Characterization of a murine model of endothelial dysfunction induced by chronic intraperitoneal administration of angiotensin II

Endothelial dysfunction (ED) is a key factor for the development of cardiovascular diseases. Due to its chronic, life-threatening nature, ED only can be studied experimentally in animal models. Therefore, this work was aimed to characterize a murine model of ED induced by a daily intraperitoneal administration of angiotensin II (AGII) for 10 weeks. Oxidative stress, inflammation, vascular remodeling, hypertension, and damage to various target organs were evaluated in treated animals. The results indicated that a chronic intraperitoneal administration of AGII increases the production of systemic soluble VCAM, ROS and ICAM-1 expression, and the production of TNFα, IL1β, IL17A, IL4, TGFβ, and IL10 in the kidney, as well as blood pressure levels; it also promotes vascular remodeling and induces non-alcoholic fatty liver disease, glomerulosclerosis, and proliferative retinopathy. Therefore, the model herein proposed can be a representative model for ED; additionally, it is easy to implement, safe, rapid, and inexpensive.


Results
Chronic i.p. administration of AGII induces endothelial dysfunction by increasing of soluble VCAM. It has been documented that the levels of both endothelium-bound and soluble adhesion molecules increase in endothelial dysfunction 15 . To confirm this in our model, soluble levels of VCAM were measured in plasma samples from AGII-treated mice and from control animals. As shown in Fig. 1, notably increased VCAM levels were found in plasma from AGII-treated mice with respect to untreated animals (24.5 vs. 10.8 pg/mL, respectively).
Chronic i.p. administration of AGII induces a prooxidant condition by increasing NOX2 and NOX4 transcription. Oxidative stress plays an important role in the pathophysiology of ED and associated cardiovascular diseases (CVDs), like SAHT, atherosclerosis, diabetes, cardiac hypertrophy, heart failure, and ischemia-reperfusion 16 . NADPH oxidases seem to be especially important for redox signaling, and this protein family could be a specific therapeutic target 16 . Six homologous forms of NADPH oxidases constitute the NOX family, which shares the capacity to transport electrons across the plasma membrane to generate superoxide and other downstream ROS. From this family, NOX2 and NOX4 are expressed in the endothelium, and AGII has been reported to induce their transcription. To analyze the expression of the NOX2 and NOX4 homologues in kidney, a real-time quantitative PCR assay was performed, with specific primers for NOX2 and NOX4 mRNAs. Primer specificity was confirmed by PCR amplification of the cloned cDNAs, where no cross-hybridization to other NOX cDNAs was observed (data not shown). The results of PCR analysis are shown in Fig. 2. The expression of NOX2 ( Fig. 2A) and NOX4 (Fig. 2B) mRNA showed a 4-and fivefold increase, respectively, in the kidneys of AGII-treated mice with respect to control animals (P < 0.05). www.nature.com/scientificreports/ AGII increased eNOS transcription levels. Both eNOS and iNOS produce nitric oxide (NO), but under different conditions. The former is an endothelial constitutive enzyme, and in this case, NO helps maintaining tissue homeostasis. The latter enzyme produces NO as a response to the activation of pro-inflammatory cells after damage or inflammation 17 . AGII has been reported to influence the expression of all three NO synthase (NOS) isoforms in the long term. To determine whose NOS isoform expression was modified by AGII, the expression of both eNOS and iNOS mRNA in mouse kidney was quantified by RT-PCR. After AGII administration, eNOS expression ( Supplementary Fig. S1A online) showed a 2.5-fold increase (P < 0.05) with respect to the control group. iNOS expression also increased, but not significantly (see Supplementary Fig. S1B online).
The main ROS source is xanthine oxidase. Oxidative stress is a key trait of ED. To determine whether ROS were produced in this model, and which enzyme is responsible for its production, various enzymes reported as ROS producers were evaluated in kidney extracts by quantitative DHE oxidation. Protein extracts from control or AGII-treated mice were exposed to a set of substrates before measuring ROS production (Fig. 3). ROS production increased almost 2 times in AGII-treated animals with respect to controls when the extracts were exposed to xanthine (P < 0.01). To confirm this result, the extract was incubated with xanthine and its competitive inhibitor allopurinol, which reduced ROS production. No differences were observed when the extracts were incubated with NADH, arginine, nor succinate. These data suggest that xanthine oxidase is involved in ROS production in response to chronic AGII administration, as expected in an ED model.

Figure 2.
Chronic AGII administration increases the expression of NOX2 and NOX4 mRNA in kidney. Data are reported as mean ± SD (n = 4 for each group) and analyzed with the Mann-Whitney U-test (P < 0.05); *Indicates differences with respect to controls. Figure 3. ROS production. AGII failed to induce changes in ROS production when NADH, arginine, or succinate were used as enzyme substrates, while using xanthine as a substrate increased significantly the xanthine oxidase production of ROS; this increase was prevented by incubation with the inhibitor allopurinol. Xan (Xanthine); Allop (Allopurinol). Eight kidney specimens were analyzed per treatment. Data are reported as mean ± SD and analyzed with the Mann-Whitney U-test (P < 0.05); *Indicate differences with respect to controls.
in the endothelium is a major contributor to the development of CVD 18 . It has been reported that CVD patients show increased plasma levels of cytokines like the tumor necrosis factor-alpha (TNFα) and interleukin 6 (IL6), as well as adhesion molecules like the intercellular adhesion molecule-1 (ICAM-1), the vascular cell adhesion molecule-1 (VCAM-1), and E-selectin, among other inflammatory markers 2 . Thus, inflammation, a central mechanism in the progression of ED and CVD 3 , was also evaluated in this model. Pro-inflammatory status was assessed in mouse kidney samples obtained after the last BP measure. Cytokine concentrations were determined by ELISA in tissue homogenates; ICAM-1 was detected by immunohistochemistry, and ICAM-1 mRNA was quantified by RT-PCR. As shown in Fig. 4, AGII-treated mice showed significantly increased TNFα (25.2%, On the other hand, a sixfold increase was observed in the expression of ICAM-1 mRNA in AGII-treated animals with respect to controls (P < 0.05, Supplementary Fig. S2G online). ICAM-1 was detected in perirenal fat tissue (Fig. S2D), renal capsule cells (Fig. S2E), and renal interstice cells (Fig. S2F), accompanied by an inflammatory cell infiltrate in renal capsule tissue (Fig. S2E), in the group treated with AGII, while in the control group ( Fig. S2A-C) it was not detected. These results indicate that i.p. AGII administration induces an inflammatory condition, as expected in ED.

Chronic AGII administration leads to hypertension and vascular remodeling. Hypertension
induction is one of the most important parameters expected in this model, because it is the most common clinical parameter of ED. AGII is actively involved in the pathophysiology of SAHT, since it triggers increased ROS production by endothelial cells 19 . By reducing NO bioavailability, higher ROS levels increase BP through a twoway mechanism: by altering the inner and outer diameter of blood vessels, thus reducing their lumen (vascular remodeling) 20 ; and by affecting the capacity of blood vessels to dilate as normal 17 . Acting together, both effects result in hypertension.
The measurements and calculations performed in histological slides of the BHA of the portal triad in mouse liver are shown in Fig. 5 (panels A, E). Vascular remodeling was assessed by determining medial thickness, lumen percentage, and the media/lumen ratio ( Fig. 5B-D, Fig. 5F-H). Medial thickness was calculated by subtracting the luminal BHA area from the total BHA area. As shown in Fig. 5, neither the total BHA vascular area, medial thickness, nor lumen area (panels B, C, and F respectively) were affected by AGII administration, but a twofold decrease was observed in lumen percentage with respect to control mice (P < 0.05, Fig. 5G). Medial thickness (Fig. 5D) and the media/lumen ratio (Fig. 5H) showed a 1.3-and 3.7-fold increase, respectively, which indicate hypertrophic vascular remodeling 18 . These results suggest that chronic AGII administration induces vascular remodeling, a key part of ED, hindering the flow of blood through the vessels and thus promoting hypertension. Data are reported as mean ± SD and analyzed with the Mann-Whitney U-test (P < 0.05); *Indicate differences with respect to controls.  Fig. 6. BP was measured at the beginning (week 0), the middle (week 5), and the end of experiment (week 10). A gradual increase of systolic BP (SBP, Fig. 6A) and diastolic BP (DBP, Fig. 6B) was observed in AGII-treated mice. On week 5, SBP and DBP increased by 15% (Fig. 6A) and 16% (Fig. 6B), respectively, with respect to control mice (P < 0.05). On week 10, SBP increased by 25% (Fig. 6A) and DBP increased by 24% (Fig. 6B, P < 0.05) with respect to control. These results indicate that this model is suitable to study SAHT and its causes, including vascular remodeling, and it also points to the effect of AGII on NO-dependent vascular function.
Chronic AGII administration provides a promising model of non-alcoholic fat liver disease, glomerulosclerosis, and proliferative retinopathy. Historically, AGII has been regarded as a primary factor of tissue damage 21 in different organs: it increases ROS production; it also promotes upregulation  . Blood pressure was measured in mice (n = 12 per group) on weeks 0, 5, and 10. Both blood pressure values were increased in AGII-treated mice on weeks 5 and 10, and the animals were regarded as hypertensive. Data are expressed as mean ± SD and were analyzed by ANOVA-Tukey test (P < 0.05). *Indicate differences with respect to control. www.nature.com/scientificreports/ in the expression of cytokines, cell adhesion molecules, and profibrotic factors like the transforming growth factor-β (TGFβ), which increases the synthesis of extracellular matrix proteins and promote macrophage activation and infiltration 22 , among other harmful responses. AGII treatment induced kidney damage resembling glomerulosclerosis 23 , hepatic damage similar to non-alcoholic steatohepatitis (NASH) 24 , and eye damage resembling proliferative retinopathy 25 . Glomerulosclerosis is a condition frequently resulting from ED; thus, AGII-induced kidney damage was assessed on Masson trichrome-stained perirenal fat tissue, capsules, perivascular connective tissue, and cortical glomeruli. The type of cells infiltrating the capsule was determined by PAS staining; finally, and a morphometric analysis of glomeruli was performed. As shown in Fig. 7, AGII-treated mice exhibited signs of renal damage, including renal capsule thickening (Fig. 7F) due to edema, fiber deposition, and infiltration of mononuclear cells, mainly lymphocytes and macrophages, characteristic of CKDs like glomerulosclerosis 26 . Additionally, perivascular fibrosis (Fig. 7G), congestion of the tubulointerstitial zone ( Fig. 7H) and points of mononuclear infiltration (macrophages and lymphocytes) among perirenal fat adipocytes (Fig. 7E) were observed, while in the control group these characteristics were not detected ( Fig. 7A-D).
In AGII-treated mice, the renal capsule thickening, which may be associated to the infiltration of inflammatory cells, mostly lymphocytes, macrophages, plasmatic cells, and fibroblasts, but also basophiles and neutrophils to a lesser extent ( Supplementary Fig. S3B), while these cells were not observed in the control group (Fig. S3A). As shown in Supplementary Fig. S3 online, an increase in glomerular area (37.0%, Fig. S3D-E), glomerulus vascular region hypertrophy (52.1%, Fig. S3F), and increased mesangial area (16.1%, Fig. S3G), all histopathological signs of glomerulosclerosis, were observed in AGII-treated mice compared to the control group (Fig. S3C, E-G).
On the other hand, three pathologic liver alterations were found in AGII-treated mice: 1) Steatosis with a perivascular pattern stemming from the central vein of the hepatic lobule (Fig. 8F); 2) thickening of trabeculae (Fig. 8H) and Glisson's capsule (Fig. 8I-J), and 3) lymphocytic microabscesses with central necrosis (Fig. 8G). All these alterations, steatohepatitis, inflammation, and fibrosis are pathological traits of non-alcoholic steatohepatitis (NASH), while these characteristics were not observed in the control group ( Fig. 8A-E). The thickening of trabeculae and Glisson's capsule is due to fiber deposition and the infiltration of mononuclear cells like lymphocytes, macrophages, fibroblasts, and fibrocytes, which account for fibrosis ( Fig. 8H-J), indicate the chronicity of the inflammatory event and the ongoing repairing process.
With respect to the eye, neovascularization is a hallmark of proliferative retinopathy, where the growth of abnormally formed blood vessels leads to hemorrhage, vision loss, and blindness 27 . As shown in Supplementary  Fig. S4, no significant differences in the number of blood vessels (Fig. S4A,B) were found between control and AGII-treated mice. However, AGII administration increased the formation of neovessels (P < 0.05, Fig. S4C,D).

Discussion
The implementation of an endothelial dysfunction (ED) model induced by the daily administration of a subchronic dose of AGII is reported herein. The model performance was confirmed by increased levels of the soluble adhesion molecule, sVCAM, a parameter known to increase during this pathology 15 . ED is characterized by alterations in endothelium activity in which oxidative stress and inflammation act simultaneously, leading to vascular remodeling, limited vasorelaxation, and fibrosis; in turn, these signs are linked to CVDs like hypertension, coronary artery disease, chronic heart failure, peripheral artery disease, cerebral vascular disease, diabetes, and chronic renal failure 4,28 . Interestingly, liver, kidney, and eye damage similar to that reported in non-alcoholic steatohepatitis (NASH), glomerulosclerosis, and retinopathy was observed. The severity of these diseases makes www.nature.com/scientificreports/ necessary to have a well-characterized animal model that represents ED pathophysiological frame. The factor that triggers the pathophysiological process is the cornerstone of the experimental model. Thus, we proposed an overstimulation of RAAS as the triggering factor for permanent endothelial damage, and a chronic administration of sub-effective doses of AGII was the factor chosen to accomplish this. AGII has been used to induce hypertension in various animal models for ED. The main differences between those models and the one herein reported is the route for AGII administration and the dose employed. The main advantages of this model are its low cost and ease of implementation since no additional equipment or materials are required. While the route of administration could indeed be uncomfortable for the mouse, the animals did not show any clinical or behavioral alteration suggestive of a poorer quality of life due to the drug intraperitoneal administration. On the other hand, the model herein described is minimally invasive, with no anesthesia, shaving, nor surgery required for AGII infusion as it is the case with osmotic pump implantation. As reported by Morton et al. 29 , if the inoculum is kept within pH limits (7.4) that do not cause alteration in the serosa of abdominal organs, and adequate volumes are inoculated (10 ml/kg), a daily intraperitoneal administration is safe. Both points were safely met by this model. In contrast with other models, we applied a dose of 0.1 µg/kg/day in a volume of 250-300 µl at a pH of 7.4 for 10 weeks to induce a condition analogous to the chronic degenerative process and the possible complications in human patients. Other models administer a higher dose within a shorter period; for instance, the dose used in the osmotic pump model ranges from 280 to 3200 µg/kg/day 10,30-32 depending on the author, with an overall duration of 1-4 weeks 10,31 , while a dose of 6 µg/kg/day was used in the model of AGII-impregnated pellets for 21-60 days 11 . Despite the variable doses and induction times, most models coincide in that systolic blood pressure values increased by about 40 mmHg 10,11 , as in our model. In addition to this parameter, we also showed that sVCAM, NOX2, NOX4, and iNOS expression, as well as ROS production, vascular remodeling, cellular infiltration, fibrosis, mesangial expansion, and liver inflammation increased after AGII intraperitoneal administration, along with the concentration of TNFα, IFNγ, IL1β, VCAM-1, ICAM-1 and TGFβ.
Some of these parameters have also been reported as increased in aorta 10,32,33 samples, peripheral blood cells 10 , heart 11 , kidney 31 , or liver tissues 34 in osmotic pump-or pellet-based models. This suggests that the condition observed in our model is similar to those found in those reported models, with the advantages of a lower cost and not requiring additional materials. As reported, AGII induced the production of ROS (such as superoxide) through NAD(P)H oxidase, xanthine oxidase, mitochondrial enzymes, and uncoupled NOS. These molecules are responsible for nitric oxide degradation linked to endothelial dysfunction 3,35 . As observed in our study, among these four enzymes, only xanthine oxidase (XO) produced ROS, a result in line with the work by Landmesser 35 , who reported a marked increase in endothelial XO levels and in XO-dependent endothelial superoxide production after AGII administration, suggesting that AGII-activated XO is a major superoxide source in coronary diseases resulting from ED. It is interesting to consider that, although eNOS and NADPH did not participate in ROS generation, a significant increase in mRNA expression for eNOS, NOX2, and NOX4 was observed, probably because superoxide ion is rapidly degraded by superoxide dismutase to produce hydrogen peroxide (H 2 O 2 ), as previously reported 36 ; H 2 O 2 is an extremely potent stimulus for the gene expression of eNOS 37 , NOX2 (mediated by JnK), and NOX4 (which proceeds through P38MAPK) [38][39][40] .
With respect to the inflammatory status observed in our model, AGII is known as a proinflammatory stimulant that attracts immune cells to the vascular wall, enhancing the production of cytokines and adhesion molecules 2,3 . Indeed, in our model, AGII-treated mice showed increased levels of inflammatory cytokines, ICAM-1 expression, and the infiltration of immune cells (macrophages, lymphocytes, plasmatic cells, basophils, and neutrophils) in the kidney, a well-known target organ for ED that plays a crucial role in the pathogenesis of ED and SAHT 5 . Moreover, the combined effect of AGII, oxidative stress, and inflammation induces fibrosis, as well as the proliferation and migration of vascular smooth muscle cells (VSMC), which further contribute www.nature.com/scientificreports/ to vascular remodeling ang hypertension 3,4 . The observed increase in blood pressure can be explained in part because AGII and ROS activate the ERK1/2 41,42 , JNK and NF-kB signaling pathways [43][44][45] , leading to the activation of matrix metalloproteinases (MMPs) and the production of growth factors like the epidermal growth factor (EGF), platelet-derived growth factor (PDGF), and insulin-like growth factor (IGF) 4,8 . These molecules promote the accumulation of extracellular matrix proteins (collagen and fibronectin) and the proliferation and migration of VSMC, which result in vascular remodeling 4,18 . While we did not study these signaling pathways, they are expected to be involved because of the high ROS levels observed upon AGII administration. High blood pressure values also depend on the levels of NO, which induces vasorelaxation. In this case, NO will not be available because it can react with ROS to produce molecules like peroxynitrite (ONOO − ); furthermore, peroxynitrite by itself affects NO production by uncoupling eNOS, so it will stop synthesizing NO, producing more O 2 − instead 46 . In our model, hypertension was induced by vasoconstriction and vascular remodeling, which could be related with a lower NO bioavailability.
The fact that the histopathological damage observed in this model resembled those found in NASH, glomerulonephritis, and human retinopathy, could be one of the most interesting results of our work, since they confirm the link between AGII and ED in the origin and progression of pathological damage. Since all affected organs are highly vascularized 47 , the status of the endothelium is key for the organ health. On the other hand, all three diseases have oxidative stress and inflammation as common traits in their origin and progression 24,25,48 .
The use of the intraperitoneal route to administer AGII implies that the changes observed, besides endothelial dysfunction, could be due to the effect of AGII on peritoneal tissues like the visceral adipose tissue, in which it induces differentiation and dysfunction, favoring thus a prooxidant and proinflammatory status 49 .
Oxidative stress and inflammation, induced by both AGII and ED, promote the progression of liver diseases like NASH 50,51 , whose histopathological characteristics, such as steatosis, cellular infiltrate, and fibrosis, were evident in the liver of AGII-treated mice. This prooxidative and proinflammatory environment induces metabolic changes in hepatocytes; indeed, TNFα 52 and ROS 50 induce insulin resistance, favoring lipolysis and increasing the serum levels of free fatty acids. In addition, both TNFα and ROS activate SREP-1, which in turn leads to lipogenesis 50 . On the other hand, AGII itself affects hepatic stellate cells (HSC), Kuffer cells, and hepatocytes 53,54 . The accumulation of lipids and their simultaneous degradation induce an erratic condition in hepatocytes that results in three events: increased ROS levels, cell death by necrosis, and inflammation 24 . The presence of a cellular infiltrate enriched with neutrophils and lymphocytes that accompanied necrosis and inflammation 24 , along with Kuffer cells, which are activated by AGII 54 , was observed in our model, as shown in Fig. 8. Fibrosis, a consequence of chronic liver damage in which extracellular matrix is produced to replace dead cells 53 , was also induced in out model. It has been proposed that fibrosis results from to ED (due to TGFβ activity), while AGII 53 induces the transformation of HSC into myofibroblasts, which eventually produce liver fibrosis 55 . These results indicate that ED, acting together with AGII, play an important role in the pro-steatotic environment that leads to NASH, further supporting our model of endothelial dysfunction induced by AGII.
Mice treated with AGII in our model also developed glomerulosclerosis. The pro-oxidant, pro-inflammatory and profibrotic environment induced by ED has been reported to induce an excessive accumulation of the extracellular matrix in the glomeruli, along with increased levels of vasoconstrictor substances, and an accumulation of type-I collagen and fibronectin. All these changes damage nephrons and may result in their loss 48 . It is known that AGII increases the synthesis of matrix molecules in the kidney, and it could stimulate macrophage infiltration, which favors the progression of sclerosis and promotes interstitial fibrosis, resembling glomerulosclerosis on histopathological examination 2,3 . The results herein reported are interesting because, while the mechanisms underlying the effects of AGII on renal cells are not fully understood, they highlight the effect of this octapeptide on nephrons, either directly or indirectly.
Finally, the ocular damage observed in our model was similar to that reported for diabetic retinopathy 25 . Diabetic retinopathy is due to a microangiopathy affecting retinal precapillary arterioles, capillaries, and venules. The damage is due to both microvascular leakage from a breakdown of the inner blood-retinal barrier and microvascular occlusion 56 . ED has been reported to contribute to ocular damage by oxidative stress 25 and the persistent low-grade inflammation that accompanies it 57 ; both conditions were found in mice treated intraperitoneally with AGII. On the other hand, the renin-angiotensin system is a causative factor of diabetic microvascular complications, inducing a variety of tissular responses, including vasoconstriction, inflammation, oxidative stress, cell hypertrophy and proliferation, angiogenesis, and fibrosis. All the components of the renin-angiotensin system, including angiotensin type-1 and type-2 receptors, have been identified in the retina of humans and rodents 58 . These results are of interest since, although the origin of this disease is not well understood 57 , our results indicate the relevance of ED and AGII in its establishment, which may be useful to develop novel treatment approaches.
In summary, our results show that chronic administration of AGII by a daily intraperitoneal injection leads to endothelial dysfunction characterized by prooxidant and proinflammatory activity, vascular remodeling, hypertension, and damage to organ targets associated with fibrosis.

Materials and methods
Animals and housing. All experiments were performed on 10-weeks-old, male C57BL/6 J mice. The animals were housed in groups of five in standard polypropylene cages. The mice were maintained in a 12:12-h light-dark cycle, fed with standard rodent chow and allowed freshwater ad libitum. All procedures were performed in accordance to the guidelines established by the official Mexican regulation NOM-062-ZOO-1999 (technical specifications for production, care and use of laboratory animals). Experimental protocols were reviewed and approved by the Ethical Committee for the Care and Blood pressure measurement. Systolic and diastolic BP was measured daily within the 11:00-15:00-h interval, to minimize circadian cycle-related variability 59 , under surgical anesthesia (xylazine, 10 mg/kg, i.p.); the animals were kept on a heat pad at 33-34 °C and under a cotton towel to prevent heat loss. Systolic and diastolic blood pressure was measured in mice by a non-invasive method, using a digital plethysmograph tail-cuff placed on the mouse tail; this tail-cuff is connected to a computerized system for data acquisition (LE5002 LETICA®, Panlab, Barcelona, Spain). This equipment closely approximates blood pressure measurement by a sphygmomanometer. The systolic and diastolic BP were measured at the beginning (baseline) and every 5 weeks (on week 0, 5, and 10) until the end of the experiment. Two weeks before the baseline BP measurement and the first AGII application, the mice were subjected to two training sessions, one week apart. Mice with a BP increase of 15% or higher with respect to the baseline (time 0) were considered as hypertensive.
Histopathology. After the last BP measurement, the mice were anesthetized with sodium pentobarbital (30 mg/kg, i.p.) and perfused with ice-cold PBS (NaCl 140 mM, KCl 2 mM, K 2 HPO 4 1.15 mM). The kidneys and livers were removed and fixed in Zamboni's solution. Then, the tissues were dehydrated and embedded in paraffin. Tissue sections (5 μm) were transferred to poly-L-lysine-coated slides (Sigma) and were deparaffinized and rehydrated. For histopathological studies, the slides were stained with either the Masson trichrome (kidney and liver), hematoxylin-eosin (eyes), or periodic acid-Schiff (PAS) stain (kidney). The Masson trichrome method combines hematoxylin stain with a cytoplasmic stain and a selective stain for connective tissue. The PAS stain, which detects polysaccharides (glycogen and mucosal substances like glycoproteins, glycolipids, and mucins) in tissues to identify alterations in the basement membrane, was used to assess the expansion of the mesangial matrix by the presence of increased amounts of PAS-positive material in the mesangial region 60  www.nature.com/scientificreports/ Glomerulosclerosis or hyalinization was defined as the disappearance of cellular elements from the tuft, a collapse of capillary lumen, and a folding of the glomerular basement membrane with entrapment of amorphous material, as proposed by Raij et al. 60 . Glomerular injury was also analyzed according to the method proposed by that author, with minor modifications. Briefly, 50-60 cortical glomeruli were evaluated in PAS-stained kidney slides from each group under a 100X objective. Digitized images were analyzed with Metamorph v.6.1. The mesangial area was calculated by subtracting the capillary area from the total area. The results were expressed as percentage with respect to the total area. Vascular remodeling was evaluated in the portal triad in Masson-stained liver slides. Total and luminal areas of the branch of hepatic artery (BHA) were measured in 10 slides per group under the 40X objective. The area corresponding to vessel thickness was determined by subtracting these two values. Lumen percentage, medial thickness percentage, and media/lumen ratio were calculated to assess vascular remodeling.

Cytokine and sVACM quantification by ELISA.
Kidneys were weighed and frozen at − 80 °C until used.
The organs were macerated in a frozen mortar with ice-cold PBS-PMSF (0.1%) 1:5 w/v. The suspensions were centrifuged, and supernatants were recovered and frozen at − 20 °C until used. Various ELISA kits were used to determine cytokine concentration, following the manufacturer's instructions. OptEIA Mouse IL1β, IL4, IL6, IL10, IFNγ, and TNFα ELISA kits were purchased from BD (Franklin Lakes, NJ, USA), while mouse IL17A and TGFβ ELISA kits were purchased from Applied Biosystems (Foster City, CA, USA). sVCAM was provided by Abacam (Waltham, MA, USA). Briefly, 96-well flat-bottomed ELISA plates were coated with the respective capture antibody and incubated overnight at 4 °C. Non-specific binding sites were blocked by incubating for 30 min at RT with PBS-5% fetal bovine serum. Aqueous kidney extracts were added and incubated for 2 h at RT. Then, the plates were incubated with the corresponding detection anti-cytokine-HRP antibodies for 30 min at RT. Bound complexes were detected by reaction with tetramethylbenzidine substrate after 30 min of incubation in the darkness. The reaction was stopped with H 2 SO 4 2 N and absorbance was measured at 450 nm at 37 °C in a VERSAmax ELISA plate reader (Molecular Devices). Cytokine and sVACM concentration was calculated according to standard curves for each cytokine and reported as pg/mg protein.
Statistical analysis. Data were analyzed using the software InStat (GraphPad, San Diego, CA, USA). Data are reported as mean ± standard deviation (SD) and analyzed with Mann-Whitney U-test. Analysis of blood pressure was performed using ANOVA-Tukey test. P < 0.05 was considered as statistically significant.