Galectin-3 in the pathogenesis of heart failure: a causative mediator or simply a biomarker?

First described over three and a half decades ago, galectin-3 has gone by many names and is found in many tissue types. Encoded by a single gene, LGALS3, galectin-3 is a 30-kDa β-galactoside-binding lectin predominantly located in the cytoplasm and the extracellular space. Galectin-3 expression is markedly upregulated in injured and remodeling tissues and is predominantly expressed by infiltrating macrophages and a subset of activated fibroblasts and vascular cells, serving as a marker of inflammation and fibrosis. Because of its consistent association with tissue injury, regardless of the affected organ or etiology, galectin-3 has been extensively studied as a biomarker in a vast array of diseases, with one review literally describing it as a biomarker for everything from A, asthma, to Z, zoster-related pain, including atherosclerosis, liver and kidney fibrosis, cutaneous inflammation, and cancer (7). Galectin-3 expression is increased in the failing and remodeling myocardium, and circulating galectin-3 has been suggested as a prognostic biomarker in patients with heart failure (9). Although galectin-3 has been implicated in the regulation of inflammation, angiogenesis, and fibrosis in many tissues, whether it plays a crucial role in the pathogenesis of adverse cardiac remodeling in failing hearts remains controversial. Robust documentation of a causative role for endogenous galectin-3 in the pathogenesis of myocardial fibrosis, hypertrophy, and dysfunction could identify a promising new therapeutic target for patients with heart failure.

To explore the role of galectin-3 in heart failure, in a recently published article in the American Journal of Physiology-Heart and Circulatory Physiology, Nguyen et al. (6) used both genetic and pharmacological loss of function approaches in a mouse β₂-adrenoceptor overexpression model of cardiomyopathy (6). In agreement with previous studies, the authors observed a robust progressive increase in galectin-3 mRNA and protein levels during the 6 mo in which the mice developed cardiomyopathy. However, in contrast to other studies in heart failure models, pharmacological inhibition of galectin-3 with N-acetyllactosamine (N-Lac; 5 mg·kg⁻¹·day⁻¹ ip, 3 times/wk for 3 wk) or with modified citrus pectin (MCP) and genetic removal of the galectin-3 gene in a knockout mouse did not prevent cardiomyopathy. Specifically, the authors observed no improvement in fibrosis, dysfunction, or remodeling by echocardiography or in inflammatory and fibrotic gene profiles (although N-Lac alone had a mild suppressive effect on a handful of profibrotic genes). As both the pharmacological and genetic inhibition occurred before the development of the cardiomyopathy, the findings suggest that galectin-3 loss was unable to prevent, much less reverse, dysfunction in this model.

These negative results contribute to an ongoing controversy in the field regarding the role of endogenous galectin-3 in cardiac remodeling and the potential benefits of galectin-3 inhibition in heart failure. Several experimental studies in models of heart failure have suggested that galectin-3 loss may be beneficial, although Nguyen et al. (6) are not alone in seeing little to no advantage. Table 1 shows the major studies that examine the effect of genetic and/or pharmacological galectin-3 inhibition in clinically relevant experimental models of heart failure. The study by Yu et al. (10) used pharmacological inhibition of galectin-3 with N-Lac at the same dose and timing as used by Nguyen et al. (although for 1 wk longer) and genetic loss of function experiments in mice subjected to cardiac pressure overload through transverse aortic constriction (TAC). Both studies used only this single dose of N-Lac, so a full dose-response curve is still absent. This particular inhibitor works by binding to the carbohydrate recognition domain that is responsible for binding to β-galactosides, so if galectin-3 has signaling effects through other domains, they may not be affected. Yu et al. found that in mice that underwent TAC protocols and in hypertensive transgenic Ren2 rats, galectin-3 inhibition was beneficial, reducing fibrosis and improving function. Furthermore, in both mouse and rat models of aldosterone-induced cardiac fibrosis, galectin-3 loss attenuated fibrotic changes and reduced cardiac dysfunction (2, 4).

Table 1.

Studies investigating the role of galectin-3 in clinically relevant models of heart failure

Experimental Models	Sex and Sample Size	Strategy Used	Functional Effects of Galectin-3 Loss	Histological and Molecular Effects	Reference
1) Mouse model of angiotensin II infusion	Male WT and galectin-3 KO mice (n = 5–12 mice/group); male Ren2 rats (n = 5–12 rats/group)	1) Pharmacological inhibition with N-acetyllactosamine (5 mg/kg ip 3 times/wk)2) Genetic loss (global galectin-3 KO mice)	Genetic loss of galectin-3 attenuated systolic and diastolic dysfunction in both angiotensin II infusion and TAC models; similar protective effects were noted in Ren2 rats that received galectin-3 inhibitor	Both genetic disruption and pharmacological inhibition attenuated cardiac fibrosis	10
2) Mouse model of pressure overload induced through TAC (end points studied after 28 days)
3) Ren2 rat model of heart failure
1) Rat and mouse models of experimental hyperaldosteronism (aldosterone/salt-induced cardiac fibrosis)	Male Wistar rats (n = 9–10 rats/group); male SHRs (n = 8 rats/group); male WT and galectin-3 KO mice (n = 7 mice/group)	1) Pharmacological inhibition with oral MCP (100 mg·kg−1·day−1)2) Genetic loss (galectin-3 KO mice)	Effects on cardiac function were not studied	Galectin-3 inhibition reduced inflammation and attenuated fibrosis in hyperaldosteronism models and in SHRs; genetic loss of galectin-3 attenuated inflammation in the hyperaldosteronism model	4
2) SHRs
Rat and mouse models of experimental hyperaldosteronism (aldosterone/salt-induced cardiac fibrosis)	Male Wistar rats (n = 9–10 rats/group); male WT and galectin-3 KO mice (n = 7 mice/group)	1) Pharmacological inhibition with oral MCP (100 mg·kg−1·day−1)	Galectin-3 inhibition attenuated the hypertensive response and reduced cardiac hypertrophy and diastolic dysfunction	Both galectin-3 inhibition and genetic loss attenuated renal and cardiac fibrosis in response to aldosterone	2
		2) Genetic disruption (galectin-3 KO mice)
Rat model of obesity-related cardiomyopathy induced through administration of a high-fat diet (6 wk)	Male Wistar rats (n = 8 rats/group)	Oral MCP (100 mg·kg−1·day−1)	Galectin-3 inhibition had no significant effects on left ventricular wall thickness, chamber dimensions, or systolic function; diastolic function was not studied	Galectin-3 inhibition did not affect cardiomyocyte hypertrophy but attenuated myocardial inflammation and fibrosis and reduced oxidative stress	5
Mouse model of pressure overload induced through TAC (end points studied after 7, 28, and 56 days of TAC)	Male and female WT and galectin-3 KO mice (C57BL/6J strain, n = 10−20 mice/group)	Galectin-3 KO mice	Galectin-3 loss did not affect survival or systolic or diastolic function but was associated with accelerated hypertrophy; female galectin-3 KO mice had delayed chamber dilation	Galectin-3 loss did not affect fibrosis and cardiomyocyte hypertrophy in the pressure-overloaded heart	3
Mouse model of isoproterenol-induced heart failure	Male mice (n = 10−12 mice/group)	Oral MCP (100 mg·kg−1·day−1)	Galectin-3 inhibition attenuated systolic dysfunction and dilative remodeling in isoproterenol-treated animals	Galectin-3 inhibition reduced cardiomyocyte hypertrophy, diminished inflammation, and decreased interstitial fibrosis	8
Rat model of supravalvular aortic constriction (end points studied after 6 wk)	Male adult Wistar rats (n = 7 rats/group)	Oral MCP (100 mg·kg−1·day−1)	Galectin-3 inhibition abrogated the reduction in left ventricular end-diastolic diameter observed in pressure-overloaded hearts without affecting wall thickness and ejection fraction	Galectin-3 inhibition attenuated cardiomyocyte hypertrophy, decreased fibrosis, diminished inflammatory gene expression, and attenuated the increase in cardiac brain natriuretic peptide and atrial natriuretic peptide levels	1

KO, knockout; MCP, modified citrus pectin; SHRs, spontaneously hypertensive rats; TAC, transverse aortic constriction; WT, wild type.

In contrast, Frunza et al. (3) found no significant effects of global genetic galectin-3 loss on survival, cardiac fibrosis, and dysfunction after pressure overload. Although the pressure overload protocol was similar to the Yu et al. (10) study and the same galectin-3 knockout mouse line was used, several differences in experimental design (such as sex and age of the mice studied, analysis of different time points, and differences in the intensity of the pressure load) may explain the conflicting observations. Further complicating the issue, Frunza et al. discovered galectin-3 expression in atrial myocytes at baseline and in a subset of ventricular myocytes after pressure overload in a mosaic pattern specifically adjacent to areas of fibrosis. The significance and potential role of galectin-3 expression by cardiomyocytes were not explored in this study and remain unclear. In broad agreement with other studies, however, Frunza et al. also observed a robust increase in macrophage galectin-3 expression in the TAC model.

What is the basis for conflicting findings in studies that have examined the role of galectin-3 in heart failure? In some cases, contrasting conclusions between studies reflect different interpretations of experimental findings. For example, in two studies, the effects of galectin-3 inhibition on cardiac remodeling are supported exclusively by histological data, despite the absence of significant differences in functional end points (1, 5). In other cases, differences in protocol standardization, the timing of disease progression in various models, challenges related to dosing, delivery, and off-target effects of pharmacologic inhibitors, and sex-specific actions may explain disparate findings. Inconsistencies between studies using different models of heart failure may reflect context-dependent actions of galectin-3. The cell biological basis for heart failure in each model may be a major determinant of the relative significance of galectin-3. Considering its well-documented effects on fibroblast and immune cell phenotype, galectin-3 may play more important roles in heart failure associated with prominent inflammation and fibrosis. In contrast, the effects of galectin-3 may be less prominent in the genetic model of cardiomyocyte-specific β₂-adrenergic receptor overexpression used by Nguyen et al. (6).

It should be emphasized that negative studies, such as that by Nguyen et al. (6), are important for revealing the dynamic topography of a subject. Unfortunately, studies with negative findings often remain unpublished. Not surprisingly, investigators with negative data cannot attract funding and are often discouraged from pursuing their research. These patterns of funding and publication create a bias against negative studies, leading to selective publication of positive investigations that can develop an oversimplified view of a field. This typically creates an overly optimistic perspective of the potential impact of new therapies, often followed by painful translational challenges and failures. Publication of all rigorous investigations, both positive and negative, is critical to paint a more realistic translational landscape.

Human heart failure exhibits remarkable pathophysiological heterogeneity and cannot be recapitulated by a single animal model. Just as one animal study is not sufficient to predict the effectiveness of a therapeutic approach in human patients, the article by Nguyen et al. (6) does not mean that galectin-3 inhibition will not work. What it does, however, is to document that in a cardiomyopathy caused by an exaggerated β₂-adrenergic response, galectin-3 is not critical for dysfunction, adverse remodeling, and progression of disease. From a translational perspective, this type of information is valuable for identification of pathophysiologically distinct patient subpopulations that may be unlikely to respond to galectin-3 inhibition. Clinical trials rarely get do-overs if the patient population originally targeted does not respond significantly. Qualifying (negative) studies, those that find the circumstances where an approach does not work, are critical for identifying which therapeutic strategies deserve further development and which ones should be abandoned.

GRANTS

J. A. Kirk’s laboratory is supported by National Heart, Lung, and Blood Institute (NHLBI) Grant R01-HL-136737 and by American Heart Association Grant 14SDG20380148. N. G. Frangogiannis’ laboratory is supported by NHLBI Grants R01-HL-76246 and R01-HL-85440 and by Department of Defense Grants PR151134 and PR151029.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

J.A.K. and N.G.F. drafted manuscript; J.A.K. and N.G.F. edited and revised manuscript; J.A.K. and N.G.F. approved final version of manuscript.