Circulating miRNome profiling data in Behçet's syndrome

We conducted a screening analysis to assess the presence of a characteristic extracellular circulating microRNAs (ci-miRNAs) profile in Behçet's syndrome (BS). Total RNA was extracted from platelets-free plasma (PFP) samples obtained from 16 BS patients and 18 healthy controls. Ci-miRNAs profiling was conducted by using dedicated Agilent microarray hybridization and data extraction technology. Statistical analysis of data extracted from microarray scanning revealed the deregulation of 36 ci-miRNAs, which turned out be differentially expressed between BS patients and healthy controls. Detailed experimental methods and data analysis were described here. The raw and normalized microarray data were deposited into Gene Expression Omnibus (GEO) under accession number GSE145191.


a b s t r a c t
We conducted a screening analysis to assess the presence of a characteristic extracellular circulating microRNAs (ci-miRNAs) profile in Behçet's syndrome (BS). Total RNA was extracted from platelets-free plasma (PFP) samples obtained from 16 BS patients and 18 healthy controls. Ci-miRNAs profiling was conducted by using dedicated Agilent microarray hybridization and data extraction technology. Statistical analysis of data extracted from microarray scanning revealed the deregulation of 36 ci-miRNAs, which turned out be differentially expressed between BS patients and healthy controls. Detailed experimental methods and data analysis were described here. The raw and normalized microarray data were deposited into Gene Expression Omnibus (GEO) under accession number GSE145191.

Value of the Data
• The provided data represent the extracellular ci-miRNAs profile of patients affected by Behcet's syndrome, a rare systemic vasculitis. • The data provide basis for the prediction of deregulated ci-miRNAs association with pathogenic pathways in Behçet's syndrome via in silico target analysis. • The data could potentially provide guidance for the discovery of non-invasive biomarker candidates for BS diagnosis by validation in a larger and independent cohort.

Patients and healthy controls
Sixteen BS patients and 18 HC were included in the present study as the microarray screening cohort.
No statistically significant difference in both age and sex between the two groups was detected.
Independently from the different baseline clinical features and from the active manifestations at the enrollment, our cohort well represented the different phenotypes of BS (namely the mucocutaneous and articular, the ocular and neurological and the vascular manfestations) [1] . Data presented as mean with interquartile range or number ( n ) and relative percentage when applicable. No statistically significant differences were found between groups for mean age and sex ratio evaluated by Student's T test and Chisquare test respectively. BS = Behçet's syndrome patients; HC = heathy controls; IQR = Interquartile Range.
Demographic features of patients and controls included the microarray screening cohort are reported in Table 1 .

Deregulated ci-miRNAs identification
RNA samples were subjected to miRNA profiling using the dedicated Agilent microarray technology. The statistical analysis of the whole miRNome data revealed the presence of 36 DE ( p < 0.05; -1 > log 2 FC > 1) ci-miRNAs between BS patients and HC, 17 down-regulated and 19 upregulated (listed in Table 2 ). The maximum up-regulation and down-regulation fold change values were, respectively, 2.45 and -2.19. The identified profile mainly comprised human sequences (indicated by the "hsa" prefix) with only 7 entities being of viral origin.
Complete microarray data (both as raw and processed datasets) are available in the NCBI's Gene Expression Omnibus (GEO) database under series accession number GSE145191, in accordance with the MIAME guidelines. For each sample, the overall miRNA profile is available both as raw and normalized mean fluorescence intensity values (contained respectively in the GEO Supplementary file and Sample table) associated to each microarray probe.
Samples characteristics and GEO data description are provided in Table 3 .

Patients and healthy controls
Sixteen BS patients, fulfilling the International Criteria for Behçet's Disease (ICBD) [2] , were recruited at the Florence Behçet Center (Azienda Ospedaliero-Universitaria Careggi), while 18 healthy controls (HC) were enrolled as blood donors at the Transfusional Medicine Centre of Azienda Ospedaliero-Universitaria Careggi.
All subjects were free from any laboratory or clinical sign of malignancies, infections or other immune-mediated diseases at blood sampling.

Sample collection and handling
Eight ml of peripheral blood were collected from each subject in BD Vacutainer K2 EDTA tubes (BD, Franklin Lakes, NJ, USA) by standard venipuncture. Among all commonly used anticoagulants, EDTA was selected due to its reported minimal effect on the circulating microRNA (ci-miRNA) profile during sample preparation step, as well as for the absence of interference with downstream applications [3] . Platelets Free Plasma (PFP) was obtained from peripheral blood samples by a double centrifugation protocol (1500 g for 15 min at room temperature followed by careful supernatant collection and centrifugation at 13,0 0 0 g for 3 min to eliminate platelets). Supernatants were carefully collected (making sure not to disturb the pellet) and finally aliquoted into fresh 1.5 ml RNAse-free tubes and stored at -80 °C until use. The PFP plasma preparation was selected considering that the accurate and reliable measurement of extracellular ci-miRNAs is dependent on the removal of residual platelets prior to freezing plasma sample. Any sample showing clots or signs of hemolysis (red/pink plasma discoloration against a white background, indicative of severe hemolyzed samples) by visual inspection was excluded from the analysis, considering the well-known ability of this features to significantly alter plasmatic miRNA quantification [4] . All blood samples were sent to our laboratory and processed within 2 h from collection, since several evidences reported how peripheral blood cells, including erythrocytes, can contribute to extracellular miRNAs found in plasma and serum following longer term storage [5] .

RNA extraction and quality control
Total RNA was extracted starting from a volume of 0, 25 ml of plasma. Trizol-LS reagent was used, following the manufacturer's protocol. RNA concentration and purity were assessed using the NanoDrop ND-20 0 0 spectrophotometer (Thermo Fisher Scientific Inc., Waltham, Massachusetts, USA). Qualitative sample analysis was performed using the Agilent 2100 Bioanalyzer system and Small RNA Assay Chips (5067-1548, Agilent Technologies, Santa Clara, California, USA). Only Samples showing acceptable concentration and quality were included in the analysis.

Microarray hybridization
Circulating miRNome profiling was performed using the Agilent Human miRNA 8 × 15k Microarray kit v3.0 and the miRNA Complete Labeling and Hyb Kit (Agilent Technologies, Santa Clara, CA, USA) following the manufacturer instructions.
Briefly, For each sample, a total amount of 100 ng of total RNA was dephosphorylated, 3 end-labelled with Cy3-pCp and dried. Cy3-labeled RNA in hybridization buffer was hybridized overnight (20 h, 55 °C) to Agilent Human microRNA Microarray Chips using the recommended hybridization chamber and oven. Following hybridization, the microarrays were washed with the Agilent Gene Expression Wash Buffer 1 for 5 min at room temperature followed by a second wash with preheated Agilent Gene Expression Wash Buffer 2 (37 °C) for 5 min before scanning.

Data extraction, pre-processing and differential expression analysis
Following microarray chips scanning procedure expression data were extracted from the TIFF images obtained from the G2565 Agilent's Microarray Scanning System (using a scan protocol with a resolution of 3 μm and a dynamic range of 20 bits) using the integrated Agilent Feature Extraction (AFE) v9.1 software (Agilent Technologies, Santa Clara, CA, USA). In order to perform data pre-processing and differential expression analysis, the processed information contained in the data source file were the following: gTotalGeneSignal, gMeanSignal (also contains background information), gIsGeneDetected, ControlType, ProbeName, and GeneName. Files were then processed using the R library AgiMicroRna v3.12 package, available on Bioconductor repository ( https://bioconductor.org/packages/release/bioc/html/AgiMicroRna.html ) [6] .
This package allows for raw data pre-processing using a variant of the robust multi-array average (RMA) algorithm that has been specifically implemented for Agilent miRNA microarrays.
This pre-processing method has been shown to have better precision than the one recommended by Agilent [5] .
Firstly, probe summarization was based on the median expression value for the replicated probes.
Median background values were then subtracted from the median expression values obtained from the AFE software. After obtaining the normalized total gene signal, probe signals were filtered based on the quality flags that AFE algorithm attaches to each feature. The quantile method was then applied in order to perform inter-array normalization ("normalizeBetweenArrays" function) in order to compensate for systematic technical differences between probe affinity and arrays.
After background correction, statistical significance of differences between study groups (BS patients vs HC) was assessed comparing mean microarray fluorescence intensity values using the Limma (Linear models for Microarray Data) differential expression t-test. A 2.0 times fold expression cutoff was applied to minimize the effect of probe background signal.
Finally, log2 transformation was used to obtain standardized expression values. Deregulated (DE) miRNAs were identified from the microarray analysis on the basis of a p < 0.05 by two-tailed t-test and log 2 -scale fold change > 1 or < -1.
To account for discrepancies in miRNA nomenclature due to the different miRBase releases, array features names were converted to MIMATIDs (unique mature miRNA accession number) using the mapping file available in miRBase ( http://www.mirbase.org ).

Statistical analysis
Unless otherwise stated, categorical variables were presented with counts and proportions, while continuous ones as the mean ± standard error of the mean (SEM) or median with IQR (interquartile range). All statistical tests were two tailed with a significance level of 0.05.

Ethics Statement
This study was conducted according to the Helsinki Declaration and approved by the local ethical committee of Azienda Ospedaliero-Universitaria Careggi, Florence, Italy (protocol number CE 13972). A signed written informed consent was obtained from each participant enrolled in the study.