Abstract
Background and objectives Occlusive renovascular disease and hypertension may progress to CKD. Circulating levels of several biomarkers, including fibroblast growth factor (FGF)-23, Klotho, and soluble urokinase plasminogen activator receptor (suPAR), are altered in patients with CKD, but their role in essential hypertension (EH) and renovascular hypertension (RVH) remains unclear.
Design, setting, participants, & measurements Levels of FGF-23, Klotho, suPAR, plasminogen activator inhibitor (PAI)-1, tissue factor, and tissue factor pathway inhibitor (TFI) were measured in the inferior vena cava and renal vein of hypertensive patients with atherosclerotic renal artery stenosis (n=12) or age-matched participants with EH (n=12) and relatively preserved renal function. Single-kidney blood flow was measured to calculate renal release of markers. For control, peripheral vein levels were measured in healthy volunteers (HVs; n=12).
Results FGF-23 levels did not differ among the groups, whereas Klotho levels were lower in participants with RVH and EH than in HVs, and suPAR levels were elevated in patients with RVH compared with HVs and patients with EH (6.1±1.5 versus 4.4±1.9 and 3.2±1.2 ng/ml, P<0.05). PAI-1 levels were higher in patients with RVH than in patients with EH, but tissue factor and TFI levels were not statistically significantly different. After adjustment for GFR, Klotho levels remained decreased in both RVH and EH, and suPAR and PAI-1 levels remained elevated in RVH. eGFR correlated inversely with systemic and renal vein suPAR levels, and directly with systemic Klotho levels.
Conclusions Klotho levels are low in hypertensive patients, whereas suPAR and PAI-1 levels are specifically elevated in RVH, correlating with GFR. Klotho, PAI-1, and suPAR may be markers of kidney injury in hypertensive patients.
Introduction
Hypertension is known to induce target organ injury and may lead to deterioration of renal function. Atherosclerotic renal artery stenosis (ARAS) often accelerates hypertension and may manifest with ischemic renal disease (1), which is in turn linked with increased risk of progression to ESRD (2,3). In addition to plaques in the main renal artery, atherosclerosis can directly compromise the intrarenal parenchyma and vessels, and the severity of parenchymal damage is an important prognostic factor for renal function (4).
In recent years, several novel markers of renal injury in CKD have been identified. Fibroblast growth factor (FGF)-23 is a 32-kD protein secreted mainly by osteocytes, but also by the kidney in diabetic nephropathy (5), and its levels inversely correlate with renal function (6). However, renal production has not been shown in other forms of CKD, including ARAS. Klotho is a protective protein that is expressed predominantly by the kidney, and it ameliorates renal injury in experimental GN and ischemic injury (7–9), may limit hypertension (10) and endothelial dysfunction (11), and acts as a cofactor for FGF-23 (12,13).
In addition, altered fibrinolytic system activity has been implicated in loss of renal function and is causally linked to development of atherosclerosis and its cardiovascular complications (14). Plasminogen activation is pivotal not only for fibrinolysis, which is determined by the interaction between plasminogen activator (PA) and plasminogen activator inhibitor (PAI), but also for diverse biologic processes involved in atherogenesis, including cell adhesion, migration, and angiogenesis (15–18). Accumulating data show that the urokinase plasminogen activator (uPA) membrane receptor and PAI are highly expressed in atherosclerotic plaques of human vessels (19,20). The soluble urokinase plasminogen activator receptor (suPAR) circulates in the plasma, and its level correlates with atherosclerosis, cardiovascular risk, and subclinical organ damage (21). Furthermore, suPAR levels increase in CKD associated with some glomerular diseases (22,23). The degree to which levels of these ubiquitous biomarkers are altered in patients with essential hypertension (EH) or renovascular hypertension (RVH) remains unknown.
This study was undertaken to test the hypothesis that FGF-23 and suPAR are elevated and Klotho is decreased in RVH. For this purpose, we measured the levels of these markers and several related factors in samples obtained from the kidney vein and/or systemic circulation of patients with EH or RVH.
Materials and Methods
Patient Populations
This study was approved by the Mayo Clinic Institutional Review Board and adhered to the Declaration of Helsinki. Written informed consent was obtained from all participants. We prospectively enrolled 24 hypertensive patients in this study, of which 12 patients with EH and 12 patients with RVH were identified, from August 2008 to October 2010. ARAS was defined using imaging criteria including renal artery Doppler ultrasound velocity acceleration, and/or magnetic resonance/computed tomography angiography with evident stenosis >60% and/or poststenotic dilation (24). We excluded patients with an eGFR (calculated with the Modification of Diet in Renal Disease equation) <30 ml/min per 1.73 m2, uncontrolled hypertension (systolic BP >180 mmHg, despite antihypertensive therapy), diabetes requiring medications, recent cardiovascular event (myocardial infarction, stroke, congestive heart failure within 6 months), pregnancy, or kidney transplant. Normotensive healthy control participants (systolic BP <130 and diastolic BP <80 mmHg) were prospectively recruited through the Mayo Clinic Biobank, and matched to patients with EH and RVH according to age, weight, and body mass index (BMI).
Hypertensive patients were all treated with blockers of the renin-angiotensin system, and spent 3 days at the clinical research unit, where they consumed a controlled diet (sodium, 150 mEq/d) and measurements were taken.
Clinical Data Collection and Laboratory Measurement
Clinical data collected by physical examination or via the electronic medical records included age, sex, height, weight, BMI, use of medications, systolic BP, and diastolic BP. Serum creatinine, eGFR, and serum lipid levels were determined by standard procedures. To calculate the gradient and net renal release of each marker, single-kidney blood flow (RBF) was measured on day 3 at the clinical research unit using multidetector computed tomography.
Blood Sampling and Measurement of RBF
Renal vein (RV) and vena cava samples were obtained in hypertensive patients before measurement of RBF. A guide catheter was placed through the femoral or internal jugular vein (using a 6F sheath) and blood samples were collected from the right and left RVs and inferior vena cava (IVC). For injecting contrast media, the catheter was then positioned at the superior vena cava. Measurement of single-kidney RBF in RVH and EH patients utilized multidetector computed tomography (Somatom Sensation-64; Siemens Medical Solutions, Germany) to obtain 45 consecutive scans over the renal hilum (5-mm–thick slices), over 2–3 minutes, as previously described (25,26). Perfusion was calculated from time-attenuation curves obtained in the kidney after contrast injection (iopamidol-370, 0.5 ml/kg, up to 40 ml and 10 ml/s). In addition, cortical and medullary volumes were assessed using stereology. Single-kidney volume was calculated by summing cortical and medullary volumes, and single-kidney RBF by multiplying kidney volume by perfusion (25,27).
Measurement of Biomarker Levels
Levels of FGF-23 (catalog no. EZHFGF23-32K; EMD Millipore), Klotho (catalog no. 27998; Immuno-Biologic Laboratories), suPAR (catalog no. CSB-E04752H; Cedarlane), vascular endothelial growth factor (VEGF) (catalog no. MPXHCYTO-60K; Millipore Luminex), PAI-1 (catalog no. HCVD1-67AK; Millipore Luminex), and TNF-α (catalog no. MPXHCYTO-60K; Millipore Luminex) were measured following the manufacturer’s protocol in the RV and IVC in the RVH and EH groups, and in an antecubital vein in healthy volunteers (HVs). In patients with ARAS with bilateral stenoses, measurements were taken at the more severe side. In addition, tissue factor (TF) and tissue factor pathway inhibitor (TFI) (catalog nos. ab108903 and ab108904, respectively; Abcam) levels were measured in the systemic circulation.
Using the difference between infrarenal IVC and RV levels as an index of the net release of these markers within the affected kidney (25), we estimated gradient (RV-IVC) and net release (gradient×RBF) for each marker in hypertensive patients.
Statistical Analyses
Results are expressed as means±SD for normally distributed variables and medians (interquartile range) for non-normally distributed variables. For comparison of two means of independent samples, the t test or Mann–Whitney test was used; for three means of independent samples, ANOVA or the Kruskal–Wallis test followed by post hoc Bonferroni analysis was used. The chi-squared test or Fisher’s exact test was utilized for categorical variables as appropriate. To adjust levels, gradients, and net release by eGFR, we used analysis of covariance. Skewed data were transformed to logarithmic values. Spearman rank correlation analysis was used to test for associations between markers and other variables. A two-tailed P value of ≤0.05 was considered statistically significant.
Results
Patient Characteristics
Table 1 shows the characteristics of patients included in this study. Systolic BP was higher in patients with RVH compared with HVs, but was not different from patients with EH. There were no differences in antihypertensive regimens between patients with RVH and EH, and none were taking vitamin D supplementation. The serum lipid profile was similar among the groups, although triglyceride levels tended to be higher in patients with RVH than in HVs (P=0.09). Serum creatinine levels were increased and eGFR reduced in patients with RVH compared with HVs and patients with EH, but urinary protein excretions were similar. Single-kidney RBF was lower in the stenotic kidney compared with the EH kidney.
Clinical, laboratory, and demographic data in HVs and patients with RVH or EH
Systemic and RV Levels of CKD Biomarkers
Supplemental Figure 1 shows levels of FGF-23, Klotho, suPAR, and PAI-1 in HVs as well as participants in the EH and RVH groups. Neither systemic nor RV levels of FGF-23 were different among the groups (P=0.92 and P=0.24, respectively). The systemic level of Klotho was similarly reduced in the EH and RVH groups compared with HVs (P=0.02 and P=0.01, respectively), and its RV level did not differ between EH and the stenotic RVH kidney (P=0.94). Systemic suPAR and PAI-1 levels were elevated only in RVH compared with EH (P<0.03 and P=0.02, respectively), and the PAI-1 level was lower in patients with EH compared with HVs (P=0.04). RV levels of suPAR and PAI-1 were also higher in the stenotic RVH than in EH (P=0.004 and P=0.05, respectively), whereas systemic levels of TF, TFI, or their ratio (TF/TFI) were not different among the groups (Figure 1).
Systemic levels of TF, TFI, and TF/TFI ratio showed no significant difference among HVs and patients with EH or RVH. TF, tissue factor; TFI, tissue factor pathway inhibitor.
To examine whether differences among the groups were determined by renal function, all levels, gradients, and net release data were also adjusted for eGFR (Supplemental Figure 1). Systemic and RV levels of FGF-23 remained similar among the groups (P=0.32 and P=0.94), whereas systemic levels of Klotho remained reduced in the EH and RVH groups compared with HVs (P=0.02 and P=0.03, respectively), and its RV level similar between in EH and RVH (P=0.87). The systemic level of suPAR remained higher in patients with RVH than in HVs (P=0.004), and also tended to be elevated in patients with EH compared to HVs (P=0.06). However, the difference in systematic suPAR levels between patients with ARAS and patients with EH was abolished by adjustment for eGFR (P=0.53), and its RV levels in the EH and stenotic RVH groups became indistinguishable (P=0.22). The systemic level of PAI-1 remained elevated (P=0.04) and its RV level also tended to be higher in patients with RVH compared with patients with EH (P=0.06). Systemic levels of TF, TFI, or TF/TFI ratio remained similar in all of the groups (Figure 1).
Renal Gradient and Net Release of CKD Biomarkers
In EH kidneys and stenotic kidney of RVH patients, eGFR-unadjusted RV levels of FGF-23 and Klotho were significantly higher than their systemic levels (P=0.03 and P<0.001, respectively), resulting in positive gradients (Supplemental Figure 2). By contrast, RV levels of suPAR were reduced compared to their systemic levels (P=0.004) and showed a negative gradient in the RVH group, suggesting renal uptake or urinary excretion, and PAI-1 did not show any gradient (P=0.81). After eGFR adjustment of their levels, RV levels of FGF-23 and Klotho remained higher than their systemic levels (P=0.01 and P<0.001, respectively), and suPAR retained a negative gradient (P=0.01). However, gradients across the kidneys and net release of most markers were not significantly different between EH kidneys and stenotic kidney of RVH patients, except for the FGF-23 gradient was that greater in RVH compared with EH (Supplemental Figure 2, P=0.05), and Klotho release tended to be lower (P=0.06). Furthermore, eGFR-adjusted gradients and net release of all markers were not different among the two groups (Supplemental Figure 2). In patients with RVH, RV levels of suPAR tended to be lower in the less severely affected contralateral kidney (CLK) compared with the stenotic kidney (P=0.07), whereas other markers did not differ between them (Supplemental Figure 2). Yet both eGFR-unadjusted and eGFR-adjusted gradients and release of suPAR in the stenotic kidney were higher (P=0.04) than in the CLK, as was net release of PAI-1 (P=0.05), which remained significant after eGFR adjustment (P=0.01).
Correlation with Renal Function and Hemodynamics in Hypertensive Patients
Table 2 shows bivariate analysis for correlations between these markers and clinical variables. eGFR correlated directly with systemic levels of Klotho, and inversely with RV levels of FGF-23 and with both systemic and RV levels of suPAR. RV levels of FGF-23 and both systemic and RV levels of suPAR also correlated inversely with RBF (Figure 2).
Bivariate correlation analysis between biomarkers and other variables in patients with EH or RVH (n=12 each)
Correlation between biomarkers and RBF. RV levels of FGF-23 and both systemic and RV levels of suPAR also correlated inversely with RBF. RBF, renal blood flow.
Upon analysis with partial correlation with control of other variables (BMI, triglycerides), systemic levels of suPAR remained inversely correlated with eGFR (r=−0.34, P=0.02), but Klotho did not (P=0.13). After adjustment for eGFR, only the RV level of suPAR remained inversely correlated with RBF (r=−0.47, P=0.03).
Related Cytokine Levels in Hypertensive Patients
Both systemic and RV levels of TNF-α were increased in ARAS compared with EH, whereas VEGF levels were not different among the groups (Table 3). Bivariate analysis showed that the RV level of VEGF directly correlated with the systemic or RV level of Klotho (r=0.52, P=0.04, and r=0.61, P=0.02, respectively). Systemic TNF-α correlated directly with RV levels of suPAR (r=0.52, P=0.02). Systemic and RV levels of TNF-α did not correlate with either the systemic or RV level of Klotho, but were inversely related to net release of Klotho (r=−0.53, P=0.02; and r=−0.56, P=0.01, respectively), and RV TNF-α tended to correlate inversely with Klotho gradient, (P=0.07).
Cytokine levels in HVs and patients with RVH or EH
Discussion
This study shows that systemic and RV levels of suPAR and PAI-1 are elevated in patients with RVH but not EH, although the elevated suPAR levels were partly explained by reduced GFR. A reduction in levels of the renoprotective factor Klotho in both EH and RVH suggests propensity for kidney damage in hypertension, and positive cross-kidney eGFR-adjusted gradients and net release of Klotho and FGF-23 may imply a renal source. PAI-1 and TNF-α were distinctly elevated in patients with RVH compared with EH. Overall, these observations imply that suPAR and Klotho, but not FGF-23, may be useful markers for renal injury in hypertensive patients, whereas PAI-1 and TNF-α may be capable of distinguishing kidney injury in ARAS from EH.
Recent studies in ARAS revealed important insights into the pathogenesis of target organ injury related to renal atherosclerosis and ischemia, and suggest that reduced perfusion triggers an inflammatory and fibrotic process within the kidney that has systemic ramification. In EH, increased intraglomerular pressure eventuates in renal fibrosis and glomerulosclerosis. In this study, we measured levels of several emerging biomarkers proposed as indices of kidney injury in various forms of renal disease. To assess whether they originated within the kidney, we also quantified single-kidney RBF and determined both their cross-kidney gradient and renal release (25,26).
uPA is a serine protease and is an important component of the fibrinolytic system, converting plasminogen to the active enzyme, plasmin. suPAR is released from the plasma membrane by cleavage of the glycosylphosphatidylinositol anchor (28), and has been extensively studied as a biomarker of inflammation or immune activation. In addition, circulating suPAR has been reported to positively associate with carotid atherosclerosis and increased cardiovascular risk in CKD (29,30), and is also reported to be involved in renal damage that might progress to CKD (31). suPAR was previously postulated to represent a specific biomarker for primary FSGS, but recently has been identified in other glomerular diseases and negatively correlated with GFR (23,30,32). In this study, the systemic level of suPAR in RVH was higher than observed in primary FSGS (approximately 4.6 ng/ml) (22,33). The suPAR level in ARAS remained elevated compared with HVs after eGFR adjustment, and both systemic and RV levels correlated with eGFR. Furthermore, the stenotic kidney RV level of suPAR showed direct correlation with its RBF and was increased compared with the CLK. Therefore, the suPAR level might implicate specific ARAS-associated kidney damage and consequent functional decline.
PAI-1, a potent inhibitor of fibrinolysis, may lead to excessive vascular fibrin accumulation and neointimal and thrombus formation (34), and is elevated in clinical situations linked to accelerated atherosclerosis (18,35). In support of suppression of the fibrinolytic system, we found that PAI-1 levels were distinctively elevated in ARAS. Although its RV level was no longer higher than in EH after eGFR adjustment, the systemic levels of PAI-1 remain higher than in EH. However, given the unaltered TF and TFI levels as well as the correlation of suPAR level with TNF-α, the elevated levels of suPAR and PAI-1 in ARAS may reflect their involvement in progression of atherosclerosis and renal dysfunction rather than fibrinolysis alone. The relation between single-kidney RBF with its effluent venous level of suPAR might also reflect affected kidney injury. Interestingly, we found a lower systemic level of PAI-1 in patients with EH compared with HVs. Because angiotensin II increases PAI-1, its inhibition in our patients might have reduced the level of PAI-1 (36,37) and might have blunted the difference between patients with ARAS and HVs. The PAI-1 level in ARAS remained high after eGFR adjustment but TF and TFI were not, implying that besides inhibiting fibrinolysis, other biologic process of PAI-1 might contribute to progression of atherosclerosis in ARAS.
VEGF plays an important role in maintaining peritubular and glomerular capillaries, and activates the uPA/suPAR system to increase vascular permeability (38). However, we found no difference in VEGF levels among the groups. TNF-α upregulates expression of PAI-1 (39), and was elevated in patients with ARAS, although its levels correlated only with RV suPAR levels.
FGF-23 purportedly increases in parallel with a decline in renal function (40–42), to compensate for hyperphosphatemia (43), or due to its retention in the circulation (40).Our study implies that in addition to progressive diabetic renal disease in rats (5), FGF-23 might also be produced in the poststenotic human kidney in ARAS. The most important determinant for FGF-23 in CKD is decreased GFR, and elevated FGF-23 levels predict progression of CKD (44) and cardiovascular events. For these studies, we excluded patients with CKD grade 4 and above, and did not detect increased FGF-23 levels. However, neither the systemic nor RV level of FGF-23 in ARAS was different from patients with EH or HVs. The modest decrease in GFR in ARAS might have masked any difference among the groups.
Ample experimental evidence indicates that Klotho is renoprotective. Klotho ameliorates apoptosis in renal tubule cells (9), and improves kidney function, tubulointerstitial injury (45), and endothelial function (10). Being expressed mainly in the kidney, Klotho production might fall as kidney function declines (46). This protein plays an important role in maintaining endothelial integrity in association with VEGF, and its deficiency might impair VEGF-mediated angiogenesis and endothelial function through disturbed nitric oxide–dependent VEGF signaling (47,48). Although we did not find a change in VEGF levels, we observed a correlation between the RV level of Klotho and VEGF, which may reflect an interaction between them. Klotho deficiency also increases PAI-1 expression (49), but we did not observe this relationship. We did not find a relationship between levels of TNF-α and Klotho either, except an inverse relationship with net release of Klotho, although TNF-α might repress Klotho expression of Klotho via IFN-γ (50). Indeed, decreased levels of Klotho in both ARAS and EH and the direct correlation of Klotho with GFR might implicate it in the pathogenesis of kidney injury in hypertensive disorders. This observation suggested Klotho as a potential early biomarker for detecting renal injury or reduced renal function, which are not presented with a conventional marker such as plasma creatinine.
This study has some limitations. First, this study is limited by its cross-sectional nature with a relatively small study population, as well as by the lack of longitudinal observation. Second, we do not have available plasma levels of calcium, phosphate, or vitamin D, and cannot exclude their effects on FGF-23 and Klotho levels. However, because severe CKD was excluded, GFR-dependent changes in calcium/phosphate homeostasis were likely negligible. Third, eGFR was decreased in ARAS compared with patients with EH or HVs, and despite adjustment might affect the levels of biomarkers. Finally, we do not have adequate tissue samples to assess glomerular injury. Additional prospective population studies are needed to define injury biomarkers in hypertensive patients.
In conclusion, plasma (systemic and RV) levels of Klotho are reduced in hypertensive patients, whereas suPAR and PAI-1 are both elevated only in patients with RVH. Our results imply that inhibited fibrinolysis and other biologic activities of suPAR/PAI-1 may contribute to the pathophysiology and potentially serve as a biomarker for renal damage in patients with ARAS.
Disclosures
None.
Acknowledgments
This study was partly supported by grants from the National Institutes of Health (DK73608, HL121561, DK100081, HL123160, DK104273, HL92954, and C06-RR018898) and the Mayo Clinic Center for Regenerative Medicine.
Footnotes
Published online ahead of print. Publication date available at www.cjasn.org.
This article contains supplemental material online at http://cjasn.asnjournals.org/lookup/suppl/doi:10.2215/CJN.07290714/-/DCSupplemental.
- Received July 21, 2014.
- Accepted November 10, 2014.
- Copyright © 2015 by the American Society of Nephrology
References
In this issue
