Dynamic molecular and histopathological changes in the extracellular matrix and inflammation in the transition to heart failure in isolated volume overload

Abstract

Left ventricular (LV) volume overload (VO) causes eccentric remodeling with inflammatory cell infiltration and extracellular matrix (ECM) degradation, for which there is currently no proven therapy. To uncover new pathways that connect inflammation and ECM homeostasis with cellular dysfunction, we determined the cardiac transciptome in subacute, compensated, and decompensated stages based on in vivo hemodynamics and echocardiography in the rat with aortocaval fistula (ACF). LV dilatation at 5 wk was associated with a normal LV end-diastolic dimension-to-posterior wall thickness ratio (LVEDD/PWT; compensated), whereas the early 2-wk (subacute) and late 15-wk (decompensated) ACF groups had significant increases in LVEDD/PWT. Subacute and decompensated stages had a significant upregulation of genes related to inflammation, the ECM, the cell cycle, and apoptosis. These changes were accompanied by neutrophil and macrophage infiltration, nonmyocyte apoptosis, and interstitial collagen loss. At 15 wk, there was a 40-fold increase in the matricellular protein periostin, which inhibits connections between collagen and cells, thereby potentially mediating a side-to-side slippage of cardiomyocytes and LV dilatation. The majority of downregulated genes was composed of mitochondrial enzymes whose suppression progressed from 5 to 15 wk concomitant with LV dilatation and systolic heart failure. The profound decrease in gene expression related to fatty acid, amino acid, and glucose metabolism was associated with the downregulation of peroxisome proliferator associated receptor (PPAR)-α-related and bioenergetic-related genes at 15 wk. In VO, an early phase of inflammation subsides at 5 wk but reappears at 15 wk with marked periostin production along with the suppression of genes related to PPAR-α and energy metabolism.

Fig. 1.

Venn diagram showing the number of genes differentially expressed in aortocaval fistula (ACF) rats by microarray analysis and validation of the microarray results by real-time quantitative (q)PCR. A and B: upregulated (A) and downregulated (B) genes in 2-, 5-, and 15-wk ACF rats. C: comparison of gene fold changes by microarray and real-time PCR of 35 altered genes identified by microarray at 2 wk with Pearson's correlation coefficient (r) and a linear regression formula. See Supplemental Table S1 for gene annotation.

Fig. 2.

Heat map demonstrating extracellular matrix (ECM)-related genes identified by DAVID analysis among upregulated genes and expression of the ECM in the left ventricle (LV) of sham-operated (sham) and ACF rats. A: upregulated genes related to ECM turnover at 2, 5, and 15 wk. See Supplemental Tables S2 and S3 for gene annotation. B and C: LV interstitial collagen measured by picric acid sirius red (PASR) staining (B) and hydroxyproline content (C) in age-matched sham and ACF rats at 2, 5, and 15 wk. D: representative images of perivascular and endocardial collagen staining by PASR. Excessive collagen accumulation was observed in perivascular LV areas in ACF rats. E: representative gel zymography image demonstrating matrix metalloproteinase (MMP)-2 activity in sham and ACF rats at 2, 5, and 15 wk. Results are expressed as fold changes compared with the corresponding control. Bars = 20 μm. n = 5 rats/group except for the 15-wk ACF group, where n = 4. *P < 0.05 and **P < 0.01 vs. sham rats.

Fig. 3.

Expression of periostin in the LV of sham and ACF rats. A and B: representative Western blots with tubulin loading controls (A) and immunohistological staining (B) of periostin in age-matched sham and ACF rats. Endo, endocardium; V, vessel.

Fig. 4.

Heat map demonstrating upregulated inflammation-related genes identified by DAVID analysis and inflammatory cell infiltration and protein expression of inflammation-related genes. A: upregulated genes related to inflammation and the immune response at 2, 5, and 15 wk. See Supplemental Tables S2 and S3 for gene annotation. B and C: numbers of neutrophils (myeloperoxidase positive; B) and macrophages (CD68 positive; C) in ACF and age-matched shams at each time point. D and E: representative Western blots demonstrating the expression of NF-κB p65 (D) and heme oxygenase-1 (HO-1; E) with densitometric analysis after normalization to tubulin. n = 5 rats/group except for the 15-wk ACF group, where n = 4. *P < 0.05 and **P < 0.01 vs. sham rats.

Fig. 5.

Real-time PCR analysis of genes involved in bioenergetics and determination of apoptotic cells. A: heat map representing gene expression fold changes in ACF rats versus age-matched sham rats by microarray and real-time PCR at 2, 5, and 15 wk. The gene expression fold changes were imported into GeneSpring for heat map production. Fold changes were set to “null” when they had P values of >0.05 and are shown in gray. B: quantification of TUNEL-positive nonmyocyte nuclei in ACF and sham rats. n = 5 rats/group except for the 15-wk ACF group, where n = 4. *P < 0.05 and **P < 0.01 vs. sham rats.

Fig. 6.

Hypothesis for the dynamic molecular and histopathological changes in the ECM and inflammation in the transition to heart failure in isolated volume overload. In the early phase of inflammation, MMP activation transitions to LV “compensation” with stabilization of gene programs of inflammation and MMP activation as the LV end-diastolic dimension-to-posterior wall thickness ratio (LVEDD/wt) normalizes. Subsequent to this point, there is marked periostin production along with downregulation of genes related to peroxisome proliferator-activated receptor (PPAR)-α and energy metabolism. The marked upregulation of periostin can explain ECM dysregulation and LV dilatation, while gene programs related to PPAR-α downregulation connect the intensified inflammatory response with the profound depression of bioenergetics and metabolism genes in the progression of LV wall stress (σ) and heart failure in volume overload. FA, fatty acid.