Klotho deficiency is an early biomarker of renal ischemia–reperfusion injury and its replacement is protective

Abstract

Klotho is an antiaging substance with pleiotropic actions including regulation of mineral metabolism. It is highly expressed in the kidney and is present in the circulation and urine but its role in acute kidney injury (AKI) is unknown. We found that ischemia–reperfusion injury (IRI) in rodents reduced Klotho in the kidneys, urine, and blood, all of which were restored upon recovery. Reduction in kidney and plasma Klotho levels were earlier than that of neutrophil gelatinase-associated lipocalin (NGAL), a known biomarker of kidney injury. Patients with AKI were found to have drastic reductions in urinary Klotho. To examine whether Klotho has a pathogenic role, we induced IRI in mice with different endogenous Klotho levels ranging from heterozygous Klotho haploinsufficient, to wild-type (WT), to transgenic mice overexpressing Klotho. Klotho levels in AKI were lower in haploinsufficient and higher in transgenic compared with WT mice. The haploinsufficient mice had more extensive functional and histological alterations compared with WT mice, whereas these changes were milder in overexpressing transgenic mice, implying that Klotho is renoprotective. Rats with AKI given recombinant Klotho had higher Klotho protein, less kidney damage, and lower NGAL than rats with AKI given vehicle. Hence, AKI is a state of acute reversible Klotho deficiency, low Klotho exacerbates kidney injury and its restoration attenuates renal damage and promotes recovery from AKI. Thus, endogenous Klotho not only serves as an early biomarker for AKI but also functions as a renoprotective factor with therapeutic potential.

Acute kidney injury (AKI) is a formidable clinical problem resulting primarily from ischemia or nephrotoxins, with outcomes ranging from full recovery to the development of chronic kidney disease with dialysis dependence, or death.^1,2 Epidemiological data estimates an annual incidence of 22–620 per million with overall immediate mortality of up to 23 and 79%.³ This unacceptably high mortality has improved only modestly despite advances in our understanding of the pathogenesis and improvements in dialysis. Even if patients survive the acute insult, long-term prognosis after AKI is still far from optimal,⁴ with survival rates projected as 46–74, 55–73, 57–65, and 65–70% at 90 days, 6 months, 1 year, and 5 years, respectively.⁵ The annual incidence of end-stage renal disease from AKI survivors has increased from 0.4 to 4.9 per 100,000 between 1998 and 2002.⁶ Patients who had an episode of AKI are at higher risk for the development of chronic kidney disease.⁷ The poor outcome can be attributable to both delayed diagnosis due to the lack of sensitive and specific early biomarkers and limited efficacious specific therapies.

Klotho was originally identified as an aging suppressor gene^8,9 with pleiotropic functions. It encodes a single-pass transmembrane protein and is expressed in the kidney, where it functions as a coreceptor for fibroblast growth factor-23 (FGF23) and has a critical role in phosphate metabolism.^10–12 The extracellular domain of Klotho is cleaved on the cell surface by membrane-anchored proteases, and released into blood,^13–15 urine,^16,17 and cerebrospinal fluid.¹⁵ Secreted Klotho protein has multiple functions distinct from those of membrane Klotho, including regulation of multiple ion channels^16–19 and oxidative stress.²⁰ Recent studies demonstrated that Klotho might also be renoprotective in chronic glomerulonephritis.²¹ Data on Klotho in AKI are limited. One study showed that renal Klotho mRNA and protein is decreased 1 day after ischemia reperfusion and adenoviral delivery of Klotho led to modest 15% decrease in serum creatinine (Cr) concentration after ischemia.²² It will be important to know whether there is a systemic deficiency of Klotho in AKI and what is the time profile of the decrease. Furthermore, whether lower Klotho expression makes animals more susceptible to ischemic injury has not been addressed. This study was conducted to definitively address four unanswered questions: (1) Is AKI a temporary state of systemic Klotho deficiency? (2) Is decreased Klotho in blood and urine an early biomarker of AKI? (3) Do Klotho expression levels affect the outcome and course of AKI? Most importantly, (4) Does Klotho administration ameliorate kidney damage in AKI and/or promote recovery from AKI?

RESULTS

AKI from IRI is an acute state of Klotho deficiency

After bilateral renal arterial occlusion (30 min in mice; 40 min in rats), all animals showed at least two- to threefold increase in blood urea nitrogen and/or plasma Cr above baseline (Figure 1, Table 1) with no mortality before termination of study (<10 days after ischemia–reperfusion injury (IRI)). Plasma Cr peaked on day 1, started to decline on day 2, but still remained higher on day 7 compared with Sham rats (Figure 1a). Endogenous Cr clearance dropped dramatically on day 1 and started to recover on day 2, (Figure 1b). IRI also affected renal tubular function as evident by the reduced urinary osmolality and increased fractional excretion of Na (Figure 1c and d). Histology on day 2 showed widespread protein casts in dilated tubules (Figure 1e). Proximal tubular cells lost their brush border membrane and nuclei, and became flat and eosinophilic, compatible with cell necrosis. There were erythrocyte congestion and some intrarenal extravasation in the medulla.

Figure 1. Characterization of acute kidney injury (AKI) in rats.

AKI and Sham rats were killed at 1, 2, and 7 days after ischemia–reperfusion injury (IRI) and blood and urine were collected. (a) PCr, plasma creatinine; (b) Cl_Cr, creatinine clearance; (c) U/Posm, urine-to-plasma osmolality; and (d) FE_Na, fractional excretion of Na⁺. Values are expressed as means±s.e.m. of six rats in each group. Statistical differences for a, b, c, and d on days 0, 1, 2, and 7 were analyzed by analysis of variance (ANOVA) followed by Student–Newman–Keuls test, and significant differences were accepted when *P<0.05, **P<0.01 vs Sham group; ^#P<0.05; ^##P<0.01 vs AKI rats at day 0. Kidneys were harvested for hematoxylin and eosin (H and E) staining (e, top panel) on days 1, 2, and 7 after IRI. Kidney histological scores (e, bottom panel) from H and E staining (e, top panel) were obtained by a renal pathologist blinded to the experimental conditions using a semi-quantitative pathological scoring system as described in Methods section. Scores are expressed as mean±s.e. of four samples in each group. Statistical differences were analyzed by ANOVA followed by Student–Newman–Keuls test, and significant differences were accepted when P<0.05 or less between two groups. A, AKI; S, Sham.

Table 1.

Plasma and urine chemistry of AKI rats on day 2 after IRI

	Sham	AKI
Body weight (g)	323.1±10.2	331.5±11.5
Cr (mg/dl)	0.43±0.08	2.31±0.45**
BUN (mg/dl)	27.3±7.3	104.1±13.5**
ClCr (ml/min/Kg)	1.20±0.23	0.29±0.05*
U/P osmolality	5.32±1.03	2.47±0.55**
FENa (%)	0.11±0.03	0.62±0.11*

Abbreviations: AKI, acute kidney injury; BUN, blood urea nitrogen; Cr, creatinine; Cl_Cr, Cr clearance; FE_Na, fractional excretion of Na; IRI, ischemia–reperfusion injury; U/P, urine-to-plasma.

Data are represented as means±s.e.m.; n=4. Significant differences were accepted when *P<0.05, **P<0.01 vs Sham; by unpaired Student’s t-test.

The kidney has the highest expression of Klotho among all organs in the rodent^8,14 and is likely a major source of circulating Klotho. Both Klotho protein (Figure 2a and b) and mRNA (Figure 2c) were reduced in kidneys of AKI rats on days 1 and 2. Previous studies never examined Klotho levels in plasma or urine in AKI. Similar to the low erythropoietin state in blood in acute kidney diseases from a variety of insults,²³ we found substantial decrease in both plasma and urine Klotho protein in AKI rats, suggesting a state of ‘pan-Klotho-deficiency’ (Figure 2d and e). By the seventh day, Klotho in the kidney, plasma, and urine of AKI rats returned to the levels comparable with Sham rats (Figure 2a–e), indicating that AKI induces acute and transient systemic Klotho deficiency. Because the human plasma Klotho assay was not ready and access to kidney biopsies were understandably limited at the time of the study, only urine Klotho levels were measured. Fresh urine samples of 17 AKI patients (Table 2) had notably lower Klotho protein than normal healthy volunteers (Figure 2f, Supplementary Figure S1 online, Table 3).

Figure 2. Klotho levels in acute kidney injury (AKI) rats from 1 to 7 days post-ischemia–reperfusion injury (IRI).

(a) Representative immunohistochemistry for Klotho protein in kidneys from three rats in each group on days 1, 2, and 7 post-IRI. Four μm frozen kidney sections were stained for Klotho with rabbit anti-Klotho (KMDC)⁴⁹ (1/50 dilution) and for β-actin. (b) Representative immunoblot (30 μg protein per lane) for Klotho in total kidney lysate from AKI and Sham rats at days 1, 2, and 7 after IRI with rat anti-Klotho antibody (KM2076).⁴⁹ The upper panel is a representative blot. The lower panel is a summary of densitometric analysis of all samples from Sham and AKI rats. (c) Klotho transcripts in kidney. One μg of total RNA from kidneys of AKI and Sham rats on day 1, 2, and 7 after IRI was used to generate complementary DNA using oligo-dT primers and real-time PCR for Klotho and cyclophilin was performed. Quantification of Klotho transcripts was calculated as described in Methods section. (d) Representative immunoblot for immunoprecipitated plasma Klotho. One hundred μl of heparinized blood from AKI and Sham rats on days 1, 2, and 7 were immunoprecipitated (IP) with 4 μl of rabbit anti-Klotho antibody (KMDC), and the immune complex was immunoblotted with rat anti-Klotho antibody (KM2076, 1/3000 dilution). The band of immunoglobulin heavy chain (IgG-HC) was shown for IP control. The upper panel is a representative blot. The lower panel is summary of densitometric analysis of all samples from Sham and AKI rats. Values are expressed as means±s.e.m. of six rats in each group. Statistical differences were analyzed by analysis of variance (ANOVA) followed by Student–Newman–Keuls test and significant differences were appreciated when **P<0.01 vs Sham group; ^#P<0.05; ^##P<0.01 vs AKI group at day 1; ^£P<0.05; ^££P<0.01 vs AKI group at day 2. (e) Representative immunoblot for urinary Klotho from rats a volume of fresh urine containing 10 μg of creatinine was immunoblotted with rat anti-Klotho antibody (KM2076). Lane 1 was loaded with 20 fmoles of recombinant mouse Klotho protein (rMKl). The top panel is a representative blot for Klotho in urine samples of AKI rats at days 0, 1, 2, and 7. The bottom panel is summary of densitometric quantification of data from immunoblot. (f) Representative immunoblot for Klotho protein in urine samples of normal subjects (open circle) and of AKI patients (closed circle). A volume of concentrated human urine containing 50 μg of creatinine was loaded. Top panel shows membrane blotted with rat anti-Klotho antibody. Values for individual subjects are shown in bottom panel. Values are expressed as means±s.e.m. of six rats in each group. Statistical differences for b, c, and e were analyzed by ANOVA followed by Student–Newman–Keuls test, and significant differences were accepted when P<0.05 or less between two groups.

Table 2.

Klotho protein in urine of AKI patients

	Normal subject	AKI
Age (year)	47.7±3.1	44.8±4.3
(n)	(14)	(17)
Gender (F/M)	7/7	7/10
Cr (mg/dl)#	0.73±0.05	3.76±0.58**
Urinary Klotho (pmoles/l)	20.66±1.81	2.52±0.76**
Urinary Klotho/Cr (fmoles/mg)	25.38±4.08	4.85±1.69**

Abbreviations: AKI, acute kidney injury; Cr, creatinine.

^#Blood creatinine when AKI patients’ urines were collected for urinary Klotho determination. Significant differences were accepted when *P<0.05, **P<0.01 vs normal subjects by unpaired Student’s t-test.

Table 3.

Clinical profile and urinary Klotho protein in AKI patients

ID	Age (years)	Gender	Potential contributors to AKI													Scr (mg/dl)		Klotho in urine		Outcome
			Pre- renal	Sepsis	Nephro- toxin	Lupus	Hyper- tension	DN	HUS	OU	Solitary kidney	CKD	Graft rejection	Pre- eclamp- sia	HRS	Base- line	Peak	Klotho (pmoles/l)	Klotho/ Cr (fmoles/mg)	Short term <3 months	Scr (mg/dl)
IN-1	78	F	+													0.89	2.73	1.31	1.28	FR	0.87
IN-2	41	M			+											1.17	3.77	5.90	10.41	FR	0.84
IN-3	53	M		+												1.10	2.43	3.34	4.96	FR	0.97
IN-4	34	M	+													0.96	1.79	10.27	19.75	PR	1.49
IN-5	21	F							+							1.0	3.17	0	0	Death	4.67
IN-6	37	F		+												0.36	2.62	6.87	20.09	FR	0.62
IN-7	68	M								+						0.75	5.57	4.9	3.2	PR	1.33
IN-8	68	M			+											0.49	5.58	0.15	0.12	FR	0.81
IN-9	55	M			+											1.05	9.76	2.1	2.10	PR	4.78
IN-10	22	F				+										0.68	1.71	0.10	0.07	Death	0.84
IN-11	55	F													+	1.2	2.25	0.30	0.60	Death	3.5
IN-12	20	F												+		0.89	1.38	7.34	9.38	PR	1.29
IN-13	60	M			+		+	+			+	+				1.98	4.70	0	0	Death	2.90
IN-14	49	F	+													0.80	4.05	0	0	CKD4	3.64
IN-15	28	M											+	+		2.20	7.89	0.28	0.56	CKD4	4.5
IN-16	48	M		+												0.77	1.92	2.97	6.51	Death	2.86
IN-17	55	M			+											1.10	1.86	7.9	12.8	FR	0.98

Abbreviations: AKI, acute kidney injury; CKD, chronic kidney disease; DN, diabetic nephropathy; FR, full recovery; HRS, hepatorenal syndrome; HUS, hemolytic uremic syndrome; OU, obstructive uropathy; PR, partial recovery; Scr, serum creatinine.

Klotho deficiency is an early biomarker for AKI induced by IRI

To explore the time course of changes in Klotho, we measured plasma and renal Klotho in very early stages (1, 3, 5, and 24 h post-IRI) of AKI. Plasma Cr was slightly increased at 5 h, which eventually reached statistical significance after 24 h (Supplementary Figure S2A online). Mild alteration of kidney morphology was detectable at 5 h post-reperfusion with slight degeneration in proximal tubules and mild hemorrhage in the interstitium at 5 h. After 24 h, there were marked necrosis and loss of brush border membrane in proximal tubules with higher kidney pathological scores (Supplementary Figure S2B,C online). Interestingly, Klotho levels in plasma and kidney were moderately increased at 1 h post-reperfusion, but immediately and sustainably decreased after 3 h (Figure 3a and b and Supplementary Figure S2D online). Thus, changes in blood and renal Klotho levels preceded changes in plasma Cr or kidney histology (Figure 3 and Supplementary Figure S2A-D online). Plasma and kidney neutrophil gelatinase-associated lipocalin (NGAL) levels started to rise 5 h after reperfusion (Figure 3c and d and Supplementary Figure S2E online). Thus, Klotho might be one of the earliest biomarkers for kidney injury, at least for AKI induced by IRI.

Figure 3. Klotho and neutrophil gelatinase-associated lipocalin (NGAL) in blood and kidney in acute kidney injury (AKI) rats from 1 to 24 h.

(a) Representative immunoblot for immunoprecipitated (IP) Klotho (upper panel) in blood of AKI and Sham rats at 1, 3, 5, and 24 h post-reperfusion (three rats at each time point); immunoglobulin G (IgG) heavy chain (IgG-HC) band is shown for IP control. One hundred μl of heparinized blood was immunoprecipitated with rabbit anti-human Klotho antibody (KMDC) and immunoblotted for Klotho protein with rat anti-Klotho antibody (KM2076). (b) Representative immunoblot for Klotho in total kidney lysates of AKI and Sham rats at 1, 3, 5, and 24 h post-reperfusion (three rats at each time point). Thirty μg of protein were loaded in each lane and subjected to SDS-polyacrylamide gel electrophoresis (PAGE) followed by immoblotting with rat anti-Klotho antibodies. (c) Representative immunoblot for NGAL in blood of AKI and Sham rats at 1, 3, 5, 24 h post-reperfusion (three rats at each time point). Four μl of heparinized blood from rats were mixed with Laemle’s buffer and subjected to immunoblot with goat anti-NGAL antibody (1/200 dilution) (Santa Cruz Biotechnology, Santa Cruz, CA, USA). (d) Representative immunoblot for NGAL in total kidney lysates of AKI and Sham rats at 1, 3, 5, and 24 h post-reperfusion (three rats at each time point). Thirty μg of protein were loaded in each lane and subjected to SDS-PAGE followed by immoblotting with goat anti-NGAL antibody.

Klotho deficiency accentuates and Klotho overexpression attenuates AKI

If AKI is a state of acute Klotho deficiency, the logical and most critical question is whether this is merely a marker of renal damage or a contributor to renal dysfunction. To address this, we induced AKI in mice with three different Klotho expression levels. It was impossible to induce experimental AKI in the homozygous knockout mice (Kl^−/−) because of 100% surgical mortality so we compared Klotho haploinsufficient (Kl ^+/−) mice with their wild-type (WT) counterparts. Kl^+/− mice had lower levels of Klotho protein in the plasma, kidney, and urine at baseline, which became undetectable after AKI (Figure 4a–d). Kl ^+/− mice developed more severe renal dysfunction (Table 4) and more extensive histological abnormalities (Supplementary Figure S3C,D online), such as tubular casts, tubule dilation, loss of brush border membrane, and interstitial hemorrhage than WT mice (Supplementary Figure S3G,H online).

Figure 4. Klotho protein expression in acute kidney injury (AKI) mice with different genetic background.

(a) Representative immunoblot for Klotho in total kidney lysate from four mouse kidneys each group at day 2 post-ischemia–reperfusion injury (IRI) (upper panel). Thirty μg of protein was immunoblotted for Klotho protein with rat anti-Klotho antibody. Lower panel is a summary of densitometric quantification of the data. (b) Representative fluorescent immunohistochemistry for Klotho in kidney sections of four mice in each group at day 2 post-IRI. Four μm of frozen kidney sections were stained for Klotho with rat anti-Klotho antibody. (c) Plasma Klotho levels. One hundred μl of heparinized blood from mice on day 2 were incubated with rabbit anti-Klotho antibody to precipitate Klotho from the blood. The immune complex was subjected to immunoblot with rat anti-Klotho antibody. Upper panel is a representative blot for Klotho and immunoglobulin G heavy chain (IgG-HC). Lower panel is summary of densitometric analysis of all samples. Values are expressed as means±s.e.m. of four mice in each group. Statistical analysis was conducted similar to that shown in (a). (d) Representative immunoblot for Klotho in urine from four mice in each group. On day 2, a volume of fresh urine containing 10 μg of creatinine was blotted using rat anti-Klotho antibody (upper panel). Lower panel is summary of densitometric analysis of immunoblots for urine Klotho from four mice in each group. (e) Kidney histological scores. Scores are expressed as mean±s.e. of six samples in each group. Statistical differences on day 2 were analyzed by analysis of variance followed by Student–Newman–Keuls test, and significant differences were accepted when P<0.05 or less between groups. Kl^+/−, Klotho haploinsufficiency; WT, wild type; and Tg-Kl, transgenic overexpression of Klotho.

Table 4.

Kidney function in AKI mice of Kl^+/− group

	WT		Kl+/−
	Sham	AKI	Sham	AKI
Cr (mg/dl)	0.13±0.03	0.29±0.06*	0.17±0.10	0.33±0.05**
ClCr (μl/min per g)	12.80±2.01	4.84±1.63**	15.50±1.80	3.34±0.12*,#

Abbreviations: AKI, acute kidney injury; Cr, creatinine; Cl_Cr, clearance creatinine; WT, wild type.

Data at day 2 are represented as mean±s.e.m.; n=8. Significant differences were accepted when *P<0.05, **P<0.01 vs Sham; ^#P<0.05 vs WT AKI by analysis of variance followed by Student–Newman–Keuls test.

Transgenic Klotho-overexpressing (Tg-Kl) mice had higher renal (Figure 4a and b), plasma (Figure 4c), and urinary Klotho (Figure 4d) at baseline. Upon induction of ischemia–reperfusion, Klotho protein decreased but still maintained levels comparable with those in Sham WT mice (Figure 4a–d). Tg-Kl mice were more resistant to injury (Table 5) with fewer casts in the renal tubules and almost intact brush border membrane in proximal tubules (Supplementary Figure S3K,L online) compared with WT AKI mice (Supplementary Figure S3G,H online). Kidney histological scores were significantly lower in AKI Tg-Kl mice and higher in AKI Kl ^+/− mice compared with their WT counterparts (Figure 4e). The histological findings in Tg-Kl mice are similar to the results from one previous study, in which Klotho was increased using viral gene delivery.^22,24 It is noteworthy that Tg-Kl mice at baseline did not have higher Cr clearance than Sham WT mice (Table 5), so the beneficial effect of Klotho overexpression on renal IRI is not simply a ‘glomerular filtration rate enhancing’ effect.

Table 5.

Kidney function in AKI mice of Tg-Kl group

	WT		Tg-Kl
	Sham	AKI	Sham	AKI
Cr (mg/dl)	0.27±0.01	0.48±0.01*	0.33±0.03§	0.35 ±0.04*,#
ClCr (μl/min per g)	14.23±2.14	5.65±0.57**	9.68±3.19§	7.49 ±0.50*,#

Abbreviations: AKI, acute kidney injury; Cr, creatinine; Cl_Cr, clearance creatinine; WT, wild type.

Data at day 2 are represented as mean±s.e.m.; n=8. Significant differences were accepted when *P<0.05, **P<0.01 vs Sham; ^#P<0.05 vs WT AKI; ^§P<0.05 vs WT Sham by analysis of variance followed by Student–Newman–Keuls test.

Klotho administration post-ischemia rescues rat kidney from AKI

In clinical settings, practitioners rarely have the chance to anticipate AKI but are usually confronted with the syndrome after established injury. Thus, any meaningful therapy has to be efficacious when given after AKI develops. We next tested whether administration of Klotho after AKI may still exert therapeutic benefits in AKI. We administered recombinant Klotho protein (soluble extracellular domain of Klotho protein at 0.01 mg/kg body weight) intraperitoneally 30 min after the ischemic injury. Single bolus injection of Klotho protein 30 min post-injury was effective in attenuating AKI (Figure 5a–f). Klotho treatment post-IRI significantly lessened the elevation of plasma Cr and blood urea nitrogen, and decline in Cr clearance (Figure 5a–c) on first 2 days compared with vehicle injection. The abnormal blood and urinary parameters returned to near basal levels (Figure 5a–c) after 7 days of IRI.

Figure 5. Klotho administration post-ischemia–reperfusion injury (IRI) in acute kidney injury (AKI) rats.

AKI and Sham rats treated with Klotho (Kl) or vehicle (Veh) 30 min post-reperfusion were killed at 0, 1, 2, and 7 days after IRI and blood and urine were collected for plasma creatinine (Cr) (a), blood urea nitrogen (BUN) (b), Cl_Cr (Cr clearance) (c), urine excretion rate (d), and urinary osmolality (e). Data are expressed as mean±s.e.m. (n = 4) in each group. Statistical significance was analyzed by analysis of variance (ANOVA) followed by Student–Newman–Keuls test, for data presented in a–e and significant differences were accepted when *P<0.05, **P<0.01 vs Sham-vehicle rats; ^#P<0.05; ^##P<0.01 vs AKI-vehicle rats; ^§P<0.05, ^§§P<0.01 vs Sham-Klotho rats. Sprague–Dawley rats were killed on days 1, 2, and 7 after IRI and the kidneys were harvested for hematoxylin and eosin staining, and kidney histological scores (f) were obtained. Data are expressed as mean±s.e.m. (n = 4) in each group. Statistical significance was analyzed by ANOVA followed by Student–Newman–Keuls test, and significant differences were accepted when P<0.05 or less between two groups. A, AKI rats; S, Sham rats.

On ad libitum water consumption, AKI rats were oliguric on day 1 and 2, and polyuric on day 7 compared with the Sham group, whereas AKI-Klotho rats had milder oliguria on day 1 and less polyuria on day 7 (Figure 5d). The defect in urinary concentrating ability was also ameliorated by Klotho treatment (Figure 5e).

Kidney histology showed milder damage in Klotho-treated rats compared with vehicle-treated rats (Supplementary Figure S4 online). On day 7, the kidney of vehicle-treated rats started to regenerate, but still had numerous dilated renal tubules and granular casts. In contrast, Klotho-treated rats had near-normal kidney morphology (Supplementary Figure S4 online) by day 7. The histopathological scores for Klotho-treated AKI rats were significantly lower than the vehicle-treated AKI rats, but higher than sham rats (Figure 5f) on day 1. The pathological score decreased with time in the vehicle-treated AKI rat, but the score declined more rapidly in Klotho-treated AKI rats (Figure 5f), suggesting Klotho may also enhance kidney recovery from AKI.

NGAL protein (Figure 6a; Supplementary Figure S5A online) and mRNA (Figure 6b) were upregulated in the kidneys with AKI, but the increase was much less in Klotho-treated rats. As NGAL has been considered as a real-time indicator of active kidney damage,^25–27 these results suggest that Klotho treatment led to less severe damage by IRI. However, an interesting finding was that Klotho significantly upregulated endogenous renal NGAL mRNA and protein expression in the Sham animals in the absence of ischemia (Figure 6a and b). NGAL has been proposed as a renoprotective substance and administration of exogenous NGAL attenuates kidney damage and improves renal function induced by ischemia.²⁸ Augmentation of endogenous NGAL expression in kidney by Klotho may be one potential mechanism whereby Klotho confers protective effect against ischemic renal injury.

Figure 6. Neutrophil gelatinase-associated lipocalin (NGAL) and Klotho response to ischemia–reperfusion injury (IRI) in acute kidney injury (AKI) rats: effect of Klotho treatment.<.

br>Representative immunoblot for NGAL protein in total kidney lysate from four rats in each group on days 1, 2, and 7 (a). Thirty μg of protein was immunoblotted for NGAL with goat anti-NGAL antibody (Santa Cruz Biotechnology). (b) Abundance of NGAL transcript in kidney. Total RNA was extracted from rat kidney on days 1, 2, and 7 after IRI. Complementary DNA was generated. Relative quantification of NGAL transcripts was calculated as 2^−(ΔΔCt) by normalization to cyclophilin followed by ratio to Sham-vehicle rats. Relative NGAL transcripts in treated rats vs Sham-vehicle rats are expressed as mean±s.e.m. of four samples in each group. Statistical significance was assessed by analysis of variance (ANOVA) followed by Student–Newman–Keuls test, and significant differences were accepted when P<0.05 or less between two groups. Immunoblot for Klotho protein in total kidney lysate from rats on days 1, 2, and 7. Representative immunoblot for Klotho and β-actin in total kidney lysate of six rats in each group (c, upper panel). A summary of densitometric quantification of all samples is shown in lower panel. Results are expressed as mean±s.e.m. of six rats in each group. Statistical significance was assessed by ANOVA followed by Student–Newman–Keuls test, and significant differences were accepted when *P<0.05, **P<0.01 vs Sham-vehicle rats; ^#P<0.05; ^##P<0.01 vs AKI-vehicle rats; ^§P<0.05,^§§P<0.01 vs Sham-Klotho rats. Infrared dye-labeled Klotho (IR-Kl) in blood and urine of Sham or AKI rats (d). Rats were intravenously injected once with IR-Kl and 4 μl of plasma at 5, 10, 30, 60 min, and at 24 and 48 h post-injection and of fresh spot urine at 20 min and 24 and 48 h post-injection were collected and fractionated on 7.5% SDS-polyacrylamide gel electrophoresis and scanned by Odyssey Infrared Imaging System (Lincoln, NE, USA). IR-Klotho in the kidneys of Sham and AKI rats (e): After a single bonus intravenous injection of IR-Klotho, rats were killed at the indicated time points for examination of labeled IR-Klotho in kidneys. Four μm cryosections of non-fixed kidneys were prepared and scanned directly using Odyssey Infrared Imaging System. A, AKI; S, Sham rats.

We examined renal Klotho protein levels after administration of exogenous Klotho. AKI rats treated with Klotho protein had milder decline in Klotho protein content in the kidney (Figure 6c; Supplementary Figure S5B online) compared with AKI rats without Klotho treatment on day 1 and 2. This can potentially stem from renal accumulation of injected Klotho protein and/or preservation of endogenous Klotho expression caused by injected exogenous Klotho. It should be noted that Sham rats that received exogenous Klotho also had higher Klotho protein levels (Figure 6c), but lower mRNA expression (Supplementary Figure S5C online) in the kidneys compared with Sham rats given vehicle in the absence of ischemia.

To determine whether delayed administration of Klotho exerts renoprotective effects in AKI, we injected Klotho intraperitoneally into rats 60 min after IRI and compared them with those treated 30 min after IRI. Delayed Klotho treatment blunted the increase in blood Cr (Supplementary Figure S6A online) and decrease in Cr clearance (Supplementary Figure S6B online), and decreased proteinuria (Supplementary Figure S6C online). The degree of oliguria on the first and second day and the degree of polyuria at seventh day were also decreased (Supplementary Figure S6D online). Moreover, the increase in fractional excretion of Na was also diminished by later Klotho treatment (Supplementary Figure S6E online). The alleviation of histological scores (Supplementary Figure S6F online) was similar to those of physiological parameters (Supplementary Figure S6A–E online). Therefore, Klotho treatment at 60 min post-injury was still effective but the renoprotection was less efficient than that of Klotho treatment 30 min post-IRI. Therefore, early intervention conveys better outcome than delayed treatment, although the delayed treatment still has the capacity of renoprotection.

The more abundant renal Klotho protein in Klotho-treated animals might be of either exogenous or endogenous in origin because our detection methods do not distinguish native versus injected Klotho. To test this possibility, we labeled recombinant Klotho with an infrared tag and injected it into rats to track its fate. We found substantial levels of exogenous Klotho protein in the blood circulation within 5 min after a single intravenous bolus (Figure 6d), which diminished over the next 2 days. Klotho appeared to stay longer in the blood of AKI rats compared with Sham rats, as it was still detectable in blood after 48 h. In contrast to plasma, we only detected traces of injected Klotho in the urine after 20 min of injection in Sham rats and not in AKI rats; no exogenous Klotho was recovered in the urine after 24 and 48 h. Kidneys of Sham rats retained the labeled Klotho up to 2 days (at termination study) (Figure 6e). Kidneys of AKI rats have more Klotho in the renal cortex compared with Sham rats. This is compatible with the model in which exogenously administered Klotho protein can enter and stay in the kidney, thereby potentially activating protective pathways supporting its therapeutic potential.

DISCUSSION

The purpose of this study was to (1) determine whether AKI is an acute and transient state of deficiency in both membrane and secreted Klotho; (2) determine whether Klotho is a potential biomarker for renal injury; (3) determine whether Klotho has any pathogenic role in renal injury or protection from IRI and if so, replacement of Klotho can be a therapeutic option. We showed that AKI is a state of acute Klotho deficiency in the blood, urine, and kidney. Human urine Klotho measurements also support this notion. In addition to being a biomarker for renal injury, Klotho levels correlated with resilience to kidney damage in AKI. This is consistent with the previous data indicating that forced expression of exogenous Klotho by an adenoviral vector reduced apoptosis and kidney damage in IRI.^22,24 A single bolus injection of recombinant Klotho protein is effective in ameliorating AKI even when administered after the insult. This study has identified Klotho protein as a potential biomarker and therapeutic agent for AKI.

Clinical diagnostic potential

The current use of plasma Cr suffers from the reliance on accumulation of Cr in the blood after reduction of glomerular filtration and hence is notoriously slow. Moreover, renal hypoperfusion without true parenchymal damage (prerenal failure) can sometimes lead to a syndrome very similar to that of AKI and challenges distinction even by the most seasoned expert clinicians. Several biomarkers of AKI are being investigated with the most evidence centered on kidney injury molecule-1 (Kim-1)²⁹ and NGAL.^25–27 Kim-1 expression in kidney and excretion in urine are sensitive and specific for kidney injury as well as predictors of outcome.³⁰ The elevation of NGAL in the blood, urine, and kidney is an indicator of active kidney damage^25–27,31 and is also associated with severity of kidney damage.^32,33 NGAL in kidney is induced early post-IRI (Figure 3c and d and Supplementary Figure S2D online).³⁴ Conditions that activate leukocytes such as cerebrovascular ischemia,³⁵ inflammation,³⁶ and cancer³⁷ also upregulate NGAL in plasma, urine, and local tissues. Kim-1 is a more specific marker for kidney injury, but its upregulation in IRI-induced AKI rodent model occurs at 9 h post-IRI.²⁹ This study indicates that the decreased Klotho is detectable 5 h post-IRI in AKI rodent model. The time course of Klotho decrease remains to be confirmed in human clinical studies. The exploration of Klotho specificity in diagnosis of AKI is beyond the scope of this study.

Clinical therapeutic potential

Currently, supportive care including renal replacement therapy remains the core of clinical management.¹ Replacement therapy is unlikely to significantly affect the disease course of AKI per se and intensive therapy appeared not to consistently or dramatically increase survival rate or improve long-term outcome.⁵ Our studies suggest that Klotho may have a role in both reducing kidney damage and promoting kidney recovery. Klotho deficiency worsens ischemia–reperfusion-induced injury in hindlimbs and overexpressing Klotho protein attenuates IRI-induced tissue damage through decreasing apoptosis,^22,24 or promotes tissue regeneration through angiogenesis.^38,39 These studies increased Klotho level through either genetic manipulation or viral-based Klotho delivery system, and none of them used Klotho replacement. As transgenic expression can have a myriad of secondary effects, we showed that protection can be conferred by giving recombinant Klotho thus raising the potential for clinical use. At least in the rodent model, there appears to be a time-dependent efficacy of exogenous Klotho replacement and that early Klotho supplementation may be critical for AKI treatment. Klotho may function more efficiently as a preventive than a repair agent.

Mechanism of reduction of Klotho expression in AKI and Klotho’s renoprotective action

IRI leads to rapid and pronounced reduction of Klotho mRNA and protein in kidney of animals. But the mechanism of how Klotho is downregulated is not known. One in vitro study showed that oxidative stress decreases Klotho mRNA and protein in a cultured cell line.⁴⁰ We showed that Klotho reduction in vivo takes place only 3 h post-IRI. It was suggested that tumor necrosis factor and interferon-γ in mice with colitis reduces renal Klotho mRNA and protein through activation and induction of nitric oxide production.⁴¹ But whether an increase in tumor necrosis factor and interferon-γ in AKI⁴² results in Klotho downregulation is not known.

Klotho is expressed prominently in renal distal tubules⁸ and to a lesser extent in renal proximal tubules.¹⁷ Although the transmembrane form of Klotho functions as a coreceptor for FGF23 signaling,¹⁰ the extracellular domain of Klotho is released^43,44 into blood, urine, and the cerebrospinal fluid^15,16,19 and can potentially function as an endocrine factor to affect distant organs. We found higher blood FGF23 and inorganic phosphate in Kl^+/− mice compared with WT and Tg-Kl mice (Supplementary Figure S7 online). IRI significantly increase blood inorganic phosphate compared with Sham in all groups, except in Tg-Kl animals. Of note is that IRI did not significantly alter blood levels of FGF23, except in the Kl⁺⁻ mice, where IRI slightly suppressed FGF23. Therefore, unlike chronic states of Klotho deficiency, the acute Klotho deficit in AKI is not associated with high FGF23.

In an immune-mediated glomerulonephritis model, increasing Klotho improved renal function and reduced mitochondrial DNA fragmentation, superoxide anion generation, lipid peroxidation, and apoptosis, suggesting that Klotho may ameliorate mitochondrial oxidative stress.²¹ Klotho also inhibits apoptosis and senescence in vascular endothelial cells,⁴⁵ and upregulation of heat-shock protein-70 by Klotho is associated with inhibition of apoptosis.²⁴ Another possible mechanism whereby Klotho deficiency during AKI worsens kidney damage may be abnormal activation of calpain and degradation of cytoskeletal elements.⁴⁶ The fact that soluble Klotho significantly increases NGAL protein and mRNA in the kidney of Sham rats suggests that Klotho activates NGAL. NGAL has a protective role in ischemic kidney injury.^28,47 We proposed that the upregulation of NGAL by Klotho is one mechanism of renoprotection conferred by Klotho. The full spectrum of mechanisms of Klotho’s protective effect remains to be studied.

MATERIALS AND METHODS

Human study

The study was approved by the Institutional Review Board at the University of Texas (UT) Southwestern Medical Center. The study population consisted of 17 AKI patients recruited randomly in Parkland Memorial Hospital affiliated with UT Southwestern. The diagnosis was based on 50% acute rise in blood Cr or to a concentration above 2.0 mg/dl from baseline.^1,2 Normal control consisted of 14 healthy volunteers with no known health problems and on no medications, enrolled from the staff and students at the UT Southwestern Medical Center (Tables 2 and 3). All subjects gave informed consent. For measurement of urinary Klotho protein, 4 ml fresh urine was concentrated to 0.2 ml through Amicon Ultra-4 filters with 100 kDa cutoff (Millipore, Billerica, MA, USA), and immediately mixed with Laemmli sample buffer and stored at −80 °C. Urine samples with identical Cr content along with recombinant murine Klotho (standard curve) protein of known concentration were subject to immunoblot and specific signals based on bands on films were obtained with free Image J program (NIH, Washington DC, USA).

Animal models

Male Spraque–Dawley rats (250–350 g) were used for Experiments 1 and 3. Male mice (30–40 g; genetic background for Tg-Kl mice = mixture of C57BL/6J and 129, and for Kl^+/− mice = mixture of C57BL/6J and C3H/J) were used for Experiment 2. Lines included the following: (1) transgenic mice overexpressing Klotho, EFmKL46 (Tg-Kl);^8–9 and (2) heterozygous Klotho-deficient mice (Kl^+/−).⁸ WT littermates were generated during crossbreeding for Tg-Kl and Kl^+/− mice. The age of Kl^+/−, Tg-Kl, and WT mice ranged from 12 to 16 weeks. All animal work was conducted following the Guide for the Care and Use of Laboratory Animals by The National Institutes of Health and was approved by the Institutional Animal Care and Use Committee at the UT Southwestern. AKI model was generated using bilateral renal IRI.⁴⁸ Under anesthesia, renal arteries were clamped with arterial clips (30 min Experiment 2; or 40 min Experiments 1 and 3). After clips were removed, the kidneys were observed for 5 min to ensure reperfusion. Sham animals underwent laparotomy of the same duration without arterial clamping. At predetermined times after reperfusion, 24-h urine was collected. Under anesthesia, blood was drawn and kidneys were harvested, instantly snap-frozen in liquid nitrogen, and stored at −80 °C for further processing. Plasma and urine chemistry of animals were analyzed using a Vitros Chemistry Analyzer (Ortho-Clinical Diagnosis, Rochester, NY, USA).

Experiment 1

Sprague–Dawley rats were randomly assigned to Sham or AKI. Twenty-four-hour urine samples and blood were collected at 1, 3, 5 h and on days 1, 2, and 7 after surgery, respectively.

Experiment 2

Kl^+/− mice, Tg-Kl mice, and their WT littermates were randomly divided into Sham and AKI for each genotype. Twenty-four-hour urine samples were collected and blood samples were drawn on day 2.

Experiment 3

To study the therapeutic effect of Klotho administration post-IRI on AKI, Sprague–Dawley rats were randomly divided into either Sham or AKI, and the animals in each group were randomly allocated into vehicle or Klotho treatment. Animals in the Klotho group received one bonus intraperitoneal injection of recombinant mouse Klotho protein (0.01 mg/kg body weight) 30 or 60 min after reperfusion; rats in vehicle group received the same volume of Klotho buffer (150 mM NaCl and 10 mM HEPES pH 7.4). Twenty-four-hour urine and blood were collected on days 1, 2, and 7 after surgery.

See Supplementary Information for detailed methods on kidney histology and histopathology, immunohistochemistry, immunoblot, real-time reverse transcription-PCR, clearance study on labeled Klotho in normal rats, blood FGF23 determination using enzyme-linked immunosorbent assay.

Statistical analysis

Data are expressed as the means±s.e.m. (n = 3 or more unless indicated otherwise). Statistical analysis was performed using Student’s unpaired or paired t-test, or analysis of variance followed by all pairwise multiple comparison procedures using Student–Newman–Keuls test when appropriate. A value of P≤0.05 was considered statistically significant.

Supplementary Material

1. Figure S1.

Relationship of urinary Klotho to creatinine ratio with eGFR of normal (open circle) and of AKI patients (closed circle).

2. Figure S2.

Changes in kidney function and in renal histology of AKI rats: Time course of post-reperfusion.

3. Figure S3.

Kidney histology in AKI mice with different genetic Klotho background.

4. Figure S4.

Renal histological alterations induced by IRI: Klotho administration 30 min post-ischemia.

5. Figure S5.

NGAL protein, Klotho protein, and mRNA abundance in the kidneys of AKI rats treated with Klotho 30 minutes post-reperfusion.

6. Figure S6.

Comparison of Klotho replacement on AKI rat between 30 and 60 min post-reperfusion.

7. Figure S7.

Blood FGF23 and Pi in AKI mice with different genetic Klotho background.