Increased urinary Angiotensinogen/Creatinine (AGT/Cr) ratio may be associated with reduced renal function in autosomal dominant polycystic kidney disease patients

From January 2011 to February 2012, a total of 304 ADPKD patients above 18 years regularly visiting our ADPKD clinic were enrolled in the study. ADPKD was clinically diagnosed according to the Unified Criteria for Ultrasonographic Diagnosis of ADPKD proposed by Pei et al [17]. Among them, 71 patients were excluded from the analysis due to following reasons: 34 patients older than 60 years old, 5 patients on hemodialysis, 4 patients who received kidney transplantation, 12 patients with cancer, 8 patients currently on immunosuppressive therapy for the various reasons, 1 patient with previous unilateral nephrectomy, 2 patients who did not take computed tomography (CT) exam, 4 patients with chronic viral hepatitis, and 1 pregnant patient. As a result, a total of 233 patients were included in an analysis.

Upon enrollment, the demographic information and laboratory parameters were investigated. Systolic and diastolic blood pressures were measured at every visit and the number and types of anti-hypertensive medications were also reviewed. Upon enrollment, blood samples were withdrawn and serum Cr was measured by Jaffe method. The eGFR was calculated using chronic kidney disease epidemiology (CKD-EPI) equation [18, 19]. The chronic kidney disease (CKD) stage was defined according to the eGFR estimated by CKD-EPI equation as follows: stage I (>90 mL/min/1.73 m²), stage II (60-90 mL/min/1.73 m²), stage IIIA (45-60 mL/min/1.73 m²), stage IIIB (30-45 mL/min/1.73 m²), stage IV (15-30 mL/min/1.73 m²), and stage V (<15 mL/min/1.73 m²) [20]. To evaluate systemic RAS activity, plasma renin activity (PRA) and basal aldosterone were measured on resting state. Albumin and Cr were measured from random urine samples.

Of 233 patients, 200 patients underwent CT within 1 year from their enrollment time points. Thirty-three patients were excluded from the analysis because they underwent CT scans apart from urinary biomarker measurement (>1 year). Among them, 189 patients had available height information to calculate htTKV (mL/m) [21]. We have calculated TKV from 3- to 5-mm thickness three-dimensional contrast-enhanced CT of the kidney and bladder using a multi-detector CT scanner (Somatom Sensation 16, SIEMENS; Light speed Ultra 8, GE; Brilliance CT 64, Philips; Somatom Definition, SIEMENS). The CT examination was performed either with contrast agent or without depending on patient’s renal function. In order to acquire contrast images, we administered contrast material right after pre-contrast image acquisition and acquire post-contrast images about 17-20 s later when the Hounsfield Unit (HU) at the descending aorta reaches 100HU. The htTKV was defined as the sum of the left and right renal volumes divided by height of the patient.

Urine samples were separately collected for the measurement of urinary AGT, mixed with 1 mL of 10 mM Tris buffer, centrifuged at 3000 g for 10 min, and stored at -80 °C until measurement. Urinary AGT was measured by commercial sandwich enzyme-linked immunosorbent assay (ELISA) (Immuno-Biological Laboratories, Co., Ltd., Gunma, Japan). Urine samples were diluted 10 times before the measurement. To evaluate whether urinary AGT is a better biomarker for renal function in ADPKD, we also measured urinary NAG and β2MG. The activity of urinary NAG was measured by a spectrophotometric assay under a 340-nm wavelength using a TBA 200 FR biochemical analyzer (Toshiba, Tokyo, Japan) [22]. The urinary excretion level of β2MG was measured using a radioimmunoassay kit (Beckman Coulter, Prague, Czech Republic) [23]. To compensate for the production of concentrated or dilute urine samples, the urinary biomarker levels were expressed based on urinary Cr content.

One normal kidney specimen and two polycystic kidney specimens were collected from radical nephrectomized kidneys (Table 1). Each kidney specimen was fixed in 4 % paraformaldehyde (PFA) for 24 h and paraffin embedded. The 4-μm sections were prepared for immunohistochemistry as described before [24]. Antigen retrieval was performed with Tris-EDTA buffer, pH9.0, at 100 °C for 15 min. The slides were rinsed three times with PBS, followed by the addition of 3 % H₂O₂ for 10 min at room temperature to block endogenous peroxidase. The slides were rinsed three times with PBS and then blocker with universal blocking buffer. The primary antibodies were added at the dilution indicated in Table 2. Slides were incubated either for 1 h at room temperature or overnight at 4 °C, followed by the addition of the secondary antibodies. Sections were stained with DAB substrate and counterstained with hematoxylin, and viewed under a Leica microscope.

Statistical analyses were performed using SPSS, version 19.0 (SPSS Inc., www.​spss.​com). The variables that did not follow normal distribution were log-transformed before analysis. To investigate whether urinary biomarkers were correlated with either eGFR or htTKV, linear regression analyses were performed. To adjust for potential confounders, multiple regression analysis was performed. We also performed independent t-test and the analysis of variance (ANOVA) test to compare continuous variables between groups. The P-value < 0.05 was considered statistically significant.

This study was approved by the Institutional Review Board of Seoul National University Hospital (H-0901-046-269). The procedures for the use of the human kidney specimens were approved by Seoul National University Hospital Institutional Review Board (H-0701-033-195). All participants provided written informed consent before the study in accordance to the Declaration of Helsinki.

A total of 233 ADPKD patients were selected for this study. Baseline clinical characteristics are summarized in Table 3. The mean age was 43.3 ± 9.7 years and had slight female predominance (54.1 %). The majority of them were either diagnosed with hypertension or on blood pressure lowering therapy (n = 217, 93.1 %), and 153 patients were taking angiotensin converting enzyme inhibitor (ACEi) and/or angiotensin receptor blocker (ARB). Mean serum Cr and eGFR were 1.1 ± 0.5 mg/dL and 80.5 ± 27.1 mL/min/1.73 m², respectively. The median value of urinary AGT/Cr was 13.7 (Interquartile range, IQR 7.5 – 35.1) μg/g. The CT scan was taken in 189 subjects. Measured htTKV at baseline was 705 mL/m (Min. 119, Max. 3436) and the mean time gap between urinary biomarker measurement and CT scan was 4.7 ± 4.0 months.

To evaluate the association between each urinary biomarker and renal functional and structural markers, a linear regression analyses were performed. Urinary AGT, NAG, and β2MG were compared with eGFR, serum Cr, and htTKV. Urinary AGT/Cr was negatively correlated with eGFR (r ²  = 0.162, P < 0.001)(Fig. 1). When compared with urinary NAG/Cr and β2MG/Cr, urinary AGT/Cr showed the best correlation with eGFR demonstrating the greatest Pearson’s correlation coefficient (r ²  = 0.162 vs. 0.111 vs. 0.138, P < 0.001) (Additional file 1: Figure S1). Furthermore, when we performed ANOVA among CKD stage groups, the mean values of urinary AGT/Cr increased according to CKD stages (CKD stage I-II (n = 186), 27.8 ± 58.5 μg/g; CKD stage IIIA (n = 22), 56.0 ± 61.1 μg/g; CKD stage IIIB (n = 15), 89.0 ± 89.5 μg/g; CKD stage IV-V (n = 9), 95.3 ± 108.9 μg/g) (Fig. 2).

Among 189 patients with available htTKV measurement, urinary AGT/Cr also demonstrated a better correlation with htTKV (r ²  = 0.107, P < 0.001) than urinary NAG/Cr and β2MG/Cr (r ²  = 0.089 and r ²  = 0.08, P < 0.001) (Additional file 1: Figure S1). The patients with larger htTKV > 750 mL/m (n = 90) showed greater urinary AGT/Cr (50.9 ± 76.7 μg/g) compared with those with smaller kidneys < 750 mL/m (23.5 ± 48.2 μg/g, P = 0.003) (Fig. 2). These results indicated that urinary AGT/Cr shows a good correlation with both concurrent eGFR and htTKV.

Since RAS activation is a major cause of hypertension in ADPKD patients, urinary AGT/Cr and PRA levels were compared among normotensive and hypertensive groups. As shown in Fig. 3, both urinary AGT/Cr and PRA were tended to be higher in hypertensive subjects compared to normotensive subjects (P > 0.05). Among hypertensive subjects, urinary AGT/Cr was significantly increased in the advanced CKD stages (III-V) compared to early CKD stages (I-II) (28.6 ± 60.3 vs. 93.2 ± 139.3 μg/g, P < 0.001). This remained statistically significant when we compared urinary AGT/Cr among the subgroup population who were not taking RAS blocker medications (29.9 ± 54.0 vs. 106.7 ± 217.7 μg/g, P = 0.009, data not shown). On the other hand, PRA could not differentiate those with decreased renal function from those with normal renal function (7.5 ± 8.7 vs. 7.4 ± 10.2 ng/mL/hr, P = 0.928). In concordance to previous studies, our results suggest that local RAS (represented by urinary AGT/Cr) as well as systemic RAS (represented by PRA) are activated in hypertensive patients. However, urinary AGT/Cr showed better association with CKD stages among hypertensive subjects compared to PRA. Meanwhile, RAS blockers, ACEi and/or ARB, did not affect either urinary AGT/Cr levels or PRA/aldosterone levels (Additional file 1: Figure S2).

To evaluate clinical variables associated with eGFR, we performed a linear regression analysis. Age, gender, and hypertension were evaluated as demographic variables. The eGFR at the initial hospital visit, plasma hemoglobin, serum uric acid, htTKV, random urine albumin to Cr ratio, urinary AGT/Cr were included as laboratory parameters. In univariate analysis, higher urinary AGT/Cr was associated with reduced eGFR as well as old age, hypertension, low initial eGFR, low plasma hemoglobin, high serum uric acid, large htTKV, and microalbuminuria (Table 4). Since the association between urinary AGT/Cr and eGFR may be confounded by other clinical variables, multivariate regression analyses were performed as the next step. The stepwise selection method was used to define independent factors. Age, hypertension, initial eGFR, plasma hemoglobin, serum uric acid, htTKV, random urine albumin to Cr ratio, urinary AGT/Cr were included in the final model. Old age, low initial eGFR, low plasma hemoglobin, high serum uric acid, and large htTKV were the independent factors associated with reduced eGFR. However, urinary AGT/Cr was not an independent factor associated with reduced eGFR.

In order to investigate the origin of AGT expression in polycystic kidneys, immunohistochemistry was performed using normal and polycystic kidney tissues. In the normal kidney, AGT was not expressed in any of proximal tubules, glomeruli, or vessels. On the other hand, in case I (PKD-CKD), AGT was strongly expressed in proximal tubules and cyst-lining epithelial cells (Fig. 4). Of note, the staining intensity of AGT was greater in the proximal tubules compressed by nearby cysts. In case II, PKD-end-stage renal disease (ESRD), AGT was also expressed in the proximal tubules; however, its intensity was slightly reduced than that of case I. Nevertheless, strong expression of AGT was also noted at cyst-lining epithelial cells in case II.

Expression levels of other intrarenal RAS components such as AngII, Ang-(1-7), ACE, ACE2, and chymase were investigated using immunohistochemical staining (Fig. 4, Table 5). Immunohistochemitry results of case I demonstrated that all other intrarenal RAS components including AngII, Ang-(1-7), ACE2 and chymase but ACE expression were augmented in the polycystic kidneys compared to the normal kidney. The AngII expression was moderately increased in both proximal and distal tubules. The Ang-(1-7) expression was markedly increased in proximal and distal tubules and glomeruli. The ACE2 expression was markedly increased in the proximal tubular cells of polycystic kidneys. However, the ACE2 expression level in the distal tubules was similar to that of normal kidney. Of note, the expression level of chymase, an alternative enzyme which converts AngI to AngII, was moderately increased in both proximal and distal tubules. Meanwhile, ACE expression levels were decreased in polycystic kidney tissue compared to the stain intensity in normal kidney tissue. Immunochemistry results of case II showed similar distribution of expression to case I, but cyst-lining epithelial cells were positively stained especially for Ang-(1-7). The staining pattern of AngII, ACE2 and chymase in polycystic kidney tissue was rather patchy with less staining intensity compared to those in case I.