Differential mucosal expression of Th17-related genes between the inflamed colon and ileum of patients with inflammatory bowel disease

The expression level of the proinflammatory cytokine IL8 was equal in colonic and ileal controls, while the expression level of TNF (TNFα) was slightly higher in ileal control samples than in colonic control samples (P = 0.037) (Figure 1A).

The expression levels of the Th17 effector cytokines IL17A, IL17F, IL21, IL22 and IL26 were comparable between colonic and ileal control samples (Figure 1B); however, the expression of these cytokines in each individual sample was not consistent. In one ileal sample and none of the colonic samples, all five cytokines were expressed. In one colonic sample and four ileal samples, three to four cytokines were expressed, while in eight colonic samples and seven ileal samples, one to two cytokines were expressed. In four colonic samples and three ileal samples, none of the five cytokines were detected.

We next examined the expression of genes involved in the differentiation of Th17 cells. TGFβ (P = 0.0005) and STAT3 (P = 0.007) were expressed at higher levels in ileal control samples than in colonic control samples, while IL6, IL1β and IL23A expression levels were similar in colonic and ileal control samples (Figure 1C).

Expression of genes involved in the recruitment of Th17 cells was also assessed. CCR6 expression was higher in ileal control samples than in colonic control samples (P = 0.0008), while CCL20 expression was similar between colonic and ileal controls (Figure 1D).

Although all samples were taken from endoscopically inflamed mucosa, we quantified inflammation by measuring the expression of the proinflammatory cytokines IL8 and TNFα. The expression level of IL8 has been shown to be associated with the grade of inflammation [35, 36]. IL8 was strongly induced in inflamed colonic samples from UC (P < 0.0001) and CD patients (P = 0.0004) and in inflamed ileal samples from CD patients (P < 0.0001) (Figure 2A). The expression level of IL8 was significantly higher in UC samples than in colonic CD samples (P = 0.017). Expression of TNFα was only significantly increased in UC samples (P = 0.0002); however, a tendency for increase was observed in ileal CD samples (P = 0.052) (Figure 2A). Expression levels of TNF α were significantly higher in UC samples than in colonic CD samples (P = 0.028).

In UC samples, IL17A (P = 0.0002), IL21 (P = 0.0021), IL22 (P = 0.0003) and IL26 (P = 0.0017) were strongly induced, and IL17F (P = 0.046) was weakly induced, while in colonic CD samples, IL17A (P = 0.0084) and IL22 (P = 0.001) were strongly induced, and IL21 (P = 0.047), IL26 (P = 0.048) and IL17F (P = 0.011) were weakly induced (Figure 2B). In ileal CD samples, IL22 (P = 0.0015) was strongly induced, while IL17A (P = 0.013), IL26 (P = 0.029) and IL17F (P = 0.019) were weakly induced, and expression levels of IL21 were similar to those in ileal controls. Furthermore, in the colon, expression levels of IL17A (P = 0.005) and IL21 (P = 0.036) were significantly higher in UC samples than in CD samples. In CD samples, expression of IL21 (P = 0.0097) was significantly higher in the colon than in the ileum.

IL1β (UC, P≤0.0001; CD, P = 0.0009), IL6 (UC, P≤0.0001; CD, P = 0.0007), TGFβ (UC, P = 0.0002; CD, P = 0.0001) and IL23A (UC, P = 0.0006; CD, P = 0.003) were strongly induced in both UC and colonic CD (Figure 2C). STAT3 was only slightly induced in colonic CD (P = 0.015) and UC (P = 0.018). Although a strong induction of the Th17 differentiation factor IL1β (P = 0.0007) and a weak induction of IL6 (P = 0.015) were observed in ileal samples, TGFβ, IL23A and STAT3 were not induced. In the colon, IL23A was expressed at higher levels in UC samples than in CD samples (P = 0.027). In CD samples, TGFβ (P = 0.0008) and IL23A (P = 0.010) were expressed at higher levels in the colon than in the ileum.

Expression levels of both CCR6 (UC, P = 0.0004; CD, P = 0.011) and CCL20 (UC; P = 0.0003, CD; P = 0.0009) were significantly increased in colonic CD and UC samples but not in ileal CD samples (Figure 2D).

Two sets of primers for RORC were developed. The RORC primers detect both the full-length transcript and the shorter T cell-specific isoform, while the RORC1 primers detect only the full-length transcript. A strong correlation was found between RORC1 and RORC expression levels (R = 0.776, P < 0.001). When comparing the RORC1 and RORC levels between healthy controls and UC and CD patients, higher significance levels were achieved using the primer detecting both transcripts. This suggests a stronger role for the T cell-specific RORC2 isoform.

In colonic samples from CD and UC patients, RORC mRNA expression levels were comparable to those in healthy controls, while in ileal CD samples, expression levels of RORC were lower than those in control samples (P = 0.0019) (Figure 3).

A total of 72 macrodissected intestinal tissue samples from 22 healthy controls, 12 UC patients and 26 CD patients were obtained during colonoscopy with a Single-Use Biopsy Forceps Radial Jaw3 (Boston Scientific, El Coyol, Costa Rica) (Table 1). The size of a biopsy specimen was usually between 2-4 mm² with an estimated average weight of 6.4 mg. UC and CD patients were diagnosed based on clinical, endoscopic and histological criteria. Patients characteristics, medication intake and the Montreal classification which gives an overview of the subclassification of IBD patients is shown in Table 1[48]. Mucosal inflammation was defined as the presence of endoscopic signs of disease activity. Samples from healthy controls were taken from the ileum and sigmoid of patients who underwent colonoscopy to screen for cancer or polyps. All biopsies collected during colonoscopy were immediately stored in RNALater (Ambion, Cambridgeshire, UK) at -80°C. The study was in accordance with the guidelines of the Helsinki Declaration (1964 and amended in 1975, 1983, 1989, 1996 and 2000) of the World Medical Association. Informed consent was obtained from all patients, and the protocol was approved by the local Ethics Committee of Ghent University Hospital (EC UZG 2004/242).

Total RNA was extracted from 2-3 pooled mucosal samples using an RNeasy Mini Kit (Qiagen, Westburg BV, The Netherlands) with on-column DNAse treatment (Qiagen). Needle homogenization was performed. The total RNA was quantified using spectrophotometry (Nanodrop; Thermo Scientific, Wilmington, USA) and ranged from 150 ng to 16.8 μg with an average of 5.9 μg total RNA. The quality of the RNA, expressed as RNA quality indicator (RQI), was checked by automated electrophoresis (Experion, Bio-Rad, Hercules, California) and ranged from 7.4 to 10 with an average of 8.6. Starting from 20 ng of total RNA, the WT-Ovation RNA Amplification System (Nugen Technologies Inc., San Carlos, USA) was used to the letter of the manufacturer's instructions, generating approximately 6 μg of cDNA. First strand cDNA is prepared from total RNA using both oligo-dT and random hexamer primers and reverse transcriptase. After the generation of double strand cDNA, a DNA amplification step developed by NuGEN was performed. The cDNA was diluted to 50 μl.

PCR amplification reactions were carried out in a total volume of 8 μl containing 2× SYBR Green I Master Mix (Eurogentec, Seraing, Belgium), 3 μl 1/100 cDNA (~3.75 ng) and 250 nM forward and reverse primers (BioLegio, Nijmegen, The Netherlands). All reactions were performed in 384-well plates (LightCycler 480 Multiwell Plates 384, white and LightCycler 480 Sealing Foils from Roche) on the CFX384 real-time PCR detection system (Bio-Rad, Hercules, California), followed by a regression Cq value determination method. Cycling conditions were as follows: 95°C for 10 min followed by 45 cycles of 95°C for 10 s and 60°C for 30 s, followed by a dissociation curve analysis from 60 to 95°C. Because instrument and liquid handling variations were shown to be minimal using the Tecan Freedom Evo robot for pippeting (< 6% CV in 0.5 μl and <3% CV in 2 μl), and a large number of biological replicates were used, no PCR replicates were carried out. Primers containing neither SNPs nor secondary structures were designed for Glyceraldehyde 3-phosphate dehydrogenase (GAPDH), Succinate dehydrogenase complex subunit A (SDHA), Hypoxanthine-guanine phosphoribosyltransferase (HPRT), IL8, TNF, TGFB1, IL1B, IL6, IL23A, CCR6, STAT3, IL17A, IL17F, IL21, IL22, IL26, CCL20 and RORC (Table 2). For RORC, two sets of primers were designed, RORC, which detects the mRNA levels of both isoforms, and RORC1 which detects only the full-length transcript. BLAST searches confirmed that only the target genes were 100% covered. A 6 point 4-fold standard dilution series (highest concentration; 32 ng/μl) of a cDNA mixture of all samples included in the study diluted in 5 ng/μl tRNA (Roche) was used to test the PCR efficiency of the primers. The dynamic range had to cover at least 3 orders of dilution. Only primers with an efficiency between 88% and 112% were retained (Table 2). Correlation coefficients of the targets were between 0.9843<R² < 1, with a mean of 0.9942 (Table 2). The PCR efficiency for each gene was calculated according to the equation E = 10 (-1/slope). Each sample has been revised for a melting-curve with a single sharp peak with a high correlation between the observed and the expected Tm (mean variation of 0.9°C). Samples with other patterns than a single sharp peak at the expected Tm, defined as multiple peaks, a single broader peak or a shoulder peak, were omitted. Cq values of samples with flattened melting-curves were set as 45. An amplification signal in the no template control (NTC) was ignored as long as the difference in Cq value between the NTC and the highest Cq >5. Although the pre-amplification method of NuGEN does not amplify genomic DNA, possible gDNA contamination was assessed using intronic primers. We confirmed that gDNA was undetectable in a dilution of up to 32 ng/μl cDNA [49]. The mRNA expression level of each gene was determined in Excel by using the comparative 2-(delta delta Cq) method and normalized to the geometric mean of the stably-expressed reference genes GAPDH, SDHA and HPRT as determined by geNorm [50].

According to the MIQE guidelines, the minimum information for publication of quantitative real-time PCR experiments was provided [51].