Author + information
- Carolina Gil-Cayuela, PhD,
- Miguel Rivera, MD, PhD,
- Ana Ortega, PhD,
- Estefanía Tarazón, PhD,
- Juan Carlos Triviño, PhD,
- Francisca Lago, PhD,
- José Ramón González-Juanatey, MD, PhD,
- Luis Almenar, MD, PhD,
- Luis Martínez-Dolz, MD, PhD and
- Manuel Portolés, PhD∗ ()
- ↵∗Cardiocirculatory Unit, Instituto de Investigación Sanitaria La Fe, Hospital Universitario La Fe, Avenida Fernando Abril Martorell, 106, 46026 Valencia, Spain
Ventricular remodeling, a process involving morphological changes in cardiomyocytes and the extracellular matrix (ECM), has been linked to worsening of clinical status and heart failure (HF) development. In HF, remodeling initially occurs as a compensatory response to preserve the structural integrity of the myocardium, but progressive collagen deposition can lead to cardiac fibrosis and impairment of diastolic and systolic function (1). Thus, we analyzed changes in the expression profile of collagen-related genes in patients with ischemic cardiomyopathy (ICM) and examined the relationships with left ventricular (LV) dysfunction.
We obtained 23 human LV tissue samples from patients with ICM (13 men; mean age, 54 ± 8 years) with end-stage HF undergoing heart transplantation and control donors (CNT) (8 men, 2 women; mean age, 47 ± 16 years). All control donors had normal LV function (ejection fraction >50%) and no history of cardiac disease. As an established condition for inclusion in the study, all selected samples displayed a 260/280 nm absorbance ratio >2.0 and RNA integrity number ≥9. Transcriptome-level differences between ICM and CNT samples were investigated by means of large-scale screening of 23 heart samples with the use of RNA-sequencing technology and further validated by means of real-time (RT-PCR) and Western blot analysis (Figure 1). These data have been deposited in the NCBI Gene Expression Omnibus (GEO) (retrieved by use of the GEO Series accession No. GSE55296). We performed quantitative RT-PCR assays in duplicate with the use of TaqMan technology in the ViiA7 The Fast Real-Time RT-PCR System (Applied Biosystems, Foster City, California); collagen IV (COL4A5, Hs00166712_m1), collagen XIV (COL14A1, Hs00964045_m1), and collagen XVI (COL16A1, Hs00156876_m1) was obtained from TaqMan. Housekeeping genes GAPDH (Hs99999905_m1), PGK1 (Hs99999906_m1), and TFRC (Hs00951083_m1) were used as reference. Regarding the Western blot analysis, tissue samples were transferred into Lysing Matrix D tubes designed for use with the FastPrep-24 homogenizer (MP Biomedicals, Santa Ana, California). Protein samples were separated by tris-acetate electrophoresis on 3% to 8% polyacrylamide gels under no-reducing conditions and transferred to a PVDF membrane with the use of the iBlot Gel Transfer Device (Life Technologies, Carlsbad, California) for Western blot analyses. The primary detection antibodies used were anti-COL4A5 rabbit polyclonal (sc-11360) obtained from Santa Cruz Biotechnology, anti-COL14A1 rabbit polyclonal (ab-101464) obtained from Abcam, and anti-COL16A1 rabbit polyclonal (A-96190) obtained from Sigma; monoclonal anti-GAPDH antibody (ab-9484) from Abcam was used as a loading control. We must note that patients with end-stage HF are under heavy medical treatment, and some therapies might influence mRNA levels.
We compared ICM and CNT samples, and we found significantly increased mRNA levels of 11 collagen genes, such as COL9A1, COL11A2, COL14A1, and COL16A1 (p < 0.05 for all), not previously described in the cardiac remodeling process. A heat map and hierarchical clustering analysis identified 2 divergent gene expression profiles, showing a clear separation of the ICM and CNT groups (Figure 1). We also found significant relationships between LV dysfunction and the gene expression levels of COL4A5 (fractional shortening, r = −0.694, p < 0.05) and COL16A1 (LV end-systolic diameter and LV end-diastolic diameter, r = 0.678, p < 0.05, and r = 0.687, p < 0.05, respectively).
For validation of the novel regulated transcripts, we performed quantitative RT-PCR and Western blot analysis, focusing on those related significantly with the LV function: COL16A1 and COL4A5, and COL14A1, newly described. These results (Figure 1) confirm that patients with ICM showed direction changes in expression identical to what we found in the RNA-Seq analysis.
Collagens types IX, XIV, and XVI belong to the fibril-associated collagens with interrupted triple helices class, highly expressed in tissues that have high mechanical stress, such as heart tissue. Collagen IX is thought to be required for several processes, including eye and heart development, and its overexpression has been observed in pathologies such as pectus excavatum (2). In light of these studies, our results suggest that overexpression of COL9A1 could lead to increased cardiac tissue stiffness and may be potentially involved in human heart development and remodeling.
Previous studies suggest that collagen XIV is required to establish and maintain an organized ECM environment in the developing myocardium and plays an important role in regulating cardiomyocyte growth and cardiac fibroblast survival (3). During replacement fibrosis, cardiomyocytes undergo hypertrophic adaptive changes, whereas myofibroblasts remain at the site of injury; this results in collagen deposition and scar formation. We propose that collagen XIV expression could promote this process.
Collagen XVI, a minor collagen component of connective tissues, is thought to act as a linker protein, helping to organize the large fibrillar networks that modulate ECM integrity and stability. A recent study proposed that in Crohn’s disease, increased collagen XVI expression would promote formation and maturation of focal adhesion contacts on intestinal subepithelial myofibroblasts (4). This would retain the cells at the inflammation site, promoting fibrotic responses in the tissue and prolonging disturbances of cellular and ECM homeostasis. We suggest that dysregulated collagen XVI expression could play a similar role in cardiac tissue by keeping myofibroblasts at the inflammation site and promoting pathological remodeling.
Furthermore, collagen XI promotes the nucleation of type I and II fibrils and is required for myocardial morphogenesis (5). The overexpression of COL11A2 may be related to the formation of heterotypic fibrils with collagen I, involved in cardiac remodeling.
Our findings propose that these collagen genes may have novel roles in the remodeling process, regulating the size increase of cardiomyocytes and the survival of myofibroblasts at the inflammation site and assisting the organization in fibers of other collagens, such as type I and III. All these processes jointly may facilitate the development of cardiac fibrosis and, consequently, ventricular dysfunction. Inhibition of collagen remodeling can lead to improved cardiac function, demonstrating the relevance of new insights into the compensatory remodeling mechanism. Thus, our findings give new theoretical support for the treatment of patients with ischemic cardiomyopathy.
Please note: This work was supported by grants from the National Institute of Health “Fondo de Investigaciones Sanitarias, Instituto de Salud Carlos III,” Madrid, Spain [PI10/00275, PI13/00100]; European Regional Development Fund (ERDF); and RETICS, Madrid, Spain [RD 06/0003/1001, 12/0042/0003]. The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- 2015 American College of Cardiology Foundation