Unique mechanisms of connective tissue growth factor regulation in airway smooth muscle in asthma: Relationship with airway remodelling

Abstract Neovascularization, increased basal membrane thickness and increased airway smooth muscle (ASM) bulk are hallmarks of airway remodelling in asthma. In this study, we examined connective tissue growth factor (CTGF) dysregulation in human lung tissue and animal models of allergic airway disease. Immunohistochemistry revealed that ASM cells from patients with severe asthma (A) exhibited high expression of CTGF, compared to mild and non‐asthmatic (NA) tissues. This finding was replicated in a sheep model of allergic airways disease. In vitro, transforming growth factor (TGF)‐β increased CTGF expression both in NA‐ and A‐ASM cells but the expression was higher in A‐ASM at both the mRNA and protein level as assessed by PCR and Western blot. Transfection of CTGF promoter‐luciferase reporter constructs into NA‐ and A‐ASM cells indicated that no region of the CTGF promoter (−1500 to +200 bp) displayed enhanced activity in the presence of TGF‐β. However, in silico analysis of the CTGF promoter suggested that distant transcription factor binding sites may influence CTGF promoter activation by TGF‐β in ASM cells. The discord between promoter activity and mRNA expression was also explained, in part, by differential post‐transcriptional regulation in A‐ASM cells due to enhanced mRNA stability for CTGF. In patients, higher CTGF gene expression in bronchial biopsies was correlated with increased basement membrane thickness indicating that the enhanced CTGF expression in A‐ASM may contribute to airway remodelling in asthma.

However, in silico analysis of the CTGF promoter suggested that distant transcription factor binding sites may influence CTGF promoter activation by TGF-b in ASM cells. The discord between promoter activity and mRNA expression was also explained, in part, by differential post-transcriptional regulation in A-ASM cells due to enhanced mRNA stability for CTGF. In patients, higher CTGF gene expression in bronchial biopsies was correlated with increased basement membrane thickness indicating that the enhanced CTGF expression in A-ASM may contribute to airway remodelling in asthma.

K E Y W O R D S
airway remodelling, airway smooth muscle, asthma, connective tissue growth factor 1 | INTRODUCTION Asthma is a common, chronic respiratory disease affecting more than 300 million people worldwide. 1 The main characteristics of asthma are airway inflammation, airway hyper-responsiveness and airway remodelling. 2 The structural changes in the airways, termed airway remodelling, include increased airway smooth muscle (ASM) bulk, increased basal membrane thickness and vascular expansion. 3,4 The extent of airway way remodelling correlates with several clinical features of asthma [5][6][7][8] and agents that normalize the remodelling response potentially improve asthma symptoms. [9][10][11] Once considered a manifestation of chronic inflammation, recent studies have identified remodelling is a separate but parallel component of the asthmatic process. 12 In asthma, the observed increase in ASM cell bulk and contractility directly mediates airway narrowing and is central to the process of airway remodelling. Increased ASM number correlates with increased reticular basement membrane (BM) thickness and eosinophilia, but not neutrophilia. 13 Further, the secretory profiles of ASM from asthmatic patients differ significantly from those of nonasthmatic patients suggesting that paracrine signalling from the ASM in may have as much to do with airway remodelling as their contractile state (reviewed in Ref. [14]). Ultimately, the increased deposition of extracellular matrix (ECM) proteins by ASM cells in asthma is key to the airway narrowing that takes place. 12,[15][16][17][18] Often this is a response to an imbalance in the cytokines/growth factors present in their local milieu. 3 In lung tissue, ASM cells are a potent source of connective tissue growth factor (CTGF), a member of the cysteine-rich 61, CTGF, nephroblastoma (CCN) family of proteins. 19 Our previous studies have shown greater CTGF expression in primary asthmatic (A)-ASM cells than non-asthmatic (NA)-ASM cells after TGF-b treatment. [20][21][22] CTGF controls ECM deposition and ultimately airway biomechanics through changes to collagen deposition which increase ECM density and airway stiffness. 23 Indeed, the increased stiffness of the matrix in which asthmatic ASM cells are embedded promotes a more proliferative and pro-inflammatory ASM phenotype. 14 The mechanisms underlying the differential regulation of CTGF expression in A-ASM are not currently known. Studies in other systems have reported that CTGF induction by TGF-b is regulated through interactions of transcription factors with promoter elements directly upstream of the promoter start site. [24][25][26][27] In this study, we investigated the mechanisms that enhance TGF-b induction of CTGF release from A-ASM cells and the potential links to airway remodelling in asthma.

| Detection of CTGF by immunohistochemistry
Sections from archived paraffin-embedded lung tissue were obtained from non-, mild and severely asthmatic patients 3,4 as well as a sheep model of allergic airways disease that had been previously described. 30

| CTGF promoter constructs
A Gluc-on reporter plasmid containing the full-length CTGF promoter (À1500 to +200 bp) driving expression of a secreted Gaussia    F I G U R E 1 Connective tissue growth factor (CTGF) expression is increased in house dust mite (HDM)-induced allergic airway disease in sheep lungs. CTGF expression in a model of allergic airway disease was assessed by immunohistochemistry in HDM-and saline-exposed (sham control) lung segments from the same sheep 30 (n = 5). Isotype-matched negative control antibody on serial sections shown for comparison. Representative images shown for each group former asthma patients. All patients originated from cohorts investigated earlier by our research group, and a set of previously acquired clinical data is available 31,32 . The study protocol was approved by the University Medical Center Groningen medical ethics committee. All patients gave their written informed consent.
For full patient information and details relating to RNA isolation and sequencing, refer to Appendix S1.

| Tissue specific genetic elements indicate CTGF regulation in lung tissue is unique
To assess whether our promoter construct was indeed inducible by | 2831 different to other tissues, we transfected NIH-3T3 cells and examined luciferase activity. Previous reports have shown that CTGF promoter-luciferase reporter constructs increase activity 2-to 4fold when NIH-3T3 cells are stimulated with TGF-b. 25,27,33,34 Indeed, treatment of transfected NIH-3T3 fibroblasts with TGF-b induced a 2-fold increase in luciferase expression compared to unstimulated cells ( Figure S1) indicating our CTGF promoter construct (À400 to +200) was indeed inducible but just not in human ASM cells.

TGF-b and to determine whether CTGF regulation in ASM was
Having found that the CTGF promoter regulation in human ASM cells differed from that reported in other cell lines, 25,27,33,34 we investigated regions of transcription factor binding activity surrounding the CTGF transcriptional start site to look for additional regulatory elements. This analysis was conducted by investigating H3K27Ac binding (a marker of transcription factor binding). For this analysis, we used human lung fibroblasts as previously no differences in gene expression were detected between lung fibroblasts and ASM cells, indicating highly similar gene expression regulation. 35 There was a strong region of activity immediately upstream of the CTGF transcriptional start site in HUVECs (À1300 to À200 bp), which was less active in human lung fibroblasts ( Figure 4A). This region spanned the À1500 bp promoter construct we had analysed ( Figure 4B), and contained several validated SMAD and TGF-b response elements previously reported to drive CTGF expression in other species (Figure 4C, Table 3). 25,27,36 However, an alternate genomic region 5 0 to the (À1300 to À200 bp) site (À4200 to À2400 bp) showed robust H3K27Ac binding in lung cells but relatively low activity in HUVECs ( Figure 4A). These findings suggest that this region may be responsible for the alternative regulation of CTGF expression in human lung cells. A-ASM cells was more stable than NA-ASM cells providing a possible explanation for the differential expression of TGF-b-induced CTGF between A-and NA-ASM cells.

| CTGF gene expression is unchanged in mild asthmatic patients but relates to BM thickness
Immunohistochemical staining showed CTGF protein expression was concentrated in the ASM area in human lung tissue, with enhanced detection visible in asthmatic tissues, particularly from severe asthma patients ( Figure 6A). No difference in CTGF mRNA expression (Figure 6B) was detected between bronchial biopsies derived from mild to moderately severe asthma patients (n = 69) and healthy controls (n = 77). In addition, within asthmatic patients, we found a significant relationship between CTGF expression and BM thickness (b AE SE 0.472 AE 0.174, P = .008, Figure 6C) suggesting ASMderived CTGF expression may influence airway narrowing and remodelling in asthma. In contrast, higher CTGF expression in asthmatic patients was not associated with lower FEV1% predicted (Figure 6D), more severe bronchial hyper-responsiveness ( Figure 6F) or higher % eosinophil levels in sputum ( Figure 6E).

| DISCUSSION
This is the first study that has focused on the molecular regulation In our study, we observed greater CTGF expression in airway tissues taken from lung segments of sheep chronically exposed to  CTGF, connective tissue growth factor; TGF, transforming growth factor.
F I G U R E 5 Connective tissue growth factor (CTGF) mRNA stability is enhanced in A-ASM cells. NA-(n = 4) and A-ASM cells (n = 5) were treated with TGF-b (1 ng/mL) with actinomycin D (10 lg/mL) added after 8 h for up to 16 h. CTGF mRNA expression was measured by Q-PCR to assess the rate of turnover. *Means significant difference in CTGF mRNA expression to time 0, *P < .05, ***P < .001, ****P < .0001. #P < .05 indicates a significant difference between NA-and A-ASM. ASM, asthmatic airway smooth muscle; TGF, transforming growth factor previously shown TGF-b induced CTGF through activation of the extracellular signal-regulated kinase (ERK) and phosphatidylinositol 3-kinase (PI3K) signalling pathways in ASM cells. ERK is known to be linked to SMAD2/3 activation 36 and likely targets the traditional core promoter region (À1300 to À200 bp) of CTGF which appears to be SMAD sensitive ( Figure 4). However, PI3K has not been associated with SMAD 2/3 signalling to date and therefore may target the alternative promoter regulatory region we have identified in this study. Moreover, how these distant genomic elements are recruited to the core promoter to modulate CTGF expression is unknown and may be influenced by epigenetic modification of histones (particularly K 27 acetylation) which is readily acknowledged to be different in asthmatic and healthy airways. 40,41 Further research is necessary to identify the transcriptional regulatory elements, potentially within the À4200to À2400-bp region, activated by these alternative signalling pathways for driving CTGF expression.
The similarities in basal promoter activity in A-and NA-ASM were just as surprising as the lack of TGF-b responsiveness in the CTGF promoter. We identified that the basal promoter in ASM is located between À100 and À400 bp. This region contains predicted binding We also showed that CTGF mRNA stability was enhanced in A-ASM cells. Chowdhury and colleagues previously showed this to be mediated by p38 in bladder smooth muscle cells. 37 However, our previous data conclusively showed this pathway is not involved with CTGF regulation by TGF-b in A-ASM. 22 The mechanism underlying this enhanced CTGF mRNA stability is currently unknown.
The dysregulation of CTGF in asthmatic airways may have profound consequences for disease progression, as suggested by the association of CTGF gene expression levels with BM thickening in our patient cohort. One limitation in our study is that the biopsies from which we obtained the gene signal were of a mixed cell population and we have no information about the ASM content in each biopsy. This may have altered the CTGF gene signal as CTGF is also expressed by airway epithelial cells and fibroblasts [43][44][45][46][47] and it is not known if these levels are also altered in asthma. If the epithelial gene expression of CTGF is not increased in asthma, this may have reduced the strength of the association we observed with BM thickness. The BM is 2-to 3-fold thicker in asthmatic compared to healthy airways and is associated with increased airway resistance, limitations to airflow and decreased lung function. 48,49 Association of BM thickening with poor clinical outcome is somewhat controversial with adults 32,50 but shows better correlation in children. [50][51][52] The thicker BM of asthmatic airways also has an altered elastic modulus compared to healthy airways. 53 The increased ECM stiffness that accompanies such a change is likely to contribute to the pro-remodelling environment found in asthmatic airways as stiffer matrices promote angiogenesis 54,55 and ASM cell proliferation. 53 Indeed, stiffer matrices may also enhance CTGF expression through Taz activation, 56,57 completing a positive feedback loop in the asthmatic airway that would co-ordinate all aspects of airway remodelling (ASM bulk, neovascularization and BM thickening).
In conclusion, our data strongly suggest that the unique regulatory mechanisms that underpin the enhanced CTGF expression in A-ASM are pivotal for the development of airway remodelling. Thus, CTGF represents an underappreciated target for future therapeutic intervention addressing an aspect of disease pathogenesis currently not effectively treated by existing approaches.

CONFLI CTS OF INTERES TS
The authors confirm there are no conflict of interests in this study. F I G U R E 6 Connective tissue growth factor (CTGF) expression and correlations with clinical indices in asthmatic patients. A, CTGF expression was assessed by immunohistochemistry in human lung tissue (n = 5 for healthy control, mild asthma and severe asthma). Representative images shown for each group. B-E, CTGF mRNA expression fragments per kilobase million (FPKM) was detected in bronchial biopsies from healthy controls and mild asthmatic patients (B). A linear model comparing the association between CTGF expression in asthmatic bronchial biopsies and BM thickness (lmol/L) (C), FEV1% predicted (D), % of sputum eosinophils (E) and PC20 mg/mL (F) was conducted correcting for age, gender and smoking status. b, correlation co-efficient; P, significance value of the correlation. BM basement membrane, FEV1% predicted forced expiratory volume in 1 s percentage predicted, PC20 the concentration of methacholine needed to produce a 20% fall in FEV(1) from baseline. See Ref. [32] (Table 1) for lung function on this cohort