Collagen type XIX regulates cardiac extracellular matrix structure and ventricular function

The cardiac extracellular matrix plays essential roles in homeostasis and injury responses. Although the role of fibrillar collagens have been thoroughly documented, the functions of non-fibrillar collagen members remain underexplored. These include a distinct group of non-fibrillar collagens, termed, fibril-associated collagens with interrupted triple helices (FACITs). Recent reports of collagen type XIX (encoded by Col19a1) expression in adult heart and evidence of its enhanced expression in cardiac ischemia suggest important functions for this FACIT in cardiac ECM structure and function. Here, we examined the cellular source of collagen XIX in the adult murine heart and evaluated its involvement in ECM structure and ventricular function. Immunodetection of collagen XIX in fractionated cardiovascular cell lineages revealed fibroblasts and smooth muscle cells as the primary sources of collagen XIX in the heart. Based on echocardiographic and histologic analyses, Col19a1 null (Col19a1N/N) mice exhibited reduced systolic function, thinning of left ventricular walls, and increased cardiomyocyte cross-sectional areas—without gross changes in myocardial collagen content or basement membrane morphology. Col19a1N/N cardiac fibroblasts had augmented expression of several enzymes involved in the synthesis and stability of fibrillar collagens, including PLOD1 and LOX. Furthermore, second harmonic generation-imaged ECM derived from Col19a1N/N cardiac fibroblasts, and transmission electron micrographs of decellularized hearts from Col19a1N/N null animals, showed marked reductions in fibrillar collagen structural organization. Col19a1N/N mice also displayed enhanced phosphorylation of focal adhesion kinase (FAK), signifying de-repression of the FAK pathway—a critical mediator of cardiomyocyte hypertrophy. Collectively, we show that collagen XIX, which had a heretofore unknown role in the mammalian heart, participates in the regulation of cardiac structure and function—potentially through modulation of ECM fibrillar collagen structural organization. Further, these data suggest that this FACIT may modify ECM superstructure via acting at the level of the fibroblast to regulate their expression of collagen synthetic and stabilization enzymes.


Introduction
The myocardial extracellular matrix (ECM) provides structural context and facilitates force transmission. Yet, despite advancements in our understanding of the ECM in cardiovascular biology, the function of many matrix components remains largely unresolved. The earliest studies of cardiac ECM composition focused on the chief structural components of the interstitial matrix-namely, the fibrillar collagens [1]. In the heart, these consist predominantly of collagen types I and III, which, respectively, form dense, structural, rod-like fibers [2,3] and finer, elastic fiber networks [2,4,5]. Although these components of the cardiac ECM have well-characterized roles in cardiovascular health and disease [6], the importance of their lesser abundant associates, the non-fibrillar collagens, remains obscure.
Non-fibrillar collagens do not form fibrillar bundles but can associate with collagen types I and III fibrils to influence cardiac ECM organization [7,8]. Other studies implicate non-fibrillar collagens and, for some, their bioactive peptide derivatives ("matrikines") as mediators of cardiovascular cell function [9][10][11][12]. Despite major improvements in our understanding of non-fibrillar collagens in cardiac biology, many of the groups that make up this family have remained underexplored. One such group includes the Fibril-Associated Collagens with Interrupted Triple helices (FACITs). The FACITs include eight distinct members, specifically, collagens type IX, XII, XIV, XVI, XIX, XX, XXI, and XXII. Of these, six display transient and/or sustained expression in the mammalian heart, including types IX [13], XII [14], XIV [15], XVI [16], XXI [17], and XXII [18]. Though knowledge of their function in the heart is considerably limited, reports of the heightened expression of a number of FACITs during cardiac development (e. g., types IX [13], XIV [15], XVI [16]), as well as injury (e.g., XII [19], XIV [20], XVI [8], XXI [8]) suggest potentially important biological roles for these low abundance collagens. Though detailed insights regarding the function of FACITs is limited, they may influence cardiomyocyte size, myocardial organization, and ventricular function. For instance, collagen XIV (encoded by Col14a1) null mice exhibit altered production of other collagen types and remodeling enzymes, disorganized ventricular collagen fibrils, diminished ventricular performance, and reduced cardiomyocyte growth [15]. Such insights support the premise that FACITs could be important arbiters of cardiomyocyte function. This also raises questions as to whether other members of this group share redundant or distinct biological roles in such processes.
Though a majority of collagen types that make up the FACIT family have confirmed cardiac localization in the mammalian heart, the distribution of one particular FACIT, type XIX, has remained controversial. For example, an early study surveying the expression patterns of collagen type XIX found that its mRNA was ubiquitously expressed in tissues during mouse embryogenesis (limbs, vertebrae, heart, brain, tail, kidneys, calvaria, lung, muscle, skin, and intestine), but its expression became constrained to only a few non-cardiac tissues in adulthood [21]. Conversely, recent studies have detected collagen XIX in both whole tissue RNA extracts [22] and mesenchyme protein isolates [23] from adult murine hearts. There are no other reports on the biological importance of collagen XIX in the heart, except the recent discovery that it is elevated upon cardiac injury [22]. Perhaps due to its exceptionally scarce abundance in the ECM (≈10 −6 % of dry tissue weight) [24] and limited reports of its detection in adult myocardium, the biological significance of collagen XIX has gone unnoticed. Here, we addressed this gap in knowledge by examining the impact of Col19a1 deletion on cardiac structure and function.

Fibroblasts are a major source of collagen XIX in the adult mammalian heart
Though collagen type XIX is generally recognized as an extra-cardiac collagen that is not abundant in the adult mammalian heart [21], Col19a1 mRNA and its encoded alpha-1 chain protein were detected in separate studies using total RNA from adult mouse hearts [22] and protein extracts derived from isolated cardiac fibroblasts [23]. To gain a more complete view of the potential source of collagen XIX in the murine heart, collagen XIX immunoblots were performed using total protein lysates derived from fractionated cardiac cell types, which included endothelium, smooth muscle, ventricular myocytes, and fibroblasts ( Fig.  1A). Immunoblots showed that collagen XIX is expressed in both cardiac fibroblasts and coronary smooth muscle, but absent in cardiomyocytes and endothelium (Fig. 1A). Anticollagen XIX antibody specificity was confirmed in accompanying immunoblots utilizing Col19a1 N/N cardiac fibroblasts, in which neither Col19a1 mRNA nor protein was produced (Fig. 1B,C). Examination of Col19a1 mRNA and protein expression in wildtype cardiac fibroblasts revealed transcript levels to be of relatively low abundance, with mean cycle threshold (Ct) values of ≈32 in RT-PCR analyses (Fig. 1B). Yet, collagen XIX protein was consistently detected in wildtype cardiac fibroblasts via immunoblot (Fig. 1C). These data suggest that cardiac fibroblasts and smooth muscle cells are the predominant source of collagen XIX in the adult mammalian heart-a predicted observation as both cell types are considered the chief producers of ECM components in cardiac tissue [25].

Col19a1 nullizygous mice display reduced systolic function and ventricular remodeling
To uncover the biological significance and/or potential functional roles for this non-fibrillar collagen in the heart, collagen XIX nullizygous mice (Col19a1 N/N , denoted as "N/N") were examined for evidence of underlying cardiovascular structural and/or functional phenotypes. Initial examination of Col19a1 null animals at 18-20 weeks of age revealed minor decreases in body mass relative to wildtype (Col19a1 +/+ , denoted as "+/+") littermates ( Fig. 2A). Changes in body weight were not accompanied with differences in heart mass to tibia length quotients (Fig. 2B). Blinded echocardiographic analyses were employed to evaluate the consequences of Col19a1 ablation on postnatal ventricular volumes, dimensions, and cardiac performance. Volume analyses showed no difference in left ventricular end diastolic volumes (EDV) between Col19a1 null and littermate controls; however, they did reveal modest, yet statistically significant increases in left ventricular end systolic volumes (ESV) in Col19a1 nullizygous animals ( Fig. 2C-E). Despite said volume differences, cardiac stroke volumes remained comparable between Col19a1 null and wildtype animals (Table 1). Further, as there were no changes in heart rate between Col19a1 null and wildtype animals (Table 1), cardiac output remained comparable ( Table   1). Indices of cardiac function, including fractional shortening (measured from short axis M-mode echocardiograms) and left ventricular ejection fraction (measured from parasternal long axis B-mode echocardiograms), revealed decrements in systolic function in Col19a1 null animals compared to wildtype ( Fig. 2F and Table 1). Col19a1 null animals also showed evidence of mild left ventricular remodeling (Table 1). End diastolic left ventricular internal diameters were similar among groups; however, end systolic internal diameters were larger in Col19a1 null animals, compared to wildtype (Table 1). Accompanying these changes in luminal diameter, Col19a1 null animals displayed statistically significant reductions in both left ventricular anterior and posterior wall thickness in end diastole and systole (Table 1), relative to littermate controls. Together, these data suggest a novel role for collagen XIX in cardiac structure and function.

Cardiomyocyte hypertrophy, but not overt fibrosis, in collagen XIX-deficient mice
Myocardial tissue sections were stained and evaluated for cardiomyocyte cross-sectional area in 18-20-week-old Col19a1 null and wildtype littermate mice (Fig. 3). Automated computational analyses were performed without regional bias on transverse, mid-ventricular tissue sections derived from both wildtype and Col19a1 null mice (Fig. 3A). Analyses revealed that cardiomyocyte cross-sectional areas were significantly greater in Col19a1 null mice compared to controls (Fig. 3B). Further, accompanying histograms displaying the size distribution of the more than 13,000 cardiomyocytes measured in each group revealed a discernible shift in the abundance of smaller cardiomyocytes (<250 μm 2 ) to larger cardiomyocytes (>250 μm 2 ) in Col19a1 null hearts relative to wildtype (Fig. 3C).
This suggests collagen XIX influences cardiomyocyte growth and/or maturation, but the mechanism is unknown. Because this particular FACIT [26] and others [27] may regulate collagen fibril abundance, organization, and stability, alterations in ECM dynamics may be one mechanism by which collagen XIX exerts its biological influence on cardiac growth and structure. To this end, myocardial collagen abundance was evaluated in hearts of 18-20-week-old mice (Fig. 4). Transverse myocardial tissue sections stained with picrosirius red/fast green (Fig. 4A) revealed no apparent differences in myocardial collagen content among collagen XIX-deficient and littermate control animals. Further supporting these observations, accompanying hydroxyproline assays utilizing cardiac tissue homogenates also failed to detect any alterations in total myocardial collagen levels between Col19a1 null and wildtype (Fig. 4B)-excluding bulk fibrosis as a contributor to the cardiovascular phenotypes in Col19a1 null mice.
Because collagen XIX is considered a basement membrane (BM) associated protein [24], transverse myocardial tissue sections were stained for nidogen 1 to ascertain the impact of collagen XIX deletion on myocardial basement membrane morphology ( Supplementary  Fig. S1). Compared to wildtype, Col19a1 null hearts showed no discernable irregularities in basement membrane zone morphology ( Supplementary Fig. S1). In addition, there was no evidence of basement membrane expansion in intracellular regions or alterations in staining intensity ( Supplementary Fig. S1). Overall, these data indicate that the cardiomyocyte hypertrophy observed in Col19a1 null animals is neither connected to perturbations in basement membrane assembly nor interstitial collagen accumulation.

Collagen XIX regulates cardiac fibroblast collagen synthetic and stabilization enzyme expression
Reports that FACITs can participate in the modulation of ECM dynamics [15,26,27] led us to test whether alterations in ECM architecture may underlie structural differences in collagen XIX deficient animals. Because fibroblasts are a source of collagen XIX in the heart ( Fig. 1), we next examined the effects of collagen XIX depletion on the expression of structural collagens (Col1a1, Col3a1, and Col4a1) and ECM remodeling enzymes (MMPs, TIMPs, LOXs, and PLODs) in isolated cardiac fibroblasts (Fig. 5). In agreement with our histological analyses of Col19a1 nullizygous mice, which showed no evidence of myocardial fibrosis, siRNA-mediated depletion of Col19a1 had no effect on mRNA levels of key structural collagens (Col1a1, Col3a1, and Col4a1), relative to non-target-transduced controls (siNT) (Fig. 5A,B). Though collagen transcripts remained unchanged, a number of enzymes responsible for ECM turnover (e.g., Timp2) and the normal assembly and cross-linking of collagen fibrils (i.e., Plod1 and Lox), were differentially expressed in siCol19a1-transduced fibroblasts relative to siNT controls ( Fig. 5C-E). Of these, Plod1, Lox, and Timp2 mRNA levels were significantly elevated in siCol19a1-transduced fibroblasts compared to controls. Congruent with these observations, the expression of PLOD1 and LOX proteins were markedly elevated in both isolated cardiac fibroblasts (Fig. 6A) and in whole hearts (Fig. 6B) of Col19a1 null mice. These results indicate that collagen XIX does not directly impact cardiac mesenchyme collagen production itself, but in fact modulates the expression of key enzymes involved in their synthesis and stabilization-suggesting potential roles for this FACIT in the regulation of fibrillar collagen assembly and structure.
Collagen XIX modulates ECM fibrillar collagen structure/organization PLOD1 catalyzes hydroxylation of lysine residues in collagen alpha chains (critical for the stability of intermolecular crosslinks) [28][29][30][31], and LOX participates in the oxidative deamination of lysine/hydroxylysine residues (important for the formation of covalent pyridinoline cross-links between individual collagen fibers and stabilization of heterotypic fibrils) [32][33][34]. Hence, we hypothesized that collagen XIX may participate in the regulation of cardiac fibroblast collagen assembly/organization. To this end, fibrillar collagen structure/ organization was evaluated in matrices derived from Col19a1 null and wildtype fibroblasts in culture (Fig. 7). Fibroblast-produced matrices were imaged by second harmonic generation (SHG) microscopy and subsequently subjected to quantification. There, imaged collagen fibers were digitally extracted, processed, and quantified ( Fig. 7A-D); providing values pertaining to ECM collagen fiber alignment, fiber width, and fiber length ( Fig. 7E-G). While there were no discernible differences in cell morphology among these populations of fibroblasts ( Supplementary Fig. S2), matrices produced from Col19a1 null cells exhibited pronounced reductions (≈46%) in CurveAlign program-computed alignment coefficients compared to wildtype controls (Fig. 7E). Though alterations in ECM organization were observed, CT-FIRE program dimensional measurements of extracted fibers revealed no changes in collagen fiber widths or lengths among matrices sourced from Col19a1 null and wildtype fibroblasts (Fig. 7F,G). In addition, transmission electron micrographs of decellularized mouse hearts (Fig. 8A) revealed pronounced alterations in the spatial organization and morphology of collagen fibers in Col19a1 null hearts compared to wildtype animals. Specifically, Col19a1 null hearts exhibited significant heterogeneity in both the shape and size of individual collagen fibers (Fig. 8B, top row). Assessment of longitudinalimaged fibers in Col19a1 null hearts showed evidence of abnormal packing and reduced alignment of individual collagen fibrils (Fig. 8B, middle row). In addition to these structural abnormalities, inspection of transverse-imaged fibers uncovered reductions in collagen fibril diameters in null animals relative to wildtype littermates (Fig. 8B, bottom row). These results, tabulated in accompanying histograms incorporating the dimensions of the more than 4000 individually measured collagen fibrils, showed a near 15% decrease in mean fibril diameter in Col19a1 nulls relative to their wildtype counterparts (Fig. 8C). Collectively, these data indicate that collagen XIX has important functions in ECM fibrillar collagen structure/organization in the mammalian heart.

De-repression of the focal adhesion kinase signaling pathway accompanies cardiomyocyte hypertrophic growth in Col19a1 nullizygous animals
The biological underpinnings of the cardiovascular phenotypes manifested (reduced systolic function and increased cardiomyocyte size) in Col19a1 null mice are not immediately clear; however, because ECM-dependent integrin signaling influences cardiomyocyte proliferation and hypertrophy [35], we posited that perturbation in ECM structure could contribute to cardiomyocyte hypertrophy observed in collagen XIX knockout animals. To begin to address this idea, analyses of the activation of canonical downstream integrin receptor pathways (i.e., PI3K/AKT, MEK/ERK, and SRC/FAK) were evaluated in Col19a1 null and wildtype mouse hearts ( Fig. 9). Immunoblots revealed no significant change in the expression levels of phosphorylated AKT (pAKT) or ErK (pERK) among wildtype and mutant hearts (Fig. 9); however, a significant increase in pFAK was observed in Col19a1 null animals ( Fig. 9). Elevated pFAK levels in Col19a1 null animals were not associated with increased myocardial expression of the dominant cardiomyocyte β integrin, specifically integrin β1 ( Supplementary Fig. S3), which comprises the integral β subunit component of the major heterodimeric binding receptors for collagen (α1β1), fibronectin (α5β1), and laminin (α7β1) on cardiomyocytes [36]. These results suggest that the loss of Col19a1 results in de-repression of the focal adhesion kinase pathway-a pathway that is central for the development of load-induced [37] and compensatory [38] cardiomyocyte hypertrophy. While such data implicate enhanced fAk signaling, independent of changes in integrin β1 expression, as a potential molecular pathway underlying this hypertrophic response, primary stimuli driving these phenomena in global Col19a1 nullizygous in adult animals are undefined. Given that the cardiac ECM itself is a critical regulator of cellular processes in both organ development and homeostasis, we believe the contributing mechanisms may also be attributed to alterations in ECM-dependent signaling influencing cardiomyocyte growth and/or survival during early postnatal cardiac development in collagen XIX knockout animals. To address this, neonatal cardiomyocyte proliferation was comparatively assessed in Col19a1 null and wildtype mouse hearts at postnatal day 3 via immunohistochemical detection of phospho-histone H3 (pH3) (Fig. 10). Collagen XIX-deficient neonatal hearts displayed conservation of cardiomyocyte proliferation with both groups exhibiting comparable numbers of pH3 + myocytes per unit area (Fig. 10B) and per myocardial tissue section (Fig. 10C). In addition, myocardial apoptosis was evaluated in neonatal hearts by in situ terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) (Fig. 11). Relative to their wildtype counterparts, neonatal Col19a1 nullizygous hearts exhibited an approximate 6-fold increase in both the number of apoptotic cells per unit area (Fig. 11B) and per myocardial tissue section (Fig. 11C). Collectively, these findings suggest that apoptosis may contribute to compensatory hypertrophic growth of surviving cardiomyocytes in collagen XIX knockout animals.
Overall, these data provide evidence that fibroblast-derived collagen XIX plays a role in the regulation of cardiac structure and function-and that it may do so via modifying ECM collagen fiber organization. These data also suggest that collagen XIX may regulate ECM superstructure via acting in a juxtacrine fashion to regulate the expression of genes involved in collagen synthesis/stability in cardiac fibroblasts.

Discussion
Since the earliest studies of the cardiac extracellular matrix nearly four decades ago [1], the field has made significant progress in both clarifying matrix compositional complexity and uncovering the role of its various fibrillar and non-fibrillar collagen components in homeostasis and disease [6]. Nevertheless, our knowledge concerning the biological function of many of its individual components remains limited. We have developed a long-range program to interrogate the specific involvement of heretofore underexplored regulators of the cardiac ECM, including distinct groups of non-fibrillar collagens, such as the FACIT family members. FACITs play fundamental roles in ensuring the stability of the extracellular matrix and its fibrillar collagen network, through regulating the polymerization and size of fibrillar collagens [39] and directing their structural arrangement [15]. Herein, we addressed the role of collagen XIX in shaping the cardiac ECM and, ultimately, the structure and function of the heart. Through the use of Col19a1 nullizygous animals, we revealed that collagen XIX guides the alignment of the stromal-derived filamentous matrix, which provides structural and spatial context for parenchymal cells within the mammalian heart. Col19a1 null animals exhibited alterations in the expression of collagen synthetic/ stabilization enzymes and disorganization of ventricular collagen fibrils. These findings were accompanied by thinning of the myocardial wall, expansion of cardiomyocyte crosssectional area, and reductions in systolic function. Collectively, such insights support the broader concept that FACITs can impact the structure and function of the heart.
Although not generally recognized to be expressed in adult myocardium, we found that type XIX collagen is produced by fibroblasts and smooth muscle cells, which is consistent with expression profiles indicating that this protein localizes to the basement membrane of other tissues [24]. Our expression analyses largely agree with more recent investigations that have detected collagen XIX in whole heart tissue RNA extracts [22] and cardiac mesenchyme protein isolates [23]. Consistent with collagen XIX being a scarcely abundant collagen, accounting for approximately 10 −6 % of dry tissue weight [24], Col19a1 mRNA transcript and protein levels in isolated cardiac mesenchyme were low-showing late cycle threshold values (average Ct values ≈32) in RT-PCR amplification reactions. Likewise, collagen XIX immunoblots required extended durations of membrane exposure (3-5 min) for adequate visualization. Considering that this FACIT has such low expression levels in cardiac fibroblasts, and that fibroblasts account for less than 25% of the myocardial tissue volume [40], may help to explain why collagen XIX protein is not easily detected using traditional antibody-based approaches. Notwithstanding its low level of expression, targeted reduction of Col19a1 in cardiac fibroblasts increased expression of collagen synthetic/stabilization enzymes (e.g., PLOD1 and LOX). In addition, Col19a1 deletion reduced alignment of fibroblast-produced collagen matrices, suggesting that collagen XIX can function to finetune fibroblast matrix-modifying capacity. These observations were mirrored in the hearts of global Col19a1 N/N mice, wherein collagen synthetic/cross-linking enzymes PLOD1 and LOX were elevated and accompanied by myocardial collagen disorganization and smaller fibrils. Although we did not address the degree to which the elevated expression of PLOD1 and LOX enzymes contribute to the observed changes in cardiac ECM architecture or fibril morphology in Col19a1 nullizygous animals, our finding that PLOD1 expression negatively correlates with fibril diameter is consistent with other studies and the established function of PLOD1 [41].
Our results highlight a potential relationship between FACIT-dependent alterations in matrix geometry and the regulation of myocardial growth and function. And while our data point to collagen XIX as an important regulator in this process, similar roles have been described for another FACIT in the heart. In fact, collagen XIV (Col14a1) knockout mice, too, display altered expression of remodeling enzymes, disorganized ventricular collagen fibrils, and reductions in ventricular function [15]; however, contrary to our Col19a1 N/N mice, which show enhancement in cardiomyocyte size, Col14a1 −/− mice exhibit reductions in cardiomyocyte size in postnatal stages [15]. While we have yet to understand the fundamental mechanisms underlying their distinct impacts on cardiac physiology in vivo knockout models, such findings highlight potentially diverse roles for these FACIT members in the establishment and/or maintenance of the cardiac ECM. These observations also implicate FACIT-mediated modifications in ECM collagen architecture regulate cardiac structure and function. Though the precise mechanism(s) remain to be fully explicated, we deduce cellular integrins as major effectors in these processes. Cellular integrins play essential roles in cell sensing and adaptation to physical constraints imposed by the ECM. Both ECM spatial organization and ECM mechanical properties can influence cellular integrin-mediated signaling [42,43]. Thus, collagen XIX deletion may contribute to the cardiac phenotypes through ECM architectural-dependent alterations in integrin receptor signaling. If true, such changes could impact cardiomyocyte growth and/or survival, which could result from alterations in myocardial mechanical properties and/or modifications in ECM/myocyte integrin receptor engagement. It is known that RGD-and Col-dependent integrins regulate numerous cardiomyocyte biological processes, including, growth [44], force production [45,46], and cellular apoptosis [47,48]. This may help to explain the conspicuous induction of cellular apoptosis in early neonates and the augmented cardiomyocyte hypertrophic growth observed in adult Col19a1 nullizygous animals. As the current study utilized conventional global collagen XIX knockout animals, it is difficult to ascertain the extent to which the cardiovascular phenotypes derive from the impact of collagen XIX deficiency on early cardiac development and homeostasis. Nevertheless, the loss of myocardial cells at early postnatal timepoints (due to increased apoptosis) may contribute indirectly to the development of compensatory myocyte hypertrophy in adult Col19a1 nullizygous animals. Notwithstanding, as the ECM is also a critical regulator of cellular processes in organ homeostasis, we surmise that deficits in cardiac function and alterations in myocyte growth also be credited to perturbations in ECM-dependent signaling influencing cardiomyocyte growth and/or function in adult animals.
Beyond ECM-dependent mechanisms, collagen XIX-derived matrikines themselves exert biological activity. In cancer cells [49] and neurons [50], the non-collagenous C-terminal domain (NC1) of collagen XIX (cleaved by plasmin) can modulate cell responses via binding and signaling through RGD-dependent integrins (including αvβ3 and α5β1). Interestingly, RGD integrin signaling regulates cardiac fibroblast activation [51], cardiomyocyte growth [44], and force production [45,46]. Hence, we speculate that the mechanisms by which collagen XIX regulates cardiac function are manifold and may include both changes in ECM architecture/biomechanics (reducing their engagement with the ECM) and NC1 matrikine-mediated antagonism of cardiomyocyte integrin receptor activity. In light of the extensive evidence showing ECM/integrin receptor signaling as important mediators of cardiomyocyte biology [35,38,45], it is plausible that NC1 matrikine may directly regulate cardiomyocyte growth and/or function through its direct interactions with cardiomyocyte integrins in adult animals. FAK, a major kinase involved in the integrin intracellular signaling cascade, is activated by various growth factors and by ligation of all β1-, β3-, or β5-containing integrins [52]. The finding that hearts sourced from Col19a1 null animals display de-repression of FAK signaling supports the notion that collagen XIX may inhibit postnatal cardiomyocyte growth by antagonizing myocyte integrin receptors and/or decreasing ECM/integrin receptor engagement.
Our study employed multiple design elements to promote rigor, demonstrate transparency, and enhance reproducibility. Mice were studied in a blinded manner using a priori application of ear tags, which were decoded only after the data were analyzed. No mice were excluded; all mice were reported herein; statistical design was rationalized before the experiments. Yet in terms of limitations, there were a few. To test the concept that FACITs can influence cardiac structure and function, we used a global Col19a1 deficient mouse. There is no human condition related to global deficiency of Col19a1. Thus, we avoided making unreasonable translational claims. Most of our conclusions presumed fibroblasts were the sole/primary source of Col19a1 expression; however, given that smooth muscle cells also produce collagen XIX, it may also regulate vascular structure, and by extension, function. It is thus also possible that the developmental impact of constitutive Col19a1 deficiency differed from (or was additive to) its role in the adult heart. Both possibilities will be addressed in future studies in which we use inducible, fibroblast-specific promoters.
In summary, we showed that collagen XIX impacts the structure and function of the heart. In addition, we showed that collagen XIX guides alignment, but not abundance, of fibrillar collagen. Such insights have important implications for cardiac development and the potential role of collagen XIX during acute, active tissue remodeling. In future studies, we will explore the role of collagen XIX during infarct-induced heart failure and test its role in reversible models of remodeling such as pregnancy-induced or exercise-elicited cardiac growth.

Experimental procedures
All animal experiments were performed in accordance with the Guide for the Care and Use of Laboratory Animals published by the U.S. National Institutes of Health (Revised 2020) and were approved by the University of Louisville Institutional Animal Care and Use Committee (Louisville, KY, USA).

Animals
Collagen XIX null mice (Col19a1 N/N ) on a C57BL/6 J background were kindly provided by Dr. Michael Fox (Fralin Biomedical Research Institute at Virginia Tech) and are previously described [50]. These mice harbor a Col19a1 null mutation (denoted "N19") that was generated by deleting exon 4 of collagen XIX [26]. All experiments in the study utilized collagen XIX nullizygous (Col19a1 N/N ) and wildtype littermate control (Col19a1 +/+ ) mice at 18-20 weeks of age. Mice were housed in a pathogen-free facility under a standard 12 h light/dark cycle with ad libitum access to food and water. Genomic DNA was isolated from tails using an EZ Tissue/Tail PCR Genotyping Kit (EZ BioResearch LLC) according to the manufacturer instructions and genotyping was performed as previously described [50].

Rigor, transparency, and reproducibility
A total of 20 mice were required for completion of in vivo experiments. This number was calculated via implementation of a statistical power analysis using SigmaStat Software (SigmaStat, Systat Software, Inc.). Based on preliminary experiments comparing cardiac function among Col19a1 null and wildtype mice, we determined a 10.7% minimum detectable difference in means. Further, with these groups, a desired power of 80%, and an alpha of 0.05 (95th percentile), we calculated a requisite number of no fewer than 7 animals per group (Col19a1 +/+ and Col19a1 N/N ). No animals were excluded from the study.
All experiments were performed by a blinded observer-unapprised as to animal genotype/ group assignment during outcome assessment and/or data analysis.

Echocardiographic assessment of cardiac function
Col19a1 N/N and Col19a1 +/+ littermates (both sexes) were subjected to echocardiography at 18-20 weeks of age. Transthoracic echocardiography of the left ventricle was performed as previously described [53]. The sonographer was blinded to mouse genotype at the time of assessment. Transthoracic echocardiography was performed using a Vevo 3100 echocardiography system. Mouse body temperature was maintained at 36.5-37.5 °C using a rectal thermometer interfaced with a servo-controlled heat lamp. Mice were anesthetized with 2% isoflurane and subsequently maintained under light anesthesia (≈1.5% isoflurane) during echocardiography procedures. Using the Vevo rail system, mice were placed in a supine position on an examination board interfaced with the Vevo 3100. M-mode, short-axis views of the left ventricle at the mid-papillary level allowed determination of left ventricular fractional shortening (FS), which was calculated by:

Cells, cell culture, and treatment
All cell lines were maintained under standard incubation conditions at 37 °C with 5% atmospheric CO 2 and passaged using TrypLE™ (ThermoFisher Scientific) at 70-80% confluence.

Cardiovascular cell isolation/procurement
Cardiac fibroblast isolation: cardiac fibroblasts were isolated from both male and female adult C57BL/6 J mice (18-20 wk) according to a previously established protocol with minor modification [23]. In brief, whole hearts were excised from isoflurane anesthetized mice, transferred to sterile 10 cm dishes, and mechanically minced using a razor blade. Minced myocardial tissues were transferred to a sterile 50 mL conical tube and washed via gravity sedimentation in 1 × PBS to remove red blood cells. Subsequently, tissues were enzymatically digested in 10 mL of filter sterilized (0.22 μm filter) collagenase II (1600 U/mL; Worthington Hills) at 37 °C for 25 min, with gentle agitation every 5 min. Enzymatic digestion was halted with the addition of 40 mL of mouse fibroblast complete medium [DMEM/F12, 10% FBS (VWR Life Science Seradigm Fetal Bovine Serum), 20 ng/mL bFGF (Peprotech), 20 ng/mL EGF (Peprotech), 1 × ITS (Thermo Fisher Scientific), 100 U/mL penicillin-streptomycin (Thermo Fisher Scientific), and 1 × GlutaMax (Thermo Fisher Scientific)]. Cells were then harvested by centrifugation at 600 × g for 10 min and then washed two times in complete medium. Lastly, cell pellets were gently resuspended in mouse fibroblast complete medium and grown in culture for 3-5 days prior to use.

Immunoblotting
Immunoblotting was performed according to a previously described protocol [55] with minor modification of total protein isolation procedures. For the isolation of total protein extracts from cardiac tissue, flash-frozen myocardial specimens were suspended in RIPA lysis buffer (Thermo Fisher Scientific, Cat. 89,900) containing 1 × Halt™ Protease and Phosphatase Inhibitor Cocktail (Thermo Fisher Scientific, Cat. 78,444) at a ratio of 40 μL per mg of tissue. Tissue segments were then mechanically dissociated on ice using a handheld, electric tissue homogenizer followed with 10 s of sonication to facilitate dissociation of finer tissue aggregates. Samples were then incubated on ice for 30 min with periodic agitation and finally centrifuged at 14,000 × g for 15 min at 4 ° C to remove insoluble material. To prepare total pro-ein extracts from fibroblasts in culture, cells were trypsinized, washed twice with ice-cold 1 × PBS, and collected via centrifugation (500 × g for 5 min at 4 °C). Resulting cell pellets were suspended in 80-100 μL of RIPA lysis buffer containing protease and phosphatase inhibitors (as above) and mechanically lysed via repeated trituration on ice. Samples were then placed on ice for 30 min with intermittent agitation and then centrifuged at 14,000 × g for 15 min at 4 °C to remove insoluble cell debris. Resultant total protein lysates (from myocardial tissues and cells in culture) were analyzed for total protein content using a Pierce™ BCA Protein Assay Kit (Thermo Fisher Scientific; Cat. 23,225), according to the manufacturer's instructions. Protein lysates were electrophoretically resolved on 4-12% Bis-Tris polyacrylamide gels (Thermo Fisher Scientific; Cat. NP0335), transferred to PVDF membranes (Thermo Fisher Scientific; Cat. LC2005), and blocked in 5% milk 1 × TBS-T overnight. Membranes were then rinsed with 1 × TBS-T, incubated with primary antibodies in antibody dilution buffer (5% BSA, 1 × TBS-T) at 4 °C, and lastly washed 5 times with 1 × TBS-T for 5 min each. Afterward, membranes were then incubated with HRP-conjugated secondary antibodies in 5% milk, 1 × TBS-T at room temperature and again washed 5 times with 1 × TBS-T for 5 min each. A detailed list of antibodies with corresponding dilutions is available in Supporting Information Table S1. HRP was detected using SuperSignal® West Pico Chemiluminescent Substrate (Thermo Fisher Scientific; Cat. 34,580) and digitally imaged using a Thermo Fisher Scientific myECL Imager. Images were subject to densitometric quantification using the ImageJ (version 1.53n) [56].

Quantitative real-time PCR
Gene expression assays were performed as previously [55] with minor modification. Total RNA was isolated using a PureLink® RNA Mini Kit (Thermo Fisher Scientific) per the manufacturer's instructions. 1000 ng of total RNA was reverse transcribed using the SuperScript™ IV VILO™ First-Strand Synthesis System (Thermo Fisher Scientific) containing both random hexamer and oligo(dT)18 primers, according to the manufacturer's protocol. Quantitative PCR was performed with PerfeCTa® SYBR® Green FastMIX® in a QuantStudio ® 5 real-time PCR system (Thermo Fisher Scientific). Transcript specific primers and respective annealing temperatures are listed in Supporting Information Table  S2. PCR was performed with an initial denaturation step of 95 °C for 10 min, followed by 40 cycles of 15 s denaturation at 95 °C and 1 min annealing/extension at 56-60 °C. Relative expression/fold expression was calculated according to the ΔΔC T method for quantitative real-time PCR using Actb or Gapdh as an internal reference control where indicated.

General histology and morphometry procedures
Hearts of isoflurane anesthetized mice were accessed by a thoracotomy and arrested in diastole via injection of 1 mL potassium chloride solution (3 M KCl in 1 × PBS) into the apex. Hearts were then removed and cannulated through the aorta using a Langendorff apparatus and retrogradely perfused with 1 × PBS for 3 min to clear blood within the coronary circulation, and perfused with 10% neutral buffered formalin solution for 10-15 min. Perfusion pressures were maintained between 60 and 80 mmHg. Following perfusion-fixation, total heart weight was measured on an analytical balance and tibia length determined using precision calipers. Heart weight (mg) to tibia length (mm) ratios were calculated and reported. Each heart was then cut into 3 transverse slices (each 3 mm thick), processed, paraffin-embedded, sectioned at 4 μm intervals, and subsequently utilized in histological and immunohistochemical analyses (detailed below).

Determination of cardiomyocyte cross-sectional area
Mid-papillary level transverse myocardial sections were heated at 80 °C for 30 min, deparaffinized with xylene, and stepwise rehydrated by successive rinsing in decreasing concentrations of ethanol (100, 96, 90, and 80%) prior to placing in deionized water. Antigen retrieval was then performed by submerging tissue sections in citrate retrieval buffer [2.4 g/L sodium citrate tribasic dehydrate (S4641, Sigma-Aldrich), 0.35 g/L citric acid (C0759, Sigma-Aldrich), pH 6.0] for 10 min at approximately 100 °C. Sections were allowed to cool for 10 min on ice and washed in deionized water prior to application of diluted antibodies and wheat germ agglutinin (WGA) (detailed in Supporting Information Table S3). Slides were incubated with α-sarcomeric actin primary antibody in antibody diluent reagent ( Entire myocardial sections were imaged, without regional bias, using a Keyence BZ-X810 microscope with 49,000-UF1 C195675 (DAPI), 49,008_UF1 C195528 (Texas Red), 49,009-UF1 C195677 (Cy5) filter cubes. Cardiomyocyte cross-sectional areas were calculated in an automated fashion using the Keyence BZ-X800 Analyzer Software with the following conditions: nucleated cardiomyocytes (DAPI + ), circularity ≥0.50, and area range of 50-600 nm.

Nidogen 1 basement membrane staining
As above, transverse myocardial tissue sections heated at 80 °C, deparaffinized with xylene, and stepwise rehydrated. Samples were then permeabilized with 0.2% Triton X-100 in 1 × PBS for 15 min, rinsed with deionized water for 3 min, and subjected to citrate buffer retrieval procedures (as detailed above). Sections were cooled on ice for 10 min, rinsed in deionized water, and blocked with 2% donkey serum (Immuno Reagents Inc., Cat. SP-072-VX10) for 2 h at room temperature. Subsequently, slides were washed once with 1 × PBS and then incubated with diluted nidogen 1 antibody in diluent reagent (Thermo Fisher Scientific, Cat. 003,218) overnight at 4 °C. The following day, slides were washed with 1 × PBS (three times) and incubated with Alexa Fluor 488 conjugated secondary antibody in antibody diluent reagent for 1 h at 37 °C. After three washes with 1 × PBS, slides were incubated with α-sarcomeric actin antibody for 1 h at 37 °C, washed three times with 1 × PBS, and finally incubated with Alexa fluor 633 secondary antibody in antibody diluent reagent for 1 h at 37 °C. Sections were then counterstained with 1 μg/mL DAPI (D3571, Thermo Fisher Scientific) for 10 min and incubated with 0.1% Sudan black solution (Sigma-Aldrich, Cat. 199,664) in 70% ethanol for 15 min to reduce autofluorescence signal. Sections were washed with 1 × PBS (three times), rinsed in deionized water, and mounted under glass coverslips using PermaFluor Aqueous Mounting Medium (ThermoFisher Scientific, Cat. TA-030-FM). Whole myocardial tissue sections were imaged using a Keyence BZ-X810 microscope with 49,000-UF1 C195675 (DAPI), 49,002-UF1 C195676(GFP), 49,009-UF1 C195677 (Cy5) filter cubes. All antibody information and dilutions used are detailed in Supporting Information Table S3.

Neonatal mouse cardiomyocyte proliferation and myocardial apoptosis
Isoflurane anesthetized mouse pups at postnatal day 3 were decapitated, their hearts quickly removed, and placed in 10% neutral buffered formalin solution overnight. The following day neonatal hearts were processed via an automated tissue processor, paraffin-embedded, and sectioned at 4 μm intervals. Coronal plane cardiac tissue sections were deparaffinized, rehydrated, permeabilized, and subjected to antigen retrieval procedures (as described above). Slides were then blocked with 2% donkey serum (Immuno Reagents Inc., Cat. SP-072-VX10) for 2 h at room temperature and rinsed with 1 × PBS prior to application of diluted antibodies (detailed in Supporting Information Table S3). For the assessment of cardiomyocyte proliferation, slides were incubated with anti-phospho-histone H3 antibody, diluted in antibody diluent reagent (Thermo Fisher Scientific, Cat. 003,218), overnight at 4 °C. Sections were then washed three times with 1 × PBS, incubated with Alexa Fluor 488 secondary antibody in diluent reagent for 1 h at 37 °C, and again washed three times with 1 × PBS. Slides were subsequently incubated with anti-α-sarcomeric actin antibody for 1 h at 37 °C, washed three times with 1 × PBS, and incubated with Alexa Fluor 633 secondary antibody in antibody diluent reagent for 1 h at 37 °C. After three washes with 1 × PBS, slides were counterstained with 1 μg/mL DAPI (D3571, Thermo Fisher Scientific) for 10 min and then incubated with 0.1% Sudan Black solution (Sigma-Aldrich, Cat. 199,664) in 70% ethanol for 15 min. Sections were then washed with 1 × PBS (three times), rinsed in deionized water, and mounted under glass coverslips using PermaFluor Aqueous Mounting Medium (Thermo Fisher Scientific, Cat. TA-030-FM). Entire myocardial sections were imaged, without regional bias, using a Keyence BZ-X810 microscope with 49,000-UF1 C195675 (DAPI), 49,002_UF1 C195676 (GFP), 49,009-UF1 C195677 (Cy5) filter cubes. For the evaluation of myocardial apoptosis, coronal tissue sections were subjected to TUNEL (terminal deoxynucleotidyl transferase dUTP nick-end labeling). Tissue sections were stained using the Click-iT™ TUNEL Assay for In Situ Apoptosis Detection kit, Alexa Fluor 488 dye (Thermo Fisher Scientific, Cat. C10617), with strict adherence to the manufacturer's protocol. Resultant sections were imaged, without regional bias, using a Keyence BZ-X810 microscope with 49,000-UF1 C195675 (DAPI) and 49,002_UF1 C195676 (GFP) filter cubes. The number of pH3 + or TUNEL + cells were enumerated in ImageJ (version 1.53n) [56] by a blinded observer unaware of genotype.

Myocardial hydroxyproline content
Myocardial collagen content was enumerated using a colorimetric hydroxyproline assay kit (Abcam, Cat. Ab222941) according to the manufacturer's instructions. Briefly, 100 mg of isolated heart tissue was homogenized in 74 μl of distilled water using a 1 mL Dounce tissue grinder (Sigma-Aldrich, Cat. DWK885300). Next, 100 μL of 10 N NaOH was added to homogenates, heated for 1 h at 120 °C, and neutralized via the addition of 100 μL of 10 N HCl. Samples were then centrifuged at 10,000 × g to remove insoluble material. Resultant supernatants were placed into 96 well plates (10 μL per well) and crystallized by heating at 65 °C. Subsequently, 100 μL of oxidation solution (consisting of 96 μL oxidation buffer and 4 μL chloramine T concentrate) was added to each well and incubated at room temperature for 20 min. Then 50 μL of developer solution was added to each well and placed at 37 °C for 5 min. Afterward, 50 μL of 4-(dimethylamino)benzaldehyde (DMAB) solution was pipetted per well and incubated at 65 °C for 45 min. Sample absorbance values were measured at 560 nm on a SynergyMx plate reader (BioTek) and compared to kit provided standards to calculate total myocardial hydroxyproline content in μg per mg of wet tissue.

In vitro collagen matrix organization
Approximately 300,000 cardiac fibroblasts were plated in 10 cm tissue culture grade dishes and grown for ten days in fibroblast complete medium. Cells were then counterstained via the addition of Alexa Fluor 555 conjugated WGA diluted in 1 × PBS (1 μL/mL). After a 5 min incubation, plates were rinsed with 1 × PBS, and subjected to second harmonic generation (SHG) imaging using a Nikon A1R MP+ multiphoton microscope with an Apo LWD 25 × immersion objective. The excitation laser was tuned to 950 nm and SHG signal was collected through a FITC bandpass filter with an emission wavelength of 525 nm. Square images were acquired with 1024 × 1024 pixels of resolution with a scan speed of approximately 15 s. Resultant SHG images were subjected to computational analyses using MATLAB software framework packages, CurveAlign and CT-FIRE [57].

Transmission electron microscopy of decellularized hearts
Hearts of isoflurane anesthetized mice were accessed by a thoracotomy. Hearts were then removed, aortas positioned onto a cannula attached to a peristaltic pump (Grainger, Cat. 20LX53), and affixed via a suture loop. Hearts were retrogradely perfused with 1 × PBS for 10 min followed by decellularization solution A (1% SDS, 1 × PBS) for 24 h.

Supplementary Material
Refer to Web version on PubMed Central for supplementary material.    A Immunoblots quantifying collagen XIX protein expression in wildtype cardiac fibroblasts transfected with Col19a1 siRNA (siCol19a1) or non-target control (siNT). β-tubulin is a loading control. Graphs report mean collagen XIX protein expression (relative to βtubulin loading controls) ± SEM. Immunoblot markers denote molecular weight standards (in kDa) resolved with experimental protein lysates. Gene expression assays employing transcript-specific primers targeting B collagens (Col1a1, Col3a1, and Col4a1), C matrix metalloproteinases (Mmp2, Mmp3, Mmp13), D tissue inhibitors of metalloproteinases (Timp1, Timp2, Timp3), and E collagen synthetic (Plod1, Plod2, P4ha1)/stabilization enzymes (Lox, Loxl1, Loxl2). Graphs report mean mRNA expression (relative to siNT controls) ± SEM. Gapdh was used as an internal reference control. All experiments were performed using three biological replicates (N = 3 independent cardiac fibroblast isolations [2 female, 1 male]), with each cell line performed in triplicate. Statistics: Data normality was assessed using the Shapiro-Wilk test. Data were analyzed by an unpaired, two-tailed Student's T-test. p < 0.05 was considered statistically significant. Amido black is a loading control. Immunoblot markers denote molecular weight standards (in kDa) resolved with experimental protein lysates. Graphs report mean protein expression (relative to loading control) ± SEM. Statistics: Data normality was assessed using the Shapiro-Wilk test. Data were analyzed by an unpaired, two-tailed Student's T-test. p < 0.05 was considered statistically significant. Cardiac fibroblasts isolated from Col19a1 wildtype (+/+) and null (N/N) mice were grown to confluence for 7 d A Resultant collagen matrices were imaged via second harmonic generation microscopy (SHG; Nikon A1R MP+ multiphoton). Fibrillar collagen alignment was quantified using CurveAlign/CT-FIRE. B Image generated by CurveAlign showing the location (red dots) and orientation (green lines) of extracted fibers. C Heatmap indicating regional fiber alignment (red denotes areas of greater alignment). D CT-FIRE extracted fibers overlaid on master image (A). Information for entire fibers are shown and highlighted by different colored lines for contrast. Graphs illustrate mean fibrillar collagen E alignment, F width, and G length ± SEM in wildtype (+/+; N = 3 independent fibroblast isolations A Representative immunoblots assessing the expression of phosphorylated and total AKT, ERK, and FAK proteins in myocardial tissues sourced from Col19a1 wildtype (+/+) and null (N/N) mice at 18-20 wk. β-tubulin is a loading control. Immunoblot markers denote molecular weight standards (in kDa) resolved with experimental protein lysates. Those with asterisks (*) indicate predicted molecular weights of visualized protein targets. B Graphs report total kinase levels relative to β-tubulin ± SEM (left column) and total phosphorylated kinase levels relative to total kinase protein ± SEM (right column). N = 6 independent biological replicates per group [3 female, 3 male]. Statistics: Data normality was assessed using the Shapiro-Wilk test. pERK/ERK immunoblots data were analyzed by an unpaired, two-tailed Student's T-test. pAKT/AKT and pFAK/FAK data were analyzed using the nonparametric Mann-Whitney U test. p < 0.05 was considered statistically significant. A Coronal myocardial tissue sections from Col19a1 wildtype (+/+) and null (N/N) mice at postnatal day 3. Sections were stained with antibodies against phospho-histone H3 (pH3; proliferation marker; green) and α-sarcomeric actin (αSA; myocytes; red). Nuclei were counterstained with DAPI (nuclei; blue). Sections incubated with only secondary antibodies [2° Ab (ctrl)], in the absence of anti-pH3 antibody, were included as negative staining controls. Graphs depict the mean number of pH3 + cells per B unit area (mm 2 ) ± SEM or C tissue section ± SEM in wildtype (+/+; N = 4) and null (N/N; N = 4) neonates. Statistics:

Funding sources
Data normality was assessed using the Shapiro-Wilk test. Data in B was analyzed by an