Abstract
Crohn’s disease (CD) and ulcerative colitis (UC) are inflammatory bowel diseases (IBD) presumably caused by dysregulated immune responses to the gut microbiota. Genetic association studies have implicated dozens of chromosomal regions or loci in IBD susceptibility. The next challenge is to explain the individual role of each of these modest effect loci in disease state. We have previously identified MAST3 as an IBD susceptibility gene through genetic fine-mapping of the 19p linkage region. Testing MAST3 in a reporter assay provided preliminary evidence that MAST3 modulates the activity of inflammation-related transcription factor nuclear factor kappa B (NF-κB). Here, we further characterized the function of MAST3 through an examination of the influence of the modulation of MAST3 expression on endogenous genome-wide expression patterns. More specifically, we looked at differential gene expression resulting from overexpression and knockdown of the MAST3 gene in epithelial and macrophage cell lines. We highlight a group of genes whose expression is modulated by MAST3 and correlate their expression with NF-κB activity. Their expression was found to be enriched in inflamed mucosal tissue of UC patients, confirming the importance of these genes in IBD. These MAST3-regulated genes are central to mucosal immune responses. Among them are pro-inflammatory cytokines (e.g. CCL20, IL8), regulators of NF-κB (e.g. TNFAIP3, LY96, NFKBIA), genes involved in interferon-induced defense against pathogen invasion (e.g. IFIT1, ISG15) and genes involved in cell adhesion and/or migration (e.g. CD44, TMOD1). Taken together, these results confirm MAST3 as a modulator of the inflammatory response through regulation of immune gene expression in the gut of IBD patients.
Keywords: MAST3, Inflammatory Bowel Disease, NF-κB
Introduction
Inflammatory bowel diseases (IBD) refer to Crohn’s disease (CD) and ulcerative colitis (UC), two common inflammatory diseases of the gastrointestinal tract most prevalent in urban areas of North America and Europe. Current consensus assigns the causes of IBD to environmental factors - such as cigarette smoking, appendicitis, hygiene, and gut microbiota - combined to genetic predisposition that contribute to the development of a dysregulated immune response to the gut microbiota leading to mucosal inflammation (1). Genome-wide genetic association studies have identified 99 IBD susceptibility regions or loci and some of them contain multiple genes (2, 3). Targeted association studies, such as ours (4), have also identified several susceptibility genes. However, estimation of heritability explained by these loci suggests that many more remain to be discovered. The important challenge is to identify individual role of the IBD genes to better understand the molecular mechanisms underlying disease onset and chronic inflammation.
In a previous study, we identified the microtubule associated serine/threonine protein kinase-3 (MAST3) gene as a genetic risk factor for inflammatory bowel disease (IBD) through association fine-mapping of the chromosome 19p region (also known as linkage region IBD6) (4). MAST3 is one of the least studied members of the MAST kinase family. Most members of the MAST family are ubiquitously expressed but their expression is highest in the brain (5). MAST2, the most studied member of this family, is involved in immune reactions, more specifically in the regulation of NF-κB (6). Based on this knowledge, we performed a knockdown of MAST3 in HEK293 cells and tested transcription factor NF-κB activity through reporter assays. We showed that a knockdown of MAST3 decreased significantly and specifically receptor TLR4 stimulated activity of NF-κB (4).
In the current study, we aimed to further characterize the role of MAST3 by modulating its expression, via overexpression and knockdown (KD) of the gene, in ex vivo cell culture systems and determining the impact on genome-wide endogenous gene expression. Specifically, we first compared transcriptional profiles of mock-transfected and MAST3-overexpressing HEK293 cells using genome-wide microarray analysis. Our genome-wide analysis highlighted a group of 28 MAST3-regulated genes whose expression levels were increased by ≥2-fold. In order to assess specificity of these results, we knocked down MAST3 in immune cells (THP1) and determined the expression levels of a subset of the genes identified in the overexpression experiment. We observed that for 7/9 genes tested, the expression was significantly reduced and thus the results of our KD confirmed our previous overexpression model results. To assess the importance of the MAST3-regulated genes in the clinical presentation of IBD, we compared gene expression in colonic mucosal tissue of healthy controls and of UC patient non-inflamed and inflamed regions. We found a significant enrichment of the expression of the MAST3 gene set in inflamed vs non-inflamed tissues (P=0.005).
The MAST3-regulated genes coordinate immune responses in intestinal inflammation. They include pro-inflammatory cytokines, regulators of NF-κB, interferon induced genes, and genes involved in adhesion and cell migration and a majority of these genes are regulated by NF-κB which suggests that MAST3 controls their expression via the NF-κB pathway.
Furthermore, the results of our expression study in patients biopsies suggest that MAST3 can regulate the expression of a set of genes that discriminates inflamed from non-inflamed tissues and that the overexpression of these genes is part of the clinical manifestation of the disease.
Materials and Methods
Cell culture
THP1 cells (ATCC TIB-202) were cultivated in RPMI medium supplemented with 10% fetal bovine serum, 2-mercaptoethanol (final concentration 0.05 mM), and penicillin (100 units/mL)-streptomycin (100 ug/mL). TLR4 293-hTLR4A-MD2-CD14 HEK cells (InvivoGen #293-htlr4md2cd14) were cultivated in Dulbecco’s modified Eagle medium supplemented with 10% fetal bovine serum, penicillin (100 units/mL)- streptomycin (100 ug/mL), blasticidin (10 ug/mL) and HygroGold (50 ug/mL, InvivoGen) at 37°C in a 5% CO2 atmosphere.
MAST3 overexpression model
At day 0, twelve aliquots of 1×106 TLR4 293-hTLR4A-MD2-CD14 HEK cells (InvivoGen, San Diego, CA, USA) were diluted in 2ml of complete medium (without antibiotics) and plated in individual wells of two 6-well plates. At day 1, cells were transiently transfected using 6ug of pCMV-XL6 vector containing MAST3 cDNA (SC316595 Origene, Rockville, MD, USA) or 6ug of pCMV-AC-GFP vector (PS100010 Origene, Rockville, MD, USA) and 15 uL of lipofectamin 2000 (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. Cells were collected 24h later (at day 2) and frozen. For the time course assay, cells were similarly transfected, and collected and frozen at 0h, 8h, 12h and 24h after transfection and used for protein or RNA extraction.
MAST3 knockdown model
Small hairpin RNA (shRNA) was introduced into THP1 cells by spinoculation with Mission lentiviral particles (Sigma, St-Louis, MO, USA) according to manufacturer’s instructions. At day 1, aliquots of 50,000 cells were diluted in 100 uL of complete medium. Cells were transduced with lentiviral transduction particules carrying a vector with shRNA targeting MAST3 and a puromycin resistance gene (clone ID TRCCN0000200003 from Sigma, St-Louis, MO, USA, CCGGCGAGCCTTTCTGCCGACACAGCTCGAGCTGTGTCGGCAGAAAGGCTCG TTTTTTG) or control transduction particules carrying a vector with non-target shRNA and a puromycin resistance gene(SHC002V Sigma, St-Louis, MO, USA) at MOI =10. Three independent aliquots of cells were transduced with MAST3 shRNA and three others with non-target control shRNA. Cells with virus containing medium were centrifuged at 800 g for 30 minutes at 32 °C. Virus containing medium was aspirated and cell pellets were resuspended in 100 ul of complete media by gently pipetting the pellets up and down, and each resuspended pellet was transferred to its own well in a 96-well tissue culture plate. The plate was returned to the tissue culture incubator 37°C in a 5% CO2 atmosphere. At day 4, puromycin (final concentration 0.5mg/mL) was added to the medium. At day 12, all cells in the non-transduced control wells were dead. Cells were further propagated in medium containing puromycin and pools of transduced THP1 clones were used for differentiation. The 6 pools of transduced THP1 cells (3 MAST3 knockdowns, 3 controls) were each separated in aliquots. One was kept untreated, one was treated with phorbol myristate acetate (PMA; P1585 Sigma, St-Louis, MO, USA) (10 ng/mL) for 48h and the other ones were treated with PMA (10 ng/mL) for 48h and with lipopolysaccharide (LPS; L2654 Sigma, St-Louis, MO, USA) (10 ng/mL) for 3h, 6h or 24h. Cells were then collected for proteins and RNA extraction.
RNA extraction, quantification and reverse-transcription (RT) in cell models
RNA was extracted using the RNeasy Plus Mini Kit (Qiagen, Valencia, CA, USA) according to the manufacturer’s instructions. RNA was quantified using a Bioanalyzer 2100 (Agilent, Santa Clara, CA, USA). All samples had RIN above 9.5. RNA samples (2 ug per rnx) were reverse-transcribed (random hexamers) using the High Capacity cDNA Reverse Transcription kit (ABI, Carlsbad, CA, USA).
Proteins and Western Blots
Proteins were extracted from cell lysate using RIPA Buffer (899100 Thermo, Waltham, MA, USA) according to the manufacturer’s instructions. Proteins were quantified using Pierce BCA Protein Assay Kit (23225 Thermo, Waltham, MA, USA) according to the manufacturer’s instructions. The absorbance was read on a Synergy 2 plate reader (Biotek, Winoosky, VT, USA). Western blots were performed according to standard methods, antibodies used for immunoblotting were anti-MAST3 (ab64952, Abcam, Cambridge, MA, USA), anti-actin (ab3280, Abcam, Cambridge, MA, USA), and goat anti-rabbit-HRP (CLCC43007, Cederlane, Burlington, NC, USA).
qPCR
Gene expression was monitored with 10 ng of total cDNA and SybrGreen (Invitrogen Carlsbad, CA, USA) based qPCR using the Mx3005P Real-Time PCR system (Agilent, Santa Clara, CA, USA). Cycling conditions were set at 95°C for 10 minutes, then 40 cycles of 95°C for 30 seconds and 60°C for 1 minute and 1 cycle of 55°C for 30 seconds and 95°C for 30 seconds. Samples were run in duplicates. Gene expression levels were normalized to the HPRT1 gene expression levels. Human gene specific primer sequences used were: MAST3 forward (5′GCAGCGAAGTGGACTATGG3′), MAST3 reverse (5′GATGGTATTCAGGAGAGATGGG3′), HPRT1 forward (5′ TGGCGTCGTGATTAGTGATG3′), HPRT1 reverse (5′CAGAGGGCTACAATGTGATGG3′) (Other sequences in Supplementary Table 4).
MAST3 model of overexpression: Whole-genome gene expression analysis
Twelve RNA samples from TLR4 293-hTLR4A-MD2-CD14 HEK cells (6 MAST3 overexpressed samples and 6 controls) were sent to Genome Quebec Innovation Center (Montreal, QC, Canada) (GQ) to be analyzed using the HumanHT-12 v4 Expression BeadChip (Illumina, San Diego, CA, USA). Preliminary data was analyzed at GQ using the Genome Studio Software (Illumina, San Diego, CA, USA). We performed further quality control assessment and data normalization using the Lumi library included in the Bioconductor package (7, 8). First, data were adjusted by log2 transformation and global mean normalization. Then, differentially expressed genes between cells overexpressing the MAST3 gene and mock-transfected cells were determined using a one-tailed Welch’s t-test calculated using the MeV software (9). P values were corrected for multiple testing using the adjusted Bonferroni correction. Genes with significant expression change (Padjusted<0.05) were further studied.
A query of the Database for Annotation, Visualization and Integrated Discovery (DAVID) (10) was used to identify functional pathways based on Gene Ontology biological processes.
NF-κB activity Assays
NF-κB activity was monitored using ELISA strip plates from the Transcription Factor kit for NF-κB p65 (89858 Thermo, Waltham, MA, USA) according to the manufacturer’s instructions. For THP1 cells and TLR4 HEK cells, 6 and 20 ug of total protein lysate was used per well, respectively. For each of the three biological replicates at each point, luminescence from two independent wells was read on a Synergy 2 plate reader (Biotek, Winoosky, VT, USA).
Whole-Genome Expression in Patient Biopsies
Biopsies were collected from patients suffering from ulcerative colitis (UC) and healthy controls at Royal Victoria and Maisonneuve-Rosemont Hospitals in Montreal (QC, Canada). Patients had to be older than 18 year old and diagnosed with UC for at least 5 years. Patients under biological treatment (i.e.: Infliximab) were excluded from the study. Biopsies were collected during routine exams (patients) and during cancer screening exams (controls). Written informed consent was obtained from all participants and ethics approval was granted in each of the participating institutions. For patients, biopsies was taken from inflamed (on the margin of ulcerations) and noninflamed region. Biopsies were taken from the rectum (controls and most patients) and colon (some patients) and stored in RNAlater (Qiagen, Valencia, CA, USA) until RNA extraction. RNA was extracted in the following 24h using the RNeasy Plus Mini Kit (Qiagen, Valencia, CA, USA) according to the manufacturer instructions. RNA was quantified using a Bioanalyzer 2100 from Agilent. RIN was above 8. One ug of RNA was sent to GQ and whole-genome expression was evaluated using the Human Genome U133 Plus 2.0 Array (Affymetrix, Santa Clara, CA, USA).
Statistical analyses for the enrichment analyses
Expression data from the biopsies was normalized using quantile normalization with the RMA algorithm within each analyzed pair of classes (control+ non-inflamed and non-inflamed+inflamed) (11, 12). Differential expression between each class was tested using Welch T tests. Data structure (inflamed and non-inflamed samples that came from the same individual and batches of Affymetrix (Santa Clara, CA, USA) chips hybridized on 2 separate occasions) was corrected using permutations (1000 permutations). Enrichment in the MAST3-regulated gene set was tested using Fisher’s exact test.
For assessment of the differential expression of individual genes, P values were further adjusted for FWER using a Bonferonni-type correction that incorporates correlation within the data.
Results
Overexpression model identifies a MAST3-regulated gene set involved in immune response
In order to gain insight into the function of the IBD susceptibility gene MAST3, we overexpressed MAST3 cDNA in HEK293 cells (chosen because these cells are easily and efficiently transfected) and analysed the resulting MAST3 transcriptional signature using genome-wide cDNA microarray. The endogenous expression of MAST3 is relatively low in these cells, and was measured by reverse transcription quantitative polymerase chain reaction (RT qPCR) to be several orders of magnitude lower than in cells overexpressing MAST3. MAST3 protein levels at 24h post-transfection were increased by 60-fold in these cells (Supplementary Figure 1). Unsupervised hierarchical clustering of expression data clearly discriminates between cells overexpressing the MAST3 gene and mock-transfected cell control (data not shown). In total, 195 probes showed significant (Pcorrected<0.05) differential expression (Supplementary Table 1). This represents 0.4% of all interrogated probes. The differentially expressed probes represent 146 different validated protein coding genes, 24 genes encoding small nuclear RNAs and 15 encoding hypothetical proteins. Of the 185 differentially expressed genes, 113 showed higher expression and 72 showed lower expression in cells overexpressing the MAST3 gene compared to mock-transfected cells. Underexpressed genes (in cells overexpressing MAST3) consisted mostly of small nuclear RNAs and hypothetical proteins, so we decided to focus on the overexpressed genes. We were most interested in the top 28 genes with 2-fold or greater increases in expression (Table 1). We validated the microarray expression data by testing the expression of 5 genes randomly chosen among the top 28 using RT qPCR (Supplementary Figure 2).
Table 1.
Differentially expressed genes (Fold increase>2) in the MAST3 overexpression model
| Gene | chr | P value | Fold increase | Function |
|---|---|---|---|---|
| MAST3 | 19 | 3.48E-04 | 42.00 | |
| CCL20 | 2 | 1.31E-10 | 38.13 | C, NP |
| IL8 | 4 | 6.09E-06 | 13.89 | C, NP |
| CD44 | 11 | 1.13E-02 | 6.41 | AM, NP |
| IFIT2 | 10 | 1.20E-03 | 6.23 | I |
| IER3 | 6 | 4.05E-04 | 5.28 | RN, NP |
| NFKBIA | 14 | 4.41E-02 | 4.98 | RN, NP |
| IL32 | 16 | 8.70E-04 | 4.60 | C, NP |
| ISG15 | 1 | 6.82E-04 | 3.95 | I |
| TNFAIP3 | 6 | 5.81E-04 | 3.83 | RN, NP |
| NFKBIZ | 3 | 8.69E-05 | 3.74 | RN, NP |
| IFIT1 | 10 | 8.40E-06 | 3.66 | I |
| CD70 | 19 | 5.98E-04 | 3.55 | C, RN |
| TNC | 9 | 3.97E-02 | 3.49 | AM, NP |
| SLC25A24 | 1 | 3.82E-05 | 3.07 | |
| CXCL2 | 4 | 5.62E-03 | 2.92 | C, NP |
| TNFRSF9 | 1 | 1.40E-03 | 2.81 | RN, NP |
| IFIT3 | 10 | 1.78E-04 | 2.80 | I |
| SGPP2 | 2 | 3.89E-03 | 2.61 | NP |
| TNF | 6 | 2.77E-04 | 2.55 | C, NP |
| GAL | 11 | 2.53E-04 | 2.44 | NP |
| CD83 | 6 | 5.69E-07 | 2.41 | NP |
| TNFRSF12A | 16 | 3.84E-03 | 2.35 | RN, AM |
| PMAIP1 | 18 | 1.17E-03 | 2.28 | I |
| NFKB1 | 4 | 2.87E-04 | 2.26 | RN, NP |
| LY96 | 8 | 1.27E-02 | 2.22 | RN |
| CDK6 | 7 | 5.01E-04 | 2.18 | NP |
| CALCB | 11 | 4.38E-03 | 2.17 | |
| TMOD1 | 9 | 7.18E-05 | 2.00 | AM |
AM: cell adhesion and/or migration. chr: chromosome. C: pro-inflammatory cytokine. I: induced by interferons. NP: regulated by NF-κB. RN: regulator of NF-κB.
This set of 28 genes (Table 1) whose expression appears to be regulated by MAST3 shows a significant enrichment (P<0.01) of functional involvement in defense response (10-fold), inflammatory response (16-fold), response to wounding (11-fold), regulation of apoptosis (7-fold), and regulation of programmed cell death (7-fold) (Supplementary Table 2). The importance of this group of genes in the immune response can be further recognized in their known function based on the literature: they encode pro-inflammatory cytokines, regulators of the activity of NF-κB, proteins involved in adhesion and cell migration, and proteins induced by interferon in response to pathogen invasion or tumor progression. Additionally, 61% of these genes are regulated by NF-κB via functionally validated binding sites in their promoter (Figure 1 and Table 1) suggesting that MAST3 may act through this pathway.
Figure 1. MAST3-regulated genes are involved in critical immune functions.
Genes in the MAST3-regulated gene set (genes upregulated by 2-fold or more in MAST3 overexpressed cells) were classified according to their function recovered from literature searches. The functions of the genes mainly cluster around the activity of NF-κB. Sixty-one percent of the 28 genes have experimentally confirmed NF-κB binding site in their promoter and are regulated by NF-κB. Thirty-two percent regulates the activity of NF-κB. Twenty-one percent are pro-inflammatory cytokines. Other functions include cell adhesion and migration and interferon mediated defense against pathogen invasion and tumor progression. Two genes out of the 28, CALCB and SLC25A24 do not seem to be directly involved in immune functions and were not included in the diagram.
The median fold-increase in gene expression in the MAST3-regulated gene set (top 28 genes) in MAST3 overexpressing cells compared to mock-transfected cells is 3. However, two genes are overexpressed by greater than 10-fold, CCL20 (P=1.31×10−10, 38-fold) and IL8 (P=6.09×10−6, 14-fold), suggesting that MAST3 levels have a greater influence on the expression of these genes than the other upregulated genes. The CCL20 chemokine is the only known ligand of the CCR6 receptor and CCR6 signaling is involved in the recruitment of dendritic cells, and other antigen presenting cells as well as CD4+T cells to the sites of epithelial inflammation(13). The chemokine IL-8 is produced predominantly in the lamina propria of the colon and its level correlates with mucosal inflammation (14).
Knockdown of MAST3 expression in immune cells confirms gene expression profile
In order, to further study the effect of MAST3, we created a MAST3 knockdown (KD). We were interested in confirming that the observed gene expression patterns from our overexpression model were reversed in the knockdown. Additionally, we wanted to study immune gene expression in an immune cell context so we selected monocyte-like THP1 cells. Our KD was created by transducing THP1 cells with MAST3 small hairpin RNA (shRNA).
Interestingly, the KD has different effects on MAST3 levels at different stages of cell differentiation and stimulation (Supplementary Figure 3). The KD has the greatest effect on MAST3 levels in cells differentiated by phorbol 12-myristate 13-acetate (PMA) and stimulated by lipopolysaccharide (LPS) for 24h (a decrease of 75% of mRNA levels and of 92% of MAST3 protein levels). In order, to maximize the effect of the MAST3 KD on endogenous gene expression, we studied THP1 cells treated with PMA and LPS for 24h. We tested a subset of the MAST3-regulated gene set. Of the 9 genes tested, 7 are significantly underexpressed (P<0.05) in our KD model compared to control and 2 show no significant difference (Table 2). Overall, this downregulation of gene expression in MAST3 KD cells is consistent with the upregulation observed in the cells that overexpress the MAST3 gene. Notably, IL8 is highly expressed in PMA and LPS treated THP1 (4800-fold increase from unstimulated THP1 stage) and its expression level is reduced to 55% of the control expression in the KD (P=0.01).
Table 2.
Effect of the MAST3 stable knockdown (KD) on expression of genes in stimulated THP1 cells confirms results from the overexpression model.
| Gene | Effect of MAST3 KD (% of expression of non target KD control)* | P |
|---|---|---|
| IL8 | 55% | 0.01 |
| IER3 | 41% | 0.02 |
| CD70 | 76% | 0.03 |
| CCL20 | no detected effect | 0.80 |
| ISG15 | 75% | 0.02 |
| LY96 | 40% | 0.01 |
| NFKBIA | 92% | 0.02 |
| TNF | no detected effect | 0.47 |
| TNFAIP3 | 75% | 0.02 |
-: no significant effect of the KD of MAST3 on expression.
: THP1 cells were differentiated using PMA and treated with LPS for 3h (IER3, TNF) or 24h (other genes).
Our expression study in cells overexpressing the MAST3 gene highlighted a group of upregulated genes that are regulated by transcription factor NF-κB. We were interested in exploring the extent of the effect of a MAST3 KD on the NF-κB dependent endogenous gene expression but were somewhat limited by the HEK293 cell model (fibroblast like human embryonic kidney cells). The THP1 cell model allowed us to study the expression of additional NF-κB regulated genes not necessarily expressed in HEK293 cells. We tested the expression of several interleukins, TGFβ1 and members of the NF-κB pathway (Table 3). As expected, based on our previous results, the KD of MAST3 decreases the expression of a majority (4/6) of the pro-inflammatory genes tested. The KD also upregulates the expression of anti-inflammatory cytokine gene IL10 to 210% of non target KD control (P=0.003). These additional results in our THP1 MAST3 KD model support the results from the overexpression model and suggest a prominent pro-inflammatory role for MAST3.
Table 3.
MAST3 stable knockdown (KD) influences the expression of immune genes in stimulated THP1 cells.
| Gene | Effect of MAST3 KD (% of expression of non target KD control)* | P |
|---|---|---|
| IL10 | 210% | 0.003 |
| IL12B | 200% | 0.03 |
| IL18 | 220% | 0.04 |
| IL1a | 60% | 0.01 |
| IL1b | 70% | 0.03 |
| TGFB1 | 65% | 0.03 |
| TRAF6 | 65% | 0.002 |
: THP1 cells were differentiated using PMA and treated with LPS for 6h (IL10, IL1b, TGFB1) or 24h (IL12B, IL18, IL1A, TRAF6).
NF-κB activity correlates to MAST3-regulated gene expression
We were interested in studying the relationship between NF-κB activity and the MAST3-regulated gene set expression. We have previously shown that a KD of MAST3, in LPS stimulated HEK293 cells that constitutively express the TLR4 receptor, decreases the activity of transcription factor NF-κB (4). Thus, we first wished to confirm these findings in our two new cell models of overexpression in HEK cells and KD in THP1 cells. Then, we used the IL8 and CCL20 genes as proxies for the MAST3-regulated gene set to study the temporal correlation between NF-κB activity and MAST3 gene set expression.
We tested the activity of NF-κB at different times post-transfection MAST3 (Figure 2A). We observed a significant increase in activity specific to MAST3 beginning 12 hours post-transfection. The activity of NF-κB in cells overexpressing the MAST3 gene, at 12 hours and 24 hours post-transfection is 159% (P=0.01) and 354% (P=0.02) that of mock-transfected cells, respectively. This increase in NF-κB activity is correlated with significant MAST3 mRNA (data not shown) and protein level increases (Supplementary Figure 1).
Figure 2. MAST3 regulates gene expression through NF-κB.
A. The overexpression of MAST3 increases the activity of NF-κB. The graph shows the activity of NF-κB tested in HEK293 cells that overexpress the MAST3 gene at different times post-transfection (MAST3 over). The increase in activity is significantly detected after 12h (*P=0.01). B. The knockdown of MAST3 decreases the activity of NF-κB. The graph shows the activity of NF-κB tested in THP1 cells at different stages of differentiation and stimulation following different treatment. The decrease in activity is significantly detected in phorbol 12-myristate 13-acetate (PMA) differentiated cells after 6h of lipopolysaccharide (LPS) stimulation (*P=0.02). C. and D. Expression of MAST3-regulated genes IL8 (C) and CCL20 (D) post-transfection in cells overexpressing the MAST3 gene. Expression is normalized to HPRT1 expression. Error bar represent standard deviation of 3 independent biological replicates. KD MAST3: knockdown MAST3, KD NT: non target knockdown control.
The activity of NF-κB in our THP1 knockdown confirms the results of our previous study in HEK293 cells. A significant decrease in NF-κB activity (70% of control, P=0.02) is detected in THP1 cells differentiated with PMA and treated with LPS for 6h (Figure 2B). Importantly, these results correlate with decreased MAST3 mRNA (67.2% decrease) and protein (86% decrease) levels (Supplementary Figure 3).
The expression levels of MAST3-regulated genes CCL20 and IL8 genes are induced early and significant differences between MAST3 overexpressed cells and control can be observed as soon as 8h post-transfection (Figure 2C and 2D). Yet, the most significant fold differences in expression between cells overexpressing the MAST3 gene and mock-transfected cells are observed at 12 and 24h (10 and 32-fold increases respectively), and thus correlate temporally with the increases in NF-κB activity.
MAST3-regulated gene set expression is enriched in inflamed tissues of UC patients
To establish the importance of the MAST3-regulated gene set in the clinical presentation of the disease, we tested its expression in human intestinal mucosa tissues from patient biopsies. In order to maximize our statistical power to detect differential expression by limiting noise, we defined the MAST3-regulated gene set as the group of genes that had been overexpressed by 3-fold or greater in our cell model of MAST3 overexpression (Table 1, top 14 genes). We were mostly interested in answering two questions: 1. Are there differences in gene expression between intestinal tissue from healthy people and normal looking intestinal tissue from patients affected with UC? 2. Are there differences between normal looking intestinal tissues and inflamed tissue from UC patients? In order to answer these questions, we analyzed the differential gene expression profiles between biopsies of controls (BC) and patients’ non-inflamed (BPN) and inflamed (BPI) regions of the colon and rectum. Specifically, we were interested in discovering whether there was an enrichment of our MAST3-regulated genes as part of the inflammatory process in IBD patients.
There were no detectable differences between BC (n=16) and BPN (n=15) (Table 4 and Supplementary Table 3). In contrast, we observed significant differences between BPN (n=19) and BPI (n=15). Of all the probes of the array, 34.5% (8498 out of 24607) showed differential expression (P<0.05). In our gene set, 71.4% of genes (10 out of 14) showed differential expression (P<0.05), which represent a significant 2-fold enrichment (P=0.005) (Table 4). After correction for multiple testing, we examined the expression patterns for individual genes. In general, expression of genes from the MAST3-regulated gene set was increased in the BPI compared to BPN. However, this trend did not extend to MAST3 expression for which the difference between BPI and BPN did not reach statistical significance. For 5 genes the increase in expression was significant, CD44 (P=0.001), TNC (P=0.017), ISG15 (P=0.02), NFKBIZ (P=0.02) and IL8 (P=0.031) (Table 5).
Table 4.
MAST3-regulated gene set enrichment tests in patient biopsies
| BC vs BPN | BPN vs BPI | |||
|---|---|---|---|---|
| MAST3-regulated gene set | All genes* on microarray | MAST3-regulated gene set | All genes* on microarray | |
| P*<0.05 | 3 | 7045 | 10 | 8498 |
| Total | 14 | 24066 | 14 | 24607 |
| % | 21.4 | 29.2 | 71.4 | 34.5 |
| Enrichment P | 0.825 | 0.005 | ||
Probe sets on the Affymetrix chip are used in lieu of genes. Some genes are represented by more than one probe set.
P*= the empirical P after 1000 permutations.
BC=biopsies from healthy control. BPN=biopsies from patients taken from non-inflamed region. BPI=biopsies from patients taken from inflamed region.
Table 5.
Comparison of the expression of the MAST3-regulated genes between BPN and BPI
| Exp. BPN | Exp. BPI | P* | |
|---|---|---|---|
| CD44 | 8.01 | 8.98 | 0.001 |
| TNC | 6.90 | 8.31 | 0.017 |
| ISG15 | 9.02 | 9.96 | 0.02 |
| NFKBIZ | 10.56 | 11.74 | 0.02 |
| IL8 | 5.89 | 7.66 | 0.027 |
| NFKBIA | 10.57 | 11.11 | 0.061 |
| TNFAIP3 | 9.47 | 9.96 | 0.07 |
| IER3 | 10.22 | 10.89 | 0.137 |
| CCL20 | 10.33 | 11.57 | 0.27 |
| IFIT2 | 8.12 | 8.56 | 0.505 |
| SLC25A24 | 11.61 | 11.44 | 0.609 |
| CD70 | 6.72 | 6.82 | 1 |
| IFIT1 | 7.81 | 7.80 | 1 |
| IL32 | 9.05 | 9.09 | 1 |
Exp.= mean expression. BPN=biopsies from patients taken from non-inflamed region. BPI=biopsies from patients taken from inflamed region. P*= Empirical P corrected for multiple testing.
Discussion
Genome-wide and targeted association studies have already identified numerous modest effect genes and estimations of cumulative heritability predict that they will continue to identify new genes for some time in the future. In order to use these genetic findings to benefit patients, functional studies are essential. The challenge is to explain the individual role of each of these genes in the pathogenesis of complex diseases.
Following the identification of MAST3 as an IBD gene (4), we were interested in studying its functional effect in the context of the immune system. Since MAST3 had not been studied extensively, we decided to use a non-targeted approach to examine the effect of perturbing the expression of MAST3 (overexpression) in a well-characterized cell system. The hypothesis being that overexpression of MAST3 will have an influence on expression of genes in related biological pathways. We report here our results that highlight a group of MAST3-regulated genes heavily involved in immune responses. These genes are functionally linked to NF-κB as a majority of them are regulated and/or regulators of the NF-κB pathway. We show that the activity of NF-κB is correlated temporally with MAST3-regulated gene expression implying that MAST3 acts on the NF-κB pathway to trigger immune reactions through changes in gene expression. The MAST3 gene set is upregulated in colonic biopsies of inflamed regions from patients with UC as compared to non-inflamed regions. This suggests that MAST3 contributes to the dysregulation of the immune response to microbiota underlying IBD through modulation of gene expression.
To examine the MAST3-regulated gene set in a clinically relevant context, we studied genome-wide expression in colonic biopsies from control and UC patients. We observed a significant increase in the expression of the MAST3-regulated genes in inflamed vs non-inflamed tissues. Since 20% of the genes in this set are encoding proinflammatory cytokines, including the key chemokines CCL20 and IL8, this emphasizes the role of the inflammatory process in the initiation and/or maintenance of the disease state. Interestingly, IL8 is a chemokine that is important in innate immunity and highly regulated by MAST3. IL8 expression has been shown to be increased in both UC and CD patients and IL8 mRNA and protein levels correlate to inflammation severity (14). Mutations in NOD2, the first gene associated to CD (15), were shown to regulate the expression of IL8 (16).
In addition, a few of the MAST3-regulated genes are involved in the regulation of the NF-κB pathway which seems to also implicate MAST3 in a feedback mechanism to prevent uncontrolled inflammation. NFKBIA is among the NF-κB regulators within this gene set. NFKBIA encodes protein IκBα which sequesters the NF-κB complex in the cytoplasm therefore inhibiting its nuclear localization and transcription factor functions (17). Another of these regulators is TNFAIP3 encoding the protein A20 which acts on the NF-κB pathway through interaction with TRAF6 (18). Overexpression of A20 has been shown to block TLR4 activation of NF-κB (19). MD-2, encoded by gene LY96, is the co-receptor to TLR4. The binding of LPS to TLR4 and subsequent NF-κB activation is dependent on the formation of the TLR4-MD-2 complex(20). IER3 has been shown to interact with NF-κB and inhibits its activation, thus interfering with downstream gene expression (21).
Several IBD genes influence immune responses through modulation of the NF-κB pathway. Notably, a recent study into the effect of the overexpression of IBD associated variant NOD2L1007finsC on global gene expression underscores some NF-κB regulated genes found in the MAST3-regulated gene set(22). NOD2L1007finsC was found to upregulate the expression of pro-inflammatory cytokine gene IL8 and CXCL2 and of NF-κB regulator genes NFKBIA, IER3 and TNFAIP3. Our combined findings give weight to the hypothesis that the dysregulation of a limited number of biological pathways is involved in the pathogenesis of IBD. NOD2 is the first identified and among the most systematically replicated CD genes. As opposed to MAST3, it is also among the most studied functionally. NOD2 has been involved in the regulation of immune gene expression and epithelial barrier integrity (23). Here we show that another IBD risk factor, MAST3, is modulating the expression of genes and regulating the activity of NF-κB in a similar way and thus that its variants could translate into similar cellular consequences. It is possible that there is an inflammation gene set that is regulated by several IBD genes such as NOD2 and MAST3. The upregulation of the expression of this inflammation gene set in mucosal gut tissues could trigger the conditions for the development of IBD. In this context of interconnectivity between IBD genes, functional studies that aim to decipher disease-causing pathways are becoming essential companions to genetic studies in the quest to improve our understanding of the molecular mechanisms underlying IBD.
Among MAST3-regulated genes, a few - TNF, TNFRSF9 and CCL20/CCR6 - are located in IBD associated loci which also underlines the interconnectivity of the IBD genes. The TNF gene is on chromosome 6p21 in the major histocompatibility complex (MHC) region and has been associated to CD following a metanalysis of genome-wide associated studies (24). The TNF gene encodes tumor necrosis factor α, a pro-inflammatory cytokine that induces cell proliferation and differentiation and that is found in increased amounts in patients suffering from IBD (25, 26). Biological therapies that target TNF have been used for several years to treat refractory IBD. Notably, the anti-TNF humanized antibody Infliximab has shown efficacy in reducing intestinal inflammation and maintaining remission, preventing fistulae, reducing the need for steroids treatment, reducing hospitalization and need for surgery, and improving quality of life (27–30). The TNFRSF9 gene is located in a locus recently associated to UC (2). TNFRSF9 encodes a cell surface transmembrane protein that is involved in cell differentiation and activation (31, 32). TNFRSF9 has been shown to activate NF-κB in a TRAF-dependent manner (33). CCL20, the most upregulated gene in our overexpression model, has not been associated to IBD. However, it encodes the sole ligand of receptor CCR6 which is encoded by a CD associated gene (24). CCL20 and CCR6 are involved in the regulation of mucosal immunity through the recruitment of antigen presenting cells to inflammation sites and homing of dendritic cells and CD4+ T cells to the lymphoid tissue of the gut (13). There is much interest in CCR6 because it is the only receptor expressed on every TH17 cell subsets (34) and TH17 driven immune responses are hypothesized to be responsible for CD pathogenesis(35).
Our genome-wide expression profiling is a first step into the investigation of the role of MAST3 in the immune response. Several signal transducers acting between TLR4 and NF-κB are phosphorylated on serine and/or threonine. It is possible that the MAST3 kinase modulates the activity of NF-κB through the phosphorylation and activation of one of these molecules. Also, MAST3 might physically interact and modulate the pathway through stabilization and chaperoning of a target. MAST2, a protein from the same family, has been shown to regulate NF-κB through the formation of a complex with TRAF6 (6). Additional functional studies will be needed to pinpoint the precise target of MAST3 in the NF-κB pathway. In the meantime, this study reinforces the importance of this pathway in IBD and identifies an IBD risk factor, MAST3, as a regulator of NF-κB activity and immune gene expression.
Supplementary Material
Supplementary Figure 1. Effect of MAST3 overexpression on MAST3 protein levels at different times post-transfection. Protein levels are relative to actin. Error bars represent the standard deviation between 3 independent biological replicates.
Supplementary Figure 2. RT qPCR validation of expression chip results. The same RNA samples used on the chip were tested by qPCR. All gene tested were validated. The fold increase in expression is indicated on each bar chart. Expression is normalized to HPRT1 expression. Error bars represent the standard deviation between 6 independent biological replicates.
Supplementary Figure 3. Effect of a knockdown of MAST3 on MAST3 protein levels. Protein levels are relative to actin. There were no detectable MAST3 proteins in THP1 non stimulated cells. Error bars represent the standard deviation between 3 independent biological replicates. NT KD Control: non target knockdown control, KD MAST3: knockdown MAST3
Supplementary Table 1. All gene significantly differentially expressed (P<0.05) in cells overexpressing the MAST3 gene.
Supplementary Table 2. Major Gene Ontology biological processes involving genes upregulated in our MAST3 overexpression model.
Supplementary Table 3. Comparison of the expression of gene from the MAST3-regulated gene set between biopsies from healthy control (BC) and biopsies from patients taken from non-inflamed region (BPN).
Supplementary Table 4. Sequences of the RT qPCR primers.
Acknowledgments
The authors would like to thank Guillaume Lettre and Marcia Budarf for their careful review of the manuscript. This work was funded by a grant from the U.S. National Institute of Diabetes and Digestive and Kidney Diseases (DK062432) awarded to JDR. CL holds a doctoral training award from the Fond de Recherche en Santé du Québec.
References
- 1.Khor B, Gardet A, Xavier RJ. Genetics and pathogenesis of inflammatory bowel disease. Nature. 2011;474:307–317. doi: 10.1038/nature10209. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Anderson CA, Boucher G, Lees CW, et al. Meta-analysis identifies 29 additional ulcerative colitis risk loci, increasing the number of confirmed associations to 47. Nature genetics. 2011;43:246–252. doi: 10.1038/ng.764. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Franke A, McGovern DP, Barrett JC, et al. Genome-wide meta-analysis increases to 71 the number of confirmed Crohn’s disease susceptibility loci. Nature genetics. 2011;42:1118–1125. doi: 10.1038/ng.717. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Labbe C, Goyette P, Lefebvre C, et al. MAST3: a novel IBD risk factor that modulates TLR4 signaling. Genes Immun. 2008;9:602–612. doi: 10.1038/gene.2008.57. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Garland P, Quraishe S, French P, et al. Expression of the MAST family of serine/threonine kinases. Brain Res. 2007 doi: 10.1016/j.brainres.2007.12.027. [DOI] [PubMed] [Google Scholar]
- 6.Xiong H, Li H, Chen Y, et al. Interaction of TRAF6 with MAST205 regulates NF-kappaB activation and MAST205 stability. The Journal of biological chemistry. 2004;279:43675–43683. doi: 10.1074/jbc.M404328200. [DOI] [PubMed] [Google Scholar]
- 7.Gentleman RC, Carey VJ, Bates DM, et al. Bioconductor: open software development for computational biology and bioinformatics. Genome Biol. 2004;5:R80. doi: 10.1186/gb-2004-5-10-r80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Du P, Kibbe WA, Lin SM. lumi: a pipeline for processing Illumina microarray. Bioinformatics. 2008;24:1547–1548. doi: 10.1093/bioinformatics/btn224. [DOI] [PubMed] [Google Scholar]
- 9.Saeed AI, Bhagabati NK, Braisted JC, et al. TM4 microarray software suite. Methods Enzymol. 2006;411:134–193. doi: 10.1016/S0076-6879(06)11009-5. [DOI] [PubMed] [Google Scholar]
- 10.Huang da W, Sherman BT, Lempicki RA. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc. 2009;4:44–57. doi: 10.1038/nprot.2008.211. [DOI] [PubMed] [Google Scholar]
- 11.Bolstad BM, Irizarry RA, Astrand M, et al. A comparison of normalization methods for high density oligonucleotide array data based on variance and bias. Bioinformatics. 2003;19:185–193. doi: 10.1093/bioinformatics/19.2.185. [DOI] [PubMed] [Google Scholar]
- 12.Gautier L, Cope L, Bolstad BM, et al. affy--analysis of Affymetrix GeneChip data at the probe level. Bioinformatics. 2004;20:307–315. doi: 10.1093/bioinformatics/btg405. [DOI] [PubMed] [Google Scholar]
- 13.Ito T, Carson WFt, Cavassani KA, et al. CCR6 as a mediator of immunity in the lung and gut. Exp Cell Res. 2011;317:613–619. doi: 10.1016/j.yexcr.2010.12.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Daig R, Andus T, Aschenbrenner E, et al. Increased interleukin 8 expression in the colon mucosa of patients with inflammatory bowel disease. Gut. 1996;38:216–222. doi: 10.1136/gut.38.2.216. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Hugot JP, Chamaillard M, Zouali H, et al. Association of NOD2 leucine-rich repeat variants with susceptibility to Crohn’s disease. Nature. 2001;411:599–603. doi: 10.1038/35079107. [DOI] [PubMed] [Google Scholar]
- 16.Li J, Moran T, Swanson E, et al. Regulation of IL-8 and IL-1beta expression in Crohn’s disease associated NOD2/CARD15 mutations. Hum Mol Genet. 2004;13:1715–1725. doi: 10.1093/hmg/ddh182. [DOI] [PubMed] [Google Scholar]
- 17.Ferreiro DU, Komives EA. Molecular mechanisms of system control of NF-kappaB signaling by IkappaBalpha. Biochemistry. 2010;49:1560–1567. doi: 10.1021/bi901948j. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Heyninck K, Beyaert R. The cytokine-inducible zinc finger protein A20 inhibits IL-1-induced NF-kappaB activation at the level of TRAF6. FEBS letters. 1999;442:147–150. doi: 10.1016/s0014-5793(98)01645-7. [DOI] [PubMed] [Google Scholar]
- 19.O’Reilly SM, Moynagh PN. Regulation of Toll-like receptor 4 signalling by A20 zinc finger protein. Biochem Biophys Res Commun. 2003;303:586–593. doi: 10.1016/s0006-291x(03)00389-9. [DOI] [PubMed] [Google Scholar]
- 20.Park BS, Song DH, Kim HM, et al. The structural basis of lipopolysaccharide recognition by the TLR4-MD-2 complex. Nature. 2009;458:1191–1195. doi: 10.1038/nature07830. [DOI] [PubMed] [Google Scholar]
- 21.Arlt A, Rosenstiel P, Kruse ML, et al. IEX-1 directly interferes with RelA/p65 dependent transactivation and regulation of apoptosis. Biochimica et biophysica acta. 2008;1783:941–952. doi: 10.1016/j.bbamcr.2007.12.010. [DOI] [PubMed] [Google Scholar]
- 22.Billmann-Born S, Till A, Arlt A, et al. Genome-wide expression profiling identifies an impairment of negative feedback signals in the Crohn’s disease-associated NOD2 variant L1007fsinsC. J Immunol. 2011;186:4027–4038. doi: 10.4049/jimmunol.1000085. [DOI] [PubMed] [Google Scholar]
- 23.Kobayashi KS, Chamaillard M, Ogura Y, et al. Nod2-dependent regulation of innate and adaptive immunity in the intestinal tract. Science. 2005;307:731–734. doi: 10.1126/science.1104911. [DOI] [PubMed] [Google Scholar]
- 24.Barrett JC, Hansoul S, Nicolae DL, et al. Genome-wide association defines more than 30 distinct susceptibility loci for Crohn’s disease. Nature genetics. 2008;40:955–962. doi: 10.1038/NG.175. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Braegger CP, Nicholls S, Murch SH, et al. Tumour necrosis factor alpha in stool as a marker of intestinal inflammation. Lancet. 1992;339:89–91. doi: 10.1016/0140-6736(92)90999-j. [DOI] [PubMed] [Google Scholar]
- 26.Reinecker HC, Steffen M, Witthoeft T, et al. Enhanced secretion of tumour necrosis factor-alpha, IL-6, and IL-1 beta by isolated lamina propria mononuclear cells from patients with ulcerative colitis and Crohn’s disease. Clin Exp Immunol. 1993;94:174–181. doi: 10.1111/j.1365-2249.1993.tb05997.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Hanauer SB, Feagan BG, Lichtenstein GR, et al. Maintenance infliximab for Crohn’s disease: the ACCENT I randomised trial. Lancet. 2002;359:1541–1549. doi: 10.1016/S0140-6736(02)08512-4. [DOI] [PubMed] [Google Scholar]
- 28.Sands BE, Anderson FH, Bernstein CN, et al. Infliximab maintenance therapy for fistulizing Crohn’s disease. N Engl J Med. 2004;350:876–885. doi: 10.1056/NEJMoa030815. [DOI] [PubMed] [Google Scholar]
- 29.Lichtenstein GR, Bala M, Han C, et al. Infliximab improves quality of life in patients with Crohn’s disease. Inflamm Bowel Dis. 2002;8:237–243. doi: 10.1097/00054725-200207000-00001. [DOI] [PubMed] [Google Scholar]
- 30.Targan SR, Hanauer SB, van Deventer SJ, et al. A short-term study of chimeric monoclonal antibody cA2 to tumor necrosis factor alpha for Crohn’s disease. Crohn’s Disease cA2 Study Group. N Engl J Med. 1997;337:1029–1035. doi: 10.1056/NEJM199710093371502. [DOI] [PubMed] [Google Scholar]
- 31.Jiang D, Chen Y, Schwarz H. CD137 induces proliferation of murine hematopoietic progenitor cells and differentiation to macrophages. J Immunol. 2008;181:3923–3932. doi: 10.4049/jimmunol.181.6.3923. [DOI] [PubMed] [Google Scholar]
- 32.Halstead ES, Mueller YM, Altman JD, et al. In vivo stimulation of CD137 broadens primary antiviral CD8+ T cell responses. Nat Immunol. 2002;3:536–541. doi: 10.1038/ni798. [DOI] [PubMed] [Google Scholar]
- 33.Arch RH, Thompson CB. 4–1BB and Ox40 are members of a tumor necrosis factor (TNF)-nerve growth factor receptor subfamily that bind TNF receptor-associated factors and activate nuclear factor kappaB. Mol Cell Biol. 1998;18:558–565. doi: 10.1128/mcb.18.1.558. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Lim HW, Lee J, Hillsamer P, et al. Human Th17 cells share major trafficking receptors with both polarized effector T cells and FOXP3+ regulatory T cells. J Immunol. 2008;180:122–129. doi: 10.4049/jimmunol.180.1.122. [DOI] [PubMed] [Google Scholar]
- 35.Brand S. Crohn’s disease: Th1, Th17 or both? The change of a paradigm: new immunological and genetic insights implicate Th17 cells in the pathogenesis of Crohn’s disease. Gut. 2009;58:1152–1167. doi: 10.1136/gut.2008.163667. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary Figure 1. Effect of MAST3 overexpression on MAST3 protein levels at different times post-transfection. Protein levels are relative to actin. Error bars represent the standard deviation between 3 independent biological replicates.
Supplementary Figure 2. RT qPCR validation of expression chip results. The same RNA samples used on the chip were tested by qPCR. All gene tested were validated. The fold increase in expression is indicated on each bar chart. Expression is normalized to HPRT1 expression. Error bars represent the standard deviation between 6 independent biological replicates.
Supplementary Figure 3. Effect of a knockdown of MAST3 on MAST3 protein levels. Protein levels are relative to actin. There were no detectable MAST3 proteins in THP1 non stimulated cells. Error bars represent the standard deviation between 3 independent biological replicates. NT KD Control: non target knockdown control, KD MAST3: knockdown MAST3
Supplementary Table 1. All gene significantly differentially expressed (P<0.05) in cells overexpressing the MAST3 gene.
Supplementary Table 2. Major Gene Ontology biological processes involving genes upregulated in our MAST3 overexpression model.
Supplementary Table 3. Comparison of the expression of gene from the MAST3-regulated gene set between biopsies from healthy control (BC) and biopsies from patients taken from non-inflamed region (BPN).
Supplementary Table 4. Sequences of the RT qPCR primers.