Motor neuron death in ALS – programmed by astrocytes?

Abstract

Motor neurons in ALS die via cell-autonomous and non-cell autonomous mechanisms. Using adult human astrocytes and motor neurons, Re et al (2014) discover that familial and sporadic ALS derived human adult astrocytes secrete neurotoxic factors that selectively kill motor neurons through necroptosis, suggesting a new therapeutic avenue.

Amyotrophic lateral sclerosis (ALS), commonly referred to as Lou Gehrig’s disease is an adult onset neurodegenerative disease manifested by degeneration of motor neurons in motor cortex, brain stem and spinal cord, resulting in muscle paralysis and ultimately death (Rowland and Shneider, 2001). While the disease is dominantly inherited in approximately 10% of patients (termed familial), most often the disease is sporadic with largely unknown genetic aetiology. Numerous hypotheses have been proposed to account for the disease associated motor neuron loss including conformational instability of proteins triggering neurotoxicity, perturbations of RNA processing, mitochondrial dysfunction, defective axonal transport, excitotoxicity and inflammation (reviewed in (Rothstein, 2009). However, the precise molecular mechanism(s) that target motor neurons for death and result in the pathological manifestations still remain elusive. Studies reporting mutations in the SOD1 gene encoding the antioxidant enzyme, Cu/Zn superoxide dismutase-1 provided the first genetic link to ALS (Rosen et al., 1993). Since then, studies have implicated both cell autonomous and non-cell autonomous mechanisms in SOD1 mediated toxicity (Clement et al., 2003). The identity of such cell types that modulate motor neuron survival came from studies using cultured motor neurons derived from embryonic spinal cord or differentiated from embryonic stem cells (Di Giorgio et al., 2008; Marchetto et al., 2008; Nagai et al., 2007). In these studies, co-cultured motor neurons were less likely to survive when they were on astrocytes that express mutant SOD1. Exposure of cultured motor neurons to conditioned media derived from astrocytes also affected survival of motor neurons likely through soluble factors released by mutant SOD1 expressing astrocytes. In co-culture systems with motor neurons, astrocytes derived from sporadic ALS (sALS) post-mortem spinal cord neural progenitor cells (NPCs) also selectively killed motor neurons (Haidet-Phillips et al., 2011), suggesting a shared astrocyte dependent disease mechanism between familial and sporadic forms of ALS. While these studies point to astrocytes as deadly neighbors that exacerbate motor neuron damage, the mechanism underlying astrocyte mediated toxicity towards motor neurons has remained ambiguous.

In this issue of Neuron, Re et al (2014) report that human adult astrocytes derived from motor cortex and spinal cord of sALS patients have detrimental effects on human embryonic stem cell derived motor neurons by triggering a form of regulated necrosis, termed necroptosis. By contrast, survival of these motor neurons was not affected by astrocytes derived from patients with Alzheimer’s disease or chronic obstructive respirator disorder (COPD). Co-culturing motor neurons with other cell types like fibroblasts did not have any marked effect on their survival either. Furthermore, non-motor neuron cell types of GABAergic origin were also unaffected by sALS derived astrocytes. Thus, in line with previous studies, there seems to be selective vulnerability of motor neurons to astrocyte mediated toxicity in sALS.

In transgenic familial ALS (fALS) mouse models, SOD1 mutations result in ALS like phenotypes even in the presence of endogenous SOD1 suggesting that gain of toxic functions play a causal role in the disease. Accordingly, shRNA mediated knockdown of SOD1 in fALS astrocytes fully rescues motor neuron targeted toxicity (Haidet-Phillips et al., 2011). Previous studies also argued for a pathogenic role for WT SOD1 towards motor neuron survival in sALS, potentially through aberrant post-translational modifications of WT SOD1 that may cause it to undergo conformational changes and aggregate. (Rotunno and Bosco, 2013) (Deng et al., 2006). Re et al. test this possibility by knocking down WT SOD1 in sALS astrocytes independently using four different shRNA constructs and examining the resulting effect on survival of motor neurons. Surprisingly, loss of motor neurons was not mitigated by knockdown of WT SOD1 in sALS astrocytes under these conditions. Astrocyte targeted silencing of TAR DNA binding protein 43 (TDP 43), mutations in which have been genetically linked to ALS, also failed to mitigate motor neuron degeneration. Furthermore, survival of cultured motor neurons was affected by exposure to media conditioned for 7 days with sALS derived astrocytes that induced signs of DNA fragmentation, caspase-3 activation and loss of plasma membrane integrity. These observations suggest that motor neuron death was mediated by the release of soluble factors specific to sALS astrocytes. Next, Re et al provide evidence that sALS as well as fALS derived astrocytes kill motor neurons through a Bcl-2 associated × (Bax) protein dependent death pathway. Bax is a mitochondrially translocated pro-apoptotic protein that induces cell death in response to apoptotic stimuli by permeabilizing the outer mitochondrial membrane which results in release of cytochrome C and activation of caspases. However, Re et al show that while Bax inhibition protects motor neurons exposed to sALS astrocytes, pharmacological inhibition of the initiator caspase 8 or executioner caspases 3 and 7 that are activated downstream of cytochrome C release did not promote survival of motor neurons. Thus, it appears that sALS astrocytes triggered death of motor neuron is mediated by Bax in a caspase independent manner. Such Bax dependent toxicity also seems to underlie fALS mediated motor neuron death as primary mouse motor neurons lacking Bax are resistant to mutant SOD1 induced toxicity in this study.

While activation of apoptotic pathways have been long implicated in the demise of motor neurons in ALS, it is becoming increasingly clear that apoptosis is not the only cellular mechanism that regulates cell death. Necrotic cell death, traditionally viewed as a passive process is now gaining recognition, at least in part, as a cell death program (Galluzzi et al., 2012). There are different forms of necrotic cell death that are defined biochemically. While caspases are executioners of apoptosis, they play no positive role in necrotic cell death. Re et al. show that in the absence of caspase activation, motor neuron demise triggered by sALS as well as fALS astrocytes might involve necroptosis, a specialized form of necrotic cell death that involves the activation of the kinase domain of receptor interacting protein 1 (RIP1) and subsequent recruitment of a mixed lineage kinase domain-like protein (MLKL) (Kaczmarek et al., 2013). Re et al show that treatment of motor neuron/astrocyte co-cultures with the tryptophan based allosteric inhibitor of RIP1, necrostatin-1 (Nec-1) or RIP1 specific shRNA prevented the fALS and sALS astrocyte induced loss of motor neurons. They observe similar neuroprotective effects towards motor neurons when MLKL was pharmacologically inhibited using necrosulfanamide (NSA) (Figure 1). In Parkinson’s disease (PD), parthanatos, another necrotic cell death program plays a role in the demise of dopamine neurons (Lee et al., 2013). Thus, non-traditional cell death mechanisms are emerging as important players in chronic neurodegenerative diseases. Since the pharmaceutical focus on conventional cell death pathways has not been fruitful as disease modifying therapies, it will be important to determine if these necrotic programmed cell death pathways modify ALS, PD and other neurodegenerative diseases in humans.

Figure 1.

Non-cell autonomous mediators of cell death derived from activated astrocytes contribute to necroptosis mediated motor neuron demise. Presence of mutant SOD1 in astrocytes derived from familial ALS patients affect motor neuron viability through secretion of neurotoxic factors that target motor neurons for death. Through comparative analysis of astrocytes derived from familial and sporadic ALS patients, Re et al show that the secretion of such toxic factors may not be triggered by mutant SOD1 per se. Rather astrocytes derived factors induce motor neuron death via necroptosis involving RIP1 and MLKL under both conditions. Accordingly, pharmacological inhibition of RIP1 and MLKL using necrostatin (Nec-1) and necrosulfanamide (NSA) promotes motor neuron survival. While the identity of such motor neuron specific neurotoxins is yet to be verified, this finding offers new insight into ALS pathogenesis.

Although familial and sporadic forms of ALS have common disease pathologies, a major challenge in ALS research has been to evaluate similarities and differences in the mechanisms that cause the detrimental effects in the respective scenarios. The development of humanized ALS models from patients afflicted with the disease of known and unknown aetiology have helped make strides towards this end. For instance, together with earlier studies, the findings from Re et al identify astrocytes, the largest population of cells in the CNS, to have toxic non-cell autonomous effects on motor neurons in both forms of ALS. Astrocytes closely interact with neurons and when activated by neuronal insults proliferate and exhibit morphological and gene expression changes. This phenomenon referred to as astrogliosis is a hallmark feature of ALS. Thus, the presence of reactive astrocytes in the motor neuron vicinity poses similar pathological challenges in both forms of ALS. Additionally, both forms of ALS are mediated by release of astrocyte specific toxic factors that target motor neurons. Identification of such astrocyte derived factors can have far reaching implications in terms of pathogenesis and therapeutic interventions. Few earlier studies have attempted to identify such soluble mediators of cellular toxicity but with little success. Di Georgio and colleagues (2008) identified prostaglandin D2 (PGD2) as an astrocyte-derived mediator of motor neuron death in sALS. However, blocking PGD2 has offered only modest rescue of motor neuron loss suggesting the involvement of additional factors in this pathway. In a study by (Marchetto et al., 2008), activation of inflammatory responses in fALS derived astrocytes was observed with concomitant increase in reactive oxygen species (ROS) and pro-inflammatory cytokines like nitric oxide. NF-ΚB and kinase (MAPK, JNK, AKT) signaling pathways are also activated in fALS and sALS astrocytes (Haidet-Phillips et al., 2011). The identification of necroptosis mediated cell death as a common feature of fALS and sALS further necessitates the identification of factors and processes that serve as initiators, modulators or executioners of necroptosis in the context of ALS.

The identification of astrocytes as motor neuron killers has important implications for stem cell based therapies for ALS. Given the toxic effects motor neurons endure, replacement of damaged motor neurons with wild type neurons may appear as a promising repair strategy. However, engraftment of healthy motor neurons onto to ALS spinal cord that harbors a hostile cellular milieu surrounded by damaged astrocytes may not have beneficial effects. The presence of mutant SOD1 in astrocytes marked them as mediators of motor neuron targeted toxicity. However, failure to rescue motor neuron loss by SOD1 suppression in both fALS and sALS raises a number of interesting questions for further investigation. For instance, what are the cell autonomous and non-cell autonomous stimuli that activate astrocytes culminating in motor neuron demise? Re et al show that GABAergic interneurons are spared from astrocyte mediated toxicity. However, ventro-lateral and medio-lateral spinal interneurons are also known to degenerate in ALS prior to motor neuron loss (Martin et al., 2007). This raises questions about other cell types that are susceptible to astrocytes that are on an attack rampage. The availability of humanized ALS co-culture systems are advantageous to study the exclusive contribution of astrocytes (or perhaps other suspected cell types) without any interference from other cell types. However, in light of the fact that secreted soluble factors are mediating the toxic effects in motor neurons, it prompts to question whether these factors originate exclusively from astrocytes or can they be derived from other cell types that are also susceptible to astrocytes? Future complementary work using in vitro culture systems and animal models will hopefully help identify biomarkers for diagnosis and screen for drugs that have neuroprotective effects with potential clinical relevance with regards to ALS.