Urinary Klotho Excretion: A Key Regulator of Sodium Homeostasis in Chronic Kidney Disease Stage 2–4

Background

α-Klotho, an anti-aging protein encoded by the klotho gene, is a 130-kDa transmembrane β-glucuronidase mainly expressed in the renal tubule [1]. Soluble α-klotho (hereafter referred to as klotho), a circulating protein shed from the extracellular domain of membrane form klotho by secretase, is present in the blood, urine, and cerebrospinal fluid [2–4]. Although klotho was found to decrease in patients with early chronic kidney disease (CKD) [5,6], there was no correlation between serum klotho and estimated glomerular filtration rate (eGFR) in patients with stages 2 to 4 CKD [7]. Moreover, Akimoto et al showed that urinary klotho levels, but not serum klotho levels, were linked to functional nephron numbers [3]. Urinary klotho was demonstrated to increase in the first day after cardiac surgery and to be a potential early biomarker for acute kidney injury [8]. Therefore, not only the mechanism for the decline of serum klotho but also the role of urinary klotho in CKD patients remain controversial and in need of investigation.

Mounting evidence shows that klotho can regulate electrolyte homeostasis by modifying the ion transport process. Klotho can prevent endocytosis of transient receptor potential cation channel, subfamily V, member 5 [9,10] and renal outer medullary potassium channel 1 [11] to increase renal calcium reabsorption and potassium excretion in the distal renal tubule. Cooperating with fibroblast growth factor receptor (FGF23), klotho can also promote endocytosis of sodium-dependent phosphate cotransporter type IIa to increase phosphate excretion in the proximal renal tubule [12]. Moreover, klotho deficiency was demonstrated to develop salt-sensitive hypertension in klotho haplo-deficiency mice. An upregulation of thiazide-sensitive sodium chloride cotransporter (NCC) and aldosterone synthesis was also observed in klotho-deficient mice [13,14]. However, all these findings were demonstrated either in vitro or in vivo.

Since all of these ion transport proteins are located on the apical membrane (lumen side) of the renal tubular cells, we hypothesized that urine klotho, instead of serum klotho, impacts renal electrolyte handling. In this study, we aimed to investigate the relationships of either serum or urine klotho on renal electrolyte handling in patients with stages 2 to 4 CKD.

Material and Methods

PATIENTS:

This cross-sectional study was approved by the Institutional Review Board for the Protection of Human Subjects at Dalin Tzu Chi General Hospital in compliance with the Helsinki Declaration (IRB: B10501013). The estimated glomerular filtration rate (eGFR) was calculated by using the Chronic Kidney Disease Epidemiology Collaboration equation [15]. Patients with CKD stages 2 to 4 who visited outpatient clinics of nephrology in Dalin Tzu Chi General Hospital in 2016 were screened, and a total of 53 patients were enrolled. All participants had their blood pressure measured after a 10-min rest by using a standard mercury sphygmomanometer. Patients who experienced episodic acute kidney injury, acute coronary syndrome, aortic dissection, acute cerebrovascular stroke, severe infection, or trauma 3 months before enrollment, or who had a life expectancy of less than 3 months were excluded. Patients who had a systolic blood pressure ≥140 mmHg, diastolic blood pressure ≥90 mmHg, or currently used anti-hypertensive medication(s) were regarded as having hypertension. Patients with a fasting plasma glucose level ≥126 mg/dL or who used glucose-lowering agents were regarding as having diabetes mellitus. The use of angiotensin converting enzyme inhibitor/angiotensin II receptor blockers, calcium channel blockers, and diuretics in patients was recorded.

LABORATORY PARAMETERS:

All the fasting blood, spot urine, and 24-h urine specimens were collected between 7 AM and 10 AM for analysis. Blood samples were immediately centrifuged at 3000×g for 10 min and sent for analysis. Blood urea nitrogen, serum creatinine, total cholesterol, low-density lipoprotein cholesterol, triglyceride, high-density lipoprotein cholesterol, albumin, uric acid, sodium, potassium, calcium, phosphorus, magnesium, glucose, hemoglobin (Hb), and HbA1c, were measured by an autoanalyzer (Siemens Advia 1800; Siemens Healthcare GmbH, Erlangen, Germany). Serum intact parathyroid hormone was determined by an enzyme-linked immunosorbent assay (ELISA) kit (Diagnostic Systems Laboratories, Webster, TX, USA). Serum 25 hydroxyvitamin D was measured with an ELISA kit (Crystal Chem, Inc., Chicago, IL, USA). Serum intact FGF23 and C-terminal FGF23 (cFGF23) were measured by an ELISA kit (Immutopics, Inc., San Clemente, CA, USA). Serum and urine klotho levels were determined by ELISA (Immuno-Biological Laboratories Co., Ltd, Gunma, Japan). Creatinine, total protein, klotho, sodium, potassium, calcium, phosphorus, and magnesium were checked in the spot urine sample, and 24-h urine klotho and total protein were measured. The inter- and intra-assay coefficients of variation for intact parathyroid hormone was 6.8% and 7.5%, for intact FGF23 was 6.5% and 2.6%, for cFGF23 was 4.7% and 2.4%, respectively. The coefficient of variation for 25 hydroxyvitamin D was lower than 10%.

We calculated the fractional excretion of each electrolyte, such as sodium, potassium, calcium, phosphate, and magnesium, with the following formula: FEX = [(urine X) × (serum creatinine)]/[(serum X) × (urine creatinine)] (where X represents the electrolyte). Fractional excretion of klotho (FEKlotho) was calculated with the following formula: FEKlotho = [(urine klotho) × (serum creatinine)]/[(serum klotho) × (urine creatinine)].

STATISTICAL ANALYSIS:

Continuous data were evaluated with the Kolmogorov-Smirnov test for normal distribution before statistical testing. Categorical data were expressed as numbers and percentages (%). Continuous data normally distributed were expressed as mean±standard deviation, or for others were expressed as median and 25^th–75^th interquartile range (IQR 25–75). The one-way analysis of variance (ANOVA) test and Kruskal-Wallis H test were used for differences across CKD stages in parametric and nonparametric continuous variables, respectively. The Pearson chi-square test was used for categorical variables. Pearson correlation was used for parametric continuous variables and the Spearman correlation was used for non-parametric or categorical variables. Because klotho, FEKlotho, urine protein-to-creatinine ratio (UPCR), fractional excretion of sodium (FENa), 24-h urine protein, uric acid, fractional excretion of calcium (FECa), C-terminal and intact FGF23, and fractional excretion of potassium (FEK) were skewed, natural logarithmic(ln) transformations were performed to achieve normality. If correlation analyses were statistically significant for klotho, FEKlotho, and FENa, further univariate linear regression analyses were performed. Variables significantly correlated in univariate regression were selected for multivariate stepwise analyses. A P value less than 0.05 was considered statistically significant. All data were analyzed using IBM SPSS 24 (version 24.0, Armonk, NY, USA).

Results

CHARACTERISTICS OF PATIENTS WITH CKD:

Demographic data of these 53 patients, stratified by CKD stages, are shown in Table 1. Among them, 68% were men and the mean age was 71.1±10.5 years. Mean eGFR was 45.5±15.2 ml/min/1.73 m2 and mean 24-h urine protein was 226.8 mg (25th–75th percentage 114.4–838.6 mg). FENa, FEK, fractional excretion of phosphate (FEPi), fractional excretion of magnesium (FEMg), serum intact parathyroid hormone, and cFGF23 were increased, while Hb and 25 hydroxyvitamin D were decreased in patients with advanced CKD. FECa revealed a trend to decline in advanced CKD, but this trend was insignificant. Serum klotho, intact FGF23, and 24-h urine klotho excretion show no differences.

CORRELATIONS BETWEEN SERUM KLOTHO AND OTHER PARAMETERS IN CKD:

When we examined the correlations between serum klotho and other parameters in these patients, our data showed that serum log-klotho was positively correlated with Hb (β=0.271, P=0.049) and negatively correlated with log-uric acid (β=−0.311, P=0.024) and log-FEKlotho (β=−0.352, P=0.01). We performed univariate linear regression analyses by adopting serum log-klotho as a dependent variable and found log-FEKlotho remained the only predictor of serum log-klotho (β=−0.085, P=0.02; Table 2). Our findings suggested that serum klotho level was negatively regulated by FEKlotho in CKD patients, indicating that an increase in urinary klotho excretion resulted in a decline of serum klotho levels.

CORRELATIONS BETWEEN FENA AND CLINICAL PARAMETERS:

Furthermore, we also conducted correlation analyses between FENa and other parameters. Correlation analysis shows FENa was positively correlated with FEK (β=0.66, P<0.0001), FECa (β=0.344, P=0.012), FEPi (β=0.692, P<0.0001), FEMg (β=0.766, P<0.0001), cFGF23 (β=0.377, P=0.005), FEKlotho (β=0.48, P<0.0001), and UPCR (β=0.318, P=0.02), while negatively correlated with eGFR (β=−0.528, P<0.0001), Hb (β=−0.464, P=0.00047) (data are not shown). The results of univariate and multivariate analyses with log-FENa as a dependent variable are shown in Table 3. Multivariate stepwise regression after adjustment of eGFR, Hb, log-FEK, log-FECa, FEPi, FEMg, log-UPCR, log-cFGF23, and log-FEKlotho showed log-FENa was positively correlated with log-FEKlotho (β=0.138, P=0.032), FEPi (β=0.027, P=0.002), and FEMg (β=0.133, P<0.0001). FENa was positively regulated by FEKlotho, suggesting that an increase in urinary klotho excretion can enhance urinary sodium wasting.

Discussion

The findings from our cross-sectional study in patients with CKD stages 2 to 4 are summarized as followed. First, serum klotho is inversely correlated with FEKlotho, indicating the increase in urinary klotho excretion can result in a decline of serum klotho levels. Second, FENa is positively correlated with FEKlotho, implying that an increase in urinary klotho excretion can enhance urinary sodium wasting.

In agreement with our findings in CKD patients, Hu et al found a negative correlation between FEKlotho and serum klotho levels in their animal studies. They demonstrated that the kidney regulates serum klotho levels either by shedding klotho from the renal tubule into the blood or by transcytosis from the basolateral to the apical side of the renal proximal tubule [2], indicating that the kidney is responsible not only for klotho production but also for klotho clearance from circulation.

Many studies have shown that circulating klotho decreased in advanced CKD [5,6,16–18]. However, in contrast, there was no decrease in serum klotho in our CKD patients. In addition, Seiler et al [7] also failed to find this association in advanced CKD. Accordingly, we hypothesized that there are many other factors beyond eGFR, such as renal tubule dysfunction, daily protein excretion, and electrolyte imbalance, that can also influence serum klotho level.

The regulation of sodium excretion was achieved by controlling transporters in renal tubules. NCC is a major sodium transporter channel located in the apical membrane of the distal renal tubule [19]. The abundance and expression of NCC in the apical membrane results in sodium reabsorption. Zhou et al showed that the expression of NCC was upregulated in klotho mutant heterozygous mice rather than in the wild-type mice. Although the functional changes of NCC were not assessed in their study, their finding still implied that klotho repressed the expression of NCC and therefore led to urine sodium wasting [13]. Our study, for the first time, demonstrated the relationship between urine klotho and urinary sodium excretion. However, the mechanism of how klotho impacts NCC still needs to be elucidated. Another study by Zhou et al showed klotho gene deficiency can also upregulate CYP11B2 expression in the adrenal cortex, thereby increasing aldosterone synthesis. Aldosterone can further enhance sodium reabsorption via the decreased abundance of NCC [14].

To the best of our knowledge, the present study is the first study to show that FEKlotho, as a surrogate of urine klotho elimination, is associated with urine sodium excretion in CKD patients. Still, this study had several limitations. First, to investigate the impact of eGFR on serum klotho level, we failed to include patients with all stages of CKD and healthy individuals as control. Second, the single-center study design and a limited number of patients create selection bias in our study. Third, we did not evaluate the dietary phosphate, calcium, and sodium content in our patients.

Conclusions

In conclusion, our study indicated that FEKlotho is negatively correlated with serum klotho levels and positively correlated with FENa, indicating an increase in urinary klotho excretion can result in a decline of serum klotho levels and an increase in sodium wasting. These findings imply that urine excretion of klotho determines both klotho and sodium homeostasis in patients with stages 2 to 4 CKD.