A Land of Controversy: Fibroblast Growth Factor-23 and Uremic Cardiac Hypertrophy

Abstract

Cardiac hypertrophy is a common feature in patients with CKD. Recent studies revealed that two phosphate regulators, fibroblast growth factor-23 and α-Klotho, are highly involved in the pathophysiologic process of CKD-induced cardiac hypertrophy. With decreasing renal function, elevated fibroblast growth factor-23 and decreased α-Klotho may contribute to cardiac hypertrophy by targeting the heart directly or by inducing systemic changes, such as vascular injury, hemodynamic disorders, and inflammation. However, several studies have demonstrated that disturbances in the fibroblast growth factor-23/α-Klotho axis do not lead to cardiac hypertrophy. In this review, we describe the cardiac effects of the fibroblast growth factor-23/α-Klotho axis and summarize recent progress in this field. In addition, we present not only the main controversies in this field but also provide possible directions to resolve these disputes.

Cardiovascular complications of CKD seriously affect the prognosis of patients with CKD.¹ Cardiac hypertrophy is the most common of such complications and underlies the pathology for other heart conditions. However, there are still many unknowns regarding the pathophysiology of CKD-induced cardiac hypertrophy.² To date, several factors are considered to be involved in the pathogenesis of uremic cardiac hypertrophy, including hemodynamic overload, systemic inflammation, accumulation of uremic toxins, and other factors.³ Recent findings suggest that two proteins that are highly involved in phosphate metabolism, fibroblast growth factor-23 (FGF-23) and its coreceptor α-Klotho play pivotal roles in cardiovascular abnormalities (especially cardiac hypertrophy) in the setting of CKD. However, some researchers dispute this, arguing that disturbance in the FGF-23/α-Klotho axis is unable to induce cardiac hypertrophy. This review will focus on the FGF-23/α-Klotho axis in uremic cardiac hypertrophy, present controversies about their roles in cardiovascular abnormalities, and provide direction for future research.

Brief Introduction to the FGF-23/α-Klotho Axis

FGF-23

FGF-23, a newly discovered FGF-19 subfamily member, is an approximately 32 kDa glycoprotein that is mainly secreted by osteocytes or osteoblasts.⁴ After removal of the signal peptide and glycosylation, intact FGF-23 (iFGF-23) is secreted into the blood (Figure 1).⁵ iFGF-23 can be cleaved at R¹⁷⁶XXR¹⁷⁹/S¹⁸⁰ into inactive approximately 18 kDa amino-terminal and approximately 12 kDa carboxy-terminal (cFGF-23) fragments, and the latter can compete with iFGF-23 for binding to the FGF receptor (FGFR), thus exerting certain biologic functions (Figure 1).^6,7 Unlike other FGF subfamilies, the FGF-19 subfamily lacks a heparin binding domain, which is replaced by a Klotho binding domain. Hence, low affinity to heparin sulfate enables FGF-19 subfamily members to enter the bloodstream and function as endocrine factors.^8–10 On the other hand, the lack of a heparin binding domain renders these proteins incapable of forming the FGF-FGFR-heparin sulfate complex and activating downstream signaling.¹¹ Klotho, a single-pass transmembrane protein, interacts with FGF-19 subfamily members and FGFR to form the FGF-FGFR-Klotho complex, and then activates FGFR signaling.^12,13

Figure 1.

The maturation and disintegration of FGF-23. FGF-23 precursor contains 251 amino acids (aa) and can be divided into a signal sequence (24 aa), a FGFR binding domain (155 aa), and an α-Klotho binding domain (72 aa). After removal of the signal sequence, iFGF-23 is secreted into the blood or cleaved between aa 179 and 180 into inactive amino-terminal FGF-23 and active cFGF-23.

α-Klotho

Klotho is a membrane protein that demonstrates antiaging effects.¹⁴ It can be divided into three subtypes: α-Klotho, β-Klotho, and γ-Klotho.^15–17 Among these, α-Klotho and β-Klotho interact with FGF-23 and FGF-19/FGF-21, respectively, and assist the binding of FGF and FGFR.^15,17,18 However, the biologic function of the third member, γ-Klotho, remains unclear. In mammals, α-Klotho (abbreviated as Klotho below) is considered to exist in two forms in vivo: membranous Klotho (mKlotho) and secreted Klotho (sKlotho). The secreted enzymes A disintegrin and metalloproteinase-10 (ADAM-10), ADAM-17, and β-site amyloid precursor protein cleaving enzyme-1 can incise the extracellular domain of mKlotho, which then generates sKlotho.^19,20 Alternative splicing of Klotho mRNA generates sKlotho as well.²¹

FGF-23 exerts its functions mainly through FGFR-1; however, FGF-23 needs the assistance of mKlotho to activate FGFR-1.¹² The expression of mKlotho is tissue specific, predominantly in distal convoluted tubules of the kidney and chief cells in the parathyroid gland.²² In the kidney, FGF-23 suppresses the reabsorption of phosphate and inhibits the activation of 1,25-dihydroxyvitamin D₃.²³ In the parathyroid gland, FGF-23 restricts the synthesis of parathyroid hormone.²⁴ In the heart, Klotho is expressed in the sinoatrial node, but the mRNA level of Klotho is fairly low or even undetectable in the myocardium.²⁵ Studies using either PCR or RNA sequencing confirmed low levels or even a lack of Klotho mRNA in mouse and human heart,²² whereas Klotho protein can be detected by Western blotting.²⁶ Such results may indicate that the actions of FGF-23 and Klotho in cardiac tissue are different from those in the kidney and parathyroid gland.

Cardiac Effects of FGF-23 in CKD

CKD is a common result of various kidney and urologic diseases, with a gradual decrease in the number of functional nephrons and excretory function. As a result of impaired excretory capacity, hyperphosphatemia and hyperparathyroidism lead to increased circulating levels of FGF-23. Often, worse kidney function is associated with higher circulating FGF-23 concentrations.^27–29 Although the increase in FGF-23 helps maintain a normal serum phosphate level, epidemiologic studies revealed that a high circulating level of FGF-23 is an independent risk factor for cardiovascular disease and may increase mortality in patients with CKD.^30–35

Uremic Cardiac Hypertrophy and FGF-23

The primary link between FGF-23 and cardiovascular complications in CKD is left ventricular hypertrophy (LVH).^33,36 The detailed mechanism of FGF-23–induced LVH was first revealed by Faul et al.³³ A series of studies confirmed that hypertrophic growth of cardiomyocytes can be induced by FGF-23, and that this effect does not require FGFR-1/Klotho but does need the presence of FGFR-4.^26,33,37 Specifically, FGF-23 binds to FGFR-4 in cardiomyocytes and results in stimulation of the phospholipase Cγ (PLCγ)/calcineurin pathway,³⁷ which is an important mediator of cardiac hypertrophy in response to a majority of pathologic stimuli (including CKD conditions).^38,39 FGFR-4 null mice fed a high-phosphate diet³⁷ and 5/6 nephrectomized rats treated with an FGFR-4 antagonist³⁷ or a calcineurin inhibitor⁴⁰ all demonstrated no obvious cardiac hypertrophy. These findings further indicate an important role of FGFR-4/PLCγ/calcineurin signaling in FGF-23–induced cardiac hypertrophy. Our studies also revealed that microRNA-30, which is an endogenous post-transcriptional inhibitor of calcineurin,⁴¹ attenuates CKD-induced LVH (unpublished data). In addition, exogenous microRNA-30 supplementation also inhibits FGF-23–induced cardiomyocyte hypertrophy (unpublished data). Recently, Han et al.⁴² found that mice with cardiomyocyte FGFR-4 deletion exhibit resistance to FGF-23–induced cardiac hypertrophy, which provides further evidence for the vital role of cardiac FGFR-4 in FGF-23–induced cardiac hypertrophy.

The myocardium of patients with CKD²⁶ and CKD rats⁴³ can secrete FGF-23 and exert a paracrine effect, suggesting there may be a local regulatory system of FGF-23. Leifheit-Nestler et al.²⁶ confirmed that FGF-23 promotes cardiac hypertrophy from multiple perspectives and, in a follow-up study, Leifheit-Nestler et al.⁴³ further revealed that myocardial FGF-23 significantly promotes LVH in patients with CKD in a paracrine manner. In addition, FGF-23 released by cardiomyocytes also stimulates fibroblasts to secrete TGF-β, connective tissue growth factor, and other profibrotic factors.⁴⁴ These studies indicate that the paracrine role of FGF-23 in cardiac tissue may not be ignored in CKD-induced cardiac hypertrophy.

Controversies

Although a number of studies^33,36 indicate that FGF-23 increases the incidence of cardiovascular abnormalities, there also are studies that do not seem to support this conclusion. X-linked hypophosphatemia (XLH), which causes an increased concentration of FGF-23, can be used as a convenient model to investigate the role of FGF-23 in the myocardium. Recently, Pastor-Arroyo et al.⁴⁵ found that the XLH mouse model, which has normal kidney function, exhibits high circulating FGF-23 and hypophosphatemia and has no activation of the calcineurin pathway in the myocardium; they also did not observe LVH or abnormal cardiac function in these mice. Although this result is consistent with the results from some groups,^46,47 other studies have observed significant cardiac hypertrophy in patients with XLH or XLH models.^48–50 The reasons for these differences may depend on factors other than FGF-23, including age, diet, and genetic background (Table 1). In addition, XLH and CKD exert different physiologic effects, such as hypophosphatemia in XLH versus hyperphosphatemia in CKD, and these factors may significantly affect the myocardial effects of FGF-23 (Figure 2). For example, Liu et al.⁵¹ showed that overexpression of FGF-23 leads to hypotension and cardiac hypertrophy, and that these abnormalities can be normalized by a high-phosphate diet.

Table 1.

The comparison of the different high circulating FGF-23 models

Model (genetic background)	Circulating FGF-23 (model:control)a	Blood-Related Parameters	Cardiac Changes	Reference
80 μg/kg per d FGF-23 IOCV for 5 d (C57BL/6J)	(c + i) FGF-23 ↑ (3.2:1)	/	LVH	33
Klotho heterozygous (/)	(c + i) FGF-23 ↑ (3.0:1)	Phosphate ↑	LVH	33
		Active VitD3 ↑
5/6 nephrectomy (Sprague Dawley)	(c + i) FGF-23 ↑ (12.5:1)	Creatinine ↑	LVH	33
		CCreatinine ↓
		BUN ↑
		Systolic pressure ↑
5/6 nephrectomy (Sprague Dawley)	(c + i) FGF-23 ↑ (3.5:1)	Creatinine ↑	LVH	37
		CCreatinine ↓
		BUN ↑
		Systolic pressure ↑
		Diastolic pressure ↑
		Phosphate →
2.0% phosphate diet for 12 wk (C57BL/6)	(c + i) FGF-23 ↑ (6.0:1)	BUN →	LVH	37
75 μg/kg per d FGF-23 IP for 5 d (C57BL/6)	FGF-23 ↑ (/)	Systolic pressure ↑	LVH	42
2.0% phosphate diet for 3 mo (C57BL/6)	iFGF-23 ↑ (5.6:1)	/	LVH	121
5/6 nephrectomy (Sprague Dawley)	iFGF-23 ↑ (/)	CCreatinine ↓	LVH	121
		BUN ↑
		Systolic pressure ↑
		Diastolic pressure ↑
		Phosphate ↑
PhexC733R (C3Heb/FeJ)	iFGF-23 ↑ (12.8:1)	Phosphate ↓	No changes	45
		Calcium →
		sKlotho ↓
		Active VitD3 ↓
		PTH ↑
		BUN →
		Creatinine →
		Systolic pressure ↓
		Cholesterol →
Hyp (C57BL/6J)	/	Systolic pressure →	Smaller heart	47
Hyp (C57BL/6)	iFGF-23 ↑ (14.4:1)	Aldosterone ↓	LVH	50
Transgenic human FGF-23 in liver (C57BL/6J)	/	Systolic pressure ↓	LVH	51
		Phosphate ↓
		Epinephrine ↑
		Norepinephrine ↑
		Dopamine ↑
		Ang-II ↑
		Cortisol ↑
		Glucose ↓

IOCV, injection of caudal vein; (c + i) FGF-23, C-terminal and intact FGF-23; /, not mentioned; VitD₃, vitamin D₃; C_Creatinine, creatinine clearance; →, no significant difference between model group and control group; IP, intraperitoneal injection; PTH, parathyroid hormone; Hyp, X-linked semidominant mutation of Phex gene that causes hypophosphatemia and high circulating FGF-23, whereas does not affect kidney functions.

^aModel:control is calculated by average value of each group.

Figure 2.

The cardiac effects of elevated FGF-23 is significantly different in CKD and XLH. Increased FGF-23 is observed in both CKD and XLH. However, the increase in FGF-23 occurs secondary to hyperphosphatemia, whereas it is idiopathic and induces hypophosphatemia in XLH. Elevated FGF-23 induces cardiac hypertrophy in CKD, whereas no cardiac changes are observed in XLH. In CKD, perhaps the relative levels of circulating FGF-23 and phosphate determine the prohypertrophic effects of FGF-23. Moreover, it is unclear whether any other factors might affect the prohypertrophic effect of FGF-23.

Remarkably, there are few studies that seem to refute or support the prohypertrophic effect of FGF-23. Shalhoub et al.⁵² found that administration of FGF-23–neutralizing antibody in CKD rats did not attenuate cardiac hypertrophy. On the other hand, ferric citrate administration from 6 to 10 weeks of age reduces iFGF-23 and significantly improves cardiac function in Col4a3 knockout mice (a model of progressive CKD).⁵³ These two seemingly contradictory results may be due to interference of other factors related to cardiac hypertrophy. For example, FGF-23 neutralization facilitates aortic calcification and severe hyperphosphatemia,⁵² which are prohypertrophic stimuli. In addition, it is possible that ferric citrate improves cardiac function by alleviating iron deficiency and anemia rather than by lowering FGF-23.⁵³ Therefore, these two studies were unable to prove or exclude the direct cardiac role of FGF-23.

Cardiac Effects of Klotho in CKD

The kidney’s distal tubules are a main source of Klotho and sKlotho.²² In the setting of CKD, deficiency in Klotho and sKlotho can be caused by a reduction in functional nephrons and inhibition of Klotho expression. CKD-related factors such as angiotensin II (Ang-II),^54,55 inflammatory cytokines,^56,57 phosphate,⁵⁸ deficiencies in vitamin D⁵⁹ and erythropoietin,^60,61 or even FGF-23⁶² can trigger the suppression of Klotho.

Uremic Cardiac Hypertrophy and Klotho

Animal experiments established a clear relationship between Klotho and the cardiovascular system. Klotho-null mice develop CKD-like cardiovascular phenotypes, such as vascular calcification, cardiac hypertrophy, and fibrosis.^14,33,58,63 However, Klotho mRNA is not detected in the myocardium, aside from the sinoatrial node²⁵; thus, loss of Klotho in the myocardium may not directly contribute to cardiac hypertrophy.^22,33 Leifheit-Nestler et al.²⁶ revealed the presence of Klotho protein in cardiac muscle, which suggests that the heart can take up sKlotho from the circulation. A series of studies found that sKlotho inhibits the prohypertrophic signaling pathway and exerts unique biologic functions that are independent of FGF-23.^64–67 For example, sKlotho directly binds to the type II TGF-β receptor,⁶⁸ blocking TGF-β1 signaling, and TGF-β1⁶⁹ is highly involved in cardiac hypertrophy.

Intriguingly, sKlotho suppresses cardiac hypertrophy that is induced by Ang-II,⁵⁸ TGF-β,^58,67 or isoprenaline^58,67,70; thus, sKlotho demonstrates a broad spectrum antihypertrophic effects. sKlotho inhibits the production of reactive oxygen species (ROS), and sKlotho deficiency may lead to an increase in ROS in myocardial tissue and induce cardiac hypertrophy.⁷¹ In addition, several signaling pathways that may be involved in CKD-induced cardiac injury, including the PI3K,⁶⁵ WNT,^72,73 and NF-κB pathways,⁷⁴ can be regulated by sKlotho. Recently, a specific receptor for sKlotho, monosialoganglioside, was identified; sKlotho binds to it and inhibits raft-dependent PI3K signaling.⁷⁵ This illustrates that sKlotho can inhibit the translocation of transient receptor potential cation channel, subfamily C, member 6 (TRPC6), an ion channel that is involved in the formation of cardiac hypertrophy.^65,76,77

Controversies

Although animal studies have established a clear relationship between Klotho and the cardiovascular system, some clinical studies have failed to detect an independent association between sKlotho level and left ventricular mass index in patients on hemodialysis.^78–80 In patients with CKD from stage 1 to 5 (predialysis), however, sKlotho demonstrates a significant inverse association with left ventricular mass index.⁸¹ Another parameter of cardiac hypertrophy, the wall thickness of the left ventricle, has an inverse association with sKlotho in patients on dialysis, whereas such an association does not exist in predialysis patients.^79,80 As mentioned by Zhang et al.,⁸⁰ such discrepancies may be caused by dialysis, which is associated with such factors as interdialytic weight gain, ultrafiltration, dialysate, and dialyzers that do not apply to nondialysis patients with CKD. Furthermore, sKlotho also does not associate with pathologic changes in the heart or predict cardiovascular risk in patients with heart disease who have normal renal function.⁸²

Surprisingly, a study by Sugiura et al.⁸³ found increased sKlotho levels in patients with CKD. Authors of a subsequent study that found a significant increase in sKlotho level in CKD hypothesized that the reliability of assay kits is the principal reason for these unexpected results.⁸⁴ Indeed, a high concentration of bilirubin (250 mg/L) can significantly reduce the measured value of sKlotho.⁸⁴ In addition, different kits contain different sKlotho antibodies, which are directed to different epitopes, and such differences may obviously influence the measurements. Currently, the best way to measure sKlotho in human samples remains controversial. Thus, the association between sKlotho and cardiovascular risk needs careful consideration.

Combined Cardiac Effects of FGF-23 and Klotho in CKD

As a known scenario of cardiac hypertrophy, CKD is accompanied by a low level of sKlotho and a high level of FGF-23 in circulation.^85,86 Although FGF-23 induces cardiac hypertrophy via FGFR-4 in the absence of Klotho,^33,37 sKlotho may exert cardioprotective effects without the presence of FGF-23.^68,77,87 Whether an increased ratio of circulating FGF-23 to sKlotho in CKD results in more serious cardiac hypertrophy is unknown.⁸⁸

It seems that normalization of circulating FGF-23 and phosphate via a low-phosphate diet containing only 0.2% inorganic phosphate does not block the exacerbation of cardiac hypertrophy in CKD mice with heterozygous Klotho deletion.⁶⁶ Hence, the reduction in sKlotho may be involved in CKD-induced cardiac hypertrophy regardless of FGF-23 level. Six et al.⁸⁹ found that Klotho deficiency is detrimental to the endothelium and that Klotho sufficiency prevents the damaging effects of FGF-23 and hyperphosphate. In addition, Hu et al.⁵⁸ revealed that a positive correlation between FGF-23 levels and cardiac hypertrophy and fibrosis occurs only with relatively low sKlotho levels. Han et al.⁴² found that sKlotho injection inhibits FGF-23–induced cardiac hypertrophy, providing further evidence for the antihypertrophic effect of sKlotho. Collectively, we may infer that beyond the levels of FGF-23 or sKlotho individually, the ratio of these two proteins is a key contributor to uremic cardiac hypertrophy. According to this hypothesis, the paracrine effect of FGF-23 may further lead to an increase in the ratio of FGF-23 to sKlotho locally and aggravate cardiac hypertrophy.^26,43

To our knowledge, Klotho is not only a coreceptor of FGF-23 but also a converter of FGF-23/FGFR interaction. Mechanistically, FGF-23 has a higher affinity for FGFR-4 than for FGFR-1 in the absence of Klotho, but it shows a nearly 20-fold increase in affinity for FGFR-1 when Klotho is present.⁹⁰ Recent studies have revealed that sKlotho acts as a coreceptor for FGF-23⁹¹ and a protective factor for the myocardium through “pathway conversion,” and these were previously considered as essential functions of mKlotho (Figure 3).^12,42 Specific cardiac ablation of Fgfr-4 blocks FGF-23–induced cardiac hypertrophy, and sKlotho administration converts FGF-23–induced PLCγ/calcineurin signaling into mitogen-activated protein kinase (MAPK) signaling.⁴² In addition, sKlotho leads to only a small increase in activation of MAPK activity in response to FGF-23,⁹² which means that this pathway conversion may not induce robust activation of MAPK signaling and result in cardiac hypertrophy. Unfortunately, however, these studies were not reproduced in CKD models. Theoretically, high FGF-23 and low sKlotho levels in CKD may lead to the interaction of FGF-23 and FGFR-4 and activate the calcineurin pathway, which can cause cardiac hypertrophy.³⁷

Figure 3.

FGF-23 activates MAPK signaling in the presence of sKlotho, whereas activates calcineurin signaling in the absence of sKlotho. (A) Under physiologic conditions, sKLOTHO functions as a coreceptor of FGF-23 and leads to MAPK signaling activation. Importantly, sKlotho only leads to a small increase in activation of MAPK activity in response to FGF-23, which means it may not induce robust activation of MAPK signaling and result in cardiac hypertrophy. (B) In CKD, decreased sKlotho blunts the interaction between FGF-23 and FGFR-1, and FGF-23 preferentially binds to FGFR-4, thus resulting in calcineurin cascade activation and pathologic cardiac hypertrophy.

Indirect Cardiac Effects of FGF-23 and α-Klotho in CKD

Effects on Vessels

Vascular lesions, another pathologic feature of patients with CKD, are closely related to cardiac impairment. Animal models or patients with genetic Klotho deficiency develop extensive arteriosclerosis and vascular calcification, which are similar to CKD-related vascular injury.^14,93 In CKD mice, Klotho preservation not only attenuates hyperphosphatemia but also suppresses the activity of the NaPi-3 group of sodium-coupled transporters (Pit-1 and Pit-2), thus inhibiting calcification and dedifferentiation of vascular smooth muscle cells.⁹⁴ In addition, Klotho deficiency is highly involved in endothelial dysfunction, which is mediated by increased apoptosis and reduced production of endothelial nitric oxide synthase (eNOS).^95–97 In vivo treatment with Klotho leads to alleviation of endothelial dysfunction, high BP, and perivascular fibrosis.⁹⁸

It is not surprising that FGF-23 deficiency results in a phenotype in blood vessels similar to that of CKD-related vascular injury, given the disturbance in phosphate metabolism.^99,100 Conversely, high circulating FGF-23 also results in vascular injury. Studies in patients with CKD found that high FGF-23 is associated with vascular dysfunction.^101–103 In endothelial cells, FGF-23/Klotho/FGFR-1 promotes the production of eNOS via the AKT-dependent pathway. FGF-23 stimulates both FGFR-dependent ROS formation via activation of NADPH oxidase 2, as well as ROS degradation via activation of SOD2 and catalase.⁹⁶ However, the absence of Klotho blunts FGF-23–induced eNOS, SOD2, and catalase generation, whereas FGF-23–induced ROS synthesis is unaffected.⁹⁶ The balance between the generation and degradation of ROS may be lost in CKD, thus leading to the successive production of ROS and vascular injury. Therefore, FGF-23’s vascular effects, as they affect cardiomyocytes, are probably also determined by the ratio of FGF-23 to sKlotho.

Effects on Kidney

Elevated FGF-23 has been described as a hypertension-related factor, and studies suggest that prohypertensive effect may partially be due to the increase in sodium reabsorption via sodium chloride cotransporters in the distal tubules.^48,104 In contrast, other work demonstrated that FGF-23 overexpression results in hypotension, and hypophosphatemia caused by high levels of circulating FGF-23 may mediate such BP decreases.⁵¹ To illustrate this inconsistency, the balance in circulating levels of FGF-23 and phosphate should be considered.

Effects on Immunocytes

In recent years, inflammation has been considered as a consequence of disturbance of the FGF-23/Klotho axis,¹⁰⁵ suggesting another way FGF-23 and Klotho can indirectly affect the heart. Elevated FGF-23 acts on FGFR-2 on neutrophils in the absence of Klotho; this deactivates β₂-integrin and impairs recruitment of neutrophils to infection sites,^106–108 which may partially explain the impaired host defense resulting from neutrophil dysfunction in patients with CKD. Studies also have demonstrated that FGF-23 inhibits the production of 1,25-dihydroxyvitamin D₃ and results in monocyte dysfunction (increasing susceptibility to bacterial infection)¹⁰⁹ and macrophage activation,¹¹⁰ which are closely related to chronic inflammation. In addition, FGF-23 promotes the release of proinflammatory factors in hepatocytes under CKD conditions and may participate in chronic inflammation.¹¹¹

Given Klotho’s role as an anti-inflammatory effector,¹¹² its deficiency in CKD may also play a role in the generation of inflammation. The cardiovascular pathogenic role of chronic inflammation in CKD has been demonstrated¹¹³; hence, the association between FGF-23 and inflammation may greatly enhance the significance of FGF-23 in uremic cardiac hypertrophy.

Summary and Future Perspectives

As endocrine factors, FGF-23 and Klotho exert effects on various tissues in the absence or presence of one another. In this review, we described the direct and indirect effects of FGF-23 and Klotho on cardiac tissues, including hypertrophy-related effects, vessel-related effects, and inflammation-related effects (Figure 4). However, in light of continuing controversies regarding the cardiac effects of FGF-23 and Klotho, further investigations are needed.

Figure 4.

Direct and indirect effects of FGF-23 and KLOTHO on myocardium. In CKD, high FGF-23 level and Klotho/sKLOTHO deficiency lead to several changes in blood vessels, inflammatory responses, and the kidney, including vascular injury, inflammation, hypertension, and active 1,25-dihydroxyvitamin D₃ (1,25-[OH]₂-VitD₃) deficiency. These changes can induce cardiac hypertrophy directly and indirectly. Moreover, disturbances in FGF-23/sKLOTHO may induce cardiac hypertrophy and fibrosis directly. 25-OH-VitD₃, 25-hydroxyvitamin D₃; Cl⁻, chloride ion; Na⁺, sodium ion.

The discrepancy in cardiac changes in XLH and CKD is the most controversial aspect of FGF-23–induced cardiac hypertrophy, and several mechanisms may account for this discrepancy. The process of FGF-23 cleavage may explain this discordance. Under physiologic conditions, iFGF-23 and its cleavage are in dynamic equilibrium, but in CKD, iFGF-23 cleavage may be impaired, leading to iFGF-23 accumulation and several pathologic changes.^114–116 Mechanistically, cFGF-23, which is the cleavage product of iFGF-23, can bind to FGFR-1/Klotho competitively, thereby inhibiting hypophosphatemia induced by a high circulating level of iFGF-23.¹¹⁷ A recent study found that cFGF-23 attenuates fibrosis and inflammation in diabetic nephropathy, although the underlying mechanisms are not clear.⁷ It is possible that cFGF-23 interferes with the interaction between FGF-23 and FGFR-4. Notably, the cleavage of iFGF-23 seems normal in XLH.⁴⁵ Hence, abnormal cleavage of iFGF-23 may contribute to uremic cardiac hypertrophy.

The largest difference between CKD and XLH is that XLH exhibits significant hypophosphatemia instead of hyperphosphatemia. Interestingly, overexpression of FGF-23 leads to hypotension and cardiac hypertrophy, with activation of the renin-angiotensin system (RAS) and the sympathetic nervous system. The latter phenomena (cardiac hypertrophy and activation of the RAS and sympathetic nervous system) are considered a stress response to hypotension and can be mitigated by a high-phosphate diet.⁵¹ However, Pastor-Arroyo et al.⁴⁵ observed hypotension in XLH models, but there were no cardiac changes in these models. This inconsistency may be caused by different degrees of stress responses due to genetic background (Table 1). As mentioned above, FGF-23 results in hypotension, and hypotension induces a stress response and leads to cardiac hypertrophy. However, whether FGF-23 leads directly to a stress response and promotes cardiac hypertrophy as a compensatory response requires further study.

According to current understanding, the FGF-23/Klotho axis interacts with the RAS^44,50 and ROS,⁹⁶ which are all constituents of the stress response. It is possible that elevated FGF-23 in CKD can act as a stress hormone, similar to another member of the FGF-19 subfamily, FGF-21,^118,119 a speculation that can be partially verified in cardiomyocytes. Incubation of cardiomyocytes with FGF-23 results not only in cardiomyocyte hypertrophy, but also augments the contractility of cardiomyocytes, and the hypertrophic condition seems to be a compensatory response for increased contractility.¹²⁰ An in vivo study further revealed that FGF-23–induced cardiac/cardiomyocyte hypertrophy is reversible.¹²¹ This suggests that FGF-23–induced hypertrophy is a reversible response to increased contractility, and the disappearance of the stressor (FGF-23) results in a decreased stress response (mitigation of cardiac hypertrophy). Moreover, FGF-23 upregulates the intracellular expression of Ang-II in cardiomyocytes and shares a common mechanism of calcium-dependent cardiomyocyte hypertrophy with Ang-II. This further indicates a close interaction of FGF-23 and the stress response in cardiomyocytes.¹²²

XLH is also an abnormal condition and can trigger a stress response as well. Several studies reported a high rate of LVH in patients with XLH or XLH models.^48–50 One possibility is that severe hypotension and hypophosphatemia in XLH trigger a stress response and strengthen FGF-23–induced hypertrophy; in turn, FGF-23 enhances the stress response, forming a positive feedback loop and ultimately leading to significant cardiac hypertrophy. CKD can be regarded as a stress state, wherein the stress response aims to correct manifestations of the disorder such as hyperphosphatemia and uremic toxin accumulation. Klotho partially blunts the stress response by inhibiting the RAS activation, ROS generation, and even inflammation.^71,112,118 Thus, downregulation of Klotho in CKD may be a consequence of promoting the stress response, although it reduces the excretion of phosphate.

Collectively, we hypothesize that severe hyperphosphatemia, hypophosphatemia, and a high circulating level of FGF-23 trigger the stress response, thus downregulating Klotho/sKlotho and ultimately enhancing the stress response. The balance of FGF-23, sKlotho, and phosphate actually represents the degree of the stress response. Under this hypothesis, FGF-23 is not a detrimental factor for the heart but rather a hormone to induce the stress response and protect the body from injurious factors. In some disease states, such as pathologic cardiac hypertrophy, the secretion of FGF-23 from damaged organs may represent a protective response.¹²³ Obviously, however, the continuous stress response leads to organ damage, including cardiac hypertrophy and vascular injury.¹²⁴ Under this scenario, persistent FGF-23/sKlotho disturbance converts protective effects into detrimental effects. Thus, we suspect that the FGF-23/Klotho axis is closely related to the stress response, and that elucidating the detailed relationship of the stress response and FGF-23/Klotho in uremic cardiac hypertrophy may offer a new research direction.

Disclosures

All authors have nothing to disclose.

Funding

A. Li was supported by National Natural Science Foundation of China grants 81270825 and 81770727, Guangdong Province Universities and Colleges Pearl River Scholar Funded Scheme grant 2017, Science and Technology Planning Project of Guangdong Province grant 2017A010103041, and Key Project of Guangzhou Science Technology and Innovation Commission grant 201804020054.