We used ALK1 heterozygous mice to evaluate the role of ALK1 in the early changes of the ureteral obstruction. Alk1 ^+/− mice were generated as previously described (Oh et al., 2000). Adult Alk1 ^+/− mice were kept in the pathogen-free facilities for genetically modified mice of the Animal Experimentation Service, University of Salamanca. Genotype analysis was performed by PCR with DNA isolated from mouse tail biopsies and using the primers previously reported (Oh et al., 2000).

Unilateral ureteral obstruction (UUO) is an experimental model of renal injury, which causes tubular cell injury, inflammation and fibrosis. UUO has been used as a model for the events that take place during chronic kidney disease (Ucero et al., 2014).

UUO was performed during 3 days, as we aim to evaluate the early changes of this experimental approach. The unilateral ureteral obstruction (UUO) was performed as previously described (Rodríguez-Peña et al., 2002; Grande et al., 2009). In brief, 8 weeks old male mice were anesthetized with Isoflurane (Schering-Plough, Madrid, Spain). After laparatomy, we used non-reabsorbable 5-0 silk to ligate the left ureter. To generate sham operated mice (SO), we manipulated the left kidney ureter without ligation.

In this study, 5 mice were included in each experimental group. Animals were kept under controlled ambient conditions in a temperature controlled-room with a 12 h light/dark cycle, and were reared on standard chow (Panlab, Barcelona, Spain) and water ad libitum. In all procedures, mice were treated in accordance with the Recommendations of the Helsinki Declaration on the Advice on Care and Use of Animals referred to in: law 14\/2 007 (3 July) on Biomedical Research, Conseil de l´Europe (published in Official Daily N. L358/1-358/6, 18-12-1986), Government Spanish (Royal Decree 223/1 988, (14 March) and Order of 13-10-1989, and Official Bulletin of the State b. 256, pp. 31349-31362, 28-10-1990). The procedure was approved for the Bioethics committee of the University of Salamanca and Consejería de Agricultura y Pesca (Junta de Castilla y León).

Obstructed (O) and contralateral kidneys (NO), as well as kidneys of sham operated mice (SO), were removed 3 days after surgery after perfusion with heparinized saline solution at 37°C in order to eliminate red blood cells from the tissue and to avoid endogenous peroxidase signals in immunohistochemistry procedures. Next, kidneys were halved longitudinally in order to use one half for protein extraction and analysis and the other half for stainings and immunohistochemistry. Renal samples for protein extraction were frozen in liquid nitrogen and stored at −80°C. Renal samples for histological studies were fixed for 24 h in formaldehyde, transferred to ethanol 70% and then embedded in paraffin.

Western blot was performed for protein levels analysis in mouse kidney tissue. Tissue protein extracts were homogenized in lysis buffer (150 mmol/L NaCl, 1% Igepal CA-630, 10 mmol/L MgCl₂, 1 mmol/L EDTA, 10% glycerol, 1 mmol/L Na₃VO₄, 25 mmol/L NaF, l mmol/L PMSF, 10 mg/ml aprotinin and 10 mg/ml leupeptin) containing 25 mmol/L HEPES, pH 7.5 and centrifuged at 14000 g during 20 min at 4°C. Supernatants were recovered and the protein concentration was quantified using Lowry Method (BioRad). Lysates (100 µg per lane) were loaded onto polyacrylamide gels and the proteins were transferred to PVDF membranes (Millipore, Billerica, MA, United States) by electroblotting. Next, membranes were blocked in bovine serum albumin (BSA) 3% and were incubated overnight at 4°C with the following antibodies: rabbit anti-collagen type I (dilution 1:1,000) and rabbit anti-fibronectin (1:1,000) from Chemicon International (Temecula, CA, United States); rabbit anti-ACVRL1 (ALK1) (1:1,000) from Abgent (Derio, Spain); mouse anti-α-SMA (1:5,000) from Sigma-Aldrich (Madrid, Spain); mouse anti-PCNA (1:1,000) from Transduction Laboratories (Madrid, Spain); and mouse anti-GAPDH (1:40000) from Ambion (Barcelona, Spain). After overnight incubation with the primary antibodies, membranes were incubated with the corresponding horseradish peroxidase-conjugated secondary antibodies during 60 min (1:10000) and were developed using ECL chemiluminescence reagent (Amersham Biosciences, Barcelona, Spain). Developed signals were recorded on X-ray films (Fujifilm Spain, Barcelona, Spain) for densitometric analysis (Scion Image software, Frederick, MD, United States). GAPDH (Ambion, Barcelona, Spain) was used as loading control. GAPDH was incubated in the same membrane that the protein of interest when it was possible.

3 µm sections were cut from paraffin-embedded samples and stained with haematoxylin-eosin, picrosirius, and Masson’s trichrome. Sirius red staining was evaluated by a quantitative scoring system, Fiji (https://imagej.net/software/fiji/), released as open source under the GNU General Public License in 12 randomly selected fields (200X) per experimental group.

Immunohistochemistry was performed on buffered formalin-fixed, paraffin-embedded tissues as previously described (Rodríguez-Peña et al., 2002; Grande et al., 2010) Briefly, 3 µm sections were deparaffinized in xylene and rehydrated in graded ethanols (100, 80, 70 and 50%) before antigen retrieval with sodium citrate buffer pH = 6.0, Then, primary antibodies were incubated overninght. Primary antibodies were: mouse anti-alpha smooth muscle actin (α-SMA, dilution 1:100, from Sigma-Aldrich), CD31 (Abcam) and rabbit anti-S100A4 (1:100 from Chemicon International, Temecula, CA, United States), mouse anti-VEGF (1:100 from Santa Cruz Biotechnology. After that staining continued with the peroxidase-antiperoxidase method. Tissue sections were incubated with the corresponding horseradish peroxidase-conjugated secondary antibodies during 60 min (1:250). After three washes in phosphate-buffered saline (PBS: 0.81% NaCl, 2.6 mM H₂KPO₄, 4.1 mM HNa₂PO₄), sections were sequentially incubated with the Novolink Polymer Detection System (Novocastra, Newcastle, United Kingdom) using 3,3′-diaminobenzidine (Biogenez, San Ramón CA, United States) as chromogen. Sections were counterstained with haematoxylin and were dehydrated and cover slipped. Endogenous peroxidase was blocked by incubation in 3% hydrogen peroxide. For an adequate optimization of the method, negative controls were prepared without primary antibodies. VEGF staining was evaluated using Fiji as mentioned above in 10–15 randomly selected fields (400X) per experimental group.

Paraffin-embedded tissues were cut in 3 µm sections. Heat-induced antigen retrieval was performed in sodium citrate buffer pH 6.00 in a microwave owen, and washed with PBS. Sections were incubated with endomucin (Santa Cruz Biotechnology) in combination or not with anti-α-SMA (Sigma Aldrich) overnight at 4°. Following three washes in PBS, sections were incubated with anti-rat 488 Alexa and anti-mouse 546 (Molecular Probes, Barcelona, Spain), diluted 1:200 for 60 min at room temperature, washed in PBS, and stained with 2 µM Hoechst 33258 (Molecular Probes, Barcelona, Spain). Slides were rinsed in PBS and mounted in Prolong anti-fade (Invitrogen, Barcelona, Spain). Confocal images were photographed using a Zeiss Axiovert 200 M microscope and a Zeiss LSM 510 confocal module. All images were obtained with identical parameters for intensity, pinhole aperture, etc. Image manipulation of immunofluorescence analysis was performed using the same settings in all the samples and pictures shown in this manuscript, and following the image manipulation guidelines from the journal.

Abundance of peritubular capillaries was assessed by counting the number of CD31 or endomucin-positive microvessels in peritubular areas in the kidney cortex across 5 different fields per kidney. Glomerular capillaries were not considered for the analysis.

Apart from the blood vessel density analysis, vascular rarefaction was assessed by the individual quantitation of endomucin+ endothelial cells across 5 different fields per kidney (SO and O kidneys form Alk1 ^+/+ and Alk1 ^+/− mice). Myofibroblast emergence was also assessed by the individual number of α-SMA+ cells excluding vascular smooth muscle cells. Hoechst counterstaining helped us to identify individual cells. Endothelial-myofibroblast transition was assessed by counting the double endomucin+ and α-smooth muscle actin α-SMA+ cells versus total endomucin + cells across 5 fields per kidney.

Data are expressed as mean ± standard error of the mean (SEM). The Kolmogorov-Smirnov test was used to assess the normality of the data distribution. Comparison of means was performed by two way analysis of variance (ANOVA) with Tukey’s HSC post hoc test. Data was analyzed using Graph Pad Prism software 9.0. A “p” value lower than 0.05 was considered statistically significant.

O kidneys from both Alk1 ^+/+ and Alk1 ^+/− mice show the histological events that take place after the UUO: Tubular cell injury, tubular dilation, inflammation; as shown with the Haematoxylin-eosin staining (Figure 1A) and tubule-interstitial fibrosis, as shown with the Masson’s trichrome staining (Figure 1B). While we observed no differences in kidney injury between both genotypes, we detected lower tubule-interstitial fibrosis in Alk1 ^+/− mice.

One of the most representative features of obstructive nephropathy is the accumulation of ECM proteins in the tubular interstitium, such as collagens (collagen I or collagen III), fibronectin or laminin. After our analysis of the picrosirius red staining, we observed increased levels of collagens in O kidneys form Alk1 ^+/+ mice but not in O kidneys from Alk1 ^+/− mice (Figure 2).

There is also an increase in the expression of ECM proteins (collagen I, fibronectin) in O kidneys from Alk1 ^+/+ but not in O kidneys from Alk1 ^+/− mice evaluated by western blot (Figure 3).

Renal myofibroblasts emerge and proliferate in the first steps of obstructive nephropathy. After 3 days of UUO we observe an increase in the presence of myofibroblasts in the tubular initerstitium of Alk1 ^+/+ mice, evaluated by immunostaining of the myofibroblast markers α-smooth muscle actin (α-SMA) (Figure 4A) and FSP1/S100A4 (Figure 4C) and by α-SMA western blot analysis (Figure 4B) in O kidneys from Alk1 ^+/+ mice. However, we barely observe those increases in O kidneys from Alk1 ^+/− mice. (Figures 4A–C). As α-SMA is not only a specific maker for myofibroblasts, because it is highly expressed in vascular smooth muscle cells (VSMCs) we quantified the α-SMA+ individual cells by immunofluorescence with Hoechst counterstaining excluding VSMCs, and we show that O kidneys from Alk1 ^+/+ mice show a higher number of α-SMA+ myofibroblasts than O kidneys from Alk1 ^+/− mice (Figure 5). In O kidneys from Alk1 ^+/− mice the presence of α-SMA positive cells is reduced and correlates with the lower ECM deposition observed in these animals. Moreover, the increase in cell proliferation is lower in O kidneys from Alk1 ^+/− mice, assessed by proliferating cell nuclear antigen (PCNA) expression (Figure 4D). Taken all these together, we can suggest that ALK1 heterozygosity leads to a lower kidney fibrosis due to a reduced abundance and proliferation of myofibroblasts.

As mentioned before, microvascular rarefaction is a feature of tubule-interstitial fibrosis and it contributes to the progression of hypoxia and tissue fibrosis. In the early stages of UUO, there is a vessel regression phase in which endothelial cells undergo apoptosis and pericyte adhesion is disrupted (Kida et al., 2014). Several studies show that ALK1 is involved in vessel maturation and quiescence (Akla et al., 2018; Viallard et al., 2020).

We observe a decrease in blood vessel density in O kidneys from Alk1 ^+/+ mice, assessed by immunostaining of the endothelial markers CD31 (Figure 6A) and endomucin (Figure 6B), similar to that previously described in other studies performed in the UUO model (Kida et al., 2014). However, blood vessel density was maintained in Alk1 ^+/− mice after UUO, suggesting that ALK1 heterozygosity protects from vascular rarefaction in the UUO early stages.

As stated before, myofibroblasts in the obstructed kidney emerge from different origins such as proliferating local resident fibroblasts, bone marrow derived cells or vascular cells. Vascular endothelial cells and pericytes can transdifferentiate into myofibroblasts. To dissect the myofibroblast cells that arise from endothelial cells we have double-immunostained kidney sections with an endothelial marker (endomucin) and a myofibroblast marker (α-SMA). Thus, cells with double positive staining for endomucin and α-SMA are endothelial cells being transdifferentiated into myofibroblasts. We observed that these double stained cells are more abundant in O kidneys from Alk1 ^+/+ mice than in O kidneys from Alk1 ^+/− mice (Figure 7). This finding indicates that the lower abundance of myofibroblasts observed in ALK1 heterozygous mice is due to a lower transdifferentiation from endothelial cells and this correlates with the PTC stability after 3 days UUO in Alk1 ^+/− mice.

ALK1 was described by David et al. (2008) as a molecule which induces quiescence and inhibits endothelial cell proliferation and migration (Lamouille et al., 2002; David et al., 2008). Angiogenesis is a process linked with the development of tubule-interstitial fibrosis after UUO. We found an increase of ALK1 expression in O kidneys from both Alk1 ^+/+ mice which may indicate the beginning of the regression phase of angiogenesis after obstruction. As expected, we do not observe increased levels of ALK1 receptor in Alk1 ^+/− mice after UUO, suggesting a pro-angiogenic effect in these animals (Figure 8A). To elucidate the differences in the angiogenic process during vascular rarefaction after UUO, we have analyzed the levels of one of the most important angiogenic factors, vascular endothelial growth factor (VEGF). We detected no differences in VEGF expression after UUO in Alk1 ^+/+ mice but we observed a higher expression in O kidneys from Alk1 ^+/− mice (Figure 8B). Taken all these together, we suggest that the lower ALK1 levels and increased levels of VEGF in O kidneys from Alk1 ^+/− mice may correlate with a delay of the regression phase of angiogenesis that prevents endothelial-pericyte detachment, contributing to microvascular preservation and impaired endothelial to myofibroblast transdifferentiation.