Activated cardiac fibroblasts are a primary source of high-molecular-weight hyaluronan production

Abstract

During acute myocardial infarction, the composition of the extracellular matrix changes remarkably. One of the most notable changes in the extracellular matrix is in the accumulation of collagen; however, hyaluronan rivals collagen in its abundance. Yet, the extent to which specific cells and enzymes may contribute to such accumulation has been largely unexplored. Here, we hypothesized that activated cardiac fibroblasts produce hyaluronan via hyaluronan synthase 2 (HAS2). We show that hyaluronan accumulates following myocardial infarction and persists through at least 4 wk. Our analyses of failing heart RNA sequencing data suggest that fibroblasts are the cells most changed in the expression of HAS2. Given these insights, we used HAS2 gain- and loss-of-function approaches to examine the extent to which activated cardiac fibroblasts produce hyaluronan. Transforming growth factor β (TGFβ)-induced activation of fibroblasts caused a significant increase in Has2 mRNA and concomitant accumulation of hyaluronan >1 MDa in size. Deletion of Has2 abrogated TGFβ-induced production of hyaluronan. In addition, overexpression of Has2 was sufficient to cause an increase in hyaluronan accumulation in the absence of TGFβ-induced activation. Our data indicated negligible impacts of Has2 on proliferation, migration, and collagen production. Exposing fibroblasts to exogenous hyaluronan also had minimal impact on fibroblasts. We also assessed whether fibroblast-borne Hyal2 plays a role in the degradation of hyaluronan, and our data indicated little impact of Hyal2 on hyaluronan accumulation (or even any impacts on the transcriptional profile of fibroblasts). Activated fibroblasts produce high-molecular-weight hyaluronan via Has2, which occurs independent of other fibroblast functions.NEW & NOTEWORTHY Activated cardiac fibroblasts produce copious quantities of collagen, and much is known about this process. They also produce hyaluronan, which is abundant in the extracellular matrix, but less is known about hyaluronan. Here, we identify cardiac fibroblasts as major producers of hyaluronan and, specifically, that they produce high-molecular-weight hyaluronan via HAS2. This has important implications for ventricular remodeling and for metabolic regulation of activated fibroblasts, as they produce this abundant matrix component.

Keywords: extracellular matrix; fibrosis; heart failure; myocardial infarction.

Figure 1.. Hyaluronan accumulates in the heart 7 d and 28 d post-MI.

Cardiac tissue from male and female mice were sectioned and stained with hyaluronan binding protein (HABP) 7 d after MI or sham and imaged with epifluorescence microscopy. (A,C) Representative images taken after staining with DAPI (blue), which stains for nuclei and HABP (red), which is hyaluronan binding protein. (B) Quantification of stained tissue shows a significant increase in hyaluronan accumulation in the total left ventricle and especially infarct zone (IZ) but not remote zone (RZ), 7 d post-MI. Sham n=10 (5F, 5M), MI n=8 (4F, 4M). unpaired t-test with Welch's correction. (D) Quantification of stained tissue shows a significant increase in hyaluronan accumulation in the total left ventricle and especially infarct zone (IZ) but not remote zone (RZ), 28 days post MI. Sham n=8 (4F, 4M), MI n=8 (6F, 2M). t-test with Welch's correction. (E) HA ELISA shows an increase in HA in MI tissue in both 7d and 28 MI hearts. n=7–10 Unpaired t-test (F) Representative images taken after echocardiographs were done and analyzed. (G) Quantification of measured echocardiogram showing a significant decrease in end-diastolic volume and end-systolic volume in the left ventricle 7 d post-MI. Sham n=20, MI n=14.

Figure 2.. Collagen accumulates in the heart 7 d and 28 d post-MI.

Cardiac tissue from male and female mice were sectioned and stained with picrosirius red 7 days after MI. (A,C) Representative images are shown; reddish color indicates collagen. (B) Quantification of stained tissue shows a significant increase in collagen accumulation in the total left ventricle and especially infarct zone (IZ) but not remote zone (RZ) 7 days post MI. Sham n= 10 (5F, 5M), MI n=8 (4F, 4M). t-test with Welch's correction. (D) Quantification of stained tissue shows a significant increase in collagen accumulation in the total left ventricle, especially infarct zone (IZ) 28 days post MI. An increase in collagen was also seen in the remote zone. Sham n=8 (4F, 4M), MI n=8 (6F, 2M). t-test with Welch's correction.

Figure 3.. High molecular weight hyaluronan accumulates in the heart post-MI.

Hyaluronan was isolated from male and female mouse hearts after 7 d MI or 28 d MI. Samples were electrophoresed and stained with Stains-All to indicate hyaluronan (blue). (A) Representative assessment of the size and abundance of hyaluronan isolated from mouse hearts. (B) Quantification of stained agarose gels (n=6 per group). HA^HMW increases chronically in the infarct zone (IZ) after myocardial infarction (MI) compared to naïve hearts, 7 d infarct zone (IZ), and 28 d remote zone (RZ). Quantification of stained agarose gels (n=6 per group). One-way ANOVA test with Tukey's multiple comparisons test.

Figure 4.. High molecular weight hyaluronan accumulation is increased in activated cardiac fibroblasts.

Hyaluronan was isolated from the media of cultured cardiac fibroblasts. Samples were electrophoresed and stained with Stains-All to indicate hyaluronan (blue). All fibroblasts were from WT mice. (A) Representative gel showing relative size and abundance of hyaluronan. High molecular weight hyaluronan (HA^HMW) was loaded as a molecular weight marker/positive control. (B) Quantification of stained agarose gels. Stimulation of fibroblasts with TGFβ for 24 h caused accumulation of HA^HMW compared to naïve fibroblasts n=5 (3F, 2M). Paired t-test. (C) Heat map illustrating Log2 fold changes of genes involved in hyaluronan metabolism, hyaluronan receptors, and ECM proteins from activated fibroblasts stimulated with TGFβ for 24 h compared to naïve fibroblasts. Quantification of Color: Blue color indicates downregulation of genes and maroon color indicates up-regulation of genes. Significance: p<0.05 = *, p<0.01 = **, p<0.001 = ***, p<0.0001 = ****, p<0.00001 = ***** and NS = not significant.

Figure 5.. Has2 is necessary and sufficient for TGF-induced HA production.

Hyaluronan was isolated from the media of cultured cardiac fibroblasts. Samples were electrophoresed and stained with Stains-All to indicate hyaluronan (blue). All fibroblasts were from Has2^fl/fl mice. (A) Representative gel showing relative size and abundance of hyaluronan. Overexpressing Has2 in fibroblasts via adenovirus (Ad-Has2) increased HA accumulation in the media compared to Ad-GFP. Deletion of Has2 attenuated accumulation of HA in the media compared to Ad-GFP. Deletion of Has2 attenuated TGF-induced accumulation of HA in the media compared to Ad-GFP TGFβ. (B) Quantification of stained agarose gels n=6 (3F, 3M). (C) Relative mRNA expression of Has2 is significantly decreased in both naïve and activated fibroblasts with Has2 deletion via adenovirus (Ad-Cre). Has2 is significantly increased in fibroblasts with Has2 overexpression via adenovirus (Ad-Has2). n=6 (3F, 3M). Ratio paired t-test, parametric w/ SEM.

Figure 6.. Depletion of glucose causes a decrease in hyaluronan accumulation in activated fibroblasts.

Hyaluronan was isolated from the media of cultured cardiac fibroblasts. Samples were electrophoresed and stained with Stains-All to indicate hyaluronan (blue). All fibroblasts were taken from WT mice. (A) Representative gel showing relative abundance of hyaluronan. Naïve fibroblasts treated with low or high glucose concentrations showed no significant differences compared to normal glucose concentrations. (B) Quantification of stained agarose gels. n=5–6 (2–3F, 2–3M) (C) Representative gel showing relative abundance of hyaluronan. TGFβ-activated fibroblasts treated with low or high glucose concentrations showed no significant differences compared to normal glucose concentrations. (D) Quantification of stained agarose gels. n=6 (3F, 3M). Paired t-test between treatment and control. (E) Representative gel showing relative size and abundance of hyaluronan. Quantification of stained agarose gels (F) Naïve and (G) TGFβ-activated fibroblasts treated with no glucose showed no significant differences compared to normal glucose concentrations. n=3 (3F). (H) Representative gel showing relative size and abundance of hyaluronan. Quantification of stained agarose gels (I) naïve fibroblasts treated with no glucose showed no significant differences compared to normal glucose concentrations. (J) TGFβ-activated fibroblasts treated with no glucose showed a significant decrease in HA^HMWcompared to normal glucose concentrations. n=3 (3F). Paired t-test between treatment and control.

Figure 7.. Hyal2 alone does not degrade HA derived from cardiac fibroblasts.

Hyaluronan was isolated from the media of cultured cardiac fibroblasts. Samples were electrophoresed and stained with Stains-All to indicate hyaluronan (blue). All fibroblasts were from Hyal2^fl/fl mice. (A) Representative gel showing relative abundance of hyaluronan. (B) Quantification of stained agarose gels. Overexpressing Hyal2 in fibroblasts via adenovirus (Ad-Hyal2) does not change HA accumulation in the media compared to Ad-GFP. Deletion of Hyal2 via adenovirus (Ad-Cre) attenuated TGF-induced accumulation of HA in the media. n=8 (4F, 4M). paired t-test. (C) Relative mRNA expression. Hyal2 is significantly decreased in both naïve and activated fibroblasts with Hyal2 deletion via adenovirus (Ad-Cre). Hyal2 is significantly increased in fibroblasts with Hyal2 overexpression via adenovirus (Ad-Hyal2). n=6–8 (3-5F, 3M). paired t-test

Figure 8.. Little impact of HA on fibroblast collagen secretion.

(A) Linear regression analysis of the association between hyaluronan accumulation (HABP) and fibrotic area (Picro Sirius). 7 d MI hearts showed no significant relationship between an increase in HABP area and fibrotic area. (n=8, p = 0.1734, R2 = 0.2845) (B) Linear regression analysis of the association between hyaluronan accumulation (HABP) and fibrotic area (Picro Sirius). 28 d MI hearts showed a significant relationship between an increase in HABP area and increase in fibrotic area. (n=8, p = 0.0003, R2 = 0.8987) (C) Cardiac fibroblasts were treated and protein was isolated. All fibroblasts were from Has2^fl/fl mice. Representative Western Blot. (D) Quantification of Western blot showed that overexpression or deletion of Has2 in naive fibroblasts via adenovirus (Ad-Cre and Ad-Has2) showed no significant difference in collagen protein levels compared to GFP control. n=6 (3F, 3M) Paired t-test TGF-activated fibroblasts showed a significant increase in collagen protein levels compared to bFGF control. n=6 (3F, 3M) Paired t-test. (E) Quantification of hydroxyproline assay. There was no significant difference in hydroxyproline levels in Has2 deleted and overexpressed naïve fibroblasts compared to GFP control. n=5 (3F, 2M) Paired t-test.