To profile circulating lncRNAs, 25 patients (19 females and 6 males) aged 27.07±14.76 years, and 25 healthy volunteers (18 females and 7 males) aged 33.25±11.26 years, were recruited from the Dermatology Unit of the University of Campania Luigi Vanvitelli (Caserta, Italy), and venous blood samples were collected from each participant in the study. Dysregulated lncRNAs in patients with HS were identified and used for further analysis and validation. The association between the selected lncRNAs and HS characteristics and prognosis was then investigated in the circulation of an expanded cohort of patients with HS (n=25) and healthy controls (n=25). Written informed consent was obtained from all the participants. The entire trial was conducted in agreement with the guidelines of The Ethics Committee of the University of Campania Luigi Vanvitelli, who also approved the research (approval no. 12478/20). The blood samples were handled in accordance with the guidelines of the Declaration of Helsinki.

A venous blood sample (10 ml) was collected and analyzed from each participant for the purposes of lncRNA identification and quantification. For human whole blood isolation, Lymphoprep™ (Axis-Shield Diagnostics, Ltd.) was used in accordance with the manufacturer's instructions.

From whole blood samples, total RNAs were extracted using Invitrogen® TRIzol™ reagent (cat. no. 15596-026; Thermo Fisher Scientific, Inc.), according to the manufacturer's instructions. RNA isolated with this protocol was extracted from all white cells (polymorphonuclear leukocytes and mononuclear cells) and subjected to a DNase digestion step. A concentration of 1 µl DNase I (2 U; RNase-free, cat. no. AM2222-24; Thermo Fisher Scientific, Inc.) for up to 10 µg RNA was used, in line with the manufacturer's instructions. The RNA integrity was subsequently evaluated using standard denaturing 1% agarose gel electrophoresis and the SYBR green dye (Qiagen, Inc.).

For lncRNA profiling, human RT² lncRNA PCR assays (Qiagen, Inc.) were applied. The RT² First Strand Kit (Qiagen, Inc.) was used for RT, which included an exclusive genomic DNA elimination step to remove any remaining contamination. Prior to hybridization to the chip (Human lncFinder RT2 lncRNA PCR Array; cat. no. 330721; Qiagen Inc.) RT² SYBR Green Mastermix for RT-qPCR (Qiagen, Inc.) was used. Microarrays were scanned using an Agilent microarray scanner (Agilent Technologies, Inc.), directed by GenePix Pro 6.0 software (Axon Instruments; Molecular Devices, LLC). Scanned images (in the .tiff format) were imported into Agilent Feature Extraction software v.12.2 (Agilent Technologies, Inc.) for grid alignment and expression data analysis. Expression data were corrected and normalized (according to the quartiles) using the robust multiarray average algorithm included in the Agilent software. Differentially expressed lncRNAs were then identified using fold change as the filter. The microarray data were selected applying threshold values of >2 and <−2-fold change under false discovery rate protection (P<0.05), and >25,000 human long noncoding transcripts were analyzed.

RT-qPCR quantification of the expression of lncRNAs and lncRNAs co-expressed with mRNAs was performed using the PowerTrack SYBR Green Master Mix (Qiagen, Inc.) using an Applied Biosystems™ QuantStudio 5 instrument (Thermo Fisher Scientific, Inc.), following precisely the manufacturer's protocol with the following thermocycling conditions: Enzyme activation at 95°C for 2 min 1X, denaturation phase at 95°C for 14 sec, followed by annealing/extension at 60°C for 60 sec; the final two phases were repeated 40 times. Reactions were performed in a mixture (10 µl) containing 1 µl cDNA template, 5 µl 2X PowerTrack SYBR Green Master Mix and 0.5 µM each of sense and antisense primers. Table I shows the sequences of primers used for the RT-qPCR analysis. RT-qPCR reactions were performed in triplicate for each sample, and the specificity of the PCR products was estimated using the dissociation curve. β2-microglobulin was used as the internal control. For quantitative results, expression of each lncRNA was represented as the fold change using the 2^−ΔΔCq method, and then analyzed for statistical significance (22,23).

Through searching three databases, RNAinter (http://www.rnainter.org/), LncRRIsearch (http://rtools.cbrc.jp/LncRRIsearch/) and NPinter (http://bigdata.ibp.ac.cn/npinter4), potential mRNA targets for each lncRNA were identified with adjusted P-value was set to <0.05. To determine the roles of differentially expressed mRNAs, Gene Ontology (GO) analysis was applied to categorize mRNAs according to their involvement in biological processes, cellular components or molecular function. Subsequently, differentially regulated mRNAs were uploaded into the Database for Annotation, Visualization and Integrated Discovery (https://david.ncifcrf.gov) for annotation and functional analysis, including gene set enrichment analysis and mapping gene sets to the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway. The KEGG database (http://www.genome.ad.jp/kegg) and BioCarta (https://www.hsls.pitt.edu/obrc/index.php?page=URL1151008585) were then used to predict the potential functions of these target genes in the pathways.

The lncRNA-microRNA regulatory network was defined as described in a previous report (24). The Pearson correlation method was used for miRNA analysis, and those pairs of miRNAs with significant correlations (≥0.90) were chosen to construct the network. Co-expression networks were generated using Cytoscape v.3.10.1 (https://cytoscape.org/); in this representation, each lncRNA corresponded to a node, and the connection of two genes was depicted by an edge, indicating a strong correlation (i.e., either positive or negative).

SPSS version 22.0 software (IBM Corp.) was used to analyze the data. Continuous variables were presented as the mean ± SD, or as the median (interquartile range). For each subject, >2-fold changes in the expression of distinct lncRNAs were calculated by dividing the standardized expression levels of the lncRNAs. Differences in the expression levels of the lncRNAs between the HS groups and the healthy controls, with at least 5 biological replicates/group, were analyzed using unpaired Student's t-test, and P<0.05 was considered to indicate a statistically significant value.

The identification of circulating lncRNA signatures in patients with HS was performed in a pilot study (unpublished data). The aim was to identify any significant changes in lncRNA expression in the blood samples of patients with HS and non-HS participants. LncRNA arrays were performed using RNA extracted from the total population of white cells. A 2-fold expression difference was used as the cut-off, and a total of 133 lncRNA transcripts were found to be dysregulated (109 lncRNA transcripts were upregulated, and 24 lncRNA transcripts were downregulated; each P<0.05) in patients with HS compared with non-HS participants (Fig. 1). Using high-signal intensity and ≥6-fold deregulation of gene expression as filters, eight lncRNA candidates were selected. Only three of these eight lncRNA candidates (namely, lncRNA-TINCR, lncRNA-RBM5-AS1 and lncRNA-MRPL23-AS1) were found to be consistently dysregulated in the cohort of blood samples (25 patients with HS and 25 healthy non-HS control participants). The expression levels of circulating lncRNA-TINCR lncRNA-RBM5-AS1 and lncRNA-MRPL23-AS1 in patients with HS were found to be significantly increased compared with non-HS participants (Fig. 2; P=0.004, P=0.002 and P=0.002, respectively).

As shown in Fig. 2, the expression levels of these three lncRNAs in the peripheral blood was consistently found to be higher in HS blood compared with that of the healthy participants; therefore, the origin of the total RNA was investigated. Further analysis of the data showed that no dysregulated molecules appeared to be sex specific (data not shown). The large standard error values observed for the other lncRNAs is possibly due to the grouping of the samples, given the inherent variability in expression levels among those groups. It is noteworthy that a larger standard error was associated with small expression values, which may be less reproducible. In fact, at low levels of expression, minor technical variations or biological fluctuations can result in relatively large proportional changes, making it challenging to consistently measure these small quantities accurately across samples; however, the absence of dysregulation of these lncRNAs is clear for these groups.

Gene enrichment analysis was performed to determine which genes were involved in the GO categories of biological process, cellular component and molecular function. As shown in Fig. 3, the most enriched GO terms, targeted by lncRNA-co-expressed mRNAs, were found to be responses to regulation of biological and cellular processes (GO: Biological process), cellular and anatomical entities (GO: Molecular function) and intracellular anatomical structures (GO: Cellular component). Examples of relevant functional terms are ‘Cytokine receptor activity’, ‘Non-coding RNA processing’ and ‘Histone deacetylase complex’.

KEGG pathway analysis revealed an involvement of lncRNA-co-expressed mRNAs in pathways such as ‘Toll receptor signaling pathway’, ‘Cadherin signaling pathway’, ‘RHOA signaling’, ‘CCKR signaling pathway’ and ‘ERK5 signaling pathway’ (Fig. 4).

IPA of lncRNA-co-expressed mRNAs revealed that MYC appeared in several of the pathways associated with the lncRNAs co-expressed mRNAs, followed by YAP1 and SEPTIN7. To gain an understanding of the involvement of these genes in HS etiology, RT-qPCR was performed to measure the expression level of these targets. Both SEPTIN7 and IL-13 were found to be upregulated in the blood of patients with HS, whereas YAP1 and MYC1 exhibited significantly lower expression levels compared with healthy controls (Fig. 5). Taken together, these results suggested that such mRNAs in association with lncRNAs may be involved in the inflammatory state, the maintenance of cellular morphology and tissue development, or the progression of HS.

The lncRNA-miRNA co-expression network was constructed using Cytoscape software (Fig. 6), and the potential interactions between miRNAs and lncRNAs were investigated. LncRNA-TINCR, lncRNA-RBM5-AS1 and lncRNA-MRPL23-AS1 were found to interact with 145 miRNAs, and these were involved in three relevant pathways: i) RhoA signaling and signaling by Rho family GTPases involving Septin3 and Septin7; ii) glucocorticoid receptor signaling/STAT3 and; iii) macrophage alternative activation signaling. Among these pathways, the RhoA signaling and signaling by Rho family GTPases involving Septin3 and Septin7 appears to be the most enriched one. The network indicated that each lncRNA was associated with a large number of target miRNAs, suggesting an inter-regulation of lncRNAs and miRNAs only in the HS state.