The role of the extracellular matrix in neuronal development and perinatal brain injury

Abstract

Hyaluronan (HA) is a glycosaminoglycan (GAG) that forms an integral part of the extracellular matrix (ECM). HA is synthesised by HA synthases (HAS1–3) and degraded by hyaluronidase (Hyal) enzymes. It is abundant in the brain, particularly in the cerebral cortex, where it forms the major scaffold of ECM structures around neurons, termed perineuronal nets (PNNs). Although HA and PNNs are important for neuronal function in the mature brain, the specific expression and functions of HA and PNNs in the developing CNS under physiological and pathophysiological conditions remain largely unknown. The specific aims of this thesis were to: (1) characterise HA, HAS, Hyal, and PNN expression in developing cortical neurons in vitro, (2) determine the short-term effect of HAS inhibition on neuronal morphology in vitro, and (3) determine the longer term effect of selective HAS knockdown on neuronal morphology in vitro. The final aim (4) was to assess the effects of cerebral hypoxia-ischemia (HI) on PNN and interneuron expression in the term-equivalent fetal sheep. Primary cortical neuronal cultures were established from embryonic age (E)16 rats (Aims 1–3). For Aim 1, neurons were collected at 0–21 days in vitro (DIV) and processed for quantitative real-time PCR (qPCR) or immunocytochemistry. HA, HAS, and Hyal expression were assessed at each time point. Next, the role of neuronal HA production in neuronal development was assessed by pharmacological blockade of HA synthesis with 4-methylumbelliferone (4-MU) (Aim 2) or transfection with short hairpin RNA (shRNA) constructs for HAS2 or HAS3 knockdown (Aim 3). For these studies, neurons were imaged live or labelled with structural markers, and then traced using Neurolucida software to assess morphological changes at DIV1–3 or DIV7. Aim 4 utilised the near-term (0.85 gestation) fetal sheep model of HI induced by 30 min bilateral carotid artery occlusion. Brains were perfused and collected after 7 d, and expression of interneurons and PNNs was quantified in the parasagittal cortex. Cortical neurons were capable of independent HA synthesis throughout development (DIV0–21), with expression of HA and HASs on lamellipodia, filopodia, immature neurites, differentiated processes, and synapses (Chapter 3). Notably, HA was expressed earlier and more widely than PNNs. Neurons also expressed Hyal1–3 mRNA. In Chapter 4, pharmacological inhibition of HAS activity with 4-MU reduced HA expression and impaired lamellipodia and filopodia at DIV1–3. In Chapter 5, selective HAS2 or HAS3 knockdown reduced HA expression and impaired the growth and complexity of differentiated processes at DIV7. In Chapter 6, HI in the near-term fetal sheep reduced the density of interneurons, PNNs, and decreased the percentage of surviving interneurons with PNNs. This thesis provides novel evidence that cortical neurons themselves synthesise HA independent of glia, and demonstrates that this HA important for the development of lamellipodia and filopodia, as well as process outgrowth and arborisation, of cortical neurons in vitro. Together, these data indicate that neuronal HA production represents a novel mechanism by which neuronal morphology is regulated throughout critical developmental stages. This thesis also suggests that loss of cortical interneurons and PNNs may contribute to the pathogenesis of HI, and that PNN protection or restoration may improve outcomes after perinatal brain injury.