Rodent models of AKI-CKD transition

Abstract

Acute kidney injury (AKI) is a contributing factor in the development and progression of chronic kidney disease (CKD). Despite rapid progresses, the mechanism underlying AKI-CKD transition remains largely unclear. Animal models recapitulating this process are crucial to the research of the pathophysiology of AKI-CKD transition and the development of effective therapeutics. In this review, we present the commonly used rodent models of AKI-CKD transition, including bilateral ischemia-reperfusion injury (IRI), unilateral IRI, unilateral IRI with contralateral nephrectomy, multiple episodes of IRI, and repeated treatment of low-dose cisplatin, diphtheria toxin, aristolochic acid, or folic acid. The main merits and pitfalls of these models are also discussed. This review provides helpful information for establishing reliable and clinically relevant models for studying post-AKI development of chronic renal pathologies and the progression to CKD.

INTRODUCTION

Acute kidney injury (AKI) is a major kidney disease with high morbidity and mortality around the world (38, 68). Besides its acute consequence, recent epidemiological and experimental studies have provided substantial evidence that AKI contributes significantly to the development and progression of chronic kidney diseases (CKD) (2, 6, 17, 21, 59, 61). Moreover, the severity and frequency of AKI are closely associated with its progression to CKD (24, 55, 57). AKI is pathologically characterized by injury and death of tubular epithelial cells (34). After the initial injury, surviving tubular epithelial cells undergo dedifferentiation for proliferation and kidney repair. In this process, normal repair can restore tubular epithelial integrity and function; however, incomplete or maladaptive repair results in the development of chronic pathologies (e.g., interstitial fibrosis) and the progression to CKD (4, 23, 59, 61). Emerging evidence suggests that after severe or episodic AKI, normal tubular epithelial cells undergo a phenotype change with persistent production and secretion of profibrotic factors, leading to atrophic proximal tubules and renal interstitial fibrosis (20, 23, 41, 61, 67). However, it is largely unclear how these cells become atrophic and assume a profibrotic phenotype.

Animal models are commonly used for the research of AKI-CKD transition as well as assessing potential therapies. Especially, chronic pathologies develop in kidneys after renal ischemia-reperfusion injury (IRI). As such, rodent models of bilateral IRI (bIRI), unilateral IRI (uIRI), unilateral IRI with contralateral nephrectomy (uIRIx), and multiple episodes of IRI have been used or considered for studying AKI-CKD transition and related postinjury pathological alterations. In addition, nephrotoxicity may have long-term adverse effects in kidneys leading to CKD. In this regard, recent studies have begun to reveal the development of chronic renal pathologies after repeated treatment of low-dose cisplatin, diphtheria toxin (DT), aristolochic acid (AA), and folic acid (FA) in mice. In this review, we present these commonly used rodent models of AKI-CKD transition (Fig. 1) and highlight their merits and limitations.

Fig. 1.

Schematic diagram of rodent ischemi-reperfusion injury (IRI) models for acute kidney injury (AKI)-chronic kidney disease transition. bIRI, bilateral IRI; uIRI, unilateral IRI; uIRIx, unilateral IRI with contralateral nephrectomy; multiple IRI, multiple episodes of IRI.

IRI MODELS OF AKI-CKD TRANSITION

IRI is one of the leading causes of AKI, which occurs in clinical situations including vessel occlusion in major surgeries, postoperative hypoperfusion, bleeding, dehydration, shock, and sepsis. Renal IRI consists of a temporary impairment and subsequent restoration of oxygen and nutrient delivery to kidney cells, initiating a cascade of deleterious cellular responses leading to tubular cell injury and death, inflammation, and vascular dysfunction (1, 4, 34, 49). In addition to these acute changes, post-IRI kidneys may develop chronic renal pathologies and CKD over a few weeks to several months in rodents, providing clinically relevant research models of AKI-CKD transition. At present, commonly used rodent IRI models for AKI-CKD transition research include bIRI, uIRI, and uIRIx with contralateral nephrectomy.

bIRI Model of AKI-CKD Transition

bIRI is induced by blockage of blood flow to both kidneys, of which the impact on renal hemodynamics is more relevant to human pathophysiology. In 2001, Basile et al. (3) developed a rat model of bIRI to study the long-term effect of AKI. In their study, male rats were subjected to 60-min bilateral renal ischemia followed by 4, 8, or 40 wk of reperfusion. They found that tubular morphology essentially returned to normal at 4 and 8 wk postinjury but tubulointerstitial fibrosis developed at 40 wk, which was associated with transforming growth factor-β1 (TGF-β1) expression. Interestingly, there was significant reduction in peritubular capillary density in the inner stripe of the outer medulla at each of the postinjury time points examined. It is also notable that the rats started to have proteinuria from week 16 postinjury. The results suggest that severe ischemic injury results in a permanent alteration of renal capillary density that may contribute to the development of renal fibrosis (3). In 2010, Yang et al. (67) examined several mouse models of fibrosis. For bIRI, they did bilateral renal artery clamping for 30 min at 37.0°C for moderate IRI or for 32 min at 37.5°C for severe IRI in BALB/c mice. Renal function analysis showed that serum creatinine in the moderate IRI model increased at early stage and then returned to normal levels at day 7 after the initial injury, and these mice showed no fibrotic changes by day 42, suggesting a complete restoration of renal integrity. In contrast, the mice in the severe IRI model had sustained high levels of serum creatinine that did not return to baseline at day 42 after injury, and these mice developed interstitial fibrosis as indicated by histological examination and immunostaining of fibrosis markers. Collectively, these findings indicate that severe bilateral IRI leads to chronic changes in kidneys with some of the characteristic features of AKI-CKD transition, especially tubulointerstitial fibrosis (67). The bIRI model has been verified and used for AKI-CKD transition (or more specifically post-AKI fibrosis) study by other investigators (12, 22, 65). For example, Xiao et al (65) recently compared the long-term effects of different durations of bIRI. In this study, bIRI was induced in male C57BL/6 mice by bilateral clamping of renal arteries for 20 or 30 min at 37°C followed by reperfusion for 1, 3, or 10 days. It was shown that 20 min of bIRI triggered mild AKI, while 30 min of bIRI caused much more severe AKI. With mild IRI, serum creatinine was significantly elevated at 1 day of reperfusion but returned to the baseline at day 3. However, in severe IRI, serum creatinine level remained significantly higher than the baseline even at 10 days after injury, suggesting a persistent kidney failure. Consistent with these functional data, the renal histology of mild IRI was completely restored in 10 days, whereas renal damage was still evident 10 days after severe IRI. These results suggest that the severity of injury is a key determinant dictating the divergent outcomes of AKI. We also tested multiple ischemic time and temperature gradients, showing that 22–25 min of ischemia in male C57BL/6 mice resulted in a moderate damage with full recovery within 1 wk, while 30 min of ischemia caused serious kidney damage resulting in significance animal death in days after reperfusion (46, 63). Body temperature is a critical determinant of ischemic injury and recovery. According to our experience, the use of a homeothermic system with rectal probe is essential (36, 63). In addition, Marschner et al. (36) used an egg-breeding device during the ischemia time and an infrared lamp for the wake-up phase to help maintain body temperature, resulting in a significant reduction of interindividual variability following renal IRI. Therefore, by adjusting and maintaining ischemic time and body temperature, a moderate (fully recoverable or reversible) AKI model or a severe injury with progressive AKI-CKD model can be established in rodents.

The bIRI model may be more relevant to human pathophysiological situations, because often both kidneys are affected in human patients. In addition, renal functional changes are readily measurable in the bIRI model by collecting urine and serum samples at desired time points. The biggest problem of the bIRI model is its variation or inconsistency. As discussed above, if AKI is too severe, mice may die in the acute injury phase, and if too mild, kidneys may fully recover and do not progress to chronic pathologies or CKD (52, 63). Thus it is critical to control bIRI conditions including body temperature and ischemic duration. In this regard, variations in half minute or 0.5° would result in significant differences in the severity of AKI (31, 46, 60, 63, 70). Furthermore, the choice of anesthetic drugs also has an effect on the degree of renal damage. During our selection process (63), it was found that the dose of ketamine/xylazine required for an hour of anesthesia caused significant animal death. Isoflurane can be conveniently controlled, but it has protective effects against renal IRI (39). In our experience, pentobarbital provides relatively reliable anesthesia for mouse renal IRI surgery, and animal loss is minimal if the dose is controlled well (starting from 50 mg/kg and supplemented when necessary). In addition, different mouse strains and even different colonies of the same mouse strain have differences in their susceptibility to ischemic AKI (5). For example, a recent study showed that C57BL/6 mice were more sensitive to ischemic-reperfusion damage than 129/Sv mice (35). Furthermore, our previous study showed that different colonies of C57BL/6 mice may have differences in their susceptibility to renal IRI (62). These factors are particularly important for studying the chronic or long-term effects of AKI. In our experience, small (5–10%) differences in initial injury measured by blood urea nitrogen may induce much bigger (20–50%) changes in the development of chronic pathologies, such as renal fibrosis. Additionally, the bIRI model has the inherent issue of balance between the right and left kidneys. In other words, small differences in injury between these two kidneys may lead to significant differences between them in chronic pathologies weeks later.

uIRI Model of AKI-CKD Transition

uIRI is induced by blockage of blood supply to one kidney. In this model, the contralateral kidney is intact and functional, allowing long-term animal survival for researching AKI to CKD transition. In 2011, Zager et al. (71) subjected CD-1 mice to 30 min of uIRI at 37°C and assessed renal pathologies at 1 day, 1 wk, or 3 wk respectively. At 1 day post-ischemia, there was widespread necrosis in proximal tubules, while overt tubular dilatation was not observed and the interstitium revealed mild inflammatory cells infiltration. At 3 wk, there were renal fibrosis, inflammation, and interesting histone modifications (71). This model has been used by other investigators to study chronic pathologies after initial IRI. For example, Lech et al. (32, 33) clamped the left renal pedicle of C57/BL6 mice at 37°C for 45 min and examined the postischemic response. Within days, they noticed acute pathological changes, such as brush border loss, cast formation, and tubular dilatation. At 10 wk, there was a loss of kidney weight that was associated with tubular atrophy, interstitial fibrosis, and inflammation (32, 33). Danelli et al. (9) did 60 min of uIRI in RMB mice and followed up for 2 days, 2 wk, or 6 wk after reperfusion. Depending on the time point post-IRI, the pathological changes in kidney tissue ranged from early inflammation and tubular necrosis at 2 days to long-term organ and tubular atrophy, persistent inflammation, and fibrosis development at 2 to 6 wk. Of note, while the focus of this line of research is on the injured kidney, the contralateral kidney may have cellular and molecular changes that are functionally compensatory following uIRI.

The model of uIRI has the notable advantage for prolonged observation. In general, the observation time of bIRI in mice varies from a few days to 2 wk, but uIRI allows for much longer time studies without significant animal loss (7, 31, 37, 71). Moreover, due to the presence of functional contralateral kidney, the animals in uIRI can survive from severe renal ischemia induced by relatively long clamping time, which produces more consistent AKI resulting in more reproducible chronic pathologies, including interstitial fibrosis. In addition, the same duration of ischemia induces more renal fibrosis in uIRI than in bIRI or uIRI with contralateral nephrectomy (31). These features make uIRI a much more reliable model for investigation of postischemic AKI-CKD transition. The pitfall of the uIRI model is the difficulty of monitoring the renal functional decline as an indication of the progression of kidney injury and deterioration, because the contralateral kidney is functional and compensatory. One way to partially circumvent this problem is to remove the contralateral kidney 1 day before the death of the animal and then collect urine and blood samples at the time of the death of the animal for blood urea nitrogen and creatinine measurement (52). Nonetheless, monitoring changes of renal function at several time points in the same animal is difficult and requires more sophisticated techniques such as ^99mTc-MAG3 imaging (48, 56).

uIRIx Model of AKI-CKD Transition

The uIRIx model includes the removal of the contralateral kidney on the basis of uIRI, allowing functional evaluation of the IRI-injured kidney. Early studies by Finn et al. (18, 19) induced IRI in rats by 60 min of complete unilateral renal artery occlusion. They found that if the contralateral kidney was removed before ischemia, reflow of blood to the postischemic kidney would be better and conducive to recovery and, as a result, renal tubular structure and function would be better preserved, although they did not further apply this model to AKI-CKD transition research (18, 19). In this regard, Yang et al. (67) performed right nephrectomy 3 days after uIRI of the left kidney. They found that at 42 days after uIRI, the group with nephrectomy had significantly less fibrosis than uIRI-only mice. These findings indicated that the uIRIx model not only could progress into chronic renal pathologies but also could withstand greater intensity of ischemic treatment, thereby making the long-term model more stable (67). Multiple causes may underlie the better outcome of uIRIx, but a major factor is the renal hemodynamics in the IRI kidney. In the presence of the functional contralateral kidney, lower blood reflow would be restored to the IRI-injured kidney, resulting in more severe and prolonged “persistent ischemia” for tubular damage and the development of interstitial pathologies including fibrosis. Skrypnyk et al. (52, 53) further compared moderate and severe uIRI with nephrectomy in mice. By clamping the renal pedicle for 26 min at 38°C with a simultaneous contralateral nephrectomy in BALB/c mice, they found only 50–60% of mice developed renal insufficiency 24 h after injury, which was fully recovered later. In the severe injury model, mice were subjected to ischemia for 30 min with contralateral nephrectomy 8 days later, which developed consistent kidney injuries and postinjury renal fibrosis at day 28. These observations suggest that adequate duration of ischemia is critical to the continued injury and incomplete recovery in this model (52, 53).

The uIRIx model has several notable merits. Compared with bIRI, this model has less variability and allows for longer ischemic time to induce consistent kidney injury for studying its progression to chronic pathologies. Compared with uIRI without nephrectomy, this uIRIx model allows researchers to assess renal function of the animals at multiple time points to indicate kidney injury and repair. However, this model is associated with significant animal death if the initial injury is severe. At our hands, ~30% mice were lost within 2 wk after 30 min of left uIRI with right nephrectomy. We also noticed bigger variations in this model as compared with uIRI only.

Repeated IRI Model of AKI-CKD Transition

On the basis of developing fibrotic damage by multiple episodes of tubular damages (20, 30, 55), similar results might be expected after repeated IRI insults. In 1984, Zager et al. (69) reported a study of repeated bIRI in female Sprague-Dawley rats. When compared with single injury, repeated bIRI did not induce significant declines in glomerular filtration rate (GFR) and had no significant deterioration in renal histopathology. Conversely, they found that, instead of increasing the susceptibility to additional ischemic injury, repeated IRI had a protective effect (69). The phenomenon of protection afforded by a prior episode of IRI is known as “preconditioning”, which has been verified by many other studies (28). For example, Park et al. (45) further extended the second IRI to 8 or 15 days later and found that AKI induced by this IRI was remarkably less than a single IRI without prior injury. Another interesting observation from this study is that ischemia of one kidney provided protection against the second injury in this kidney but not in the contralateral kidney, indicating the involvement of intrinsic mechanisms of protection by preconditioning (45). In view of the protective effect of preconditioning, it seems unlikely for repeated IRI to induce the deterioration of kidney tissues resulting in the progression to CKD. However, it is noteworthy that multiple episodes (more than 2 times) of IRI have not been seriously tested in rodent models. In addition, the duration of ischemia and the interval between injury episodes need to be thoroughly taken into consideration.

NEPHROTOXIC MODELS OF AKI-CKD TRANSITION

The nephrotoxic effect of drugs is an important cause of kidney damage and is commonly used for the study of acute and chronic kidney injury. In this section, we discuss several nephrotoxic drugs that have been used to study AKI-CKD transition in recent years.

Repeated Low-Dose Cisplatin Model of AKI-CKD Transition

Cisplatin is a commonly used chemotherapy drug. However, cisplatin is also recognized for its side effects in normal tissues, especially nephrotoxicity in kidneys, that limit its use and efficacy for cancer treatment. In the past decade, there has been intensive research about the pathogenesis of cisplatin nephrotoxicity, revealing the mechanisms of tubular cell death and inflammation (13, 40, 42, 44). Notably, most of those studies were focused on acute toxicity of cisplatin in kidneys using the models of a single cisplatin injection at a relatively high dosage. However, in clinical settings, cisplatin is often given to cancer patients for several rounds or cycles. While acute nephrotoxicity has been well documented within 1–2 wk of cisplatin chemotherapy, chronic effects of repeated cisplatin treatment may occur. In 2011, we established both xenograft and syngeneic tumor models in mice to test the effects of repeated low-dose cisplatin treatment. In these tumor-bearing animals, weekly injection of 10 mg/kg cisplatin for 3–4 wk attenuated tumor growth and induced obvious renal toxicity (43). In 2016, three studies examined the long-term sequelae of repeated cisplatin treatment at low dosages. Torres et al. (58) gave two doses (15 mg/kg each) of cisplatin to C57BL/6 mice 2 wk apart and then documented the changes in renal function (GFR), fibrosis, histology, and morphology at 9 and 16 wk. They found that the treated mice showed 30% decreases of kidney weight and a persistent decline of ~50% in GFR at the 9-wk time point. Plasma creatinine had a greater than threefold increase at 9 wk and progressed further at 16 wk. What is more, there were myofibroblast proliferation and other CKD features at 9 wk that persisted into 25 wk, but no significant changes in glomeruli during cisplatin induction. Remarkably, “atubular” glomeruli appeared, as a result of degeneration of proximal tubules (58). Sharp et al. (50) compared the effects of intraperitoneal injections of one high dose of cisplatin (25 mg/kg) and repeated low dose of cisplatin (7 or 9 mg/kg once a week for 4 wk) in FVB/n mice. They noticed that inflammatory chemokines and cytokines were highly induced, but levels of cell death were lower in the repeated low-dose model. In addition, the repeated low-dose model had increased levels of fibrotic markers (fibronectin, TGF-β, and α-smooth muscle actin) and interstitial fibrosis after 4 wk (50). Similarly, Noiri et al. (29) administered 10 mg/kg cisplatin intraperitoneally to C57/BL/6 mice once weekly for three times at 0, 1, and 3 wk. This model showed tubule dilatation accompanied by brush-border loss and moderate renal interstitial fibrosis that appeared at 3-wk time point and further deteriorated at 4 wk (29). These findings indicate that repeated low-dose cisplatin injections may induce chronic renal pathologies, providing a feasible model for studying AKI-CKD transition. Indeed, the latest work by Sharp et al. (51) followed up the mice for 6 mo after initial repeated cisplatin treatment. Remarkably, these mice developed some of the characteristics of CKD, including glomerular pathologies and endothelial dysfunction (51).

To sum up, repeated administration of low-dose cisplatin can induce some of the key features of CKD in both tubulo-interstitium and glomerulus, and if observation time is long enough, it may lead to real CKD. Because of the repeated injection of low-dose cisplatin, long period of observation, and variations in laboratory environments, the published studies are for reference only. In our experience, the cisplatin dosage and treatment regimen need to be titrated carefully for each study. To control its stability and efficacy, cisplatin should be freshly prepared and dissolved for injection. Conditions also need to be titrated for different strains and sub-strains of animals. For instance, C57/BL6 mice tend to be more resistant to fibrosis than FVB/n mice (29). In addition, gender and age are also important factors of consideration in experimental design (15, 64). The repeated low-dose cisplatin model has several advantages. Obviously, this model avoids the severe toxicity of the acute model and provides a condition to investigate the long-term sequelae of cisplatin exposure. Importantly, the repeated cisplatin administration in this model is in line with the clinical regimen of cisplatin chemotherapy in cancer patients. In addition, this is a nephrotoxic model of inducing chronic renal pathologies and even CKD, which is complementary to the post-IRI models for studying AKI-CKD transition. Technically, urine and blood samples can be drawn at different time points in this model to monitor renal function to follow the process of progression into CKD.

Repeated DT Model of AKI-CKD Transition

Experimental studies of CKD after acute injury are often limited to current animal models, which often target a variety of renal cell types at the same time, such as epithelial cells, endothelial cells, and inflammatory cells. To specifically determine the role of epithelial injury, a highly selective tubular injury model was established. This model utilizes multiple insults of DT to induce renal tubular lesions and injuries of the proximal tubules, leading to adaptive repair with capillary vessels loss, interstitial fibrosis, and glomerulosclerosis (20, 30). For instance, Grgic et al. (20) used Six2-Cre-LoxP technology to establish a DT-inducible model of tubular injury. In this model, sublethal doses of DT were given three times at a 1 wk interval to induce injury of the S1 and S2 segments of proximal tubular epithelial cells, which resulted in maladaptive repair with interstitial fibrosis and capillary loss. Remarkably, these mice developed glomerulosclerosis, suggesting that repeated injury in renal tubules may directly lead to CKD (20). More recently, Takaori et al. (55) generated a similar model, where repeated administration of DT induced multiple episodes of mild tubular injury in mice that were followed by the development of characteristic features of CKD, including fibrosis, atubular glomeruli, glomerulosclerosis, and reduced erythropoietin production (55). Together, these studies indicate that multiple episodes of tubular damages can result in CKD, providing a novel way to target tubule epithelium to study the mechanisms of AKI-CKD transition.

AA Model of AKI-CKD Transition

Aristolochic acid nephropathy (AAN) is a progressive renal tubulointerstitial nephritis characterized by progressive proximal tubule atrophy and interstitial fibrosis in clinic (10, 11). Due to its characteristic of an early phase of AKI leading to CKD (25), AA may be used to develop a model for the progression of AKI to CKD (66, 67). In this regard, Yang et al. (66, 67) gave one dose (5 mg/kg) of AA to male BALB/c mice aged 8 to 10 wk for 5 days intraperitoneally to examine renal interstitial fibrosis. More recently, Jadot et al. (25) gave 8-wk-old C57Bl/6J male mice 3.5 mg/kg AA daily for 4 days for analysis at 5, 10, and 20 days later. These mice showed significant renal damages and typical histopathological features of AAN, including interstitial cell infiltration and tubulointerstitial fibrosis by morphometric analysis in sirius red staining and quantitative RT-PCR measurements. The results demonstrate that repeated AA administration results in AKI that may progress toward CKD (25).

FA Model of AKI-CKD Transition

FA is a potent nephrotoxic inducer of kidney fibrosis (16, 26, 27, 54). Stallons et al. (54) recently examined mitochondrial homeostasis in FA-induced AKI in mice, which developed renal fibrosis rapidly. In this study, a single dose (250 mg/kg) of FA induced renal expression of fibrosis markers α-smooth muscle actin and TGF-β1 at days 6 and 14. Picro sirius red staining and collagen 1A2 immunohistochemical staining revealed mature collagen deposition as well, further suggesting the development of chronic kidney pathologies (54). Similarly, Jiang et al. (26) showed that a single dose (250 mg/kg) of FA induced massive collagen deposition in renal tubulointerstitium on day 28, accompanied by increased accumulation of fibronectin. More recently, Florin et al. (8) injected male Balb/c mice with a higher dose of FA (375 mg/kg), resulting in kidney fibrosis and increase in col1agen 1 and fibronectin expression at day 14. Collectively, these data indicate that FA induces an early change of AKI leading to chronic kidney pathologies including fibrosis.

CONCLUSIONS

In this review, we have provided a brief overview of the commonly used animal models for studying AKI-CKD transition. As discussed, each of the models has its own merits and pitfalls (Tables 1 and 2). Among the IRI models, we prefer the model of uIRI without nephrectomy, which provides the consistency and reliability and the capacity of long-time observation. The models of bIRI and uIRIx permit the monitoring of renal function at multiple time points, but in our experience, these two models generally have bigger variations and significant animal loss, particularly for long-term study. The nephrotoxic models, especially repeated low-dose cisplatin treatment, are good supplementary to uIRI for understanding the common, basic mechanism of AKI-CKD transition. To this end, it is important to titrate cisplatin doses and treatment regimen to have an optimal condition for inducing chronic renal pathologies. It is also noteworthy that investigation of the long-term sequelae of cisplatin treatment in kidneys is critical to the discovery of novel strategies for reducing the side-effects of chemotherapy in cancer patients.

Table 1.

Published mouse IRI models for studying AKI-CKD transition

Type of IRI/Animals	Ischemic Conditions	Observation Time Points	CKD Changes	Merits	Pitfalls	Reference
bIRI
Male Sprague-Dawley rats	60 min	4, 8, 16, and 40 wk postinjury	Moderate IRI: renal function recovers within 1 wk. Severe IRI: lead to CKD changes	Human pathophysiological modes; easy to monitor	High mortality rate in early phase; variation and inconsistency	(3)
Male Tie2-GFP mice (FVB/NJ background)	20 or 30 min at 37°C	Varied between 1 day and 30 days				(12, 22, 65)
Male BALB/c mice aged 8 to 10 wk	30 min at 37.0°C (moderate IRI) or for 32 min at 37.5°C (severe IRI)	1,7, and 42 days				(67)
Male C57BL/6 mice 22–25 g	20 min (moderate IRI) or 30 min (severe IRI) at 37°C and 38°C	1, 3, and 10 days				(65)
uIRI
Male CD-1 mice	Left kidney ischemia for 30 min at 37°C	1 day, 1 wk, and 3 wk	Tubule atrophy, significant interstitial fibrosis and other CKD changes in IRI kidney	Longer observation time with less animal loss; more reliable	Impossible to monitor renal function at different time points of the same animal	(71)
Male IRAK-M–deficient mice (C57BL/6 background)	Unilateral ischemia for 45 min at 37°C	1, 5, 10, 15, 21, 35, 49, and 70 days				(32, 33)
RMB mice	Unilateral ischemia for 60 min at 37°C	2 days, 2 wk, and 6 wk				(9)
Male BALB/c mice aged 8 to 10 wk	Left kidney ischemia for 30 min at 37°C	1,7, and 42 days				(67)
uIRIx
Male Sprague-Dawley rats	Left kidney ischemia for 60 min at 37–38°C with nephrectomy at 14 days	4 wk	Progressive atrophy and permanent loss of function in injured kidney. (except for moderate IRI model)	Available to measure kidney function	More animal loss and variability	(18, 19)
BALB/c background	Moderate: ischemia for 26 min at 38°C with nephrectomy. Severe: ischemia for 30 min at 38°C with nephrectomy at 8 days	4 wk				(52, 53)
Male BALB/c mice aged 8 to 10 wk	Left kidney ischemia for 30 min at 37°C with right nephrectomy at 3 days	1,7, 42 days				(67)
Multiple IRI
Male BALB/c mice weighing 20–25 g	Initial bilateral or unilateral ischemia for 5,15 or 30 min and a second ischemic insult imposed 8 days later for 30–35 min or 15 days later for 30 min	8 or 15 days after the first surgery	Multiple episodes of tubular injury show severe sustained interstitial fibrosis. But no direct evidence for multiple IRI	Simulate patient situation of multiple episodes of AKI	Multiple IRI models of AKI-CKD transition need to be further examined	(20, 45)
Female Sprague-Dawley rats	40 min of bIRI and either 18 or 48 h later, rechallenged with either 25 or 40 min of bIRI	2 days postsurgery				(69)

AKI-CKD, acute kidney injury-chronic kidney disease; IRI, ischemia reperfusion injury; bIRI, bilateral IRI; uIRI, unilateral IRI; uIRIx, unilateral IRI with contralateral nephrectomy; multiple IRI, multiple episodes of IRI.

Table 2.

Repeated low-dose cisplatin models for studying CKD progression

Mouse Strain	Repeated Dosing Regimen	Main Pathophysiological Changes	Ref.
Athymic nude mice C57BL/6 mice	1) 10 mg/kg CDDP weekly for 4 wk	1) Inhibition of tumor growth by cisplatin treatment	(43)
	2) Four tumor mouse models tested: ovarian tumor xenograft, syngeneic ovarian tumor, testicular tumor xenograft, and breast cancer xenograft models	2) Cisplatin treatment time-dependent increases in BUN, serum creatinine, and renal histopathology
FVB/n mice	1) 7 mg/kg CDDP weekly for 2 wk	1) Dose 2 of CDDP: subtle inflammation	(50, 51)
	2) 7 mg/kg CDDP weekly for 4 wk3) 7 mg/kg CDDP weekly for 4 times, and age for 6 mo	2) Dose 4 of CDDP: no change in kidney function, but a significant increase in inflammation, endothelial dysfunction, and the development of interstitial fibrosis
		3) 6 mo post-CDDP treatment: inflammation, worsened interstitial fibrosis and glomerular pathologies
C57BL/6 mice	15 mg/kg CDDP twice with 2 wk interval, and observation at 9 and 25 wk	1) Tubular degeneration and “atubular” glomeruli	(58)
		2) Reduction of glomerular filtration rate, significant cell turnover, increased apoptosis and endothelial rarefaction at 9 wk
		3) Myofibroblast proliferation, mild fibrosis at 9 wk, and persisted into 25 wk
		4) Macrophage activity increased transiently around 9 wk, normalizing by 25 wk
L-FABP Tg mice (C57BL/6 background)	10 mg/kg CDDP once a week for 3 times at 0, 1, 3 wk, and observation at 4 wk	1) Remarkably increased BUN, SCr, and urinary L-FABP	(29)
		2) Tubule dilatation accompanied by brush-border loss and moderate renal interstitial fibrosis, and fibrotic change with a dose-dependent tendency
C57BL/6 mice and CD4−/− mice (C57BL/6 background)	Weekly 10 mg/kg CDDP for 4 wk; syngeneic lung cancer model	1) Inhibition of tumor growth by cisplatin treatment	(47)
		2) Increases in BUN, serum creatinine, and renal injury during cisplatin treatment
		3) Fibrosis, tubular dilatation, and glomerular collapse.

BUN, blood urea nitrogen; CDDP, cisplatin; Scr, serum creatinine; L-FABP, liver-type fatty acid binding protein.

GRANTS

The work was supported partly by National Natural Science Foundation of China Grants 81720108008 and 81430017, National Institutes of Health, and Department of Veterans Administration. Z. Dong is a recipient of Senior Research Career Scientist of Department of Veterans Administration.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

Y.F. and Z.D. conceived and designed research; Y.F. analyzed data; Z.D. interpreted results of experiments; Y.F. and Z.D. prepared figures; Y.F. drafted manuscript; Y.F., C.T., J.C., G.C., D.Z., and Z.D. edited and revised manuscript; Y.F. and Z.D. approved final version of manuscript.