Increased urinary angiotensin converting enzyme 2 and neprilysin in patients with type 2 diabetes

Abstract

Angiotensin converting enzyme 2 (ACE2) and neprilysin (NEP) are metalloproteases that are highly expressed in the renal proximal tubules. ACE2 and NEP generate renoprotective angiotensin (1-7) from angiotensin II and angiotensin I, respectively, and therefore could have a major role in chronic kidney disease (CKD). Recent data demonstrated increased urinary ACE2 in patients with diabetes with CKD and kidney transplants. We tested the hypothesis that urinary ACE2, NEP, and a disintegrin and metalloproteinase 17 (ADAM17) are increased and could be risk predictors of CKD in patients with diabetes. ACE2, NEP, and ADAM17 were investigated in 20 nondiabetics (ND) and 40 patients with diabetes with normoalbuminuria (Dnormo), microalbuminuria (Dmicro), and macroalbuminuria (Dmacro) using ELISA, Western blot, and fluorogenic and mass spectrometric-based enzyme assays. Logistic regression model was applied to predict the risk prediction. Receiver operating characteristic curves were drawn, and prediction accuracies were calculated to explore the effectiveness of ACE2 and NEP in predicting diabetes and CKD. Results demonstrated that there is no evidence of urinary ACE2 and ADAM17 in ND subjects, but both enzymes were increased in patients with diabetes, including Dnormo. Although there was no detectable plasma ACE2 activity, there was evidence of urinary and plasma NEP in all the subjects, and urinary NEP was significantly increased in Dmicro patients. NEP and ACE2 showed significant correlations with metabolic and renal characteristics. In summary, urinary ACE2, NEP, and ADAM17 are increased in patients with diabetes and could be used as early biomarkers to predict the incidence or progression of CKD at early stages among individuals with type 2 diabetes.

Keywords: ACE2; ADAM17; NEP; chronic kidney disease; diabetic nephropathy; type 2 diabetes.

Fig. 1.

Urinary and plasma ACE2 activity in ND, Dnormo, Dmicro, and Dmacro subjects using a fluorogenic enzyme assay. A: urinary and plasma ACE2 activity was measured in 10–20μl of samples, *P < 0.01 vs ND B: ACE2 activity assay was validated using MLN-4760 (ACE2 inhibitor), ZPP (Prolyl endopeptidase/prolyl carboxypeptidase inhibitor), and thiorphan (NEP inhibitor). Urinary ACE2 activity without addition of inhibitors (control) was set at 100%. *P < 0.0001 vs. control. Each bar represents mean ± SE. ACE2, angiotensin converting enzyme 2; Dmacro, Type 2 diabetic patients with macroalbuminuria; Dmicro, Type-2 diabetic patients with microalbuminuria; Dnormo, Type 2 diabetic patients with normoalbuminuria; ND, nondiabetics; NEP, neprilysin; ZPP, Z-prolyl-prolinal.

Fig. 2.

Immunoblot analysis of urinary ACE2, NEP, ADAM17, and albumin in ND, Dnormo, Dmicro, and Dmacro subjects. Immunoblots were depicted using two patients, and the semiquantitative analysis is performed using four to six patients per group. A: urinary ACE2 expression was determined in ND and patients with diabetes with various degrees of albuminuria. Kidney lysate (2 μl) and urine (2 μl) from diabetic mice were used as positive controls. In mice, full length of ACE2 (100 kDa) was observed both in kidney and urine. In addition, fragmented ACE2 bands (75 kDa and 65 kDa) were seen in urine obtained from mice, whereas in humans, a 120-kDa immunoreactive band, as well as fragmented ACE2 immunoreactive bands at 75 kDa, ~70 kDa, and ~50 kDa were seen only in patients with diabetes. Fragmented urinary ACE2 bands (50 kDa) were quantified, and significant increase in urinary ACE2 expression was observed in patients with diabetes compared with ND (*P < 0.0001). B: urinary NEP expression was determined in ND and patients with diabetes using the volume equivalent to 10 μg creatinine. Kidney lysate (2 μl) and urine (2 μl) from diabetic mice were used as positive controls. In mice, full length of NEP (94 kDa) was observed in both kidney and urine. In addition, fragmented NEP immunoreactive bands at 70 kDa and 50 kDa were seen in urine obtained from mice, whereas in humans, full-length as well as fragmented NEP immunoreactive bands were seen at ~110 kDa, ~70 kDa, and ~50 kDa with increased intensity in patients with diabetes. Fragmented NEP immunoreactive bands were not observed in ND patients. Fragmented urinary NEP bands (70 kDa) were quantified, and significant increase in urinary NEP expression was observed in patients with diabetes compared with ND (*P < 0.02). C: urinary ADAM17 expression in ND and patients with diabetes according to volume equivalent to 10 μg creatinine. Kidney lysate (2 μl) and urine (2 μl) from diabetic mice were used as positive controls. The diabetic mouse kidney shows several immunoreactive bands for ADAM17 (~93 kDa, ~65 kDa, and ~55 kDa). A 55-kDa immunoreactive band was observed in urine obtained from diabetic mice. In humans, a 70-kDa immunoreactive band for urinary ADAM17 was only seen in patients with diabetes with increased intensity in patients with diabetes compared with ND subjects (*P < 0.0001). D: urinary albumin was determined in volume equivalent to 10 μg creatinine. A predominant band for albumin at 66 kDa was detected in kidney and urine obtained from diabetic mice and patients with diabetes. Significant increase of urinary albumin expression was observed in patients with diabetes compared with ND (*P < 0.002). ACE2, angiotensin converting enzyme 2; ADAM17, a disintegrin and metalloproteinase 17; Dmacro, patients with type 2 diabetes with macroalbuminuria; Dmicro, patients with type 2 diabetes with microalbuminuria; Dnormo, patients with type 2 diabetes with normoalbuminuria; ND, nondiabetics; NEP, neprilysin.

Fig. 3.

Urinary and plasma NEP ELISA in ND, Dnormo, Dmicro, and Dmacro subjects. A: urinary NEP concentration in ND and patients with diabetes. *P < 0.03 vs. ND. B: plasma NEP concentration in ND and patients with diabetes. *P < 0.05 vs Dmicro. Values are the mean ± SE. Dmacro, Type 2 diabetic patients with macroalbuminuria; Dmicro, Type 2 diabetic patients with microalbuminuria; Dnormo, Type 2 diabetic patients with normoalbuminuria; ND, nondiabetics; NEP, neprilysin.

Fig. 4.

Mass spectrometric based enzyme activity assay for urinary ACE2 and NEP in ND, Dnormo, Dmicro, and Dmacro subjects. A: urinary ACE2 activity was measured in volume equivalent to 25 µg of creatinine. Samples were incubated for 2 h at 37°C in 0.4 M MES buffer pH 6.75 containing 0.1 mg/ml Ang II. B: urinary ACE2 activity inhibition in the presence of specific ACE2 inhibitor, MLN-4760 (0.1 mM) C: urinary NEP activity was measured in volume equivalent to 25 µg of creatinine incubated for 2 h at 37°C in 0.4 M MES buffer pH 6.75 containing 1 mg/ml Ang I. D: urinary NEP activity inhibition in the presence of specific NEP inhibitor thiorphan (0.1 mM). ACE2, angiotensin converting enzyme 2; And I, angiotensin I; Ang II, angiotensin II; Dmacro, patients with type 2 diabetes with macroalbuminuria; Dmicro, patients with type 2 diabetes with microalbuminuria; Dnormo, patients with type 2 diabetes with normoalbuminuria; m/z, mass-to-charge ratio; ND, nondiabetics; NEP, neprilysin.

Fig. 5.

Linear regression analysis of urinary ACE2 and NEP with renal functional parameters. (A–D): urinary ACE2 activity was significantly and positively correlated with HbA1C (r = 0.42, **P < 0.0003), BUN (r = 0.28, *P < 0.014), and SCr (r = 0.25, *P < 0.03) but negatively correlated with eGFR (r = −0.22, *P < 0.04). (E–H): a linear regression analysis of urinary NEP activity was only significantly and positively correlated with HbA1C (r = 0.33, *P < 0.01) but no correlation with eGFR, BUN, or SCr. ACE2, angiotensin converting enzyme 2; eGFR, estimated glomerular filtration rate (ml/min/1.73 m²); HbA1C, glycated hemoglobin; BUN, blood urea nitrogen (mg/dl); NEP, neprilysin; SCr, serum creatinine (mg/dl).

Fig. 6.

ROC curves to predict HbA1C (>6.5) and eGFR (<60 ml/min/1.73 m²). A: ROC curves for urinary ACE2 (AUC = 0.6117) and urinary NEP (AUC = 0.6211) for the prediction of HbA1C (>6.5%). B: ROC curves for urinary ACE2 (AUC = 0.6117) and plasma NEP (AUC = 0.6211) for the prediction of eGFR (<60 ml/min/1.73 m²). C: the addition of age, urinary ACE2, and urinary NEP to the model for predicting HbA1C (>6.5) in all the patients (AUC = 0.7367). D: inclusion of age, urinary ACE2, and plasma NEP to the model for predicting eGFR (<60ml/min/1.73 m²) in all the patients (AUC = 0.76). ACE2, angiotensin converting enzyme 2; AUC, areas under the ROC; eGFR, estimated glomerular filtration rate; HbA1C, glycated hemoglobin; NEP, neprilysin; ROC, receiver operating characteristic.