Search

Arterial hypertension can lead to structural changes within the heart including left ventricular hypertrophy (LVH) and eventually heart failure with preserved ejection fraction (HFpEF). The initial diagnosis of HFpEF is costly and generally based on later stage remodeling; thus, improved predictive diagnostic tools offer potential clinical benefit. Recent work has shown predictive value of multibiomarker plasma panels for the classification of patients with LVH and HFpEF. We hypothesized that machine learning algorithms could substantially improve the predictive value of circulating plasma biomarkers by leveraging more sophisticated statistical approaches. In this work, we developed an ensemble classification algorithm for the diagnosis of HFpEF within a population of 480 individuals including patients with HFpEF, patients with LVH, and referent control patients. Algorithms showed strong diagnostic performance with receiver-operating-characteristic curve (ROC) areas of 0.92 for identifying patients with LVH and 0.90 for identifying patients with HFpEF using demographic information, plasma biomarkers related to extracellular matrix remodeling, and echocardiogram data. More impressively, the ensemble algorithm produced an ROC area of 0.88 for HFpEF diagnosis using only demographic and plasma panel data. Our findings demonstrate that machine learning-based classification algorithms show promise as a noninvasive diagnostic tool for HFpEF, while also suggesting priority biomarkers for future mechanistic studies to elucidate more specific regulatory roles. NEW & NOTEWORTHY Machine learning algorithms correctly classified patients with heart failure with preserved ejection fraction with over 90% area under receiver-operating-characteristic curves. Classifications using multidomain features (demographics and circulating biomarkers and echo-based ventricle metrics) proved more accurate than previous studies using single-domain features alone. Excitingly, HFpEF diagnoses were generally accurate even without echo-based measurements, demonstrating that such algorithms could provide an early screening tool using blood-based measurements before sophisticated imaging.

Left ventricular pressure overload (LVPO) can develop from antecedent diseases such as aortic valve stenosis and systemic hypertension and is characterized by accumulation of myocardial extracellular matrix (ECM). Evidence from patient and animal models supports limited reductions in ECM following alleviation of PO, however, mechanisms that control the extent and timing of ECM regression are undefined. LVPO, induced by 4 wk of transverse aortic constriction (TAC) in mice, was alleviated by removal of the band (unTAC). Cardiomyocyte cross-sectional area, collagen volume fraction (CVF), myocardial stiffness, and collagen degradation were assessed for: control, 2-wk TAC, 4-wk TAC, 4-wk TAC + 2-wk unTAC, 4-wk TAC + 4-wk unTAC, and 4-wk TAC + 6-wk unTAC. When compared with 4-wk TAC, 2-wk unTAC resulted in increased reactivity of collagen hybridizing peptide (CHP) (representing initiation of collagen degradation), increased levels of collagenases and gelatinases, decreased levels of collagen cross-linking enzymes, but no change in CVF. When compared with 2-wk unTAC, 4-wk unTAC demonstrated decreased CVF, which did not decline to control values. At 4-wk and 6-wk unTAC, CHP reactivity and mediators of ECM degradation were reduced versus 2-wk unTAC, whereas levels of tissue inhibitor of metalloproteinase (TIMP)-1 increased. ECM homeostasis changed in a time-dependent manner after removal of LVPO and is characterized by early increases in collagen degradation, followed by a later dampening of this process. Tempered ECM degradation with time is predicted to contribute to the finding that normalization of hemodynamic overload alone does not completely regress myocardial fibrosis. NEW & NOTEWORTHY In this study, a murine model demonstrated persistent interstitial fibrosis and myocardial stiffness following alleviation of pressure overload.

In human heart failure and in murine hearts with left-ventricular pressure overload (LVPO), increases in fibrosis are associated with increases in myocardial stiffness. Secreted protein acidic and rich in cysteine (SPARC) is shown to be necessary for both cardiac fibrosis and increases in myocardial stiffness in response to LVPO; however, cellular sources of cardiac SPARC are incompletely defined. Irradiation and bone marrow transfer were undertaken to test the hypothesis that SPARC expression by bone marrow-derived cells is an important mediator of fibrosis in LVPO. In recipient SPARC-null mice transplanted with donor wild-type (WT) bone marrow and subjected to LVPO, levels of fibrosis similar to that of WT mice were found despite the lack of SPARC expression by resident cells. In recipient WT mice with donor SPARC-null bone marrow, significantly less fibrosis versus that of WT mice was found despite the expression of SPARC by resident cells. Increases in myocardial stiffness followed a similar pattern to that of collagen deposition. Myocardial macrophages were significantly reduced in SPARC-null mice with LVPO versus that of WT mice. Recipient SPARC-null mice transplanted with donor WT bone marrow exhibited an increase in cardiac macrophages versus that of SPARC-null LVPO and donor WT mice with recipient SPARC-null bone marrow. Expression of vascular cellular adhesion molecule (VCAM), a previously identified binding partner of SPARC, was assessed in all groups and with the exception of WT mice, increases in VCAM immunoreactivity with LVPO were observed. However, no differences in VCAM expression between bone marrow transplant groups were noted. In conclusion, SPARC expression by bone marrow-derived cells was critical for fibrotic deposition of collagen and influenced the expansion of myocardial macrophages in response to LVPO. NEW & NOTEWORTHY Myocardial fibrosis and the resultant increases in LV and myocardial stiffness represent pivotal consequences of chronic pressure overload (PO). In this study, a murine model of cardiac fibrosis induced by PO was used to demonstrate a critical function of SPARC in bone marrow-derived cells that drives cardiac fibrosis and increases in cardiac macrophages.

Chronic iron overload results in heart and liver diseases and is a common cause of morbidity and mortality in patients with genetic hemochromatosis and secondary iron overload. We investigated the role of tissue inhibitor of metalloproteinase 3 (TIMP3) in iron overload-mediated tissue injury by subjecting male mice lacking Timp3 (Timp3(−/−)) and wild-type (WT) mice to 12 wk of chronic iron overload. Whereas WT mice with iron overload developed diastolic dysfunction, iron-overloaded Timp3(−/−) mice showed worsened cardiac dysfunction coupled with systolic dysfunction. In the heart, loss of Timp3 was associated with increased myocardial fibrosis, greater Timp1, matrix metalloproteinase (Mmp)2, and Mmp9 expression, increased active MMP-2 levels, and gelatinase activity. Iron overload in Timp3(−/−) mice showed twofold higher iron accumulation in the liver compared with WT mice because of constituently lower levels of ferroportin. Loss of Timp3 enhanced the hepatic inflammatory response to iron overload, leading to greater neutrophil and macrophage infiltration and increased hepatic fibrosis. Expression of inflammation-related MMPs (MMP-12 and MMP-13) and inflammatory cytokines (IL-1β and monocyte chemoattractant protein-1) was elevated to a greater extent in iron-overloaded Timp3(−/−) livers. Gelatin zymography demonstrated equivalent increases in MMP-2 and MMP-9 levels in WT and Timp3(−/−) iron-overloaded livers. Loss of Timp3 enhanced the susceptibility to iron overload-mediated heart and liver injury, suggesting that Timp3 is a key protective molecule against iron-mediated pathology. NEW & NOTEWORTHY In mice, loss of tissue inhibitor of metalloproteinase 3 (Timp3) was associated with systolic and diastolic dysfunctions, twofold higher hepatic iron accumulation (attributable to constituently lower levels of ferroportin), and increased hepatic inflammation. Loss of Timp3 enhanced the susceptibility to iron overload-mediated injury, suggesting that Timp3 plays a key protective role against iron-mediated pathology.

To investigate the role of secreted protein acidic and rich in cysteine (SPARC) in age-related cardiac inflammation, we studied six groups of mice: young (3–5 mo old), middle-aged (10–12 mo old), and old (18–29 mo old) C57BL/6 wild-type (WT) and SPARC-null (Null) mice ( n = 7–10/group). Cardiac function and structure were determined by echocardiography. The left ventricle was used for cytokine gene array and macrophage quantification by immunohistochemistry. Macrophage infiltration increased with age in WT ( n = 5–6/group, P < 0.05 for young vs. old), but not in Null. Proinflammatory markers ( Ccl5, Cx3cl1, Ccr2, and Cxcr3) increased in middle-aged and old WT, whereas they were increased only in old Null compared with respective young ( n = 5–6/group, P < 0.05 for all). These results suggest that SPARC deletion delayed age-related cardiac inflammation. To further assess how SPARC affects inflammation, we stimulated peritoneal macrophages with SPARC ( n = 4). SPARC treatment increased expression of proinflammatory macrophage M1 markers and decreased anti-inflammatory M2 markers. Echocardiography ( n = 7–10/group) revealed an age-related increase in wall thickness of the left ventricle in WT (0.76 ± 0.02 mm in young vs. 0.91 ± 0.03 mm in old; P < 0.05) but not in Null (0.78 ± 0.01 mm in young vs. 0.84 ± 0.02 mm in old). In conclusion, SPARC deletion delayed age-related increases in macrophage infiltration and proinflammatory cytokine expression in vivo and in vitro. SPARC acts as an important mediator of age-related cardiac inflammation by increasing the expression of macrophage M1 markers and decreasing M2 markers.

While the term “fibrosis” can be misleading in terms of the complex patterns and processes of myocardial extracellular matrix (ECM) remodeling, fibrillar collagen accumulation is a common consequence of relevant pathophysiological stimuli, such as pressure overload (PO) and myocardial infarction (MI). Fibrillar collagen accumulation in both PO and MI is predicated on a number of diverse cellular and extracellular events, which include changes in fibroblast phenotype (transdifferentiation), posttranslational processing and assembly, and finally, degradation. The expansion of a population of transformed fibroblasts/myofibroblasts is a significant cellular event with respect to ECM remodeling in both PO and MI. The concept that this cellular expansion within the myocardial ECM may be due, at least in part, to endothelial-mesenchymal transformation and thereby not dissimilar to events observed in cancer progression holds intriguing future possibilities. Studies regarding determinants of procollagen processing, such as procollagen C-endopeptidase enhancer (PCOLCE), and collagen assembly, such as the secreted protein acidic and rich in cysteine (SPARC), have identified potential new targets for modifying the fibrotic response in both PO and MI. Finally, the transmembrane matrix metalloproteinases, such as MMP-14, underscore the diversity and complexity of this ECM proteolytic family as this protease can degrade the ECM as well as induce a profibrotic response. The growing recognition that the myocardial ECM is a dynamic entity containing a diversity of matricellular and nonstructural proteins as well as proteases and that the fibrillar collagens can change in structure and content in a rapid temporal fashion has opened up new avenues for modulating what was once considered an irreversible event - myocardial fibrosis.

Myocardial fibrillar collagen is considered an important determinant of increased ventricular stiffness in pressure-overload (PO)-induced cardiac hypertrophy. Chronic PO was created in feline right ventricles (RV) by pulmonary artery banding (PAB) to define the time course of changes in fibrillar collagen content after PO using a nonrodent model and to determine whether this time course was dependent on changes in fibroblast function. Total, soluble, and insoluble collagen (hydroxyproline), collagen volume fraction (CVF), and RV end-diastolic pressure were assessed 2 days and 1, 2, 4, and 10 wk following PAB. Fibroblast function was assessed by quantitating the product of postsynthetic processing, insoluble collagen, and levels of SPARC (secreted protein acidic and rich in cysteine), a protein that affects procollagen processing. RV hypertrophic growth was complete 2 wk after PAB. Changes in RV collagen content did not follow the same time course. Two weeks after PAB, there were elevations in total collagen (control RV: 8.84 ± 1.03 mg/g vs. 2-wk PAB: 11.50 ± 0.78 mg/g); however, increased insoluble fibrillar collagen, as measured by CVF, was not detected until 4 wk after PAB (control RV CVF: 1.39 ± 0.25% vs. 4-wk PAB: 4.18 ± 0.87%). RV end-diastolic pressure was unchanged at 2 wk, but increased until 4 wk after PAB. RV fibroblasts isolated after 2-wk PAB had no changes in either insoluble collagen or SPARC expression; however, increases in insoluble collagen and in levels of SPARC were detected in RV fibroblasts from 4-wk PAB. Therefore, the time course of PO-induced RV hypertrophy differs significantly from myocardial fibrosis and diastolic dysfunction. These temporal differences appear dependent on changes in fibroblast function.

Cardiac interstitial fibrillar collagen accumulation, such as that associated with chronic pressure overload (PO), has been shown to impair left ventricular diastolic function. Therefore, insight into cellular mechanisms that mediate excessive collagen deposition in the myocardium is pivotal to this important area of research. Collagen is secreted as a soluble procollagen molecule with NH2- and COOH (C)-terminal propeptides. Cleavage of these propeptides is required for collagen incorporation to insoluble collagen fibrils. The C-procollagen proteinase, bone morphogenic protein 1, cleaves the C-propeptide of procollagen. Procollagen C-endopeptidase enhancer (PCOLCE) 2, an enhancer of bone morphogenic protein-1 activity in vitro, is expressed at high levels in the myocardium. However, whether the absence of PCOLCE2 affects collagen content at baseline or after PO induced by transverse aortic constriction (TAC) has never been examined. Accordingly, in vivo procollagen processing and deposition were examined in wild-type (WT) and PCOLCE2-null mice. No significant differences in collagen content or myocardial stiffness were detected in non-TAC (control) PCOLCE2-null versus WT mice. After TAC-induced PO, PCOLCE2-null hearts demonstrated a lesser collagen content (PCOLCE2-null TAC collagen volume fraction, 0.41% ± 0.07 vs. WT TAC, 1.2% ± 0.3) and lower muscle stiffness compared with WT PO hearts [PCOLCE2-null myocardial stiffness (β), 0.041 ± 0.002 vs. WT myocardial stiffness, 0.065 ± 0.001]. In addition, in vitro, PCOLCE2-null cardiac fibroblasts exhibited reductions in efficiency of C-propeptide cleavage, as demonstrated by increases in procollagen α1(I) and decreased levels of processed collagen α1(I) versus WT cardiac fibroblasts. Hence, PCOLCE2 is required for efficient procollagen processing and deposition of fibrillar collagen in the PO myocardium. These results support a critical role for procollagen processing in the regulation of collagen deposition in the heart.

Secreted protein, acidic, and rich in cysteine (SPARC) is a matricellular protein that functions in the extracellular processing of newly synthesized collagen. Collagen deposition to form a scar is a key event following a myocardial infarction (MI). Because the roles of SPARC in the early post-MI setting have not been defined, we examined age-matched wild-type (WT; n=22) and SPARC-deficient (null; n=25) mice at day 3 post-MI. Day 0 WT ( n=28) and null ( n=20) mice served as controls. Infarct size was 52 ± 2% for WT and 47 ± 2% for SPARC null ( P=NS), indicating that the MI injury was comparable in the two groups. By echocardiography, WT mice increased end-diastolic volumes from 45 ± 2 to 83 ± 5 μl ( P < 0.05). SPARC null mice also increased end-diastolic volumes but to a lesser extent than WT (39 ± 3 to 63 ± 5 μl; P < 0.05 vs. day 0 controls and vs. WT day 3 MI). Ejection fraction fell post-MI in WT mice from 57 ± 2 to 19 ± 1%. The decrease in ejection fraction was attenuated in the absence of SPARC (65 ± 2 to 28 ± 2%). Fibroblasts isolated from SPARC null left ventricle (LV) showed differences in the expression of 22 genes encoding extracellular matrix and adhesion molecule genes, including fibronectin, connective tissue growth factor (CTGF; CCN2), matrix metalloproteinase-3 (MMP-3), and tissue inhibitor of metalloproteinase-2 (TIMP-2). The change in fibroblast gene expression levels was mirrored in tissue protein extracts for fibronectin, CTGF, and MMP-3 but not TIMP-2. Combined, the results of this study indicate that SPARC deletion preserves LV function at day 3 post-MI but may be detrimental for the long-term response due to impaired fibroblast activation.

Advanced age, independent of concurrent cardiovascular disease, can be associated with increased extracellular matrix (ECM) fibrillar collagen content and abnormal diastolic function. However, the mechanisms causing this left ventricular (LV) remodeling remain incompletely defined. We hypothesized that one determinant of age-dependent remodeling is a change in the extent to which newly synthesized procollagen is processed into mature collagen fibrils. We further hypothesized that secreted protein acidic and rich in cysteine (SPARC) plays a key role in the changes in post-synthetic procollagen processing that occur in the aged myocardium. Young (3 mo old) and old (18–24 mo old) wild-type (WT) and SPARC-null mice were studied. LV collagen content was measured histologically by collagen volume fraction, collagen composition was measured by hydroxyproline assay as soluble collagen (1 M NaCl extractable) versus insoluble collagen (mature cross-linked), and collagen morphological structure was examined by scanning electron microscopy. SPARC expression was measured by immunoblot analysis. LV and myocardial structure and function were assessed using echocardiographic and papillary muscle experiments. In WT mice, advanced age increased SPARC expression, myocardial diastolic stiffness, fibrillar collagen content, and insoluble collagen. In SPARC-null mice, advanced age also increased myocardial diastolic stiffness, fibrillar collagen content, and insoluble collagen but significantly less than those seen in WT old mice. As a result, insoluble collagen and myocardial diastolic stiffness were lower in old SPARC-null mice (1.36 ± 0.08 mg hydroxyproline/g dry wt and 0.04 ± 0.005) than in old WT mice (1.70 ± 0.10 mg hydroxyproline/g dry wt and 0.07 ± 0.005, P < 0.05). In conclusion, the absence of SPARC reduced age-dependent alterations in ECM fibrillar collagen and diastolic function. These data support the hypothesis that SPARC plays a key role in post-synthetic procollagen processing and contributes to the increase in collagen content found in the aged myocardium.

Cardiac tissue from mice that do not express secreted protein acidic and rich in cysteine (SPARC) have reduced amounts of insoluble collagen content at baseline and in response to pressure overload hypertrophy compared with wild-type (WT) mice. However, the cellular mechanism by which SPARC affects myocardial collagen is not clearly defined. Although expression of SPARC by cardiac myocytes has been detected in vitro, immunohistochemistry of hearts demonstrated SPARC staining primarily associated with interstitial fibroblastic cells. Primary cardiac fibroblasts isolated from SPARC-null and WT mice were assayed for collagen I synthesis by [3H]proline incorporation into procollagen and by immunoblot analysis of procollagen processing. Bacterial collagenase was used to discern intracellular from extracellular forms of collagen I. Increased amounts of collagen I were found associated with SPARC-null versus WT cells, and the proportion of total collagen I detected on SPARC-null fibroblasts without propeptides [collagen-α1(I)] was higher than in WT cells. In addition, the amount of total collagen sensitive to collagenase digestion (extracellular) was greater in SPARC-null cells than in WT cells, indicating an increase in cell surface-associated collagen in the absence of SPARC. Furthermore, higher levels of collagen type V, a fibrillar collagen implicated in collagen fibril initiation, were found in SPARC-null fibroblasts. The absence of SPARC did not result in significant differences in proliferation or in decreased production of procollagen I by cardiac fibroblasts. We conclude that SPARC regulates collagen in the heart by modulating procollagen processing and interactions with fibroblast cell surfaces. These results are consistent with decreased levels of interstitial collagen in the hearts of SPARC-null mice being due primarily to inefficient collagen deposition into the extracellular matrix rather than to differences in collagen production.