Background

Each year, millions of neonates are affected by intrauterine hypoxia, of which the most recognized form is hypoxic ischemic encephalopathy (HIE) in full-term infants (Lee et al., 2013; Stanaway et al., 2018). Prenatal hypoxia is one of the leading worldwide causes of long-term brain injury (Lee et al., 2013; Stanaway et al., 2018), and HIE is characterized by an intrapartum loss of oxygen and nutrients. Approximately 40% of affected children develop a neurodevelopmental disability (NDD), such as autism or epilepsy (Ferriero, 2004; Lee et al., 2013; Stanaway et al., 2018). Therapeutic hypothermia is the mainstay of current therapy but is only available to a fraction of children who have moderate to severe acute HIE and are near a facility capable of initiating it in the immediate perinatal period (Ahearne et al., 2016). Children initially classified as having mild HIE may still experience adverse outcomes (de Haan et al., 2006; van Kooij et al., 2010; Eunson, 2015; Reiss et al., 2019; Schreglmann et al., 2020; Finder et al., 2020). These mildly affected children do not qualify for therapeutic hypothermia. In addition to term infant HIE, HIE in preterm infants is a poorly understood entity that predisposes children to NDDs (Schmidt and Walsh, 2010; Gopagondanahalli et al., 2016). Preterm infants with HIE are also not eligible for therapeutic hypothermia. Given the need for targeted therapies, we must develop animal models that can recapitulate the full spectrum of phenotypes exhibited by children affected by prenatal hypoxia.

Existing models of transient hypoxia in rodents have focused primarily on its effect on the postnatal brain (postnatal day 7–10) (Rice et al., 1981; Sun et al., 2016) because this postnatal period is considered to correlate neuroanatomically to term human infants (Semple et al., 2013). Models such as the Vannucci model involve unilateral ligation of the carotid artery with subsequent exposure to hypoxia (Rice et al., 1981; Sun et al., 2016). In addition to recapitulating human neuroanatomy, this postnatal method makes studying the combined effects of hypoxia and ischemia more technically feasible.

While there are benefits to postnatal rodent models of transient hypoxia, these have limitations. Importantly, postnatal models do not account for maternal and placental factors that may modulate the fetal response to hypoxia. The in utero environment is hypoxic at baseline (Trollmann et al., 2008), and many studies investigating the hypoxic response have revealed complex interactions between oxygen control and injury outcome (Tomita et al., 2003; Chen et al., 2008; Sheldon et al., 2009; Sheldon et al., 2014; Arthuis et al., 2017; Xu et al., 2019; Peebles et al., 2020; Zhang et al., 2021). Additionally, postnatal models involving carotid artery ligation followed by hypoxia result in severe injury, characterized by significant levels of cell death (Rice et al., 1981), and not the milder injury seen in many human patients (Lee et al., 2013). Finally, while the postnatal rodent brain correlates anatomically with term human infants, some functional networks mature faster in postnatal rodents than in term infants (Feather-Schussler and Ferguson, 2016). At birth, mice exhibit neuroanatomy analogous to infants born at 23–26 weeks gestation (Semple et al., 2013). However, they do not display respiratory or feeding issues common in children born before 34 weeks gestation (Zhao et al., 2019). Additionally, mice are ambulatory around postnatal day 10, a milestone human children do not achieve until approximately 12 months old (Feather-Schussler and Ferguson, 2016; Aziz et al., 2018).

Although previous mouse models of prenatal hypoxia have been described (Baud et al., 2004; Mallard and Vexler, 2015), many are primarily used to study the effects of chronic hypoxia throughout gestation. Other prenatal models of transient injury are more challenging to execute (such as late gestation uterine artery ligation) (Kubo et al., 2017). Sheep models of transient and mild hypoxia-only injury require specialized equipment, ample resources, and are not amenable to genetic manipulation (McClendon et al., 2017; McClendon et al., 2019). Here, we characterize a protocol to induce transient, mild, hypoxic injury in prenatal mice.

Previously, we have used this model to characterize the effect of late gestation (embryonic day 17.5) transient prenatal hypoxia (5% FiO₂) on long-term neurodevelopmental and anatomical features in mice (Cristancho et al., 2022). Our studies showed increased levels of hypoxia-inducible factor 1 alpha (a marker of hypoxic exposure) in the brains of fetal mice exposed to prenatal hypoxia. Mice exposed to hypoxia had decreased weights at postnatal day 2. Hypoxic exposure also led to decreased seizure threshold in mice. In addition, both male and female hypoxic mice showed abnormalities in grip strength and repetitive behaviors. Finally, male hypoxic mice had increased anxiety-like behaviors, while female hypoxic mice showed abnormalities in social interaction. We hope that future researchers may use this model to investigate the mechanisms underlying mild and preterm HIE and elucidate the maternal–placental–fetal factors underlying neurodevelopmental outcomes in preterm children exposed to transient hypoxia.