The present protocol describes a rodent model of newborn hypoxic-ischemic injury for identifying early changes in cerebral tissue by gross morphology and magnetic resonance imaging. This has benefits over existing models, which can be used to study late injury but do not allow the evaluation of reproducible early changes.

Perinatal hypoxic-ischemic encephalopathy (HIE) is an acute disease that may afflict newborns, resulting in variable long- and short-term neurodevelopmental outcomes. Early diagnosis is critical to identifying infants who may benefit from intervention; however, early diagnosis relies heavily on clinical criteria. No molecular or radiological tests have shown promise in detecting early cerebral injury. Studies have shown that magnetic resonance imaging (MRI) can show changes in both blood flow/ischemia and metabolic disruption. However, they have all been used to evaluate the secondary phase of the disease (>12 h) after the onset of the injury. Early diagnosis is critical to rapidly starting therapeutic hypothermia in eligible infants, which is currently recommended to be initiated within 6 h of birth. The rat model of hypoxic-ischemic injury was developed in 1981 and has been validated and used extensively to study changes in brain perfusion, cerebral injury markers, and morphology. However, it has primarily been used as a "late model", evaluating injury several days after the initial ischemic insult. The model has been known to have poor sensitivity in evaluating reliable and reproducible early cerebral changes. The objective of this study was to develop a reliable model to study early gross morphological and radiological markers of HIE using pathological staining and cerebral magnetic resonance imaging/magnetic resonance spectroscopy.

Hypoxic ischemic encephalopathy (HIE) is a devastating condition resulting from various factors in newborn infants¹. Perinatal asphyxia and/or the disruption of cerebral blood flow may result in focal or global ischemic changes in the brain². The occurrence rate is approximately 1.6 in 1,000 live births but may be as high as 12.1 in 1,000 live births in developing countries³. This condition results in high mortality (20%-50%), while 25% of those who survive are likely to suffer from a long-term neural disability such as mental retardation, epilepsy, or cerebral palsy⁴. The only therapeutic intervention proven effective in mild to moderate injury is therapeutic hypothermia, which must be initiated within 6 h of birth^5,6,7,8,9. While this may help prevent the metabolic changes that lead to secondary injury, there may also be potential for side effects such as hypotension, thrombocytopenia, prolonged coagulation time, intracranial hemorrhage, dysrhythmias, fat necrosis, and serum electrolyte imbalance^4,5. Early diagnosis of HIE in babies is often difficult as the criteria are subjective and rely heavily on physical exam findings, which evolve over time. Magnetic resonance imaging may show changes reflective of injury several days to weeks after injury. However, morphologic changes in T1/T2 MRI can be normal in up to two-thirds of moderate encephalopathy, the category of infants most likely to benefit from therapeutic hypothermia¹⁰. As per recent reports, magnetic resonance spectroscopy (MRS) may show early changes correlating with neonatal HIE¹¹. However, no standardization or validation has been performed to date.

Many investigators rely on animal models to evaluate potential diagnostic or therapeutic interventions for cerebrovascular injury. The most frequently used method to create an infarct is ligating rodents' middle cerebral artery^12,13. While often used to study adult ischemic stroke, this is technically challenging in neonatal rodents due to the small size and the fragility of the pups at the age equivalent to human newborn disease. Furthermore, it does not represent the global cerebral ischemic changes likely to be seen in HIE. The Rice-Vanucci Model¹⁴ of unilateral carotid artery ligation in rats has been used since the 1980s as a cost-effective rodent model to study hypoxic-ischemic brain injury. However, there is large variability in early cerebrovascular changes and high mortality in earlier experiments. Most studies report the cerebral injury in long-term changes (i.e., after 24 h of injury), which are more consistent. This study aimed to develop an approach to evaluate early (within 6 h) molecular and radiological changes in a rat model of HIE. The protocol was designed to ensure ischemia at an early (term newborn equivalent) age and to increase the survival of the pups, especially during exposure to hypoxia. MRI/MRS were used to evaluate radiological evidence of altered flow, cerebral tissue changes, and metabolic changes within 6 h of injury. Gross morphological evaluation of the infarct areas was also performed. Further validation of the reproducibility was conducted by repeating the experiments in multiple litters.

All the experimental procedures were approved by the Oklahoma Medical Research Foundation (OMRF) Institutional Animal Care and Use Committee (protocol IACUC #17-17). Pregnant female Sprague-Dawley rat pups at E14 were used for the present study. The animals were obtained from a commercial source (see Table of Materials).

1. Animal preparation

1. Acclimatize the animals in the animal facility prior to delivery of the litters.
2. Maintain all rats on a12 h light/darkcycleand feed standard ratchow.
3. After delivery of the litters, keep the pups with their respective dams. Use both sexes for the experiments.
4. At postnatal age 10 (P10), randomize the pups to either sham or HIE groups.
   ​NOTE: The experimental timeline is depicted in Figure 1. All the experiments were performed on the same day at P10.

2. Carotid artery ligation (CAL) for the experimental HIE group

1. Place the rat pups on a warming pad.
2. Initiate anesthesia with 4% isoflurane in oxygen (0.6 LPM) using a nasal cone until the pinch reflex disappears. Drop the flow of gas to 0.5%-2% for the maintenance of anesthesia. Ensure the pups are unconscious without suppressing the respiratory drive.
3. Mark the pups on the tail for identification and gently restrain them with tape on all four limbs.
4. Shave the neck area and sterilize with 70% iodine-povidone solution swabs (see Table of Materials).
5. Using an #11 blade, make a 1 cm midline neck incision through the skin. Carefully dissect the left parotid gland and fascia until the left carotid artery is exposed. Gently mobilize the vessel using hemostats to free it from the fascia.
6. Using a small hemostat, carefully pass two 5-0 sutures (see Table of materials) around the vessel, 0.5 cm apart, and tie them tightly.
7. Using small scissors, cut the artery in between the two sutures to ensure the discontinuity of blood flow. Close the skin and fascia with 5-0 silk sutures (see Table of Materials).
8. Inject buprenorphine (10 units in 800 μL of sterile normal saline) intraperitoneally and another 800 μL of normal saline subcutaneously in the back of the neck to prevent dehydration.
   NOTE: The procedure must be completed within 10-12 min.
9. Return the pups to the cages with their dams and allow the pups to awaken and recover for 1-2 h.

3. Sham surgical procedure for the control group

1. Place the rat pups on a warming pad.
2. Initiate anesthesia with 4% isoflurane in oxygen (0.6 LPM) using a nasal cone until the pinch reflex disappears. Drop the flow of gas to 0.5%-2% for the maintenance of anesthesia. Ensure the pups are unconscious without suppressing the respiratory drive.
3. Mark the pups on the tail for identification and gently restrain them with tape on all four limbs.
4. Shave the neck area and sterilize with 70% iodine-povidone solution swabs.
5. Using an #11 blade, make a 1 cm midline neck incision through the skin and then close it with 5-0 silk sutures.
6. Follow the same hydration, analgesia, and postoperative care as for the HI group (steps 2.8-2.9).

4. Hypoxia exposure for both the CAL and sham groups

1. Prepare the clear plexiglass hypoxia chamber (see Table of Materials) by attaching tubing to the chamber lid to provide continuous airflow of 6 LPM of the hypoxia gas mixture (8% oxygen, 92% nitrogen).
2. Place a blue absorbent pad in the chamber and immediately place the rat pups from both groups in the chamber. Allow the pups to remain in the hypoxia chamber for 45 min.
3. Immerse the chamber in a water bath with continuously flowing warm water to keep the temperature set at 37 °C inside the chamber.
4. Hydrate the pups with an oral saline solution via gavage of 600 μL before being placed in the hypoxia chamber and 600 μL at the end of the 45 min.
5. Remove all the pups to the cages with their dams (experimental and sham) and allow them to recover for 2 h in a room next to the small animal imaging facility.

5. Magnetic resonance imaging and spectroscopy

1. Perform MRI and MRS to identify and evaluate the radiological and metabolic markers 4 h after the end of carotid artery ligation. Perform the procedure under anesthesia with continuous cardiovascular monitoring at the small animal imaging facility.
2. Anesthetize each animal (with 1.5% isoflurane and 0.7 L/min oxygen) and place it in the MR probe (see Table of Materials) in a supine position on a blue absorbent pad covering a heating pad. Monitor the respiration rate of the animals continuously using an abdominal pneumatic pillow (see Table of Materials).
3. Use a head surface coil as a signal receiver and transmit radiofrequency pulses to the sample through a quadrature volume coil (72 mm inner diameter, see Table of Materials).
4. Perform MRI to evaluate both the changes in cerebral blood flow (CBF) and the changes in water diffusion constants (ADC) following previously published methods^15,16,17. Perform MRI morphology (T1 and T2), diffusion, and perfusion to determine the brain's most affected and least perfused areas.
   NOTE: The mean values for perfusion and diffusion (ADC) in each group are compared between the ligated side and the control side (intact carotid artery side).
5. Perform MRS following previously published methods^15,17 and analyze using an in-house coded program using Mathematica software (see Table of Materials).
6. Scale the MR spectra in parts per million (ppm) by calibrating against the water peak (4.78 ppm). Identify the major brain metabolic peaks as N-acetylaspartate (NAA) at 2.02 ppm, choline (Cho) at 3.22 ppm, creatine (Cr) at 3.02 ppm, and myo-inositol at 3.53 ppm.
   ​NOTE: The peak area measurements of the metabolites are used to calculate the following ratios: NAA to Cho (NAA/Cho), Cr to Cho (Cr/Cho), and Myo-Ins to Cho (Myo-Ins/Cho)¹⁵.

6. Serum and cerebral tissue analysis

1. Perform blood sampling at 5.5 h following the carotid artery ligation or sham procedure according to previously published methods¹⁸.
2. Anesthetize the pups again with 4% isoflurane.
3. Using a sharp #11 blade, make an abdominal incision, followed by a diaphragmatic incision to expose the heart.
4. Perform blood sampling via cardiac puncture as previously described¹⁸. Briefly, insert a 32 G needle on a 1 mL syringe into the right heart chamber and gently aspirate 1 mL of blood.
5. Allow the whole blood to coagulate, followed by centrifugation at 1,000 x g for 15 min at 4 °C. The serum gets separated into clean microcentrifuge tubes.
6. Decapitate the whole pup head for the gross assessment of cerebral pathology and then immerse it in ice for 2 min.
7. Make an incision on the dorsal scalp from the base of the skull to the tip of the nose and peel the skull bones from around the brain. Remove the intact whole brain into a clean Petri dish.
8. Mark the right side of the brain with a non-toxic marker. Position the brain with the cephalad surface upward so that both hemispheres are visible. Using an ice-cold razor blade, slice the brain into four equal sections parallel to the coronal plane.
9. Immerse the brain sections in 2,3,5-triphenyltetrazolium chloride (TTC, see Table of Materials) solution in a Petri dish covered with foil to prevent photosensitization and incubate for 15 min at 37 °C.
   NOTE: Infarcted areas are delineated as white areas devoid of red TTC stain.
10. If infarcts are subtle or difficult to detect, inject 0.5-1 mL of TTC (1%) in phosphate-buffered saline directly into the right heart after abdominal incision and thoracotomy, and allow to perfuse for 2 min prior to the decapitation of the pup.
11. Store the brain tissue and serum at −80 °C if further analysis is required.

The present protocol to produce and evaluate early cerebral changes after HIE was easy to implement and allowed gross pathological and radiological visualization of cerebral injury within 6 h of insult in rat pups at P10. The experimental plan is depicted in Figure 1. Both sexes were analyzed together, and 24 animals from five litters were examined in each group. Animal mortality was very low, with 99% survival of animals until the terminal experiments were performed.

Areas of ischemia are clearly visualized 6 h after CAL and hypoxia on brain sections stained with TCC. These lesions are seen as a clear white area of injury devoid of stain, depicted in Figure 2. No cardiac perfusion of TCC was necessary for any samples to visualize the injury. Many animals had multiple areas of infarcts denoted. The majority were on the cortical surface of the brain.

A decrease in MR perfusion on the ligated side of the brain compared to the perfusion on the contralateral or non-surgical side of the brain was seen. An 80% decrease in perfusion on the ligated side in the HIE group was detected 3 h after CAL and hypoxia. As expected, there was no difference between both sides of the brain in the sham group. Representative MR results are shown in Figure 3. A similar pattern was seen in the diffusion scan between the two sides of the brain. A 10% decrease in ADC was seen on the ligated side of the brain in the HIE group that was detected 3 h after CAL surgery. Again, there was no difference between both sides in the sham group.

The results of the MRS analysis are shown in Figure 4. MRS metabolites are normalized to creatine, the most stable of all the metabolites. There were no statistically significant differences between the groups, but there was a clear trend for N-acetylcysteine (NAA) to decrease by 30%, while myo-inositol increased more than two-fold in the HIE compared to the sham group. There was no change in either choline or taurine.

Figure 1: Experimental timeline for rat pups at P10. Both the experimental group and the sham control group are shown. Please click here to view a larger version of this figure.

Figure 2: Representative image of a coronal brain section after staining with TCC. The area of the infarct shows the absence of staining (white), which is marked in the dotted line. Abbreviation: TCC = 2,3,5-triphenyltetrazolium chloride. Please click here to view a larger version of this figure.

Figure 3: Representative MR results. (A) The MR perfusion and (B) apparent diffusion coefficient (ADC) of the carotid artery ligation (CAL) side of the brain in both groups (which has either the CAL in HI or just the neck incision in sham) compared to the perfusion of the contralateral or non-surgical side of the brain. The side of the brain with CAL has a statistically significant decrease in relative cerebral blood flow (RCBF) and ADC compared to the non-ligated side (n = 6 per group). The comparison of the CAL and sham groups was analyzed using unpaired two-tailed t-tests with **** p < 0.0001, ** p < 0.01. The error bars show the standard deviation (SD). Please click here to view a larger version of this figure.

Figure 4: MRS analysis results. For MRS metabolites, there were no statistically significant differences between the groups, but there was a trend toward a decrease in NAA between the ligated and non-ligated sides. Myo-inositol increased more than two-fold in the HIE compared to the sham group. There was no change in either choline or taurine. Abbreviations: MRS = magnetic resonance spectroscopy; NAA = N-acetylcysteine; Cr = creatine; Cho = choline; Glu = glucose; Tau = taurine, Myo-Ins = myo-inositol. n = 6 per group. The comparison of the CAL and sham groups was analyzed using unpaired two-tailed t-tests with *p < 0.05. The error bars show the standard deviation (SD). Please click here to view a larger version of this figure.

A research protocol in newborn rat pups was successfully designed to visualize and analyze early markers of cerebral injury in HIE. To date, there is a lack of objective assessment tools to detect early cerebral injury in the newborn population. After HI injury, there is a phase (1-6 h) in which the impairment of cerebral oxidative metabolism has the potential to partially recover before the failure of mitochondrial function¹⁹, which is irreversible. This latent phase is the therapeutic window for neuroprotective interventions such as therapeutic hypothermia⁶. Most published clinical trials advocate for the initiation of therapy within the first 6 h^5,7, and most institutions do not offer this therapy past that time. However, cooling must be used selectively as there is a potential for serious adverse effects²⁰. These include dysrhythmias, hypotension, platelet dysfunction and thrombocytopenia, persistent pulmonary hypertension, coagulation disturbances, and subcutaneous fat necrosis²¹. Therefore, the decision to begin treatment must be made selectively only for those infants who would have the most benefit.

The use of rodents in the study of HIE is both cost-efficient and easy to perform. A variety of postnatal days (PND) have been suggested to study cerebral ischemic injury in rats, ranging from P7 to P13. The original Rice-Vanucci model¹⁴ consisted of unilateral common carotid artery ligation in rat pups at P7. At 4-8 h later, the pups were exposed to 8% oxygen at 37 °C for 3.5 h. The brain tissue was then examined for morphological changes. A literature search revealed that most investigators use this model to evaluate markers of injury after 24 h or longer. The acquisition of consistent and reproducible results at shorter intervals has proven to be difficult. In addition, high mortality has been observed by investigators due to dehydration, as well as difficulty in maintaining body temperature during exposure to hypoxia. This protocol was designed to ensure improved survival of the rat pups, as well as a clear delineation of brain injury. This is critical to evaluate potential diagnostic tools and other early therapies for preventing encephalopathy. For these experiments, P10 in Sprague-Dawley rat pups was chosen as the timepoint, as this most closely represents the term newborn period while ensuring the survival of the animals. The maintenance of hydration is paramount. An oral electrolyte solution was gavage fed both before and after hypoxia, and a normal saline solution was injected intraperitoneally and subcutaneously. Another factor was the strict maintenance of body temperature at 37 °C. These key steps ensured >90% survival of the rat pups in both groups at the time of terminal evaluation.

Other authors have reported a wide variation in the degree of injury observed after unilateral carotid artery ligation. While this model successfully reproduces neuronal injury, it was also observed that the degree of injury is variable as the contralateral (unligated side of the brain) provides blood flow in variable amounts (via the circle of Willis) to restore blood flow to the ischemic area. However, the variation in injury severity correlates well with what is seen in human babies and serves as a relevant, clinically equivocal animal model to study disease.

The confirmation of infarcted areas is necessary by gross visualization as often areas of injury correlate with prognosis and may help identify areas of neuronal tissue for further analysis. For the visualization of infarcted neuronal tissue, 2,3,5-triphenyltetrazolium chloride (TTC) is a commonly used compound as it differentiates viable tissue from infarction macroscopically^22,23. TCC is reduced to form a red formazan product mainly in the mitochondria. The intensity of the red stain correlates with the number and functional activity of mitochondria²⁴; therefore, TTC-unstained brain tissue is infarcted and TTC-stained tissue viable. Gross morphological evaluation of brain tissue after TCC staining showed clear areas of infarct and injury. The pathological data shown did not require cardiac perfusion with TCC; however, it may be performed, especially when a more subtle injury is visualized.

Magnetic resonance imaging has been used extensively in human babies to determine the areas and severity of injury affected by infarcts and injury after HIE. However, most studies evaluated injury at a "late" stage (i.e., after 48 h of life)^10,25,26. The recent emergence of MRS and its ability to target specific bio-metabolites affected during HIE has made it a valuable prognostic tool²⁷ in evaluating these infants. Again, no studies have evaluated its use in the early stages of the disease. Using a combination of MRI/MRS techniques, these data show that early changes and injury markers are apparent in the rodent model, which may identify infants with cerebrovascular involvement and target them for early therapies. Results from this study show that N-acetylaspartate (NAA) levels are decreased in infarcted tissue. This is similar to previous studies on rats and humans at different stages after injury. For example, Looji et al. found a 30% decrease in NAA levels 5 h after injury in adult human subjects¹⁶. Lally et al. investigated MRS metabolite changes 4-14 days after injury in human newborns as prognostic factors for determining the neurodevelopmental outcome. They found that the population with neurodevelopmental delay had lower NAA than the less-affected subjects²⁷. Cheong et al. found similar results on days 1-3 after injury²⁸. In the present study, it is seen that a decrease in NAA levels can occur within 1 h after the HI injury and can be used as a very early marker of diagnosis.

Choline was found to be decreased early after injury in the HIE group. Our findings are similar to other studies that showed a mild decrease in choline in MR spectroscopy, which was also directly related to the severity of the neurodevelopmental outcome in the newborn with HI injury^27,29. This contrasts with the findings of Guo et al., who observed an increased choline/Cr ratio in human subjects from 0-15 days of age⁵. However, levels of choline/Cr seem to vary according to age, sex, and time after the injury. Serial-timed MRS may be needed to study the variation in choline levels reported by different studies.

The only metabolite found to be increased in the current data from early MRS spectroscopy was myo-inositol. Myo-inositol is one of the least-studied metabolites in HIE-related MRS. Van de Looji et al.²⁸ found that myo-inositol levels decreased 4 days after injury, but these levels increased significantly after that. These differences might be partially explained by the findings of Shibasaki et al., who linked the change in myo-inositol levels to the severity of HI injury²⁹. On comparing myo-inositol expression in newborns sustaining HI injury with different outcomes, they found that, despite the early increase in myo-inositol expression, it drops within 2 weeks of injury, with this drop much more significant in the group severely affected by the HI injury. Therefore, this may be a useful marker of the severity of the injury.

In conclusion, maintaining hydration and body temperature were the key elements in successfully establishing a rat model to evaluate early cerebral injury in HIE. This will be an invaluable tool to help define markers and for the evaluation of early injury in term newborn infants. The early HIE rat model allows cerebral injury assessment by all modalities tested, including MRI, MRS, and gross pathological examination. Hopefully, these studies will be of value in the early identification of infants that will benefit the most from lifesaving therapies.

There are no conflicts of interest for any of the authors.

We thank the veterinary staff of the Oklahoma Medical Research Foundation for their expertise and assistance in modifying the animal care protocols.