In recent years, improvements in obstetric and neonatal intensive care techniques have led to a significant increase in the survival rate of preterm infants, and the incidence of brain injury in preterm infants has been increasing year by year. About 35% of survivors have chronic neurological deficits such as cerebral palsy, epilepsy, learning difficulties and cognitive abnormalities.
Hypoxia-ischemia and infection are the main causes of brain injury in preterm infants. According to the neuropathology, they can be divided into.
(i) white matter injury (WMI), including periventricular leukomalacia (PVL) and diffuse cerebral white matter injury;
(ii) Non-parenchymal hemorrhage, including periventricular-intraventricular hemorrhage, subarachnoid hemorrhage, and choroid plexus hemorrhage;
(iii) other areas of injury, such as brain parenchyma, cerebellum, brainstem, etc.
Intraventricular hemorrhage (IVH) was the most predominant form of brain injury in preterm infants 10 years ago, while WMI is currently the most common type of brain injury, mostly occurring in preterm infants between 23 and 32 weeks of gestational age, and is the main cause of severe neurological sequelae. Among them, PVL is more prevalent in WMI, characterized by periventricular white matter cysts and coagulative necrosis, manifested by damage to oligodendrocyte precursor cells, diffuse damage and impaired formation of myelin sheaths, reactive gliosis and microglial cell activation. However, recent neuroimaging studies have shown that diffuse white matter injury surpasses the occurrence of PVL and predominates.
In the study of brain injury in preterm infants, the production of animal models is a prerequisite for basic research. With the help of animal models, the etiology, pathology, pathogenesis, and treatment and prognosis of brain injury in preterm infants can be studied in more depth and detail. There are many animal models of brain injury in preterm infants, and the current status of their research is reviewed in this paper as follows.
1. Commonly used animals for studying brain injury in preterm infants
It is very difficult to choose an animal model of brain injury in preterm infants that best fits the human developmental process, because there are differences in brain structure and developmental stages in different species, and different regions of the brain develop at different rates. Therefore, to date, there is no animal model that best mimics the disease in humans. Regardless of the form of injury, the key to animal selection is the species of animal and its age of maturation of the central nervous system.
In general, it is much easier to make models that approximate the distribution and morphological characteristics of the injury to human WMI in polycephalic gyrus animals such as rabbits, dogs, pigs, cats, and sheep than in rodents. However, in recent years, hypoxic-ischemic brain damage (HIBD) models in rats from 1 to 7 d postnatally have been more developed, because the models established using small animals are more advantageous in observing long-term neuropathological conditions and behavioral aspects.
The rodent is the most popular model for experimental research because of its rapid reproduction and low cost. Newborn rats at 1 d (postnatal day 1, P1), 2 d (postnatal day 2, P2), 3 d (postnatal day 3, P3), and 5 d (postnatal day 5, P5) are approximately equivalent to viable preterm infants born at 18-20, 20-24, 24-28, and 28-32 weeks of gestational age, respectively. The postnatal day 7 (P7) neonatal rat has been the animal model for full-term infants, but recent studies have shown that it is equivalent to near-full term (approximately 32-36 weeks), with full term being about 10 d after birth.
Models established using neonatal rats provide physiological parameters that more closely resemble the human intrauterine or prenatal situation. Non-human primate models are more valuable for research than lower-grade animals, such as rhesus monkeys, macaques, and baboons. The premature baboon is a model with a high similarity to human neonatal maturation and forms of brain damage and can help provide evidence of factors associated with preterm birth in humans. Fetal monkeys have been used to establish animal models of HIBD, but their application to basic experimental research is limited because they are expensive and limited in number.
2. Animal model of brain injury in preterm infants caused by hypoxia-ischemia
In 1960, Levine first used ligation of the common carotid artery and hypoxia to make an animal model of HIBD, and in 1981 Rice applied this technique to neonatal rats and succeeded. The HIBD model was made by ligating the unilateral common carotid artery in P7 neonatal rats and placing them in a hypoxic chamber with limited time hypoxia. The model was able to observe tissue damage not only in the cerebral cortex, hippocampus, striatum and thalamus, but also in the subcortical and periventricular white matter, which was soon widely used in laboratories worldwide and became the classic neonatal HIBD model. Subsequently, scholars have focused their research on HIBD on white matter brain injury and established various animal models of HIBD in preterm infants.
The commonly used models of HIBD, such as bilateral common carotid artery occlusion or unilateral common carotid artery ligation combined with 6% to 8% oxygen hypoxia for 1.5 to 3.5 h, can cause significant white matter damage. sheldon et al. ligated the unilateral common carotid artery in P1 neonatal rats and then administered 6% oxygen hypoxia for 3.5 h. They found corresponding tissue damage in the ipsilateral cerebral hemisphere. Reactive changes in astrocytes and microglia were seen in the first day after injury, with loss of neurons and edema in the cortex and caudate nucleus in the second day, reduced edema in the third day, and extensive reactive changes in the area of injury with significant loss of white matter on the damaged side in the 10th day. This model simulates HIBD in the very immature brain, is reproducible, and helps to elucidate the mechanisms of neuronal loss and PVL.
Alfreda to P3 neonatal rats inhaled 5% oxygen for 60 min and found that 14 of 30 animals in the model died after 24 h. Nine of the survivors developed brain damage, and the degree of damage was consistent with surviving preterm infants with brain damage. Uehara et al. established a PVL model by ligating the bilateral common carotid arteries in P5 neonatal rats, and 90.9% of the cerebral white matter was altered at 7 days of age, including coagulative necrosis and cystic damage in and around the internal capsule, and the internal capsule and The white matter was more susceptible to damage than the cerebral cortex when the cerebral perfusion pressure decreased by about 25% in rats.
Therefore, this model is useful to study the white matter damage caused by low perfusion, and Goñi-de-Cerio et al. ligated the immature fetal lambs with unilateral common carotid artery for 1 h. After 3 h of execution, they found that neuronal necrosis was mainly seen in the midbrain, pons, deep cerebellar nuclei, and basal nuclei, while apoptotic cells were widely seen in the white and gray matter and were not fixed in one brain region. Some fetal lambs showed extensive asymmetric changes in the cell membrane and impaired mitochondrial integrity. This model confirms that early forms of cell death after hypoxia-ischemia, such as apoptosis and necrosis, are brain region-specific and have significant kinetic characteristics, suggesting the importance of rescuing cells, especially apoptotic cells, in therapeutic strategies.
The disadvantage of the above model is that it cannot simulate intrauterine fetal hypoxia due to prenatal factors such as placental insufficiency, umbilical cord factors or maternal causes, which have more complex and diverse pathogenic mechanisms. The intrauterine HIBD model is commonly established by clamping the bilateral uterine arteries of pregnant rats. The histopathological changes of the brain were neuronal degeneration, edema and necrosis, glial nodule formation, and liquefied brain tissue necrosis to form cavities. A model of HIBD in pigs was established. This method resulted in enlarged ventricles, reduced cortical, striatal and hippocampal volumes, reduced neuronal numbers and inhibited dendritic and axonal growth in fetal pigs.
Haan et al. clamping the umbilical cord of pregnant ewes for 10 min resulted in transient asphyxia in near-full-term fetal lambs, with histologically visible neuronal necrosis in the hippocampal region and depressed neuronal activity on EEG. Derrick et al. produced an animal model of hypertonic cerebral palsy using rabbits with in utero placental insufficiency. P1 surviving rabbits were given the corresponding neurobehavioral tests, and the results of the video tests showed that the hypoxic group had significant deficits in multiple response activities, sucking and swallowing coordination, and increased muscle tone and active flexion/extension movements of the limbs. Histopathological examination revealed specific acute brain injury features in the subcortical motor pathways including the basal ganglia and thalamus.
Persistent damage to the caudate nucleus and thalamus was significantly associated with hypertonic dyskinesia. This suggests that prenatal ischemia and hypoxia lead to hypertonicity and abnormal motor control. The results provide a unique behavioral model to study the mechanisms and sequelae of perinatal brain injury due to prenatal hypoxia-ischemia.Inder et al. To study the effects of preterm birth and neonatal intensive care on the immature brain, 16 baboons of 125 d gestational age (184 d at term) were selected and given two weeks of intensive care, and neuropathological wave spectroscopy showed white matter damage, intracranial hemorrhage and ventricular enlargement, which was similar to that of human The results of brain injury in preterm infants were highly similar.
3. Animal models of infection-induced brain injury in preterm infants
3.1 It is generally accepted that perinatal infection and maternal-fetal inflammation are closely related to preterm birth and brain injury. Inflammation is an important pathogenesis of brain injury and is consistent with other significant consequences of brain injury and dysplasia. Relevant epidemiological data and multiple animal models confirm this view, but the causal relationship between the two remains unclear.
To explore the mechanism of the effect of infection on the immature brain, many neonatal animal models have been used for the study of WMI. For example, prenatal or postnatal administration of microorganisms such as bacteria (Colletotrichum or Gardia vaginalis), viruses (sheep plague virus) or bacterial products (lipopolysaccharides) have been used to model WMI. Lipopolysaccharide is the most commonly used infectious substance for making immature murine, feline, dog, and sheep WMI models, and the mechanism of its induction of brain injury is not fully understood and may involve an immune response, which in turn initiates a systemic inflammatory response leading to systemic hypoglycemia, blood clotting, inadequate cerebral perfusion, and activation of inflammatory cells in the central nervous system.
Administration of lipopolysaccharide to sheep up to 65% full term gestational age produced a similar distribution of WMI to that produced by umbilical cord occlusion. Despite the different morphological changes, lipopolysaccharide resulted in more inflammatory cell infiltration into the brain and local microglia/macrophages than the extensive microglia response provoked by cerebral perfusion deficit. Moreover, small doses of lipopolysaccharide injections per se had no side effects in P7 rats but increased the extent of brain damage after hypoxia-ischemia, suggesting a susceptibility of bacterial products to immature brain.
If lipopolysaccharide 0.5 mg/kg was administered intraperitoneally to pregnant SD rats of 18-19 d gestational age , 7 d after birth, the developing brain was myelinated abnormally and white matter cell death was evident. Moreover, the brain damage caused by small doses of lipopolysaccharide administration will affect into adulthood. Other bacterial products such as lipopolysaccharide have also been used in studies of immature brain tolerance, suggesting that the innate immune system has different responses to various ligands, and this needs to be further explored.
4. Other models of brain injury in preterm infants
Hyperoxia, high temperature, and injection of excitatory amino acid receptor agonists have been used to create models of brain injury in preterm infants. For example, animals were exposed to 80% oxygen from birth until 5 d postnatally and adjusted twice daily to create a hyperoxic brain injury model, and similar changes to the aforementioned brain injury models could be found. In contrast, Nuñez et al. used exogenous fly muscarinic alcohol, a selective γ-aminobutyric acid receptor agonist injected into P1 neonatal rats to produce a model of brain injury in preterm infants. At 7 and 21 d postnatally, a decrease in hippocampal neurons was seen. Behavioral functions associated with the hippocampus were reduced, such as the ability to learn in the water maze, until adolescence.
Chen et al. injected excitotoxic gossypolysine from the bilateral lateral ventricular horn and the white matter between the forelimb sensorimotor cortices in P5 and P7 rats, and by 14 d postnatally, the P5 group had intact cortical gray matter but showed focal reduced myelination and regional cystic degeneration. In the P7 group myelination was less impaired, but extensive immunoreactive deficits of neuro-microfilaments were present. Testing of their corticospinal function at 28 d postnatally revealed that all rats improved with age, with no difference in forebrain transection area on histological examination, but a significant enlargement of periventricular area, especially in the P7 group.
The density of anti-myelin immunoreactivity in the corpus callosum was significantly lower in the P5 group than in the anterior union. This model, especially the P5 group, functionally mimicked the periventricular white matter injury model, also suggesting that P7 is the period when dentary development is equivalent to human corticospinal maturation and susceptible to periventricular white matter injury, whereas P5 is the period when oligodendrocyte precursor cells are more sensitive to excitotoxicity.
Intraventricular hemorrhage in the matrix of the germinal layer is a common neurological problem in preterm infants. In the IVH group, more neutrophils and microglia were seen around the ventricles and more periventricular apoptotic and degenerated neurons were seen than in the control group. immunolabeling for Beta amyloid precursor protein and neurofilaments suggested axonal damage.
Neurobehaviorally the IVH group had poorer and more unstable walking ability than controls. No manifestations of acute systemic toxicity have been found. Also, the same findings of periventricular apoptosis and cellular infiltration in IVH were found on autopsies from preterm infants, but not in the IVH-free group. Thus, this model can be used as an animal model of IVH in preterm infants and can provide evidence of acute brain injury, laying the foundation for the prevention of IVH and posthemorrhagic complications.
The establishment of an animal model of brain injury in preterm infants enables us to conduct in-depth studies on the pathophysiological mechanisms underlying the occurrence of brain injury in preterm infants, to gain a clearer understanding of the process of developmental changes in tissue injury and its outcome, and also helps to conduct validation of therapeutic strategies and treatments and open up further directions for research and development. At present, there is no single animal model that can simulate all brain injuries in preterm infants, and each model has its advantages and disadvantages, so a suitable animal model should be selected for research focus. It is also believed that more animal models closer to clinical practice will be established in the future, so that our research on brain injury in preterm infants will be improved.