SEARCH

SEARCH BY CITATION

Keywords:

  • Alzheimer's disease;
  • centronuclear myopathy;
  • Charcot-Marie-Tooth neuropathy;
  • cytoskeleton;
  • dynamin-2;
  • vesicles trafficking

Abstract

  1. Top of page
  2. Abstract
  3. Dynamin structure and catalytic mechanism
  4. Dynamin isoforms and synaptic vesicle recycling
  5. Dynamin-2 controls endocytosis at the post-synaptic membrane
  6. Role of dynamin-2 in neurosecretion
  7. Role of dynamins in neuronal process extension
  8. Dynamin-2 mutations in Charcot-Marie-Tooth neuropathy
  9. Dynamin-2 mutations in centronuclear myopathy associated with abnormalities in the nervous system
  10. Tissue-specific phenotypes of disease-related dynamin-2 mutations
  11. Dynamin-2 in AD
  12. Conclusions and perspectives
  13. Acknowledgements
  14. References
Thumbnail image of graphical abstract

Dynamin-2 is a pleiotropic GTPase whose best-known function is related to membrane scission during vesicle budding from the plasma or Golgi membranes. In the nervous system, dynamin-2 participates in synaptic vesicle recycling, post-synaptic receptor internalization, neurosecretion, and neuronal process extension. Some of these functions are shared with the other two dynamin isoforms. However, the involvement of dynamin-2 in neurological illnesses points to a critical function of this isoform in the nervous system. In this regard, mutations in the dynamin-2 gene results in two congenital neuromuscular disorders. One of them, Charcot-Marie-Tooth disease, affects myelination and peripheral nerve conduction, whereas the other, Centronuclear Myopathy, is characterized by a progressive and generalized atrophy of skeletal muscles, yet it is also associated with abnormalities in the nervous system. Furthermore, single nucleotide polymorphisms located in the dynamin-2 gene have been associated with sporadic Alzheimer's disease. In the present review, we discuss the pathogenic mechanisms implicated in these neurological disorders.

Abbreviations used
AD

Alzheimer's disease

APOE

apolipoprotein E

APP

amyloid precursor protein

amyloid beta protein

BSE

bundle signaling element

CME

clathrin-mediated-endocytosis

CNM

centronuclear myopathy

DNMBP

dynamin-binding protein

EOAD

early-onset AD

GED

GTPase effector domain

LOAD

late-onset AD

PH

pleckstrin homology domain

PRD

proline and arginine-rich domain

PS1

presenilin1

PS2

presenilin2

Dynamin is a 100 kDa GTPase involved in membrane and cytoskeleton remodeling (Ferguson and De Camilli 2012). In mammals, dynamin is encoded by three different genes, DNM1, DNM2, and DNM3; located in chromosomes 9 (Newman-Smith et al. 1997), 19 (Züchner et al. 2005), and 1 (Nagase et al. 1999), respectively. All dynamin isoforms can undergo alternative splicing of exons, thus giving rise to four to thirteen spliced variants which appear to differ in their intracellular distribution (Cao et al. 1998). As we discuss below, all three dynamin isoforms are expressed in the nervous system, where their most well-described role is in membrane fission during clathrin-mediated-endocytosis (CME) of synaptic vesicle membranes and membrane proteins. In addition to its role in CME, dynamin functions have been extended to other cellular processes such as caveolin-dependent internalization (Henley et al. 1998), vesicle budding from Golgi membranes (Jones et al. 1998) and endosomes (Nicoziani et al. 2000), regulation of microtubule stability (Thompson et al. 2004), actin cytoskeleton dynamics (Mooren et al. 2009; Gu et al. 2010), and fusion processes (De la Vega et al. 2011; Reid et al. 2012; Leikina et al. 2013).

Whether each dynamin isoform specializes in a specific cellular function, or whether they exert overlapping roles in the nervous system is still unclear, as evidence arguing in favor of redundant (Raimondi et al. 2011) as well as differential roles of the three isoforms (Liu et al. 2011a) have been presented. However, it is interesting that unlike dynamin-1, which is mainly pre-synaptic (Powell and Robinson 1995), dynamin-2 is expressed at both pre- and post-synaptic densities (Okamoto et al. 2001; Tanifuji et al. 2013), where it participates in synaptic vesicle recycling (Tanifuji et al. 2013) and signaling receptor internalization (Carroll et al. 1999; Kabbani et al. 2004). These evidences support an overall role of dynamin-2 in the maintenance of synaptic transmission. Interestingly, mutations in the dynamin-2 gene have been related with two congenital neuromuscular disorders: Charcot-Marie-Tooth neuropathy and Centronuclear myopathy (Durieux et al. 2010a), whereas single nucleotide polymorphisms located in DNM2 have been shown to associate with sporadic Alzheimer's disease (AD) (Aidaralieva et al. 2008), thus pointing to a critical function of dynamin-2 over that of the other isoforms in the nervous system.

In the current review, we discuss the relevance of dynamin-2 in the function of the nervous system and the mechanisms implicated in the pathologies associated with dynamin-2 gene mutations.

Dynamin structure and catalytic mechanism

  1. Top of page
  2. Abstract
  3. Dynamin structure and catalytic mechanism
  4. Dynamin isoforms and synaptic vesicle recycling
  5. Dynamin-2 controls endocytosis at the post-synaptic membrane
  6. Role of dynamin-2 in neurosecretion
  7. Role of dynamins in neuronal process extension
  8. Dynamin-2 mutations in Charcot-Marie-Tooth neuropathy
  9. Dynamin-2 mutations in centronuclear myopathy associated with abnormalities in the nervous system
  10. Tissue-specific phenotypes of disease-related dynamin-2 mutations
  11. Dynamin-2 in AD
  12. Conclusions and perspectives
  13. Acknowledgements
  14. References

The three isoforms of dynamin share a common primary structure consisting of five structural domains (Fig. 1a): a large amino-terminal GTPase (G-domain), a middle domain, a GTPase effector domain (GED), a pleckstrin homology domain (PH) that binds inositol phospholipids, and a carboxy-terminal proline and arginine-rich domain (PRD) that allows dynamin to interact with SH3-domain-containing proteins (Cao et al. 1998). The G-domain contains four G-motifs (G1 to G4), which are highly conserved among small and large GTPases (Praefcke and McMahon 2004) and are needed for the binding and hydrolysis of GTP. Only one GTP molecule is bound per G-domain, but the sequences that contribute to its interaction are distributed along the complete domain (http://www2.mrc-lmb.cam.ac.uk/Dynamin/GTPbinding-residues.html). The G1 motif (also called P-loop) co-ordinates the GTP phosphates, a threonine in the G2 motif is implicated in its hydrolysis, a glycine residue in the G3 motif forms a hydrogen bond with the γ-phosphate of GTP, and the G4 motif plays a role in guanine and ribose co-ordination (http://www2.mrc-lmb.cam.ac.uk/Dynamin/GTPbinding-residues.html). The GTP binding and hydrolysis is accompanied by cyclic conformational changes that regulate dynamin assembly and disassembly (Marks et al. 2001).

image

Figure 1. Disease-linked dynamin-2 mutations are mainly located in the middle and pleckstrin homology (PH) domains. (a) Dynamin is a multimodular enzyme comprising five highly conserved structural domains: a N-terminal GTPase domain (G-domain) required for binding and hydrolysis of GTP, a middle domain, a PH domain that mediates lipid interaction, a GTPase effector domain (GED) that regulates GTPase activity, which together with the middle domain is involved in dynamin oligomerization. Finally, a C-terminal proline rich domain (PRD) is present, which is required for interaction with SH3-domain-containing proteins. (b) A ‘T-shape’ dimer appears to be the structural unit of dynamin oligomers (Chappie et al. 2011). In this configuration, the PH domains (yellow) of neighboring monomers act as the ‘legs’ that insert dynamin into lipid membranes. Each ‘stalk’ region formed by the middle (green) and GED (red) domains interacts with the other in a crossed fashion, orienting the respective G-domains (blue) in opposite directions. Most of the mutations identified in CNM-patients (at the left) localize at the middle and C-terminal α-helix PH domains, both of which are implicated in dynamin oligomerization. Most of the CMT-linked mutations (at the right) clustering at the N-terminal region of the PH domain are involved in the insertion of dynamin into lipid membranes. Color code for structural domains and letters indicating the mutations are the same as those shown in A. In this diagram, PRD is omitted.

Download figure to PowerPoint

Flanked by the PRD and PH domains, GED functions as a regulatory region that favors dynamin oligomerization and stimulates its GTPase activity (Muhlberg et al. 1997). An essential arginine residue located at GED (R725) was shown to contribute to the GTP hydrolysis (Sever et al. 1999), suggesting a role of GED as an intra-molecular GTPase-activating protein. A carboxy-terminal helix from GED together with helices at the amino- and carboxy-terminal of the G-domain form the helical bundle called ‘bundle signaling element’ (BSE), which seems to sense and transmit assembly-dependent conformational changes to the G-domain (Chappie et al. 2009). Helices from the middle domain and from the aminoterminal of GED compose the so called ‘stalk’ region (Chappie et al. 2010; Ford et al. 2011). Crystallographic studies show that stalks of neighboring dynamin-1 monomers interact in a crossed fashion, yielding a ‘T-shape’ dimer oriented with their G-domains in opposite directions (Chappie et al. 2011; Faelber et al. 2011). This dimer configuration is schematized in Fig. 1b.

In solution, dynamin undergoes a predominantly tetrameric conformation, and exhibits a basal GTPase activity (Muhlberg et al. 1997). However, in the presence of anionic lipid scaffolds it assembles into helical arrays (Sweitzer and Hinshaw 1998; Stowell et al. 1999). This dynamin self-assembly potently stimulates its GTPase activity, up to 100-fold (Warnock et al. 1996), and many different factors that induce dynamin oligomerization can also act as effectors of its catalytic activity. For instance, low-ionic strength solutions (Hinshaw and Schmid 1995), negatively charged tubular templates, such as microtubules (Maeda et al. 1992; Warnock et al. 1997) and actin bundles (Mooren et al. 2009; Gu et al. 2010) promote dynamin oligomerization and hence potentiate its catalytic activity. These properties make dynamin a peculiar GTPase which, unlike conventional G-proteins, does not require GTPase-activating proteins and guanine nucleotide-exchange factors (Gasper et al. 2009). Instead, dynamin, together with septins and the signal recognition particle, could be classified as ‘G-proteins activated by nucleotide-dependent dimerization’ (Gasper et al. 2009). In fact, crystallographic studies indicate that the dimerization of G-domains triggers the increased GTPase activity of dynamin (Chappie et al. 2010). The same authors later showed that the G-domain dimerization is facilitated by the helical assembly of dynamin, where G-domains of adjacent rungs are balanced to form a dimer (Chappie et al. 2011). On the other hand, as Morlot and Roux (2013) indicated, dynamin shares more properties with motor ATPases than conventional GTPases. In fact, dynamin (i) displays a non-negligible ATPase activity; (ii) the dynamin GTPase domain has some common structural features with kinesins; and (iii) similar to motor proteins, the nucleoside triphosphate hydrolysis triggers a power stroke; this is the mechanical motion of a lever arm that in dynamin is equivalent to the BSE (Morlot and Roux 2013).

Inositol phospholipids such as phosphatidylinositol 4,5-bisphosphate (PtdIns (4,5) P2) can also stimulate dynamin GTPase activity (Salim et al. 1996; Lin et al. 1997) by binding directly to its PH domain (Zheng et al. 1996). It has been reported that the capability of dynamin to insert itself into lipid membranes, via a hydrophobic variable-loop (VL1) located in the PH domain, is essential for its self-assembly-induced activation and for membrane scission (Ramachandran et al. 2009). This finding points to a role of the PH domain as an effector of dynamin GTPase activity (Bethoney et al. 2009). Interestingly, Kenniston and Lemmon (2010) proposed that, by interacting with the adjacent GED, the PH domain plays an auto-inhibitory role by preventing dynamin self-assembly in the absence of lipid membranes. It seems that dynamin lipid-binding could relieve this GED/PH interaction by favoring molecular rearrangements that promote dynamin self-assembly and GTPase activation (Kenniston and Lemmon 2010). SH3 domain-containing molecules that interact with the PRD also promote dynamin self-assembly in vitro (Muhlberg et al. 1997), and potentiate its GTPase activity, as described for amphiphysin, an active partner of dynamin during endocytosis (Yoshida et al. 2004).

The most well-known function of dynamin is its participation in membrane fission (for review see Schmid and Frolov 2011). Although the exact mechanism by which the fission process occurs is still under debate, some interesting models have been proposed. The most accepted idea is that dynamin acts as a mechano-enzyme that uses GTP hydrolysis to generate the driving-force required to separate membranes (Hinshaw 2000). One of the first models (‘pinchase’ model) proposed that dynamin self-assembles, forming spirals around the neck of vesiculating membranes that, as a result of the GTP hydrolysis, constrict the membranes (Sweitzer and Hinshaw 1998; Kozlov 1999). A second model (‘poppase’ model) proposed that the dynamin helical structure extends after GTP hydrolysis, stretching and finally cutting the vesiculating membranes (Stowell et al. 1999). On the other hand, microscopy-based live assays have shown that the hydrolysis of GTP results in a ‘twisting’ of dynamin-coated membrane tubules (‘twistase’ model), which contributes to a longitudinal tension that favors membrane scission (Roux et al. 2006). Given that point mutations of an essential arginine residue at GED (R725A) inhibited GTP hydrolysis, yet it had no negative effects neither on dynamin oligomerization nor on receptor-mediated endocytosis (Sever et al. 1999), it was suggested that the GTP-bound state of dynamin represents an ‘active form’ (Sever et al. 1999) that has a ‘regulatory’ function, and which probably activates other mediators of the endocytotic machinery (Newmyer et al. 2003), but it would not be a mechano-enzyme per se. Nevertheless, other dynamin mutants (T65A, R66A, and T141Q) which are capable of binding but not hydrolyzing GTP have been shown to inhibit endocytosis (Marks et al. 2001), thus supporting the role of a GTPase activity of dynamin itself during membrane scission. In 2009, Mettlen and co-workers proposed a dual role of dynamin during endocytosis that reconciles the previous models (Mettlen et al. 2009). These authors suggested that, initially, an unassembled dynamin is recruited to the endocytotic pits acting as a monitor of the coat assembly in a manner dependent on its basal GTP-ase activity, which could be negatively regulated by the GED (Mettlen et al. 2009). At later stages, dynamin would assemble itself into spiral structures around the neck of invaginated membranes enhancing its GTPase activity and finally promoting membrane fission (Mettlen et al. 2009). Unlike other classical GTPases, the ‘switch’ from one to another functional state of dynamin would not be the GDP dissociation and GTP-binding/hydrolysis, but rather the transition from the unassembled to the assembled state (Mettlen et al. 2009).

Nowadays, crystallography studies provide a more complete picture of dynamin function during membrane fission. When dynamin oligomerizes around of a pit neck, G-domains of adjacent helical rungs dimerize, leading to GTP hydrolysis. This GTP hydrolysis triggers conformational changes in BSE, which transmits the change to the rest of the structure causing the constriction of the dynamin helix (Roux et al. 2006; Mears et al. 2007; Chappie et al. 2011). Faelber et al. (2013) suggested that multiple rounds of GTP-binding and hydrolysis are required to constrict the pit neck, to consequently increase the membrane curvature (Faelber et al. 2013). The augmented membrane curvature would in turn elevate the local elastic energy, reducing the energy barrier required for the spontaneous fission at the boundary between the dynamin ring and bare membranes (Morlot et al. 2012).

Dynamin isoforms and synaptic vesicle recycling

  1. Top of page
  2. Abstract
  3. Dynamin structure and catalytic mechanism
  4. Dynamin isoforms and synaptic vesicle recycling
  5. Dynamin-2 controls endocytosis at the post-synaptic membrane
  6. Role of dynamin-2 in neurosecretion
  7. Role of dynamins in neuronal process extension
  8. Dynamin-2 mutations in Charcot-Marie-Tooth neuropathy
  9. Dynamin-2 mutations in centronuclear myopathy associated with abnormalities in the nervous system
  10. Tissue-specific phenotypes of disease-related dynamin-2 mutations
  11. Dynamin-2 in AD
  12. Conclusions and perspectives
  13. Acknowledgements
  14. References

Endocytic recycling of secretory vesicles after exocytosis is critical for the maintenance of synaptic transmission and membrane homeostasis (Cárdenas and Marengo 2010). The first evidence of dynamin participating in membrane recycling and endocytosis of secretory vesicles arose from experiments using the shibire mutated Drosophilia melanogaster model. These insects present a neuromuscular junction dysfunction that induces paralysis and alterations in development and locomotion (Poodry et al. 1973; Poodry and Edgar 1979). This phenotype is expressed as a decreased number of synaptic vesicles, and, most interestingly, the mutated gene was found to be the ortholog of the mammalian dynamin protein (Chen et al. 1991). Accumulation of omega-shaped structures at the presynaptic zone in shibire flies suggested an incapacity of the vesicles to achieve the fission process (Koenig and Ikeda 1989). Later studies using GTPase defective dynamin mutants (Clark et al. 1997), non-hydrolyzable guanine-nucleotide analogs (Yamashita et al. 2005) and pharmacological inhibition with dynasore (Newton et al. 2006) demonstrated the essential role of dynamin during synaptic vesicle recycling in different experimental models.

Regarding the contribution of the different dynamin isoforms in synaptic vesicle recycling, data arising from the De Camilli laboratory showed that neuronal dynamin-1 knock-out mice were able to form functional synapses and support limited synaptic vesicle endocytosis under spontaneous neuronal activity, affecting synaptic transmission only under high levels of stimulation (Ferguson et al. 2007). These authors suggested that other dynamin isoforms were likely compensating for the absence of dynamin-1. This hypothesis was partially verified with the generation of a dynamin-1 and -3 double knock-out mouse model (Raimondi et al. 2011), in which the absence of dynamin-3 worsened the phenotype produced by the loss of dynamin-1, but without resulting in a significant impact on neuronal development and differentiation (Raimondi et al. 2011). Remarkably, nerve terminals of those double knock-out animals were still able to recycle synaptic vesicles suggesting that a dynamin-independent mechanism for synaptic vesicle endocytosis could compensate for the lack of the isoforms 1 and 3. Supporting a dynamin-independent mechanism, a GTP-independent compensatory endocytosis of synaptic vesicles has been described in Calyx of Held synapses (Xu et al. 2008). Despite the latter, the persistence of synaptic transmission in dynamin-1 and -3 double knock-out animals also suggests that dynamin-2 may be sufficient to support basic synaptic functions (Raimondi et al. 2011). In this regard, the ability of dynamin-2 to partially rescue activity-induced synaptic vesicle recycling in neurons of dynamin-1 knock-out mice (Ferguson et al. 2007) further supports this notion, arguing in favor of the presence of overlapping functions for the three dynamin isoforms in the nervous system.

A recent report by Tanifuji et al. (2013) using an interfering RNA-mediated knockdown strategy in rat superior cervical ganglion neurons, showed that the three dynamin isoforms are involved in synaptic vesicle membrane recycling. However, they contribute differently to maintain releasable vesicle pools under different stimulation paradigms. The authors demonstrated that dynamin-1 directs vesicle recycling with a fast kinetics, during and after high levels of synaptic activity. Conversely, dynamin-3 mostly replenished the readily releasable vesicle pools at a slow rate, in a manner independent of the action potential frequency. Yet, after firing, dynamin-3 mediated the recovery more rapidly than dynamin-1. Dynamin-2 exhibited dual properties: it mediated vesicle recycling during high frequency stimulation, and acted quickly after an action potential firing. How dynamin isoforms differentially modulate the recycling of specific synaptic vesicle pools is still elusive, but the underlying mechanisms may allow synapses to selectively respond under different physiological conditions.

Dynamin-2 controls endocytosis at the post-synaptic membrane

  1. Top of page
  2. Abstract
  3. Dynamin structure and catalytic mechanism
  4. Dynamin isoforms and synaptic vesicle recycling
  5. Dynamin-2 controls endocytosis at the post-synaptic membrane
  6. Role of dynamin-2 in neurosecretion
  7. Role of dynamins in neuronal process extension
  8. Dynamin-2 mutations in Charcot-Marie-Tooth neuropathy
  9. Dynamin-2 mutations in centronuclear myopathy associated with abnormalities in the nervous system
  10. Tissue-specific phenotypes of disease-related dynamin-2 mutations
  11. Dynamin-2 in AD
  12. Conclusions and perspectives
  13. Acknowledgements
  14. References

Endocytosis regulates cell signaling by controlling the number of receptors available for activation at the cell surface (Sorkin and Von Zastrow 2009). Clathrin-dependent and -independent endocytotic pathways seem to mediate agonist-induced internalization of several receptors including G-protein-coupled receptors, growth factor receptors, and ionotropic receptors, thus implying the participation of dynamin in such processes (McMahon and Boucrot 2011).

Given that dynamin-2 is highly enriched at the post-synaptic density (Okamoto et al. 2001), this isoform probably plays an important role in regulating post-synaptic membrane protein turnover. In this regard, Carroll et al. (1999) demonstrated that AMPA-induced internalization of GluR1-containing receptors is significantly inhibited in cultured hippocampal neurons expressing the dynamin-2 mutant K44A, indicating a pivotal role of dynamin-2 in controlling glutamate receptor recycling and excitatory synaptic transmission. A role of dynamin-2 in regulating dopaminergic transmission has also been suggested, as this protein co-localizes with D2 dopamine receptors at the soma of striatal neurons (Kabbani et al. 2004). Interestingly, the expression of a GTPase defective dynamin-2, but not dynamin-1 mutants, suppressed dopamine-induced D2 receptor internalization (Kabbani et al. 2004), thus pointing to an isoform-specific role of dynamin-2 during the turnover of D2 receptors. Dynamin-2 has also been implicated in the neuronal uptake of endocannabinoids through an endocytotic mechanism that is independent of clathrin, but dependent on detergent resistant structures such as caveolae and lipid rafts (McFarland et al. 2004). In this regard, internalization and processing of the endocannabinoid receptor agonist anandamide was significantly inhibited in neuronal differentiated CAD cells expressing an interfering RNA against dynamin-2 (McFarland et al. 2008), thus strongly supporting the participation of dynamin-2 in the neuromodulatory effects of endocannabinoids in the nervous system.

Dynamin-3 is also present in post-synaptic densities, where it forms a complex with the post-synaptic scaffold protein Homer and the metabotropic glutamate receptor mGluR5 (Gray et al. 2003). However, a dominant-negative dynamin-3 isoform did not inhibit mGluR5 endocytosis (Gray et al. 2003), but conversely, the disruption of its association with Homer impaired the recycling of AMPA receptors in hippocampal neurons (Lu et al. 2007).

Role of dynamin-2 in neurosecretion

  1. Top of page
  2. Abstract
  3. Dynamin structure and catalytic mechanism
  4. Dynamin isoforms and synaptic vesicle recycling
  5. Dynamin-2 controls endocytosis at the post-synaptic membrane
  6. Role of dynamin-2 in neurosecretion
  7. Role of dynamins in neuronal process extension
  8. Dynamin-2 mutations in Charcot-Marie-Tooth neuropathy
  9. Dynamin-2 mutations in centronuclear myopathy associated with abnormalities in the nervous system
  10. Tissue-specific phenotypes of disease-related dynamin-2 mutations
  11. Dynamin-2 in AD
  12. Conclusions and perspectives
  13. Acknowledgements
  14. References

In addition to playing a role in the recycling of synaptic vesicles (Raimondi et al. 2011; Tanifuji et al. 2013) and signaling receptor turnover (Carroll et al. 1999; Kabbani et al. 2004), dynamin-2 also appears to be involved in the exocytotic release of transmitters at the nervous system. Evidences arising from neuroendocrine experimental models shed light on the role of this isoform during neurosecretion. In this regard, experiments in neuroendocrine pituitary corticotrope cells have showed that dynamin-2 interacts with G-proteins of the Trans-Golgi network via its PH domain, controlling vesicle trafficking as well as constitutive and CRH-induced secretion of β-endorphins (Yang et al. 2001). Dynamin also appears to modulate the quantal release of transmitters in adrenal chromaffin cells (Graham et al. 2002; González-Jamett et al. 2010). The mechanism is not completely understood but it has been proposed that dynamin acts at two different steps during exocytosis: (i) By favoring the expansion of the fusion pore (Anantharam et al. 2011; González-Jamett et al. 2013), and (ii) by allowing the re-closure of the vesicle (Holroyd et al., 2002; Tsuboi et al. 2004). The involvement of dynamin during neurosecretion could be mediated by its interaction with different partners. In this regard, we have recently demonstrated that endogenous dynamin-2 directs a Ca2+-dependent polymerization of the cortical actin cytoskeleton, where, as we proposed, dynamin-2 and actin act in concert regulating the fusion pore expansion and quantal release in neuroendocrine chromaffin cells (González-Jamett et al. 2013). Actin is dynamically rearranged under exocytosis-inducing stimuli (Malacombe et al. 2006), and it also controls the access of secretory vesicles to the plasma membrane in neuroendocrine cells (Vitale et al. 1993; Gasman et al. 2004; Wollman and Meyer 2012), in Calyx of Held synapses (Sakaba and Neher 2003) and in neuromuscular junctions (Cole et al. 2000). Furthermore, actin disruption partially reduced evoked-transmitter release in snake motor terminals (Cole et al. 2000), thus suggesting a role of actin filaments during neurotransmitter exocytosis. Given that dynamin-2 regulates actin organization in different cell types (Ochoa et al. 2000; Lee and De Camilli 2002; Yamada et al. 2009, 2013; Taylor et al. 2012; González-Jamett et al. 2013) by promoting actin filaments formation (Mooren et al. 2009; Gu et al. 2010), it is plausible that a dynamin-2/actin-dependent mechanism may also regulate neurotransmitter release in the nervous system.

Role of dynamins in neuronal process extension

  1. Top of page
  2. Abstract
  3. Dynamin structure and catalytic mechanism
  4. Dynamin isoforms and synaptic vesicle recycling
  5. Dynamin-2 controls endocytosis at the post-synaptic membrane
  6. Role of dynamin-2 in neurosecretion
  7. Role of dynamins in neuronal process extension
  8. Dynamin-2 mutations in Charcot-Marie-Tooth neuropathy
  9. Dynamin-2 mutations in centronuclear myopathy associated with abnormalities in the nervous system
  10. Tissue-specific phenotypes of disease-related dynamin-2 mutations
  11. Dynamin-2 in AD
  12. Conclusions and perspectives
  13. Acknowledgements
  14. References

Growing evidences highlight the importance of the three dynamin isoforms during neuronal morphogenesis and maturation of the nervous system. In this regard, an increase in dynamin expression can be observed in neural cells during post-natal development (Nakata et al. 1991) where it appears to play a role during neuritic process elongation (Nakata et al. 1991). This role of dynamins appears to be dependent on the ability of dynamin to drive actin dynamics, as well as on its association with the actin-binding protein cortactin (Gray et al. 2005; Kurklinsky et al. 2011; Yamada et al. 2013).

Regarding the role of dynamin-1 in neuritogenesis, Torre and co-workers observed a direct correlation between increasing dynamin-1 levels and neurite formation in hippocampal neurons in N1E-115 neuroblastoma cells. Conversely, dynamin-1 expression was significantly decreased in response to neurite retraction induced by serum addition (Torre et al. 1994). Furthermore, an antisense oligonucleotide-mediated reduction in dynamin-1 levels disrupted neurite formation in hippocampal neurons (Torre et al. 1994). Dynamin-1 also appears to drive growth cone filopodia, in association with the actin-binding protein cortactin (Yamada et al. 2013). The underlying mechanism involved a dynamin-1/cortactin complex that stabilizes actin filament bundles and favors a growth cone filopodia formation (Yamada et al. 2013). As growth cones are critical in supporting axonal guidance during the quest for synaptic targets, a dynamin-mediated proper growth cone formation could be necessary for synaptogenesis during development.

Also, a recent report highlights the role of dynamin-2 in neuronal growth cone migration (Kurklinsky et al. 2011). The latter authors observed that dynamin-2 co-localized with cortactin in contact points of migrating growth cones of neonatal rat hippocampal neurons, influencing the area of spreading, as well as the attachment and motility of the growth cones (Kurklinsky et al. 2011).

A role of dynamin-3 in dendritic spine maturation and morphological plasticity has been proposed in view of the fact that the spliced variant of dynamin-3, Dyn3baa, which is highly enriched at post-synaptic densities (Gray et al. 2003), reportedly favored dendritic filopodia formation in a manner dependent on its GTP-ase activity and association with cortactin (Gray et al. 2005). The Dyn3baa-cortactin complex was also found to regulate the actin cytoskeleton dynamics in developing neurons and promote the morphogenesis of dendritic spines (Gray et al. 2005). The authors suggested that a functional link between dynamin-3 and cortactin is required for the actin-dependent axonal growth cone formation in developing neurons.

Dynamin-2 mutations in Charcot-Marie-Tooth neuropathy

  1. Top of page
  2. Abstract
  3. Dynamin structure and catalytic mechanism
  4. Dynamin isoforms and synaptic vesicle recycling
  5. Dynamin-2 controls endocytosis at the post-synaptic membrane
  6. Role of dynamin-2 in neurosecretion
  7. Role of dynamins in neuronal process extension
  8. Dynamin-2 mutations in Charcot-Marie-Tooth neuropathy
  9. Dynamin-2 mutations in centronuclear myopathy associated with abnormalities in the nervous system
  10. Tissue-specific phenotypes of disease-related dynamin-2 mutations
  11. Dynamin-2 in AD
  12. Conclusions and perspectives
  13. Acknowledgements
  14. References

Charcot-Marie-Tooth disease (CMT) is a peripheral neuropathy characterized by loss of the sensation of touch, development of neuropathic pain, muscular weakness, and atrophy (Patzkó and Shy 2011). This disease is comprised of a demyelinating form (CMT1) with defective nerve conduction, and an axonal form (CMT2) where the breakdown of neuronal axons affects the amplitude, but not the velocity of nerve conduction (Claeys et al. 2009). In addition, dominant intermediate subtypes of CMT denoted DI-CMTA, DI-CMTB, DI-CMTC, and DI-CMTD have been identified, which exhibit characteristics of both demyelinating and axonal forms of the disease (Tanabe and Takei 2012).

Specific mutations in dynamin-2 have been linked to rare intermediate forms of CMT with a wide spectrum of severities, some of them associated with comorbidities such as neutropenia and early-onset cataracts (Claeys et al. 2009). Most of dynamin-2 CMT-linked mutations are localized in the N-terminal region of the PH domain in a ‘lipid-binding-pocket’ (Böhm et al. 2012), that includes three hydrophobic variable loops (VL1-VL3) that are involved in the insertion of dynamin into lipid membranes and also in membrane scission (Bethoney et al. 2009; Ramachandran et al. 2009). Congruently, the CMT-related mutant K562E was shown to be incapable of binding phospholipids (Kenniston and Lemmon 2010). Some exceptional CMT-related dynamin-2 mutations that are not located into the PH structural domain have also been described, such as the mutation G358R localized within the middle domain and the microdeletion T855-856del in the PRD (Claeys et al. 2009). How these and other CMT-linked mutations affect dynamin function and dynamin-dependent cellular processes is still poorly understood. It has been proposed that some of them negatively impact on CME. For instance, the over-expression of the mutants K562E, G358R, G537C, and L570H strongly impairs clathrin-mediated internalization of transferrin receptors in Schwann cells and motor neurons, and reduce myelination in a dorsal root ganglia culture model (Sidiropoulos et al. 2012). The mutants K558E and K562E also disturb CME when over-expressed in COS-7 and COS-1 cells, respectively (Tanabe and Takei 2009; Koutsopoulos et al. 2011). Nevertheless, an impairment in CME cannot be the common mechanism underlying all dynamin-2 CMT-linked mutations, since the over-expression of the middle-domain-mutant G358R did not alter CME in COS-1 cells (Koutsopoulos et al. 2011). The expression of the dynamin-2 mutant L570H, as well as the triple deletion D555del3 (D551del3 according to Züchner et al. 2005) in a dynamin-2 conditional null mouse fibroblast cell line that mimics the heterozygous condition, also had no effect on CME (Liu et al. 2011b). Yet, it significantly inhibited clathrin-independent endocytosis and post-Golgi vesicle trafficking (Liu et al. 2011b), thus further dismissing CME as the main affected mechanism in CMT. Remarkably, Tanabe and Takei (2009) observed that, besides having no effect on CME, the triple deletion D555del3 [named D551del3 by the authors, according to the sequence P50570-3 described in UniProt (http://www.uniprot.org/uniprot/P50570)] displayed a strong association to microtubules, and caused increased tubulin acetylation and aberrant ‘decorations’ that affected microtubule dynamics when over-expressed in COS-7 cells (Tanabe and Takei 2009). These findings suggest that other mechanisms, distinct from those of CME, could be also impaired in CMT-patients.

In summary, different dynamin-2-dependent processes could be implicated in the pathological mechanism underlying CMT, which probably depend on the type of mutation. Therefore, additional studies are necessary to understand how each one of these mutations affects dynamin-2 lipid-binding properties or GTPase activity, and how those altered properties impact on specific dynamin-dependent cellular processes.

Dynamin-2 mutations in centronuclear myopathy associated with abnormalities in the nervous system

  1. Top of page
  2. Abstract
  3. Dynamin structure and catalytic mechanism
  4. Dynamin isoforms and synaptic vesicle recycling
  5. Dynamin-2 controls endocytosis at the post-synaptic membrane
  6. Role of dynamin-2 in neurosecretion
  7. Role of dynamins in neuronal process extension
  8. Dynamin-2 mutations in Charcot-Marie-Tooth neuropathy
  9. Dynamin-2 mutations in centronuclear myopathy associated with abnormalities in the nervous system
  10. Tissue-specific phenotypes of disease-related dynamin-2 mutations
  11. Dynamin-2 in AD
  12. Conclusions and perspectives
  13. Acknowledgements
  14. References

Centronuclear myopathy (CNM) is a congenital condition clinically characterized by a progressive and generalized weakness and atrophy of the skeletal muscles (Jeannet et al. 2004; Fischer et al. 2006; Susman et al. 2010). Histologically, CNM is distinguished by an abnormal central location of muscle fiber nuclei, predominance and hypotrophy of Type 1 muscle fibers, and a radial distribution of the sarcoplasmic strands around the central nuclei (Jeannet et al. 2004).

Mutations in the gene encoding dynamin-2 have been demonstrated to underlie autosomal dominant mild forms of CNM (Cowling et al. 2012), although the severity spectrum of the disease varies from mild, with onset in adulthood, to severe with onset in the neonatal period (Bitoun et al. 2005, 2007, 2009).

An important number of dynamin-2 CNM-linked mutations are located in the C-terminal helix of the PH domain (Böhm et al. 2012), which appears to coordinate lipid binding with the GTPase activity (Kenniston and Lemmon 2010). In this regard, one of these mutations, A618T, exhibits a slightly higher basal GTPase activity that, compared with wild-type dynamin, is significantly enhanced in the presence of lipids (Kenniston and Lemmon 2010). Other mutations in the same region (S619L/W and V625del) display a dramatically increased basal GTPase activity that is not affected by the presence of lipids, thus pointing to a different lipid sensitivity of these mutants. Unlike A618T, which exists mainly as a tetramer in solution, S619L/W and V625del exhibit an enhanced propensity to self-assembly, forming more stable high order structures, a property that may explain their elevated basal GTPase activity (Kenniston and Lemmon 2010). The increased catalytic activity of these mutants suggests a critical regulatory function of the C-terminal helix of the PH domain on dynamin GTPase activation. Analogous regions in the PH domain of other GTP-ases, such as the Arf6-guanidine-nucleotide exchange-factor Grp1, have also been shown to be involved in modulating their own GTPase activity (Dinitto et al. 2007). The latter is probably related to the close proximity between G- and PH domains in their primary structure (Solomaha and Palfrey 2005). In this regard, Keniston and Lemmon proposed a mechanism through which the interaction of the PH domain and GED plays an auto-inhibitory role, thus preventing dynamin self-assembly in the absence of phospholipids (Kenniston and Lemmon 2010). In this context, CNM-linked mutations would mimic the effect of lipid-binding, in turn relieving this autoinhibition and favoring inter-domain interactions that promote dynamin oligomerization and GTPase activation.

Other dynamin-2 CNM-linked mutants (E368K, R369W, and R465W) clustered into the middle domain, which is critical for dynamin oligomerization (Ramachandran et al. 2007), have also shown a marked tendency to self-assemble, forming highly stable oligomers that exhibit an increased resistance to depolymerize in the presence of guanine nucleotides or high ionic strength solutions. These mutants also display an enhanced GTP-ase activity (Wang et al. 2010). An exacerbated dynamin activity could affect the GTP hydrolysis cycle and consequently impact on dynamin-dependent cellular processes.

Regarding the effects of CNM-linked mutations on dynamin-2-dependent cellular processes, the over-expression of the CNM-related mutants R465W, V625del, E650K (Bitoun et al. 2009), R522H, S619L, and P627H (Koutsopoulos et al. 2011) in COS cells reportedly impairs CME. Nevertheless, fibroblasts from patients carrying mutations R465W or S619L (Koutsopoulos et al. 2011), or from a heterozygous R465W knock-in mouse, a model for CNM (Durieux et al. 2010b), display a normal CME, suggesting that this process is not the main mechanism affected in CNM. In fact, data from Sandra Smith′s group, using an experimental design that mimics the heterozygous state, demonstrated that the CNM-linked mutations R465W and E368K significantly impair clathrin-independent endocytosis and vesicle trafficking from the Golgi to the plasma membrane without affecting CME (Liu et al. 2011b). Interestingly, and as aforementioned, these authors observed the same effects with the CMT-related mutants D555del3 and L570H, suggesting a common pathological mechanism causing CMT and CNM (Liu et al. 2011b).

Although CNM is intrinsically a muscle disorder, some clinical reports have shown that the peripheral nervous system is also affected in the disease (Mouren et al. 1982). In 2007, Echaniz-Laguna and co-workers observed a deficient peripheral nerve conduction in CNM-patients harboring the dynamin-2 middle-domain mutation E368Q (Echaniz-Laguna et al. 2007), suggesting a parallel nerve involvement. Similarly, Fischer et al. (2006) reported an impaired axonal peripheral nerve conduction in CNM-patients carrying the R369Q mutation (Fischer et al. 2006), thus supporting the notion of overlapping phenotypes in CNM and CMT. However, it is necessary to mention that a specific loss of function of dynamin-2 in the skeletal muscle could also compromise peripheral nerves, as observed by Tinelli et al. (2013). Using a cell type-specific gene ablation strategy in mice, these authors showed that dynamin-2 loss in the skeletal muscle is not only associated with a reduction in muscle mass and number of muscle fibers but it also leads to an altered neuromuscular junction organization and axonal degeneration (Tinelli et al. 2013). The over-expression of the CNM-linked mutation S619L in larvae of developing zebrafish, which currently express two orthologs genes of dynamin-2 (Gibbs et al. 2013a), also induced motor dysfunction accompanied by neuromuscular junction disorganization (Gibbs et al. 2013b), further suggesting that a defective neuromuscular transmission underlies the myopathy.

The central nervous system also could be affected in CNM (Serratrice et al. 1978). In fact a limited intelligence quotient (Böhm et al. 2012) and even ‘border-line mental retardation’ (Echaniz-Laguna et al. 2007) have been described in several CNM-cases, suggesting a concomitant impairment in the cognitive function in these patients. A correlation between congenital muscle diseases and cognitive impairment has also been observed in other neuromuscular disorders such as Duchenne′s and Myotonic dystrophies (D'Angelo and Bresolin 2006), both of which are associated with mutations in genes that, like dynamin-2, are important for the nervous system function (i.e., dystrophin and myotonic dystrophy protein kinase, respectively).

Since, as aforementioned, dynamin-2 plays critical roles in the nervous system; hence, an altered protein can impact on its function. Nevertheless, it does not necessarily mean that overlapping muscular and nervous affections co-exist in patients bearing CNM-linked dynamin-2 mutations.

Tissue-specific phenotypes of disease-related dynamin-2 mutations

  1. Top of page
  2. Abstract
  3. Dynamin structure and catalytic mechanism
  4. Dynamin isoforms and synaptic vesicle recycling
  5. Dynamin-2 controls endocytosis at the post-synaptic membrane
  6. Role of dynamin-2 in neurosecretion
  7. Role of dynamins in neuronal process extension
  8. Dynamin-2 mutations in Charcot-Marie-Tooth neuropathy
  9. Dynamin-2 mutations in centronuclear myopathy associated with abnormalities in the nervous system
  10. Tissue-specific phenotypes of disease-related dynamin-2 mutations
  11. Dynamin-2 in AD
  12. Conclusions and perspectives
  13. Acknowledgements
  14. References

In spite of the aforementioned findings that show some impairment in the nervous system in patients bearing CNM, it is currently accepted that both CNM and CMT dynamin-2 mutations cause tissue-specific phenotypes. How point mutations in an ubiquitously expressed and pleiotropic GTPase can selectively affect certain cellular processes leading to tissue-specific illness remains elusive. Some possible explanations are discussed below:

  • The presence of other dynamin isoforms, such as dynamin-1 and dynamin -3 in neuronal tissues, could compensate a dynamin-2 dysfunction associated to CNM-related mutations in the nervous system. In fact, an overlapping role of these dynamin isoforms has been observed for synaptic vesicle endocytosis (Raimondi et al. 2011).
  • CNM-linked dynamin-2 mutations display a significantly enhanced dynamin-2 GTPase activity (Kenniston and Lemmon 2010; Wang et al. 2010). Therefore, the pathogenic mechanism could be a consequence of a gain-of-function (Böhm et al. 2012). This could explain why fibroblasts from patients carrying the mutations R465W or S619L (Koutsopoulos et al. 2011), or motor neurons and Schwann cells expressing CNM-linked mutations (Sidiropoulos et al. 2012), exhibit a normal CME. Conversely, the skeletal muscle seems to be highly susceptible to this gain-of-function, as observed in animal models of CNM associated to dynamin-2 mutations (Durieux et al. 2010b; Cowling et al. 2011).
  • CMT mutations, particularly those localized in the N-terminal region of the PH domain, display decreased lipid binding and, consequently, dynamin functions that depend on its membrane association, such as CME (McMahon and Boucrot 2011), would be affected in turn. In some tissues such as Schwann cells, CME is highly necessary for myelination and it is significantly disrupted in the presence of CMT mutations (Sidiropoulos et al. 2012). Regrettably, at present there are no animal models of CMT-linked dynamin-2 mutations to address how those mutations impair CME and impact on other tissues.
  • The alternative splicing of dynamin-2 gives rise to four splice variants, which exhibit different intracellular distribution (Cao et al. 1998; Liu et al. 2008) and efficiency to support certain cellular functions, such as the p75 export from the Golgi (Liu et al. 2008). The variable regions of these splice variants are located in the middle domain, and none of the CNM or CMT mutations are localized in such regions (Durieux et al. 2010a). However, it is still unknown if some of these splice variants are predominantly expressed in a particular tissue and if their functions are affected similarly or distinctly by disease-linked dynamin-2 mutations.

Therefore, additional studies are necessary to understand the relationship between specific mutation sites, their impact on biochemical properties and functions of dynamin-2 splice variants, and the resulting pathological phenotype.

Dynamin-2 in AD

  1. Top of page
  2. Abstract
  3. Dynamin structure and catalytic mechanism
  4. Dynamin isoforms and synaptic vesicle recycling
  5. Dynamin-2 controls endocytosis at the post-synaptic membrane
  6. Role of dynamin-2 in neurosecretion
  7. Role of dynamins in neuronal process extension
  8. Dynamin-2 mutations in Charcot-Marie-Tooth neuropathy
  9. Dynamin-2 mutations in centronuclear myopathy associated with abnormalities in the nervous system
  10. Tissue-specific phenotypes of disease-related dynamin-2 mutations
  11. Dynamin-2 in AD
  12. Conclusions and perspectives
  13. Acknowledgements
  14. References

AD is the most common form of senile dementia, and it is characterized by a gradual loss of memory and other cognitive functions (Huang and Mucke 2012). It is widely accepted that the progressive accumulation of extracellular amyloid beta protein (Aβ) in senile plaques in brain regions that are important for memory and cognition are a leading cause of AD (Selkoe 2001). However, the neurofibrillary tangles, composed by hyperphosphorylated forms of tau, are other types of abnormal protein deposits that have been deemed as causative of neurodegeneration in AD patients (Cárdenas et al. 2012). Two main forms of AD have been described, a dominantly inherited early-onset AD (EOAD) and a late-onset AD (LOAD). EOAD is linked to specific missense mutations in the amyloid precursor protein (APP), presenilin1 (PS1) and presenilin2 (PS2) genes, the latter being components of the γ-secretase enzymatic complex that cleave APP to produce Aβ (Selkoe 2001). LOAD is classified as a polygenetic disease that represents the majority of cases, which includes sporadic forms of AD (Kamagata et al. 2009). The allele ε4 of the apolipoprotein E (APOE), mapped in the long arm of chromosome 19, has been suggested as an important risk factor (Borgaonkar et al. 1993), since ApoE-ε4 allele carriers undergo significantly greater accumulation of cortical extracellular Aβ (Selkoe 2002). The 19p13.2 region in the short arm of human chromosome 19 has been also identified as a susceptible locus (Wijsman et al. 2004), a region that includes the DNM2 gene (Aidaralieva et al. 2008). Interestingly, single nucleotide polymorphisms located into the intron-1 and 3-untranslated region of DNM2 gene significantly associate with LOAD in ApoE-ε4 non-carrier patients, indicating that DNM2 could also constitute a genetic factor contributing to LOAD (Aidaralieva et al. 2008).

In 2009, Kamagata and co-workers demonstrated that the expression of DNM2 gene appears to be reduced in LOAD patients (Kamagata et al. 2009) and that the over-expression of the dominant negative mutant of dynamin-2 K44A in SH-SY5Y neuroblastoma cells induced APP accumulation at the plasma membrane and increased Aβ secretion (Kamagata et al. 2009). These authors also found high levels of PS1 in the plasma membrane of DNM2K44A-expressing cells, suggesting a pathological mechanism in which DNM2 dysfunction may reduce APP endocytosis, thus leading to its accumulation in the plasma membrane, its degradation by PS1 and an increased production of Aβ (Kamagata et al. 2009) (Fig. 2). In agreement with this idea, previous studies demonstrated that Aβ can be produced by the γ-secretase complex directly at the cell surface, and that the over-expression of the dominant negative mutant of dynamin-1 K44A can induce the retention of the full-length APP at the plasma membrane, increasing Aβ secretion into the extracellular medium (Chyung and Selkoe 2003).

image

Figure 2. Ribbon type representation of dynamin-2 structure including disease-causing mutations. Centronuclear myopathy (CNM) and CMT-linked dynamin-2 mutations are displayed on the RasMOL generated crystal structure model for human dynamin-2, based on PDB 3ZVR (Ford et al. 2011). The GTPase (blue), middle (green), PH (yellow), and GED (red) domains are highlighted. CNM-related and CMT-related point mutations are labeled in red and blue, respectively. This representation lacks proline and arginine-rich domain (PRD).

Download figure to PowerPoint

While dynamin dysfunction can result in an increased Aβ secretion as a consequence of an impaired endocytosis, it is yet unclear why Aβ accumulation in LOAD patients is associated with a reduced expression of dynamin-1 and -2 (Yao et al. 2003; Kamagata et al. 2009). Kelly et al. (2005) showed that Aβ decreased the levels of both dynamin-1 and -2 in cultured hippocampal neurons. The authors proposed a bimodal mechanism that involves calpain-mediated proteolysis and down-regulation of dynamin expression (Kelly et al. 2005). Low levels of dynamin-1 were also observed in the hippocampus of an animal model for AD, the transgenic mouse Tg2576, probably as a consequence of an altered calpain activation (Kelly et al. 2005). These and other evidences suggest that Aβ-mediated dynamin dysfunction is associated with an increased Aβ secretion, further contributing to senile plaque accumulation and to the synaptic dysfunction observed in AD patients (Fig. 3). It is interesting that chromosome 10q24, which harbors the gene encoding dynamin-binding protein (DNMBP), a scaffold protein that connects dynamin-1 with actin-regulating proteins, has also been associated with LOAD (Kuwano et al. 2006), although the evidences for DNMBP as a genetic factor in LOAD are still controversial (Minster et al. 2008).

image

Figure 3. Amyloid beta protein (Aβ)-induced dynamin-2 depletion leads to amyloid precursor protein (APP) accumulation in the plasma membrane, increasing Aβ secretion. During normal dynamin-2 activity, APP is internalized via dynamin-dependent endocytosis and later digested in endosomal/lysosomal compartments. LOAD patients display low levels of dynamin-2 (Kamagata et al. 2009). These low dynamin levels impair endocytosis and promote accumulation of APP and PS1 at the plasma membrane level, favoring APP degradation by the γ-secretase complex, and consequently increasing Aβ production and favoring senile plaque formation (Chyung and Selkoe 2003; Kamagata et al. 2009). In turn, Aβ accumulation reduces dynamin-2 expression (Kelly et al. 2005), thus generating a positive feedback that causes further progression of this pathological mechanism. This diagram is based on that published by Kamagata et al. (2009).

Download figure to PowerPoint

Conclusions and perspectives

  1. Top of page
  2. Abstract
  3. Dynamin structure and catalytic mechanism
  4. Dynamin isoforms and synaptic vesicle recycling
  5. Dynamin-2 controls endocytosis at the post-synaptic membrane
  6. Role of dynamin-2 in neurosecretion
  7. Role of dynamins in neuronal process extension
  8. Dynamin-2 mutations in Charcot-Marie-Tooth neuropathy
  9. Dynamin-2 mutations in centronuclear myopathy associated with abnormalities in the nervous system
  10. Tissue-specific phenotypes of disease-related dynamin-2 mutations
  11. Dynamin-2 in AD
  12. Conclusions and perspectives
  13. Acknowledgements
  14. References

All three dynamin isoforms present in mammals are expressed in the nervous system, where they control critical neuronal processes such as the recycling of synaptic vesicles, neurite morphogenesis, and the turnover of signaling receptors. Nevertheless, the notion that each dynamin isoform specializes in given processes is not supported at present, as studies in dynamin-1 and -3 knock-out animals have shown that they are still capable to sustain synaptic transmission (Raimondi et al. 2011). The latter points to overlapping roles of the three dynamin isoforms in the nervous system. Nevertheless, the knock-out of dynamin 2, the only ubiquitously expressed dynamin isoform, leads to an early embryonic lethality, suggesting that this isoform plays an essential role during the embryonic development (Ferguson et al. 2009). In this regard, and contrary to prior thought, dynamin-2 seems to not only perform a housekeeping role in neuronal tissues but also exert an essential role in nervous system function.

Hitherto, more than twenty mutations have been identified in the gene encoding for dynamin-2, leading to CNM and CMT (see Figs 1 and 2). However, the underlying pathological mechanisms are still elusive, and additional studies are still necessary to better understand how these mutations impact on dynamin-2 structure and function. The mechanism by which some mutations lead to a tissue-specific disease is also unclear, while others (E368Q; R369Q) appear to cause a disorder that involves both muscular and nerve impairment (Fischer et al. 2006; Echaniz-Laguna et al. 2007). Probably, the location of the mutation specifically affects a given dynamin-2 biochemical property and hence impacts on dynamin-2-dependent processes differently. Perhaps, certain dynamin-2-dependent processes are difficult to compensate by some tissues, rendering them more susceptible to be deregulated by specific disease-causing mutations. Certainly, much work is needed to understand the cellular processes in specific tissues where dynamin-2 participates, and how those are affected by point disease-related mutations.

Finally, dynamin-2 expression appears to play a critical role in AD, as it also regulates APP endocytosis and its processing by the γ-secretase complex. In turn, Aβ accumulation appears to reduce dynamin-2 levels (Kelly et al. 2005), thus producing a positive loop that worsens the pathology. However, additional experiments are required to establish the link between dynamin-2 and AD.

As research progresses toward understanding one of the final frontiers in disease, such as establishing the link between genetic mutation and different ailments, it is clear that the understanding of the cellular function and signaling pathway checkpoints is a critical step in the quest for relevant therapies. Indeed, as gene therapy involving gain of function, such as the transfection of the full native gene or sequences that lack the mutation (i.e., exon skipping) appears as the final solution, the development of lead compounds that target signaling mechanisms can provide for more therapeutical targets. It is possibly from a combination of both approaches that a final solution may arise.

References

  1. Top of page
  2. Abstract
  3. Dynamin structure and catalytic mechanism
  4. Dynamin isoforms and synaptic vesicle recycling
  5. Dynamin-2 controls endocytosis at the post-synaptic membrane
  6. Role of dynamin-2 in neurosecretion
  7. Role of dynamins in neuronal process extension
  8. Dynamin-2 mutations in Charcot-Marie-Tooth neuropathy
  9. Dynamin-2 mutations in centronuclear myopathy associated with abnormalities in the nervous system
  10. Tissue-specific phenotypes of disease-related dynamin-2 mutations
  11. Dynamin-2 in AD
  12. Conclusions and perspectives
  13. Acknowledgements
  14. References