon-gamma category (normalized enrichment score (NES) = 1.51298, P = 0.00142) ( Fig. 4A ) and the JAK-STAT signaling pathway category (NES = 0.07600, P = 0.31173) ( Fig. 4B ) in IMQ-treated samples compared to control samples. The genes controlled by the JAK-STAT signaling pathway were analyzed using the IPA database. Our results showed that the IL-6/JAK/STAT signaling pathway ( Fig. 5A ), and human leukocyte antigen (HLA) mediated T cell activation pathway ( Fig. 5B ) plays a role in promoting HaCaT keratinocyte proliferation and inflammation by IMQ. Additionally, an analysis was conducted to predict the target genes and their association with inflammation and/or pro-inflammatory responses in HaCaT cells treated with IMQ and psoriasis. The findings are shown in Fig. 6 .

3.3. Effect of IMQ on IL-6 and TNF-α mRNA levels on HaCaT keratinocytes

We used quantitative real-time PCR (QPCR) analysis to study how IMQ affects the expression of IL-6 and TNF-α cytokines in HaCaT cells. Fig. 7 shows that when HaCaT cells were treated with 100 μM IMQ, there was an increase in the mRNA expression of IL-6 and TNF-α.

3.4. Predict possible diseases and phenotype by artificial intelligence/machine learning (AI/ML)

The results presented in Fig. 8 , Table 2 , and Supplementary Table S2 ( https://www.biomedicinej.com/cgi/editor.cgi?article=1468&window=additional_files&context=biomedicinej ) provides a comprehensive understanding of the genetic connections between the important genes in IMQ-treated HaCaT cells and various illnesses by artificial intelligence/machine learning (AI/ML) method. These diseases include multiple sclerosis, mixed hematopoietic and lymphoid cancer, lympho-proliferative malignancy, polycythemia vera, glomerulonephritis, alopecia areata, tularemia, thymus gland tumor, chronic myelo-proliferative neoplasm, chronic malignant hematological neoplasm, lymphoma, relapsing-remitting multiple sclerosis, neuromuscular disease, anthrax, septic shock, ulcer, bacterial peritonitis, macrophage activation syndrome or hemophagocytic lymphohistiocytosis, hemophagocytic lymphohistiocytosis, and respiratory distress syndrome.

4. Discussion

Psoriasis is a common skin disease [ 5 ] characterized by persistent chronic inflammation and immune-mediated inflammation, which can affect more than just the skin [ 1 , 5 ]. In Taiwan, the occurrence rate of psoriasis is 0.24% [ 42 ]. Psoriasis is characterized by the presence of erythematous plaques on silvery scales. These plaques can cause discomfort, irritation, and skin damage [ 56 ]. The etiology of psoriasis includes genetic factors [ 42 ], environmental factors [ 57 , 58 ], innate and adaptive immune responses [ 59 – 61 ], lifestyle choices [ 6 , 62 , 63 ], and therapeutic drugs [ 6 , 62 ]. The pathogenesis of psoriasis involves several key processes: (1) excessive proliferation of keratinocytes, fibroblasts, and endothelial cells, (2) activation of the immune system, (3) initiation of an inflammatory response, and (4) stimulation of new blood vessel development (neovascularization or angiogenesis) [ 5 , 42 , 64 ]. To investigate the causes of psoriasis and identify new therapies, it is important to develop suitable in vitro and in vivo models of psoriasis [ 42 ]. This study aimed to create an in vitro model of psoriasis-like inflammation in keratinocytes by using imiquimod.

In this study, the application of IMQ greatly boosted the viability of HaCaT keratinocytes, as demonstrated in Fig. 2A . Fig. 2B shows that IMQ treatment did not result in significant cell death; instead, an increase in the number of cells was observed. This in vitro model offers a potential explanation for the atypical growth and inflammation of keratinocytes observed in psoriasis. Our findings are consistent with those of previous research in which HaCaT cells were treated with IMQ (100 μM) to induce inflammation and proliferation resembling psoriasis [ 40 ]. IMQhas recently shown potential for the treatment of various dermatological diseases [ 65 , 66 ]. Aldara™, a topical creamcontaining 5% IMQ, has been approved by the FDA as an immune response agent in clinical settings. IMQ is indicated for the treatment of external anogenital warts and basal cell carcinomas [ 67 ]. The Mechanism of Action (MOA) of IMQ involves its binding to Toll-like receptors, specifically TLR7/TLR8, in humans. This binding interferes with adenosine receptor (AR) signaling and activation of nuclear factor-kappa B (NF-kB) [ 68 , 69 ]. Activation of protein activity triggers the generation of proinflammatory cytokines and chemokines, leading to an active immune response through the interaction of antigen-presenting cells (APC) with MHC molecules [ 68 , 70 , 71 ]. Our findings also indicate that certain genes in the NF-kB signaling pathway (CASP1, CD74, EDNRA, GPR89A/GPR89B, HMOX1, IFIT5, IL1A, IRF1, RIPK2, STAT1, TICAM2, TRIM22, TRIM5, UACA, UBD, and ZC3HAV1) as well as MHC molecules (HLA-DMB, HLA-DRA, and HLA-DPA1) are involved in regulating the immune response in HaCaT cells treated with IMQ. The results are summarized in Supplementary Table S1 ( https://www.biomedicinej.com/cgi/editor.cgi?article=1468&window=additional_files&context=biomedicinej ).

A previous study developed an in vitro model of psoriasis using differentiated HaCaT cells generated using IMQ. This model was used to quickly test potential drugs for psoriasis treatment [ 40 ]. The cause of psoriasis-like inflammation induced by IMQ was comparable to that observed in a mouse model of psoriasis [ 30 ]. To assess the potential signaling pathways of IMQ in HaCaT cells, a comprehensive analysis of the entire transcriptome was conducted in both the IMQ-exposed and control groups. The data presented in Fig. 3 demonstrates a significant correlation between IMQ-induced changes in gene expression and cellular proliferation, anti-apoptotic processes, and inflammatory cytokine production. In our bio-informatics network analysis, we categorized these functions according to distinct gene expression patterns, including: (1) immune responses (immune system, cellular immune response, humoral immune response, and cytokine signaling); (2) inflammatory signaling (signaling influenced by pathogens and cytokine signaling); (3) Cellular proliferation and cellular death (proliferation and expansion of cells, growth and programmed cell death.); and (4) Diseases or other conditions (such as disease-specific pathways, cancer, cellular stress and damage, secondary messenger signaling, and the ingenuity toxicity list pathway). GSEA considers variations in the number of gene sets and correlations between gene sets and the expression dataset. This technique uses a combination of statistical approaches to aggregate data from individual genes into a gene set. This allows the identification of circumstances in which all genes in a pre-determined set exhibit minor but coordinated change [ 55 ]. Fig. 4 demonstrates that GSEA analysis revealed a notable increase in enrichment in the response to IFN-γ and JAK-STAT signaling pathways in the IMQ-treatment group compared to the control group. To the best of our knowledge, our study is the first to document that IMQ causes psoriasis-like inflammation in keratinocytes. The results of the associated genes involved in inflammatory reactions, IMQ treatment, and psoriasis are shown in Fig. 6 . This study presents evidence that confirms the importance of the IL-6/JAK/STAT signaling pathway and human leukocyte antigen (HLA) mediated T cell activation pathway in various physiological processes and its involvement in psoriatic-like cell formation ( Fig. 5 ).

The etiology of psoriasis involves up-regulation of several pro-inflammatory cytokines, including IL-6, IL-8, IFN-γ, and TNF-α. The correlation between IL-6 and TNF-α, and the development of psoriasis has been widely recognized [ 72 , 73 ]. Figs. 5 and 6 presents the findings that provide evidence for a model explaining keratinocyte inflammation in psoriasis. This model suggests that a particular group of inflammatory products are produced at elevated levels. Moreover, certain inflammatory cytokines, such as IL-6 and TNF-α, were significantly increased owing to mRNA up-regulation. However, our results did not demonstrate an increase in IFN-γ levels, as observed in both whole transcriptome sequencing and QPCR (data not shown). Consequently, we inferred that the increase in IL-6 and TNF-α level was a result of JAK-STAT activation ( Fig. 7 ). This study differs from previous research [ 40 ].

Constructing models using artificial intelligence/machine learning (AI/ML) and making predictions about potential diseases and phenotypes is a significant approach in the fields of bio-informatics and artificial intelligence (AI) [ 74 – 76 ]. Polygenic risk scores (PRS) are frequently employed in clinical settings to evaluate an individual’s genetic susceptibility to a specific disease or trait relative to similar groups [ 77 ]. In this study, we utilized artificial intelligence/machine learning (AI/ML) to predict potential illnesses and phenotypes in a model of IMQ-induced psoriasis-like inflammation in keratinocytes. The findings presented in Fig. 8 , Table 2 and Supplementary Table S2 ( https://www.biomedicinej.com/cgi/editor.cgi?article=1468&window=additional_files&context=biomedicinej ) indicate that autoimmune diseases such as multiple sclerosis, alopecia areata, and relapsing-remitting multiple sclerosis, as well as cancer types such as mixed hematopoietic and lymphoid malignancy, are significant diseases associated with IMQ-induced psoriasis-like inflammation in keratinocytes. This finding further validates the notion that genes associated with psoriasis may play a role in the development of other autoimmune or cancer-related conditions.

5. Conclusion

The current study aimed to provide a more comprehensive understanding of the characterization and mechanism of action (MOA) of keratinocytes treated with IMQ using an in vitro model. Additionally, recently discovered genes and their associated network pathways can be used to assess the effectiveness of potential drugs for psoriasis and other skin conditions. This study offers valuable insights into the potential implications of improving dermatological treatment.

Supplementary Information

Acknowledgement

The authors would like to thank the Office of Research and Development, China Medical University (Taichung, Taiwan) for providing Medical Research Core Facilities to perform the experiments and data analysis. We also thank Mr. Chih-Ping Sung (Welgene Biotech Co., Ltd, Taipei, Taiwan), Mr. Chang-Wei Li (AllBio Science, Inc., Taichung, Taiwan) and Mr. Kuan-Wen Chen and Mr. Yao-Wei Jheng (GGA Corporation, Molecular Science and Digital Innovation Center, Taiwan) for their assistance and equipment for their assistance.

List of Abbreviations

Artificial intelligence

Dulbecco’s Modified Eagle medium

Dimethyl sulfoxide

Fetal bovine serum

Gene Set Enrichment Analysis

Genome-wide association studies

Human leukocyte antigen

Interferon-gamma

Interleukin

Imiquimod

Ingenuity Pathway Analysis

Machine Learning

Mechanism of action

3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide

Normalized enrichment score

Nuclear factor-kappa B

Polygenic risk scores

Quantitative real-time PCR

Single nucleotide polymorphisms

Tumor Necrosis Factor-α

Toll-like receptors

Funding Statement

This work was supported in part of the project (DMR-113-109) from China Medical University Hospital, Taiwan

Footnotes

Author contributions: LTW, JSY, SCT and FJT were responsible for the overall conception and design. TJH, HPC, YRP, TYL and YNJ contributed to data analysis and interpretation. SCT, YRP, and YJC performed the analysis of the bio-informatics network and interpreted the data. JSY, TYL and FJT performed the interpretation of Artificial intelligence/machine learning (AI/ML) disease pathways prediction analysis results. All authors have read and approved the final version of the manuscript.

Conflicts of interest: The authors declare that they have no competing interests.

Funding: This work was supported in part of the project (DMR-113-109) from China Medical University Hospital, Taiwan.

References

1. Medovic MV, Jakovljevic VL, Zivkovic VI, Jeremic NS, Jeremic JN, Bolevich SB, et al. Psoriasis between autoimmunity and oxidative stress: changes induced by different therapeutic approaches. Oxid Med Cell Longev. 2022;2022:2249834. doi: 10.1155/2022/2249834.

2. Jin S, Wang Y, Qie C, Yang L, Wu Y, Zhang T, et al. Single-cell RNA sequencing reveals the immune cell profiling in IMQ induced psoriasis-like model. J Inflamm Res. 2022;15:5999–6012. doi: 10.2147/JIR.S379349.

3. Koo J, Marangell LB, Nakamura M, Armstrong A, Jeon C, Bhutani T, et al. Depression and suicidality in psoriasis: review of the literature including the cytokine theory of depression. J Eur Acad Dermatol Venereol. 2017;31:1999–2009. doi: 10.1111/jdv.14460.

4. Zhao J, Li T, Wang J. Association between psoriasis and dementia: a systematic review. Neurologia (Engl Ed) 2024;39:55–62. doi: 10.1016/j.nrleng.2023.12.005.

5. Wride AM, Chen GF, Spaulding SL, Tkachenko E, Cohen JM. Biologics for psoriasis. Dermatol Clin. 2024;42:339–55. doi: 10.1016/j.det.2024.02.001.

6. Potestio L, Lauletta G, Tommasino N, Portarapillo A, Salsano A, Battista T, et al. Risk factors for psoriasis flares: a narrative review. Psoriasis (Auckl) 2024;14:39–50. doi: 10.2147/PTT.S323281.

7. Teshima R, Saito-Sasaki N, Sawada Y. Generalized pustular psoriasis and systemic organ dysfunctions. Int J Mol Sci. 2024:25. doi: 10.3390/ijms25116270.

8. Kim H, Choi MR, Jeon SH, Jang Y, Yang YD. Pathophysiological roles of ion channels in epidermal cells, immune cells, and sensory neurons in psoriasis. Int J Mol Sci. 2024:25. doi: 10.3390/ijms25052756.

9. Juma SN, Liao J, Huang Y, Vlashi R, Wang Q, Wu B, et al. Osteoarthritis versus psoriasis arthritis: physiopathology, cellular signaling, and therapeutic strategies. Genes Dis. 2024;11:100986. doi: 10.1016/j.gendis.2023.04.021.

10. Kim TG, Kim DS, Kim HP, Lee MG. The pathophysiological role of dendritic cell subsets in psoriasis. BMB Rep. 2014;47:60–8. doi: 10.5483/BMBRep.2014.47.2.014.

11. Grine L, Dejager L, Libert C, Vandenbroucke RE. An inflammatory triangle in psoriasis: TNF, type I IFNs and IL-17. Cytokine Growth Factor Rev. 2015;26:25–33. doi: 10.1016/j.cytogfr.2014.10.009.

12. Albanesi C, Madonna S, Gisondi P, Girolomoni G. The interplay between keratinocytes and immune cells in the pathogenesis of psoriasis. Front Immunol. 2018;9:1549. doi: 10.3389/fimmu.2018.01549.

13. Andersen CSB, Kvist-Hansen A, Siewertsen M, Enevold C, Hansen PR, Kaur-Knudsen D, et al. Blood cell biomarkers of inflammation and cytokine levels as predictors of response to biologics in patients with psoriasis. Int J Mol Sci. 2023:24. doi: 10.3390/ijms24076111.

14. Bruno M, Davidson L, Koenen H, van den Reek J, van Cranenbroek B, de Jong E, et al. Immunological effects of anti–IL-17/12/23 therapy in patients with psoriasis complicated by Candida infections. J Invest Dermatol. 2022;142:2929–29239e8. doi: 10.1016/j.jid.2022.05.1083.

15. Yang BY, Cheng YG, Liu Y, Liu Y, Tan JY, Guan W, et al. Ameliorates imiquimod-induced psoriasis-like dermatitis and inhibits inflammatory cytokines production through TLR7/8-MyD88-NF-kappaB-NLRP3 inflammasome pathway. Molecules. 2019;24 doi: 10.3390/molecules24112157.

16. Ahmad F, Alam MA, Ansari AW, Jochebeth A, Leo R, Al-Abdulla MN, et al. Emerging role of the IL-36/IL-36R Axis in multiple inflammatory skin diseases. J Invest Dermatol. 2024;144:206–24. doi: 10.1016/j.jid.2023.11.004.

17. Tsiogka A, Gregoriou S, Stratigos A, Soulaidopoulos S, Rompoti N, Panagakis P, et al. The impact of treatment with IL-17/IL-23 inhibitors on subclinical atherosclerosis in patients with plaque psoriasis and/or psoriatic arthritis: a systematic review. Biomedicines. 2023:11. doi: 10.3390/biomedicines11020318.

18. Zhang J, Zhao S, Xing X, Shang L, Cao J, He Y. Effects of neuropeptides on dendritic cells in the pathogenesis of psoriasis. J Inflamm Res. 2023;16:35–43. doi: 10.2147/JIR.S397079.

19. Mesas-Fernandez A, Bodner E, Hilke FJ, Meier K, Ghoreschi K, Solimani F. Interleukin-21 in autoimmune and inflammatory skin diseases. Eur J Immunol. 2023;53:e2250075. doi: 10.1002/eji.202250075.

20. Ceribelli A, Motta F, Vecellio M, Isailovic N, Ciccia F, Selmi C. Clinical trials supporting the role of the IL-17/IL-23 Axis in axial spondyloarthritis. Front Immunol. 2021;12:622770. doi: 10.3389/fimmu.2021.622770.

21. Bunte K, Beikler T. Th17 cells and the IL-23/IL-17 Axis in the pathogenesis of periodontitis and immune-mediated inflammatory diseases. Int J Mol Sci. 2019;20 doi: 10.3390/ijms20143394.

22. Syed TA. A review of the applications of imiquimod: a novel immune response modifier. Expet Opin Pharmacother. 2001;2:877–82. doi: 10.1517/14656566.2.5.877.

23. Alsabbagh MM. Imiquimod-induced psoriasis model: induction protocols, model characterization and factors adversely affecting the model. Acta Biochim Pol. 2023;70:729–33. doi: 10.18388/abp.2020_6426.

24. Patel U, Mark NM, Machler BC, Levine VJ. Imiquimod 5% cream induced psoriasis: a case report, summary of the literature and mechanism. Br J Dermatol. 2011;164:670–2. doi: 10.1111/j.1365-2133.2010.10124.x.

25. Balak DM, Hajdarbegovic E. Drug-induced psoriasis: clinical perspectives. Psoriasis (Auckl) 2017;7:87–94. doi: 10.2147/PTT.S126727.

26. Li R, Wang J, Wang X, Zhou J, Wang M, Ma H, et al. Increased betaTrCP are associated with imiquimod-induced psoriasis-like skin inflammation in mice via NF-kappaB signaling pathway. Gene. 2016;592:164–71. doi: 10.1016/j.gene.2016.07.066.

27. Luckner B, Essfeld F, Ayobahan SU, Richling E, Eilebrecht E, Eilebrecht S. Transcriptomic profiling of TLR-7-mediated immune-challenge in zebrafish embryos in the presence and absence of glucocorticoid-induced immunosuppression. Ecotoxicol Environ Saf. 2023;266:115570. doi: 10.1016/j.ecoenv.2023.115570.

28. Flutter B, Nestle FO. TLRs to cytokines: mechanistic insights from the imiquimod mouse model of psoriasis. Eur J Immunol. 2013;43:3138–46. doi: 10.1002/eji.201343801.

29. Bozrova SV, Levitskii VA, Nedospasov SA, Drutskaia MS. [Imiquimod: the biochemical mechanisms of immunomodulatory and anti-inflammatory activity]. Biomed Khim. 2013;59:249–66. doi: 10.18097/pbmc20135903249.

30. Schon MP, Schon M. TLR7 and TLR8 as targets in cancer therapy. Oncogene. 2008;27:190–9. doi: 10.1038/sj.onc.1210913.

31. Oldfield V, Keating GM, Perry CM. Imiquimod: in superficial basal cell carcinoma. Am J Clin Dermatol. 2005;6:195–200. doi: 10.2165/00128071-200506030-00006. discussion 01-2.

32. Meyer T, Nindl I, Schmook T, Ulrich C, Sterry W, Stockfleth E. Induction of apoptosis by Toll-like receptor-7 agonist in tissue cultures. Br J Dermatol. 2003;149(Suppl 66):9–14. doi: 10.1046/j.0366-077x.2003.05632.x.

33. Zhang S, Li HJ, Yang CM, Liu L, Sun XY, Wang J, et al. Inflammatory and immunomodulatory effects of tripterygium wilfordii multiglycoside in mouse models of psoriasis keratinocytes. Chin J Integr Med. 2024;30:222–9. doi: 10.1007/s11655-023-3599-y.

34. Guo Y, Mao W, Bai N, Jin L, Tang S, Lin X, et al. Integrated network pharmacological analysis revealed that Smilax glabra Roxb. alleviates IMQ-induced psoriatic skin inflammation through regulating T cell immune response. J Ethnopharmacol. 2024;325:117836. doi: 10.1016/j.jep.2024.117836.

35. Chen X, Fu C, Zheng Y, Li X, Liao Y, Zheng Y, et al. Intermittent fasting alleviates IMQ-induced psoriasis-like dermatitis via reduced gammadeltaT17 and monocytes in mice. Arch Dermatol Res. 2024;316:176. doi: 10.1007/s00403-024-02886-5.

36. Matsuno A, Sumida H, Nakanishi H, Ikeyama Y, Ishii T, Omori I, et al. Keratinocyte proline-rich protein modulates immune and epidermal response in imiquimod-induced psoriatic skin inflammation. Exp Dermatol. 2023;32:2121–30. doi: 10.1111/exd.14960.

37. Shobeiri SS, Dashti M, Pordel S, Rezaee M, Haghnavaz N, Moghadam M, et al. Topical anti-TNF-a ssDNA aptamer decreased the imiquimod induced psoriatic inflammation in BALB/c mice. Cytokine. 2023;172:156406. doi: 10.1016/j.cyto.2023.156406.

38. Swindell WR, Michaels KA, Sutter AJ, Diaconu D, Fritz Y, Xing X, et al. Imiquimod has strain-dependent effects in mice and does not uniquely model human psoriasis. Genome Med. 2017;9:24. doi: 10.1186/s13073-017-0415-3.

39. Zhang WJ, Song ZB, Bao YL, Li WL, Yang XG, Wang Q, et al. Periplogenin induces necroptotic cell death through oxidative stress in HaCaT cells and ameliorates skin lesions in the TPA- and IMQ-induced psoriasis-like mouse models. Biochem Pharmacol. 2016;105:66–79. doi: 10.1016/j.bcp.2016.02.001.

40. Varma SR, Sivaprakasam TO, Mishra A, Prabhu S, MR, PR Imiquimod-induced psoriasis-like inflammation in differentiated Human keratinocytes: its evaluation using curcumin. Eur J Pharmacol. 2017;813:33–41. doi: 10.1016/j.ejphar.2017.07.040.

41. Yuan X, Li N, Zhang M, Lu C, Du Z, Zhu W, et al. Taxifolin attenuates IMQ-induced murine psoriasis-like dermatitis by regulating T helper cell responses via Notch1 and JAK2/STAT3 signal pathways. Biomed Pharmacother. 2020;123:109747. doi: 10.1016/j.biopha.2019.109747.

42. Yang JS, Liu TY, Lu HF, Tsai SC, Liao WL, Chiu YJ, et al. Genome-wide association study and polygenic risk scores predict psoriasis and its shared phenotypes in Taiwan. Mol Med Rep. 2024:30. doi: 10.3892/mmr.2024.13239.

43. Yang YH, Chen YS, Chuang WT, Yang JS. Bifurcated polymorphic transition and thermochromic fluorescence of a molecular crystal involving three-dimensional supramolecular gear rotation. J Am Chem Soc. 2024;146:8131–41. doi: 10.1021/jacs.3c12454.

44. Chiang JH, Tsai FJ, Lin TH, Yang JS, Chiu YJ. Tremella fuciformis inhibits melanogenesis in B16F10 cells and promotes migration of human fibroblasts and keratinocytes. In Vivo. 2022;36:713–22. doi: 10.21873/invivo.12757.

45. Chen LJ, Yu TC, Huang BH, Tso KC, Song YF, Yin GC, et al. Synthesis of novel chitosan/sodium hyaluronate/iridium hydrogel nanocomposite for wound healing application. Int J Biol Macromol. 2024;270:132351. doi: 10.1016/j.ijbiomac.2024.132351.

46. Chiu YJ, Tsai FJ, Bau DT, Chang LC, Hsieh MT, Lu CC, et al. [Corrigendum] Next-generation sequencing analysis reveals that MTH-3, a novel curcuminoid derivative, suppresses the invasion of MDA-MB-231 triple-negative breast adenocarcinoma cells. Oncol Rep. 2024:52. doi: 10.3892/or.2024.8750.

47. Tsai YF, Yang JS, Chiu YJ, Tsai CW, Bau DT, Chang WS. Gadodiamide induced autophagy and apoptosis in human keratinocytes. In Vivo. 2022;36:603–9. doi: 10.21873/invivo.12743.

48. Wang CH, Yang JS, Chen CJ, Su SH, Yu HY, Juan YN, et al. Protective effects of Jing-Si-herbal-tea in inflammatory cytokines-induced cell injury on normal human lung fibroblast via multiomic platform analysis. Tzu Chi Med J. 2024;36:152–65. doi: 10.4103/tcmj.tcmj_267_23.

49. Cheng YD, Lu CC, Hsu YM, Tsai FJ, Bau DT, Tsai SC, et al. In Silico and in Vitro studies of taiwan chingguan yihau (NRICM101) on TNF-alpha/IL-1 beta-induced human lung cells. Biomedicine (Taipei) 2022;12:56–71. doi: 10.37796/2211-8039.1378.

50. Chiu YJ, Tsai FJ, Bau DT, Chang LC, Hsieh MT, Lu CC, et al. Next-generation sequencing analysis reveals that MTH-3, a novel curcuminoid derivative, suppresses the invasion of MDA-MB-231 triple-negative breast adenocarcinoma cells. Oncol Rep. 2021:46. doi: 10.3892/or.2021.8084.

51. Yang JS, Liu TY, Chen YC, Tsai SC, Chiu YJ, Liao CC, et al. Genome-wide association study of alopecia areata in taiwan: the conflict between individuals and hair follicles. Clin Cosmet Invest Dermatol. 2023;16:2597–612. doi: 10.2147/CCID.S428788.

52. Chiu YJ, Li CY, Wang TH, Ma H, Chou TY. Comparative transcriptomic analysis reveals differences in gene expression and regulatory pathways between nonacral and acral melanoma in Asian individuals. J Dermatol. 2024;51:659–70. doi: 10.1111/1346-8138.17187.

53. Mahawan T, Luckett T, Mielgo Iza A, Pornputtapong N, Caamano Gutierrez E. Robust and consistent biomarker candidates identification by a machine learning approach applied to pancreatic ductal adenocarcinoma metastasis. BMC Med Inf Decis Making. 2024;24:175. doi: 10.1186/s12911-024-02578-0.

54. Liu S, Wu J, Yang D, Xu J, Shi H, Xue B, et al. Big data analytics for MerTK genomics reveals its double-edged sword functions in human diseases. Redox Biol. 2024;70:103061. doi: 10.1016/j.redox.2024.103061.

55. Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005;102:15545–50. doi: 10.1073/pnas.0506580102.

56. Chi CC, Wu YW, Chao TH, Chen CC, Chen YJ, Cheng HM, et al. 2022 Taiwanese Dermatological Association (TDA), Taiwanese Association for Psoriasis and Skin Immunology (TAPSI), and Taiwan Society of cardiology (TSOC) joint consensus recommendations for the management of psoriatic disease with attention to cardiovascular comorbidities. J Formos Med Assoc. 2023;122:442–57. doi: 10.1016/j.jfma.2022.10.010.

57. Bhagwat AP, Madke B. The current advancement in psoriasis. Cureus. 2023;15:e47006. doi: 10.7759/cureus.47006.

58. Mateu-Arrom L, Puig L. Genes (Basel) 2023. Genetic and epigenetic mechanisms of psoriasis; p. 14.

59. Zalesak M, Danisovic L, Harsanyi S. Psoriasis and psoriatic arthritis-associated genes, cytokines, and human leukocyte antigens. Medicina (Kaunas) 2024:60. doi: 10.3390/medicina60050815.

60. Cianciulli A, Calvello R, Porro C, Lofrumento DD, Panaro MA. Inflammatory skin diseases: focus on the role of suppressors of cytokine signaling (SOCS) proteins. Cells. 2024:13. doi: 10.3390/cells13060505.

61. Syed NH, Mussa A, Elmi AH, Jamal Al-Khreisat M, Ahmad Mohd, Zain MR, et al. Role of MicroRNAs in inflammatory joint diseases: a review. Immunol Invest. 2024;53:185–209. doi: 10.1080/08820139.2023.2293095.

62. Liu S, He M, Jiang J, Duan X, Chai B, Zhang J, et al. Triggers for the onset and recurrence of psoriasis: a review and update. Cell Commun Signal. 2024;22:108. doi: 10.1186/s12964-023-01381-0.

63. Diaz MJ, Tran JT, Montanez-Wiscovich M. From the Cochrane Library: lifestyle changes for treating psoriasis. J Am Acad Dermatol. 2024;90:e35–7. doi: 10.1016/j.jaad.2023.09.023.

64. Gupta R, Debbaneh MG, Liao W. Genetic epidemiology of psoriasis. Curr Dermatol Rep. 2014;3:61–78. doi: 10.1007/s13671-013-0066-6.

65. Micali G, Lacarrubba F, Nasca MR, Schwartz RA. Topical pharmacotherapy for skin cancer: part I. Pharmacology. J Am Acad Dermatol. 2014;70:965e1–e12. doi: 10.1016/j.jaad.2013.12.045. quiz 77-8.

66. Vidal D. Topical imiquimod: mechanism of action and clinical applications. Mini Rev Med Chem. 2006;6:499–503. doi: 10.2174/138955706776876131.

67. Wagstaff AJ, Perry CM. Topical imiquimod: a review of its use in the management of anogenital warts, actinic keratoses, basal cell carcinoma and other skin lesions. Drugs. 2007;67:2187–210. doi: 10.2165/00003495-200767150-00006.

68. Schon MP, Schon M. Imiquimod: mode of action. Br J Dermatol. 2007;157(Suppl 2):8–13. doi: 10.1111/j.1365-2133.2007.08265.x.

69. Baulon E, Vautravers A, Rodriguez B, Nisand I, Baldauf JJ. [Imiquimod and immune response modifiers in gynaecology]. Gynecol Obstet Fertil. 2007;35:149–57. doi: 10.1016/j.gyobfe.2006.12.016.

70. Hanna E, Abadi R, Abbas O. Imiquimod in dermatology: an overview. Int J Dermatol. 2016;55:831–44. doi: 10.1111/ijd.13235.

71. Garland SM. Imiquimod. Curr Opin Infect Dis. 2003;16:85–9. doi: 10.1097/00001432-200304000-00004.

72. Alajroush WA, Alrshid AI, Alajlan AH, Alsalamah YB, Alhumaidan MI, Alhoumedan AI, et al. Psoriasis and metabolic disorders: a comprehensive meta-analysis of million adults worldwide. Cureus. 2024;16:e52099. doi: 10.7759/cureus.52099.

73. Sengupta R, Roy M, Dey NS, Kaye PM, Chatterjee M. Immune dysregulation and inflammation causing hypopigmentation in post kala-azar dermal leishmaniasis: partners in crime? Trends Parasitol. 2023;39:822–36. doi: 10.1016/j.pt.2023.07.005.

74. Nunez R, Dona I, Cornejo-Garcia JA. Predictive models and applicability of artificial intelligence-based approaches in drug allergy. Curr Opin Allergy Clin Immunol. 2024 Aug 1;24:189–94. doi: 10.1097/ACI.0000000000001002.

75. Alabi RO, Elmusrati M, Leivo I, Almangush A, Makitie AA. Artificial intelligence-driven radiomics in head and neck cancer: current status and future prospects. Int J Med Inf. 2024;188:105464. doi: 10.1016/j.ijmedinf.2024.105464.

76. Molfino NA, Turcatel G, Riskin D. Machine learning approaches to predict asthma exacerbations: a narrative review. Adv Ther. 2024;41:534–52. doi: 10.1007/s12325-023-02743-3.

77. Brown MA. Polygenic risk scores. Semin Arthritis Rheum. 2024;64S:152330. doi: 10.1016/j.semarthrit.2023.152330.

Associated Data

Supplementary Materials