Porcine models of acute kidney injury

Abstract

Pigs represent a potentially attractive model for medical research. Similar body size and physiological patterns of kidney injury that more closely mimic those described in humans make larger animals attractive for experimentation. Using larger animals, including pigs, to investigate the pathogenesis of acute kidney injury (AKI) also serves as an experimental bridge, narrowing the gap between clinical disease and preclinical discoveries. This article compares the advantages and disadvantages of large versus small AKI animal models and provides a comprehensive overview of the development and application of porcine models of AKI induced by clinically relevant insults, including ischemia-reperfusion, sepsis, and nephrotoxin exposure. The primary focus of this review is to evaluate the use of pigs for AKI studies by current investigators, including areas where more information is needed.

INTRODUCTION

Acute kidney injury (AKI) is a frequent complication of illness, characterized by a dramatic decrease in the glomerular filtration rate (GFR), a major indicator of renal function, with an incidence of 0.21%, occurring in nearly 5% of in-patients (1). The pathogenic factors of AKI are complex, including volume depletion, ischemia-reperfusion (IR) injury, sepsis, trauma, and nephrotoxins (2, 3). Preclinical studies in animal models and clinical trials have been carried out to improve short-term and long-term outcomes of patients with AKI, with limited progress. Treatment for AKI is still largely based on removing or reversing the proximal causes, supportive care for volume and electrolyte management while waiting for renal recovery, and renal replacement therapy if recovery is prolonged or fails to occur. Since there are still no approved specific treatments for AKI beyond treating the underlying cause, a considerable portion of patients with AKI progress to chronic kidney disease (CKD) (4). In-hospital mortality is also high.

A better understanding of the pathogenesis of AKI and its associated complications is essential to reduce the associated morbidity and mortality. Much of the current knowledge of AKI at the molecular level comes from experiments using small animals. However, those experiments often fail to simulate the corresponding clinical conditions, making it difficult to extrapolate information obtained from small animals to humans. Compared with the commonly used laboratory animals (mice, rats, or rabbits), larger animals like pigs possess physiological advantages in size equivalency to humans as well as pathophysiological responses, metabolic process, and immunology characteristics that closely resemble those of humans.

The use of pigs in biomedical study began in the early 1970s (5). In recent years, the application and interest of pigs in all fields of medical research have become increasingly widespread since genetic similarities between Sus scrofa domestica and Homo sapiens were identified (6, 7). Pigs share more similarities in size, anatomic structure, and physiology with human than small animals and act as a principal mammalian model in research into auto- and xenotransplantation, cardiovascular diseases, digestive disorders, respiratory-related diseases, diabetes, urinary system diseases, orthopedic and muscular disorders, dermatology, and neurological sequelae (8–13). Furthermore, pigs have been used for surgical training and practice (14) and preclinical evaluation of medical devices as well as pharmacokinetics, pharmacodynamics, and toxicology studies (15–17).

This article provides a comprehensive overview of up-to-date approaches to the development of porcine models for AKI research and discusses the characteristics of porcine models currently available. Furthermore, we compare the advantages and disadvantages of large and small animal models of AKI and discuss areas where more information is needed to define the role for pigs in research into causes and potential treatments for AKI and disease (18).

CHARACTERISTICS OF EXPERIMENTAL PIGS

The urinary system of swine is anatomically and physiologically similar to that of humans, characterized by multilobular structures as opposed to the unilobular and unipapillary kidneys of rodents and dogs (Fig. 1) (8, 19–21). Porcine kidneys are multirenculate and multipapillate, with a calyceal system similar that in human kidneys. The total number of nephrons in swine kidneys is greater than in human kidneys (1.6−4.6 × 106 vs. 0.2−2.0 × 106), while the number of nephrons in rodents is approximately two orders of magnitude lower than in humans (9−18 × 103 in mice and 13−26 × 103 in rats) (12). The human and pig renal vasculature share many similarities. The porcine renal artery divides into cranial and caudal branches, whereas human renal arteries are divided into anterior and posterior branches. A horizontal branch in the ventral-mid kidney is described in both species. In the inferior pole of the kidney, the caudal branch fell into ventral and dorsal branches in 85% of swine and 62% of human kidneys (20). Porcine physiological function of the urological system and urodynamic parameters are also similar to humans (22). Taking advantage of similarities in the renal vasculature, researchers have designed pig models of renal vascular disease by inducing renal artery stenosis and hypertension that lead to progressive microvascular dysfunction and loss of the cortex and medulla and that correlated with loss of renal function (23). The model demonstrated reduction in renal blood flow, elevated serum creatinine, albuminuria, and increased urinary levels of kidney injury molecule-1 (KIM-1) and neutrophil gelatinase-associated lipocalin (NGAL).

Figure 1.

The kidney architecture and vasculature in the human, miniature pig, and mouse. A: human kidneys are about 10−12 cm in length, with multiple calyxes. B: porcine kidneys are around 8 cm in length and have a similar structure to that of humans with multiple renal papillae and pyramids. C: mouse kidneys are about 1 cm in length, with only a single calyx. Renal arteries are originated from the abdominal aorta and divide into cranial and caudal branches at the hilar level in humans, miniature pigs, and mice. Furthermore, they split intro branches to supply the upper and lower poles and the middle segment of the kidney. Branching patterns of the renal vein in humans, miniature pigs, and mice are similar to those of the renal artery.

Domestic pigs, farm pigs, or large white pigs are frequently used in biomedical studies. However, fully grown adult pigs can weigh up to 250 kg, which limits their general experimental application (20). As a solution, several breeds of miniature pigs have been developed to make better laboratory models. Miniature pigs usually weigh 15−35 kg at 4−6 mo, with kidney size resembling that of the human kidney (20). They reach sexual maturity at 4−5 mo old or so. Different strains are often named after places where they were originally bred, such as the Yucatan, Hanford, Gottingen, Aachen, Sinclair, Bama, Diannan, and Tibet miniature pigs (24–30). Miniature pigs can be maintained at a lower cost than large pigs since they require less housing space, fodder, and drugs in pharmacological studies. The savings can be significant when the drug is under investigation only and manufactured at a small scale. In addition, piglets or porklings have been used in some studies, usually related to pediatrics, neonatology, or developmental biology, including research of renal diseases. In conclusion, there are many different types and strains of pigs that can be used as model animals, among which miniature pigs can be bred exclusively for scientific research since they have extra advantages compared with livestock pigs.

IR-INDUCED AKI MODEL

In humans, kidney damage from ischemic injury can commonly be seen as a result of reduced blood flow for a number of reasons, including structural or functional renal artery stenosis, hypoperfusion as a result of hemorrhagic, hypovolemic (including increased insensible losses from full-thickness burns), septic or cardiogenic shock, prolonged cold ischemic time in kidney transplantation, and during cardiopulmonary bypass in cardiac surgery. Effects of AKI due to hemorrhagic shock (31, 32), AKI produced by full-thickness burn injuries (33, 34), and cardiopulmonary bypass (31) have been studied in large animal models. In human and animal models, the principal feature of ischemia-induced AKI is renal blood flow decrease, resulting in a lower level of oxygen delivery to kidney tissue, ultimately leading to capillary endothelial cell damage, tubular cell detachment, apoptosis, and necrosis (35). IR kidney injury is characterized by inflammation and a sophisticated cascade of intracellular mechanisms leading to AKI in both animal models and humans (36).

Cells cultured under conditions of hypoxia or in medium deprived of glucose can be used as in vitro models of renal ischemic injury. However, few models of renal ischemia in vitro have been previously studied, and in vitro AKI studies cannot represent the global context expected to be observed because of cellular, focal, or systemic inflammatory, immune, and humoral factors involved in the process of ischemia and reperfusion in vivo. Characteristics including kidney development, anatomy, immunological patterns, urinary concentrating mechanism, and susceptibility to ischemic and IR injury indicate that the porcine kidney resembles the human kidney more than kidneys of small animals and can serve as a superior choice to study kidney IR consequences in humans (20).

Since IR is a leading cause of AKI, IR-induced AKI swine models are most frequently used in relevant studies (37–47). The methods used to establish such models are variable. Ischemia in porcine kidneys can be induced by laparoscopic or open surgery and closing the artery only, the artery and vein, or the whole renal pedicle, depending on the study. Bilateral or unilateral ligation, with or without contralateral nephrectomy, remains unstandardized (38, 48, 49). IR modeling can also be performed through interventional procedures, and the renal artery can be temporarily occluded by introducing and inflating a balloon. However, this method requires the use of a contrast agent, and the possible influence of the contrast agent on the kidney must be considered carefully (50). Moreover, the proper ischemic time is still under controversy. Generally, it was accepted by most researchers that the ischemic duration of <30 min would be not enough to cause kidney injury, and maintaining ischemia for over 120 min would probably result in irreversible damage to the kidney and even undesirable animal casualties (51). According to literature reports, renal function in pig models showed significant impairment around 1−3 days after reperfusion and began to recover gradually afterward (39, 51, 52).

Compared with small animals, pigs also have obvious advantages in the operation of IR modeling besides the anatomic and physiological resemblance to human kidneys. In laparoscopic surgery, the same instruments used in human adults or children can be used in small pigs. There is currently a lack of relevant equipment and microinstruments specifically designed for mice. In addition, because the pig has a more macroscopic anatomy structure, the renal artery, renal vein, and ureter can be easily separated during surgery, and subtle anatomic structures like arteriovenous branches can be clearly observed without a surgical microscope or other equipment.

Several examples are given below as well as in Table 1 to illustrate the specific application of the IR-induced AKI porcine model. Barrera-Chimal et al. (53) successfully established an IR model with white male pigs by clamping bilateral renal pedicles for 1 h. The renal function impairment induced from IR was evidenced through a rise in serum creatinine level, a substantial increase in the excretion of kidney injury-associated biomarkers (NGAL and KIM-1) in urine, and a loss of brush-border integrity and tubular dilation from biopsy and histopathological analysis on day 7 after IR (53). They then used this model to investigate the role of vascular mineralocorticoid receptor antagonism in IR-mediated AKI, and the beneficial results provided a rational basis for future clinical trials. Gozdzik et al. (68) used suprarenal aortic cross-clamping for 2 h to induce organ IR injury. This method not only involved the kidney but other distal organs. Similarly, multiorgan IR injury is reported during extracorporeal cardiopulmonary resuscitation (69–71). Although there were no generally accepted standardized procedures for modeling, IR-induced AKI in a porcine model has been widely used and has played an essential role in exploring the mechanisms and possible prevention of or therapeutic strategies for treating AKI.

Table 1.

Summary of porcine AKI-related literature concerning ischemia, infection, and nephrotoxicity

First Author	Country	Publication Year	Journal	Animal	Age, mo	Body Weight, kg	Method	Sample Size	Outcome	Reference
Ischemia-reperfusion
Jayle	France	2006	Kidney International	Large white pigs	3−4	30−35	Clamping left renal pedicle for 45/60 min with a vascular clamp followed by a right nephrectomy	60	Creatinine clearance rate significantly decreased at day 1 and begin to recover afterward	(52)
Barrera-Chimal	France	2017	Journal of the American Society of Nephrology	White male pigs	NA	42−51	Clamping both renal pedicles for 60 min with a nontraumatic vascular clamp	NA	Plasma creatinine significantly increased at day 1 and begin to recover afterward	(53)
Silberstein	United States	2013	BJU International	Yorkshire farm pigs	1.5−2	27−36	Unilateral renal ischemia with left hilar clamping of timed duration (15, 30, and 60 min)	12	Median urine volume was lower (P = 0.04) at 6, 12, 24, and 48 h and median NGAL concentration was higher (P = 0.04) at 12 and 48 h compared with the control group	(37)
Baldwin	United States	2004	Urology	Farm pigs	NA	46−69	Unilateral laparoscopic right nephrectomy and, 2 wk later, left hilar clamp for 0, 30, 60, or 90 min	16	SCr level rose initially on days 2 and 4 in the 60- and 90-min ischemia groups and returned to baseline by day 7	(38)
Bechara	Brazil	2016	Acta Cirurgica Brasileira	Pigs	NA	20	30 min of warm ischemia by arteriovenous/arterial clamping of left kidney vessels	24	Renal warm ischemia of 30 min by arterial clamping did not caused significant glomerular damage, but arteriovenous clamping caused significant glomerular loss in a swine model	(49)
Brunswig-Spickenheier	Germany	2010	Stem Cells and Development	German hybrid pigs	3	29−45	Two balloon catheters were introduced into both renal arteries via the femoral arteries, and both renal arteries were occluded for 110 min by balloon inflation	24	Successful induction of AKI by balloon catheter occlusion of renal arteries and subsequent reperfusion were monitored angiographically as well as hemodynamically	(39)
Amdisen	Denmark	2016	PLoS One	Landrace pigs	NA	42	Unilateral renal ischemia-reperfusion injury was induced by clamping the left renal artery over a 2-h period	20	Unilateral kidney ischemia-reperfusion resulted in an immediate ∼50% GFR reduction, associated with a 4-fold increase in urinary NGAL-excretion	(54)
Gardner	United Kingdom	2016	American Journal of Physiology-Renal Physiology	Pigs	NA	55−65	Bilateral renal artery cross-clamping (40 min) with 48-h postsurgical recovery	35	40 min of bilateral renal artery occlusion led to an increase (48%) in plasma creatinine that peaked at 24 h and an increase in the urinary albumin-to-creatinine ratio that peaked at 2 h	(40)
Malagrino	Brazil	2017	Journal of Proteomics	Pigs	2−3	15−20	A balloon catheter was introduced and inflated for 120 min (ischemia), followed by deflation (reperfusion) for 24 h	5	Only two animals showed levels of SCr above 150% of baseline	(42)
Matejkova	Germany	2013	Intensive Care Medicine	Cross-breed of Rapacz farm pig	13−20	53−85	120 min of aortic occlusion by an aortic balloon catheter	20	Ischemia-reperfusion caused AKI with reduced creatinine clearance, increased fractional Na+ excretion and NGAL levels, and moderate to severe glomerular and tubular damage	(45)
Castellano	Italy	2018	Frontiers in Immunology	Large white pigs	4	20	Clamping the renal artery for 30 min followed by reperfusion	10	Oliguric kidney injury	(46)
Sepsis
Duburcq	France	2018	Shock	Large white pigs	2	18−25	Administration of 5 μg/kg/min Escherichia coli LPS intravenously	10	Endotoxic shock	(55)
Lima	The Netherlands/Italy	2018	Critical Care Medicine	Crossbred Landrace × Yorkshire pigs	2	28	Sepsis was induced through continuous intravenous infusion of LPS at 2 µg/kg/h	15	Increase in creatinine during shock and resuscitation	(56)
Castellano	Italy	2014	Critical Care	Domestic swine	7	58	10 mL of saline solution containing 300 μg/kg LPS was infused	28	Acute tubulointerstitial fibrosis occurred within 9 h from LPS injection	(57)
Vassal	France	2015	Acta Anaesthesiology Scandinavica	Piglets	2.5−3.5	20−25	Intravenous infusion of Pseudomonas aeruginosa	17	A significant decrease in GFR was observed	(58)
Matejovic	Czech Republic	2016	Shock	Domestic Pigs	NA	32− 0	Continuous infusion of Pseudomonas aeruginosa	17	43% reduction in glomerular filtration	(59)
Barth	Germany	2008	Critical Care Medicine	Domestic Pigs	NA	NA	0.5 g/kg of autologous feces were suspended in 100 mL saline and cultivated at 38°C until the Matijevic induction of peritonitis	20	Septic shock	(60)
Merz	Germany	2018	Intensive Care Medicine Experimental	Pigs	15−30	69	1 g/kg of autologous feces was collected during premedication, dissolved in 500 mL of 0.9 % saline, incubated at 38 °C for 12 h, and then infused into the abdominal cavity	13	Septic shock	(61)
Wang	China	2015	International Urology and Nephrology	Hybridized pigs	NA	23.5−27.9	Cecal ligation and puncture and clamping of the renal artery	17	Significantly higher renal function marker creatinine and blood urea nitrogen 24 h after surgery	(62)
Nephrotoxicity
Cui	China	2019	Shock	Chinese experimental mini-pigs	3	8−10	80 mg/kg gentamicin sulfate intramuscularly daily for 10 days	18	AKI was well developed after 10 days of drug infusion	(63)
Robbins	United Kingdom/United States	1992	Cancer Chemotherapy Pharmacology	Large white pigs	10	NA	Single intravenous infusion of 1.5, 2, or 2.5 mg/kg (i.e., 90, 120, and 150 mg/m2) cisplatin	12	Significant cisplatin-induced reduction in GFR over the 24-wk period following drug treatment	(64)
Santiago	Spain	2016	PLoS One	Maryland pigs	2−3	9	Infused cisplatin to piglets through the peripheral vein in the Plow ear at doses of 2, 3, and 5 mg/kg	41	SCr significantly rose 48 h after cisplatin administration and largely fell back to the predose level in a week	(65)
Wu	China	2019	Biochemical and Biophysical Research Communications	Miniature pigs	3	10.4 ± 0.23	Injection of 25 mL/kg iohexol (350 mg iodine/mL) after dehydration	16	Increased blood levels of blood urea nitrogen, SCr, serum and urinary β-microglobulin, and injured kidney tissue, dilated tubules and foamy degeneration	(66)
Xu	China	2020	International Immunopharmacology	Miniature pig	3	NA	Iohexol (350 mg/mL) was injected via the external jugular vein at 25 mL/kg	10	Increased serum levels of blood urea nitrogen, SCr, and β-microglobulin and urinary levels of β-microglobulin and renal fibrosis and injury scores	(67)

AKI, acute kidney injury; GFR, glomerular filtration rate; LPS, lipopolysaccharide; NA, not applicable; NGAL, neutrophil gelatinase-associated lipocalin; SCr, serum creatinine.

IR injury is an inevitable event and important cause of graft failure in kidney transplantation (72). Appropriate approaches to prevent and reduce IR damage in transplanted kidneys have not been fully elucidated. The ischemic phase duration, storage temperature, perfusion solution components, and preservation medium can all exert effects on ex vivo kidneys (73, 74). Large animal experiments or ex vivo experiments based on large animals will contribute to the progress in preclinical renal transplantation research and thereby possess potential to improve the prognosis of kidney recipients.

SEPSIS-ASSOCIATED AKI MODELS

As one of the most common causes of mortality in intensive care units, sepsis is a state of whole body inflammation caused by infection affecting around 20−50% of critical patients, with AKI a common complication (75). Gram-negative bacteria and the components of their cytoderm, in particular, lipopolysaccharide (LPS), play significant roles in the pathogenesis of sepsis-associated AKI (76). The pathophysiology characteristics of septic AKI include intrarenal hemodynamic alterations, capillary endothelial cell dysfunction, renal interstitial inflammatory cell infiltration, intraglomerular thrombosis, and tubular obstruction caused by exfoliated dead tubular cells or necrotic- and apoptotic-derived cellular debris (57). The large population of patients with sepsis and AKI has led to a strong need for further research, and studies concerning sepsis-associated AKI in porcine models have been carried in multiple medical subspecialties (55, 57, 59, 60, 77–81). Sepsis models developed using small animals showed mixed results when translated into clinical trials or practice (82). The response of mice to inflammatory stressors like endotoxins is quite different from human beings (83). In comparison, pigs reacted to infection in a manner that is similar to humans in terms of cytokine and immune cell profiles (84) and presented symptoms characteristic of human infections (85). As such, pigs appear to be an appropriate animal model for sepsis-induced AKI.

Several infectious options are available for the study of septic AKI in pigs. Commercialized Escherichia coli LPS can be administered intravenously (55–57, 79, 80, 86). In several studies, live E. coli or Pseudomonas aeruginosa was directly injected to pigs intravenously (59, 80, 87–89). These approaches are highly practicable, and it is easy to control the dose of endotoxin or bacteria, but there is a certain gap with the actual situation encountered in clinical practice. Another sepsis-related model of AKI in pigs is to induce peritonitis and subsequent multiorgan damage, including kidney injury. Autologous feces have introduced through drains into the pig abdomen after suspension in saline, with or without in vitro culturing first (58, 60, 61, 78, 79, 81, 90–92). Surgically, cecal ligation and puncture (CLP) results in abdominal infection, and then fecal peritonitis could induce sepsis and AKI (62, 93). The mortality rate during CLP modeling is relatively high. Many animals die of acute lung failure, and the degree of kidney damage is difficult to manage. Nevertheless, this model mimics changes in humans with digestive tract perforations and has the potential for clinical translational value.

For an E. coli peritoneal injection porcine model, without intervention, death usually occurred at 7.6 ± 0.5 h (89). AKI was demonstrated by a 30% rise in serum creatinine concentrations and subtle histological changes on renal biopsy 18−22 h after the administration of bacteria or feces in some reports (but not all studies) (81, 94). Castellano et al. (57) analyzed renal fibrosis with Masson trichrome staining and demonstrated extensive collagen deposition surrounding peritubular capillaries after LPS administration. Diffuse glomerular thrombi and tubular vacuolization were also observed via hematoxylin and eosin staining (57). Moreover, serum KIM-1 and cystatin C levels significantly increased 6 h after LPS infusion before serum creatinine began to rise (57). In addition, inflammatory factors are key indicators in the evaluation of infection-induced AKI models. IL-6 and TNF-α levels significantly elevated 12−18 h after LPS infusion (80).

Clinically, septic AKI affects ∼40−50% of all patients with sepsis (80). Therefore, inducement of sepsis does not mean the successful development of septic AKI. This is a concern for the establishment of septic AKI pig models. For example, in 2011, Benes et al. (78) used two procedures to develop septic AKI: either by peritonitis (n = 13) induced by injecting autologous feces into the abdominal cavity after in vitro culture or through intravenous administration of live P. aeruginosa (n = 15). The results showed that AKI developed only in 14 pigs (50%), 62% in peritonitis, and 40% in the bacteria infusion model.

Another concern regarding porcine septic AKI animal models is the complex and heterogeneous nature of sepsis (61, 95). Simon et al. (79) developed a septic shock pig model caused by peritonitis induced through intraperitoneal infusion of 1.0 g/kg autologous excrement after incubation with 100 mL of 0.9% saline for 12 h at 38°C. According to their report, liver, kidney, and heart dysfunction were all observed during resuscitation following septic shock, which can evolve out of systemic inflammatory response syndrome or multiple organ dysfunction syndrome. Thus septic AKI models were not perfectly applicable to studies aimed at exploring mechanisms or interventions to treat AKI alone.

Taken together, the above considerations pose challenges for applying specific models of septic AKI models to the pig. Still, the model holds considerable potential for studying septic AKI because of multiple choices of methods and simulation of the pathophysiological process of the human body.

NEPHROTOXIC AKI MODEL

AKI induced by known nephrotoxic drugs also occurs frequently in hospitalized patients. Nephrotoxin-induced AKI is a common complication of diagnostic and therapeutic procedures and is becoming an important cause of AKI both in China and worldwide. In studies of rodent models of AKI, renal injury has been induced by contrast media, vancomycin, aristolochic acid, and folic acid (96). In large animal research, the most commonly used modeling drugs are cisplatin, gentamicin, and contrast medium (63, 64, 66, 67, 97).

Cisplatin-Induced AKI

Cisplatin is a cytotoxic platinum derivative and has been widely used as potent chemotherapeutic agent for a variety of tumors, usually in patients with solid tumors, including nonsmall cell lung cancer, gastric tumor, and prostate cancer (98). A notable side effect of cisplatin is nephrotoxicity, which induces DNA damage, mitochondrial disorders, oxidative stress, and apoptosis in kidneys. Freely filtered at glomeruli and reabsorbed by tubular epithelial cells, cisplatin dominantly affects tubular areas of the kidney. Histopathological results show signs of a cisplatin dose-dependent increase in renal injury (65). A single administration of cisplatin at high dose could lead to AKI in animal models, while repeated low-dose injections of cisplatin might result in renal fibrosis and CKD (99).

In an effort to clarify the mechanism of cisplatin-induced renal injury through large animal model studies, Santiago et al. (65) infused cisplatin into piglets through a peripheral vein in the ear at doses of 2, 3, and 5 mg/kg. According to their results, 3 mg/kg cisplatin injection led to a significant rise in plasma creatinine and urea levels without an evident increase in plasma K⁺ concentration. Serum creatinine significantly rose 48 h after cisplatin administration and fell back to the predose level in a week (65). However, the data of an earlier study in 1992 presented with different results. Robbins et al. (64) developed a large white pig AKI model by a single intravenous administration of cisplatin at doses including 1.5, 2, and 2.5 mg/kg (i.e., 90, 120, and 150 mg/m²) and concluded that GFR significantly decreased during 8−12 wk after dosing. Histopathology lesions were observed in the corticomedullary junction region and extended into the medullary area, as far as calyces and the renal pelvis in some kidneys. Most histological changes were found in the inner one-half to one-third of the kidney cortex. Occasionally, corresponding renal tubules were moderately compressed. Neither severe glomerular abnormalities nor tubular epithelial degeneration was observed in this study. Considering a lack of documented reports and the differences in pig breeds and age, it is difficult to derive precise dose or time data on cisplatin-induced AKI. This also means that more research is needed to clarify the key information such as the dose required for cisplatin modeling, timing of the injury after administration of the drug, and how severe the renal damage will be.

The clinical significance of this model lies in formulating renoprotective strategies for cancer patients who receive chemotherapy. Cisplatin-induced AKI presents with relatively shorter immunosuppressive effect on animals, so this model is particularly suitable for studies that require long-term observation of animals, such as observation of long-term prognosis or potential delayed adverse events caused by drugs or therapies. Cisplatin-induced AKI can also be used for small animals and in cell cultures, allowing researchers to conduct multilevel studies. The porcine model of cisplatin-induced AKI is worthy of further exploration. Researchers have used a cisplatin-induced AKI model in miniature pigs to examine the effects of a probiotic cocktail on preventing renal injury, hypothesizing that the probiotic mix reduced inflammation and apoptosis and thereby slowed the progression of CKD (100).

Gentamicin-Induced AKI

Gentamicin belongs to aminoglycosides family, a category of antibiotics used clinically to treat serious infections caused by gram-negative bacteria. Nephrotoxicity mediated by aminoglycosides affects 10−20% of patients under associated treatment regimes. Gentamicin accumulates in renal tubular epithelial cells, leading to nephrotoxicity and AKI. The pathogenesis of gentamicin-mediated kidney injury is not totally clear. Studies have indicated that gentamicin may act on mitochondria and result in oxidative stress and apoptosis as cisplatin does. Ultimately, it can causes congestion of glomeruli, generation of free radicals in the kidney, impairment of renal antioxidant defense mechanisms, and acute tubular necrosis (5).

Cui et al. (63, 101) developed a gentamicin-associated AKI model using Chinese experimental miniature pigs. Pigs received intramuscular administration of gentamicin for up to 10 days at doses of 5, 6.85, 10, 15, 20, 40, 60, 80, 100, and 120 mg/kg. Marked dose-dependent effects were observed in survival rates, histopathology scores, renal function, and molecular biomarkers. The researchers established a protocol for inducing AKI by intramuscularly infusing 80 mg/kg gentamicin sulfate daily for 10 days. Histopathological findings indicated that gentamicin caused tubular epithelial cell degeneration, granular or protein cast tubular lumen obstruction, and inflammatory cell infiltration into the renal parenchyma.

Contrast-Induced AKI

With the development of imaging technology, many patients suffered from kidney damage induced by contrast medium (102). In a meta-analysis enrolling 29 studies, the incidence of contrast-induced AKI was reported to be 4.4−22.1% (103). Commonly used contrast agents included sodium iothalamate, iopamidol, and iodixanolare (102). These contrast agents can induce renal tubular cell apoptosis through the mechanisms involved in cell membrane damage, increased intracellular Ca²⁺, an activated proapoptotic unfolded protein response, and decreased ATP (102). Repeated administration of contrast medium also induced a transient increase in the renal resistance index in both pig models and clinical study (104, 105).

Porcine model of contrast-induced AKI model can be established by injecting a single dose of iohexol (350 mg/mL) at 25 mL/kg. This model highly mimics clinical contrast-induced AKI and is reproducible. After AKI, increased blood levels of blood urea nitrogen, serum creatinine, and β-microglobulin and urinary levels of β-microglobulin were observed; pathological changes included dilated tubules, foamy degeneration, and renal fibrosis (66, 67). Administration of retinoic acid and miR-30c can attenuate kidney injury in this model, respectively, providing new ideas for the prevention and treatment of contrast-induced AKI.

OTHER PIG MODELS

Beside IR, sepsis, and drug toxicity, other porcine models of AKI such as radiation-induced kidney damage exist (106). However, similar to septic AKI, most of those models produced not only kidney injury but damage to other organs, including acute liver injury, lung atelectasis, and intestinal damage, making it difficult to limit multiorgan systemic damage while trying to study renal injury alone. Furthermore, porcine kidneys were often used for ex vivo research (107), abiding by three “R” principles (the consensus to reduce the number of animals used, refine their life conditions, and replace animal models by in vitro or other procedure if possible) of animal welfare, and pigs have considerable application prospects in the field of transplantation (52, 108).

Clinically significant AKI is usually unpredictable and sometimes caused by multiple insults (ischemia, infection, nephrotoxicity, etc.) and their cross talk. For example, we can study the mechanism of renal IR injury in a diabetic model or explore a solution for cisplatin-mediated AKI in a tumor-bearing pig model. Compared with single insult-induced AKI models, “multihit” models share more similarities to actual clinical situations and show more translational medicine significance (94, 109).

CKD is an undesirable outcome of AKI, but the AKI-CKD transition and its underlying mechanism remain largely unclear. Rodent models for exploring post-AKI development of chronic renal lesions and the progression mechanism to CKD have been reviewed by Fu et al. (110). Recently, Lui et al. (111) established a porcine model of CKD by partial nephrectomy and radical contralateral nephrectomy (2/3, 3/4, or 5/6 nephrectomy) using laparoscopy, and Chade et al. (23) created a CKD porcine model by inducing bilateral renal artery stenosis. Similarly, Lerman and colleagues (112–114) developed a pig model with combination of unilateral renal artery stenosis and high-cholesterol/carbohydrate diet. The interplay between hyperlipidemia and CKD was discussed in detail in these studies; it seems that combined application of both hypercholesterolemia and renal artery stenosis could amplify, upregulate, and accelerate the cellular mechanisms leading to renal vascular, glomerular, and tubulointerstitial damage. In addition to metabolic similarities, the swine is an omnivorous animal with similar nutritional requirements to humans, so diet and regime can be designed exactly according to the needs of human (115–118). Thus porcine models are particularly suitable for research of nutrition- or metabolism-related disease, including CKD.

The pig is also the optimal animal for kidney transplant research and surgeon training, especially laparoscopic or robotic surgical training (50). For example, Kobayashi et al. (119) achieved orthotopic renal transplantation using pigs weighing 20–30 kg, and their methodology was reported in detail. The surgical method used in pigs can provide a direct reference for clinical settings. Swine can be used to explore the way to improve the survival rate of grafts and solve any problems that may arise during or after kidney transplantation (120). Pigs even possess the potential as organ donors for xenotransplantation, offering a new option for patients who need renal replacement. Studies are ongoing to overcome xenogeneic rejection through creation of genetically manipulated humanized swine (121). The advantages of porcine models in this aspect are beyond the scope of small animal models, indicating a great value of pigs.

Indeed, pig kidney xenotransplantation is not that far off, with some expecting clinical trials in the next couple of years (13). The authors of a recent review in the Journal of the American Society of Nephrology suggest that it would be ethical to offer a pig kidney transplant to a patient whose life expectancy was shorter than their projected waiting time on the list for a deceased donor (13). An important part of that process is to engineer pigs genetically to remove potential xenoantigens using technologies like CRISPR/Cas9 (13). Researchers have used flow cytometry to investigate whether patients on dialysis with high degrees of human leukocyte antigen sensitization had high cross-reactivity to pig antigens. They found no IgM or IgG binding to galactose-α-1,3 galactose or N-glycolylneuraminic acid in subjects who were highly sensitized [calculated panel reactive antibodies (cPRA) = 100%] versus those who were not (cPRA = 0%) but were significantly reduced compared with normal controls (122).

Miniature pigs have also been used in preclinical models of organ preservation (123) and prevention of rejection in marginal kidneys harvested in donation after cardiovascular death (123).

ADVANCES AND LIMITATIONS OF PIG MODELS IN AKI RESEARCH

The widespread use of laboratory animals has greatly contributed to the development of biology and medical science. In nephrology, a variety of mouse and rat AKI models are used to explore the mechanism of renal injury and potentially develop novel preventative, diagnostic, or therapeutic approaches for AKI. Currently, some models are only studied in small animals, such as warfarin-related AKI (124) and rhabdomyolysis-induced AKI (125). While the use of animals in experiments remains controversial and subject to many legal and ethical restrictions, large animal models are still frequently used when effective and safe preclinical protocols directly transferable to clinic are expected. Research using pigs has played a significant role in the assessment of new medical devices or efficacy of a pharmacological therapeutic regime before their entry into clinical trials, market, and application (22). Big physiological differences between small animals and humans limit the clinical translation significance of research based on rodents, particularly when a given model has been extensively carried out in small animals only. In that case, the United States Food and Drug Administration recommends that a therapeutic regime be tested in more than one animal model before an investigative new drug application is submitted. Thus at least one of the experimental models should be a large animal model (126). The ethical policy of porcine experimentation differs among countries and regions, and it is usually more complicated than rodents. At times, the ethics of pig models can reach a level approximate to that for human research, including the requirement that studies on pigs should be based on prior in vitro or even small animal experimentation (127).

To choose an optimal laboratory animal offering structural and functional similarities to humans, other important factors like size, manipulability, information transferability, economic efficiency, ecological consideration, ethical issues, and social impact should also be evaluated (128). Pigs possess many advantages over rodents (Table 2). A major advantage is the similarity in renal structural features between human and pig kidneys. While the structure of the pig kidney does not correspond completely to that of the human kidney, the analogy between porcine and human intrarenal vascular systems as well as relationships between intrarenal arteries and renal collecting systems is close, as previously described (20). The pig body size is also approximate that of humans, especially miniature pigs, whose body weight reaches 60−70 kg in adulthood. The importance of animal size in experimental or preclinical research is usually underestimated. Humans and rodents differ from each other in total number of cells, which matters in some experiments (127). Due to the similarity in body size, surgical techniques used in pig models can be directly transferred to clinical settings, including instruments and devices, operating methods, etc. (129, 130). AKI is a common complication of surgical patients, and surgical operations are required sometimes for AKI modeling, such as IR modeling. It is easier to control the amount of blood loss and to detect cardiac and hemodynamic parameters (i.e., electrocardiograph, blood pressure, etc.) in swine than in small animals. Despite the larger size, pigs are also docile animals and, when treated properly, are often gentle and tame (131).

Table 2.

Summary of advantages and limitations of the mouse and pig

Advantages	Limitations
Mouse
Small size: single-handed operation and relatively small housing space requirements	Small size: body weight is 3 orders of magnitude different from humans
Low food intake: low feeding cost	Anatomic characteristics, physiological mechanisms, genetic background, and immune response patterns are quite different from those of the human body
Complete data: gene sequence, proteomics, and transcriptomics information available	Research results cannot be directly translated into clinical applications
Applicable reagents, test kits, antibodies, and instruments available	Sometimes unable to obtain sufficient blood or tissue samples
Strong fertility and short reproductive cycle	Repeated blood sampling is usually not allowed
	Different scales for blood biochemical parameters from humans
Pig
Large size: body weight and kidney size proximity to humans	Large size: complicated handling and large living space required
Anatomic characteristics, physiological features, genetic background, and immune response patterns are similar to those of the human body	High food intake: high feeding cost
Research results can be directly translated into clinical applications	Incomplete data: lack of full gene sequence, proteomics, and transcriptomics information
Sufficient amount of blood and tissue material in a lower total number of animals	Lack of applicable reagents, test kits, antibodies, and equipment
Repeated blood and tissues sampling	Relatively long breeding cycle
Comparable blood biochemical parameters and similar urodynamic parameters with humans	Ethical complexity

Repeated blood sampling is another advantage for porcine models of AKI. Since pigs have large total blood volumes, this makes it possible for researchers to follow their renal function and other blood biochemical indexes in long-term studies. Moreover, renal biopsy can be performed on pig models to assess the longitudinal changes in histopathology and examine the expression of injury markers after AKI by repeated sampling of kidney tissue. It is also more convenient to collect 24-h urine specimens from pigs in metabolic cages or by catheters to monitor urinary parameters after renal injury (20). The ability to provide sufficient amounts of blood, urine, and kidney tissue or other biological material from relative fewer animals conforms to the three “R” principle. The human-resembling physiology patterns of swine assure a high relevance of the information obtained in studies using this species for clinical trials. The reference values, normal range, and threshold for many blood biochemical variables, such as serum creatinine and blood urea nitrogen levels, are almost identical in pigs and humans (132). Hannon et al. (132) analyzed over 100 porcine physiological parameters under baseline conditions and concluded that most variables were comparable to those collected under corresponding conditions in humans.

The inconsistency of innate and adaptive immunity between the human and mouse raises concerns about the relevance of rodents as preclinical AKI models (133) and calls for better models. Interestingly, the number of orthologous immune gene families between the pig and human is 10 times more than between the mouse and human (134). Pigs have a large proportion of immune cells in lymphoid tissues and similar distribution of their vascular-associated lymphoid tissues as do in humans (135). The Toll-like receptor expression profile is also close between swine and humans (136). Moreover, the literature on immune markers reported an evident analogy between the human and pig in antigen pattern and the immune response (137). Since the epitopes of T cells and B cells were comparable between human and pig, de León and colleagues (138) recently initiated a project to develop vaccine by using Landrace and large white cross-bred female pigs.

Genetics influence nearly every aspect of human biology and disease including AKI pathogenesis. The pig has 18 pairs of autosomes and 2 allosomes, totaling around 2.7 Gb, ∼7% smaller than the human genome. In contrast, mouse genomes are 14% smaller than humans. The overall similarity of the transcribed genome between the pig and human is nearly 85% (139). With the rapid development of gene editing technology, a great deal of genetically engineered pig models has been established by Cre-loxP, Tet-on/off, or CRISPR/Cas systems, as reviewed by Yum et al. (140). These transgene models not only mimic human diseases in symptoms but also achieve consistency in etiology.

Although the pig has many advantages for creating an AKI model, it has certain limitations. While pig anatomy and physiology show some correlation to those of humans, distinctions between the two species exist that may be associated with different responses to certain experimental protocols (12). How pigs are handled and how experiments are carried out can lead to certain variability. In addition, the immune response and disease susceptibility are slightly different among strains and even from one pig to another in a same breed. General attributes like breed, sex, age, and body weight are also reported to have an effect on results of certain tests (141). Experimental mice are mostly inbred strains, and pigs are usually closed colonies, so the individual variability of pigs is greater than that of mice. In clinical settings, the manifestation of patients varies from individual to individual and population to population. In this respect, pigs and humans are similar. As long as one controls the variables (strain, age, sex, etc.) well, one can still evaluate differences between experimental groups and obtain reliable data from swine.

One of the most important limitations of pig AKI models lies in a lack of elaborated data at the molecular level, including but not limited to accurate DNA sequencing, gene expression patterns, transcriptome features, and proteome composition. Applicability of protein-based techniques in pigs is limited since they have been developed at a relatively lower rate in contrast to humans or rodents. This limitation is a results of an imperfect porcine peptide mass database; using homologous sequences from databases of other species (such as mice) to carry out swine-based research is extremely inefficient since sufficiently identical or high-scored peptide matches are rare (8). The systematic creation and elaboration of porcine proteome database for different systems and organs are still underway (6, 8). With increased application of pig models, those issues are expected to be solved over time. In the past few decades, researchers have bred miniature pigs to settle the inconvenience of handling large livestock pigs. Aware of the drawbacks from lack of a swine database, researchers have initiated many genomic, transcriptomic, and proteome studies in pigs (142, 143). For example, the genome of Sus scrofa was sequenced for the first time in 2012, and data are continually being updated (144). Several microarray studies were also conducted in pigs (145, 146). Manufacturers nowadays can offer a variety of pig-specific products, including but not limited to cell lines, research agents, primary antibodies, and assay kits (115).

Another inconvenience may come from the high growth rate and subsequent large size of pigs. The gestation period for female pigs is around 114 days, with 2.4 litters a year during reproductive ages and 12 piglets per litter, which means large kennels are required (20). Procedures need to be performed at a time while pigs are still at a weight that allows them to fit into laboratories and associated equipment. Animals are therefore often piglets or young adults rather than mature adults. Moreover, the diet and experimental kits for pigs are more expensive than for rodents (20). It is generally believed that rodents are the most economical models since they present with short reproductive cycles, are easy to handle by research staff, and cost less to maintain. Large animals like pigs seem to be less cost effective and more complicated to manage but in the long run often prove to be cost efficient and worthwhile by providing more reliable, transferable, and solid data (8). Furthermore, total cost and number of euthanized pigs required can be reduced by using tissue materials from slaughtered big animals for ex vivo or in vitro studies. Samples from slaughtered large animals represent an abundant and inexpensive resource of viable cell material for applied renal study (147). This makes swine a convenient, reliable, and cost-efficient model system in AKI research.

SUMMARY

The application of pigs in AKI models is increasingly gaining traction because of anatomic, metabolic, and pathophysiological similarities between pigs and humans. Different types of porcine AKI models have been developed by researchers over the past decades. We believe that even if it is impossible for swine to completely replace rodents as a firstline choice, porcine AKI models can be used as complementary or alternative models to confirm the results obtained from small animal models. These modeling approaches of pigs provide a better understanding of the mechanisms of AKI and may eventually reveal new molecular targets for the prevention, early diagnosis, and effective treatment of AKI. Obstacles remain to be overcome in order to achieve the full potential of using pigs to study human disease and translate findings into clinical trials for potential therapies. Finally, as xenotransplantation moves to the clinical from preclinical arena, knowledge about AKI, acute kidney disease AKD, and CKD in the pig kidney will become crucial.

GRANTS

This work was supported by National Natural Science Foundation of China Grants 81830021 and 82070700 (to S.Z.), National Key R&D Program of China Grant 2018YFA0108802 (to S.Z.), and the National Institute of Diabetes and Digestive and Kidney Diseases Grant 1R01-DK-113256-01A1 (to S.Z).

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

J.H. and G.B. drafted manuscript; S.Z. edited and revised manuscript; J.H., G.B. and S.Z. approved final version of manuscript.