Experiments were carried out in C57BL/6 (wild-type, WT) and hR301Q α-GAL-Tg/KO mice. Specifically, these mice are homozygous for endogenous GLA knockout and express the human transgene GLA carrying the R301Q mutation under the transcriptional control of the human GLA promoter. These mice were kindly provided by AMICUS Therapeutics and used as a model of Fabry’s disease (17–19). The mice were housed in the animal facility of the Department of Translational Medical Sciences of Federico II University of Naples (Italy) in a 12-h light–dark cycle and provided access to a commercial mouse diet and water ad libitum. Animal husbandry and other experiments were conducted under Institutional Animal Care and the use of committee-approved protocols at Federico II University (Prot. n. 971/2016-PR). Aortas were collected from 9-month-old mice. Mice were euthanized, and aortas were quickly removed, rinsed in cold phosphate-buffered saline (PBS), and processed for successive analyses: one part was immediately frozen in liquid nitrogen for a biochemistry analysis, and another part was fixed in 4% paraformaldehyde for histological analysis.

Cryopreserved aortic vessels from α-GAL-Tg/KO mice and WT mice were pulverized and homogenated on ice in a RIPA buffer supplemented with protease inhibitors by repeated sonication and incubation for 1 h. The homogenates were centrifuged for 30 min at 13,000 rpm at 4°C and protein content in the supernatants was measured using a BCA assay (Thermo Fisher).

The total protein extracts (20 µg) were separated on TGX 4%–20% precast gradient gels (Bio-Rad, Germany) and transferred onto nitrocellulose membranes. The membranes were blocked in 5% milk (for anti-angiostatin antibody) or 3% milk (for anti-VEGFα and anti-α-actinin antibodies). The membranes were then incubated with the following primary antibodies:

Polyclonal rabbit-a-angiostatin (Abcam, ab2904) 2 µg/ml in 5% milk, overnight at 4°C; monoclonal mouse-a-VEGFα (Abcam, ab1316), 5 µg/ml in tris-buffered saline (TBS)/T, 1 h at room temperature; rabbit-a-α-actinin (Cell Signaling, 3134), 1:1,000 in 5% milk, overnight at 4°C. After washing with TBS/T, the membranes were incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies (sheep anti-mouse, Jackson ImmunoResearch 515-035-003, 1:20,000 in 3% milk; goat anti-rabbit, Jackson ImmunoResearch JIM-111-035-003, 1:10,000 in 5% milk), washed again, and specific signals were detected using the ECL or Femto system (Thermo Fisher) according to the manufacturer's protocol. A densitometric analysis of the protein bands was done using ImageLab (Bio-Rad).

Paraffin sections (3 µm) of aortic vessels from WT and α-GAL-Tg/KO mice were deparaffinized and incubated in a 10 mM citrate buffer at 95°C for 40 min for antigen retrieval. Afterward, the sections were incubated for 10 min in 0.5% Triton/TBS for permeabilization, washed in TBS, and quenched in a 0.25% Sudan black solution for 30 min on a shaker in the dark. After repeated washing, the sections were blocked in 3% bovine serum albumin (BSA) in TBS for 1 h at room temperature. The sections were then incubated with primary antibodies [anti-angiostatin (Abcam ab2904), 1:20 in 3% BSA or anti-VEGFα (Abcam, ab1316), 1:400 in 3% BSA] overnight at 4°C, washed in TBS, and incubated with secondary antibodies [donkey anti-mouse Alexa Fluor 488 (A21202 life tech) or donkey anti-rabbit Alexa Fluor 488 (21206 life tech), 1:500 in 3% BSA] for 2 h at room temperature. After repeated washing in TBS, the slides were mounted with DAPI Fluoromount G (Biozol) and fluorescence signals were descriptively analyzed using a Zeiss Axiovert M200 with ApoTome.

In this retrospective cross-sectional study, a total of 48 patients with genetically verified Fabry's disease was included and compared with 22 healthy control patients. Participants were enrolled during routine visits at the outpatient clinic of the University Heart and Vascular Center Hamburg between January 2014 and December 2020. A medical history of coronary artery disease (CAD), diabetes mellitus, arterial hypertension, and stroke was confirmed by self-report or the use of corresponding medication. A diagnosis of atrial fibrillation (AF) was established by a positive history and electrocardiogram (ECG) documentation within the last 5 years prior to examination. Cardiac symptoms were classified according to the New York Heart Association (NYHA) classification. All participants received a 12-lead surface ECG, a transthoracic echocardiography, and a routine blood test including a measurement of hsTNT (high-sensitive troponin T) and NTproBNP (brain natriuretic peptide), creatinine, GFR, and albumin in urine. The serum levels of NTproBNP were assessed by using the Atellica® IM NT-proBNP assay (Siemens Healthcare, Erlangen, Germany). For the measurement of serum hsTNT levels, the Elecsys Troponin T hs STAT assay (Roche Diagnostics, Rotkreuz, Schweiz) was used. In 40 patients with Fabry’s disease and in all healthy controls, cardiac magnetic resonance (CMR) imaging was performed to determine the extent of fibrosis as a marker of Fabry-associated cardiomyopathy. The study was conducted in compliance with the principles outlined in the Declaration of Helsinki and was approved by the local ethics committee (Ethikkommission der Ärztekammer Hamburg, Nr.: PV4056). All study participants gave their written informed consent.

All study subjects underwent a comprehensive transthoracic echocardiographic examination (Philips iE33 system, Philips Healthcare, Best, Netherlands) including M-mode, two-dimensional, pulsed, and continuous-wave Doppler and tissue Doppler imaging. Structural and functional imaging parameters were measured according to the current recommendations of the American Society of Echocardiography and diastolic dysfunction was classified according to current guidelines (20). Images were analyzed using the commercially available software Syngo Dynamics (Siemens Healthcare, Erlangen, Germany).

CMR was performed on a 1.5-T scanner (Achieva, Philips Healthcare, Best, Netherlands). The imaging protocol included cine imaging and late gadolinium enhancement (LGE) imaging. LGE images were acquired using a standard phase-sensitive inversion recovery (PSIR) sequence at least 10 min after a bolus injection of gadoterate meglumine (Dotarem, Guerbet, Sulzbach, Germany) in three long-axis orientations (2CH, 3CH, and 4CH views) and a stack of short-axis slices. Typical imaging parameters were as follows: voxel size 0.98 mm³ × 0.98 mm³ × 8 mm³, echo time = 2.39 ms, time to repetition = 4.97 ms, and flip angle = 15°. The presence of LGE was assessed by using the commercially available software cvi42 (Circle Cardiovascular Imaging Inc., Calgary, Alberta, Canada).

Protein serum levels in samples from patients with Fabry’s disease and healthy controls were quantified using enzyme-linked immunosorbent assay (ELISA) kits. For measuring and analyzing VEGFα serum levels, the ELISA kit BMS277-2TEN (Invitrogen) was used, and for determining angiostatin serum levels, the ELISA kit #ELH-Angiostatin (RayBio) was used, respectively, in accordance with the manufacturer's instructions.

A statistical analysis was performed using IBM SPSS Statistics (Version 28.0, Statistical Package for the Social Sciences, International Business Machines, Inc., Armonk, NY, USA). Continuous data are given as mean and standard deviation (SD). Categorical data are given as frequencies and percentages. Outliers were identified via evaluation of the standardized residues and included whenever measurement errors could be excluded. Statistical comparisons between the two groups were performed using the Student's t-test. Associations with disease severity parameters were analyzed by using Pearson's correlation tests and corrected for age- and sex-related differences. Logistic regression was performed to investigate the relationship between clinical and laboratory markers [albuminuria, lyso-Gb3, NTproBNP, hsTNT, and interventricular septal wall diameter (IVSd)] and VEGFα and angiostatin serum levels, respectively. *p-value ≤0.05 was considered statistically significant.

Western blots with aortic lysates from six α-GAL-Tg/KO and six WT mice revealed a significant increase in VEGFα protein levels in α-GAL-Tg/KO mice compared with that in WT mice (signal intensity = 1.66 ± 0.35 vs. 0.62 ± 0.16); *p = 0.0009), as shown in Figure 1.

Angiostatin expression showed a decrease in aortic tissue from α-GAL-Tg/KO mice compared with that from WT mice (signal intensity = 0.024 ± 0.007 vs. 0.053 ± 0.02, *p = 0.038), as shown in Figure 2.

Immunohistochemical staining of aortic rings showed an adventitial VEGFα signal in α-GAL-Tg/KO mice. No VEGFα signal could be detected in WT mice aortas (Figure 3). In contrast to the results of the Western blot analysis, there were no detectable differences in angiostatin expression levels between α-GAL-Tg/KO and WT mice (not shown).

In this study, we found increased endothelial VEGFα and reduced angiostatin levels in the aortic rings of mice with genetically defective α-GAL activity, as well as increased serum levels of VEGFα in patients with Fabry’s disease, highlighting the role of endothelial cell activation and oxidative stress in Fabry's disease.

Our data demonstrate for the first time a significant increase in VEGFα protein expression in aortic lysates from α-GAL- Tg/KO mice compared with those from WT mice. In Fabry's disease, storage of Gb3, as a result of deficient α-GAL A activity, was identified to induce oxidative stress in endothelial cells (21, 22), leading to an activation of inflammatory and pro-angiogenic reactions (23) in vitro and in vivo, including changes in the pattern of circulating miRNA (24). In this context, we analyzed the expression of the most potent pro-angiogenic factor VEGFα and its counter player angiostatin in the aortic rings of the Fabry R301Q transgenic/KO mouse model and compared this with the expression in the serum of patients with Fabry’s disease and healthy controls. In line with our data, treatment of bovine aortic endothelial cells with Gb3 induced the expression of angiogenic factors such as TGfβ, VEGFR2, VEGFα, and FGF2 (3, 25), demonstrating a pivotal role of angiogenetic factors in Gb3-induced vasculopathy. Further studies using the α-GAL-Tg/KO mouse model revealed a renal overexpression of VEGF and TGFβ in the presence of Fabry nephropathy (3).

In addition, immunohistochemical staining for VEGFα was performed in the aortic rings of α-GAL-Tg/KO and WT mice, which showed a higher degree of staining of VEGFα in α-GAL-Tg/KO mice. Although smooth muscle cells have been shown to be the major source of VEGFα expression and secretion (26), and accordingly in a model of hypertensive rats, VEGFα expression was mainly distributed in the outer to middle layers of the media (27), our data indicate that the VEGFα signal in mice aortic tissue was mainly increased in the adventitial layer. Whether this is a species-specific effect is not known. However, our observations are supported by a study in human atherosclerotic arteries analyzing the localization of different VEGF isoforms, also showing the most prominent staining for VEGFα in the adventitia along with some staining of the media (28).

To reveal the relevance of VEGFα expression seen in this in vitro model, we also measured VEGFα concentrations in the serum of 48 patients with Fabry’s disease and 22 healthy controls, which showed significantly higher VEGFα concentrations in the Fabry’s disease cohort. In line with other studies demonstrating VEGFα overexpression in patients with Fabry’s disease with cutaneous and systemic manifestations, this effect was more pronounced in male patients (29). This may be explained by the fact that male patients presented with a more severe clinical phenotype and a higher lyso-Gb3 level, which is usually attributed to lower enzyme activities in male patients compared with female patients. In this context, a proteomics-based analysis recently showed a correlation of VEGFα with lyso-Gb3 and residual enzyme activity in classical Fabry’s disease (30). Although an association with a higher lyso-Gb3 level could not be confirmed in our cohort, we show that VEGFα was positively associated with albuminuria, NTproBNP, and hsTNT, as markers for organ manifestation in Fabry's disease. However, this relationship was solely seen in male patients with Fabry’s disease.

No statistically relevant influence of co-medication on VEGFα levels could be found, although there was a trend toward lower VEGFα serum concentrations in patients with Fabry’s disease under ERT compared with those in the small group of untreated patients with Fabry’s disease. This may indicate a potentially beneficial effect of ERT, which has to be further evaluated in a larger cohort of patients with Fabry’s disease.

Angiostatin is an internal fragment of plasminogen and is generated by proteolytic cleavage through matrix metalloproteinases (MMP) and urokinase plasminogen activator (uPA) (31, 32). It acts as a potent inhibitor of angiogenesis and attenuates VEGF expression, resulting in the inhibition of extracellular matrix formation and migration (10, 11). Recently, we showed altered MMP9 and angiostatin levels in the serum of patients with Fabry’s disease compared with healthy controls, suggesting a higher extracellular matrix turnover in Fabry's disease (16). Accordingly, we found a suppressed angiostatin expression in aortic rings of the α-GAL-Tg/KO mouse model. In the serum of patients with Fabry’s disease, we also observed a trend toward reduced angiostatin serum levels; however, this did not reach statistical significance.