The Novel Compound Sul-121 Preserves Endothelial Function and Inhibits Progression of Kidney Damage in Type 2 Diabetes Mellitus in Mice

Diabetic nephropathy is still a common complication of type 2 diabetes mellitus (T2DM) and improvement of endothelial dysfunction (ED) and inhibition of reactive oxygen species (ROS) are considered important targets for new therapies. Recently, we developed a new class of compounds (Sul compounds) which inhibit mitochondrial ROS production. Here, we tested the therapeutic effects of Sul-121 on ED and kidney damage in experimental T2DM. Diabetic db/db and lean mice were implanted with osmotic pumps delivering Sul-121 (2.2 mg/kg/day) or vehicle from age 10 to 18 weeks. Albuminuria, blood pressure, endothelial mediated relaxation, renal histology, plasma creatinine, and H2O2 levels were assessed. Sul-121 prevented progression of albuminuria and attenuated kidney damage in db/db, as evidenced by lower glomerular fibronectin expression (~50%), decreased focal glomerular sclerosis score (~40%) and normalization of glomerular size and kidney weight. Further, Sul-121 restored endothelium mediated vasorelaxation through increased production of Nitric Oxide production and normalized plasma H2O2 levels. Sul-121 treatment in lean mice demonstrated no observable major side-effects, indicating that Sul-121 is well tolerated. Our data show that Sul-121 inhibits progression of diabetic kidney damage via a mechanism that involves restoration of endothelial function and attenuation of oxidative stress.


Results
Animal characteristics. Animal characteristics are shown in Fig. 1 and Table 1. As expected, db/db mice had a higher body weight compared to lean mice (Fig. 1A). db/db mice gradually lost weight as of week 4. However, Sul-121 did not affect body weight at any time point. Non-fasting blood glucose levels were significantly increased in the db/db mice compared to lean mice (Fig. 1B). During the course of the experiment, non-fasting blood glucose levels in the db/db animals rose from 27.5 ± 2.5 and 27.2 ± 1.3 g (db/db vehicle and db/db Sul-121, respectively) to 35.8 ± 2.4 and 35.3 ± 1.1 mM, indicating a severe and progressive diabetic state. Treatment with Sul-121 did not affect blood glucose levels.
The effects of Sul-121 on albuminuria. Urinary albumin excretion (AER) was measured throughout the duration of the experiment. Here, db/db vehicle mice showed an increase in AER from week 4 onwards reaching 1241 ± 204 µg/day by week 8. In db/db Sul-121, progression of AER was completely attenuated and AER remained stable at approximately 500 µg/day (Fig. 1C). AER in lean mice was measured at week 8 and amounted to 13.5 ± 1.9 and 12.5 ± 1.7 µg/day for lean vehicle and lean Sul-121 mice, respectively (Table 1). Glomerular damage and renal histology. As Sul-121 had a profound effect on albuminuria, we assessed damage to the glomerulus using histological staining and Western blotting. Periodic acid-Schiff stainings of kidney sections are shown in Fig. 2, panels A-D. Subsequently, glomerular size (Fig. 2E) and glomerulosclerosis scores (Fig. 2F) were determined. Glomerular size was increased in db/db vehicle (124 ± 8% of control), which was normalized in db/db Sul-121 (98 ± 4% of control). Similarly, glomerulosclerosis score was increased in db/ db vehicle (166 ± 16% of control), which was significantly attenuated in db/db Sul-121 (140 ± 10% of control).
Glomerular damage was further assessed by immunohistochemistry. Expression of fibronectin, an extra cellular matrix (ECM) protein, was found predominantly in glomeruli ( Fig. 3A-D) and was most pronounced in db/ db vehicle (Fig. 3C). Fibronectin expression was further quantified by Western blotting in whole kidney homogenates and found to be significantly upregulated in db/db vehicle (396 ± 19% of control) compared to their lean vehicle (100 ± 51%) (Fig. 3L). In db/db Sul-121, the increased fibronectin expression was significantly attenuated (194 ± 24% of control).
Podycyte loss was assessed by immunostaining for podocyte markers WT1 and synaptopodin ( Fig. 3E-G and Supplementary Fig. 1). WT1 staining was confined to the nucleus and podocyte numbers were determined by automated counting of the average number of WT1 positive cells per glomerulus (Fig. 3M). Compared to lean vehicle (10.7 ± 0.2 cells per glomerulus), the number of WT1 positive podocytes was significantly reduced in both db/db vehicle (8.6 ± 0.5 cells) and db/db Sul-121 (8.3 ± 0.5). For synaptopodin, a trend towards lower synaptopodin staining in db/db vehicle could be observed ( Supplementary Fig. 1A,B). However, the decreased synaptopodin staining in db/db vehicle did not reach statistical significance (p = 0.13). Expression of slit diaphragm proteins was assessed by Western blotting for Nephrin ( Supplementary Fig. 1C). No significant differences could be observed between groups.
Sirius red staining and polarized microscopy was used to visualize collagen I as marker of renal interstitial damage (Fig. 3H-K). No differences in collagen I expression could be detected between groups (Fig. 3N). Additional markers for renal interstitial damage were assessed by real-time PCR (Fig. 4). Similar to the findings of the Sirius red staining, Collagen I mRNA expression was not different between groups (Fig. 4A). Smooth muscle actin expression (Fig. 4B) was increased in db/db vehicle (1.43 ± 0.31 arb. units) over lean vehicle (1.07 ± 0.09 arb. units), but not different in db/db Sul-121 (1.51 ± 0.16 arb. units). In addition, macrophage marker Macrosialin (CD68) and TNF-alpha expression was not different between groups (Fig. 4C,D). Interleukin Il-1B was upregulated in db/db vehicle (1.70 ± 0.48 arb. units) over lean vehicle (0.51 ± 0.10 arb. units) with a non-significant reduction in db/db Sul-121 (1.05 ± 0.31 arb. units) (Fig. 4E).
Effects of Sul-121 treatment on reactive oxygen species. Hydrogen peroxide, a stable metabolite of ROS, was taken as a reference for ROS levels in plasma (Fig. 5A). Plasma H 2 O 2 was significantly increased in db/ db vehicle mice (0.46 ± 0.03 µM) compared to lean vehicle (0.33 ± 0.05 µM). Treatment of Sul-121 reduced H 2 O 2    Glomerular protein carbonylation was determined in all groups ( Fig. 5G and Supplementary Fig. 2). Surprisingly, lean Sul-121 showed increased protein carbonylation (118.0 ± 2.6 arb. units) over lean vehicle (106.8 ± 3.0 arb. units). In db/db vehicle, protein carbonylation was further increased (129.5 ± 3.2 arb. units), whereas this was not affected by Sul-121 treatment in db/db Sul-121 (126.0 ± 4.7 arb. units). The effects of Sul-121 on ROS production were further studied in an in-vitro model of diabetes (Fig. 5H). For this, cultured mouse renal mesangial cells were exposed to conditions simulating type 2 diabetes (high glucose and insulin). Exposure to high glucose and insulin increased the fluorescence of the ROS probe H2DCFDA by approximately 70% (Fig. 5H). Interestingly, Sul-121 pre-treatment did not affect basal levels of intracellular ROS, but substantially inhibited the increase in intracellular ROS provoked by glucose and insulin (p < 0.05).
To assess the effects of Sul-121 on renal mitochondrial ROS production, mitochondria were isolated from kidneys from an additional group of lean and db/db mice. Mitochondria obtained from db/db mice showed increased ROS production (3956 ± 241 arb. units) compared to lean mice (2832 ± 315 arb. units) (Fig. 5I). Acute treatment with Sul-121 lowered mitochondrial ROS production to similar levels in both lean control (1849 ± 206 arb. units) and db/db (2302 ± 193 arb. units).
Vascular function in diabetes and the effect of Sul-121. As ROS and diabetes have profound effects on endothelial function, we next established the effects of Sul-121 treatment on endothelial function of the mouse aorta. For this, mouse aortic rings were preconstricted with phenylephrine (PE) and subsequently relaxed with stepwise incremental concentrations of acetylcholine (ACh). Concentration effect curves were constructed (Fig. 6A), demonstrating an impaired relaxation to ACh in db/db vehicle compared to lean vehicle (max. relaxation 39.2 ± 6.5% and 9.2 ± 1.6% for db/db vehicle and lean vehicle, respectively). In db/db Sul-121 mice, vasorelaxation was completely restored to lean vehicle levels (11.3 ± 2.3%). To further investigate the individual endothelial components involved in vascular relaxation, dose response curves were constructed using specific inhibitors for eNOS (L-NMMA) and cyclo-oxygenase (indomethacin). The relative contribution of each component was calculated (Fig. 6B). In db/db Sul-121 mice, total relaxation was improved mainly by an increase in the NO and prostaglandin components (274 ± 55 and 429 ± 31 AUC for db/db vehicle and db/db Sul-121, respectively). In addition, the EDHF component was higher in db/db Sul-121 (76 ± 12 AUC) compared to db/db vehicle (49 ± 6 AUC), although this did not reach statistical significance (p = 0.07).
Furthermore, we assessed the sensitivity of the smooth muscle cells to NO, using the NO donor sodium nitroprusside (SNP) (Fig. 6C). Relaxations of PE preconstricted aortic rings to SNP were significantly attenuated only in db/db vehicle (92.8 ± 1.9%) when compared to lean vehicle (98.1 ± 0.4%).

Correlations.
To assess whether the improved ROS status or the normalization of endothelial function is contributing to the inhibition of renal damage progression in diabetic mice, we plotted correlations between albuminuria and plasma H 2 O 2 levels (Fig. 7A) and between albuminuria and maximum endothelial mediated relaxation to ACh (Fig. 7B). Plasma H 2 O 2 levels correlated positively with albuminuria (R 2 = 0.38 p < 0.05) whereas maximal endothelial mediated relaxation correlated negatively with albuminuria (R 2 = 0.33 p < 0.05) using data of the combined db/db vehicle and db/db Sul-121 groups.

Discussion and Conclusion
The present study is the first to document the beneficial action of long-term subcutaneous infusion of the novel compound, Sul-121, in a model of experimental T2DM. We demonstrate that Sul-121 provides two major benefits. The first is the attenuated progression of albuminuria and the limited extent of diabetes related renal damage in the db/db model of T2DM. This was made evident by the slower advancement of albumin excretion rate, decreased focal glomerular sclerosis score, reduced glomerular ECM protein expression and normalization of glomerular size and kidney weight. Secondly, Sul-121 inhibited the development of endothelial dysfunction and normalized systemic ROS formation, which was supported by in-vitro experiments. Sul-121 treatment in lean mice had no major observable side-effects, indicating that it is a well-tolerated compound. Thus, Sul-121 inhibits the progression of experimental diabetic kidney damage via a mechanism that inhibits both oxidative stress and preserves endothelial function. Importantly, treatment with Sul-121 did not affect metabolic and diabetic parameters. In both lean control and db/db mice, Sul-121 treatment had no effects on body weight, plasma glucose levels, water consumption or urine production, indicating that the protective effects of Sul-121 in diabetes were not mediated through a less severe diabetic state.
Rather, Sul-121 treatment had profound effects on the ROS status of the diabetic mice. First, Sul-121 normalized plasma H 2 O 2 levels to levels of lean controls. H 2 O 2 is considered to be the stable end product of short-lived reactive oxygen radicals and plasma H 2 O 2 levels are determined by the kinetics of H 2 O 2 production and elimination 16 . As urinary H 2 O 2 excretion did not differ between db/db vehicle and db/db Sul-121, this indicates that Sul-121 indeed inhibits ROS production in diabetes. These findings were supported by our in-vitro experiments using the ROS probe H 2 DCFDA. There, Sul-121 was able to inhibit high glucose and insulin induced increase in ROS production in cultured mesangial cells, indicating that Sul-121 is able to decrease ROS production in response to high glucose. Further, we demonstrated that Sul-121 was able to inhibit increased ROS production by mitochondria obtained from kidneys of diabetic db/db mice. These findings are in line with a previous study performed in adipose tissue-derived stem cells 15 , confirming the inhibitory effects of Sul compounds on mitochondrial ROS production. In addition to inhibiting ROS production, Sul-121 stimulated renal expression of superoxide dismutases SOD1 and SOD2 through a yet unknown mechanism, which may have contributed to the decreased ROS status in db/db Sul-121. Finally, Sul-121 attenuated the up-regulation of renal COX2 expression in diabetic animals. COX2 expression is known to be induced by mitochondrial ROS production, and may contribute to disturbed autoregulation of glomerular blood flow in diabetes 17 . Taken together, these data demonstrate that Sul compounds can indeed reduce ROS production in diabetes confirming our hypothesis.
Further, Sul-121 had profound effects on endothelial mediated relaxation. In our study, vehicle treated db/db mice demonstrated substantial endothelial dysfunction, as evidenced by the impaired vasorelaxation to acetylcholine. Sul-121 treatment however, completely normalized endothelial function by restoring vasorelaxation of db/db mice to the level of lean control mice. Furthermore, Sul-121 treatment restored the NO sensitivity of the smooth muscle layer in diabetes. Endothelial dysfunction in the db/db models has been reported in numerous studies and is thought to be caused by reduced nitric oxide (NO) bioavailability as a consequence of oxidative stress. In keeping with this, the restored endothelial function in Sul-121 treated mice was associated with a normalization of plasma H 2 O 2 levels in db/db mice. Furthermore, dissection of the endothelial factors involved in relaxation, demonstrated an increased contribution of NO and prostaglandins in the Sul-121 treated db/db mice. These data therefore indicate that normalization of oxidative stress and availability of NO underlie the normalization of endothelial function by Sul-121 in db/db mice.
Vascular endothelial dysfunction is often associated with renal endothelial dysfunction known to increase protein leakage over the filtration barrier, resulting in albuminuria 5 . Indeed, we found a significant correlation between vascular endothelial function and albuminuria in the present study. Furthermore, in addition to restoring vascular endothelial function, Sul-121 treatment halted the progression of albuminuria in db/db mice. Moreover, kidney hypertrophy, glomerular size, focal glomerular sclerosis scores and glomerular fibronectin expression were significantly lower in Sul-121 treated db/db over vehicle treated db/db mice, indicating a reduction in glomerular damage. Taken together these data indicate that the improved ROS status by Sul-121 treatment prevented vascular-and glomerular endothelial function, thereby attenuating the development of diabetic kidney damage.
In addition to damage to the endothelium, albuminuria may also be aggravated by loss of podocytes 18,19 and changes in the distribution and expression of slit-diaphragm proteins 20 . In the db/db model, loss of glucose sensitive podocytes is an early event and is already prominent at an age of 8 weeks, whereas the remaining majority of podocytes adapt to hyperglycemia and become resistant to diabetes-induced apoptosis 18 . Indeed, at the age of 10 weeks (at the start of treatment), the db/db mice in the present study already had increased albuminuria over lean mice, possibly reflecting this early loss of susceptible podocytes. Albuminuria in db/db mice progressed further at the age of 16 weeks (week 6 of treatment) which was prevented by Sul-121 treatment. However at the end of the study, podocyte numbers were found similarly reduced in both vehicle and Sul-121 treated db/db mice, indicating that the protective effects of Sul-121 treatment is not mediated through protection of the glucose resistant podocytes. Whether Sul-121 could protect the glucose sensitive podocytes at an earlier time point remains to be investigated in future studies.
Although the pharmacokinetics of Sul-121 are not fully known, the current study showed that Sul-121 plasma levels were significantly higher in db/db than in lean control mice. These differences cannot be simply explained by differences in body weight. As daily dosing was identical in lean control (low weight) and db/db mice (high weight), the opposite is expected, i.e. lower Sul-121 levels in db/db. Further, the administration of Sul-121 through osmotic mini pumps exclude that differences in oral absorption may underlie the higher Sul-121 levels in db/db. Moreover, db/db mice had a much higher urinary output, which seems to exclude renal clearance as the elimination route of Sul-121. Taken together, these data point towards a decreased hepatic clearance of Sul-121 in diabetes 21 explaining its increased plasma concentration in db/db mice. Thus, it would be interesting in future studies to explore the pharmacodynamics/kinetics of Sul-121 to better understand its metabolism and clearance. Finally, we cannot exclude that the lower plasma levels of Sul-121 are contributing to the lack of major adverse effects in the lean mice.
Although the current study was carefully planned and executed, there are some limitations to consider. First, body weights in the db/db mice began to fall by week 4 of the experiment which corresponds with an age of fourteen weeks. This is considerably earlier than the decline in body weight at the age 5 to 6 months reported before 22 . Moreover, plasma glucose levels at the start of the experiment (age 10 weeks) were already at 28 mM which is considerably higher that the 15 mM reported for this age 23 . In addition, renal interstitial damage and inflammation were lower than expected from literature. At present, it is unclear why the db/db mice developed such a severe diabetic phenotype with mainly glomerular damage and low interstitial damage. Possibly, the relative high percentage of sugars in the institution's standard RMH-B diet (4%) may be related to this. Regardless, this study demonstrates that Sul-121 treatment is effective even in a severely diabetic model.
Further, Sul-121 treatment caused a small (approximately 10%) but statistically significant increase in glomerular protein carbonylation in lean mice. This is surprising a no other adverse effects of Sul-121 was recorded in the present study. Furthermore, plasma H 2 O 2 levels and expression of SOD1,2,3 and COX2 were all unaffected in Sul-121 treated lean mice, providing no indications for increased oxidative stress in Sul-121 treated lean mice. In addition, our in-vitro experiments did not show an effect of Sul-121 on ROS production in control cells. Alternatively, Sul-121 might interfere with protein decarbonylation. However, the molecular pathways of protein decarbonylation are still incompletely understood and whether Sul-121 interferes with electron donation to Glutaredoxin-1 or other enzymes involved in protein decarbonlyation 24 needs to be determined in future studies. Nevertheless, the small increase in glomerular carbonylation in the absence of ROS activation appears to be without functional consequences for the kidney. In conclusion, Sul-121 limits the progression of experimental diabetic kidney damage via a mechanism that preserves endothelial function and reduces oxidative stress. Sul-121 may therefore represent a new therapeutic strategy to combat diseases characterized by endothelial dysfunction and oxidative stress, including diabetes, obesity and metabolic syndrome.

Methods
Animals. Experiments were approved by the Institutional Animal Care and Use Committee of the University of Groningen, Netherlands. Further, all methods were performed in accordance with the relevant guidelines and regulations. 16 male db/db (strain JAX000642) and 16 lean controls were purchased from Harlan, UK and were assigned into four groups: db/db vehicle (diabetic control), db/db Sul-121 (treated diabetic group), lean vehicle (non-diabetic control) and lean Sul-121 (to study effect of drug only). Osmotic mini pumps were implanted at the age of 10 weeks and replaced at 14 weeks. Mice were placed in metabolic cages for 24 hours every 2 weeks. Mice were housed conventionally with free access to standard pelleted food (RMH-B, Hope Farms, Woerden, Netherlands) and water. In the db/db vehicle group, one mouse became lethargic at week 1 and was prematurely terminated. In the db/db Sul-121 group, one mouse died during surgery at week 2 due to complications caused by the anaesthetic. Therefore, the number of animals in both db/db groups was reduced to n = 7.
An additional group of 5 db/db and 4 lean control mice were used for assessing renal mitochondrial ROS production.

Mini-pumps and dose of Sul-121. Previous unpublished experiments demonstrated that bolus infusions
up to 30 mg/kg Sul-121 could be administered safely to mice. Therefore the maximal attainable dose of Sul-121 was chosen in the current study. For this, Alzet 2004 pumps were loaded with 200 µl of 12.6 g/l Sul-121 in 50% DMSO (or with 200 µl 50% DMSO solution as vehicle), resulting in a daily dose Sul-121 of 2.2 mg/kg/day for a 35gram mouse (range 1.8-2.82 mg/kg/day for the heaviest and lightest mouse in the study). Pumps were surgically implanted into a subcutaneous pocket in the dorsal midscapular area of the mouse. Plasma measurements. Plasma creatinine levels were determined using a Cayman Creatinine (serum) Colorimetric Assay Kit (Cayman, Ann Arbor, USA). Plasma hydrogen peroxide levels were determined using an Amplex Red H 2 O 2 kit, Life technologies (A22188), Leusden, Netherlands.

Cell culture and ROS production. Mouse mesangial cells (MMC) were cultured in DMEM containing
penicillin and streptomycin supplemented with 10% fetal calf serum. ROS production was measured in 96 wells plates using the probe H2DCF-DA (C2938; Molecular probes). MMCs were first incubated in DMEM containing 1 g/l glucose 24 hours. Near confluent mesangial cells were loaded with dye for 1 hour at 37 °C. Afterwards, cells were washed and incubated with or without Sul-121 (100 nM) for 30 minutes. ROS production was stimulated by incubation in high glucose (4.5 g/l) and insulin (0.1 unit/ml) for 1 hour. Fluorescence was measured using a Synergy H4 microplate reader.
Renal mitochondrial ROS production was measured in an additional cohort of lean (n = 4) and db/db (n = 5) mice. For this, the kidney was excised and immediately stored in cold saline for no more than 2 hours. Subsequently, the kidney was cut into smaller pieces and a 100 mg kidney fragment was transferred into an eppendorf vial containing 0.5 ml of cold extraction buffer (10 mM HEPES pH 7.5, 200 mM mannitol, 70 mM sucrose and 1 mM EGTA). The tissue was homogenized and subsequently centrifuged at 600*g for 5 minutes. The supernatant was transferred to a new tube and centrifuged at 11.000 g for 10 minutes. The pellet was resuspended in 0.5 ml extraction buffer and the two centrifugation steps described above, were repeated. The resulting pellet was resuspended in storage buffer (10 mM HEPES pH 7.5, 250 mM sucrose, 1 mM ATP, 80 µM ADP, 5 mM sodium succinate, 2 mM K 2 HPO 4 and 1 mM DTT). Protein concentrations were determined by a DC protein assay (Biorad). For measuring mitochondrial ROS production, 2 µg of mitochondria was transferred into 50 µl assay buffer (20 mM MOPS pH7.5, 110 mM KCl, 10 mM ATP, 10 mM KCl, 10 mM sodium pyruvate and 1 mM EGTA) supplemented with 100 µM Amplex Red en 0.2U/ml HRP (Invitrogen). To evaluate the acute effects of SUL-121 on mitochondrial ROS production, 1 µM of Sul-121 was added to the reaction mixture. The Amplex Red reaction was performed at 37 °C for two hours. Fluorescent detection was performed at 540/590 nm in a Biotek Synergy 4 plate reader.
Data analysis and calculations. Data are presented as means ± SE, and n refers to the number of animals in each group. Statistical analysis was done with SPSS version 22 for Windows (SPSS, Chicago, IL). Differences between full concentration-response curves were tested with repetitive measures ANOVA; differences between points were tested with one-way ANOVA. P values of <0.05 (two tailed) were considered statistically significant.