Impaired hemodynamic renal reserve response following recovery from established acute kidney injury and improvement by hydrodynamic isotonic fluid delivery

Abstract

Renal reserve capacity may be compromised following recovery from acute kidney injury (AKI) and could be used to identify impaired renal function in the face of restored glomerular filtration rate (GFR) or plasma creatinine. To investigate the loss of hemodynamic renal reserve responses following recovery in a model of AKI, rats were subjected to left unilateral renal ischemia-reperfusion (I/R) injury and contralateral nephrectomy and allowed to recover for 5 wk. Some rats were treated 24 h post-I/R by hydrodynamic isotonic fluid delivery (AKI-HIFD) of saline through the renal vein, previously shown to improve recovery and inflammation relative to control rats that received saline through the vena cava (AKI-VC). At 5 wk after surgery, plasma creatinine and GFR recovered to levels observed in uninephrectomized sham controls. Baseline renal blood flow (RBF) was not different between AKI or sham groups, but infusion of l-arginine (7.5 mg/kg/min) significantly increased RBF in sham controls, whereas the RBF response to l-arginine was significantly reduced in AKI-VC rats relative to sham rats (22.6 ± 2.2% vs. 13.8 ± 1.8%, P < 0.05). RBF responses were partially protected in AKI-HIFD rats relative to AKI-VC rats (17.0 ± 2.2%) and were not significantly different from sham rats. Capillary rarefaction observed in AKI-VC rats was significantly protected in AKI-HIFD rats. There was also a significant increase in T helper 17 cell infiltration and interstitial fibrosis in AKI-VC rats versus sham rats, which was not present in AKI-HIFD rats. These data suggest that recovery from AKI results in impaired hemodynamic reserve and that associated CKD progression may be mitigated by HIFD in the early post-AKI period.

NEW & NOTEWORTHY Despite the apparent recovery of renal filtration function following acute kidney injury (AKI) in rats, the renal hemodynamic reserve response is significantly attenuated, suggesting that clinical evaluation of this parameter may provide information on the potential development of chronic kidney disease. Treatments such as hydrodynamic isotonic fluid delivery, or other treatments in the early post-AKI period, could minimize chronic inflammation or loss of microvessels with the potential to promote a more favorable outcome on long-term function.

INTRODUCTION

Acute kidney injury (AKI) is a common clinical entity associated with a variety of causes. A frequent cause of AKI is cellular damage brought about by hypoperfusion of the kidney due to hypotension, volume contraction, anemia, and hemorrhage (1). Collectively, all these instigating causes are associated with a mismatch between epithelial cell energy utilization versus the nutrient and oxygen supply (2). In many cases, the operant mechanism is ischemic injury followed by reperfusion injury.

To understand the pathological mechanisms of ischemia-reperfusion (I/R) injury, our laboratories have used a renal pedicle cross-clamp model in rodents to delineate short- and long-term consequences of AKI (3, 4). In this model, renal tubular cells dedifferentiate, lose structural integrity, detach from the basement membrane, and contribute to cast formation in the tubule lumen (1, 5). Endothelial injury results in red blood cell aggregation and furthers local impairment of capillary perfusion (6–8). As epithelial and endothelial injury extends, release of damage-associated molecular pattern molecules (DAMPs) and pathogen-associated molecular pattern molecules (PAMPs) activates immune cell infiltration from the vascular space to the tissue parenchyma (9, 10). When impaired blood flow resolves, renal tubular cells undergo cellular repair processes to reestablish normal epithelial function (1).

Long-term consequences of AKI can result in the progression of chronic kidney disease (CKD) (5, 11), and potential pathophysiological mechanisms leading to progression have also been investigated in rodent I/R models. Infiltrating cells can remain in the renal parenchyma and contribute to chronic inflammation (12–15). Together with hypoxia due to capillary rarefaction, inflammatory cells can promote interstitial fibrosis, the hallmark structural feature that is highly predictive of the loss of renal function in CKD (16–18).

The initiation of CKD may be masked by the return of plasma creatinine levels to baseline following initial recovery from AKI. Plasma creatinine is a relatively insensitive measure for kidney damage and the initial stages of interstitial fibrosis could be set in motion before structural alterations result in a significant change in creatinine. In addition, since the steady-state level of this biomarker is dependent upon the time average of net production and renal excretion changes in glomerular filtration rate (GFR), acute changes in GFR in response to stimuli such as exposure to protein loads are not reflected in a single serum creatinine measurement (19). Moreover, even the resolution of GFR or steady-state creatinine to preinjury levels may be misleading since compensatory hemodynamic responses may help to preserve GFR in the face of reduced function in a proportion of individual nephrons (19–21).

The compensatory hemodynamic response to increase GFR induced by various stimuli, such as protein load, is referred to as renal reserve (22). It is possible that the resolution of GFR to preinjury levels following moderate AKI is partially attributable to activation of renal reserve capacity. Moreover, given the loss of capillary networks and the initial development of interstitial fibrosis following recovery from AKI, we hypothesized that total renal reserve capacity may become impaired following recovery from AKI, representing a physiologically measurable deficit in renal function.

In a prior communication, we showed that a retrograde bolus infusion of physiological saline through the renal vein can rapidly reverse AKI following I/R injury. This process, referred to as hydrodynamic isotonic fluid delivery (HIFD), alleviated peritubular capillary congestion, increased capillary blood flow, reduced parenchymal inflammation, and improved serum creatinine when performed 24 h following an established injury (23). These findings suggest that HIFD could also have long-term benefits in the recovery from AKI. In this communication, we evaluated the long-term effects of I/R-induced AKI and HIFD 5 wk after moderate to severe I/R injury in rats. We found that the effect of HIFD on reducing inflammatory cell infiltration was maintained for up to 5 wk after AKI, and it also attenuated the degree of capillary rarefaction and interstitial fibrosis. Interestingly, we also found that renal hemodynamic reserve responses were compromised 5 wk following AKI, but that HIFD partially preserved the reserve response. Taken together, these data suggest that renal reserve may represent an important functional biomarker of the AKI to CKD transition and its diminution may be mitigated by HIFD.

MATERIALS AND METHODS

Animals

Male Sprague-Dawley rats (250–300 g) were purchased from Envigo (Indianapolis, IN) and used for all studies. The experiments were performed in accordance with the policies of the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals. Experiments were approved by the Institutional Animal Care and Use Committee of Indiana University.

AKI Model and HIFD Treatment

To assess the effect of HIFD treatment on renal function following I/R injury, we used a model of unilateral injury to a solitary kidney and simultaneous contralateral unilateral nephrectomy. Rats were anesthetized by intraperitoneal injection of ketamine (100 mg/kg) and xylazine (5 mg/kg) and then placed on a heated pad to facilitate maintenance of body temperature. Anesthetized rats were subjected to right unilateral nephrectomy (UNX) and left unilateral renal I/R by clamping the renal pedicle for 35 min to induce AKI (Fig. 1). Sham-operated controls were subjected to anesthesia and midline incision; right UNX was performed but the left kidney was not touched. At 24 h after surgery, some rats received 0.5 mL of isotonic saline by HIFD infusion through the renal vein as described previously (23); this group is referred to as AKI-HIFD. As a control, other post-AKI rats were administered 0.5 mL via the vena cava; these rats are referred to as AKI-VC. Randomization of rats into either AKI-HIFD or AKI-VC groups was based on plasma creatinine levels obtained 24 h after I/R, such that the average values were similar between the two groups at the time of treatment (Fig. 1). Sham rats received saline only through the vena cava.

Figure 1.

Experimental design schema to investigate the effect of AKI on renal vascular reserve response and the potential effect of HIFD on mitigation. Sprague-Dawley rats were subjected right kidney unilateral nephrectomy (UNX) and left kidney I/R for 35 min. Sham animals were subjected to UNX but not I/R. Following determination of plasma creatinine, post-I/R rats received 0.5 mL of isotonic saline in the renal vein (HIFD) or in the vena cava (VC). After 5 wk of recovery, rats were assessed for renal hemodynamic responses or evaluated for GFR. AKI, acute kidney injury; GFR, glomerular filtration rate; HIFD, hydrodynamic isotonic fluid delivery; I/R, ischemia-reperfusion.

A mortality rate of 22% (12 of 55 rats) defined as death or requirement for euthanasia was observed by 48 h after I/R, with similar proportions in the HIFD or VC groups, after which all rats survived to the 5-wk time point. No mortality was observed in sham controls. Of the surviving 43 post-I/R rats, 25 were used to measure the hemodynamic responses (study 1) and 18 were used to measure GFR and for tissue analysis (study 2).

Measurement of renal hemodynamic reserve (study 1).

Rats were anesthetized using isoflurane (1.5–2% vol/vol) and then placed on a heated pad to facilitate maintenance of body temperature between 36.5°C and 37°C. A catheter (PE50) placed in the right femoral artery was used for the measurement of arterial blood pressure and collection of arterial blood. An infusion of 2% (wt/vol) bovine serum albumin (BSA, Sigma-Aldrich, St. Louis, MO) in 0.9% NaCl was administered via a catheter placed in the left femoral vein, at a rate of 2 mL/100 g/h. Following a midline incision, the left kidney and renal artery were isolated, and the kidney was placed in a stainless steel cup. A transit-time ultrasound flow probe (Type 0.7 VB, Transonic Systems, Ithaca, NY) was placed around the left renal artery to allow measurement of total renal blood flow (RBF). An arterial catheter was connected to a pressure transducer (Cobe, Arvada, CO) and the transit-time ultrasound flow probe was connected to compatible flowmeter (T108, Transonic Systems). The analog signals were digitized and continuously displayed by a data-acquisition program (BIOPAC Systems, Goleta, CA), allowing continuous sampling of mean arterial pressure (MAP, mmHg) and RBF (mL/min). Urine was collected from a bladder catheter.

After completion of the surgical preparation, an equilibration period of 25 min was followed by 15 min of baseline recording and renal reserve stimulation for up to 30 min by infusion of l-arginine (7.5 mg/kg/min in 2% BSA) at the same fluid flow rate described earlier. Values for blood flow were calculated by collecting the average signal over 10-min time bins. As rats may experience fluctuations in blood pressure or blood loss during the surgical preparation, a mean arterial blood pressure value above ≥90 mmHg was established a priori for inclusion; of 37 rats (25 AKI rats and 12 sham), 24 rats were included. At the end of the experiment, rats were euthanized, and kidneys were harvested for histological and flow cytometry analysis.

Measurement of GFR (study 2).

An additional cohort of 27 rats (18 AKI and 9 sham) was used to further evaluate the recovery of GFR using noninvasive clearance (NIC) techniques. Following 5 wk of recovery, GFR was determined by fluorescein isothiocyanate (FITC)-sinistrin clearance using a noninvasive clearance (NIC) device (MediBeacon) as previously described (24). In brief, rats were anesthetized (2.5% vol/vol isoflurane) and placed on a heated pad to facilitate maintenance of body temperature. A small area of the flank was shaved followed by application of hair removal cream. The NIC kidney device/battery unit was attached using double-sided adhesive patch and adhesive tape. A baseline recording was performed (3 min) to record background signal generated within the skin. After that, FITC-sinistrin (Fresenius Kabi Austria, Linz, Austria, 2.7 mg/100 g body wt in 0.9% NaCl solution) was administered via the tail vein. The isoflurane anesthesia was withdrawn, and the conscious rat was returned to the home cage for a period of 2 h during which the elimination kinetics of FITC-sinistrin were recorded. The excretion half-life of FITC-sinistrin was used to calculate GFR by one-compartment model with linear correction using the MPD Studio software package (Mannheim Pharma and Diagnostics) according to the manufacturer’s instructions.

Measurement of Plasma Creatinine

Blood was collected in a heparinized Eppendorf tube and centrifuged at 3000 g for 10 min. Plasma creatinine was quantified using a Point Scientific QT 180 Analyzer (Point Scientific, Canton, MI) according to the manufacturer’s instructions.

Fluorescence-Activated Cell Sorting

The kidney was bisected and ∼1/2 was processed for the presence of inflammatory cells based on methods previously described (14). In brief, after tissue digestion (2 µg/mL Liberase, Roche, Indianapolis, IN) and Percoll (Sigma, St. Louis, MO) separation, kidney mononuclear cells were stained with antibodies against CD4 (PE-Cy7) and CD8a (Alexa 647). To evaluate the T helper (Th) subtype, the cells were permeabilized with 0.1% saponin and stained with antibodies against cytokines such as IFN-γ (FITC; Th1), IL-4 (PE; Th2), and IL-17 (FITC; Th17). To evaluate regulatory T cells, mononuclear cells were stained with either CD4 or CD4 along with Foxp3 (PE). For identification of macrophages, cells were stained with an antibody against CD11b/c. The sources, clones, and catalog numbers for antibodies used in this study are shown in Table 1. Cells were scanned using a BD LSRFortessa analyzer (BD Biosciences, San Jose, CA) and signals were analyzed using FlowJo software (Tree Star, Ashland, OR). Scanning and gating strategies were exactly as described previously (14).

Table 1.

Antibodies utilized for flow cytometry

Name	Catalog	Clone	Source
Mouse anti-rat CD4 PE-Cy7	561578	OX-35	BD Biosciences
Mouse anti-rat CD8 Alexa fluor 647	561611	OX-8	BD Biosciences
Anti-mouse/rat IL-17A monoclonal antibody FITC	11-7177-80	Ebio17b7	eBioscience
Mouse anti-rat IFN-y FITC	559498	DB-1	BD Biosciences
Mouse anti-rat IL-4 PE	555082	OX-81	BD Biosciences
Mouse anti-rat CD11b/c PerCP-Cy5.5	554862	OX-42	BD Biosciences
Anti-mouse/rat FOXP3 PE	12_5773_82	FJK-16s	eBioscience

Histology and Immunohistochemistry

For histology studies, ∼1/4 of the kidney was fixed with 10% formalin, and 4 µm paraffin-embedded sections were stained with Masson’s trichrome to assess the presence of tubulointerstitial fibrosis. All sections were scanned using Aperio whole slide digital imaging system (Vista, CA) at ×20. Quantification of fibrosis was determined by measuring % positive blue stain using HALO image analysis software (https://indicalab.com/halo/) by a member of the laboratory who was blinded to the treatments.

To assess renal vascular density, ∼1/4 of each kidney was fixed in ice-cold methanol and ∼50 µm sections were obtained using a vibratome. Renal vessels were labeled using an antibody generated against cablin as we have described previously (25). Cablin was originally reported as a novel protein in the vascular basal lamina (26), and analysis of the sequence reported in this paper revealed that it was identical to the sequence for plasmalemmal vesicular-associated protein (PLVAP; Accession No. NM_031310.3), reported to be expressed in kidney capillary and venous endothelial cells (27, 28).

The immunofluorescent signal was detected using an Olympus FV1000 confocal microscope, and vascular density measurements were determined by quantifying the number of vascular structures intersecting an arbitrary grid using ImageJ software (NIH), exactly as we have described previously (25, 29). For both studies on fibrosis and cablin immunostaining, subsets of available samples from either the hemodynamic or GFR studies were randomly selected for staining and processing.

Statistical Analysis

Data are presented as means ± SE or medians [interquartile ranges (IQR)]. Statistical analyses were performed with GraphPad Prism 9.2 (GraphPad Software, La Jolla, CA). Data were analyzed using one-way analysis of variance (ANOVA) or Kruskal–Wallis one-way ANOVA test with Dunn’s multiple comparisons test or repeated measures two-way ANOVA with Bonferroni’s multiple comparison test, as described in the figures. Student’s two-sided unpaired t test was used for dichotomous comparisons, with P ≤ 0.05 considered statistically significant.

RESULTS

We have previously reported that recovery from unilateral I/R and simultaneous UNX in rats resulted in AKI with subsequent recovery of renal function by 4–8 wk, but that longer recovery (up to 20 wk) resulted in manifestation of CKD as indicated by interstitial fibrosis, development of proteinuria, and reduced creatinine clearance (30). Therefore, we sought to investigate whether renal reserve was impaired at 5 wk after I/R, following recovery of renal function but before functional evidence of CKD. Plasma creatinine levels in sham-operated UNX controls were 0.71 ± 0.3 mg/dL at 24 h, and these values decreased 5 wk following recovery to 0.47 ± 0.01 mg/dL, indicative of an expected compensatory response of the remaining uninjured kidney following UNX (Fig. 2A). I/R injury to a solitary kidney with simultaneous UNX resulted in a significant increase in plasma creatinine relative to the UNX sham control within 24 h. Similar to prior results (23), treatment of AKI rats with HIFD at 24 h resulted in significantly reduced plasma creatinine levels at 48 h relative to AKI-VC rats (Fig. 2A). Plasma creatinine returned to levels of sham-operated controls by 5 wk after I/R and was not different between AKI-VC and AKI-HIFD (Fig. 2A). Since a small difference in GFR may not be detectable by plasma creatinine, a second group of rats was used to measure the recovery of GFR by transcutaneous FITC-sinistrin clearance. In this group, GFR was not different between both AKI-VC and AKI-HIFD rats relative to sham controls at 5 wk of recovery (Fig. 2B). No differences were observed in body weights between groups, and there were no differences in the kidney weight/body weight (KW/BW) ratios between AKI-VC versus AKI-HIFD groups. However, both AKI groups had significantly greater KW/BW (14.4 ± 9.6% for AKI-VC; 13.5 ± 10.9% for AKI-HIFD) relative to sham controls (P < 0.05 by ANOVA).

Figure 2.

Recovery of renal function following I/R with or without HIFD injection. A: plasma creatinine from all rats included for either hemodynamic studies or GFR assessment, following sham surgery, AKI-VC, or AKI-HIFD treatments. Data are presented as means ± SE; *P ≤ 0.05 for 48 h AKI-HIFD vs. 48 h AKI-VC by two-way repeated measures ANOVA with Bonferroni’s multiple comparison test. For sham, n = 21; for AKI-VC, n = 26; and for AKI-HIFD, n = 17. B: a subset of rats from A was used to assess GFR 5 wk following recovery by noninvasive transcutaneous clearance. Data are presented as means ± SE, and GFR values were not different between any group (P = 0.46 by one-way ANOVA; n = 7–11/group as shown). AKI, acute kidney injury; GFR, glomerular filtration rate; HIFD, hydrodynamic isotonic fluid delivery; I/R, ischemia-reperfusion; ns, not significant; VC, vena cava.

To determine the effects of AKI on renal reserve responses at 5 wk after surgery, rats were anesthetized and instrumented to evaluate RBF responses. Of the rats included in the study, baseline MAP was slightly but significantly higher in sham-operated controls versus AKI-VC rats under isoflurane anesthesia (Fig. 3A). However, baseline RBF values did not differ significantly between sham animals and AKI-VC or AKI-HIFD rats (Fig. 3B). Baseline urine flow rates appeared slightly depressed in AKI-HIFD rats relative to sham, but the differences were not significant (Fig. 3C).

Figure 3.

Baseline renal hemodynamics in anesthetized rats 5 wk after IRI. Baseline mean arterial blood pressure (MAP; A), renal blood flow (RBF; B), and urine flow rates (C) were obtained by averaging over a 15-min recording period after equilibration. Data are presented as means ± SE. In A, *P < 0.05 in sham vs. AKI-VC by Dunn’s multiple comparisons test. For A and B, n = 7 for sham, n = 8 for AKI-VC, and n = 9 for AKI-HIFD. For C, n = 7 for sham, n = 6 for AKI-VC, and n = 9 for AKI-HIFD, with missing values in C due to failure of the bladder catheter in 2 rats. AKI, acute kidney injury; GFR, glomerular filtration rate; HIFD, hydrodynamic isotonic fluid delivery; IRI, ischemia-reperfusion injury; ns, not significant; VC, vena cava.

To evoke a reserve response, rats were subjected to l-arginine infusion (7.5 mg/kg/min). MAP did not change significantly from baseline in any group (sham = 4.8 ± 2.7%, AKI-VC = 2.0 ± 1.9%, AKI-HIFD = −0.8 ± 1.2%). However, there was a significant 22.6 ± 2.2% increase in RBF in sham-operated control rats versus baseline, which reached a plateau around 30 min of infusion. This response was significantly impaired in AKI-VC rats (Fig. 4A). The individual RBF responses versus baseline after 30 min of infusion are shown in Fig. 4B. The median RBF response was 21.4% of baseline (IQR 18–28) in sham control rats and was significantly reduced to 13.1% (IQR 11–15) in AKI-VC rats (Fig. 4B). In AKI-HIFD rats, the RBF response to l-arginine appeared partially protected, and the % RBF response was not statistically different from either the sham or AKI-VC rats (Fig. 4, A and B). In addition, urine flow rates were also significantly increased in response to l-arginine infusion relative to baseline (88.1%; IQR 57–120%) in sham control rats, which was significantly blunted in AKI-VC rats. AKI-HIFD rats showed an intermediate response that was not significantly different from sham or AKI-VC groups (Fig. 4C).

Figure 4.

Effect of AKI and HIFD renal hemodynamic reserve response induced by l-arginine. Increased renal blood flow (RBF) in response to infusion of l-arginine is shown. A: % change of RBF relative to baseline averaged over consecutive 10-min time bins for up to 30 min; data are presented as means ± SE. *P < 0.05 AKI-VC vs. sham by two-way repeated measures ANOVA with Bonferroni’s multiple comparison test. B: %RBF response as median ± IQR at 30 min, illustrating individual data points shown. C: increased urine flow rate compared with baseline presented as medians (IQR) with individual data points shown. For B and C, *P < 0.05 AKI-VC vs. sham by Dunn’s multiple comparison test and unpaired t test, respectively. The n for each group is exactly as described in Fig. 3. AKI, acute kidney injury; GFR, glomerular filtration rate; HIFD, hydrodynamic isotonic fluid delivery; IQR, interquartile range; I/R, ischemia-reperfusion; ns, not significant; VC, vena cava.

HIFD at 24 h following I/R was previously demonstrated to reduce the number of kidney inflammatory cells during the subsequent 24-h period of recovery (13). To evaluate whether HIFD resulted in sustained reductions in inflammatory cells at 5 wk of recovery, FACS was conducted on kidney mononuclear cells to measure various lymphocyte and macrophage populations. The number of Th17 cells (CD4⁺/IL17⁺) present in AKI-VC remained significantly elevated relative to sham controls after 5 wk of recovery, whereas the number of Th17 cells in AKI-HIFD rats was similar to sham controls (Table 2). Unlike Th17 cells, total CD4⁺, CD8⁺, Th1, Th2, or CD11b/C in AKI-VC rats did not remain significantly elevated relative to sham at 5 wk after AKI (Table 2).

Table 2.

Effects of HIFD on infiltration of cells 5 wk following renal I/R injury

Cell Type	Sham	AKI-VC	AKI-HIFD	P
CD4+	15.7 ± 4.9	24.42 ± 5.8	19.3 ± 4.6	0.69
CD8+	2.8 ± 0.6	7.1 ± 1.7	4.6 ± 1.1	0.22
CD4+/IL17+ (Th17)	0.53 ± 0.26	1.7 ± 0.39*	0.64 ± 0.19	0.02
CD8+IL17+	0.65 ± 0.17	1.2 ± 0.30	0.55 ± 0.14	0.30
CD4+Foxp3+	1.1 ± 0.5	2.1 ± 0.76	1.6 ± 0.59	0.55
CD4+IFN-γ+ (Th1)	0.25 ± 0.01	0.48 ± 0.11	0.33 ± 0.13	0.41
CD4+IL4+ (Th2)	0.096 ± 0.04	0.29 ± 0.08	0.13 ± 0.07	0.11
CD11 b/c	5.9 ± 2.3	6.7 ± 1.9	6.9 ± 2.5	0.94

Inflammatory cell types are expressed as number of cells per gram kidney weight × 10⁵. Values are expressed as means ± SE. P values are the outcome of Kruskal–Wallis one-way ANOVA test. AKI, acute kidney injury; HIFD, hydrodynamic isotonic fluid delivery; I/R, ischemia reperfusion; VC, vena cava.

*P < 0.05 sham vs. vena cava (Dunn’s multiple comparisons test).

Additional hallmarks of the AKI-to-CKD transition are the rarefaction of peritubular capillaries as well as the expansion of interstitial fibrosis (4, 31). Cablin/PLVAP⁺ microvessel density was significantly reduced in AKI-VC rats relative to sham-operated controls (Fig. 5); however, microvessel density in AKI-HIFD rats was similar to sham rats and preserved relative AKI-VC rats (Fig. 5B). There was also a significant increase in renal fibrosis in AKI-VC rats versus sham control rats as demonstrated by Masson’s trichrome stain, but fibrosis was not significantly increased in AKI-HIFD rats versus sham control rats (Fig. 6).

Figure 5.

Effect of IRI and HIFD on renal medullary capillary density. A: representative images of cablin-stained blood vessels shown in the renal outer medulla of sham, AKI-VC, or AKI-HIFD rats, as labeled. Higher magnification of the area outlined in white is presented below each image. Scale bar = 200 µm. B: cablin-stained kidney sections were scored for vessel density using ImageJ. Data are presented as means ± SE (n = 4–6 in each group). *P ≤ 0.05, sham vs. AKI-VC and AKI-VC vs. AKI HIFD (one-way ANOVA with Bonferroni’s multiple comparison test). AKI, acute kidney injury; HIFD, hydrodynamic isotonic fluid delivery; ns, not significant; VC, vena cava.

Figure 6.

Effect of IRI and HIFD on renal interstitial fibrosis following IRI. A: representative images of Masson’s trichrome (blue stain) cross section of kidney harvested 35 days after renal ischemia or sham surgery. Micrographs are shown for the renal outer medulla acquired by Aperio whole slide digital imaging system (scale bars = 200 µm). B: quantification of blue-stained positive fibrosis using HALO image analysis software. Data are presented as means ± SE (n = 5–10 in each group). *P ≤ 0.05, sham vs. AKI-VC (Dunn’s multiple comparisons test). HIFD, hydrodynamic isotonic fluid delivery; IRI, ischemia-reperfusion injury; ns, not significant; VC, vena cava.

DISCUSSION

Renal repair responses have been shown to be critical for recovery of renal structure and function following an episode of AKI; however, it has been recognized that incomplete or maladaptive repair responses fail to return the kidney to its pristine preinjury state and may predispose to the development of CKD (1, 11, 32). Despite this recognition, there has been little advance on the understanding of the factors that contribute to development of CKD following AKI or on potential treatments to mitigate CKD progression initiated after the establishment of injury. Clinically, there is an unmet need to identify specific patients who may be at risk for subsequent progression of CKD following AKI (20, 33, 34).

In the current report, we used a model of I/R-induced AKI in rats with a solitary kidney, which was characterized by an increase in plasma creatinine within 24 h as is typical in this model. I/R resulted in significant initial injury associated with 22% mortality in the first 2 days after surgery, but there was consistent recovery in surviving animals based on the resolution of serum creatinine and GFR at 5 wk relative to sham controls. Following recovery, several structural and functional parameters remained impaired such as urinary concentrating ability, a reduction in peritubular capillary density, an expansion of myofibroblasts, as well as a persistent presence of various immune cells (1, 4, 11–13, 32). In the current study, we also observed evidence of interstitial fibrosis 5 wk following AKI, suggesting the initiation of CKD before a loss in overall filtration function. Other investigators have suggested that a reduced number of functioning nephrons, possibly brought on by failed recovery of individual tubules, are also present following AKI depending on the severity of initial injury (35, 36). In addition, a myriad of changes in gene expression are persistently altered after AKI, which could contribute to a profibrotic environment associated with progressive CKD (37).

Despite this, a determination of whether such changes influence kidney function before functional evidence of CKD has been elusive. The hypothesis that renal reserve responses may provide insight into potential CKD progression before changes in plasma creatinine has received significant attention in recent years. This is based in part on the understanding that reserve function is invoked in the early stages of CKD, keeping creatinine or baseline GFR near normal values (20, 21). In practice, there is not yet a universally accepted and standardized approach to evaluate renal reserve in humans, but most approaches have relied on measuring changes in GFR following an oral protein load or amino acid infusion (19, 21). Interestingly, there are some reports that have demonstrated decreased renal reserve function in patients following AKI. Notably, Husain-Syed et al. (38) studied renal reserve responses in patients 3 mo following cardiac surgery. Patients who had developed AKI had a restored GFR to baseline but showed evidence of circulating kidney injury biomarkers and a reduced level of renal reserve response compared with patients who did not develop AKI. On a similar note, Fagugli et al. investigated renal reserve responses in kidney transplant patients with estimated GFR >50 mL/min/1.73 m² and reported that a subset of patients manifested an impaired reserve response relative to healthy control patients or transplant patients with intact renal reserve response. The patients with impaired renal reserve also had higher urinary thromboxane levels than transplant patients with intact reserve responses and manifested a higher filtration fraction indicative of a hyperfiltering state (39). Taken together, these reports suggest that clinical AKI or subclinical renal injury can compromise renal reserve responses. However, there have been no reports in animal models demonstrating impaired renal reserve following recovery of AKI, which may be useful to investigate mechanisms. Based on studies by our group and others that have shown impaired renal vascular responses and reduced microvascular density following AKI (1, 4, 15, 25, 31, 40), we addressed the possibility that rats may have reduced renal reserve capacity following an episode of AKI.

The measurement of reserve GFR responses in conscious rats presents challenges, since even with recent advances in noninvasive GFR measurements, there would be required washout periods of the fluorescent sinistrin probes that make successive GFR measurements before and after an oral protein load challenging. Therefore, as an alternative, we chose to measure blood flow in anesthetized rats, to allow for a simplified comparison between baseline and protein-induced responses. Several options have been suggested to induce renal reserve in rats. Previous studies have used an infusion of different amino acid solutions to investigate renal reserve responses, measuring changes in renal blood flow and/or changes in single nephron GFR by micropuncture (for a review, see Ref. 22). The basis for the increased renal blood flow in response to amino acid infusion is not entirely clear, but several possible mechanisms have been proposed, including vasodilation secondary to glucagon release, amino acid effects on renal arachidonic acid metabolism, or reduced distal sodium delivery resulting from increased proximal sodium/amino acid cotransport altering the TGF response (22). Several of these studies have demonstrated significantly impaired reserve responses in models of hypertension or diabetes (22, 41, 42), but have not been applied in the setting of AKI.

For our study, we examined different protocols to induce renal reserve based on published reports, including infusion of glycine as described previously (42, 43). However, we chose the approach described by Ruiz et al. (44) using 7.5 mg/kg/min of l-arginine infusion, as this provided the most consistent and robust increase in RBF in our hands. Notably, the increase in RBF occurs without a significant change in systemic arterial blood pressure. The basis for the kidney response to l-arginine or the impaired response to l-arginine in AKI rats is not completely known. Increased RBF in response to l-arginine has been suggested to be mediated in part by the cationic amino acid transporter CAT-1, which mediates cellular uptake of l-arginine and allows for generation of vasodilatory nitric oxide (NO) (45). Alterations in renal NO regulation have been widely reported in the genesis of kidney I/R injury (1), but its long-term regulation has not been addressed. Thus, it is possible that impaired NO signaling underlies the impaired responses reported here and will require further investigation.

Alternatively, impaired renal reserve has been suggested to reflect loss of total functioning nephrons in the setting of progressing CKD, as the remaining nephrons operate at a higher level of their initial adaptive capacity (22). This is a potential explanation for the impaired renal reserve response, as there has been a reported loss of total nephron following severe AKI due to inefficient repair (1).

Finally, impaired reserve responses might also occur secondary to alterations in vascular density or vascular tone secondary to inflammation. Our prior work has demonstrated that persistent expression of Th17 cells contributes to the AKI to CKD transition (14, 46). HIFD of saline was previously shown to improve vascular congestion, reduce inflammatory Th17 cell levels, and hasten recovery in the early postischemic period (23). The data presented in the current study suggest that HIFD also has persistent effects on reducing Th17 cell infiltration, which represents one possible mechanism to preserve renal function and structure. Based on recent studies by Lilley et al. (47), it is possible that inflammatory cytokines may have direct or indirect effects on vascular cells in the kidney to promote vasoconstriction. In addition, capillary rarefaction, which has been previously suggested to promote CKD due to enhanced hypoxia (11), was also improved by HIFD. Other investigators have shown that peritubular capillary rarefaction is a process that occurs over the course of several days following an initial kidney insult (48), and thus our data presented here suggest that improved reestablishment of perfusion at the early stages following AKI has the potential to attenuate ongoing damage to peritubular capillaries. Taken together, the improvement in vascular structure and immune cell infiltration could represent mechanisms contributing to both preserved renal reserve and reduced interstitial fibrosis.

To the best of our knowledge, this is the first report using an animal model of AKI demonstrating impaired renal vascular functional reserve. As has been suggested, it is possible that this impaired functional reserve is indicative of progressing CKD. Since renal reserve represents a functional parameter that could be measured clinically, measurement of impaired reserve responses could identify patients at risk of progression following recovery from AKI (19, 21, 33). In this report, the term “recovery” is based on clinically relevant parameters related to filtration function but point out that it is not known whether kidney repair is complete at the time studied or is ever fully realized. It is possible that a more prolonged period of recovery and potential regression of inflammation could restore renal reserve responses to levels closer to sham. In practice, assessment of renal reserve may require evaluation at longer or multiple time points after AKI to be useful.

Perspectives and Significance

The development of AKI has been linked to the development of CKD even in patients who experience recovery from the initial loss of function. The loss of renal reserve could represent a functional marker of developing CKD in patients despite recovery of baseline GFR. Here, we demonstrate in a rat model of AKI with evidence of GFR recovery that renal vascular reserve is impaired and associated with persistent capillary rarefaction, interstitial fibrosis, and inflammation. We note that early intervention with HIFD following injury can have a persistent effect on reduction of inflammation, fibrosis, and rarefaction and improves reserve response. These data support the notion that renal reserve is an important functional marker that could provide important clinical insights into the efficacy of renal repair and provides an animal model to investigate how various treatment may preserve renal reserve and long-term function.

DATA AVAILABILITY

Data will be made available upon reasonable request.

GRANTS

This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grant 1RO1DK063114 (to D.P.B.), an award from the Indiana University Showalter Scholar Trust (to D.P.B.), and Veterans Affairs Merit Review Award 1 I01 BX001736-01A1 (to R.L.B.).

DISCLOSURES

R.L.B. is the founder of AKIcept, which seeks to develop HIFD technology for clinical application and has submitted a patent on this technology. D.P.B. is a consultant for AKIcept. None of the other authors has any conflicts of interest, financial or otherwise, to disclose.

AUTHOR CONTRIBUTIONS

M.M.U., J.A.C., R.L.B., and D.P.B. conceived and designed research; M.M.U. and J.A.C. performed experiments; M.M.U., J.A.C., R.L.B., and D.P.B. analyzed data; M.M.U., J.A.C., R.L.B., and D.P.B. interpreted results of experiments; M.M.U., J.A.C., and D.P.B. prepared figures; M.M.U., J.A.C., and D.P.B. drafted manuscript; M.M.U., J.A.C., R.L.B., and D.P.B. edited and revised manuscript; M.M.U., J.A.C., R.L.B., and D.P.B. approved final version of manuscript.