Amyotrophic lateral sclerosis (ALS) is a heterogeneous neurodegenerative disease that leads to progressive muscle deterioration and commonly death within 3–5 years of symptom onset. The first gene established to cause ALS when mutated was SOD1, identified by a team of geneticists in 1990. Since then, more than 25 genes have been identified or implicated in ALS (Marangi and Traynor 2015). Though awareness of the genetic factors that cause ALS is impressive, clear understanding of the mechanism or mechanisms by which the known mutations trigger motor neuron degeneration is lacking. Several hypotheses have been proposed including protein aggregation, defective RNA processing, oxidative stress, subcellular organelle dysfunction, increased apoptosis, neuroinflammation, and glutamate excitotoxicity (Paez-Colsante et al. 2015). All are supported by evidence indicating there are many different potential ways by which motor neuron function can be compromised in ALS. That the vast majority of ALS cases (90%) are not inherited but sporadic in nature provides even more complexity to the problem.
The gene encoding alsin, ALS2, was the second gene identified to cause ALS and is mutated in a juvenile form of the disease. Unlike SOD1 and the majority of the other ALS-related genes, mutations in ALS2 are autosomal recessive and disease is due to loss of function of the alsin protein. ALS2 mutations have also been shown to cause two other neurodegenerative disorders, infantile ascending hereditary spastic paraplegia (IAHSP) and juvenile primary lateral sclerosis (JPLS). JPLS is similar to ALS but lower motor neurons are preserved while IAHSP is characterized by initial onset of spasticity followed by limb weakness. Interestingly, of the 16 mutations in ALS2 listed on Online Mendelian Inheritance in Man (OMIM), more than half are associated with IAHSP and ALS2 is the only gene identified as causing this disease (OMIM 606353). The vast majority of disease-causing mutations result in premature truncations though two missense mutations have been identified, C157Y and G540E. Next-generation sequencing of patients recently revealed that ALS2 mutations are associated with adult-onset sporadic ALS as well, though the role of alsin in causing disease in these individuals has not been established (Couthouis et al. 2014).
Due to space requirements, not all primary articles can be referenced. The reader is encouraged to consult the UCSD-Nature Molecule page for alsin (Topp 2015) and two excellent reviews by Hadano et al. (2007) and Chandran et al. (2007) for a more in-depth presentation of alsin biochemistry, Als2 gene structure and expression patterns, and the characteristics of alsin-deficient mice.
The Alsin Vps9 Domain Activates Rab5 to Promote Endosomal Fusion
Rab proteins regulate intracellular transport with, in general, specific Rabs or Rab subfamilies controlling traffic to particular organelles. Activation of the Rab and its bound effectors promotes docking and fusion of transport vesicles with the target, recruitment of subsequent Rab proteins for maturation of dynamic subcellular organelles (most evident in the endocytic system), interaction with microtubules and their motors for movement throughout the cell, and even initial packaging of the destination contents into vesicles. The Rab5 subfamily specifically regulates transport and fusion of vesicles with early endosomes and homotypic endosomal fusion.
Alsin has been shown to bind to and activate all of the Rab5 subfamily members in vitro with the exception of Rab21 (Hadano et al. 2007) and this activity is due to the Vps9 domain (Topp et al. 2004) (Fig. 2). Overexpression of the Vps9 domain of alsin leads to the formation of enlarged endosomes showing that alsin Rab5 GEF activity is also seen in vivo (Otomo et al. 2003; Topp et al. 2004). Interestingly, this activity is much weaker in the context of the full-length alsin protein suggesting that other domains of alsin autoinhibit the Vps9 domain in the cellular context. (It should be noted, however, that the Vps9 domain alone can be overexpressed at much higher levels than full-length alsin.) Neurons lacking alsin exhibit a reduction in, but not absence of endosomal fusion indicating that other Rab5 GEFs like Rabex-5 and RIN1 act in parallel to but cannot completely compensate for the absence of alsin (Devon et al. 2006; Jacquier et al. 2006).
The Alsin DH/PH Domains Promote Activation of Rac1
The Vps9 domain of alsin is not the only domain involved in GTPase modulation. The centrally located Dbl homology (DH) and pleckstrin homology (PH) domains are hallmarks for GEFs of the Rho GTPase family. Rho GTPases are best known for their regulation of the actin cytoskeleton in cells with different family members having distinct effects. For example, RhoA is associated with the formation of stress fibers and focal adhesions, cdc42 with filopodia, and Rac1 with membrane ruffles and lamellipodia. Rho GTPases are required for neural development via their regulation of the actin cytoskeleton to promote growth cone and dendritic spine formation and direct axonal path finding, and their regulation of cell survival via signaling cascades (Stankiewicz and Linseman 2014).
Several groups have shown that the DH/PH domains of alsin bind specifically to and exhibit GEF activity toward one of the Rho GTPase family members, Rac1 (Topp et al. 2004; Kanekura et al. 2005; Tudor et al. 2005). The GEF activity required the presence of the other domains of alsin and it has been argued that additional structure is required for catalytic activity because it stabilizes the protein, better orients the domains towards the Rac1 substrate, or binds to another factor required for GEF activity. Alternatively, posttranslational modification of alsin may be required, modification that can only occur in the context of the full-length protein. Further evidence supports alsin function as a Rac1 GEF. For example, motor neurons lacking alsin exhibited decreased axonal growth and cell survival and this could be overcome by overexpression of constitutively active Rac1 that only binds to GTP (Jacquier et al. 2006). Overexpression of alsin can also rescue the survival of SOD1-mutant cells in a Rac1-dependent manner (Kanekura et al. 2005). In addition, deletion of alsin suppressed cyclic-stretch induced reorientation of stress fibers in endothelial cells, similar to that observed with decreased expression of two other established Rac1 GEFs P-REX2 and alpha-PIX (Abiko et al. 2015). Last, alsin has been shown to colocalize with Rac1 and actin to membrane ruffles, growth cones, and macropinosomes, the latter a product of macroendocytosis that is stimulated by activated Rac (Topp et al. 2004; Tudor et al. 2005; Kunita et al. 2007).
Kunita et al. (2007) have argued that alsin serves not as a GEF but as an effector protein of Rac1 by showing that alsin preferentially binds to Rac1 in the GTP-bound state. Once could harmonize the apparent discrepant sets of data by saying that alsin stabilizes the GTP-bound or activated state of Rac1. Whether alsin is thought as a GEF or an effector protein of Rac1 is not a trivial matter and it impacts the way in which future studies are directed (e.g., compare GEF and effector proteins in the nucleotide-binding cycle of Rab5 in Fig. 2). In the former case, alsin would act catalytically to activate the GTPase in response to specific cellular cues. The identification of these cues would thus be essential. That overexpression of full-length alsin does not trigger “traditional” Rho GTPase cellular phenotypes is evidence against Rac1 as a GEF. In the latter case, alsin would act noncatalytically in response to Rac1 activation to transduce the signal into some effector activity. It has been argued that this effector activity is related to Rab5-mediated endocytosis but functional evidence is lacking and what has been shown is that alsin localization changes in response to overexpression of a Rac1 mutant that remains in the GTP-bound state (Kunita et al. 2007). Clearly, additional studies are necessary to further delineate the alsin-Rac1 connection.
The Alsin RCC1-like Domain (RLD) and Other Domains Mediate Protein-Protein Interactions
Due to sequence similarity with RCC1 (regulator of chromatin condensation 1), the amino-terminal domain was originally proposed to function, like RCC1, as a GEF for the nuclear shuttle protein Ran. This activity has not been demonstrated and it has been argued that this domain instead serves as a protein-protein interaction motif due to its predicted seven-bladed beta propeller structure and the presence of RCC1-like domains in a multitude of proteins without Ran GEF activity. This domain has been shown to mediate alsin’s interaction with glutamate receptor interacting protein (GRIP1), a protein responsible for transporting AMPA receptors containing the GluR2 subunit to neuronal synapses (Lai et al. 2006). Though the effect is mild, the loss of alsin causes GRIP1 and GluR2 mislocalization making neurons more susceptible to excitotoxicity, an established mechanism involved in motor neuron degeneration (Lai et al. 2006). The role of the Rab5 and Rac1 GEF domains in this process is unknown, however.
The RLD may play a role in regulation of the alsin GEF activity toward Rab5 and Rac1. While expression of the Vps9 domain alone promotes Rab5-GTP in vivo causing enlarged endosomes and the addition of the DH/PH domains enhances this process, inclusion of the RLD domain has been shown to diminish endosomal fusion (Otomo et al. 2003; Topp et al. 2004). Additional evidence to support this hypothesis comes from localization studies of individual alsin domains that show the RLD is cytosolic while the DH/PH and Vps9 domains are endosomal. It has also been observed that the RLD and C-terminal regions (MORN repeats and Vps9 domain) of alsin interact with each other when individual domains of alsin are overexpressed (Kunita et al. 2007). Thus, it is possible that the RLD binds back to autoinhibit alsin GEF activity (toward Rab5 and/or Rac1) and that binding to another protein and/or posttranslational modification of alsin is required to release this inhibitory activity. Alternatively, the RLD could both positively and negatively regulate the GEF domains in a cell signaling-dependent manner. The identification of other proteins that interact with the alsin RLD and the cellular context of the interactions should help to clarify the function of this domain. That alsin can be phosphorylated at several serine/threonine sites and acetylated at another within the RLD is intriguing (Fig. 1), though the functional significance of these modifications has not been determined.
Alsin has been shown to interact with several proteins via domains other than the RLD. For example, alsin interacts with itself (region between the MORN repeats and Vps9 domain), ALS2CL (Vps9 domain), the ubiquitously expressed transcript (DH/PH domains), and SOD1 (DH/PH domains) via the domains of alsin indicated (Hadano et al. 2007). ALS2CL (ALS2 C-like), as its name suggests, is a protein that shares sequence similarity to the C-terminal region of alsin. It has been shown to possess weak GEF activity on Rab5 in addition to binding alsin, although the significance of the latter finding is not clear since the two proteins have distinct tissue expression patterns. The alsin-SOD1 interaction is only seen in overexpression studies with mutant versions of SOD1, but it is potentially medically relevant since SOD1 mutations also cause ALS. Indeed, alsin overexpression stimulates Rac1-mediated signaling that protects cells from mutant SOD1-induced death (Kanekura et al. 2005). Furthermore, alsin deletion exacerbates the disease-associated effects of SOD1 mutation on mice in some strains (Hadano et al. 2010).
Alsin has been shown to interact with several other proteins, but the region of alsin mediating the interaction has not been identified. Examples include spartin, neurocalcin A, 14-3-3 beta, dynein axonemal assembly factor 2, and vasolin-containing protein (VCP) (Topp 2015). The significance of these interactions is unclear.
Towards an Overall Cellular Function for Alsin
Two strategies have been attempted in parallel to determine the function(s) of alsin. The first has been to confirm that the domains of alsin predicted computationally are in fact functional using biochemistry and cell biology techniques. These methods have established that the Vps9 and DH/PH domains activate Rab5 and Rac1, respectively, and form the structure of this review.
The second strategy has been to generate knockout alsin cell lines or animal models with hopes that physiological, behavioral, pathological, or cellular defects will be observed. Interestingly, alsin knockout mice have only mild phenotypes. Six different lines of mice have been generated using a few distinct knockout strategies and none of them faithfully recapitulates juvenile ALS (summarized in Cai et al. 2008). The most consistent phenotype is degeneration of the corticospinal tract, which is more congruent with hereditary spastic paraplegia and is perhaps not surprising since mutations in alsin are more commonly associated with IAHSP. Using a clever “knock-in” strategy to express GFP selectively in corticospinal motor neurons, it was recently shown that alsin-deficient mice have greatly disrupted morphology in these neurons, including vacuolated apical dendrites, shrinkage of the cell body, increased autophagy, and axonal pathology (Gautam et al. 2016).
The effect of alsin deletion in zebrafish has also been studied though not nearly as extensively as in mice. One group has shown intriguingly that alsin-deficient zebrafish exhibit phenotypes more similar to ALS including motor neuron and spinal cord deficits in addition to development defects in swimming (Gros-Louis et al. 2008). These authors further showed that alternative mRNA transcripts were present in their supposed alsin knockout mouse and that some of these transcripts could partially rescue the loss of alsin in zebrafish. There may be more to the alsin knockout mice than meets the eye.
Alsin-deficient cell lines either from knockout animals or siRNA treatment have been used to confirm the functions of individual domains. As mentioned above, these studies reveal that the loss of alsin disrupts Rab5-mediated endosomal fusion and Rac1-mediated axonal outgrowth and cell survival. Because both Rab5 and Rac1 are involved in endocytosis, the internalization and trafficking of receptors was also analyzed in alsin-deficient cells. It was shown that neurons lacking alsin have specific defects in receptor-mediated transport of IGF-1 (insulin-like growth factor 1) and BDNF (brain-derived neurotrophic factor), but not in constitutive endocytosis of transferrin and dextran (Devon et al. 2006).
Like other neurotrophins, IGF-1 and BDNF growth factors are thought to be transported in a retrograde fashion from distal axons to neuronal cell bodies in structures termed “signaling” endosomes. These endosomes connect events at the axonal synapse to the nucleus to drive sustained transcription in the presence of postsynaptic cues. The best characterized of the neurotrophic factors is NGF (nerve growth factor), which binds to its receptor, TrkA (tropomyosin receptor kinase A). NGF-bound TrkA is internalized at the synapse by both clathrin-dependent and clathrin-independent mechanisms that result in the formation of early endosomes and macropinosomes, respectively (Harrington and Ginty 2013). Though both mechanisms occur, the former is better established. The early endosomes then mature into signaling endosomes that contain activated molecules from multiple signaling cascades, including PI3K/Akt, MAP/ERK, and PLC pathways, in addition to NGF and TrkA (Harrington and Ginty 2013). Formation of signaling endosomes requires Rab proteins including the alsin target, Rab5. The signaling endosome “packages” are transported along microtubules to the nucleus to ensure that transcription is coordinated with synaptic signals. Intriguingly, both alsin and Rac1 have been identified in an NGF-TrkA signaling endosome preparation and Rac1 is required to regulate local actin dynamics post-internalization but prior to long-distance transport (Harrington and Ginty 2013).
Because motor neurons are some of the longest cells in the body approaching a meter or more in length, it has been argued that the formation of signaling endosomes is crucial to survival of these cells. That motor neurons have been observed to degenerate from the axon towards the cell body provides additional support to this hypothesis. Could alsin be a key regulator of signaling endosomes in neurons? Several lines of evidence already presented support this hypothesis. First, loss of alsin in motor neurons results in decreased axonal growth in motor neurons in culture and degenerating corticospinal axon fibers in mice (Jacquier et al. 2006; Gautam et al. 2016), both phenotypes of which can be protected in part by signaling endosomes (Harrington and Ginty 2013). Second, alsin is involved in receptor-mediated transport of neurotrophic growth factors such as IGF-1 and BDNF. Third, alsin is localized to signaling endosomes, early endosomes and macropinosomes, the latter two of which are early stages in signaling endosome formation. Fourth, Rac1 and Rab5 are both required for the formation of signaling endosomes and alsin promotes activation of each. Fifth, alsin promotes survival of cells including neurons via a Rac1-mediated signaling pathway. Taken together, the evidence suggests that alsin promotes the formation of signaling endosomes though further experimentation in neurons is required to more definitively demonstrate it.
The gene encoding alsin, ALS2, is mutated in three neurodegenerative diseases, ALS, JPLS, and IAHSP, all of which are due to motor neuron dysfunction. Alsin has several domains that play roles in cell signaling, protein-protein interaction, and membrane trafficking. Alsin possesses GEF activity towards Rab5, a protein required for membrane transport to early endosomes. Alsin is also a GEF and/or effector protein for Rac1, the activation of which drives actin cytoskeleton remodeling and cell survival signaling pathways. Alsin possesses an RLD domain with similarity to RCC1 that mediates interaction with other proteins and it is postulated that this domain serves to regulate the activities of the Rab5 and Rac1 GEF domains, perhaps via autoinhibition. If true, it is likely that the RLD binds to another protein and/or is posttranslationally modified to alleviate inhibition.
Alsin has been shown to promote receptor-mediated transport of IGF-1 and BDNF, growth factors known to be transported via signaling endosomes in a process that requires Rab5 and Rac1. Future studies should look to investigate the role of alsin in signaling endosome formation and transport in a neuronal context. No direct interaction between alsin and IGF-1 or BDNF receptors or upstream molecules in their signaling cascades has been demonstrated. The RLD is a prime target for mediation of this putative interaction and identification of proteins that interact with this domain is essential. In addition, a better understanding of the role of posttranslational modifications in regulation of alsin enzymatic function would help to better connect alsin to cellular signaling pathways, perhaps those initiated by IGF-1 and other neurotrophins. The precise spatiotemporal roles of the alsin Rab5 and Rac1 GEF domains have also not been determined. Upon cell stimulation with growth factors, does alsin function to activate Rab5 or Rac1 first? Can both GEF activities occur at the same time? Are they mutually exclusive? Structural investigation of alsin, though a technical challenge, would be very helpful in answering these and other questions surrounding alsin function.
- Devon RS, Orban PC, Gerrow K, Barbieri MA, Schwab C, Cao LP, Helm JR, Bissada N, Cruz-Aguado R, Davidson TL, Witmer J, Metzler M, Lam CK, Tetzlaff W, Simpson EM, McCaffery JM, El-Husseini AE, Leavitt BR, Hayden MR. Als2-deficient mice exhibit disturbances in endosome trafficking associated with motor behavioral abnormalities. Proc Natl Acad Sci USA. 2006;103:9595–600. doi: 10.1523/JNEUROSCI.2084-06.2006.PubMedCentralCrossRefPubMedGoogle Scholar
- Gros-Louis F, Kriz J, Kabashi E, McDearmid J, Millecamps S, Urushitani M, Lin L, Dion P, Zhu Q, Drapeau P, Julien JP, Rouleau GA. Als2 mRNA splicing variants detected in KO mice rescue severe motor dysfunction phenotype in Als2 knock-down zebrafish. Hum Mol Genet. 2008;17:2691–702. doi: 10.1093/hmg/ddn171.CrossRefPubMedGoogle Scholar
- Hadano S, Otomo A, Kunita R, Suzuki-Utsunomiya K, Akatsuka A, Koike M, Aoki M, Uchiyama Y, Itoyama Y, Ikeda JE. Loss of ALS2/Alsin exacerbates motor dysfunction in a SOD1-expressing mouse ALS model by disturbing endolysosomal trafficking. PLoS One. 2010;5:e9805. doi: 10.1371/journal.pone.0009805.PubMedCentralCrossRefPubMedGoogle Scholar
- Kanekura K, Hashimoto Y, Kita Y, Sasabe J, Aiso S, Nishimoto I, Matsuoka M. A Rac1/phosphatidylinositol 3-kinase/Akt3 anti-apoptotic pathway, triggered by AlsinLF, the product of the ALS2 gene, antagonizes Cu/Zn-superoxide dismutase (SOD1) mutant-induced motoneuronal cell death. J Biol Chem. 2005;280:4534–43. doi: 10.1074/jbc.M410508200.CrossRefGoogle Scholar
- Lai C, Xie X, McCormack S, Chiang HS, Michalak M, Lin X, Chandran J, Shim H, Shimoji M, Cookson MR, Huganir RL, Rothstein JD, Price DL, Wong PC, Martin LJ, Zhu JJ, Cai H. Amyotrophic lateral sclerosis 2-deficiency leads to neuronal degeneration in amyotrophic lateral sclerosis through altered AMPA receptor trafficking. J Neurosci. 2006;83:186–92. doi: 10.1523/JNEUROSCI.2084-06.2006.Google Scholar
- Otomo A, Hadano S, Okada T, Mizumura H, Kunita R, Nishijima H, Showguchi-Miyata J, Yanagisawa Y, Kohiki E, Suga E, Yasuda M, Osuga H, Nishimoto T, Narumiya S, Ikeda JE. ALS2, a novel guanine nucleotide exchange factor for the small GTPase Rab5, is implicated in endosomal dynamics. Hum Mol Genet. 2003;12:1671–87. doi: 10.1093/hmg/ddg184.CrossRefPubMedGoogle Scholar