Response to Quinlan and Nilsson: Astroglia sitting at the controls?

Abstract

We agree with Roy Quinlan and Michael Nilsson that selective disruption of intermediate filaments using a ‘magic bullet’, such as RNA interference (RNAi) technologies or peptide inhibition, might improve regeneration in the retina [1]. Quinlan and Nilsson also raised the issue of the wider implications of our data for neuroregeneration in humans. Our work suggests that astroglia are pivotal in determining the permissiveness of the adult mammalian CNS for at least some aspects of neuroregeneration, such as the migration or integration of immature neurons and neurite outgrowth. Consistent with this possibility, Fred Gage and colleagues suggested that adult neurogenesis occurs only at specific regions in the CNS (i.e. the subventricular zone and the dentate gyrus of the hippocampus) because astrocytes allow it there and not elsewhere [2]. Understanding on a molecular level how astrocytes induce or restrict regeneration might provide the key to controlling regeneration in the clinic. Comparison of gene expression in the regeneration-supporting environment of GFAPVim mice with that in the standard regeneration-restrictive adult CNS might lead to the identification of molecular pathways that control CNS regeneration. Naturally, such knowledge might provide targets that are more amenable to therapy than intermediate filaments themselves. Why is the regenerative capacity of the CNS limited, and is it even wise to contemplate playing with it? We would like to think that the increasing complexity of multicellular organisms – in particular, the development of the ability to acquire, process, store and use information – required tight control over the plasticity of the neuronal network. Hence, evolution might have opted for less regeneration. This might mean that a system of brakes on regeneration evolved in higher organisms. If so, where should we look for them? One such brake might be neuron-specific. The expression profiles of many genes in mammalian CNS neurons change during ontogenesis. In several cases, this change coincides with the loss of neuronal regenerative capacity around birth [3,4]. For example, in rodents Bcl-2 expression is high during an early developmental stage and declines at birth [5]. In mice, overexpression of Bcl-2 prolonged the regeneration-permissive window after birth [6]. Conceivably, the first brake on neuroregeneration is an intrinsic program that makes it very difficult for neurons to regenerate after a certain point during ontogenesis. We propose that mature astrocytes constitute a second brake. Considered since Rudolf Virchow’s time to be merely glue between the active and multifunctional neurons, astroglia increasingly appear to be at the controls. They modulate synaptogenesis [7], set the threshold for neuronal signaling [8] and have multiple roles in CNS diseases. In addition, immature glialfibrillary acidic protein (GFAP)-positive glial cells seem to be the predominant stem cells in the adult CNS [9]. Astrocytes become activated in essentially all known CNS pathologies [10]. This ‘reactive gliosis’ remains enigmatic but its complexity certainly extends beyond the questionable century-old concept of a ‘glial scar’ as a mechanical obstacle to axonal regeneration. Previously, we showed that absence of intermediate filaments in astrocytes affects the quality of post-traumatic glial scars [11,12]. In the highlighted paper [13], we show that it renders the CNS environment more permissive to the integration of transplanted cells and the outgrowth of their neurites. Thus, mature astrocytes with their characteristic GFAP expression might constitute the second brake on CNS regeneration, a brake that becomes progressively stronger as the CNS matures and ages. Is there a third brake on CNS regeneration? Many have been proposed, including the presence of myelin-associated growth-inhibiting molecules or the lack of neurotrophic and growth-supporting factors [14]. We suggest looking also in the immune system. A growing body of experimental evidence shows that the immune response can promote tissue repair in CNS injury and disease [15,16]. In the optic nerve crush model, simultaneous injury of the eye lens enhances the regeneration of axons of the ganglion cells [17]. Lens injury leads to macrophage accumulation in the vitreous body, and intravitreal injections of a monocyte-activating agent exert a comparable regeneration-promoting effect [17]. Macrophage-conditioned medium also induced axonal regeneration in vitro [18]. However, activated microglia inhibit both baseline and injury-triggered hippocampal neurogenesis in the adult [19]. Interestingly, the complement system, an evolutionarily ancient arm of humoral immunity, has recently been linked to regeneration outside the CNS [20,21]. Although the picture is still very incomplete, modulation of the immune response might prove to be important in controlling neuroregeneration in the adult. Corresponding authors: Milos Pekny (milos.pekny@medkem.gu.se), Dong Feng Chen (dfchen@vision.eri.harvard.edu). Update TRENDS in Neurosciences Vol.27 No.5 May 2004 243