Fibroblast Growth Factor 23 and αKlotho in Acute Kidney Injury: Current Status in Diagnostic and Therapeutic Applications

Fibroblast growth factor (FGF) 23 was cloned as a candidate gene of autosomal dominant hypophosphatemic rickets and a differentially expressed gene in tumor-induced osteomalacia, which ushered in a major hormone countering phosphate accumulation [1]. Beyond phosphate balance, FGF23 has been shown or postulated to regulate calcium and sodium homeostasis, iron metabolism, erythropoiesis, and inflammation [2].

αKlotho was cloned when serendipitous disruption of its locus led to premature multiorgan failure [3]. The list of cellular actions conferred by αKlotho encompasses -anti-oxidation, anti-apoptosis, pro-autophagy, pro-stem cell, and anti-fibrosis and essentially broad-based maintenance of cell health [1].

Chronic kidney disease (CKD) is a state of FGF23 excess and αKlotho deficiency, but their levels and roles in acute kidney injury (AKI) are less well defined. Many of the functions of FGF23 and αKlotho are highly relevant to the pathobiology of AKI. AKI is a systemic syndrome triggered by a sudden loss of kidney function and brings forth significant morbidity, mortality, and burden on multiple organ systems [4, 5]. Even when one evades the acute high mortality, survivors are at risk for CKD, end-stage kidney disease, and cardiovascular disease [6-9].

For a common serious syndrome with dire short- and long-term consequences, progress in diagnosis, prognosis, and therapy has been modest. We need means to diagnose AKI early, foresee the time course of recovery, and predict the risk of progression to CKD. For therapeutics, we need effective prophylaxis of AKI in high-risk situations, accelerate recovery in established AKI, prevent extrarenal complications, and forestall AKI-to-CKD transition. We will explore if FGF23 and αKlotho can fulfill these roles.

Multiple factors increase FGF23 production, including high blood pressure, phosphate loading, iron deficiency, hypoxia, and inflammation [10]. FGF23 actions are transduced by the FGF receptor 1 (FGFR1) [11, 12] and its coreceptor transmembrane αKlotho [13]. FGF23 can also exert action through binding to FGFR4 in the heart and liver, independently of αKlotho [14, 15]. Metabolism of FGF23 includes cleavage of bioactive intact FGF23 (∼30 kDa) into C-terminal (∼12 kDa) and N-terminal (∼18 kDa) fragments. FGF23 seems to be also cleared by the kidneys, possibly through filtration and/or catabolism [16]. Circulating FGF23 includes the intact hormone (iFGF23) and the N- and C-termini. In states such as iron deficiency and inflammation, increased production of FGF23 is matched by FGF23 cleavage, with high levels of FGF23 fragments in circulation [17]. Two assays used to measure circulating FGF23 levels are “C-terminal” (cFGF23) which detects both C-terminal fragments and the full-length peptide and “intact” (iFGF23) which measures only intact FGF23.

Elevated FGF23 levels have been observed in multiple studies of human AKI [10]. Plasma cFGF23 levels were 5.6-fold higher in patients with AKI versus age-matched patients without AKI [18]. Note that increased FGF23 levels detected with the cFGF23 assay (intact + C-term) may reflect increased FGF23 production and/or impaired clearance but do not inform about FGF23 bioactivity. Few studies have used both cFGF23 and iFGF23 assays in conjunction. In adult patients undergoing cardiac surgery, plasma cFGF23 levels were markedly increased (∼100-fold) postoperatively in patients who did versus did not develop AKI, while iFGF23 levels were modestly (∼2-fold) higher postoperatively in AKI versus non-AKI patients [19]. Similar findings were seen in a folic acid nephropathy mouse model [20]. Therefore, both production and clearance of FGF23 may be affected in AKI.

In 250 adult patients undergoing cardiac surgery, plasma cFGF23 levels differentially increased at the end of cardiopulmonary bypass in patients with versus without postoperative AKI, predating changes in other mineral metabolites [19]. Performance evaluation of utility of cFGF23 for early AKI diagnosis was modest (AUC = 0.78) but superior to urinary kidney injury biomarkers [19]. Similar observations were derived from cohorts of critically ill patients in which cFGF23 was measured in plasma and urine within 24–48 h of ICU admission [21, 22].

FGF23 levels may also have prognostic utility in AKI. In a large post hoc analysis (N = 1,527 patients), increased risk of 60-day mortality was noted in patients with the highest versus lowest quartiles of cFGF23 (∼3.8-fold) and iFGF23 (∼2.0-fold) [23].

The canonical mineral metabolism regulators of FGF23 production in bones seem not to play a major role in AKI [20, 24]. Furthermore, contribution of FGF23 production from other organs in acute injury/inflammation settings is not fully understood. FGF23 may augment myofibroblast activation and fibrosis via TGF-β-related pathways, opening an angle for therapeutic FGF23 blockade [25, 26]. However, off-target effects of FGF23 such as impairment of immune [27, 28], endothelial [29], and cardiac functions [15] are contenders and require further investigation. Therefore, it is still unclear if anti-FGF23 antibody (burosumab), C-terminal FGF23 (experimentally shown to interfere with FGF23 signaling at supraphysiological concentrations) [30], or specific FGFR4 inhibition have clinical applications in attenuation of incident AKI or its consequences.

αKlotho deficiency is universal in AKI animal models including ischemia-reperfusion injury (IRI) [31], unilateral ureteric obstruction [32] (UUO), sepsis induced by lipopolysaccharide (LPS) injection or cecal ligation and puncture (CLP) [33, 34], and nephrotoxins including cisplatin [35] (CP) and folic acid [36] (FA), indicating that αKlotho downregulation in the kidney is a general phenomenon after acute kidney insults (Table 1). In rats, αKlotho mRNA started to fall 6 h post-IRI-AKI and returned to baseline around 3-4 days [37]. αKlotho protein in the kidney fell at 3 h while neutrophil gelatinase-associated lipocalin (NGAL) did not increase until 5 h, so the kidney αKlotho protein is an early marker of AKI [38] (Table 1). Urine and plasma αKlotho paralleled changes in kidney αKlotho [38], decreasing dramatically at 3 h, started to recover at ∼48 h, and reached normal levels by 7 days. Soluble αKlotho may be a surrogate for kidney αKlotho.

The first human AKI study showed significantly lower urine αKlotho at the time of AKI diagnosis when compared with healthy controls [38]. A larger prospective ICU study showed that AKI patients had lower urinary αKlotho within 48 h of AKI diagnosis when compared with ICU controls [39]. Urine αKlotho concentration and urine αKlotho-to-Cr ratio inversely correlated with hospital and mechanical ventilation days. Each 1-fold higher urine αKlotho/Cr was associated with an 83% lower risk of major adverse kidney events at 90 days [39].

Heterozygous αKlotho-deficient (kl/+) mice subjected to IRI, CP, sepsis, and UUO had lower αKlothoprotein and more kidney damage than wild-type (WT) mice [34, 35, 37, 38] (Table 1). Longer ischemia time caused more severe αKlotho deficiency and kidney fibrosis. After recovery of kidney function, lower αKlotho associated with more chronic fibrosis and kidney dysfunction [40], suggesting αKlotho level can predict risk of AKI-to-CKD transition (Table 1).

Mice with genetically high αKlotho before AKI had milder kidney damage and a better renal outcome in IRI [38] and CP-induced AKI [31]. Adenoviral αKlotho gene delivery increased circulating but not kidney αKlotho prior to IRI, ameliorated injury, and kidney dysfunction [37]. αKlotho-bearing minicircle vector increased plasma αKlotho, prevented kidney damage induced by IRI, and attenuated kidney fibrosis in an UUO model [41]. αKlotho overexpression through CRISPR-Cas9 prevented kidney damage and alleviated kidney fibrosis in CP nephrotoxicity [42] and in UUO [43] (Table 2).

Clinical utility mandates efficacy if given after AKI. Recombinant αKlotho protein given to mice immediately after IRI reduced kidney damage [38], but the benefit diminishes dramatically starting 1 h after the insult. αKlotho administration immediately after UUO followed up to 14 days reduced renal fibrosis, but not hydronephrosis severity [44]. αKlotho-carrying extracellular vesicles [45] promoted kidney recovery in rhabdomyolysis-induced AKI. At 24 h after IRI, when kidney injury was fully established, serum creatinine had peaked, and endogenous αKlotho was at its lowest levels, αKlotho protein injection for 4 consecutive days preserved endogenous kidney αKlotho levels, accelerated recovery, suppressed kidney fibrosis, and protected against AKI-to-CKD transition [40] (Table 2).

In 2 CKD models, αKlotho protein was given 4 weeks after CKD induction and sustained for 12 weeks [46], or up to 16 weeks [35]. Plasma αKlotho protein levels, kidney function and fibrosis, and cardiac remodeling/cardiomyopathy were improved possibly via both direct effects and secondary effects due to renoprotection (Table 2).

The database of FGF23 is currently composed largely of human studies, which are strong but only associative in nature. In contrast, the αKlotho database is populated by compelling animal experiments but the clinical data are still in early stages of development. Regarding this pair of proteins with mineral-regulating properties and a plethora of other actions, one can envision many potential applications in human AKI (Fig. 1; Table 3). The potential applications are immense, and further research should be directed at the pathobiology of FGF23 in AKI in preclinical studies to dissect whether it is pathogenic or a biomarker (or both), that is, interventional experiments. Prospective longitudinal clinical studies using simultaneous iFGF23 and cFGF23 assays will enrich our database. Standardization of soluble αKlotho assays will enrich the human database and make it generalizable. The extensive preclinical data in αKlotho are in dire need of human translation in both diagnostics and therapeutics but one faces various hurdles in this effort.

The authors are supported by the National Institutes of Health (R01-DK091392, R01-DK092461, and R01-DK092461-S1 to OWM and MCH), the George O’Brien Kidney Research Center (P30-DK-07938 to OWM), the Charles and Jane Pak Center Innovative Research Support (to OWM and MCH), and Endowed Professor Collaborative Research Support (to OWM and MCH). JAN is a recipient of an Early Career Pilot Grant from the National Center for Advancing Translational Sciences, National Institutes of Health, through Grant UL1TR001998. The authors are grateful to the excellent secretarial assistance provided by Ms. Yesenia Aguirre.

The authors have no conflicts of interest to disclose.

Javier A. Neyra and Ming Chang Hu contributed equally.Contribution from the AKI and CRRT 2020 Symposium at the 25th International Conference on Advances in Critical Care Nephrology, Manchester Grand Hyatt, San Diego, CA, USA, February 24–27, 2020. This symposium was supported in part by the NIDDK funded University of Alabama at Birmingham-University of California San Diego O’Brien Center for Acute Kidney Injury Research (P30DK079337).