Microglia and Neonatal Brain Injury

Abstract

Microglial cells are now recognized as the “gate-keepers” of healthy brain microenvironment with their disrupted functions adversely affecting neurovascular integrity, neuronal homeostasis, and network connectivity. The perception that these cells are purely toxic under neurodegenerative conditions has been challenged by a continuously increasing understanding of their complexity, the existence of a broad array of microglial phenotypes, and their ability to rapidly change in a context-dependent manner to attenuate or exacerbate injuries of different nature. Recent studies have demonstrated that microglial cells exert crucial physiological functions during embryonic and postnatal brain development, some of these functions being unique to particular stages of development, and extending far beyond sensing dangerous signals and serving as antigen presenting cells. In this focused review we cover the roles of microglial cells in regulating embryonic vasculogenesis, neurogenesis, and establishing network connectivity during postnatal brain development. We further discuss context-dependent microglial contribution to neonatal brain injuries associated with prenatal and postnatal infection and inflammation, in relation to neurodevelopmental disorders, as well as perinatal hypoxia–ischemia and arterial focal stroke. We also emphasize microglial phenotypic diversity, notably at the ultrastructural level, and their sex-dependent influence on the pathophysiology of neurodevelopmental disorders.

INTRODUCTION

The microglial field exploded over the past 6–8 years (Tremblay et al., 2015). Literature keeps growing identifying microglial cells (microglia) as “gate-keepers” of a healthy brain microenvironment and cell–cell communications under physiological conditions. The perception of these cells as purely toxic under neurodegenerative conditions has been challenged by demonstration of the existence of a broad array of microglial phenotypes having the ability to rapidly change in a context-dependent manner to attenuate or exacerbate injuries of different nature. Recently developed tools have allowed for discrimination between microglia and monocytes and demonstration of respective markedly different gene “signatures” (Butovsky et al., 2014; Bennett et al., 2016). Together with accumulating data on the presence of microglial sexual dimorphism (Hanamsagar et al., 2017) and the modulatory role of microglial homeostasis (Sipe et al., 2016; Mildner et al., 2017; Zrzavy et al., 2017), these new findings are major milestones for reconsideration of the physiological role of microglia and their contribution to neurodegenerative diseases.

Recent studies demonstrated that microglia have a broad range of effects in embryonic and postnatal physiological brain development, including regulation of embryonic vasculogenesis, secretion of trophic factors, immuno-surveillance, oligodendrogenesis and neurogenesis. They play key roles in establishing brain connectivity network during postnatal brain development (Paolicelli et al., 2011; Arnold and Betsholtz, 2013; Butovsky et al., 2014; Hagberg et al., 2015). Microglial cells are now known to undergo a major phenotypic transformation during physiological postnatal brain development (Butovsky et al., 2014), confirming the different morphological and ultrastructural phenotypes that were previously identified (Murabe and Sano, 1982; Dalmau et al., 1998), and to develop distinctly between male and female brains (Schwarz et al., 2012). Pharmacologic depletion of microglial cells or deficiency/malformation of individual signaling pathways in these cells were shown to play context-dependent roles in perinatal infection (Baburamani et al., 2014), hypoxia–ischemia (H–I) (Hagberg et al., 2015), arterial stroke (Faustino et al., 2011; Fernandez-Lopez et al., 2016), and developmental disorders (Derecki et al., 2012; Hagberg et al., 2012). In this review we will focus on new or recently revisited concepts pertaining to structural and functional aspects of microglia in the immature brain, effects that can lead to the age- and context-dependent microglial behavior under physiological and pathophysiological conditions.

MICROGLIAL CELLS GUIDE BLOOD-BRAIN BARRIER DEVELOPMENT IN EMBRYONIC BRAIN

In mice, the origin for microglial and monocyte pools is distinct (Ginhoux et al., 2010; Schulz et al., 2012) and the establishment of local brain microglial pool temporally occurs before blood–brain barrier (BBB) development, which begins mid-gestation and before pericytes and astrocytes take their respective positions surrounding the vessels (Cahoy et al., 2008; Armulik et al., 2010; Daneman et al., 2010; Zhang and Barres, 2010). Microglia have direct effects on vasculogenesis and vascular sprouting in embryonic brain (Checchin et al., 2006; Kubota et al., 2009). They guide endothelial sprouts in fetal brain, largely via VEGF-dependent mechanisms. Microglial deficiency adversely affects vasculature development (Fantin et al., 2010). Disrupted PU.1 or SDF-1 signaling in microglia/macrophages leads to distorted vasculature and embryonic lethality (Fantin et al., 2010). Mice deficient in VEGF/VEGF receptor 2 fail vessel formation, leading to embryonic lethality (Shalaby et al., 1995). Even lack of a single VEGF allele leads to embryonic lethality despite partial blood vessel development (Carmeliet et al., 1996). Disrupted integrin and chemokine signaling was shown to adversely affect embryonic angiogenesis and BBB formation. For example, excessive vascular sprouting underlies cerebral hemorrhage in mice that lack αVβ8-integrin/transforming growth factor beta (TGFβ) signaling in the brain, whereas integrin/TGFβ signaling suppresses vascular branching/sprouting and prevents germinal matrix hemorrhage (Arnold et al., 2014). In early postnatal rat and mouse brain following stroke, pharmacological inhibition of the TGFβ1/TGFbR2 pathway selectively in microglia disrupts the BBB, which triggers hemorrhages in injured regions (Fernandez-Lopez et al., 2016). Importantly, monocytes are unable to substitute for the microglial roles in vasculogenesis in the developing brain (Checchin et al., 2006). The latter finding is consistent with different origins of microglia and monocytes and distinct signaling in these cell subtypes (Ginhoux et al., 2010; Butovsky et al., 2014). Together, these data suggest a principle role of microglia in vascular development and integrity and key role of microglial TGFβ1 signaling in modulating BBB integrity during brain development.

MICROGLIA, POSTNATAL BRAIN DEVELOPMENT AND BRAIN CONNECTIVITY

In human brain, the presence of ameboid CD68-positive microglia (i.e., feature of microglia with activated lysosomal component) has been reported as early as 13 weeks of gestation, with the highest number observed between 13 and 18 gestational weeks in the germinal matrix (Hutchins et al., 1990) and white matter (Monier et al., 2006). Topographic relationship of CD68-positive cells with growing axons was demonstrated (Cho et al., 2013). In the infant brain, ramified microglial cells were reported as the predominant type (Fujimoto et al., 1989; Cho et al., 2013).

Echoing these findings in human brain, in rodents, ameboid (considered “activated”) morphological phenotype is seen during the embryonic period and in newborn rodents when physiological death of overproduced neurons is high, followed by gradual microglial ramification during the first week of rodent’s life. Morphological and ultrastructural analyses in the neonatal rat brain conducted during the first week provided insights into this transformation of ameboid microglia into ramified cells, among the rat subventricular zone, corpus callosum, and cerebral cortex. Different stages of microglial maturation were thus described, ranging from an “ameboid” morphology in the white matter, to a “round” morphology with a few pseudopodia but devoid of processes located in the subventricular zone and nearby blood vessels, to a “primitive” morphology harboring a few processes in the cerebral cortex, hippocampus, and hypothalamus (Murabe and Sano, 1982; Dalmau et al., 1998). These “nascent” phenotypes showed microglia-distinctive long stretches of endoplasmic reticulum, numerous ribosomes, and an abundant accumulation of vacuoles and large phagocytic inclusions (Ferrer and Sarmiento, 1980; Murabe and Sano, 1982).

Microglia have been shown to regulate brain circuit connectivity in multiple ways. During embryonic neurogenesis they regulate stem cell pool via secretion of trophic factors and phagocytosis (Cunningham et al., 2013). Microglial involvement with the outgrowth of dopaminergic axons and proper positioning of interneurons in the forebrain was shown using cell-depletion approaches, as well as Cx3Cr1, CR3, and DAP1 mutants (Squarzoni et al., 2014). Findings in mouse co-culture indicate that neurogenesis does not occur without microglia (Walton et al., 2006). Microglia provide trophic support to neurons and endothelial cells, notably via their production of BDNF, IGF-1/2, TGFβ. Disrupted growth factor production in microglia interrupts cortical layer formation (Ueno et al., 2013) and synaptic refinement during the neonatal to adolescent period (Bialas and Stevens, 2013; Butovsky et al., 2014). Microglia contribute to somatosensory cortex development during the “critical period” in mice (Arnoux et al., 2013). Functional analysis of thalamocortical synapses showed that Cx3Cr1 deficiency delays maturation of postsynaptic glutamate receptors (normally occurs in postnatal day 6–8, P6-P8 mice) (Arnoux et al., 2013). Cx3cr1 knockout mice transiently have fewer microglia in the hippocampus, higher dendritic spine density and less mature physiological responses and long-term depression (LTD) (Paolicelli et al., 2011; Zhan et al., 2014). Microglia also have morphological alterations and diminished response to the migratory and activating signal ATP (Pagani et al., 2015). Phagocytosis of overproduced neurons and weak synapses by microglia, processes that govern brain connectivity, plays key role in organizing brain circuits during the postnatal period (Paolicelli et al., 2011; Hoshiko et al., 2012; Schafer et al., 2012; Bialas and Stevens, 2013; Zhan et al., 2014). Direct microglial contacts with axon terminals and dendritic spines were revealed by electron microscopy, with our quantitative analyses showing that almost all (~94%) microglial processes appose synaptic elements in adolescent mouse cerebral cortex (Tremblay et al., 2010b). 3D reconstruction also revealed that a single microglial process can make multiples contacts with multiple synaptic elements at multiple synapses, sometimes with finger-like protrusions wrapping around spines and terminals (Tremblay et al., 2010a). Clathrin-coated pits are observed inside of microglial processes or synaptic elements, at their sites of contact, suggesting reciprocal exchange of molecular signals through clathrin-mediated endocytosis of membrane-bound receptors and their ligands. The two types of morphological specializations indicated that surveilling microglia do interact functionally with excitatory synapses (Tremblay and Majewska, 2011). While dynamic microglial contacts with axon terminals and dendritic spines were also observed in vivo (Tremblay et al., 2010a), live evidence for their engulfment is currently lacking. Functional studies nevertheless demonstrate that surveilling microglia contribute to activity-dependent synaptic pruning, particularly via the classical complement cascade. Disruption of microglia-specific CR3/C3 complement signaling leads to sustained deficits of synaptic connectivity during adolescence and adulthood (Bialas and Stevens, 2013).

MICROGLIAL CELLS MEDIATE DISTINCT RESPONSES IN THE INFECTIOUS AND STERILE INFLAMMATORY SETTINGS

The fetus and newborn infant are highly sensitive to infection and chorioamnionitis (inflammation of fetal membranes) are linked to both preterm birth and later neurological deficits in the newborn (Shatrov et al., 2010; Salas et al., 2013; Soraisham et al., 2013; Shankaran et al., 2014). Similarly, infants with early or late neonatal sepsis have an increased risk of neurological morbidity. Children born very preterm with a proven bacteremia in the first weeks of life perform worse on neurocognitive tests (Bright et al., 2017) and demonstrate neurological dysfunction at school age (Kavas et al., 2017). Postmortem studies show expression of several pro-inflammatory cytokines in the brains of infants with proven infection and white matter injury, suggesting that the systemic inflammation also spreads to the brain (Kadhim et al., 2001). Thus, there is significant clinical evidence to suggest that inflammation is an important factor in neonatal brain disorders. Exactly how the brain inflammation arises, however, remains unclear. Trafficking of peripheral immune cells may contribute to the CNS inflammatory condition, but activation of resident microglia is likely to play an important role.

Based on comparisons of the timing of neurogenesis, synaptogenesis, gliogenesis, and myelination, together with age-dependent molecular and biochemical changes in rodents and humans, the rodent brain at P1–P5 corresponds to 23–32 weeks of gestation in humans, making it suitable for studies of preterm brain injury, whereas the rodent brain at P7–P10 resembles more 36–40 weeks of gestation in humans, and, thus, is better suited for studying brain injury at term (Semple et al., 2013). Brain injury and/or immune stimuli are associated with a marked induction of innate immunity via Toll-like receptors (TLRs) (Mallard et al., 2009; Stridh et al., 2011). Activation of TLRs by administration of synthetic ligands or microorganisms known to stimulate the receptors has been used extensively to study effects on microglia in the fetus or newborn. In early studies, it was reported that although expression of cytokines was increased in the brain of rat pups that had been exposed to the TLR4 agonist LPS (endotoxin) in utero, microglial numbers were either reduced or not affected (Cai et al., 2000; Bell and Hallenbeck, 2002). However, in later studies, using different markers to assess microglia (CD68), E. coli bacteria (which activate TLR-4) inoculation of uteri in rats at E17 demonstrated marked activation/morphologic transformation of microglia in pups up to two weeks after birth (Pang et al., 2005). In fetal sheep, LPS given at a gestational age which is comparable to the preterm human infant with respect to brain development, either to the fetus (Duncan et al., 2002; Mallard et al., 2003; Dean et al., 2011) or into the amniotic fluid (Nitsos et al., 2006), results in CNS inflammation, including the overall increased numbers of microglia and those with activated morphologic phenotype. Furthermore, intrauterine LPS to pregnant rabbits showed increased brain retention of (11)C-(R)-PK11195, a marker of neuroinflammation, on P1. Immunohistochemical assessment of these brains demonstrated increased numbers of microglia and morphological changes associated with an activation state in the periventricular region and hippocampus (Kannan et al., 2007). In the same model, the authors recently demonstrated that glutamate carboxypeptidase II (GCPII) is significantly increased in ionized calcium binding adaptor molecule 1 (Iba1)-positive microglia with morphological features of activation, suggesting that maternal intrauterine LPS exposure may mediate glutamate dysregulation in a microglia-dependent fashion (Zhang et al., 2016). Along the same line, intrauterine injection of LPS to E17 mice showed increased number of Iba1-positive cells with an ameboid shape in the offspring at P45, which was associated with increased synaptic strength, due in part to an increase in the probability of glutamate release from presynaptic CA3 axon terminals (Kelley et al., 2017). Thus, microglial activation following immune stimulation may not only play a role in inducing CNS inflammation, but also directly affect neurotransmission in the developing brain. Maternal immune activation by injection of Poly(I:C) at 15 was shown to lead to sustained microglial transcriptome reprograming through the adulthood (Mattei et al., 2017).

Systemic inflammation results in increased levels of cytokines and chemokines in the blood, including TNFα and IL-1β, pleiotropic cytokines known to also be produced by cells in the CNS, including microglia (Saliba and Henrot, 2001). Although these cytokines normally have important physiological and endocrine functions in the brain, when produced in excess they can have microglia-mediated cytotoxic effects. As an example, repeated systemic administration of IL-1β in P1-P5 mice results in a transient increase in microglial density and long-term myelination deficits, which are accompanied by cognitive defects (Favrais et al., 2011). Neonatal systemic activation of innate immune receptors like TLRs is also associated with brain dysfunction and can further activate microglia. We showed that administration of LPS at P5 leads to a transient increase in absolute number and cell density of Iba1-positive microglia, as well as a persistent alteration in hippocampal inflammatory status in microglia, but without driving infiltration of peripheral monocyte-derived macrophages (Smith et al., 2014). Further, repeated injections of the TLR2 agonist Pam(3)CSK (4) at P3-P11 decreased the volume of cerebral gray and white matter in association with elevated levels of several pro-inflammatory cytokines and chemokines in the brain as well as increased microglial density (Du et al., 2011). It was subsequently found that systemic TLR2 activation, but not TLR4, results in infiltration of peripheral immune cells into the brain, demonstrating that monocytes and neutrophils may contribute to the inflammatory response in the brain under specific conditions (Mottahedin et al., 2017).

It is clear from these studies that microglia are responders following antenatal and postnatal systemic infectious (e.g. TLR activation), as well as non-infectious (e.g. IL-1β) inflammation. However, the mechanisms that transfer peripheral signals to activate microglia remain obscure. Microglia are the predominant source of TLR2 expression during normal postnatal brain development (Lalancette-Hebert et al., 2017). Our recent comparison of TLR2 activation between systemic (i.p.) LPS stimulation and intracerebral IL-1β injection in P9-P10 dual-reporter luc/GFP-TLR2 mice, has demonstrated that LPS triggers strong TLR2-mediated microglial activation and increase in production of pro-inflammatory cytokines and chemokines (Lalancette-Hebert et al., 2017). In contrast, following IL-1β injection, the in vivo TLR2-luc signal is reduced and the inflammatory response is distinct and overall lessened (Lalancette-Hebert et al., 2017), suggesting that the microglia response and underlying mechanisms of inflammation may be different depending on the nature of initial systemic infectious/inflammatory stimuli.

MICROGLIA, INFECTIONS AND NEURODEVELOPMENTAL DISORDERS

Emerging clinical and experimental evidence suggests that the immune system in children with autism spectrum disorders (ASD) is altered, affecting the child’s brain development and function (Ashwood et al., 2011; Hsiao, 2013; Lucchina and Depino, 2013), and that TLR2/3/4 activation is exacerbated and contributes to the immunological component of ASD (Enstrom et al., 2010). In mice, changes in microglial gene expression were demonstrated to lead to neurodevelopmental disorders and a primary deficit in microglia to be sufficient to induce autistic ASD-like behavior (Zhan et al., 2014). Mice lacking Cx3cr1 exhibit a transient reduction of microglial number during the early postnatal period and associated deficit in synaptic pruning, as mentioned above, leading to decreased functional brain connectivity, deficits in social interaction, and increased repetitive-behavior phenotypes (Zhan et al., 2014; Fernandez de Cossio et al., 2017). A viral mimetic Poly(I:C) in mice and monkeys used as an immune challenge (Giulivi et al., 2013; Naviaux et al., 2013; Bauman et al., 2014) produced immune dysfunction and long-lasting ASD-like behavior (Giulivi et al., 2013). Compared to the offspring males from saline-injected mothers, those from immune challenged mothers displayed decreased preference for the social chamber and extremely high repetitive behavior in the marble burying test (Malkova et al., 2012). A link between the nucleotides ATP and ADP, which, when released to the extracellular space, affect innate immunity and inflammation and autistic behavior, was also established in the Poly(I:C) mouse model (Naviaux et al., 2013), suggesting the possibility that maternal immune activation mediates ASD specifically by reprograming microglial signaling. Prenatal infection leads to ASD-like behavior and altered synaptic pruning in the mouse offspring (Fernandez de Cossio et al., 2017). Literature is growing linking disruptions in microglia-mediated synaptic pruning to various neurodevelopmental and neuropsychiatric disorders including schizophrenia (reviewed in (Salter and Beggs, 2014; Neniskyte and Gross, 2017). Not only maternal immune activation, but also early life stress, have been shown to perturb the maturation of microglia in developing hippocampus (Delpech et al., 2016).

MICROGLIAL ROLE IN PERINATAL FOCAL ARTERIAL STROKE AND HYPOXIA–ISCHEMIA

Neuroinflammation is a characteristic feature of stroke progression in the adult and is a major contributor to brain injury (Iadecola and Anrather, 2011). Parenchymal, perivascular and peripheral circulating cells can contribute independently and in concert to stroke-induced production of inflammatory mediators and neuroinflammation (Iadecola and Anrather, 2011) as well as activation of endothelial cells (Osborn et al., 1989; Stanimirovic et al., 1997). Historically, in adult stroke, microglia were viewed as purely injurious, in part due to production of inflammatory mediators, reactive oxidant species (ROS) and other toxic molecules (reviewed in (Iadecola and Anrather, 2011)). Reconsideration of the microglial role in stroke is on-going following recent discoveries of the heterogeneity of the microglial pool (Lucin and Wyss-Coray, 2009), the microglial role in supporting neuronal health and neurogenesis (Walton et al., 2006), the distinct and even opposite roles of microglia and monocytes in stroke (Lambertsen et al., 2009), and the demonstration that lack of microglia disrupts neuronal network activity and exacerbates injury after stroke (Szalay et al., 2016). Furthermore, the importance of timing after stroke as a factor influencing microglial effects is being increasingly acknowledged, while the usefulness of classifying microglia into M1/M2A/2B phenotypes is being challenged as oversimplified and, as such, misleading to the in vivo situation (Martinez and Gordon, 2014; Murray et al., 2014; Ransohoff, 2016).

There is now ample evidence that the mechanisms of ischemic injury differ greatly between immature versus adult brain (reviewed in (Yager and Ashwal, 2009; Fernandez-Lopez et al., 2014)), including neuroimmune responses (Hagberg et al., 2015). Several lines of investigation have pointed to markedly different, and even opposite, effects of microglia to similar insults in neonatal as compared to adult brain (Cho et al., 2005; Lalancette-Hebert et al., 2009, 2017; Woo et al., 2012). For example, the scavenger receptor CD36 contributes to acute injury after transient middle cerebral artery occlusion (tMCAO) by enhancing ROS production in microglia with activated morphologic phenotype (Cho et al., 2005) whereas CD36 attenuates acute injury after tMCAO in P9 mice (Woo et al., 2012). Another example is ~20-fold increase in TLR2 expression measured 24 h following tMCAO in adult luc/GFP-TLR2 mice (Lalancette-Hebert et al., 2009), whereas no increase in TLR2 expression, which is predominantly seen in microglial cells, is detected 24–72 h after tMCAO in P9 luc/GFP-TLR2 mice (Lalancette-Hebert et al., 2017).

In the neonatal brain, microglial cells undergo morphologic transformation after both H–I and focal arterial stroke. Activated microglia/macrophages are thought to contribute to H–I (McRae et al., 1995; Ivacko et al., 1996; Bona et al., 1999; Xu et al., 2001; Cowell et al., 2003) and excitotoxic injury (Dommergues et al., 2003). The demonstration that genetic deletion of IL-18 or caspase-1, which are produced predominately by activated microglia, attenuates brain injury also supports the notion that microglia exert injurious effects early after H–I (Hedtjarn et al., 2002). The initial demonstration that the tetracycline derivative minocycline could markedly protect the immature brain from H–I (Arvin et al., 2002) was also attributed to suppression of toxic microglial activation. However, minocycline’s actions are not specific to microglia and its effects after neonatal H–I, either protective or injurious, were later shown to be species-dependent (Tsuji et al., 2004). After tMCAO induced in P7 rats minocycline provides only modest and short-term protection (Fox et al., 2005).

Contrary to the notion of microglial toxicity in neonatal brain following H–I, in acute neonatal stroke we demonstrated a protective role of microglia by showing that selective pharmacologic depletion of these cells by intracerebral injection of clodronate-liposomes two days before inducing tMCAO in P7 rats (Faustino et al., 2011) or P9 mice (Fernandez-Lopez et al., 2016) changes balance between caspase-3-dependent and -independent neuronal death (Faustino et al., 2011), limits engulfment and removal neuronal debris (Woo et al., 2012) and exacerbates injury (Faustino et al., 2011; Woo et al., 2012). Phagocytosis is particularly important in neonatal post-ischemic brain because the magnitude of apoptotic neuronal death is markedly higher in neonatal than in adult post-ischemic brain (Hu et al., 2000). A lack of microglia further increases the cytokine and chemokine levels induced by tMCAO in neonatal rats (Faustino et al., 2011), consistent with the notion of protective role of apoptotic debris removal. Another line of evidence for the vital role of microglial phagocytosis of dying neurons as a part of protection from neonatal stroke comes from the data that genetic deletion of the scavenger receptor CD36, receptor that contributes to several phagocytotic steps, makes acute injury worse (Woo et al., 2012; Li et al., 2015). The lack of the CD36 substantially reduces engulfment of neurons that express cleaved caspase-3 and increases production of inflammatory cytokines. Interestingly, in contrast to the observed superoxide accumulation in activated microglia after stroke and attenuation of superoxide accumulation in injured adult CD36 knockout mice, no significant increase in superoxide accumulation is detected in microglia from wild-type and CD36 knockout mice subjected to tMCAO at P9 (Woo et al., 2012). Cumulatively, these data show that CD36 may function differently in the neonate.

Another important aspect of microglial function in the neonatal brain is preservation of the neurovascular integrity and protection against hemorrhagic transformation after neonatal stroke (Fernandez-Lopez et al., 2016). Microglial depletion triggers BBB leakage and induces hemorrhages, reduces microglia-derived TGFβ1 levels and the phosphorylation and intracellular distribution of SMAD2/3, but does not significantly alter expression or intracellular redistribution of several tight junction proteins. Selective inhibition of TGFbR2/ALK5 signaling in microglia via intra-cerebral liposome-encapsulated SB-431542 delivery triggers hemorrhages after tMCAO, demonstrating that TGFβ1/TGFbR2/ALK5 signaling in microglia protects from hemorrhages. Consistent with these observations in neonatal rats, depletion of microglia before tMCAO in P9 Cx3cr1^GFP/+/Ccr2^RFP/+ mice exacerbates injury and induces hemorrhagic transformation at 24 h (Fernandez-Lopez et al., 2016). Altogether these data suggest an array of microglial effects after neonatal brain ischemia related injury. While phagocytosis of neuronal debris and the newly discovered ability of microglial cells to protect neurovascular integrity serve as protective mechanisms, the inflammatory and ROS responses can contribute to injury. It is not clear whether the divergent results relate to different models of neonatal brain injury (i.p., the presence of systemic hypoxia in H–I model) or to time-resolved differing microglial phenotypes that participate in an acute and chronic injury. It is also essentially unknown whether a particular microglial subpopulation provides endogenous cerebrovascular protection.

SEXUAL DIMORPHISM AND MICROGLIA IN HEALTHY BRAIN DEVELOPMENT AND IN DISEASED IMMATURE BRAIN

Perinatal stroke and cerebral palsy are much more common in boys than in girls (Johnston and Hagberg, 2007; Nunez, 2012). Males are also at significantly higher risk than females of developing ASD and early onset forms of schizophrenia. Importantly, sex and sex hormones may have independent effects in stroke in children (Normann et al., 2009; Vannucci and Hurn, 2009) and juvenile mice (Herson et al., 2013). Although multiple sex-dependent mechanisms have been demonstrated, pertaining to neuronal death following neonatal H–I and focal ischemia, PARP-1 versus caspase-3 dependent (Hagberg et al., 2004; Renolleau et al., 2007, 2008), the response to anti-inflammatory therapy after perinatal brain injury (Nijboer et al., 2007; Fleiss et al., 2012) and divergent roles of innate immune receptors (Pimentel-Coelho et al., 2013), the underlying sex-based responses in injured neonatal brain continue to be insufficiently understood. Several studies revealed that microglia are sexually dimorphic (Schwarz et al., 2012; Lenz et al., 2013; Hanamsagar et al., 2017) and that the number of microglial cells is significantly higher in males than in females among several brain regions, including the hippocampus and parietal cortex (Schwarz et al., 2012; Lenz et al., 2013). Injury in Cx3Cr1 knockouts was shown to be affected by sex (Pimentel-Coelho et al., 2013) and sex-specific microglial functions were implicated in ASD (Derecki et al., 2012). Thus, the existence of intrinsic sex-specific differences in the neonatal period warrants studies in both sexes.

SUMMARY AND PERSPECTIVE

The fast progressing microglial field shows that microglial phenotypes rapidly evolve with brain maturation and that their interactions with neurons, the neurovasculature and other glial cells change dynamically. Literature is growing on the role of homeostatic microglia, their receptor P2RY₁₂R in particular, in overall brain development and synaptic plasticity (Sipe et al., 2016). Microglial cells are also recognized as being sexually dimorphic, and now believed to act as a neuropathology sensor in both the developing and mature brain. However, microglial contribution to injury markedly depends on brain maturation at the time of injury, injury context and sex. The frequently used term “activated microglia” is widely applied to both injurious and beneficial functions of these cells. While evidence is growing that the “gene signature” of microglial cells differs substantially from that of monocytes and other immune cells (Butovsky et al., 2014), the lack of readily available tools to reliably distinguish microglia from differentiating monocytes in injured brain as well as dynamic changes in the microglial phenotypes in response to changed local brain microenvironment, remains a big hurdle in separating the roles of microglia from those of other immune cells in the diseased brain. It is unknown whether ultrastructural features of microglial cells in injured neonatal brain bear resemblance with dysfunctional microglia demonstrated in adult or ageing brain (Bisht et al., 2016) and whether ultrastructural morphological features are different between male and female developing brains. Such knowledge would be important for an improved understanding of neurodevelopmental disorders as well as brain and neurovascular disorders that manifest much later in life.

Abbreviations:

ASD

autism spectrum disorders

BBB

blood–brain barrier

Iba1

ionized calcium binding adaptor molecule 1

ROS

reactive oxidant species

TGFβ

transforming growth factor beta

TLRs

Toll-like receptors

tMCAO

transient middle cerebral artery occlusion