Antithrombin nanoparticles improve kidney reperfusion and protect kidney function after ischemia-reperfusion injury

Abstract

In the extension phase of acute kidney injury, microvascular thrombosis, inflammation, vasoconstriction, and vascular endothelial cell dysfunction promote progressive damage to renal parenchyma after reperfusion. In this study, we hypothesized that direct targeting and pharmaceutical knockdown of activated thrombin at the sites of injury with a selective nanoparticle (NP)-based thrombin inhibitor, PPACK (phenylalanine-proline-arginine-chloromethylketone), would improve kidney reperfusion and protect renal function after transient warm ischemia in rodent models. Saline- or plain NP-treated animals were employed as controls. In vivo 19F magnetic resonance imaging revealed that kidney nonreperfusion was evident within 3 h after global kidney reperfusion at 34 ± 13% area in the saline group and 43 ± 12% area in the plain NP group and substantially reduced to 17 ± 4% (∼50% decrease, P < 0.05) in the PPACK NP pretreatment group. PPACK NP pretreatment prevented an increase in serum creatinine concentration within 24 h after ischemia-reperfusion, reflecting preserved renal function. Histologic analysis illustrated substantially reduced intrarenal thrombin accumulation within 24 h after reperfusion for PPACK NP-treated kidneys (0.11% ± 0.06%) compared with saline-treated kidneys (0.58 ± 0.37%). These results suggest a direct role for thrombin in the pathophysiology of AKI and a nanomedicine-based preventative strategy for improving kidney reperfusion after transient warm ischemia.

acute kidney injury (AKI) affects a significant portion of hospitalized patients in the U.S. and has an attendant mortality of up to 76.8% (48). The etiology of renal failure in the hospital setting is often multifactorial and includes sustained hypoperfusion from shock; nephrotoxic agents (e.g., intravenous iodinated iso-osmolar contrast); and intraoperative transient warm ischemia (45). Although significant advances have been made in understanding the complex pathologic processes involved in AKI, therapy is still largely supportive and no specific agents are approved that target well-defined molecular mechanisms involved in the process of AKI (21).

In this work, we focus on pathological vascular events that are thought to operate in the extension phase of AKI, which involves microvascular thrombosis and vascular endothelial cell damage that promote continued tubular damage (1, 31, 43). These vascular events lead to regional microvascular nonreperfusion that impairs the recovery and regeneration of oxygen-demanding tubular cells (41). While the initial global ischemic insult may be unavoidable, we hypothesize that targeted therapies to interdict microvascular coagulation in the extension phase might be designed to operate predominantly at sites of vascular injury and inflammation to maintain perfusion, ameliorate nonreperfusion injury, and hasten renal recovery.

Prior studies have suggested that antithrombin agents, such as heparin (16), antithrombin III (30, 37), and melagatran (15), could improve recovery of renal function after transient kidney ischemia-reperfusion, yet these measures have not been adopted as routine clinical therapy for AKI. To this end, we recently have reported a potent nanoparticle (NP)-based antithrombin therapy for highly localized and safe inhibition of clotting that attends vascular damage (33). In addition to preventing coagulation-based renal injury, these antithrombin particles may have a protective effect through inhibiting thrombin-mediated inflammation via proteinase-activated receptor-1 (PAR-1) activation (9). The base NP is comprised of a stable perfluorocarbon (PFC) core surrounded by a lipid-surfactant wrapper that can be readily functionalized with existing thrombin inhibitors such as bivalirudin (34) or PPACK (phenylalanine-proline-arginine-chloromethylketone). The complete nanosystem exerts a unique therapeutic action by forming a long-lived anticlotting surface after binding to clot-associated thrombin. Safety is ensured by rapid clearance of any unbound circulating particles such that clotting parameters and bleeding times rapidly normalize after injection (30–60 min) (33). Additionally, we have previously employed this nontargeted nanostructure to map regional tissue oxygenation after ischemic kidney injury with the use of fluorine (¹⁹F) magnetic resonance imaging (MRI) (20).

Accordingly, we now propose to test the hypothesis that direct inhibition of thrombin with the use of molecularly targeted NPs will accelerate functional recovery in a rodent model of warm ischemic injury. Additionally, we take advantage of the perfluorocarbon core of this agent to perform quantitative molecular imaging in vivo with fluorine MR to document homing to the kidney, thrombin binding, and the ultimate outcomes of nonreperfusion injury and perfusion recovery. Furthermore, we examine the ability of PPACK NPs to inhibit PAR-1 activation in an in vitro model. Success of the approach would provide support for the mechanistic role of thrombin as a key regulator of acute renal injury and as a potential pharmacological target for novel nanomedicine agents. Moreover, the ability to noninvasively define the extent of renal injury and monitor early recovery could provide additional insights relevant to patient management.

METHODS

All animals were acquired from Harlan Laboratories. Adult male C57BL/6 mice (n = 16) were used for in vivo MRI of AKI. Adult male Sprague-Dawley rats were used for longitudinal evaluation of kidney function up to 7 days after ischemic injury (n = 15) or immunohistological analysis of cell injury. All procedures conformed to the guidelines of and with the approval of the Animal Studies Committee of Washington University in St. Louis.

NP Formulation

Perfluorocarbon NPs (∼250 nm) were formulated with either perfluorooctylbromide (PFOB) or perfluoro-15-crown-5-ether (CE) as previously described (33). The PFOB emulsion was composed of 20% (vol/vol) of PFOB (Exfluor Research), 2.0% (wt/vol) of a surfactant commixture, and 1.7% (w/vol) glycerin, with water comprising the balance. The CE emulsion was composed of 40% (vol/vol) of CE (Exfluor Research), 2.0% (wt/vol) of a surfactant commixture, and 1.7% (wt/vol) glycerin, with water comprising the balance. PPACK was conjugated to the PFOB NP using previously reported methods that loaded ∼13,650 PPACK moieties per particle (33).

Renal Injury and Treatment

Ischemia-reperfusion kidney injury was carried out unilaterally in mice and bilaterally in rats. Animals underwent laparotomy to produce 45 min of warm ischemia. Briefly, animals were anesthetized with a dose of ketamine (85 mg/kg) and xylazine (13 mg/kg) cocktail. PPACK NP (1 ml/kg), plain NP, or saline was injected intravenously into the tail vein at 10 min before the surgery. A mid-line abdominal incision was then performed to expose the kidney. Kidney ischemia was induced by occlusive sutures around both the renal artery and vein in mice and the renal artery alone in rats. Successful cessation of renal blood flow was confirmed visually by change of the kidney color from pink to dark purple. Animal body temperature was maintained at 37°C using a small animal heating system. The suture was released after 45 min to restore kidney blood flow, which was confirmed by the change of kidney color to pink. The surgical wound was then closed in layers and the animal recovered and was returned to the cage.

Evaluation of Functional Recovery in Rats

The bilateral rat model of kidney ischemia-reperfusion injury was used for longitudinal evaluation of the effect of PPACK NPs on kidney function. Rats were maintained for 7 days after AKI. Blood samples (200 μl) were collected before AKI and daily after AKI for 7 days. Blood was centrifuged and serum creatinine concentration was analyzed using the LIASYS 330 clinical chemistry system (AMS Diagnostics,) in the Core Research Animal Diagnostic Laboratory, Washington University School of Medicine.

Immunohistochemistry.

Rats were euthanized 1 and 7 days after AKI. The left kidney was fixed in 10% formalin, embedded in paraffin, and sliced at a 4-mm thickness for routine histological staining [hematoxylin and eosin (H&E)]. The right kidney was snap frozen and sectioned at an 8-mm thickness for immunohistochemical staining. Cell apoptosis was evaluated with a terminal deoxynucleotidyl transferase dUTP-mediated nick-end labeling (TUNEL) kit (Ab66108; Abcam). Thrombin deposition was evaluated by staining with an antithrombin primary antibody (Ab92621; Abcam) followed by an Alexa Fluor 594 secondary antibody (ab150080; Abcam).

Data analysis.

The extent of thrombin accumulation in injured kidneys was evaluated by quantifying the percentage of tissue area exhibiting thrombin-positive signal (red: wavelength = 594 nm). For each kidney, ×200 fluorescent images were taken at 10 randomly selected positions in the medulla. The thrombin-positive signal in each image was determined with the use of ImageJ software (National Institutes of Health) by detecting those pixels with red signal intensity more than twofold above green signal intensity (to exclude autofluorescent signals from kidney tissue) as well as >5 SD over background signal intensity. The ratio of thrombin-positive pixels to the total number of pixels in each image (1,032 × 1376) was calculated.

The severity of tubular cell damage was evaluated by a board-certified pathologist with specific expertise in renal histopathogy (J. P. Gaut). Blinded microscopic examination of H&E-stained slides was performed to assess tubular damage including tubular necrosis, tubular dilation, proximal tubular brush border fragmentation, and mitotic activity. Pathologic features were defined quantitatively as the percentage of tubular necrosis (%acute tubular necrosis), i.e., the ratio of tubules exhibiting necrotic cells in injured kidney.

Molecular Imaging of Kidney Injury in Mice with ¹⁹F MRI

We have previously reported high-field MRI methods for imaging and quantifying kidney oxygenation and perfusion with the use of nontargeted PFC NPs (20). As before, all MRI experiments were carried out on an 11.7T Varian scanner. Imaging was performed with a custom-built actively decoupled coil pair. The volume transmit coil (5-cm inner diameter, 10-cm long) was built to generate a homogeneous B₁ field that covers both kidneys. A surface receive coil (1 × 2 cm²) was placed under the animal's body to achieve high sensitivity to ¹⁹F signal from the two kidneys. Both transmission and receiving coils were tuned and matched at ¹⁹F frequency. Despite being slightly off-resonance for ¹H imaging, this coil pair was sufficient to generate high quality anatomic images within <1-min scan time. Because the spectral signatures of PFOB and CE are different that are readily distinguishable by ¹⁹F MRI, they can serve as distinct and separable calibration standards or as perfusion agents as outlined below.

Ex vivo MRI.

In the first study, the severity of kidney nonreperfusion after transient ischemia was quantitatively evaluated by the amount of trapped, nonflowing NPs in injured kidneys. Specifically, mice were pretreated with PFOB-core PPACK NP or plain (nondrugged) NP followed by unilateral warm ischemia. Following reperfusion for 3 h, mice were euthanized and the NPs in the blood were washed out of the kidney by intracardiac puncture and systemic perfusion with saline for 5 min. Kidneys were then removed, stored in saline, and placed in a 1-cm solenoid MR coil. A capillary tube containing 5 μl CE-core plain NPs was placed next to the kidney to serve as a calibration standard for quantification of the ¹⁹F signal as previously reported (39).

In vivo MRI.

Intrarenal reperfusion was imaged in mice pretreated with PFOB-core PPACK NP (n = 4), plain NP (n = 6), and saline (n = 6). After reperfusion for 3 h, a single dose of CE-core NP (5 ml/kg) was injected intravenously as a ¹⁹F perfusion imaging agent, due to its different ¹⁹F spectral signature compared with the therapeutic PFOB-core NP. Mice were anesthetized with a (54 mg·kg⁻¹·h⁻¹) and xylazine (3.6 mg·kg⁻¹·h⁻¹) to achieve extended anesthesia. The animal's body temperature was maintained at 37 ± 1°C using a small animal heating system (SAI). Mice were placed in the supine position and ventilated with 100% oxygen.

A gradient echo sequence was employed for ¹H imaging to locate two parallel short-axis slices sectioned through the center of the left (injured) and right (uninjured) kidneys, respectively. A respiration-gated fast spin echo sequence was used to acquire the ¹⁹F spin-density images. Imaging parameters were as follows: echo train length = 4, echo time = 11.5 ms, voxel size = 0.4 × 0.4 × 2 mm³, and field of view = 26 × 26 mm².

Image analysis.

¹⁹F MRIs were analyzed to quantify underperfused area in injured kidneys. First, the total kidney area was measured on each ¹⁹F image slice. Because CE NPs are an intravascular agent that was administrated after ischemia-reperfusion injury, the detected ¹⁹F signal from CE NPs represented reperfused tissue in injured kidneys. A region of interest within the perfused cortex was then manually defined to estimate the mean of ¹⁹F signal, S₀, and its SD. Finally, the underperfused kidney area was determined as kidney tissue with ¹⁹F signal intensity <S₀-3 ± SD.

In Vitro Analysis of PAR-1 Inhibition

Cell culture.

Human umbilical vein endothelial cells (HUVECs) were purchased from Lifeline Cell Technology (Frederick, MD) and cultured in VascuLife VEGF Endothelial Cell Culture Medium (Lifeline Cell Technology) comprised of VascuLife Basal Medium supplemented with rhVEGF (5 ng/ml), rhEGF (5 ng/ml), rhFGF (5 ng/ml), rhIGF-1 (15 ng/ml), ascorbic acid (50 μg/ml), hydrocortisone hemisuccinate (1 μg/ml), heparin sulfate (0.75 U/ml), l-glutamine (10 mM), and fetal bovine serum (2%).

Flow cytometric determination of PAR-1 cleavage.

For assaying inhibition of thrombin activity against PAR-1, HUVECs were stimulated with 10 nM human thrombin (Haematologic Technologies, Essex Junction, VT) in thrombin stimulation medium (VEGF Endothelial Cell Culture Medium without heparin sulfate) supplemented with either PBS, 10 nM PPACK, 0.8 nM plain PFOB NPs, or 0.8 nM PPACK NPs for 2 h (n = 3 per group), where the ratio of thrombin to inhibitor in PPACK-treated groups was ∼1:1. Following 2 h of stimulation, HUVECs were washed three times with Ca²⁺ and Mg²⁺-free PBS followed by harvesting with Nonenzymatic Cell Dissociation Solution (Sigma-Aldrich, St. Louis, MO). Harvested HUVECs were washed with FACS Incubation Buffer (0.5% BSA in Ca²⁺ and Mg²⁺-free PBS) and stained with 20 μl of phycoerythin-labeled SPAN12 antibody (Beckman Coulter, Brea, CA) for 30 min. The SPAN12 antibody recognizes amino acid residues 35–46 of the PAR-1 receptor, spanning the thrombin cleavage site between residues 41 and 42. Upon cleavage of PAR-1 by thrombin, SPAN12 is no longer able to bind, thus resulting in detection of only uncleaved PAR-1 receptors (4, 46). Following staining, cells were washed in FACS Incubation Buffer three times, fixed in 1% paraformaldehyde, then resuspended in FACS Incubation Buffer for flow cytometry. Ten thousand events were acquired on a BD FACScan Analytic Flow Cytometer, and data were analyzed with FlowJo Collectors Edition.

Statistics

All statistical analyses were performed using SAS software (SAS Institute). Quantitative data are expressed as means ± SD. For comparisons between two experimental groups, the significance of the difference between the means was calculated.

The unpaired Student's t-test was used to test the difference of thrombin accumulation and tubular cell necrosis between saline- and PPACK NP-pretreated kidneys. The unpaired Student's t-test was utilized to test if thrombin stimulation of HUVECs would result in a decreased presence of uncleaved surface PAR-1 receptors. One-way ANOVA was used to test the difference of underperfused kidney area among PPACK NP-treated, plain NP-treated, and saline-treated animals. Two-way ANOVA was used to test 1) the difference between trapped PPACK NPs and plain NPs and between injured and noninjured contralateral kidneys; and 2) the difference of serum creatinine concentration among PPACK NP-treated, plain NP-treated, and saline-treated animals from baseline to 7 days after AKI. When overall significance of P < 0.05 was attained by ANOVA, comparisons between the means were performed using the Freeman-Tukey test.

RESULTS

Treatment with PPACK NP Mitigates Nonreperfusion Injury and Enhances Medullary Reperfusion in Mice

Quantitative ex vivo MRI was performed to evaluate the effect of PFOB-core PPACK NP on kidney nonreperfusion after AKI (unilateral injury) in mice. Briefly, animals were pretreated with PPACK or plain NP, underwent 45-min of warm ischemia, and were euthanized at 3-h postreperfusion. Circulating NPs in perfused vessels were washed out by saline flushing in situ to permit quantification of the amount of NPs trapped in any nonreperfused microvasculature. Based on the ¹⁹F MRI, we observed that the PFC NPs were trapped predominantly in the medulla of the injured kidney (Fig. 1, A and B), directly reflecting the extent of medullary nonreperfusion at this time point. Quantitative MR spectroscopic evaluation revealed that PPACK NP pretreatment reduced the amount of trapped NPs by approximately twofold (Fig. 1C) confirming improved medullary reperfusion in these injured kidneys. Note that in this ex vivo study, all detected NPs are those that have been trapped in nonperfused blood vessels and are therefore surrogates for the extent of thrombosis and associated microvascular clotting.

Fig. 1.

Ex vivo ¹⁹F MRI of noncirculating trapped nanoparticles (NPs) in mouse kidneys at 3 h after acute kidney injury (AKI). Animals were pretreated with phenylalanine-proline-arginine-chloromethylketone (PPACK) NP or plain NP before AKI. At 3 h after AKI, mice were euthanized and the circulating NPs in blood pool were washed out by systemic saline perfusion, leaving trapped NPs in nonreperfused kidney tissues. The overlaid ¹⁹F (color) and ¹H images (grayscale) showed that trapped NPs were located mostly in the medulla of the injured kidneys (A and B). Quantitative analysis using ¹⁹F MRI showed that the amount of trapped NPs was greater in injured kidneys than that in contralateral uninjured kidneys, reflecting regional nonreperfusion after AKI. In addition, the amount of trapped NPs was lower in PPACK NP-pretreated kidneys than in plain NP-treated kidneys (C), suggesting that PPACK NP improved kidney reperfusion after AKI. Data are presented as means ± SD; n = 3 in each group. *P < 0.05, compared with contralateral uninjured kidneys. †P < 0.05, compared with plain NP-treated injured kidneys.

Given these ex vivo observations, we performed in vivo ¹⁹F MRI to directly evaluate recovery of kidney perfusion in mice. Here, a CE-core nondrugged NP that exhibits a different MR signature as contrasted with the PFOB-core therapeutic PPACK NP was administrated intravenously as the ¹⁹F MRI contrast agent. The detected ¹⁹F signal therefore directly represents the circulating CE-core based NP in perfused vessels. Figure 2A indicates that the ¹⁹F signal was detected in both cortex and medulla of the uninjured right kidney, which confirms normal kidney perfusion. In contrast, a substantial reduction of the medullary ¹⁹F signal was observed in injured kidneys confirming reduced regional perfusion (Fig. 2, B–D). Finally, animals pretreated with PPACK NP exhibited improved perfusion (Fig. 2D) as contrasted with those treated with saline (Fig. 2B) or plain NP (Fig. 2C). Quantitatively, the underperfused area was 17 ± 4% in PPACK NP-pretreated kidneys, significantly lower than the 34 ± 13% and 43 ± 12% underperfused area in saline and plain NP-treated kidneys, respectively (Fig. 2E, P < 0.05). This in vivo readout of improved medullary perfusion in PPACK NP-treated kidneys thus accords with the observation of attenuated nonreperfusion injury at 3 h by ex vivo ¹⁹F MRI.

Fig. 2.

Improved medullary reperfusion in PPACK NP-pretreated mouse kidneys detected with in vivo MRI. Mice were pretreated with perfluorooctylbromide (PFOB)-based PPACK NP, plain NP, or saline before AKI. At 3 h after reperfusion, mice were intravenously administrated a single dose of perfluoro-15-crown-5-ether (CE)-based plain NP as the imaging agent for ¹⁹F MRI. The ¹⁹F signal from circulating CE NPs of uninjured kidney illustrates normal kidney perfusion (A). In contrast, all injured kidneys exhibited reduced ¹⁹F signal in the medulla (B–D) reflecting the severity of regional nonreperfusion. Compared with kidneys pretreated with saline and plain NP (B and C), the improved medullary perfusion in PPACK NP-treated kidneys is visually appreciable (D). With the use of in vivo ¹⁹F MRI data, the ratio of nonperfused kidney area after reperfusion was determined (E). The PPACK NP-treated kidneys (n = 4) developed less regional nonreperfusion than did saline (n = 6)- or plain NP (n = 6)-treated kidneys. *P < 0.05, compared with saline or plain NP-treated kidneys. Note that the arrow in D indicates the partial overlay of ¹⁹F signal from the adjacent spleen.

Pretreatment with PPACK NP Attenuated Kidney Dysfunction After AKI in Rats

Given that PPACK NP improved kidney reperfusion within a few hours after AKI, we sought to investigate the longer term consequences of PPACK NP treatment on the recovery of kidney function. A rat model of bilateral kidney injury was utilized because it closely mimics the common functional outcome of transient warm ischemia in humans (17, 27, 28). Figure 3 plots the serum creatinine level of rats before and up to 7 days after AKI. In all animals, the serum creatinine concentration increased significantly by 24 h after AKI, reflecting compromised kidney function. PPACK NP-treated animals exhibited lower serum creatinine concentration peak by 24 h (0.68 ± 0.16 mg/dl, P < 0.05 vs. saline or plain NP treated) than did those receiving saline (1.4 ± 0.45 mg/dl) or plain NP (0.92 ± 0.48 mg/dl), suggesting that PPACK NP treatment can hasten recovery of kidney function after AKI.

Fig. 3.

PPACK NP pretreatment accelerated the recovery of kidney function after AKI. Rats undergoing bilateral ischemia-reperfusion injury exhibited an acute increase of serum creatinine concentration that peaked between 1 and 3 days after AKI. Compared with saline and plain NP-pretreated rats, PPACK NP-pretreated rats exhibited lower serum creatinine concentration as early as 1 day after AKI. Data are presented as means ± SD; n = 5 in each group. *P < 0.05, compared with saline-pretreated animals. †P < 0.05, compared with plain NP-pretreated animals.

Immunohistochemistry Revealed Less Activated Thrombin in the Renal Tissue of PPACK NP-Treated Rats

To determine whether the beneficial effect of PPACK NP might be achieved through inhibiting progressive intrarenal thrombosis that ensues even after reperfusion, we performed antithrombin staining of activated thrombin in rat kidneys 24 h after AKI. By 24 h after AKI, thrombin was detected in all kidneys (Fig. 4, A–D). The percentage of thrombin-positive area was 0.58 ± 0.37% in saline-treated kidneys and 0.11 ± 0.06% in PPACK NP-treated kidneys (Fig. 4E, P < 0.05). Thus PPACK NP pretreatment resulted in an ∼80% reduction of thrombin accumulation after AKI, suggesting that circulating and clot localized PPACK NP effectively prevents progressive thrombin activation and retention.

Fig. 4.

Intrarenal thrombin accumulation (red color) at 24 h after AKI. Thrombin was detected in both cortex (A) and medulla (B) of saline-pretreated kidneys. In contrast, PPACK NP-pretreated kidneys exhibited less thrombin-positive staining in the cortex (C) than that in the medulla (D). Insets: zoom-in view of detected thrombin. The ratio of thrombin-positive areas confirmed that PPACK NP pretreatment reduced intrarenal thrombin accumulation at 24 h after AKI (E). *P < 0.05, compared with the saline-treated group. Microscopic images were acquired under ×200 power.

The effect of PPACK NP treatment on renal cell damage was evaluated by H&E staining (Fig. 5, A–D). Pathological features included the expected cell necrosis, tubular cyst formation, and cellular apoptosis in injured kidneys. Microscopic examination of rat kidneys subjected to ischemic injury showed marked coagulative type tubular necrosis, foci of tubular dilatation, fragmentation of the proximal tubular brush border, and mitotic figures. The changes were most pronounced at the corticomedullary junction. No significant interstitial inflammation was present nor were there any observable glomerular abnormalities in any of the groups. There was no significant difference in the degree of injury between the PPACK and saline-treated animals. In contrast, microscopic examination of the normal control kidneys showed intact tubules without evidence of necrosis or tubular injury. Thus H&E staining showed that there was no appreciable effect on acute kidney necrosis scores at 24 h following treatment with PPACK NP (Fig. 5E). At 7 days after AKI, H&E staining showed that the morphology of all kidneys were recovered, and Masson's Trichrome staining showed no significant increase of fibrosis in all injured kidneys. The kidney injury score of PPACK NP-treated animals reverted to baseline levels at 7 days (0 ± 0), while the saline-treated animals appeared to exhibit more variability and a higher score (3 ± 3, P = 0.59) compared with PPACK groups.

Fig. 5.

Histopathological assessment. Hematoxylin and eosin staining reveals necrotic tubular cells and tubular cysts in both saline (A and B)- and PPACK NP (C and D)-pretreated rat kidneys at 1 day after AKI. The percentage of necrotic tubular cells, i.e., acute tubular necrosis (ATN%), was comparable between saline and PPACK NP-pretreated kidneys (E; P = not significant). Microscopic images were acquired under ×200 power.

In addition, TUNEL staining revealed cellular apoptosis in both saline and PPACK NP-treated kidneys (Fig. 6). Thus the absence of appreciable effect of PPACK NP on these pathological features suggests a primary role for thrombin inhibition that improves kidney reperfusion and functional recovery irrespective of any immediate impact on apparent cellular necrosis.

Fig. 6.

Apoptosis detection. Terminal deoxynucleotidyl transferase dUTP-mediated nick-end labeling (TUNEL) staining revealed equivalent cell apoptosis in both saline (A and B)- and PPACK NP (C and D)-pretreated rat kidneys at 24 h after AKI. The average number of TUNEL-positive cells in each image was not statistically different between saline- and PPACK NP-treated kidneys (E). Microscopic images were acquired under ×100 power.

In Vitro Analysis Verified PAR-1 Inhibition by PPACK and PPACK NP

HUVECs were cultured and treated with thrombin for 2 h in the presence of saline, free PPACK, plain NPs, or PPACK NPs. Following treatment, cells were harvested and stained with a phycoerythin-conjugated SPAN12 antibody, which recognizes the thrombin cleavage sites on the PAR-1 receptor. Upon cleavage of PAR-1 by thrombin, SPAN12 can no longer bind, thus making it a cleavage-sensitive antibody. Figure 7 demonstrated that the presence of a direct thrombin inhibitor (either free PPACK or PPACK NP) results in preservation of surface PAR-1 receptors and prevents cleavage during exposure to thrombin. Conversely, without an inhibitor, thrombin action resulted in a substantial decrease of intact-PAR-1-positive cells (P < 0.05 for saline-treated vs. baseline PAR-1 and P < 0.05 for plain NP-treated vs. baseline PAR-1). These results indicate that in addition to its antithrombin effect PPACK NP may also provide an anti-inflammatory effect through inhibiting PAR-1.

Fig. 7.

PPACK NP inhibited the cleavage of surface proteinase-activated receptor-1 (PAR-1) on human umbilical vein endothelial cells. Both free PPACK and PPACK NP treatment effectively preserved surface PAR-1 receptors on HUVECs and prevented cleavage during the exposure to thrombin. *P < 0.05, compared with cells at baseline or treated with free PPACK or PPACK NPs.

DISCUSSION

The extension phase of AKI potentially offers an opportune therapeutic window for protection of renal function after AKI. In this study, we sought to elucidate a mechanistic role for thrombin in ischemic renal dysfunction by employing a thrombin-targeted NP to inhibit both the local production and function of a molecule known to be involved in both inflammatory signaling and thrombosis. The present observations indicate that thrombin indeed appears to be an important player in the pathophysiological scenario of AKI and that its early inhibition helps to maintain renal perfusion and function. Furthermore, nanostructures carrying antithrombin agents are promising therapeutic candidates for stemming local microvascular thrombosis and mitigating nonreperfusion injury. Finally, noninvasive molecular imaging methods in concert with therapeutic perfluorocarbon nanostructures to quantify regional perfusion, and potentially tissue oxygenation as previously reported (20), could represent new diagnostic and management tools in the setting of AKI.

Although traditional therapy for AKI has been primarily supportive, the present data suggest that nanomedicine approaches could offer an alternative therapeutic platform for clinical evaluation. Targeted nanomedicine allows high drug concentrations in the intended local environments while the total drug concentration and side effects are significantly reduced (22, 50). Compared with other nanomedicine platforms, PFC NPs have several unique advantages in treating AKI. PFC NPs are comprised of a liquid perfluorocarbon core encapsulated in a lipid monolayer with a nominal size of ∼250 nm (49). PFC NPs are cleared from the circulation by the reticuloendothelial system and the perfluorocarbon component is ultimately vaporized through the lung (6). They were initially introduced as an FDA-approved blood substitute and to date no nephrotoxic effect of PFC NPs has been reported in either humans or animals (7, 42). Additionally, PFC NPs are not dependent on renal function for clearance (13).

The in vivo ¹⁹F MRI data corroborate the salutary effect of PPACK NP for improving intrarenal perfusion after AKI. Figure 2 demonstrates regional nonreperfusion in the medulla of injured kidneys after transient warm ischemia-reperfusion. This in vivo observation agreed with prior morphological and histological studies demonstrating increased red blood cell trapping and decreased capillary perfusion (26, 35, 51). In contrast, cortical blood flow appears not to be affected and even may increase after AKI. Our results indicate that PPACK NP-treated kidneys exhibit less medullary nonreperfusion injury than do those treated with saline or plain NP in the acute phase.

The longer term benefit of PPACK NP on functional recovery after AKI was evidenced by a significantly attenuated increase in the serum creatinine level in PPACK NP-treated rats than that in saline or plain NP-treated controls (Fig. 3). This therapeutic effect was detected as early as 24 h after AKI and was sustained through 7 days after AKI. Although the creatinine level in the saline group and the plain NP group appeared to exhibit minor deviations between 1 and 3 days after AKI, the difference was not statistically different. The slightly delayed, statistically insignificant onset of renal injury may have been in part due oxygen delivery by the plain PFC NPs, which is a known ability of these particles.

Because we have shown previously that antithrombin NPs form a dense coating over a nascent clot by binding to any available clot-associated thrombin, which can then prevent further clotting (33), it is possible that thrombin generation locally is blocked after the initial bolus of PPACK NP binds to the emerging damaged hypercoagulable microvascular regions. We have calculated that there are ∼13,650 PPACK molecules per particle (33), which serves as a potent reservoir for immediately deactivating any new locally generated thrombin.

Alternatively, the pharmacokinetics of these nanostructures comprises a circulating elimination half-life of 3–5 h (14), which suggests that not only locally bound NPs might be removing newly formed thrombin but there exists some potential for residual circulating NPs to extract free thrombin as they circulate through the hypercoagulable territory and ultimately are cleared in the liver.

It has been acknowledged that kidney structural changes and functional outcomes are only loosely correlated (24, 36). After acute kidney failure, the degree of kidney dysfunction is frequently more severe than the detected kidney structural injury (19). In this study, H&E and TUNEL staining exhibited no discernable differences in apparent cell damage among injured kidneys. However, a dramatic decrease of thrombin staining was noted in PPACK NP-treated kidneys. It is perhaps not surprising that evidence of functional recovery would precede tissue structural recovery, especially after only one day. Additional extensive evaluation of the healing response would be required to ascertain whether the ultimate extent of necrosis and fibrosis differed in treated vs. untreated subjects, which would be an important topic for future study given that thrombin itself promotes classical wound healing and fibrosis (12, 18), and fibrosis is considered deleterious to kidney function (3, 23).

Extensive work over the past decades has revealed that thrombin, through PAR-1 activation, can exert of host of proinflammatory effects including recruitment of inflammatory cells, vasoconstriction, and endothelial cell dysfunction. The results of in vitro work with HUVECs demonstrated inhibition of PAR-1 activation in PPACK NP-treated cells (Fig. 7). Although much of our previous work has primarily focused on the action of antithrombin nanocarriers against thrombin in preventing coagulation (33, 34, 38), we also recognize the role of thrombin as a powerful signaling activator of vasoactive agents through cleavage and activation of PAR-1, and prior work by Kaplanski et al. demonstrated the use of the direct thrombin inhibitor hirudin to inhibit upregulation of vascular cell adhesion molecule 1 in endothelial cells in response to thrombin stimulation and activation of PAR-1 (21a). As such, we sought to explore the ability of PPACK NPs to inhibit cleavage and activation of PAR-1 in response to thrombin stimulation. The in vitro results suggest that a component of the renal protective effect of PPACK NPs may be through inhibition of PAR-1 cleavage and the associated downstream pathways (i.e., inflammatory effects of thrombin) (5, 9, 10, 29, 40).

This study has several limitations. It is well known that PPACK is a nonspecific protease inhibitor relative to its intended actions against thrombin (2, 11). For example, PPACK could also inhibit other coagulation factors such as FXIIa, FIXa, FXa, or activated protein C (25), which could offer an alternative mechanism for the observed benefit related to the clot-promoting action of thrombin. However, because we have demonstrated previously that both PPACK NPs and bivalirudin NPs exhibit essentially identical effects on focal thrombosis in other studies (33, 34), it seems likely that thrombin is mechanistically involved in this model. As a proof of concept study, this report does not necessarily intend to advance PPACK as a therapeutic candidate because more specific antithrombin agents can also be formulated as we have recently shown for bivalirudin (34).

The problem of inducing a prolonged bleeding diathesis with any anticoagulant is apparent. With respect to this antithrombin NP, we have already demonstrated that systemic anticoagulation is not required for its focal anticlotting activity, and in fact, all clotting parameters and bleeding times are normal within 30–60 min after delivery whereas local anticlotting activity persists for hours (33). This is due to a combination of the rapid distribution phase and subsequent sequestration in liver of unbound particles that reduces the effective circulating level of PPACK below that required to achieve systemic anticoagulation. Parenthetically, in this study we observed no significant increase in intraoperative bleeding among the saline-, plain NP-, and PPACK NP-treated animals.

Also, based on the present data, we cannot establish the extent to which the anticlotting actions or the signaling inhibition of thrombin account for the beneficial effects. Thrombin of course may contribute to activation of inflammation messengers such as NF-κB that serve to upregulate inflammatory gene programs (44, 47). While the current results validate our hypothesis that PPACK NP treatment enhanced kidney reperfusion after AKI and improved kidney function, follow up studies to quantitatively evaluate thrombin activation and the inflammatory response at different time points after AKI will provide a more complete understanding of the mechanism of the renal protective effect of PPACK NP.

This study was not designed to address the longer term effects of acute PPACK NP treatment on kidney function and tissue structural damage, which might modulate the development of fibrosis. Previous studies have shown that surviving epithelial cells will proliferate after the initial ischemic insult, resulting in the recovery of kidney morphology and blood chemistries in as early as 1 wk after AKI (3, 32), as our functional and structural data confirm (see Figs. 5 and 3, respectively).

Both the mouse and rat model of AKI were employed in this study. Specifically, the mouse model of unilateral AKI was selected for in vivo MRI due to the limited bore size of 11.7T MR scanner (8-cm inner diameter) that does not accommodate whole body imaging of rats. For longitudinal assessment of kidney function, the rat model of AKI was selected due to the technical difficulty of daily blood sampling in mice. In addition, the rat model more closely mimics the common clinical scenario of transient warm ischemia that occurs routinely during major cardiac and vascular surgery procedures (27, 28). There are some limitations posed by including two different models of renal injury in rats and mice. Longitudinal assessment of blood chemistry was not performed in mice because daily blood draws of 5–10% total blood volume (100–150 μl) for blood urea nitrogen analysis were not practical (34a). In rats, the effect of PPACK NP on improving renal reperfusion was not confirmed by in vivo MRI. Future MRI studies to correlate intrarenal reperfusion with kidney functional outcomes in the rat model of AKI could provide more complete evidence that antithrombin activity improves kidney reperfusion, but the current observations taken together suggest that this would be expected.

Other limitations are as follows: 1) it is known that a 45-min kidney ischemia typically causes severe injury in mice. The current procedures were designed as a severe model of injury to test the value of thrombin inhibition under more extreme conditions. Although a shorter ischemia time may be useful for evaluating milder AKI in mice, the current injury model successfully induced intrarenal nonperfusion that enabled clear assessment of the effect of antithrombin PPACK NP on improving kidney reperfusion; 2) Given the lack of significant differences in morphological changes between PPACK NP-treated and saline-treated animals (see Fig. 5), other kidney injury markers, such as kidney injury molecule 1, could provide additional mechanistic data regarding antithrombin therapy in AKI. For example, our results showed PPACK NPs effectively inhibits PAR-1 cleavage in an in vitro model, suggesting this PPACK NP could provide a renal protective effect by preventing thrombin mediated signaling and not simply thrombosis. 3) Finally, the therapeutic application of this agent as a pretreatment regimen represents only a first step in evaluating its potential clinical translation. However, the data do appear to validate thrombin as a promising specific molecular target for AKI. These observations also open a path for considering other ways of mitigating thrombin's role in AKI as a preventative measure.

In summary, we report a novel nanomedicine treatment to protect kidney function after AKI in rodent models of ischemic-reperfusion injury. In vivo ¹⁹F MRI showed that treatment with thrombin-inhibiting PPACK NP acutely improves kidney reperfusion within 3 h after AKI. Treatment with PPACK NP protected the kidneys from injury as evidenced by a significantly attenuated increase in creatinine level following injury. While future studies are needed to develop a more detailed mechanistic explanation regarding thrombin's exact role in AKI, these results suggest that thrombin inhibition might represent an effective therapeutic strategy if it can be achieved with a good safety margin.

GRANTS

This study was supported by National Institutes of Health Grants R01-HL-073646, R01-HL-112303, R01-AR-056223, and R21-DK-095555.

DISCLOSURES

Dr. S. A. Wickline reports relevant equity interest in AcuPlaq, LLC. All other authors declare no financial interests.

AUTHOR CONTRIBUTIONS

Author contributions: J.C., C.V., R.U.P., J.P.G., L.H., H.Z., and S.A.W. conception and design of research; J.C., C.V., R.U.P., M.G., G.C., and H.Z. performed experiments; J.C., C.V., J.P.G., M.G., and H.Z. analyzed data; J.C., C.V., R.U.P., J.P.G., M.G., L.H., G.C., H.Z., and S.A.W. interpreted results of experiments; J.C. and C.V. prepared figures; J.C., C.V., J.P.G., and S.A.W. drafted manuscript; J.C., C.V., R.U.P., J.P.G., M.G., L.H., G.C., H.Z., and S.A.W. edited and revised manuscript; J.C., C.V., R.U.P., J.P.G., M.G., L.H., G.C., H.Z., and S.A.W. approved final version of manuscript.