Soluble Urinary Somatic Angiotensin Converting Enzyme is Overexpressed in Patients with Preeclampsia: A Potential New Marker for the Disease?

Abstract

Objective:

The aim of this study was to identify and quantify urinary Angiotensin-Converting-Enzyme (ACE) in hypertensive disorders of pregnancy.

Methods:

Urine samples were analysed by Western blot. Patients were classified into: normotensive pregnancy (N); preeclampsia and superimposed preeclampsia (PE+SPE); and gestational hypertension (GH).

Results:

Somatic ACE protein expression was higher in PE+SPE compared to N and GH. There was a positive correlation between ACE and urinary protein to creatinine ratio, systolic and diastolic blood pressures.

Conclusion:

These results indicate ACE overexpression in the urine of preeclamptic patients and suggest that it may be a new marker for the disease.

Introduction

Hypertensive disorders of pregnancy (HDP) remain one of the leading causes of maternal and fetal morbidity and mortality worldwide [1-3], and comprise a wide range of conditions, such as chronic hypertension, gestational hypertension and preeclampsia, that are clinically characterized by elevated blood pressure (BP) – systolic blood pressure (SBP) ≥ 140 mmHg and/or diastolic blood pressure (DBP) ≥ 90 mmHg [1,4-6]. Gestational hypertension (GH) refers to a new onset BP increase during pregnancy, presenting only after 20 weeks of gestation [4,6]. Preeclampsia (PE) is diagnosed when GH is followed by proteinuria and/or other maternal organ dysfunction [1,4,6]. When this condition occurs in patients with a history of hypertension prior to pregnancy, it is classified as superimposed preeclampsia (SPE) [5]. PE is a multisystemic and multifactorial disease whose pathophysiology is associated with endothelial dysfunction, placental hypoperfusion, oxidative stress, inflammation, and vasoconstriction/vasodilation imbalance [7,8].

The renin-angiotensin-aldosterone system (RAAS) has a pivotal role in the regulation of BP and fluid balance, as well as in inflammation and immune responses [9]. Angiotensin converting enzyme (ACE) generates Angiotensin II (Ang II), the most vasoactive peptide of the system, through the cleavage of Angiotensin I, and inactivates Bradykinin, an important vasodilator agent [9,10]. ACE is a zinc-dependent dicarboxypeptidase that exists in two forms: somatic and germinal [10]. Testicular ACE mRNA (3.0 kb) and endothelial ACE mRNA (4.3 kb) result from differential splicing of a single gene transcript giving rise to somatic ACE (190 kDa) and a germinal, testicular isoform (90 kDa), respectively [11,12]. Despite somatic ACE commonly being reported as an ectoenzyme anchored to the cell membrane surface observed in a variety of tissues and organs [13-29], it can also be released as a soluble enzyme in extracellular fluids, such as urine and plasma, by a process of proteolytic shedding [10,30-33]. According to molecular weight, three different ACE isoforms have been described in human urine: 65 kDa, 90 kDa and 190 kDa [34-41]. While somatic isoform (190 kDa) contains both N and C ACE catalytic sites, 65 kDa and 90 kDa contain only the N-terminal fragment [34,35]. The presence of ACE isoform of 90 kDa has been found in the urine of hypertensive patients, suggesting its potential role as a biomarker [40,41]. A great deal of attention, therefore, has been given to low-molecular weight isoforms, but urinary 190 kDa ACE has not been thoroughly studied. Urinary profile of ACE specifically in pregnancy or HDP has not been studied. This prompted us to explore whether the soluble ACE isoforms are present in the urine of pregnant women. We also wanted to compare their expression between pregnant normotensive and patients with preeclampsia.

Material and methods

Subjects and sample collection

Previously acquired urine samples from an ongoing Mayo Clinic study [42] were selected retrospectively to conduct this cross-sectional research. The study was approved by the Mayo Clinic Institutional Review Board (protocol # 2105-05). All participants gave written informed consent and aliquots of random midstream urine samples (25-50 ml, each) were collected in sterile containers within 24 hours of delivery. Urine samples were immediately stored at −80°C for future use. Thirty-four random urine samples to describe, quantify and compare soluble somatic ACE isoform expression between women with HDP and normotensive pregnancies were used for the present study.

Subjects were classified according to published criteria [1,42] into three distinct groups: 1) normotensive (NT); 2) preeclampsia and superimposed preeclampsia (PE+SPE); 3) gestational hypertension (GH). Preeclamptic and superimposed preeclamptic patients were combined into the same group due to pathophysiological similarities. Exclusion criteria included gestational diabetes and samples with inadequate protein content that would not allow further lab experiments. Patient characteristics were accessed from patient charts.

Soluble somatic ACE purification

Urine volume was measured, and pH was corrected to 8.0 with 1mol/L TRIS buffer. Samples were centrifuged at 3,000 rpm for 10 minutes. A 20 ml supernatant subsequently was separated and concentrated 10x using Amicon Ultra 15 ml Centrifugal Filters with a 30 kDa molecular weight exclusion membrane (Millipore Sigma, USA) by a series of centrifugations (3,000 rpm) against 50 mmol/L Tris-HCl, pH 8.0, containing 150mmol/L NaCl. The concentrated urine (1.0ml) then was submitted for a protein estimation assay and Western blot analysis.

Protein concentration determination

Protein concentration was determined based on the bicinchoninic acid (BCA) (Bio-Rad) protein assay method as described previously [43].

Western blot analysis

Sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) 7.5% was performed as described by Laemmly [44] using 50 μg of total urinary protein denatured and reduced under dissociating conditions. Electrophoresis was initiated at 40 V, and after thirty minutes was performed at 120 V for two hours. Polyvinylidene difluoride (PVDF) membranes (Bio-Rad, USA) activated with methanol were used. Transfer to the membrane was executed at a constant voltage of 60 V for one hour and twenty minutes. The membrane then was incubated in a 5.0% nonfat dry milk blocking solution (Bio-Rad, USA) for one hour before overnight incubation at 4°C with ACE mouse monoclonal antibody 5C5 (sc-23909, Santa Cruz, USA) (dilution 1:100). A human kidney extract (sc-363764, Santa Cruz, USA) was used as a positive control. Normalization of Western blot data was performed using Ponceau S (Bio-Rad, USA) staining for total protein. The results were analyzed with Image Lab (Bio-Rad, USA) and described in arbitrary units (A.U.).

Statistical Analyses

The absolute and relative data of continuous variables were presented as: mean ± standard deviation, for parametric variables; median with 95% confidence interval (95% CI), for nonparametric variables; and number with percentage, for categorical variables. Shapiro, Wilk, D’Agostino & Person or Q-Q (quantile-quantile) plots were used to verify the nature of the data between the variables. Data analytics was performed using Student’s t-test or ANOVA to compare parametric variables; Mann-Whitney or Kruskal Wallis for nonparametric variables; and Fisher Exact test or Chi-square test to compare categorical variables. Correlation between variables was assessed by the Spearman correlation coefficient. Null hypothesis was rejected when p<0.05.

Results

A total of 34 urine samples were available for analysis. One patient was excluded for gestational diabetes mellitus. No precipitate was observed during sample centrifugation. After determination of protein concentration, 10 samples were excluded for inadequate protein content. Total protein content was estimated using Ponceau S (Supplemental material 1) and Western blot analyses were performed on urine samples from 23 subjects. According to the stated criteria, patients were classified into three different groups: 7 patients were in the normotensive pregnancy group (NT); 9 patients were in the preeclampsia and superimposed preeclampsia group (PE+SPE) – 6 of which were classified as PE and 3 as SPE; 7 patients were in the GH group. Patient characteristics are described in Table 1.

Table 1.

Clinical characteristics of studied pregnant women.

Characteristics	Normotensive Pregnancies (n = 7)	Preeclampsia and Superimposed Preeclampsia (n = 9)	Gestational Hypertension (n = 7)	P Value
Maternal age, years	34.0 [28.0–36.0]	31.0 [26.0–36.0]	36.0 [24.0–41.0]	0.359
Gestational age at sample collection, weeks	39.4 [37.0–41.8]	34.5 [30.1–38.4]	37.2 [34.8–41.4]	0.008*
Gestational age at delivery, weeks	39.4 [37.1–41.8]	34.7 [30.5–38.4]	37.2 [35.2–41.5]	0.010*
Systolic blood pressure, mmHg	117.0 [95.0–134.0]	148.0 [138.0 – 158.0]	133.0 [118.0–149.0]	0.001*
Diastolic blood pressure, mmHg	65.0 [57.0–67.0]	82.0 [72.0–92.0]	75.0 [66.0–80.0]	<0.001*
Birth weight, g	3,540 [2,670–4,020]	2,910 [1,610-3,270]	3,150 [2,930-4,320]	0.034*
Induction, n (%)	2 (28.5)	5 (55.5)	5 (71.4)	-
Caesarean section, n (%)	1 (14.2)	4 (44.4)	1 (14.2)	-
Singleton pregnancy, n (%)	7 (100)	8 (88)	7 (100)	-
Serum creatinine, mg/dl	N/A	0.60 [0.50–0.70]	0.70 [0.50–0.90]	0.211
Urine protein-to-creatinine ratio, mg/mg	N/A	0.51 [0.33–0.63]	0.15 [0.05–0.29]	<0.001*

Data are described as mean ± standard deviation, median [95% confidence interval], or number (percentage). Statistical analysis was performed using Mann-Whitney or Kruskal Wallis test.

^*Statistically significant differences were observed between the groups for the variables: gestational age at sample collection, gestational age at delivery, systolic blood pressure, diastolic blood pressure, birth weight and urine protein to creatinine ratio. kg: kilograms. m²: square meters. mmHg: millimeters of mercury. G, grams; mg. milligrams; dl, deciliter; N/A, not available.

The soluble 190 kDa ACE isoform was identified, but neither the 65 kDa nor 90 kDa isoform were noted in the urinary samples of pregnant women. Western blot results demonstrated statistically significant differences among the three groups (Figure 1). Urinary somatic ACE isoform excretion was higher in the group of PE+SPE patients (0.3750, 95%CI [0.1483 – 0.8919] A.U.) compared to both N (0.1427, 95%CI [0.0945 – 0.1924] A.U., p=0.016) and GH (0.1339, 95%CI [0.0875 – 0.2466] A.U., p=0.022); no significant statistical difference was observed in the comparison between NT and GH groups alone.

Figure 1.

Western blot results for urinary Angiotensin Converting Enzyme isoform (190 kDa). (A) Representative image in grayscale from Western blot. First lane: positive control. Lanes 2-8: normotensive pregnancies (NT). Lanes 9-17 (PE+SPE): the first 6 lanes correspond to Preeclamptic patients (PE) and the last three lanes to Superimposed Preeclampsia (SPE). Lanes 18-24: Gestational Hypertension group (GH). (B) Patients with Preeclampsia (PE+SPE) had higher expression of 190 kDa ACE in the urine compared to their normotensive counterparts (p=0.016) and pregnant women with gestational hypertension (p=0.022). No significant difference was observed in urinary ACE expression in normotensive vs. women with GH (p=0.871). N: normotensive pregnancies. PE+SPE: Preeclampsia and Superimposed Preeclampsia group. GH: gestational hypertension patients. A.U.: arbitrary units. Data are presented as median [95% confidence interval]. Comparisons between two groups were performed using Mann-Whitney test. Kruskal Wallis test was used for whole group comparison (p value marked with “*”).

In an additional analysis, when PE and SPE were separated into two different groups, the Kruskal Wallis test was able to detect a significant statistical difference between the groups (N 0.1427, 95%CI [0.0945 – 0.1924] A.U.; PE 0.3903, 95%CI [0.1215 – 0.8919] A.U.; SPE 0.3750 [0.1828 – 1.0000] A.U.; GH 0.1339, 95%CI [0.0875 – 0.2466] A.U., p=0.047).

Spearman correlation coefficient demonstrated a moderate positive correlation between Western blot quantification for the 190 kDa ACE isoform and SBP (Figure 2A) (r = 0.495 [0.091 – 0.759], p = 0.016), and DBP (Figure 2B) (r = 0.593 [0.226 – 0.812], p = 0.002). A significant correlation also was observed between ACE expression and the urinary protein to creatinine ratio (Figure 2C) (r = 0.571 [0.088 – 0.836], p=0.022), for the GH and PE+SPE together.

Figure 2.

The Spearman correlation between Angiotensin Converting Enzyme with 190 kDa quantification and systolic blood pressure (A), diastolic blood pressure (B) and urine protein:creatinine ratio (C). Significant correlation was observed between Angiotensin Converting Enzyme quantification and systolic blood pressure and diastolic blood pressure for the three different groups. Urinary protein to creatinine ratio was analyzed only in GH and PE+SPE together and there was a significant correlation between the variable and Angiotensin Converting Enzyme quantification. ACE: Angiotensin Converting Enzyme. mg: milligrams. mmHg: millimeters of mercury. A.U.: arbitrary unit.

Discussion

Our current results indicate a higher expression of the 190 kDa soluble urinary ACE isoform in patients with PE compared to both normotensive controls and GH. We additionally observed a positive correlation of ACE with SBP, DBP and urinary protein/creatinine ratio, suggesting its role in the pathophysiology of PE. Taken together, we conclude that an increase in urinary ACE may be an indicator of PE or its severity, which offers an opportunity to study RAAS dysregulation and potential therapeutic targets in the kidneys in PE. In addition, higher expressions of 190 kDa soluble urinary ACE in both PE and SPE compared to GH raise the possibility that RAAS dysregulation at the renal level may contribute to proteinuria, which is commonly viewed as a distinctive feature of preeclamptic compared to GH phenotypes.

The circulating components of RAAS play a major role in regulating blood pressure in a healthy pregnancy. Downregulation of angiotensin I receptor expression leads to the decreased vascular sensitivity and the vasopressor response to Ang II [45]. On the other hand, placental ischemia, oxidative stress and inflammation in preeclampsia may predispose dysregulated RAAS resulting in hypertension [46]. Despite no statistical difference in circulating levels of ACE in preeclampsia [47], an increase in enzyme activity has been reported in the sera [48-51] and in some tissues, such as the placenta [52] and endothelial cells [52], compared to normotensive pregnant controls. A murine model of preeclampsia also exhibited higher expression of placental ACE [53], but elevated activity of ACE does not result in a proportional increase in circulating Ang II [48,54-57]. However, despite lower circulating Ang II levels, women with preeclampsia exhibit increased vascular sensitivity to Ang II [57,58], resulting in an excessive pressor response. Finally, animal studies using genetically upregulated ACE activity successfully recapitulated critical findings of preeclampsia in humans: hypertension, proteinuria, decreased uteroplacental blood flow, loss of fetuses, while surviving fetuses exhibited restricted growth [59].

While systemic ACE is related to basal blood pressure regulation and proper kidney development, renal ACE seems to be associated with the pathophysiology of hypertension [60-62]. The synthesis of renal Ang II occurs independently of the circulating RAAS. ACE activity has been described in the urine of healthy human subjects by a series of studies that identified somatic ACE with a high molecular weight, corresponding to 190 kDa, and a low molecular weight isoform corresponding to 65 kDa [35,36,38,40]. The presence of the ACE isoform of 90 kDa has been found in the urine of hypertensive patients and hypertensive animal models [35,36,38,41,63]. ACE is abundantly synthesized in the kidneys by proximal and distal tubular cells, collecting ducts, and kidney endothelial cells [26] and maintains activity in the urine, as indicated by its ability to catalyse specific substrates [35,38,40]. Urinary ACE- expression generally corresponds to intrarenal RAAS activity [26]. A genetically modified mouse model, in which renal ACE was nearly or completely abolished while preserving ACE expression in other tissues, was protected against salt-sensitive hypertension during renal inflammation [64]. The expression of 190 kDa ACE also increases in the kidneys upon ischemic-reperfusion injury and inflammation [65]. The present study, to the best of our knowledge, is the first to explore ACE isoforms in urine samples of women with SPE and GH.

Only few studies have reported circulating ACE levels in HDP. Serum total ACE levels did not differ when comparing PE, GH and normotensive pregnancy [47,66]. Results of the current study suggest that the 190 kDa soluble urinary ACE isoform could be involved in the process of kidney impairment and proteinuria in the PE+SPE group, as the GH group showed no such elevation in 190 kDa soluble urinary ACE and was not different from the NT group. The possible clinical relevance of this observation was underscored further by a significant correlation between ACE expression and the urinary protein to creatinine ratio.

Our study differs from previous research in that it demonstrates the alteration in a urinary level instead of in the circulating systemic ACE expression. RAAS observed at a local level in different tissues could exert effects that are partially or totally independent from the systemic RAAS [67], reacting to a specific tissue demand [68]. While our results clearly demonstrate the presence of ACE in the urine during pregnancy, and its elevation in patients with PE/SPE, its respective roles in normotensive vs preeclamptic pregnancies remains unclear and will be addressed in future studies. We must highlight that our samples were collected within 24 hours of delivery and that there was a gestational age difference at the time of delivery between the three groups, which may have affected the results. We acknowledge that intrarenal RAAS may be unbalanced by multiples gestational factors such as cardiovascular and body fluid control adaptations in each trimester and its expression could fluctuate throughout pregnancy and also according to hypertension pathophysiology or severity. Since ACE expression can be elevated in hypertensive subjects, it would be worth investigating how urinary somatic ACE isoform 190 kDa behaves in SPE patients prior to and during pregnancy, particularly if it is being considered as a possible biomarker of the disease. Nevertheless, in our additional analysis, separating PE and SPE into two different groups, we were still able to detect significant statistical differences, despite the small sample size in each group.

Western blot analysis of urinary proteins is challenging. First, the total amount of protein excreted in this type of sample tends to be much smaller than that present in the serum, plasma or tissue, especially in patients without renal impairment and in the absence of significant proteinuria [73]. Second, the typical markers used for Western blot standardization, such as housekeeping proteins (often beta-tubulin, beta-actin or GAPDH), are not considered secreted molecules [74]. The total protein transferred to the membrane therefore has been considered as a more appropriate alternative to normalization [75-78]. Notably, other methodological approaches, such as chemiluminescence immunoassay, enzyme linked immuno absorbent assay or immune-capture enzymatic precipitation assay, would be helpful to predict total ACE levels. However, they are not able to detect the different isoforms, one of the aims of this study.

Using Western blot methodology, however, we were not able to identify all three isoforms as described in the literature [35-41], but rather only the 190 kDa isoform. Our findings suggest that intrarenal RAAS, mostly urinary somatic ACE, may be involved in the kidney injury observed in the PE+SPE group. It would be worth investigating whether this ACE isoform can be detected in the early stages of HDP and/or PE and further evaluated as a predictive tool.

The major limitations of our study are its small sample size and our inability to explore polymorphic ACE expression across the HDP/PE spectrum. We also acknowledge that protein degradation might have occurred, mainly due to unpredictable disease factors and pH alterations. Previous studies have demonstrated that lower pH conditions promote alterations in ACE protein structure and a reduction in enzymatic activity [34]. Despite these limitations, our study clearly indicates differences in ACE expression between PE and normotensive pregnant women, setting the stage for future longitudinal, comparative and mechanistic studies that may shed light on RAAS studies.

Supplementary Material

Supplemental Material

Supplemental Material 1. Total protein content estimation performed using Ponceau S staining. First lane: positive control. Lanes 2-8: normotensive pregnancies (NT). Lanes 9-17 (PE+SPE): the first 6 lanes correspond to Preeclamptic patients (PE) and the last three lanes to Superimposed Preeclampsia (SPE). Lanes 18-24: Gestational Hypertension group (GH).