Collagen Pyridinoline Cross-Links Are Absent in Reversible Myocardial Fibrosis

Myocardial fibrosis is a key feature of numerous cardiomyopathies.¹ Collagen, a hallmark of fibrosis, is stabilized via cross-links initiated by the oxidation of collagen lysines to allysines. These allysines form initial/reversible cross-links that rearrange over time resulting in mature, degradation-resistant cross-links including pyridinoline (PYD) and deoxypyridinoline (DPD).

We recently reported that PYD and DPD are absent in reversible scar despite abundant collagen oxidation.² An optical probe (TMR-O) that binds to allysine was developed and showed that the deposition and oxidation of collagen was similar in myocardial infarction/injury models where fibrosis is reversible (zebrafish) or irreversible (mouse). However, the mature cross-links PYD and DPD were completely absent in the myocardial scar of the regenerating zebrafish heart.²

Here, we report that the absence of degradation-resistant collagen cross-links (DPD and PYD) is also observed in a mammalian model of reversible cardiac fibrosis.¹ A dual-hit obesity and hypertension mouse model of HFpEF that promotes interstitial fibrosis was used (experimental design is outlined in Figure A).³ Male mice (8–10 weeks old) were studied since female sex is protective in this model. In one group, mice were treated with high fat diet and N^[w]-nitro-l-arginine methyl ester (L-NAME) for six months (HFLN group).³ In a second cohort, mice were returned to normal conditions for two months after the initial 6-month HFLN treatment (Withdrawal group). Age-matched naïve mice were included as controls (Aged Naïve group). At the end of the experimental timeline, mice were injected with TMR-O (2 nmol/g, intravenously) and the hearts were excised and imaged by fluorescence reflectance imaging.² A portion of the tissue was digested to assess allysine, the biochemical target of TMR-O. Additional sections of the heart were digested for hydroxyproline content (Hyp, a measure of total collagen) as well as PYD and DPD by biochemical assays.² PYD and DPD levels were compared to those in myocardial scar from a mouse model of myocardial infarction (MI).² In the withdrawal group, one animal died after injection and was excluded from TMR-O imaging, but tissue analysis was still feasible/performed. TMR-O quantification was conducted on a short-axis slice at the mid-ventricular level and fluorescence was measured as previously described.²

Figure:

A: Study design. B: (B1) TMR-O Fluorescence images of mouse hearts and corresponding histological assessment of fibrosis by (B2) Picrosirius Red (PSR, collagen in red) and (B3)Trichrome (collagen in blue). C: TMR-O quantifies collagen oxidation/initial cross-linking in the HFLN model of myocardial fibrosis. Collagen oxidation/initial cross-linking were elevated in the HFLN group compared to Aged Naïve by (C1) allysine detection by TMR-O and (C2) biochemical assessemnt. (C1, C2) Withdrawal of HFLN resulted in decreased allysine to levels similar to Aged Naïve mice. (C3) Detection of allysine by TMR-O correlated strongly with allysine concentration. (D) Collagen accumulation in HFLN mouse hearts was reversed by HFLN withdrawal. Collagen levels assessed by (D1) hydroxyproline (Hyp), and (D2) PSR were significanly increased in HFLN hearts compared to Aged Naïve hearts. Withdrawal of HFLN resulted in a significant decrease of Hyp levels. (E) Mature Collagen Cross-Links (DPD and PYD) were abundant in mouse MI but not in the HFLN fibrosis model. (E1) HPLC traces of PYD/DPD standards (500 nM), tissue digests from infarcted mice (MI), Aged Naïve, HFLN, and Withdrawal mouse hearts. The PYD and DPD peaks are much more prominent in the MI digests. IS= Internal Standard. Quantification of (E2) DPD and (E3) PYD, normalized to hydroxyproline concentrations, shows that MI scar has 60-fold higher PYD and 10-fold higher DPD per collagen compared to HFLN hearts. (E2, E3) PYD/DPD levels are similar in Aged Naïve, HFLN and Withdrawal mice. P-values were determined by ANOVA with Tukey’s post hoc comparisons.

TMR-O fluorescence (Figure B1) was significantly higher in the HFLN group than the Aged Naïve group (p= 0.0017, Figure C1). Withdrawal of the HFLN conditions for two months resulted in a significant decrease in TMR-O fluorescence compared to the HFLN group (p= 0.033, Figure C1). The differences in TMRO signal were supported by quantitative analysis of allysine (Figure C2). TMR-O intensity and allysine concentration had a strong positive correlation (p = 0.0019, Pearson Coefficient = 0.77, Figure C3), confirming the ability of TMR-O to assess collagen oxidation.² Fibrosis was assessed histologically by picrosirius red (PSR) (Figure B2) and trichrome (Figure B3). Total collagen in the hearts was determined by quantitative analysis of hydroxyproline (Hyp). Hyp was significantly increased in the hearts of the HFLN group compared to the Aged Naïve group, (p = 0.0298, Figure D1). The Withdrawal group showed a significant decrease in myocardial Hyp compared to the HFLN group (p= 0.0484, Figure D1). The hearts of young naïve mice aged 8–10 weeks had significantly lower Hyp content compared to all other groups (p < 0.001, Figure D1). Quantification of PSR sections showed a significant increase in collagen positive area (Figure D2) that correlated strongly and positively with Hyp measurements (p = 0.0004, Pearson Coefficient = 0.74 Figure D3).

PYD and DPD were abundant in murine infarcts (Mouse MI). However, these cross-links were far less detectible in all other groups (Figure E1). Quantification of PYD and DPD levels (normalized to hydroxyproline/total collagen) confirmed that DPD were 10-fold lower, and PYD 60-fold lower in the hearts of Aged Naïve, HFLN and withdrawal mice compared to MI mice (Figure E2–E3). No significant differences were seen in the levels of PYD and DPD between the Aged Naïve, HFLN and Withdrawal groups.

Our recent work has shown that the absence of degradation-resistant cross-links PYD/DPD is a feature of reversible fibrosis in the zebrafish heart.² Here, we show that the absence of PYD/DPD is a feature of reversible myocardial fibrosis in the mammalian heart as well. Initial analysis suggests that this absence may be due to a lack of lysine hydroxylation on collagen.²