Integrative bioinformatics and experimental validation of hub genetic markers in acne vulgaris: Toward personalized diagnostic and therapeutic strategies

Acne vulgaris is a widespread chronic inflammatory dermatological condition. The precise molecular and genetic mechanisms of its pathogenesis remain incompletely understood. This research synthesizes existing databases, targeting a comprehensive exploration of core genetic markers.

[20] While high-throughput gene expression studies provide invaluable insights into the condition, a comprehensive cross-study synthesis of these genes is needed to reveal consistent genetic signatures and their functional implications, potentially offering novel therapeutic avenues.
Acne's genetic underpinnings, despite its prevalence, remain inadequately explored in terms of shared genetic markers across diverse datasets. 7,9,21This study harnesses publicly available high-throughput gene expression datasets, especially GSE6475, GSE108110, and GSE53795, aiming to discern core hub genes pivotal to acne lesion pathogenesis and potential universal genetic markers.By integrating these datasets and employing robust bioinformatics tools like weighted gene co-expression network analysis (WGCNA) and gene set variation analysis (GSVA), the research illuminates shared differentially expressed genes (DEGs) and potential diagnostic markers.These efforts illuminate the broader genetic landscape and associated pathways in acne.Hub genes play a central role within co-expression networks, influencing multiple pathways and offering diagnostic and therapeutic potential.Using a combination of protein-protein interaction (PPI) networks and functional pathway analyses, we pinpointed these genes' interactions and functional implications.Furthermore, exploring the relationship between identified hub genes and immune cell infiltration could shed light on the inflammatory intricacies of acne and inform targeted therapeutic strategies.We further assessed their diagnostic potential using receiver operating characteristic (ROC) curve analysis.Experimental validations were performed on HaCaT cells post P. acnes stimulation, particularly emphasizing the functional significance of the CXCL8 gene was in cell proliferation, migration, and inflammation.Additionally, potential therapeutic agents targeting these hub genes were identified, holding promise for innovative acne treatments, thereby bridging the genetic underpinnings with clinical applications.
Acne vulgaris is a widespread dermatological condition with profound genetic underpinnings.This research presents an integrated approach, merging computational analyses with experimental validation, to decode the genetic complexities of acne.Our study identified 104 genes, with 97 upregulated and 7 downregulated, highlighting 4 hub genes that have considerable diagnostic potential for acne.These genes showcase consistency across datasets and are closely associated with immune cell populations.By delineating these hub genes and their functional implications, the research emphasizes their potential as both diagnostic biomarkers and therapeutic targets, offering a refined molecular perspective of acne.These findings not only offer a deeper understanding of acne's genetic architecture but also pave the way for innovative therapeutic interventions, emphasizing the transformative power of genomics in dermatological research.
following Propionibacterium acnes exposure, while CXCL8 knockdown reduced proinflammatory cytokines, cell proliferation, and migration.DrugBank insights led to the exploration of retinoic acid and methotrexate, both of which mitigated gene expression upsurge and inflammatory mediator secretion.

| Microarray data acquisition and processing
Expression profiling data derived from the Affymetrix microarrays (designated GSE6475, GSE108110, and GSE53795) [22][23][24] were sourced from the Gene Expression Omnibus (GEO) public database (http:// www.ncbi.nlm.nih.gov/ geo/ ).The GEO database serves as a pivotal repository for high-throughput gene expression data and provides accessibility to myriad experimental results, thereby promoting collaborative efforts and transparency in genomics research.Its comprehensive structure allows researchers to retrieve curated datasets for meta-analysis.The distinction between the GPL571 and GPL570 platforms primarily lies in their probe design and the number of probe sets they contain, making it paramount to consider their specifications in subsequent analyses.Specifically, the GSE6475 study utilized the GPL571 platform, while both the GSE108110 and GSE53795 studies were conducted on the GPL570 platform.The distinction in platforms can influence the coverage and sensitivity of the microarray data.
Therefore, cross-platform analyses were considered with caution to account for any platform-specific bias.Within GSE6475, six paired biopsies from acne patients, encompassing both acne lesion and adjacent unaffected skin, were analyzed.GSE108110 comprised 18 such paired samples, and GSE53795 included 12 pairs.Such paired biopsy approach ensures that variations specific to individual patients can be factored out, offering a more precise comparison between affected and unaffected tissues.The selection of these specific datasets was made after a preliminary review of available data to ensure relevance to our study objectives.Furthermore, the sample size in each dataset was deemed adequate for reliable statistical analysis.It is vital to note that the platforms utilized can impact the raw data, necessitating the meticulous calibration and normalization that follow.
To ensure the rigor of the acquired data, quality control was executed using the affy package 25 in R (version 3.6.3).Quality control (QC) steps are paramount in minimizing technical biases and inconsistencies that could arise from array processing.The affy package is specifically tailored for Affymetrix array data and facilitates diverse quality assessment metrics like intensity distributions, RNA degradation plots, and others, thus ensuring the robustness of the acquired data.Subsequent data normalization employed the Robust Multiplechip Average (RMA) 26 and quantile normalization 27

| Identification of differentially expressed genes (DEGs)
In this context, differential gene analysis pertains to the isolation of genes distinctly associated with acne, based on stringent statistical thresholds.It is imperative to set strict criteria to minimize false discoveries, especially given the high dimensionality of microarray datasets.
The rationale behind such a stringent selection is to mitigate falsepositives and ensure that only genes with substantial and consistent expression differences across samples are targeted for further analysis.
Genes thus identified were earmarked for ensuing biological scrutiny and validation.This step is vital to distill the vast list of genes into a subset with potential functional implications in acne pathogenesis.
Understanding the underlying genetic causes of acne can pave the way for targeted therapeutic strategies and diagnostic markers.The limma package, 28 in conjunction with R, facilitated the discernment of DEGs between acne lesions and their matching healthy skin tissues from acne-afflicted individuals.Limma, which stands for "Linear Models for Microarray Data," employs empirical Bayes methods to stabilize variance estimates across genes, thus enhancing reliability and statistical power.This method shrinks the estimated sample variances toward a pooled estimate, resulting in more stable inference when the number of arrays is small.Genes that exhibited an absolute log 2 -fold change (|log 2 FC|) of ≥1.0 and held a p-value of <0.05 were classified as significantly differentially expressed.These criteria strike a balance between ensuring sufficient change in gene expression to warrant biological significance and minimizing the risk of identifying genes whose expression differences might be due to random variation.
The intersecting DEGs across the three GSE datasets were retained for subsequent analytical phases to provide a consolidated list of genes consistently implicated across studies.This intersection ensures consistency and reproducibility in the identified gene patterns.Visualization of these DEGs was achieved through the construction of volcano plots and heatmaps, utilizing R. Volcano plots juxtapose the significance of each gene's differential expression (p-value) against the magnitude of its change (fold change), while heatmaps provide an intuitive graphical representation of gene expression patterns across multiple samples.These graphical representations enable intuitive comprehension of the data's distribution and differential gene expression patterns.

| Gene set variation analysis (GSVA)
Gene set variation analysis (GSVA) is an unsupervised, nonparametric method designed to appraise the variability in gene set enrichment. 29,30supervised methods, like GSVA, do not rely on predefined classes or groups.Instead, they discern patterns directly from the data.GSVA shifts the focus from individual genes to pathways, offering insights into how groups of genes collectively contribute to the observed phenotypic differences.This method probed differential signaling pathway enrichment across genes in the datasets GSE6475, GSE108110, and GSE53795.GSVA distinguishes itself from other enrichment methods by assessing the differential activity of curated gene sets across samples in an expression dataset.This obviates the need for gene ranking and instead focuses on the collective behavior of genes within predefined sets, thus offering enhanced biological insight. 31llmark gene sets, encompassing 50 gene sets denoted as Gene sets H, were procured via the msigdbr package. 32These hallmark gene sets are representative collections curated from various sources to encapsulate well-defined biological states or processes, providing a comprehensive and robust repertoire for GSVA.The analytical procedure was facilitated by both the limma and GSVA packages capitalizing on their synergistic capacities to refine the analysis.

| GO and KEGG pathway enrichment analyses
Gene ontology (GO) analysis 33 serves as an analytical tool for categorizing biological processes (BP), molecular functions (MF), and cellular components (CC) pertaining to the target genes thereby enabling researchers to contextualize the broader biological implications of the identified DEGs.This ontology covers three domains: BP which describes biological goals accomplished by one or more ordered assemblies of molecular functions; MF which describes activities that can be carried out by individual gene products; and CC which describes where gene products are active.GO terms provide a consistent and controlled vocabulary for annotating genes, enabling a systematic understanding of the cellular roles and functions of individual genes in a larger biological context.
Conversely, the Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis 34 elucidates the biological metabolic pathways in which these genes participate thereby affording insights into potential therapeutic targets and intervention points.KEGG serves as a knowledge base for systematic analysis of gene functions in terms of networks of genes and molecules.It connects genomic information with higher order functional information by integrating large-scale molecular datasets into a comprehensive computational model.To present the genomic analysis results of DEGs, we employed the clus-terProfiler package 35 in R, renowned for its versatility in integrating results from both GO and KEGG, facilitating a comprehensive gene enrichment analysis.For both analyses, a threshold p-value of <0.05 was deemed indicative of statistical significance ensuring that the resultant pathways and processes are not products of mere chance, but are genuinely associated with acne pathology.

| Construction of the protein-protein interaction (PPI) network and hub genes identification
We utilized the STRING database (Version 11.5) 36 (available at: https:// www.strin g-db.org/ ) to elucidate interactions among proteins, both known and predicted.STRING, a comprehensive, curated protein-protein interaction database, contains information from experimental datasets, computational prediction methods, and public text collections, thus making it a reliable source for PPI investigation.This enabled the construction of a protein-protein interaction (PPI) network for selected genes.Given that interactions with a score below 0.400 might lead to false-positive results, differential expression genes (DEGs) were inputted into STRING, and a medium confidence interaction score (≥0.400) was designated as the threshold.The choice of a medium confidence score ensured a balanced approach between sensitivity and specificity, allowing us to capture potentially relevant interactions while filtering out less reliable ones.
Subsequently, the PPI network was orchestrated using Cytoscape 3.9.1 37 with the assistance of the CytoNCA plugin. 38Cytoscape provides a robust platform for visual representation and analysis of molecular interactions networks.By integrating the CytoNCA plugin, the network centrality measures are effectively analyzed to highlight pivotal nodes or proteins.
The molecular complex detection (MCODE) plugin 39 was employed to cluster closely related genes, thus forming modules, utilizing default parameter settings to maintain consistency and reproducibility of results.MCODE is a graph-theoretic clustering algorithm designed to detect densely connected regions in large protein interaction networks which might correspond to protein complexes.
The CytoHubba plugin 40 was then used to pinpoint the top 10 node genes across 9 algorithms (MCC, MNC, Degree, EPC, BottleNeck, Closeness, Radiality, Betweeness, and Stress).The diverse set of algorithms ensures a comprehensive evaluation of the genes' significance in the network from multiple angles, increasing the reliability of hub gene selection.The intersections of these identified hub genes were discerned with the aid of the Upset package which offers a superior alternative to traditional Venn and Euler diagrams for the visualization of intersecting sets.This step helps in consolidating the hub genes identified across different algorithms, providing a set of genes that have unanimous support.

| Assessment of immune cell infiltration in acne lesion samples
Understanding the immune landscape of skin tissue, particularly in pathological conditions such as acne, is crucial for comprehensive disease characterization and therapeutic targeting.For each skin tissue sample (both lesioned and normal), we applied single-sample gene set enrichment analysis (ssGSEA) via the GSVA package in R to compute enrichment scores. 41The ssGSEA method was chosen for its capability to determine the relative enrichment of predefined gene sets within single samples, making it apt for evaluating immune cell compositions in heterogeneous tissue samples.This involved leveraging the gene expression values from each sample and the metagenes corresponding to 24 distinct immune cell types. 42tagenes are representative gene sets for specific cell types and can provide insight into the abundance and activity of these cells in a given sample.
The normalized enrichment score (NES) obtained from ssGSEA provided a quantitative assessment of immune cell infiltration levels for each sample.NES represents a standardized score that factors in the size of the gene set, making it ideal for comparing results across different samples. 43A positive NES indicates activation, while a negative NES suggests repression.These findings were illustrated as heatmaps using the pheatmap package in the R platform to offer a concise and visually intuitive representation of immune cell prevalence in each sample. 44

| Correlation analysis between biomarkers and immune cell infiltration levels
Determining the interplay between hub gene biomarkers and immune cell infiltration can offer profound insights into the pathogenesis and potential therapeutic intervention points in acne.The relationships between our discerned hub gene biomarkers and immune cell infiltration levels were probed via Spearman's rank correlation analysis executed in R. 45 Spearman's correlation was selected over Pearson's due to its nonparametric nature, making it suitable for datasets that might not assume a linear relationship or Gaussian distribution. 46e derived correlations were visually represented in heatmaps, facilitated by the ggplot2 package.This ensures a colorful, comprehensive, and easy-to-understand representation of correlation matrices.Detailed associations between specific hub genes and immune cells were graphically presented as scatter plots utilizing the ggscatterstats package.Scatter plots provide an intuitive depiction of the relationship between two variables, allowing for visual inspection of trends and outliers.Further, correlations between diagnostic genes and immune cells were depicted using lollipop charts.Lollipop charts offer a visually engaging alternative to bar charts, clearly highlighting individual data points.

| Evaluation of the predictive value of biomarkers using ROC curve analysis
To validate the predictive accuracy of each identified hub gene across three datasets, we conducted a receiver operating characteristic (ROC) curve analysis. 47ROC analysis evaluates the diagnostic ability of a binary classifier, capturing the trade-off between sensitivity and specificity for each potential threshold.It plots the true positive rate against the false positive rate at various threshold settings.ROC curve analysis stands as a gold standard method in assessing the diagnostic accuracy of biomarkers.This was performed using the pROC package in R. The area under the curve (AUC) value is an aggregated measure of a test's performance across all possible threshold values.Hub genes exhibiting an AUC value greater than 0.7 were deemed potential diagnostic markers, offering a differentiation between acne and control tissues.

| Real-time quantitative PCR (RT-qPCR)
Total RNA was isolated utilizing the TRIzol method (T9108, TaKaRa) and then transcribed into cDNA using the reverse transcription kit (RR037A, TaKaRa) following the manufacturer's protocol.

| Enzyme-linked immunosorbent assay (ELISA)
Each well of the ELISA plate was loaded with 100 μL of sample, and the process was replicated thrice for every sample.Incubation occurred at 37°C for 2 h.After the initial incubation, the liquid was discarded from each well, followed by the addition of 100 μL of primary antibody, and another hour of incubation at 37°C.Subsequent to this, the liquid was removed from each well, which was then washed three times using wash buffer.100 μL of secondary antibody was added to each well and incubated at 37°C for 1 h.After this phase, the wells were emptied and washed three times.Each well was then loaded with 100 μL of substrate TMB and incubated at room temperature in the dark for 20 min to induce the colorimetric reaction.Finally, a stop solution (50 μL) was added to cease the reaction.

| Lentiviral transduction for gene silencing
Lentiviral-mediated gene knockdown was utilized to establish cell lines with diminished expression of the CXCL8 gene (siCXCL8-1 and siCXCL8-2), alongside corresponding control cell lines (siNC).

| Western blot (WB)
Cells were lysed using RIPA buffer containing 1% phosphatase and protease inhibitors to obtain total protein.The protein concentration was determined using a BCA Protein Assay Kit.Proteins were then separated using SDS-PAGE with varying concentrations and subsequently transferred onto PVDF membranes.Once the membranes were blocked, they were incubated overnight at 4°C with primary antibodies against CXCL8/IL-8 (ab289967, Abcam) and β-actin (ab8227, Abcam).This was followed by incubation with an HRPconjugated secondary antibody (ab205718, Abcam).After washing with TBST, protein signals were detected using ECL chemiluminescence substrate and visualized on an imaging system.All reagents related to the western blot protocol were obtained from Suzhou NCM Biotech Co., Ltd.
Cells, either untreated or subjected to various treatments/transfections, were seeded equally in 96-well plates.After 24 and 48 h of incubation, the medium was replaced with fresh culture medium containing 10% CCK-8 and incubation continued for an additional 2 h.Optical density (OD) values were measured at 450 nm using a microplate reader.

| Wound-healing assay
To ensure consistency and reproducibility, horizontal lines were marked at 0.5 cm intervals at the base of a 6-well plate, with at least four lines per well.Cells were seeded at a density of 10 6 cells/well with the goal of achieving monolayer confluence by the following day.Using the tip of a 200 μL pipette, a uniform straight scratch was gently made across the cell layer.Cells were then washed with PBS and incubated with serum-free culture medium.Photographs were taken at 0 h (beginning of the experiment) and 24 h using an inverted microscope.Horizontal lines served as references to ensure consistent measurements of wound widths at identical locations over time.

| Transwell cell migration assay
A culture medium containing 20% serum was added to the lower chamber of the Transwell apparatus (BD Corning).Cells from each group were resuspended in serum-free medium and seeded into the upper chamber.After a 24-h incubation, cells remaining in the upper chamber were gently removed with a wet cotton swab.Migrated and adhered cells in the lower chamber were fixed with 4% paraformaldehyde for 1 h, stained with crystal violet for 15 min, and subsequently imaged under a microscope in random fields of view.1F).The different log 2 FC thresholds between the datasets suggest that distinct degrees of gene expression changes were expected or observed across studies, reinforcing the need for tailored approaches in analyzing differential gene expression.These specified criteria ensured the selection of genes with high expression values and robust fold change, subsequently resulting in minimized false-positives.This adjusted p-value (after multiple testing correction) assures a rigorous standard in identifying truly significant genes.Notably, 104 genes overlapped across these datasets, which comprised 97 upregulated and 7 downregulated genes (Figure 2; Table 1).This overlapping signifies the reproducibility and robustness of the identified DEGs in different datasets, reinforcing their potential significance in the pathogenesis of acne.

| Gene set variation analysis (GSVA) across datasets
For datasets GSE6475, GSE108110, and GSE53795, GSVA was performed to elucidate the potential roles of genes associated with acne lesions, utilizing hallmark gene sets (Figure 3).GSVA, a nonparametric unsupervised method, allows for the enrichment analysis of predefined gene sets in individual samples, providing insights into pathway activities that may be related to a particular disease condition.By leveraging hallmark gene sets, which are well-defined biological states or processes, the analysis aimed to decipher the predominant cellular processes implicated in acne pathogenesis.
Notably, these genes exhibited significant enrichment in immune-

| Protein-protein interaction (PPI) network and hub modules
We uploaded the 104 DEGs to the STRING database and utilized Cytoscape for further insights.The STRING database is a comprehensive resource that offers information about known and predicted protein-protein interactions, allowing researchers to envision the functional connectivity of a particular set of proteins.
A PPI network was constructed with 92 nodes and 1148 edges, based on confidence scores ≥0.400 using the CytoNCA plugin's Betweenness method (Figure 5A).The PPI network offers a visualization of the complex interaction landscape of the identified The high interaction scores of these modules imply that they represent closely knit functional groups that might act in synergy in the pathogenesis of acne.Employing the CytoHubba plugin, top hub genes were ranked using 9 different algorithms (Figure 6A).
Different algorithms offer varied perspectives on node importance, ensuring a comprehensive understanding of network centrality.A consensus of four key hub genes-PTPRC, CXCL8, ITGB2, and MMP9-was derived using the R "UpSet" package, as delineated in Figure 6B.Identifying hub genes is essential, as these are genes central to the network and might play critical roles in the disease process, making them potential therapeutic targets or biomarkers.

| Analysis of immune cell infiltration
Given the association of GSVA results, as determined from the genes across the three datasets, and the predominance of GO terms and KEGG pathways enriched by the DEGs with inflammation and infection processes (Figure 4), we hypothesized that immune cell infiltration could be pivotal in acne pathogenesis.This hypothesis was grounded in the understanding that the immune response, often characterized by the presence and behavior of immune cells, plays a crucial role in mediating inflammatory processes, including those seen in acne. 58,59It is noteworthy to emphasize that immune cell infiltration serves as an indication of the body's defense mechanism against perceived threats. 60To discern the correlation between acne lesion and normal skin samples with respect to the 24 infiltrating immune cells, F I G U R E 2 Identification of 104 intersecting genes from DEGs across datasets GSE6475, GSE108110, and GSE53795, comprising 97 upregulated and 7 downregulated genes.This visualization aids in understanding the common genetic markers shared across all datasets, potentially pointing to the universal genetic factors influencing acne.
we employed ssGSEA.The ssGSEA method, a variant of the gene set enrichment analysis, offers a nuanced perspective on gene expression data, providing a metric that describes the degree of enrichment of predefined gene sets in each sample.Evidently, from Figure S1, there was a pronounced disparity in immune cell infiltration levels between acne lesions and normal skin across the GSE6475, GSE108110, and GSE53795 datasets.This disparity highlights a potentially crucial role of the immune response in the exacerbation of acne-related inflammation.While a general overrepresentation of immune cells in acne lesions is noted, the specific interplay between immune cell types and acne severity remains to be deciphered.In a majority of immune cell types, the acne lesions exhibited a markedly elevated level of immune cell infiltration in comparison to normal skin (Figure S1), consistent with the enrichment results depicted in Figure 4.Such upregulation in immune cell infiltration resonates with the longstanding assertion that acne is fundamentally an inflammatory condition.This upregulated infiltration emphasizes the active immune response in acne lesions, potentially driven by the presence of certain pathogenic bacteria, tissue damage, or inflammatory cytokines. 55,61 elucidate the implications of these central genes in immune cell infiltration, we examined correlations between the four hub genes and the 24 immune cells (Figure 7).Notably, in the GSE6475 dataset, aDC, macrophages, neutrophils, and Th1 cells demonstrated significant associations with the four hub genes, namely PTPRC, CXCL8, ITGB2, and MMP9 (Figure 7A).These findings suggest that there is a direct or indirect regulatory relationship between the highlighted hub genes and the specific immune cell infiltration in the acne microenvironment.Whether these genes promote or are a result of increased immune infiltration could be a topic of subsequent investigations.In the GSE108110 and GSE53795 datasets, analogous hub gene-immune cell trends were observed, with most immune cells displaying positive associations with the hub genes.However, NK cells exhibited TA B L E 1 Differential expression analysis of genes from acne lesions.A compilation of 104 differentially expressed genes (DEGs) sourced from three microarray datasets: GSE6475, GSE108110, and GSE53795.Of these, 97 genes exhibited upregulation, while 7 showed downregulation when juxtaposed against the normal skin tissue from the same acne patients.F I G U R E 3 GSVA analysis employing hallmark gene sets for all sample genes associated with acne lesions in datasets: (A) GSE6475, (B) GSE108110, and (C) GSE53795.The GSVA assessment as depicted here, offers insights into the variance in gene set enrichment across the datasets, providing a holistic view of the genetic landscape associated with acne lesions.

DEGs
an inverse correlation (Figure 7B,C).This inversion in the NK cell correlation could suggest a regulatory role, where these cells might be acting to counterbalance or suppress exacerbated immune responses in acne lesions.Scatter plots further depicted specific correlations between hub genes and neutrophils and between hub genes and NK cells.All four hub genes correlated positively with neutrophil infiltration levels but exhibited a negative relationship with NK cells across the three datasets (Figure 8).Neutrophils are frontline defenders against bacterial infections, which can be a component of acne lesions, explaining the observed positive correlation. 62,63The negative association with NK cells, which are primarily known for their antiviral and antitumor actions, warrants further exploration to understand their specific role, if any, in acne pathology. 64,65

| Receiver operating characteristic (ROC) curve analysis
To assess the diagnostic potential of the four hub genes for acne, ROC curves were plotted across the three datasets (Figure 9).The ROC curve, a graphical representation of the diagnostic ability of a binary classifier, is a reliable tool to gauge the accuracy and effectiveness of specific genes as potential diagnostic markers.The area under the curve (AUC) provides a quantitative measure of this effectiveness.A gene exhibiting an AUC value exceeding 0.7 was considered a potential diagnostic marker.Within the GSE6475 dataset, the AUC values for PTPRC, CXCL8, ITGB2, and MMP9 were 97.2%, 100%, 100%, and 100%, respectively (Figure 9A).

| Predictive analysis of key miRNAs, TFs, and drug-gene networks
Employing online databases, we aimed to predict potential miRNAs targeting our identified diagnostic genes.miRNAs are pivotal regulators of gene expression, and their interactions with the hub genes could offer insights into the intricate post-transcriptional regulatory mechanisms associated with acne. 66The miRNA-diagnostic gene regulatory network is presented in Figure 10.Notably, hsamir-124-3p was identified in both the miRNet and NetworkAnalyst-TarBase v8.0 databases, interacting with PTPRC, MMP9, and CXCL8.The recurrent identification of hsa-mir-124-3p from different databases underscores its potential regulatory significance in acne pathogenesis.
The ChEA3 database forecasted upstream TFs.As outlined in Table 2, the top 10 TFs with the optimal mean rank for the four diagnostic genes (PTPRC, CXCL8, ITGB2, and MMP9) are enumerated.Transcription factors are proteins that control the rate of transcription and can activate or suppress target genes. 67TFEC, with the most favorable mean rank, appeared to regulate all four hub genes concurrently.This underscores the potentially central role of TFEC in the regulation of these genes.Also, ZNF831, MXD1, and FOXP3 emerged as upstream TFs targeting all four diagnostic genes.Other TFs like MTF1, SP110, TBX21, NFATC2, LEF1, and CEBPB were found to target specific combinations of these genes (Table 2).The identification of upstream TFs offers invaluable information on the regulatory pathways that might be at play in acne development.
Lastly, the DrugBank database was utilized to identify potential pharmacological interventions targeting these genes.Understanding drug-gene interactions can pave the way for personalized medicine, where tailored therapeutic strategies can be devised based on an individual's genetic makeup.A plethora of drugs was identified, targeting PTPRC (9 drugs), CXCL8 (116 drugs), ITGB2 (20 drugs), and MMP9 (51 drugs) (Table S1).Intersection analysis pinpointed four drugs-arsenic trioxide, methotrexate, silicon dioxide, and tretinointhat target all four hub genes.This overlap in drug targeting suggests a potential synergistic or multi-targeted therapeutic approach to address the multifaceted nature of acne.Detailed drug-gene interactions are elaborated in Table S1, with intersecting drugs highlighted F I G U R E 9 ROC curve evaluations for the four eminent hub genes across: (A) GSE6475, (B) GSE108110, and (C) GSE53795.ROC, curves gauge the diagnostic prowess of these genes, quantifying their specificity and sensitivity in discriminating diseased from non-diseased states.
F I G U R E 1 0 MicroRNA (miRNA) regulatory networks targeting the focal genes (PTPRC, CXCL8, ITGB2, and MMP9) as deduced from two databases: miRNet and NetworkAnalyst.Understanding miRNA-gene interactions can offer insights into potential therapeutic targets and the multilayered regulation of acne pathogenesis.TA B L E 2 Predicted transcription factors associated with the four prognostic hub genes.This table collates potential regulators of the hub genes, offering insights into the upstream genetic control mechanisms that might influence the behavior of these critical genes in the context of acne.
in bold.It is worth noting that while these drugs may target the hub genes, their safety, efficacy, and suitability for acne treatment would require further clinical investigation.

| Expression validation of hub genes in human keratinocyte (HaCaT) cells
HaCaT cells are immortalized human keratinocytes which are frequently utilized as a model for skin-related studies due to their inherent resilience and retained differentiation capacity. 68HaCaT were stimulated with Propionibacterium acnes (P.acnes) to mimic the pathogenesis of acne. 10 Following incubation at 24 and 48 h, the mRNA levels of PTPRC, CXCL8, ITGB2, and MMP9 were quantitatively assessed via qRT-PCR.Compared to untreated controls, the expression of four genes was significantly upregulated in cells treated with P. acnes (Figure 11A), which corroborates our bioinformatics findings.This result highlights the congruence between experimental and computational data, solidifying the reliability of our preliminary gene selection.The elevation in gene expression signifies the potential role of these genes in the pathogenesis of acne and the potent impact of P. acnes in upregulating them.The consistency in the expression trend across these time points further strengthens the validity of our results.It is pertinent to highlight that PTPRC, ITGB2, and MMP9 are genes involved in immune response modulation and cellular adhesion, hence their upregulation potentially signifies enhanced inflammatory response and cellular interaction mechanisms in the acne model. 15,69ISA, with its specificity and sensitivity, is an ideal technique to validate protein expression levels in cell culture supernatants. 70e secretion levels of CXCL8 and MMP9 in the cell culture supernatant were assessed via ELISA, revealing a notable increase in these molecules following P. acnes exposure (Figure 11B).

| Impact of CXCL8 gene knockdown on HaCaT cell functions
2][73] Given the frequent presence of purulent inflammation in acne, the recruitment of neutrophils to inflamed sites, 54 potentially exacerbating the inflammatory response, is deemed a critical process.Neutrophils, being the primary cells of the innate immune system, contribute significantly to the early stages of inflammation. 74Their excessive recruitment can cause tissue damage, leading to the pustular form of acne lesions. 3,9To elucidate the functional relevance of the hub gene CXCL8, a siRNA-mediated gene knockdown was executed in HaCaT cells stimulated by acne.siRNA, or small interfering RNA, offers a precise method to downregulate or "knock down" the expression of specific genes, allowing researchers to study the resultant phenotype. 75The qRT-PCR and WB assays

HaCaT cells
Based on our DrugBank analysis, considerations for drug selection for cellular validation included safety, efficacy, and practical applicability.Arsenic trioxide, a known toxin, can be harmful upon prolonged or high-dose exposure. 77While arsenic trioxide has shown therapeutic efficacy in certain malignancies like acute promyelocytic leukemia, 78,79 its use for nonlethal conditions such as acne, given its inherent toxicity, is inadvisable.In contrast, silica dioxide, primarily used as a food additive and anticaking agent, lacks significant pharmacological activity, making its therapeutic efficacy for acne potentially limited. 80However, it may serve as a drug delivery system capitalizing on its ability to encapsulate and slowly release active compounds, thus enhancing their stability or delivery. 81,82tinoic acid, with established research and application in dermatological treatments, is considered an effective component for acne therapy owing to its ability to modulate keratinocyte differentiation, inhibit sebaceous gland activity, and mitigate inflammation. 83,84Methotrexate, an immunosuppressant, modulates inflammatory responses and has proven effective at low doses for certain immune-mediated diseases, suggesting potential therapeutic benefits for acne and thus presents an avenue for further exploration as an acne intervention, especially for severe or persistent forms. 85,86Methotrexate functions by inhibiting the enzyme involved in the synthesis of folic acid, thus hindering DNA synthesis and cellular replication. 87This mechanism is particularly valuable in conditions characterized by rapid cell proliferation.
Based on the known actions and safety profiles of these drugs, we selected retinoic acid and methotrexate for cellular validation while excluding arsenic trioxide and silica dioxide.Such a decision underscores the importance of a thorough benefit-risk analysis in drug selection. 88CaT cells treated with P. acnes were exposed to varying concentrations of retinoic acid and methotrexate.The CCK-8 assay revealed no significant cytotoxicity at lower drug concentrations (Figure 13A).This underscores the potential therapeutic window where the drugs can exert their beneficial effects without inducing undue cellular harm.Notably, all drugs significantly attenuated the upregulation of the four core genes' mRNA levels in a dosedependent manner offering evidence of their molecular modulating effects, potentially curtailing inflammatory responses (Figure 13B).
ELISA results demonstrated that both drugs inhibited the secretion of CXCL8 and MMP9 which might further allude to their roles in controlling neutrophilic inflammation and matrix remodeling activities, respectively (Figure 13C).Additionally, a decrease in the secretion levels of pro-inflammatory cytokines (IL-6, IL-1β, and TNFα) was observed in the drug-treated groups, suggesting potential antiinflammatory properties of the drugs in acne treatment (Figure 13D).
This implies a multifaceted therapeutic potential for retinoic acid and methotrexate, impacting not just gene expression but also inflammatory mediator secretion, further elucidating their promise as viable acne treatments.ficking. 97,98Integrins, to which ITGB2 belongs, are cell adhesion molecules that play crucial roles in interactions between cells and their surroundings. 99,100The observed association between ITGB2 and immune cells, especially macrophages and neutrophils, indicates its potential involvement in immune cell recruitment and adhesion within acne lesions.Matrix metallopeptidase 9 (MMP9) is a pivotal enzyme known for its proteolytic activity on the extracellular matrix, playing a fundamental role in tissue remodeling and inflammation. 101Our study identified amplified MMP9 expression and secretion levels, particularly in the acne context, where increased MMP9 could contribute to the degradation of dermal structures, thereby facilitating the onset and progression of inflammatory acne lesions. 102,103This heightened activity of MMP9, especially following exposure to bacterial pathogens such as P. acnes, might intensify the inflammatory response.Its association with neutrophil infiltration adds another dimension, as neutrophils release MMP9 upon activation, potentially leading to the tissue damage characteristic of inflamed acne lesions. 104 observed a pronounced interplay between hub genes and therapeutic agents like retinoic acid and methotrexate on acne-induced HaCaT cells, as evident from alterations in mRNA levels and protein secretion posttreatment, which has implications for modulating inflammatory responses in acne.Retinoic acid, a metabolite of vitamin A, is renowned for its pivotal role in skin cell differentiation, proliferation, and its long-standing therapeutic application in acne management. 83,84In our study, methotrexate, an antimetabolite and antifolate drug primarily known for its use in rheumatoid arthritis, certain cancers, and other inflammatory conditions, 86,87 displayed significant impacts on the cellular dynamics of acne-induced HaCaT cells.Notably, its influence on the mRNA levels of core genes and its modulatory effect on pro-inflammatory cytokine secretion underscores its potential as an alternative therapeutic strategy for acne, especially in cases resistant to conventional treatments.The modulation of hub genes, including markers such as CXCL8, provides compelling evidence of its therapeutic promise, suggesting avenues for personalized acne therapies.However, considering the broad systemic effects and therapeutic index of methotrexate, its application in acne therapy necessitates careful scrutiny. 105,106Additionally, these findings emphasize the need for further investigation into the molecular mechanisms of for optimized therapeutic applications, especially given their recognized benefits in other inflammatory conditions. 107,108Another certain limitation is potential biases from exclusive reliance on three GEO database platforms.In moving forward, a multifaceted approach that integrates multiple molecular pathways and cellular interactions is paramount, urging collaborative efforts to bridge molecular biology insights with clinical application for innovative acne treatment strategies.

| CON CLUS IONS
This study presents a comprehensive genetic profile associated with acne by meticulously analyzing microarray data, emphasizing the pivotal role of hub genes in acne pathogenesis.The consistent findings across multiple datasets underscore the robustness of our approach and the potential of these genes, especially PTPRC, CXCL8, ITGB2, and MMP9, for diagnostic and therapeutic interventions.It is essential, however, to proceed with cautious optimism, recognizing the need for subsequent research to delve deeper into the mechanistic roles of these genes, their interactions with environmental factors, especially immune cell dynamics, and their overall potential in shaping future therapeutic interventions.Further exploration, especially in the realm of immune cells like NK cells and broader regulatory networks, will be pivotal in translating these insights into effective clinical applications.Such endeavors will be instrumental in the evolution of personalized, targeted, and effective acne therapeutic interventions, ultimately aiming for improved clinical outcomes.
approaches to transmute fluorescence signals into quantifiable numerical values.RMA, in particular, corrects for background noise and adjusts for nonspecific binding, ensuring that variations across arrays are primarily attributable to differences in gene expression levels and not due to technical variability.Quantile normalization, on the other hand, ensures that the distribution of probe intensities is consistent across different arrays.Such normalization techniques ensure that variability introduced during sample preparation or hybridization does not unduly influence subsequent analyses.This process culminated in the derivation of the final expression matrix from the processed raw data files.Normalization is crucial in microarray analyses to minimize potential biases and variations that can arise due to non-biological factors such as array production or RNA quality.

2. 9 |
Elucidation of potential transcription factor (TF)-and miRNA-target gene regulatory networks and analysis of drug-gene interactions The regulatory roles of miRNAs and transcription factors (TFs) in gene expression modulation are paramount in various biological processes and pathologies.Unveiling these regulatory relationships can elucidate intricate molecular pathways and suggest intervention points for therapeutic strategies.miRNAs are short RNA molecules that play pivotal roles in posttranscriptional gene regulation, often leading to translational repression or target mRNA degradation.Online tools, namely miRNet (available at: https:// www.mirnet.ca/ , tissue: skin) 48 and TarBase v8.0 (accessible at: https:// www.netwo rkana lyst.ca/ ), 49 were synergistically employed to identify potential miRNAs targeting the diagnostic hub genes.These platforms provide comprehensive miRNA-target interactions, amalgamated from numerous experimental studies and supported by literature citations.TFs drive gene expression patterns, and their relationships with hub genes provide insights into transcriptional control mechanisms in acne pathogenesis.Transcription factors regulate the transcription of genes by binding to specific DNA sequences, modulating gene expression patterns.Upstream transcription factors (TFs) were predicted using the ChIP-X Enrichment Analysis Version 3 (ChEA3) portal (available at: https:// maaya nlab.cloud/ ChEA3/ ), 50 which uses large-scale chromatin immunoprecipitation (ChIP) datasets to predict TF-gene interactions.The resultant regulatory networks were depicted in both tabular format and visually via Cytoscape 3.9.1.Furthermore, potential drug interactions with the identified genes were explored using the DrugBank database 51 (available at: https:// go.drugb ank.com/ ).DrugBank is a comprehensive, highquality bioinformatics and cheminformatics resource that combines detailed drug data with comprehensive drug target and drug action information.Identifying these interactions can aid in repurposing existing drugs or guiding new drug development strategies tailored for the condition under study.By understanding the overlap of drugs targeting multiple hub genes, researchers can prioritize potential multitarget treatments.A set of potential therapeutic drugs for each hub gene was extracted, and intersections were determined to highlight promising future treatment candidates.Detailed drug-gene interactions were itemized in tabular form to provide a structured overview of potential therapeutic interventions, aiding in the translational application of our findings.
All experiments were conducted in triplicate.Data are presented as the mean ± standard deviation (SD).Differences between groups were determined using either a Student's t-test or one-way ANOVA with subsequent post hoc analysis.A p-value <0.05 was considered indicative of statistical significance.3| RE SULTS3.1 | Identification of differentially expressed genes (DEGs)In our investigation, a total of 164 DEGs were identified from 6 paired samples of acne (lesion) and healthy skin tissues (control), with 140 being upregulated and 24 downregulated, based on a fold change threshold of ≥1.0 and a p-value <0.05, as analyzed in the GSE6475 dataset using the limma package (Figure1A).The utilization of paired samples ensures a better controlled internal comparison, reducing interindividual variability and enhancing the reliability of findings.The limma package, an established tool in the bioinformatics domain, allows for linear modeling of data and aids in identifying differentially expressed genes more precisely.Additionally, from the GSE108110 dataset, 500 DEGs were identified (372 upregulated and 128 downregulated) across 18 paired samples (Figure1B).Likewise, in the GSE53795 dataset, 765 DEGs were ascertained from 12 paired samples (432 upregulated and 333 downregulated) F I G U R E 1 Visualization of differentially expressed genes (DEGs) across datasets: GSE6475, GSE108110, and GSE53795.Volcano plots graphically represent the statistical significance versus fold change, making it easier to identify DEGs of interest.(A) Volcano plot highlighting four hub genes for GSE6475.(B) Volcano plot highlighting four hub genes for GSE108110.(C) Volcano plot highlighting four hub genes for GSE53795.Heatmaps offer a visual representation of the magnitude and direction of gene expression, allowing for swift identification of patterns across samples or conditions.(D) Heatmap of 142 DEGs from GSE6475 that satisfy the criteria: expression value >5, |log 2 FC| > 1, and adjusted p-value <0.01.(E) Heatmap of 123 DEGs from GSE108110 meeting the criteria: expression value >5, |log 2 FC| > 1.5, and adjusted p-value <0.01.(F) Heatmap of 109 DEGs from GSE53795 adhering to the criteria: expression value >5, |log 2 FC| > 2, and adjusted p-value <0.01.employing the same criteria (Figure 1C).The near-equal distribution of upregulated and downregulated genes here might hint toward a balanced gene expression alteration mechanism in this dataset.The similar criteria employed across datasets ensures consistency in results and aids in the comparison between datasets.The variation in the number of upregulated and downregulated genes in different datasets could be influenced by varying pathophysiological conditions, underlying genetics, or sample quality.Figure 1A-C highlight the four key hub genes.The heatmap showcases 142 genes from the GSE6475 dataset meeting the criteria of expression value >5, |log 2 FC| > 1, and adjusted p-value <0.01 (Figure 1D), 123 genes from GSE108110 adhering to expression value >5, |log 2 FC| > 1.5, and adjusted pvalue <0.01 (Figure 1E), and 109 genes from GSE53795 fitting the criteria of expression value >5, |log 2 FC| > 2, and adjusted pvalue <0.01 (Figure

57 3. 3 |
The identification of these immunerelated pathways signifies the inflammation-centric nature of acne pathogenesis, possibly explaining the persistent inflammation observed in acne lesions.54,55Such enrichment in these pathways hints at the potential molecular mechanisms underpinning acne formation, emphasizing the pivotal role the immune system plays in this skin condition.56,Geneontology and pathway annotation analysis To delve deeper into the roles of the 104 DEGs associated with acne lesions, we conducted gene ontology (GO) and pathway analyses using the "clusterProfile" R package, employing a p-value <0.05 criterion.GO is a framework offering standardized vocabularies to describe gene product characteristics and facilitates the consistent description of gene product attributes across different databases.The most significant functional enrichments for these DEGs are illustrated in Figure 4. Biological process (BP): DEGs predominantly participated in processes such as leukocyte migration, leukocyte chemotaxis, and neutrophil migration (Figure 4A).These processes align with the histological findings of acne, where immune cells infiltrate the skin layers.Cellular component (CC): Main locations of DEGs included secretory granule lumen, cytoplasmic vesicle lumen, and vesicle membrane (Figure 4B).These cellular compartments are typically involved in immune cell signaling and mediator release, suggesting that these DEGs might play roles in immune cell activation and response.Molecular function (MF): DEGs were prominently involved in immune receptor activity, IgG binding, and serine-type endopeptidase activity (Figure 4C).Such functions reiterate the active involvement of the immune system in acne pathology.KEGG Pathways: DEGs were primarily enriched in pathways like cytokine-cytokine receptor interaction and NF-kappa B signaling (Figure 4D).The former underscores the communicative role of cytokines in acne, while the latter implies a potential link between acne and the NF-kappa B signaling pathway, which plays a central role in inflammation.

DEGs, which can
shed light on potential protein complexes and signaling cascades relevant in acne pathogenesis.A denser network suggests stronger and more numerous interactions among the proteins, which could be crucial nodes or hubs in the biological processes.Notably, genes such as PTPRC, CXCL8, and MMP9 were among the top interactors.Their position in the network suggests they may act as crucial regulators or effectors in the cellular processes linked to acne.Key modules were identified using the MCODE plugin.Module 1 exhibited the highest interaction score (13.077) with 14 genes and 170 edges (Figure 5B), followed by Module 2 (score 5.429) and Module 3 (score 4.167), detailed in Figure 6C,D, respectively.

F I G U R E 4
These high AUC values not only reinforce the diagnostic potential of the hub genes but also indicate their reliability and consistency across different datasets.These outcomes underscore the substantial diagnostic efficacy of the four hub genes in acne, suggesting their potential utility as diagnostic biomarkers in acne patients.Given their superior diagnostic accuracy, these genes could pave the way for new diagnostic techniques, eliminating the subjectivity involved in visual acne grading.Functional and pathway analysis of 104 hub genes.(A) GO analysis denoting biological processes which elucidate the broader roles these genes play in cellular life cycles.(B) GO analysis highlighting cellular components representing locations at the subcellular level where gene products are active.(C) GO analysis detailing molecular functions signifying the elementary activities, like binding or catalysis, performed by a gene product at the molecular level.(D) KEGG pathway exploration provides insights into the broader biological functions and networks the genes are involved in.

F I G U R E 5 F I G U R E 6 F I G U R E 8
Construction and identification of the PPI network: (A) PPI network of the 104 central genes with node size proportional to the number of connections.Larger nodes indicate higher interaction frequency.(B) Hub module 1 featuring 14 nodes interconnected by 170 edges.(C) Hub module 2 with 15 nodes and 76 edges.(D) Hub module 3 composed of 13 nodes connected by 50 edges.This figure visually communicates the intricate relationships and interactions among the central genes, highlighting potential key players in the network.Hub gene identification.(A) Four hub genes extracted via nine algorithmic strategies, visualized in an UpSet plot.(B) Ranking of the top 10 genes across the nine algorithms applied on 104 DEGs.This figure provides an in-depth analysis of the most impactful genes determined by various computational strategies, showcasing the congruence and distinctions across different methodologies.F I G U R E 7 Heatmap correlations among four paramount hub genes (PTPRC, CXCL8, ITGB2, and MMP9) with 24 infiltrating immune cells across datasets: (A) GSE6475, (B) GSE108110, and (C) GSE53795.Significance denoted by *p < 0.05 and **p < 0.01.By correlating hub genes with immune cell infiltrates, we can elucidate the immune responses and potential inflammatory pathways in acne lesion development.Correlation scatter plots juxtaposing PTPRC, CXCL8, ITGB2, and MMP9 with specific immune infiltrates (specifically neutrophils and NK cells) across datasets: (A) GSE6475, (B) GSE108110, and (C) GSE53795.These scatter plots offer a granular view of the gene-immune cell interrelations, underlining the strength and directionality of the associations.
The consistent elevation of CXCL8 and MMP9 in the supernatant underlines the active secretion and possibly the functional involvement of these molecules in the acne-like condition.It is critical to understand that such changes in gene expression could be indicative of the cells' response to infection, potentially leading to a pathological state similar to acne.This complements our understanding of the inflammatory cascade induced by bacterial interactions in acne-affected regions.
Figure12A,B) demonstrated that both siCXCL8-1 and siCXCL8-2 cell strains exhibited a significant reduction in transcription and translation levels of CXCL8 compared to control group siNC.This suggests a successful target-specific gene silencing activity of the employed siRNAs.This approach offers an avenue to explore the precise role and influence of CXCL8 in the complex cascade of inflammatory reactions witnessed in acne.Following successful knockdown of CXCL8, ELISA detected a substantial reduction in pro-inflammatory cytokines (including IL-6, IL-1β, and TNFα) in the cell culture supernatant implying that CXCL8 might be instrumental in orchestrating the release or synthesis of other pro-inflammatory cytokines in the acne milieu (Figure12C).The interplay between CXCL8 and these proinflammatory cytokines underscores the interconnectedness of the inflammatory pathways, where a perturbation in one component can reverberate through the entire network.72,76CCK-8 assay results further revealed a marked decline in cell proliferation rates upon CXCL8 knockdown (Figure12D).Crucially, cell migration-a vital biological function in wound healing and inflammation-was F I G U R E 11 Validation of hub genes in HaCaT cells post P. acnes stimulation.(A) Quantitative assessment of mRNA levels of PTPRC, CXCL8, ITGB2, and MMP9 at 24 and 48 h post P. acnes treatment using qRT-PCR.(B) ELISA analysis of CXCL8 and MMP9 secretion levels in the cell culture supernatant after P. acnes exposure.F I G U R E 1 2 Functional implications of CXCL8 knockdown in HaCaT cells.(A) qRT-PCR assay results showcasing the efficacy of siCXCL8-1-and siCXCL8-2-mediated knockdown of CXCL8 gene.(B) Western blot (WB) assay displaying the translation levels of CXCL8 post siRNAmediated knockdown.(C) ELISA analysis of pro-inflammatory cytokines (IL-6, IL-1β, and TNFα) secretion post CXCL8 knockdown.(D) CCK-8 assay results illustrating the effect of CXCL8 knockdown on HaCaT cell proliferation rates.(E) Scratch wound-healing assay results post CXCL8 knockdown.(F) Assessment of cell migration post CXCL8 knockdown using the transwell assay.discernibly diminished post-CXCL8 knockdown, as evidenced by both the scratch wound-healing (Figure 12E) and transwell assays (Figure 12F).The observed decline in cell migration and proliferation further underscores the gene's multifaceted role, possibly extending beyond inflammation to encompass tissue repair and regeneration.This comprehensive assessment of CXCL8 knockdown also provides a foundational understanding for potential therapeutic interventions targeting this gene.

F I G U R E 1 3
Cellular impact of retinoic acid and methotrexate on acne-induced HaCaT cells.(A) CCK-8 assay results representing the cytotoxicity levels of retinoic acid and methotrexate at various concentrations on HaCaT cells treated with P. acnes.(B) mRNA levels of the four core genes (PTPRC, CXCL8, ITGB2, and MMP9) posttreatment with retinoic acid and methotrexate.(C) ELISA findings on the secretion levels of CXCL8 and MMP9 post-drug treatment.(D) ELISA analysis of secretion levels of pro-inflammatory cytokines (IL-6, IL-1β, and TNFα) in drug-treated cells.This study delved into the gene expression profiles related to acne, leveraging Affymetrix microarrays from the Gene Expression Omnibus (GEO) database.Analyzing three specific datasets (GSE6475, GSE108110, and GSE53795), our analysis revealed differentially expressed genes (DEGs) with 104 intersecting genes identified across datasets, primarily consisting of 97 upregulated and 7 downregulated genes.This implies the potential universality of genetic markers that might influence acne pathogenesis significantly.Variations in gene expression patterns across different datasets emphasize the genetic heterogeneity in acne and the importance of multi-dataset validations.The study identified four consistently overexpressed hub genes-PTPRC, CXCL8, ITGB2, and MMP9across acne-affected and normal skin samples.These genes, known for roles in inflammation and immune modulation, further solidify the understanding of acne as an inflammatory condition.Notably, a strong positive correlation was observed between these hub genes and neutrophil infiltration levels, corroborating neutrophils' role in combating bacterial infections typical in acne lesions.89Conversely, a negative correlation with NK cells, mainly involved in antiviral and antitumor activities, suggests a more complex acne pathology and the potential for novel therapeutic targets.Further research is essential to explore this aspect and to understand the potential roles and interactions of these cellular entities in acne's pathophysiology.The diagnostic potential of the hub genes was robustly validated by the ROC curve analysis across multiple datasets, consistently showing AUC values exceeding 90%.Such high diagnostic accuracy indicates their reliability as objective molecular markers, which could revolutionize acne assessment, transitioning from the traditional subjective visual grading to a more consistent molecular-based approach. 90This study's methodological rigor was further highlighted by the use of paired samples, which effectively countered interindividual variability, ensuring the credibility of the identified differentially expressed genes (DEGs).The elucidation of the molecular dynamics of these genes was facilitated through the examination of potential miRNA regulators and associated transcription factors, offering insights into upstream genetic control mechanisms that may guide therapeutic strategies. 91Validation of the upstream regulators of these hub genes may further unveil acne's genetic foundation, potentially aiding in the development of targeted genetic interventions and tailored treatments.Cells experiments in HaCaT following P. acnes stimulation underscored the genes' differential expression and functional significance in acne pathogenesis, with CXCL8 knockdown influencing pro-inflammatory cytokine secretion, cell proliferation, migration, and wound healing.Chemokine (C-X-C motif) ligand 8 (CXCL8), commonly recognized as interleukin-8 (IL-8), emerged central significance in acne's inflammatory cascade.73,92Our study consistently revealed elevated levels of CXCL8 following P. acnes exposure.As a pivotal inflammatory chemokine, CXCL8 plays an instrumental role in neutrophil chemoattraction, activation, and degranulation.The prominence of purulent inflammation observed in acne, characterized by the excessive recruitment of neutrophils-primary responders of the innate immune system-underscores the exacerbation of the inflammatory response, potentially leading to the pustular form of acne lesions.Such findings are coherent with CXCL8's biological function and align with previous reports linking it to inflammatory dermatological conditions.93,94The other three genes also have potential utility as reliable biomarkers and therapeutic targets.Protein tyrosine phosphatase receptor type C (PTPRC), widely known as CD45, has been spotlighted for its integral role in T-and B-cell receptor signaling, crucially influencing immune responses.95,96PTPRC's surge may denote an augmented inflammatory response, potentially stimulating lymphocyte activation and differentiation, a process inherently linked with acne pathogenesis.95Integrin subunit beta 2 (ITGB2), a principal component of the integrin family and crucial for leukocyte integrins, is central to leukocyte adhesion, transmigration, and immune cell traf- methotrexate in acne and its synergy with other targeted interventions.The study not only bolsters the recognized benefits of methotrexate in other inflammatory conditions but also hints at its broader applicability in the treatment landscape of acne.Despite promising findings, it is crucial to recognize the study's limitations, particularly its reliance on in vitro HaCaT cell experiments, which may not entirely represent in vivo conditions.Nevertheless, the validation of hub genes, especially CXCL8, and their influence on cellular functions, paves the way for potential targeted treatments and encourages further research into personalized acne therapies.Retinoic acid and methotrexate demonstrated profound impacts on gene expression and inflammatory marker secretion in acne-induced HaCaT cells, emphasizing their therapeutic promise.Both compounds showcased potential in modulating hub genes' behavior, suggesting avenues