We describe how to generate a widely used surgical model of intestinal ischemia-reperfusion injury (IRI) in rodents. The procedure involves occlusion of the superior mesenteric artery followed by the restoration of blood flow. This model is useful for studies investigating occlusive causes of intestinal IRI in both veterinary and human medicine.
Intestinal ischemia-reperfusion injury (IRI) is associated with a myriad of conditions in both veterinary and human medicine. Intestinal IRI conditions, such as gastric dilatation volvulus (GDV), mesenteric torsion, and colic, are observed in animals such as dogs and horses. An initial interruption of blood flow causes tissues to become ischemic. Although necessary to salvage viable tissue, subsequent reperfusion can induce further injury. The main mechanism responsible for IRI is free radical formation upon reperfusion and reintroduction of oxygen into damaged tissue, but there are many other components involved. The resulting local and systemic effects often impart a poor prognosis.
Intestinal IRI has been the subject of extensive research over the past 50 years. An in vivo rodent model in which the base of the superior mesenteric artery (SMA) is temporarily ligated is currently the most common method used to study intestinal IRI. Here, we describe a model of intestinal IRI utilizing isoflurane anesthesia in 21% O2 medical air that yields reproducible injury, as demonstrated by consistent histopathology of the small intestines. Tissue injury was also assessed in the colon, liver, and kidneys.
Ischemia-reperfusion injury (IRI) can occur in any organ and involves two sequential components. An initial cessation of blood flow causes affected tissues to become ischemic and then subsequent reperfusion induces further cell injury. Damage from the reperfusion often exceeds that caused by ischemia1. The pathophysiology of IRI involves a complex cascade of events, the most notable of which is free radical formation upon the reintroduction of oxygen, which occurs during reperfusion2. Activation of the inflammatory cells and cytokines also plays a role2. In cases of intestinal IRI, bacterial translocation into the bloodstream following endothelial damage can lead to systemic inflammatory response syndrome2. If the damage due to IRI is severe enough, resultant systemic effects can lead to multi-organ failure3.
Cases of intestinal IRI are associated with high morbidity and mortality4,5,6. Intestinal IRI is associated with many pathologic conditions and surgical procedures in both veterinary and human medicine. In veterinary medicine, animals are especially prone to intestinal IRI conditions, such as gastric dilatation volvulus (GDV), mesenteric torsion, and colic7,8. In humans, IRI is a major and frequently occurring problem in abdominal aortic aneurysm surgery, strangulated hernias, acute mesenteric ischemia, volvulus, trauma, shock, neonatal necrotizing enterocolitis, and small bowel resection or transplantation9.
Most in vivo rodent studies of intestinal IRI involve occlusion of the base of the superior mesenteric artery (SMA), the branch of the abdominal aorta that supplies blood to the majority of the small intestines and the proximal portion of the large intestines10,11,12. Despite this model's widespread use and relative simplicity, a detailed protocol using inhalant anesthesia in 21% O2 medical air has not been published. The lack of a standard protocol poses difficulty for researchers who are unfamiliar with the procedure and prevents consistency across studies. We demonstrate the steps necessary to conduct the surgical model of intestinal IRI in 8-14-week-old male and female Swiss Webster mice. This model of intestinal IRI yields reproducible injury, as demonstrated by consistent histopathology.
The procedures described here were approved by the National Heart, Lung, and Blood Institute Animal Care and Use Committee at the National Institutes of Health and conform to the policies outlined in The Public Health Service Policy on Humane Care and Use of Laboratory Animals, The Animal Welfare Act, and the Guide for the Care and Use of Laboratory Animals.
1. Surgical setup
2. Animal preparation
3. Surgery and ischemia
4. Recovery and reperfusion
5. Euthanasia and blood collection
6. Tissue processing for histology
7. Tissue scoring
We demonstrated a model of intestinal IRI in mice that yielded consistent and reproducible results. The small intestine, proximal colon, kidneys, and liver were sectioned and stained with H&E. A veterinary pathologist graded tissue injury using the scoring systems previously mentioned (Table 1). Statistical analysis was performed using single factor analysis of variance (ANOVA) followed by Tukey's post hoc with pairwise comparisons, which determined whether or not there was a significant difference in the data within and across groups. A p-value less than or equal to 0.05 was considered the cutoff for establishing statistical significance. All statistical tests and graphing were carried out in a spreadsheet software (e.g., Microsoft Excel) with the Real Statistics Resource Pack add-on. Data are presented as the mean ± standard error of the mean (SEM).
Microscopic lesion scores of the three small intestinal segments (duodenum, jejunum, and ileum) were significantly increased for animals undergoing intestinal ischemia-reperfusion injury (IRI; N = 7) versus those that underwent sham laparotomy (Sham; N = 6) (Figure 2 and Figure 3). The standard error for these data was narrow, demonstrating consistency of the results within and across groups. Each intestinal segment in the Sham group yielded the exact same average Park/Chiu score of 0.83. The SEM for the duodenum, jejunum, and ileum in the Sham group was 0.31, 0.40, and 0.31, respectively. The average Park/Chiu scores for the duodenum, jejunum, and ileum in the IRI group were 4.07 ± 0.44, 4.14 ± 0.46, and 5.14 ± 0.40, respectively.
In this study, 50% (3/6) of the initial mice that underwent 60 min ischemia and 120 min reperfusion (60/120 group) died. Two of the three mice were submitted for necropsy. Both mice had epithelial necrosis, congestion, and hemorrhage of the small intestine. In addition, the mice had lymphocytolysis, a nonspecific change associated with physiologic stress. Neither mouse had lesions in the heart, lung, liver, or kidneys. Shortening the times to 45 min ischemia and 90 min reperfusion and adding in 400 IU/kg heparin (45/90/H group) lowered the mortality to 20% (1/5) without changing the intestinal injury scores (Figure 4). The mean Park/Chiu score for the 60/120 group was 4.56 ± 0.38 (N = 3), and the mean score for the 45/90/H group was 4.375 ± 0.38 (N = 4).
Microscopic findings indicative of injury in the proximal colon, liver, and kidney were not seen in either the 60/120 mice or the 45/90/H mice.
Table 1: Scoring systems for the intestines, kidneys, and liver. Intestinal damage was graded using the Chiu/Park system17. Kidney damage was graded using the Jablonski scoring system18,19. Liver damage was graded using the Suzuki scoring system20,21. This table is adapted with permissions from scoring systems presented in Quaedackers et al.17, Du et al.19, and Behrends et al.21. Please click here to download this Table.
Figure 1: Location and isolation of the superior mesenteric artery (SMA). (A) Normally, the SMA lies ventral to the inferior vena cava and extends toward the animal's right. It is situated between the celiac artery and the renal artery. This figure is adapted with permission from The Anatomy of the Laboratory Mouse by Margaret Cook (1965)22. (B) In this procedure, the intestines are exteriorized and flipped to the left (covered with moistened gauze in this picture), so the SMA (yellow arrow) lies to the left of the inferior vena cava (blue arrow). Abbreviations: RK = right kidney; D = duodenum. Please click here to view a larger version of this figure.
Figure 2: Small intestinal segments stained with hematoxylin and eosin. Sections of jejunum (A) and ileum (B) from mice in the Sham group featured villi that were long and thin without distortion. Sections of jejunum (C) and ileum (D) from mice in the IRI group featured areas of necrosis (asterisks) and hemorrhage with blunting and distortion of the remaining villi (arrows). The photos are from mice that underwent 45 min ischemia and 90 min reperfusion and received 400 IU/kg heparin. The photos were taken at 20x magnification with 10% zoom. Scale bar = 100 µm. Please click here to view a larger version of this figure.
Figure 3: Park/Chiu scores for small intestinal segments. Microscopic damage to all three intestinal segments (duodenum, jejunum, and ileum) for animals undergoing intestinal ischemia-reperfusion injury (IRI) was significantly increased compared to those that underwent sham laparotomy (Sham). * p < 0.05 for IRI versus Sham. Please click here to view a larger version of this figure.
Figure 4: Park/Chiu scores for small intestines undergoing 60 min ischemia and 120 reperfusion versus 45 min ischemia and 90 min reperfusion with 400 IU/kg heparin. Decreasing the times from 60 min ischemia and 120 min reperfusion (60/120) to 45 min ischemia and 90 min reperfusion with 400 IU/kg heparin (45/90/H) did not create a statistically significant difference in Park/Chiu injury scores of the small intestines of mice in the IRI group. It did, however, reduce mortality from 50% to 20%. Please click here to view a larger version of this figure.
Despite the widespread use of this intestinal IRI model, it is not without its limitations. For instance, sole occlusion of just the base of the SMA does not completely obstruct blood flow to the intestine. This is likely due to extensive collateral circulation in the mesentery, which can draw blood from neighboring branches of the abdominal aorta. In one study in cats, SMA occlusion decreased blood flow by 35% in the proximal duodenum, 61% in the distal duodenum, 71% in the jejunum and ileum, and 63% in the proximal colon. Blood flow was not reduced in the mid and distal colon, which receive much of their circulation from the inferior mesenteric artery23. In rodents, the jejunum and ileum are most often cited as the intestinal segments which incur the most significant tissue injury following SMA occlusion9.
A wide range of ischemia times after SMA occlusion have been cited in the literature, from 1 to 90 min or more. Different ischemic times result in different levels of reperfusion injury; Park et al. observed reperfusion injury when the ischemic interval was between 40 and 60 min, but not when the ischemic interval was shorter or longer24. Such results suggest that shorter times do not produce enough ischemia to incite reperfusion injury, while longer times damage the tissue so severely that it is impossible to demonstrate the reperfusion injury that follows. In addition, longer ischemic times carry the risk of increased mortality. As seen in our study, 50% (3/6) of the initial mice that underwent 60 min ischemia died after only 90 min of reperfusion. Shortening the ischemia time to 45 min lowered the mortality to 20% (1/5) without changing the tissue injury scores. Based on our study, it appears that the ideal window of ischemic damage can be attained by SMA occlusion for about 45 min.
Another variable is the reperfusion time before tissue collection. As with ischemia times, reperfusion times vary widely across studies, from 60 min to over 24 h. Several papers have reported that the intestinal mucosa incurs maximal morphologic damage at 2 to 3 h of reperfusion, with complete repair achieved at 24 h25,26,27. Collecting tissue before this 2 to 3 h window risks not capturing the full extent of the reperfusion injury, while tissues harvested closer to 24 h will have already started the repair process. We initially opted for a reperfusion time of 120 min, but then changed to 90 min in an effort to lower mortality. This change did not change tissue injury results, suggesting that a 30 min deviation from the 2 to 3 h window is acceptable.
Oxygen concentration is also an important variable in the development of IRI. Wilding et al. found that, compared to mice receiving 21% O2, those anesthetized with isoflurane delivered with 100% O2 experienced ventilation-perfusion mismatch due to atelectasis. In the same study, rats receiving 100% O2 developed acute respiratory acidosis and elevated mean arterial pressure28. Such physiologic changes are best avoided when inducing an injury such as IRI, in which a number of systemic factors are involved. Thus, 21% O2 seems to be more appropriate than 100% O2 as the carrier gas for isoflurane delivery.
The use of heparin in this protocol is open to debate. Heparin is known to have anti-coagulative and anti-inflammatory effects29. We found that changing from 60 min ischemia and 120 min reperfusion to 45 min ischemia and 90 min reperfusion with 400 IU/kg heparin did not change microscopic intestinal injury but did lower mortality. One possible explanation is that heparin prevented fatal thromboembolism to distant organs such as the lungs and brain, however we did not find evidence of this on necropsy by gross or microscopic examination of the initial two mice that died. Using shorter ischemia and reperfusion times without heparin may be just as effective at reducing mortality. If that were the case, it would be prudent to forego the use of heparin to minimize interference with IRI. However, including heparin in the protocol may be appropriate for those wishing to model surgical causes of IRI, as surgical patients often receive heparin peri-operatively.
Isoflurane has been shown to have tissue protective effects in cases of intestinal inflammation and ischemia, and its use may interfere with a clinically relevant IRI model30,31,32. However, organofluorine inhalants (i.e., isoflurane, sevoflurane) are commonly used anesthetics in both veterinary and human medicine. In addition, the length of anesthesia required for this protocol exceeds 120 min, and thus an inhalant is more appropriate than a shorter-acting injectable which would need to be re-dosed.
No microscopic lesions were present in the proximal colon, liver, or kidney. The lack of microscopic changes was perhaps due to the relatively short 90 to 120 min reperfusion time. In addition, the proximal colon has a blood supply from the inferior mesenteric artery. However, a lack of visible damage does not rule out systemic injury. Reverse transcription-quantitative polymerase chain reaction (RT-qPCR) is likely a better methodology to demonstrate systemic injury by measuring inflammatory cytokines such as TNF-α.
Several variations of this intestinal IRI model have been developed over the years. In 1990, Megison et al. demonstrated that occluding collateral vessels in addition to the SMA produced a more consistent reduction of mesenteric blood flow but an increase in the mortality rate33. A more recent study showed that, in lieu of occluding the SMA at its base, ligating its peripheral and collateral branches to induce ischemia in the distal ileum yielded reproducible injury without mortality34. Occlusion of the local arterial branches ensures maximal ischemia and may address the issue of multifocal, segmental reductions of blood flow seen with ligating the SMA just at its base. While this alternative method of modeling intestinal IRI has application for research into the local tissue effects of intestinal IRI, it is unknown whether it can accurately model the systemic inflammation and multi-organ failure which can be associated with intestinal injury.
SMA occlusion is not an appropriate model for all types of intestinal IRI. Non-occlusive mesenteric ischemia, for instance, is characterized by splanchnic hypoperfusion stemming from decreased cardiac output. Therefore, this technique would not be optimal to study intestinal IRI caused by myocardial infarction, congestive heart failure, aortic insufficiency, or renal or hepatic disease35. Kozar et al. reported that SMA occlusion is, however, a clinically relevant model for gut IRI induced by shock36. Although less economical, the use of other species such as pigs may have benefits over rodents for modeling certain intestinal injury conditions. A comprehensive review by Gonzalez et al. in 2014 describes animal models currently in use for investigating intestinal IRI9.
Despite its limitations, the technique of occluding the SMA at its base remains one of the most commonly utilized rodent models of intestinal ischemia9. As it only requires one vascular clamp and a basic setup, the surgery itself is quite simple. It also yields reproducible injury, as evidenced by the data presented here. SMA occlusion in rodents can reliably model occlusive causes of intestinal IRI and can have practical application in both veterinary and human medicine. As such, it is important that the procedures we have outlined here be carried out with consistency.
The authors have nothing to disclose.
Funding for this project was provided by the Division of Intramural Research of the National Heart, Lung, and Blood Institute, National Institutes of Health.
We would like to thank Dr. James Hawkins for his mentorship and support. We also thank Drs. Mihai Oltean and Robert Linford for their assistance in locating the superior mesenteric artery. We would like to extend our thanks to Drs. Patricia Carvalho Obeid Ellrich, Claudio Correa Natalini, and George Howell III for providing their expertise during the development of this protocol. Finally, we would like to thank Stephen Wincovitch for his assistance in acquiring the beautiful photomicrographs featured in this paper and Dr. Alicia Olivier for her help labeling and rendering the final figures.
Adson forceps | Roboz | RS-5236 | Surgical instrument |
Alm retractor | Roboz | RS-6510 | Surgical instrument |
Anesthesia machine | Datex-Ohmeda | Aestiva 5 | |
Anesthesia: isoflurane | Baxter Healthcare Corporation | NDC 10019-360-40 | Dose: 1-4%, INH |
Angiocath 20 g x 2 | Smiths Medical | 5057 | Flushing intestines with saline and formalin |
Atraumatic microvascular clip | Teleflex | 065100 | Surgical instrument |
Buffered formalin 10% | Fisher Scientific | 23-245684 | Tissue fixation |
Bupivicaine 0.25% | Hospira, Inc. | NDC 0409-1160-18 | Dose: up to 2 mg/kg drop-wise |
Buprenorphine | Par Pharmaceutical | NDC 42023-179-05 | Dose: 1 mg/kg, SQ |
Chlorhexidine scrub 2% | Vet One | 510083 | Surgical site prep |
Circulating water blanket | Cincinnati Sub Zero | Blanketrol 2 | Body temp maintenance |
Clippers – Wahl BravMini, Purple Hair clippers | Lambert Vet Supply | 008WA-41590-0438 | Surgical site prep |
Conical tubes 50 ml | Fisher Scientific | 14-432-22 | Tissue fixation and storage |
Dry ice | N/A | N/A | PCR tissue samples |
EtOH 200 proof | The Warner-Graham Company | 64-17-5 | Tissue storage |
Heparin (optional) | Meitheal Pharmaceuticals | NDC 71288-402-11 | Dose: 200-600 IU/kg |
Induction chamber | VetEquip | 941456 | |
Indus Instruments THM100 Rodent Monitor | Indus Instruments | N/A | For monitoring rodent body temperature during surgery |
Isopropyl Alcohol 70% | Humco | NDC 0395-4202-28 | For scrubbing surgical site |
Microcentrifuge Tubes: 0.6mL | Fisher Scientific | 05-408-121 | PCR tissue samples. 8 per mouse, Terminal bleed collection, serum storage |
Microsoft Excel | Microsoft | N/A | |
Nose cone | N/A | N/A | Can be homemade with syringe tube or bubble tubing |
O2 medical air 21% | Roberts Oxygen | N/A | Rate: 0.5 L/min for each L chamber volume |
Ophthalmic ointment | Akorn, Inc. | NDC 17478-062-35 | Surgical prep |
PBS pH 7.4 (1x) | ThermoFisher Scientific | 10010-031 | For tissue rinsing and making 70% EtOH |
Specimen cups | Cardinal Healthcare | C13005 | For holding tissue cassettes in formalin |
Sterile Castroviejo Needle Holder | Roboz | RS-6412 | Surgical instrument |
Sterile cotton swabs | Medline | BXTA50002Z | |
Sterile gauze | Medline | PRM21423Z | |
Sterile Micro Dissecting Scissors | Roboz | RS-5980 | Surgical instrument |
Sterile micro dissecting spring scissors | Roboz | RS-5693 | Surgical instrument |
Sterile micro forceps | Roboz | RS-5264 | Surgical instrument |
Sterile saline (0.9%) | Braun | R5201-01 | Must be warmed |
Sterile scalpel blade #15 | Cardinal Health (Allegiance) | 32295-015 | Surgical instrument |
Sterile scalpel handle | Roboz | RS-9843 | Surgical instrument |
Sterile surgical drape | Medline | DYNJSD1092 | |
Sterile surgical gloves | Medline | MSG2270 | |
Sterile surgical stapler | Roboz | RS-9260 | Surgical instrument |
Sterile surgical staples | Roboz | RS-9262 | Abdominal skin closure |
Sterile suture: Vicryl (polyglactin 910) 6-0, 27" Taper RB-1 Needle | Ethicon | J212G | Closing abdominal muscle |
Surgical tape | Medline | MMM15271Z | Securing mouse in dorsal recumbancy |
Syringe 10 ml x 2 | Medline | SYR110010 | Flushing intestines with saline and formalin |
Tissue cassettes | Fisher Scientific | 22-038-665 | Rolled intestinal segments. 4 per mouse. |
Towel or drape | Medline | GEM2140 | Water blanket cover |