Pyridoxamine reduces postinjury fibrosis and improves functional recovery after acute kidney injury

Abstract

Acute kidney injury (AKI) is a common and independent risk factor for death and chronic kidney disease (CKD). Despite promising preclinical data, there is no evidence that antioxidants reduce the severity of injury, increase recovery, or prevent CKD in patients with AKI. Pyridoxamine (PM) is a structural analog of vitamin B₆ that interferes with oxidative macromolecular damage via a number of different mechanisms and is in a phase 3 clinical efficacy trial to delay CKD progression in patients with diabetic kidney disease. Because oxidative stress is implicated as one of the main drivers of renal injury after AKI, the ability of PM to interfere with multiple aspects of oxidative damage may be favorable for AKI treatment. In these studies we therefore evaluated PM treatment in a mouse model of AKI. Pretreatment with PM caused a dose-dependent reduction in acute tubular injury, long-term postinjury fibrosis, as well as improved functional recovery after ischemia-reperfusion AKI (IR-AKI). This was associated with a dose-dependent reduction in the oxidative stress marker isofuran-to-F2-isoprostane ratio, indicating that PM reduces renal oxidative damage post-AKI. PM also reduced postinjury fibrosis when administered 24 h after the initiating injury, but this was not associated with improvement in functional recovery after IR-AKI. This is the first report showing that treatment with PM reduces short- and long-term injury, fibrosis, and renal functional recovery after IR-AKI. These preclinical findings suggest that PM, which has a favorable clinical safety profile, holds therapeutic promise for AKI and, most importantly, for prevention of adverse long-term outcomes after AKI.

acute kidney injury (AKI) usually results from ischemic, obstructive, toxic, and infectious insults (3). Severe AKI requiring dialysis affects ∼90,000 patients in the United States annually, whereas milder forms of AKI affect >1.5 million individuals per year (3, 29, 46). The incidence of AKI is increasing, particularly in the critically ill patients, with an incidence of 65–67% reported in the intensive care unit, where mortality in dialysis-dependent AKI approaches 60% (46). Mild AKI is a precursor of chronic kidney disease (CKD), and there is an increased risk of end-stage renal disease (ESRD) in patients with severe AKI (8, 11, 12, 15, 62, 67). Oxidative stress has been implicated as one of the principal drivers of renal injury and pathophysiology in all of the common causes of AKI (4, 25, 34, 44, 45, 50). It occurs when the intrinsic antioxidant systems are unable to counteract the effects of excessive production of reactive oxygen species (ROS) and reactive carbonyl species (RCS), which can cause cellular damage via chemical modification to proteins, lipids, and nucleic acids (18, 49, 58). Furthermore, the steady-state levels of oxidative stress are increased in patients with increased risk of developing AKI, including patients with CKD, diabetes mellitus, and the elderly (2, 43, 64). However, despite promising results from preclinical studies using a variety of antioxidants in different experimental models of AKI (28), there is no definitive evidence that antioxidants reduce the severity of injury, increase the rate of functional recovery, or prevent postinjury CKD or ESRD in patients with AKI (65, 69). In part, this may be the result of inadequate dosing, since most of the antioxidants tested are ROS scavengers that have to outcompete all other reactive molecules to be effective (22), and few clinical studies have evaluated dosing effects of antioxidants on markers of oxidative stress.

Pyridoxamine (PM) is a structural analog of pyridoxal phosphate (vitamin B₆) that can interfere with oxidative stress via a number of different mechanisms (33, 59). These include: scavenging of RCS, including products of lipid peroxidation; detoxification of hypohalous acids; and sequestration of catalytic transition metal ions such as copper and iron, which are essential in the production of toxic ROS and RCS (58). The combination of these unique activities along with the clinical safety of PM underlie its therapeutic promise (30, 66). PM efficacy has been demonstrated in animal models of diabetic retinopathy, obesity, heart disease, atherosclerosis, Alzheimer's disease, and other multifactorial chronic conditions in which oxidative reactions and/or carbonyl compounds confer pathogenicity (9, 10, 13, 19, 23, 48, 57, 63, 68, 70). By efficiently scavenging reactive oxidative species and RCS, PM also inhibits the formation of advanced glycation end products (59), covalently modified proteins and lipid that play an important role in promoting end-organ damage in diabetes (42). Importantly, PM has shown promise in two phase 2 clinical trials in patients with diabetic kidney disease (30, 66), and its therapeutic use is currently being evaluated in a phase 3 clinical trial in patients with kidney disease (20). The ability of PM to interfere with multiple aspects of oxidative stress is favorable for the treatment of AKI. In particular, there is extensive clinical and preclinical evidence that free iron contributes to the pathogenesis of AKI associated with crush injury (rhabdomyolysis) and ischemia-reperfusion-induced AKI (IR-AKI) (35), as well as AKI induced by sepsis and cardiac surgery (4, 5, 27). This suggests that free iron scavenging by PM may be advantageous for the treatment of AKI in a variety of different clinical scenarios. We therefore hypothesized that PM therapy will allow for more effective AKI treatment compared with antioxidant treatments evaluated thus far in preclinical and clinical trials. For these studies, we evaluated the effects of PM treatment on both short-term injury and long-term renal fibrosis and functional recovery after severe IR-AKI in mice. Our data show that pretreatment with PM provides robust dose-dependent protection from IR-AKI-induced renal injury, reduction in a long-term postinjury renal fibrosis, and improvement of renal function after AKI. These findings suggest that PM treatment will be efficacious in the prevention of adverse renal outcomes in patients with AKI.

MATERIALS AND METHODS

IR-AKI in mice.

Surgeries were performed on a water bath-heated platform at 38°C on 10- to 12-wk-old male BALB/c mice. To induce IR-AKI, mice underwent left renal pedicle clamping for 31 min, and delayed contralateral nephrectomy was performed after 8 days. Functional recovery was assessed on days 9 and 28 after initial injury, as previously described (52). As shown in Fig. 1, we performed three sets of experiments with 99 mice entering the study, and only two mice died at the time of contralateral nephrectomy in PM treatment groups. All of the assays (histology, quantitative RT-PCR, and serum creatinine assays) were performed in all the remaining mice, with the exception of five mice in which serum samples were insufficient or missing at the time of the analyses. Serum creatinine was evaluated using an enzymatic cascade assay at day 9 after IR-AKI (requires ∼7 μl of serum; Pointe Scientific, Canton, MI). We also used a mass spectrometry-based assay for serum creatinine at day 28 after IR-AKI. This assay is more sensitive to detect differences in serum creatinines between 0.1 and 0.3 mg/dl (requires ∼10 μl of serum; performed by the University of Alabama at Birmingham O'Brien Center for Acute Kidney Injury) (32, 55). PM was administered to mice at estimated doses of 500 and 1,000 mg·kg⁻¹·day⁻¹ in the drinking water (2.5 and 5 g/ml drinking water based on average 25-g mice drinking 5 ml of water/day). Treatment was started 72 h before, or 24 h after, renal injury, as indicated, and continued until the experiments were terminated. In addition, PM was administered at 200 mg/kg two times a day by gavage for 72 h after each surgery (unilateral IR-AKI and contralateral nephrectomy) to ensure proper postoperative dosing. Vehicle control mice were treated with the same volumes of vehicle, water. Protocols were approved by the Vanderbilt Institutional Animal Care and Use Committee.

Fig. 1.

CONSORT flow diagram of preclinical studies. This diagram shows how the animals in this study were used. We performed three sets of experiments with 99 mice. Of these, a total of 3 mice died: 2 mice at the time of contralateral nephrectomy, both in the pyridoxamine (PM) treatment groups, and 1 mouse died after nephrectomy in one of the PM treatment groups. All of the assays, including histology, quantitative RT-PCR, and serum creatinine assays, were performed in all of the surviving mice, with the exception of serum creatinine analyses in 5 mice for which serum sample volumes were insufficient for the analyses. Vehicle and pretreatment with 1,000 mg·kg⁻¹·day⁻¹ PM from experiment 3 were included with data from experiment 2 for analysis of long-term dose dependence follow-up.

Histological analyses.

Kidneys were harvested, and 2- to 3-mm blocks were cut transversely through the cortex and medulla and fixed in 10% formalin. After kidney blocks were embedded and sectioned, acute and chronic renal tubular injury scores were determined on periodic acid-Schiff-stained sections. For this, all nonoverlapping fields in outer medulla were semiquantitatively evaluated separately by a blinded pathologist (×200 magnification) using the following scoring system: 0 = no injury; 1 = 1–20% of area injured; 2 = 21–50%; 3 = 51–75%; and 4 = 76–100%. Acute tubulointerstitial injury was defined as an interstitial edema with loss of brush border, shedding of both necrotic and viable epithelial cells in the tubular lumen, intratubular cast formation, tubular dilation, or naked tubular basement membrane. Chronic tubulointerstitial injury was defined as a matrix-rich expansion of the interstitium with sloughing of tubular epithelial cells, tubular cast formation, tubular dilatation, tubular atrophy, or thickening of tubular basement membrane. Renal fibrosis was determined in Sirius red (SR)-stained sections, as previously described (16). Results are expressed as means ± SE of tubular injury scores (0–4) or as the percent SR staining areas (SR positive/total surface area) as a fold change vs. uninjured controls.

RNA isolation and quantitative RT-PCR.

RNA was isolated from snap-frozen kidneys, and cDNA synthesis was performed, as described previously (16). RNA quantification and integrity were determined using a NanoDrop 2000c instrument (Thermo Fisher Scientific, Waltham, MA). cDNA was amplified using previously described primers (14, 41) and labeled using SYBR Green Supermix PCR (Bio-Rad, Hercules, CA). Gene expression is expressed as relative gene expression calculated using the 2^−ddC_T method, as described (54). Gapdh mRNA was used as a loading control, since we see no changes in Gapdh mRNA expression in the kidney after injury in the IR-AKI model (16).

Western blots.

Western blots were performed on lysates from whole kidneys using the following primary antibodies: rat monoclonal antihuman and mouse collagen a2 NCI domain antibody (catalog no. 7071; Chondrex) and mouse monoclonal anti-b-actin (catalog no. A5316; Sigma). Appropriate species-specific peroxidase-conjugated secondary antibodies (Jackson Laboratories and KPL) were used. Specific bands were quantified by densitometry and normalized to β-actin loading controls. For this we imported gray-scale scans of Western blots into LI-COR Image Studio Lite and then selected each band with a uniform rectangle shape and used the program to measure signal strength for each sample.

Isoprostane and isofuran measurement.

F2-isoprostanes and isofurans were measured in the Vanderbilt University Eicosanoid Core Laboratory after lipid extraction from snap-frozen kidney samples using gas chromatography-mass spectrometry, as previously described (38).

Pyridoxamine plasma concentration measurement.

A method was developed for quantitation of PM in mouse plasma over the range of 5-5,000 ng/ml using UPLC-mass spectrometry (MS)/MS with an automated PM extraction technique. PM was extracted at 4°C from an aliquot (0.010 ml) of mouse lithium heparinized plasma supplemented with 0.05 ml of the labeled internal standard (pyridoxamine-d₃ in water-methanol 50:50) using an automated protein precipitation method with 0.6 ml of methanol. The extract was centrifuged, and 0.15 ml of the supernatant was transferred using an automated liquid-handling system (Janus MultiPROBE II EX HT; Perkin-Elmer, Woodbridge, Canada) in a 96-well microplate and evaporated to dryness at 60°C for 15 min. For derivatization, 0.1 ml of 5 mM ammonium formate in water, pH 8, and 0.02 ml of propionic anhydride were added to the dried material, and the microplate was incubated for 30 min at room temperature. At the end of incubation, 0.1 ml of 1 mM ammonium formate in 0.5% aqueous solution of formic acid was added before column injection.

The extracted samples were kept at 4°C, and an aliquot (5 μl) was injected in a liquid chromatograph equipped with an ACE Excel 2 C18-PFP, 50 × 3 mm, 3 μm analytical column (Life Science, Peterborough, Canada). The isocratic chromatographic separation was performed at room temperature at a flow rate of 0.55 ml/min using a mobile phase of water/methanol (80:20) with 1 mM ammonium formate and 0.5% formic acid. Retention times were 2.60 and 2.58 min for PM and the labeled internal standard, respectively. The analysis was performed in positive ionization mode on an API 4000 LCMSMS (AB Sciex, Toronto, Canada) equipped with an electrospray. The monitored mass transitions were 225.1 → 134.2 mass-to-charge ratio for PM and 228.1 → 137.2 for its internal standard. Data analysis was performed using Analyst 1.6.1 software (AB Sciex). The calibration curves were constructed using the PM-to-internal standard peak area ratio and a weighted (1/x²) least-squares linear regression.

Before mouse samples analysis, the matrix selectivity, the recovery of the analyte, and the internal standard, the intrabatch accuracy, and precision with a coefficient of variation of 3.01% at the lower limit of quantitation, the dilution integrity, the hemolysis effect, the stability in the matrix at room temperature and at 4°C, the stability in the extracted samples and over the two freeze-thaw cycles in the matrix at −80°C were tested as a part of the method qualification procedure. The interbatch coefficients of variation were between 1.53 and 3.26%. All of the tested parameters were within the industry standards.

Statistical analysis.

Statistical analyses were performed by one-way ANOVA for multiple between-group comparisons using Tukey's correction for post hoc pairwise between-group comparisons. The minimal level of significance was set at P ≤ 0.05, and statistical analyses were performed using GraphPad Prism 6.0.

RESULTS

PM reduces acute tubular injury and oxidative stress in a dose-dependent manner after IR-AKI.

For initial studies, we evaluated dose-dependent effects of PM pretreatment on renal injury after severe unilateral IR-AKI (Fig. 2A). PM is a water-soluble orally bioavailable small molecule that has been extensively studied in rat (1, 19) and mouse models of diabetes (24, 40, 56), at estimated doses of 200–400 mg·kg⁻¹·day⁻¹ (∼1–2 g/l in drinking water). To optimize the PM treatment effects in the context of AKI, PM was administered at estimated doses of 500 and 1,000 mg·kg⁻¹·day⁻¹. These doses were chosen for these discovery studies given the higher levels of pathological oxidative stress after AKI vs. those described in models of diabetic nephropathy, and based on earlier toxicity studies showing that rats treated short term with up to 1,000 mg/kg PM orally showed no neurologic toxicity using the Irwin test, and rats treated with up to 500 mg·kg⁻¹·day⁻¹ daily for 6 mo showed no evidence of organ toxicities [Supplemental Materials S1 and S2 (Supplemental material for this article is available online at the journal website)]. Levels of lipid peroxidation products, F2-isoprostanes and isofurans, were evaluated in the injured kidneys 3 days after IR-AKI. Measurement of the ratio of isofuran/F2-isoprostanes in tissues and body fluids is considered the gold standard for analysis of oxidative stress in vivo (21, 37). There was a significant increase in renal isofuran-to-isoprostane ratios in the vehicle-treated IR-AKI mice compared with uninjured controls (Fig. 2B). Pretreatment with PM induced a dose-dependent reduction in renal F2-isofuran-to-isoprostane ratios in the injured kidney so that the values in mice treated with 1,000 mg·kg⁻¹·day⁻¹ PM were indistinguishable from those in uninjured controls (Fig. 2B). To determine whether dose-dependent effects of PM were associated with increased levels of PM in mice, we developed a method to quantify PM in plasma samples using UPLC-MS/MS with an automated PM extraction technique (see materials and methods). PM treatment effect corresponded to a dose-dependent increase in plasma PM (Fig. 2C). To assess the effect of PM treatment on renal injury, we first evaluated renal expression of the acute kidney injury markers N-Gal and Kim1 (26, 39). As anticipated, there was a marked increase in renal N-Gal and Kim1 mRNA expression in the vehicle-treated IR-AKI group (Fig. 2, D and E). PM pretreatment induced a dose-dependent decrease in renal N-Gal but not Kim1 mRNA (Fig. 2, D and E). The PM treatment effect was confirmed by histological evaluation of renal tubular injury. Consistent with the N-Gal mRNA results, there was a significant amelioration of tubular injury at the PM dose of 1,000 mg·kg⁻¹·day⁻¹ (Fig. 2, F and G).

Fig. 2.

Dose-dependent effects of PM pretreatment on renal oxidative stress and early injury after ischemia-reperfusion acute kidney injury (IR-AKI). A: experimental design. Mice were either left uninjured or underwent unilateral renal pedicle clamping to induce AKI. The AKI mice were pretreated for 3 days with either vehicle or 500 or 1,000 mg·kg⁻¹·day⁻¹ PM in drinking water. Treatment was continued until the end of the experiment. Mice were killed, and kidneys were harvested for analysis 3 days after the initial injury. B: effect of PM treatment on renal oxidative stress. Isofuran-to-F2-isoprostane ratios were determined in renal tissues. C: plasma PM levels 3 days after injury. D and E: effect of PM pretreatment on renal injury markers. Renal Kim1 and N-Gal mRNA expression levels normalized to Gapdh mRNA. Results are expressed as fold change relative to uninjured controls. F and G: effect of PM pretreatment on acute renal tubular injury. F: representative images of the outer medulla (OM) from periodic acid-Schiff (PAS)-stained kidneys (scale bars, 50 μm). G: tubular injury scores in the OM (0–4, arbitrary units). Results are expressed as means ± SE, n = 10 mice/group; 3 uninjured controls. Data were analyzed using 1-way ANOVA with post hoc Tukey's correction for multiple pairwise comparisons. Results were only considered significant by ANOVA where P < 0.05. Post hoc analysis in these experiments is indicated above the error bars: *P < 0.05, **P < 0.01, ***P < 0.001, and #P < 0.0001 vs. uninjured controls. The differences indicated with brackets are PM-treated AKI vs. vehicle-treated AKI.

PM reduces postinjury renal fibrosis and improves long-term functional recovery after IR-AKI.

To assess the impact of PM pretreatment on more clinically relevant post-AKI endpoints, we used a model of IR-AKI in which prolonged renal pedicle clamping is followed by a delayed contralateral nephrectomy (Fig. 3A) (16, 52, 53). This model allows for analysis of PM effects on long-term functional recovery and renal fibrosis after severe IR-AKI. As seen in short-term treatment studies, there was a dose-dependent increase in plasma PM 28 days after IR-AKI (Fig. 3B). At 28 days after IR-AKI, there was decreased expression of renal fibrosis markers collagen1-α1 (Col1-α1), collagen3-α1 (Col3-α1), and α-smooth muscle actin (α-SMA) mRNAs in mice pretreated with 500 and 1,000 mg·kg⁻¹·day⁻¹ PM compared with vehicle-treated controls (Fig. 3, C–E). These effects were associated with reduced α-SMA protein expression in mice with IR-AKI treated with 1,000 mg·kg⁻¹·day⁻¹ PM by Western blot (Fig. 3, F and G). There was no significant reduction in collagen expression in PM-treated mouse kidneys (Fig. 3, F and H), but there was decreased collagen deposition, as determined by SR staining of renal tissues (Fig. 3, I and J), as well as reduced chronic renal tubular injury scores at 1,000 mg·kg⁻¹·day⁻¹ PM (Fig. 3, K and L).

Fig. 3.

Dose-dependent effects of PM pretreatment on long-term renal fibrosis after IR-AKI. A: experimental design. Mice were either left uninjured or underwent unilateral renal pedicle clamping, followed by contralateral nephrectomy 8 days after the initial surgery. The AKI mice were pretreated for 3 days with vehicle or 500 or 1,000 mg·kg⁻¹·day⁻¹ PM in drinking water. The treatment was continued until the end of the experiment. Mice were killed, and kidneys were harvested for analysis 28 days after the initial injury. B: plasma PM levels 28 days after the initial injury. C–E: effect of PM pretreatment on renal expression of fibrosis markers: collagen1-α1 (Col1-α1, C), collagen3-α1 (Col3-α1, D), and α-smooth muscle actin (α-SMA, E) mRNA levels were normalized to Gapdh mRNA. Results are expressed as fold change relative to uninjured controls. F: representative Western blot for collagen IV α₂ chain (Col IV), α-SMA, and β-actin loading control in whole kidney lysates 28 days after injury from uninjured mice, mice after IR-AKI treated with vehicle, and mice with IR-AKI treated with 1,000 mg·kg⁻¹·day⁻¹ PM in drinking water. G and H: quantification of α-SMA (G) and Col IV (H) Western blots normalized to β-actin. Results are expressed as mean ± SE fold changes compared with uninjured controls: 4 uninjured controls, 10 IR-AKI treated with vehicle; 7 IR-AKI treated with PM. I: representative images showing collagen-specific polarized light birefringence of Sirius red-stained tissues [outer medulla (OM); scale bars 50 μm]. J: quantification of Sirius red staining expressed as fold change relative to uninjured controls. K: representative images of PAS-stained kidneys (OM, scale bars 100 μM). L: chronic tubular injury scores in the OM (0–4, arbitrary units). Results are expressed as means ± SE, 8 uninjured controls, 20 vehicle treated; 10 PM treated at 500 mg·kg⁻¹·day⁻¹ and 10 PM-treated at 1,000 mg·kg⁻¹·day⁻¹. Data were analyzed using 1-way ANOVA with post hoc Tukey's correction for multiple pairwise comparisons. Results were only considered significant by ANOVA where P < 0.05. Post hoc analysis in these experiments is indicated above the error bars: *P < 0.05, **P < 0.01, ***P < 0.001, and #P < 0.0001 vs. uninjured controls. The differences indicated with the brackets are PM-treated IR-AKI vs. vehicle-treated IR-AKI.

To assess a long-term functional recovery from AKI, we evaluated serum creatinine at 9 days (i.e., 24 h after the contralateral nephrectomy) and at 28 days after IR-AKI (see Fig. 3A). As anticipated, there was a significant increase in serum creatinine in the vehicle-treated IR-AKI mice at 9 and 28 days after injury (Fig. 4, A and B). PM pretreatment had no significant effect on serum creatinine at 9 days after injury (Fig. 4A), suggesting that, despite improvement in tubular injury at early time points, PM does not improve early functional recovery after IR-AKI. To assess the long-term effects of PM on functional recovery after IR-AKI, we used a MS-based measurement that detects serum creatinine values in the lower end of the normal range described for mice (32, 55). As previously reported (32, 55), baseline levels of serum creatinine in uninjured mice were significantly lower when measured by MS vs. enzymatic cascade (EC) assay that we used day 9 after IR-AKI (MS: n = 8, mean 0.09 ± 0.009 mg/dl; EC: n = 26, mean 0.31 ± 0.015 mg/dl, t-test, P < 0.0001). Using MS-based measurement of serum creatinine, we saw a reduction in serum creatinine at 28 days after IR-AKI in mice treated with 1,000 mg·kg⁻¹·day⁻¹ PM compared with vehicle-treated controls (Fig. 4B). Taken together, these results indicate that PM pretreatment ameliorates renal tubular injury and long-term postinjury fibrosis and improves long-term functional recovery after severe IR-AKI.

Fig. 4.

Dose-dependent effects of PM pretreatment on functional recovery after IR-AKI. The experimental design was the same as in Fig. 3. Mice were pretreated for 3 days with either vehicle or PM at 500 or 1,000 mg·kg⁻¹·day⁻¹ in drinking water, which was continued until the end of the experiment. A: serum creatinine 9 days after the initial injury. Serum creatinine was measured using an enzymatic cascade assay that only detects changes in serum creatinine ≥0.3 mg/dl. B: serum creatinine 28 days after injury. Serum creatinine was measured using a mass spectrometry-based approach, which is a more sensitive assay able to detect changes in serum creatinine of ≥0.1 mg/dl. Individual data points and means for each group are shown. Data were analyzed using 1-way ANOVA with post hoc Tukey's correction for multiple pairwise comparisons. Results were only considered significant by ANOVA where P < 0.05. Post hoc analysis in these experiments is indicated above the data points: *P < 0.05, **P < 0.01, and #P < 0.0001 vs. uninjured controls. The differences indicated with the brackets are PM-treated AKI vs. vehicle-treated AKI.

Delayed treatment with PM improves renal fibrosis but not functional recovery after IR-AKI.

Having established that pretreatment with PM has optimal antioxidant effects and efficacy after IR-AKI at 1,000 mg·kg⁻¹·day⁻¹ PM, we wanted to determine whether delaying treatment with PM for 24 h after the initiating injury was also effective at the same dose after severe IR-AKI. With the use of the same contralateral nephrectomy model of IR-AKI shown in Fig. 3A, PM treatment at 1,000 mg·kg⁻¹·day⁻¹ was initiated 24 h after IR injury. Delayed treatment with PM decreased expression of some fibrotic markers (Col3-α1 and α-SMA but not Col1-α1 mRNAs) 28 days after IR-AKI (Fig. 5, A–C). This was associated with a reduction in renal collagen deposition (Fig. 5, D and E) but not in chronic renal tubular injury scores at the same time point (Fig. 5, F and G). In addition, delayed treatment with PM had no effect on early (day 9) or delayed (day 28) functional recovery after severe IR-AKI (Fig. 5, H and I). These data indicate that, while delayed treatment with PM is not as effective compared with the pretreatment regimen, it may have some beneficial effects in AKI, particularly on long-term renal fibrosis.

Fig. 5.

Effects of delayed treatment with PM on long-term renal fibrosis and functional recovery after IR-AKI. The experimental design was the same as in Fig. 3, except mice were treated with PM at 1,000 mg·kg⁻¹·day⁻¹ starting 24 h after the initial injury. A–C: effect of PM treatment on renal expression of fibrosis markers: Col1-α1 (A), Col3-α1 (B), and α-SMA (C) mRNA levels normalized to Gapdh mRNA. Results are expressed as fold change relative to uninjured controls. D: representative images showing collagen-specific polarized light birefringence of Sirius red-stained tissues (OM; scale bars 50 μm). E: quantification of Sirius red staining expressed as fold change relative to uninjured controls. F: representative images of PAS-stained kidneys (OM, scale bars 100 μM). G: chronic tubular injury scores in the OM (0–4, arbitrary units). Effect of PM treatment on serum creatinine at day 9 (H) and day 28 (I). Results are expressed as means ± SE, 8 uninjured controls, 10 vehicle- and 10 PM-treated mice. Individual data points and means for each group are shown in H and I. Data were analyzed using 1-way ANOVA with post hoc Tukey's correction for multiple pairwise comparisons. Results only indicated if ANOVA P < 0.05. Post hoc analysis in these experiments is indicated above the error bars or data points: *P < 0.05, **P < 0.01, and #P < 0.0001 vs. uninjured controls. The differences indicated with the brackets are PM-treated AKI vs. vehicle-treated AKI.

DISCUSSION

These studies provide the first evidence that PM therapy can be effective in AKI. In the IR-AKI mouse model, treatment with PM reduced the severity of renal injury, decreased postinjury fibrosis, and improved long-term functional recovery after AKI. Treatment with the drug before injury was more effective than when treatment was delayed 24 h after injury, and the effects of this pretreatment regimen on renal injury, postinjury fibrosis, and functional recovery were clearly dose dependent. Mechanism of PM action in AKI also appears to involve amelioration of renal oxidative stress, as indicated by a dose-dependent reduction in the ratio of renal isofuran/F2-isoprostane lipid peroxidation products, the key marker of tissue oxidative stress (21, 37). This mechanism of action is consistent with previous structural and functional studies of PM in vitro and in vivo (33, 59).

Other approaches have been used in the past to reduce renal oxidative stress after AKI, but none have translated into effective therapeutics. ROS scavengers, such as N-acetylcysteine (NAC) and vitamin E, have been evaluated extensively in both preclinical and clinical studies in AKI (28, 31). Results have been variable, but there has been no definitive evidence that antioxidants accelerate recovery or prevent postinjury CKD in patients with AKI (65, 69). In part, this may be the result of inadequate dosing, since most of the antioxidants tested are ROS scavengers that have to outcompete other reactive molecules to be effective (22). In this respect, it is of note that few clinical studies have evaluated antioxidant effects of these interventions, and, where this has been looked at, for example, with the use of vitamin E as an antioxidant in patients with hypercholesterolemia, it has been shown that the doses used for many of these efficacy studies have little effect on oxidative stress in vivo (47). Unlike other antioxidants, PM can scavenge different toxic RCS, including lipid peroxidation products (59, 61), which contribute to tissue injury under conditions of oxidative stress (37, 49). In addition, PM can sequester redox metal ions, including free iron, and inhibit their ability to catalyze multiple toxic oxidative pathways (35, 58, 60). Because PM therapy targets multiple oxidative reactions in vivo, in particular redox-active metal ions that amplify pathological stress responses, the beneficial effects of PM may occur at lower effective drug concentrations compared with those of antioxidants, like vitamin E and NAC. For this reason we believe that the beneficial effects of PM in ameliorating renal injury in a preclinical model of AKI are more likely to successfully translate into clinical practice than prior studies using antioxidants.

The ability of PM to sequester redox active metal ions may be of particular importance in the context of AKI since increased levels of free circulating iron have been shown to increase oxidative stress and promote AKI associated with rhabdomyolysis, sepsis, and cardiac surgery (4, 5, 7, 27). It has also been shown that AKI increases intracellular levels of free catalytic iron, which increases oxidative stress and renal injury after IR-AKI (35). Moreover, acetaminophen, which inhibits iron-catalyzed lipid peroxidation (7), attenuates serum and/or urinary isofuran/F2-isporsotane levels in patients with sepsis (27), and in patients undergoing cardiopulmonary bypass surgery (5, 51). This indicates that therapeutic strategies designed to reduce the levels and/or activity of free catalytic iron can also be effective in patients with AKI. Thus, our findings are consistent with the hypothesis that inhibition of oxidative pathways by PM results from sequestration of redox-active metal ions by PM after IR-AKI.

In the present study, we used a model of IR-AKI developed in our laboratory that gives rise to reproducible postinjury renal fibrosis with low mortality and allows for the functional evaluation of renal recovery after severe injury (52). Using this approach, we demonstrated that pretreatment with PM at 1,000 mg·kg⁻¹·day⁻¹ not only reduces the severity of acute injury but also enhances long-term functional recovery. This finding has important clinical implications, since there is increasing demand for prospective treatments in AKI to show improvements in clinically meaningful endpoints such as long-term CKD or dialysis, rather than simply improvements in early functional parameters of the kidney (6). In addition, our findings suggest that therapeutic intervention with PM is most likely to be effective when given prophylactically in clinical situations in which the timing of injury is known, such as cardiac surgery, which is modeled by IR-AKI in this study. Importantly, PM showed a favorable safety profile in preclinical and clinical trials at doses and plasma levels that are effective in AKI therapy (30, 66). Neurotoxic effects have been reported in patients taking very high doses of the PM structural analog pyridoxine, or vitamin B₆ (17). Pyridoxine lacks a 4-aminomethyl group, which plays a key role in the PM mechanism of action (59–61), and conversion of pyridoxamine to pyridoxine in tissues is negligible (36). These toxic effects have not been found for PM, even at high doses in rodents. While further studies will be required to determine whether PM has similar effects in other models of AKI such as contrast and chemotherapy-induced AKI, the present study provides proof of principal that PM treatment is effective in reducing both short- and long-term adverse renal sequelae of AKI.

GRANTS

These studies were supported by a Research Grant from NephroGenex, Inc., and National Institute of Diabetes and Digestive and Kidney Diseases Grant DK-65138 (to P. Voziyan).

DISCLOSURES

P. Voziyan, B. Hudson, R. C. Harris, and M. P. de Caestecker performed consultancy work for NephroGenex, a company that develops PM as prospective drug for treatment of AKI.

AUTHOR CONTRIBUTIONS

N.I.S., P.V., B.H., R.C.H., and M.P.d.C. conception and design of research; N.I.S., C.R.d.C., M.-C.T., and M.D. performed experiments; N.I.S., H.Y., and C.R.d.C. analyzed data; N.I.S. and M.P.d.C. interpreted results of experiments; N.I.S. and M.P.d.C. prepared figures; N.I.S. drafted manuscript; N.I.S., P.V., C.R.d.C., M.-C.T., M.D., B.H., R.C.H., and M.P.d.C. approved final version of manuscript; P.V. and M.P.d.C. edited and revised manuscript.