MicroRNA and gene signature of severe cutaneous drug hypersensitivity reactions reveal the role of miR-483-5 p / miR-28-5 p in inflammation by targeting Granulysin gene

Purpose: To build a microRNA and gene signature of severe cutaneous adverse drug reactions (SCAR), including Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN). Methods: MicroRNA expression profiles were downloaded from miRNA expression profile of patients’ skin suffering from TEN using an array comprising of 372 miRNAs; download site: www.jacionline.org. The patient samples were eight TEN, ten SJS patients and twenty-two healthy individuals. A total of 192 microRNAs were found with unique expression patterns (overexpressed) in contrast with healthy skin controls and patients. Thereafter, the following databases were used for downstream analysis: geneMANIA, DIANA-miRPath version 3, DIANA-TarBase version 7.0, Ingenuity Pathway Analysis (IPA) as well as DAVID, STRING and GENECODIS online tools. Results: Granulysin (GNLY) geneMANIA database search yielded 21 interacting genes that were 64.6 % in physical interaction, 17 % in co-expression pattern. miRBD potential microRNAs that target the 21 genes were 79 miRs. Eighteen miRs overlap between the overexpressed miRs from SJS/TEN samples and the miRs targeting the 21 genes. Moreover, Ingenuity pathway analysis IPA revealed that the microRNAs were involved in inflammation. Conclusion: Analysis of differential microRNA expressions reveals two significant DE miRs that target Granulysin (483-5p/miR-28-5p). MiR-GNLY loop interactions in hypersensitivity reactions may function as biomarkers for SCAR including SJS and TEN.


INTRODUCTION
The clinical appearances of Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN range from mild severe cutaneous adverse drug reactions (SCAR), both are rare but seldom fatal.Yet, morbidity and mortality is in rapid increase worldwide [1].The pathogenesis of SCAR is a compound and multifactorial.It is known that drug antigen interactions, genetic background, and environmental factors might be involved [2].SCAR are among the most significant serious skin conditions amid cutaneous drug-provoked hypersensitivity adverse conditions [3].SJS and TEN rely on immune mediators in their hypersensitivity reactions that involve the cutaneous tissues and layers [4].Mortality still occurs in over 5 % of hypersensitivity adverse patients with SJS and greater than 30 % of the same patients [5].Nevertheless, the mechanisms of SJS and TEN are not fully elucidated.MicroRNAs or miRs are single stranded RNAs that are capable of posttranscriptional gene regulation via targeting their Mrna [6].MicroRNAs are very important regulators in many human diseases, for instance, cancer [7].Here, microRNAs are imposed as new regulators in adverse cutaneous reactions [8].T-cell lymphocyte association with SJS and TEN was recently observed to be linked with skin reactions, in addition to the recently reported Granulysin and natural killer cells inhibitory receptors [9].Currently, Granulysin acts as the substantial player and responsible for the dispersed keratinocyte death [10] and it has a role in cytotoxic T lymphocyte conditions [11].
In addition to Granulysin, multiple reports have revealed that numerous cytokines contributing an essential role in the apoptosis (programmed cell death) and/or the cytokine inflammations.In other reports levels of SJS and TEN cytokines were considerably greater in patient serum than healthy individuals' [12].
In the current study, Granulysin is deemed to establish core regulator and mediator along with his related/co-expressed genes and microRNAs for SJS and TEN pathogenesis.The databases and software used in this work not only attain information for DEG miRs in severe cutaneous SJS and TEN reactions, but also facilitate the approach towards discovering substantial roles in the pathogenesis of SJS and TEN.

MicroRNAs array: differentially expressed in SCAR
MicroRNA expression profile were downloaded from the inclusive miRNA expression profile of patients skin suffering from TEN using an array comprising of 372 miRNAs; Ichihara et al article's Online Repository at www.jacionline.org[13].The samples were achieved from eight TEN patients, ten SJS patients with and 22 healthy volunteers.A total of 192 microRNAs were found with distinctive expressions patterns (overexpressed) in contrast between healthy skin controls and patients.Then we upload the overexpressed miRs to DIANA-miRPath v3. to obtain the significant gene list and pathways [14].

Granulsyin core gene
We loaded Granulsyin gene (GNLY) to geneMANIA database [15]; this search was important to reveal the intern actions between this core gene in SJS and TEN with other genes and microRNAs.This search yielded 21 genes that most interacting with GNLY.

MicroRNAs targeting Granulsyin gene
Characterization of the GNLY expression profiles using, DIANA-TarBase v7.0 to find the potential microRNAs targeting these genes [16].

Functional annotation by DAVID web tool
DAVID stands for: Database for Annotation, Visualization and Integrated Discovery.DAVID tool was used to annotate/confirm the function of the genes list related to Granulysin.Selection of Gene Ontology GO standings with adjusted pvalue less than 0.05.

Network and Pathway analysis by Ingenuity Pathway Analysis IPA
The twenty-one genes related to Granulysin were then investigated and characterized in KEGG (Kyoto Encyclopedia of Genes and Genomes) [17], Ingenuity Pathway Analysis (IPA) tools by the GENECODIS software [18].REACTOME, is a tools for the visualization, interpretation and analysis of pathway knowledge to support basic research, genome analysis, modeling, systems biology and education [19].

1-DIANA-miRPath 70 pathway results for the microRNA
This search yielded more than 70 significant pathways; we chose the top ten according to p Value result and number of genes targeted by the miR list loaded to DIANA-miRPath tool.All summarized in Table 1.
Two miRs were overexpressed in SJS and TEN were among the newly identified from DIANA-TarBase Search.They are miR-483-5p/miR-28-5p.This confirm the role of both microRNAs-and their target granulysin protein.

MicroRNAs targeting 21 Granulsyin-related genes
In order to further characterize the 21 Granulsyin co-expressed genes, all genes were applied to DIANA-TarBase to find the potential microRNA targeting them.
The miRs found to be potential regulators are shown in Table 2.
After loading the whole gene list of 21 gene names.More accurate list with more genes and microRNA were acquired from MirDB database (The target prediction database, miRDB) [20].
We scrutinized our search for microRNAs that are predicted to target the 21 genes and the overexpressed 193 miRs and found 18 miRs overlapping, summarized in Table 3.
Most related to SJS/TEN are Chemokine signaling pathway, inflammatory response and Natural killer cell mediated cytotoxicity.

Network/pathway analysis by ingenuity pathway analysis
IPA for 21 genes related to Granulysin were then investigated and characterized in KEGG (Kyoto Encyclopedia of Genes and Genomes) [17], Ingenuity Pathway Analysis (IPA) tools by the GENECODIS software [18].REACTOME, is a tools for the visualization, interpretation and analysis of pathway knowledge to support basic research, genome analysis, modeling, systems biology and education [19].
IPA results for Granulysin and the important 21 genes are shown in Figure 2.  In this report, several databases and web tools analyses were approached to pinpoint pathways/ network and microRNAs targeting important proteins i.e.Granulysin contributing to pathogenesis of SJS/TEN.Predominantly, Granulysin were validated to be involved in SJS and TEN before.However, never were studied from microRNA perspective.Moreover, here we prove the potential microRNA regulators of Granulysin; miR-483-5p/miR-28-5p.
REACTOME pathways [19] results for the 21 gene clusters related to Granulysin indicated that the main pathways were associated with the following: DAVID tool [21] results were agreeing with the other pathways analyses; IPA.DAVID results for the 21 gene were as following: 1-Chemokine signaling pathway 2-inflammatory response i.e.SJS/TEN 3-and Natural killer cell mediated cytotoxicity These data were generated through the use of Ingenuity Pathways Analysis, a web-delivered application that enables biologists to discover, visualize and explore therapeutically relevant networks significant to their experimental results, such as gene expression array data sets.For a detailed description of Ingenuity Pathways Analysis, visit www.Ingenuity.com.

DISCUSSION
The complete genetics behind SCAR pathogenesis remains to be identified.STS and TEN, the major clinical manifestation of SCAR, continue to cause significant morbidity and mortality.In this report, databases and tools were used to investigate the differentially expressed microRNAs in SCAR.
Obtained from miRDB were 79 miRNAs predicted to target the 21 genes that interact with the Granulysin gene, the core causal gene in SCAR.Then, by overlapping the over-expressed miRNAs and potential miRNAs that target the 21 genes, a cluster of 18 miRNAs was obtained and found by Ingenuity pathway analysis (IPA) to be related to inflammation.Two significant DE miRNAs targeting Granylysin (miR-483-5p/miR-28-5p) were found.MiR-GNLY loop interaction in hypersensitivity may be useful biomarkers for SCAR, including SJS and TEN.
After the pathway-related microRNA clusters were identified, their genetic signature was generated to understand the pathogenesis of SJS and TEN.Our hypothesis that gene/miR signature can be applied to such a complex disease agrees with the genetic signature of HLA-B, -C, -A and -DRB1 alleles and cytochrome-P450 in a previous study [22].Many significant pathways were also identified as hallmarks of SCAR, such as the necroptosis pathway that is mediated by annexin1 gene which ligates to FPR1 receptor [23].Pathways are also important signatures for SCAR.In the current report, toll-like receptor was found to be the second most important in IPA and this pathway was proved to be involved in SJS and TEN [24].There are many reports of other pathways in SCAR-related inflammatory response [25].
In terms of drugs causing SCAR, most candidates were, Allopurinol, Carbamazepine, furosemide Erythromycin and paracetamol [26].While, in this study, specific Granulysin gene/miR cluster could potentially serve as a biomarker for purposes of diagnosis and prognosis.

CONCLUSION
Highlighted in this report are 18 Granulysin genetargeting microRNAs that can provide insight into the pathogenesis of SCAR.Specifically, 483-5p/miR-28-5p, which are over-expressed in SCAR and validated to directly target the GNLY gene, are novel regulators of miRs that can modulate Granulysin gene expression.It is evident that the MiR-GNLY loop signature/control in hypersensitivity reactions can function as biomarkers for SCAR, including SJS and TEN.From the identified pathway-related microRNA clusters, a genetic signature is presented to understand the pathogenesis of SJS and TEN and will be highly valuable in the diagnosis and prognosis of SCAR.

Figure 2 :
Figure 2: The confirmed 21 genes interacting with GNLY in IPA database

Figure 3 :
Figure 3: Genes targeted by the 18 miRs analyzed by IPA database

Figure 4 :
Figure 4: The twenty-on genes major pathways in IPA database

Table 1 :
DIANA-miRPath pathway results for the microRNA list Figure 1: The twenty-one genes that interacted most with GNLY

Table 3 :
The 18 microRNAs validated by PCR to be overexpressed in SJS/TEN and targeting granulysin-related genes