Autoantibodies to angiotensin-converting enzyme 2 in patients with connective tissue diseases

As many patients as possible among those with CTD and PAH or persistent digital ischemia (vasculopathy patients) in our hospital at time of the study were enrolled. Sera from these patients were studied in comparison with those from CTD patients without vasculopathy or from healthy subjects. The ethics committee of our hospital approved this study, and written informed consent was obtained from all patients and control subjects.

Fresh serum was obtained from all of the patients and normal subjects for the present study. Each serum sample was aliquoted to avoid repeated thawing and was stocked at -20°C until assayed.

Forty-two patients with SLE, SSc, or MCTD were studied. SLE was diagnosed according to the classification criteria of the American College of Rheumatology [10]. Patients with SSc met the classification criteria for the diffuse (n = 3) or limited (n = 6) form of SSc, as described in the literature [11]. Patients with MCTD met the criteria for MCTD from Kasukawa and colleagues [12] and the original definition of MCTD by Sharp and colleagues [13].

PAH had been diagnosed in five patients with SSc, based on dyspnea on exertion, elevated plasma brain natriuretic peptide levels >100 pg/ml, right ventricular outflow and peak tricuspid regurgitant pressure gradient exceeding 30 mmHg on echocardiography, exclusion of pulmonary thromboembolism by high-resolution computed tomography or pulmonary scintigraphy, and no deteriorated lung fibrosis that could cause pulmonary hypertension.

Each of the 16 patients in this category had persistent cyanotic lesions on the digits, present for more than 2 years at the time of the present study, and a history of or existing digital ulcers or necrosis. The radial artery pulse in each patient was normal, and atherosclerotic ischemia was excluded clinically. Three of the 16 patients also had PAH.

Serum samples from 24 CTD patients who had neither PAH nor persistent digital ischemia were collected randomly (that is, independent of the activity of their CTD) on the ward or in the outpatient clinic. These samples served as disease controls.

Serum samples were obtained from 28 healthy volunteers (23 females and five males) who were free from any active or chronic diseases, including hypertension. Before bleeding, each subject was confirmed to have normal blood pressure in the arms. The mean age of the 28 subjects was 32.0 ± 9.5 years.

We established an ELISA using purified recombinant human ACE2. A plasmid DNA-encoding human ACE2 cDNA, which was kindly gifted from Dr Hyeryun Choe (Harvard Medical School, Cambridge, MA, USA), was introduced into 293 free-style cells, according to the manufacturer's protocol (Invitrogen, Carlsbad, CA, USA). Culture supernatant was harvested on day 2 after transfection, dialyzed against 25 mM Tris-Cl (pH 8.5), and applied onto a DEAE column. ACE2 was eluted with 250 mM NaCl, and was dialyzed against for further use. About 12.5 μg/ml recombinant human ACE2 were first coated overnight onto a 96-well plate with bicarbonate buffer (pH 9.6) at 4°C. The wells were then treated with a blocking buffer composed of 5% BSA in PBS and washed with buffer composed of 20 mM Tris-HCl (pH 7.5), 150 mM NaCl, and 0.1% Tween 20. The sera of the patients and normal healthy volunteers were added to the plate and incubated for 1 hour at room temperature. Bound ACE2 antibodies were detected using horseradish peroxidase-conjugated anti-human IgG antibody. The optical density at 450 nm was measured after a 30-minute incubation with SureBlue TMB microwell peroxidase substrate (Kirkegaard & Perry Laboratories Inc., Gaithersburg, MD, USA). All samples were analyzed in triplicate, and the values were normalized with the positive control.

Zinc-dependent ACE2 enzymatic activity was measured according to the methods described previously [14, 15]. Briefly, 20 μM (7-methoxycoumarin-4-yl) acetyl-APK-(2,4-dinitrophenyl)-OH (amino acids depicted by single letters) (AnaSpec Inc., Fremont, CA, USA) - a fluorogenic substrate [14] - was incubated with 1 μl test serum. Fluorescence was monitored using Safire2 (TECAN, Männedorf, Switzerland) at the excitation and emission wavelengths of 320 nm and 450 nm, respectively.

Preliminary experiments revealed that the relative fluorescence unit (RFU) value was not increased when the substrate was simply incubated without the enzyme (data not shown). Among 20 healthy humans, two samples showed significantly high autofluorescence - in these cases, the RFU values decreased during the incubation even with the substrate. We excluded these samples because it was not possible to evaluate the ACE2 enzyme activity.

The RFU values obtained by the APK substrate were almost completely inhibited by an ACE2 inhibitor, DX600 (Phoenix Pharmaceuticals Inc., Burlingame, CA, USA) (Additional file 1). Additionally, the RFU values were statistically significant at both 60 and 70 minutes of the incubation (Additional file 2). Based on these observations, we used the RFU values at 70 minutes without subtracting the residual RFU counts resistant to DX600 [15]. The relative ACE2 activity (%) in each sample was further calculated based on the level of ACE2 activity in a mixture of serum samples from 28 healthy subjects (a reference serum). The assay was performed in triplicate and the mean values of three independent measurements of the ACE2 ELISA and ACE2 activity assay were used for the statistical analysis by Student's t test.

An anti-human ACE2 goat polyclonal antibody (R&D Systems, Minneapolis, MN, USA) was used for immunoprecipitation. Serum was first treated for 15 minutes with 100 mM MgCl₂, 70 units DNase I (Takara Bio Inc., Shiga, Japan) and 500 μg/ml Ribonuclease A (Sigma, St Louis, MO, USA) at room temperature, and was then incubated with anti-human ACE2 antibody at 4°C for 2 hours. After incubation, the immune complexes were recovered using Protein G Sepharose™ 4 Fast Flow (GE Healthcare, Amersham, Buckinghamshire, UK) and subjected to SDS-PAGE, followed by western blot analysis. To detect ACE2 proteins, an anti-human ACE2 mouse monoclonal antibody was used (R&D Systems).

Five microliters of serum were incubated with Protein G Sepharose™ 4 Fast Flow in 20 mM sodium phosphate buffer (pH 7.0) for 2 hours at 4°C. The immune complexes were then washed and the IgG-bound beads eluted with 0.1 M glycine-HCl (pH 2.7). The recovered IgG was immediately neutralized with 1 M Tris-HCl (pH 8.0). IgG was detected by SDS-PAGE followed by Coomassie Brilliant Blue staining. IgG fractions were prepared from three vasculopathy patients and three healthy subjects. The effect of each IgG on the activity of recombinant human ACE2 (standard rACE2; R&D Systems) was examined as described.

The demographics of patients with CTD and PAH or persistent digital ischemia (vasculopathy patients, n = 18) and of the control CTD patients (n = 24) are presented in Table 1. The control patients had neither manifestations nor a history of these vasculopathies.

All of the patients had received a maintenance dose of steroid therapy by the time of blood sampling. Serum samples were obtained from inpatients just before starting additional immunosuppressive therapy for a disease flare-up, and from outpatients. Sera from 17 of the 18 (94%) vasculopathy patients showed reactivity to ACE2 with ELISA values above the baseline level value, which was determined as the mean ELISA score in the 28 normal control subjects. The mean ELISA score was significantly higher in the vasculopathy patients than in the control patients (Figure 1). These results suggest that anti-ACE2 antibodies in the serum are associated with constrictive vasculopathies, PAH, or persistent digital ischemia.

IgG fractions from the sera of three healthy subjects and three vasculopathy patients (Patients 1, 5, and 6) were prepared (Figure 2a). The activity of standard rACE2 was measured in 5 μg aliquots. The IgG fractions from the vasculopathy patients, but not from the healthy subjects, suppressed the rACE2 enzyme activity significantly compared with that in the absence of IgG (Figure 2b).

First, the RFU value of the ACE2 enzyme activity, which was blocked by DX600 (Additional file 1), was determined in 18 patients with vasculopathy, in 16 control patients without vasculopathy, and in 26 healthy subjects. The relative rate of serum ACE2 enzyme activity was then compared with a reference mixture composed of sera from healthy subjects. Serum ACE2 activity could not be measured in one control patient and two healthy subjects due to high fluorescence at the start of incubation, and sera from the remaining seven control patients were not available for these experiments. The mean ± standard deviation relative ACE2 activity was significantly lower in the vasculopathy patients (64.8 ± 16.4, n = 18) compared with that in the healthy subjects (86.6 ± 32.2, n = 26, P = 0.0055) and control patients (165.1 ± 99.1, n = 16, P = 0.0010).

To clarify the correlation between ACE2 activity and the presence of autoantibodies to ACE2, the ELISA scores for serum anti-ACE2 antibodies and serum ACE2 activity were plotted (Figure 3). Our data indicate that the titers of anti-ACE2 antibodies were inversely proportional to ACE2 peptidase activity in the vasculopathy patients (Figure 3c). In contrast, no correlation was observed in the healthy subjects or control patients, suggesting that the low serum ACE2 activities in most of the vasculopathy patients were due to the presence of anti-ACE2 antibodies. To assess ACE2 expression in the vasculopathy patients, we tested for ACE2 by immunoprecipitation followed by immunoblotting. As shown in Figure 3d, the ACE2 levels in Patients 1 and 5, who were deficient in ACE2 activity (Figure 3c), were comparable with those in three healthy subjects, clearly indicating that the decreased ACE2 activity in the vasculopathy patients was due to a functional deficiency. The relative ACE2 protein levels in the patients are shown in Additional file 3. The signal intensities of ACE2 protein and the IgG heavy chain shown in Figure 3d were measured and normalized.

The exacerbated SLE and progressive necrosis of the fingers and toes in Patient 1, a 25-year-old female, were treated with intravenous methylprednisolone at a dose of 500 mg/day for 3 days, followed by 30 mg/day oral prednisolone combined with three sessions of double-filtration plasmapheresis and conventional therapy, including prostanoids, stellate ganglion anesthesia, and local care. Her cutaneous lesions healed completely before discharge, without amputation of any digits. Serum tests 2 months after the start of therapeutic intervention showed that the ELISA score for anti-ACE2 antibodies was reduced markedly (P < 0.05) (Figure 4a) and the ACE2 activity had recovered significantly (P < 0.05) (Figure 4b), compared with the values before therapy. An ELISA for anti-ACE2 antibodies in the waste double-filtration plasmapheresis fluid was negative, and thus we could not evaluate whether double-filtration plasmapheresis had removed these antibodies from the patient's blood. The anti-ACE2 antibodies were probably decreased as a result of immunosuppressive therapy using high-dose pulse steroids.