Primary bronchial fibroblast cultures
Fibroblasts were grown from bronchial biopsy specimens obtained from normal (n
6) or asthmatic (n
7) volunteers by fibreoptic bronchoscopy following standard guidelines.15
The normal subjects (4:2 M:F, mean age 21 (range 20–21) years) had a mean (SD) forced expiratory volume in 1 second (FEV1
) of 97.3 (5.2)% predicted and the asthmatic subjects (4:3 M:F, mean age 21 (range 20–26)) had an FEV1
of 76.4 (7.1)% predicted. Asthmatic subjects were using β2
agonists only, as required. Primary fibroblast cultures were established as previously described16
and were used between passage 2 and 7; typically, all cells were positive for vimentin and <1% of cells were αSMA positive in the absence of TGFβ2
To induce myofibroblast differentiation, fibroblasts were seeded onto collagen coated (30 μg/ml Vitrogen in H2O; Nutacon, Leimuiden, The Netherlands) dishes and grown to 80–95% confluence before serum starvation (Ultraculture; Biowhittaker, Wokingham, UK) for 24 hours followed by treatment with TGFβ2 (Sigma, Poole, UK).
RNA extraction and reverse transcription
RNA was extracted using TRIzol reagent (Invitrogen, Paisley, UK) and contaminating DNA removed using DNase (Ambion, Austin, USA) according to the manufacturer's instructions. 1 μg total RNA was reverse transcribed using 1 mM dNTP (Invitrogen, Paisley, UK), 3 ng random hexamer primer (MWG Biotech, Milton Keynes, UK), and 100 U M‐MLV RT enzyme (Promega, Chilworth, UK) according to the manufacturers' instructions.
Real‐time qPCR analysis and normalisation using genNorm
To establish the most stable genes for normalising, control gene expression was measured in 12 fibroblast cultures before and after myofibroblast differentiation. Twelve normalising genes (Eurogentech, Seraing, Belgium) were selected for analysis; 18S ribosomal RNA (18S), 28S ribosomal RNA (28S), beta actin (ACTβ), glyceraldehyde‐3‐phosphate (GAPDH), ubiquitin C (UBC), beta‐2‐microglobulin (B2M), phospholipase A2 (A2), ribosomal protein L13a (RPL13A), succinate dehydrogenase (SDHA), hypoxanthine phosphoribosyl‐transferase 1 (HPRT1), TATA box binding protein (TBP), and hydroxymethyl bilane synthase (HMBS). To validate these assays for use with the ΔCT method, melt curve analyses were performed to determine specificity and standard curves plotted as CT value versus log dilution from serially diluted cDNA. From these data the priming efficiency was calculated according to the formula (10(−1/gradient)) − 1 and was found to be close to the theoretical maximum for each assay. Target genes were analysed in the same way and validated for ΔCT based methods of quantification.
geNorm analysis was performed using the geNorm applet (http://medgen31.ugent.be/jvdesomp/genorm/
) according to the guidelines and theoretical framework previously described.17
This approach is based on the principle that, regardless of conditions, the expression ratio of two ideal normalising genes will remain constant in all samples. Hence any variation in expression ratio between the two is indicative of one (or both) genes being variably expressed. In brief, for each normalising gene, ΔCT
calculations were performed relative to the strongest signal which was assigned the value 1 and these data used as the input for geNorm. Output files rating gene stability (M) and variation in normalising signal (NF) were collected for fibroblasts, myofibroblasts, and the combined data set for both (fig 1).
Figure 1Stability of constitutively expressed housekeeping genes measured in fibroblasts, myofibroblasts, and in a combined data set using geNorm software. The M value is calculated based on the changing ratios of genes in the analysis as previously (more ...)
For detection of the smooth muscle related genes, αSMA, HCM, calponin 1, desmin, and γ‐actin primers and fluorogenic probes labelled with the 5′ reporter dye 6‐carboxy‐fluorescein (FAM) and the 3′ quencher dye 6‐carboxy‐N,N,N′,N′‐tetramethyl‐rhodamine (TAMRA) were designed using beacon designer 2.1 (Premier Bio‐soft). The sequences were as shown in table 1.
Table 1Primer sequences used in study
For each sample the PCR reaction was performed in duplicate and contained 25 ng cDNA template, 3 pmol fluorogenic probe, 15 pmol forward and reverse primers, 12.5 μl universal qPCR master mix (Eurogentech, Seraing, Belgium), made up to 25 μl with water. RT negative samples confirmed that the signals were not due to genomic contamination. The PCR cycling conditions were 95°C for 10 minutes followed by 42 cycles of denaturation at 95°C for 15 seconds and annealing/extension at 60°C for 1 minute. Quantitation and real time detection of the PCR products was followed on an icycler IQ real time detection system (Bio‐rad, Hemel Hempstead, UK).
Based on the geNorm analysis (fig 1), the geometric mean of UBC and A2 was used as the normalising signal. Data were either analysed relative to the geometric mean of UBC and A2 using a standard ΔCT calculation or normalised to the geometric mean of UBC and A2 and expressed relative to the signal obtained for the average of the untreated time zero controls using a ΔΔCT calculation.
Cells were fixed in acetone and air dried prior to rehydration for immunocytochemical staining using the following primary antibodies: anti‐αSMA (1:500), anti‐HCM (1:100), anti‐calponin 1 (1:200), anti‐desmin (1:40) (all antibodies were from Sigma, Poole, UK) with anti‐mouse horseradish peroxidase conjugated secondary antibody (1:100; Dako, Glostrup, Denmark).
Cell lysates were prepared in SDS buffer (0.3 M Tris‐HCl pH 6.8, 50% glycerol, 25% 2‐mercaptoethanol, 10% SDS, 0.01% bromophenol blue) containing protease inhibitors. Samples were subjected to SDS‐PAGE in 12.5% polyacrylamide gels and then transferred onto PVDF membranes (Amersham) for immunoblotting using antibodies against αSMA (1:5000), CPN 1 (1:10
000), β‐actin (1:20
000) with ECL detection. Semiquantitative analysis of protein expression was performed by densitometry using Genetools software (Syngene, Cambridge, UK); during this analysis the operator was blinded to the identity of the samples. Since β‐actin was relatively stably expressed during TGFβ2
induced myofibroblast differentiation, it was used to control for protein loading. Data are expressed relative to time zero control and are presented as box and whisker plots showing the median and interquartile range with 95% confidence intervals.
Data were analysed using non‐parametric tests for within (Wilcoxon signed rank test) or between (Mann‐Whitney U test) group comparisons using SPSS for Windows Version 11.5 (Chicago, IL, USA).
Ethical approval was obtained from the Southampton and South West Hampshire joint ethics committee.