Gene name - spastin
Synonyms - D-Spastin, Dspastin
Cytological map position - 95D9
Function - enzyme
Symbol - spas
FlyBase ID: FBgn0039141
Genetic map position - 3R
Classification - ATPase activity, microtubule interacting and organelle transport (MIT) domain
Cellular location - cytoplasmic
|Recent literature||Papadopoulos, C., Orso, G., Mancuso, G., Herholz, M., Gumeni, S., Tadepalle, N., Jungst, C., Tzschichholz, A., Schauss, A., Honing, S., Trifunovic, A., Daga, A. and Rugarli, E. I. (2015). Spastin binds to lipid droplets and affects lipid metabolism. PLoS Genet 11: e1005149. PubMed ID: 25875445
Mutations in SPAST, encoding spastin, are the most common cause of autosomal dominant hereditary spastic paraplegia (HSP). HSP is characterized by weakness and spasticity of the lower limbs, owing to progressive retrograde degeneration of the long corticospinal axons. Spastin is a conserved microtubule (MT)-severing protein, involved in processes requiring rearrangement of the cytoskeleton in concert to membrane remodeling, such as neurite branching, axonal growth, midbody abscission, and endosome tubulation. Two isoforms of spastin are synthesized from alternative initiation codons (M1 and M87). This study shows that spastin-M1 can sort from the endoplasmic reticulum (ER) to pre- and mature lipid droplets (LDs). A hydrophobic motif comprised of amino acids 57 through 86 of spastin was sufficient to direct a reporter protein to LDs, while mutation of arginine 65 to glycine abolished LD targeting. Increased levels of spastin-M1 expression reduced the number but increased the size of LDs. Expression of a mutant unable to bind and sever MTs caused clustering of LDs. Consistent with these findings, ubiquitous overexpression of Dspastin in Drosophila led to bigger and less numerous LDs in the fat bodies and increased triacylglycerol levels. In contrast, Dspastin overexpression increased LD number when expressed specifically in skeletal muscles or nerves. Downregulation of Dspastin and expression of a dominant-negative variant decreased LD number in Drosophila nerves, skeletal muscle and fat bodies, and reduced triacylglycerol levels in the larvae. Moreover, reduced amount of fat stores were found in intestinal cells of worms in which the spas-1 homologue was either depleted by RNA interference or deleted. Taken together, these data uncovers an evolutionarily conserved role of spastin as a positive regulator of LD metabolism and open up the possibility that dysfunction of LDs in axons may contribute to the pathogenesis of HSP.
|Julien, C., et al. (2016). Conserved pharmacological rescue of hereditary spastic paraplegia-related phenotypes across model organisms. Hum Mol Genet [Epub ahead of print]. PubMed ID: 26744324
Hereditary spastic paraplegias (HSPs) are a group of neurodegenerative diseases causing progressive gait dysfunction. Over 50 genes have now been associated with HSP. Despite the recent explosion in genetic knowledge, HSP remains without pharmacological treatment. Loss-of-function mutation of the SPAST gene, also known as SPG4, is the most common cause of HSP in patients. SPAST (Drosophila Spastin) is conserved across animal species and regulates microtubule dynamics. Recent studies have shown that it also modulates endoplasmic reticulum (ER) stress. This study utilized null SPAST homologues in C. elegans, Drosophila, and zebrafish to tested FDA approved compounds known to modulate ER stress in order to ameliorate locomotor phenotypes associated with HSP. Locomotor defects found in all of the spastin models could be partially rescued by phenazine, methylene blue, N-acetyl-cysteine, guanabenz and salubrinal. In addition, established biomarkers of ER stress levels correlated with improved locomotor activity upon treatment across model organisms. These results provide insights into biomarkers and novel therapeutic avenues for HSP.
|Rao, K., Stone, M. C., Weiner, A. T., Gheres, K. W., Zhou, C., Deitcher, D. L., Levitan, E. S. and Rolls, M. M. (2016). Spastin, atlastin and ER relocalization are involved in axon, but not dendrite, regeneration. Mol Biol Cell 27(21):3245-3256. PubMed ID: 27605706
Mutations in over 50 genes including spastin and atlastin lead to Hereditary Spastic Paraplegia (HSP). It was previously demonstrated that reduction of spastin leads to a deficit in axon regeneration in a Drosophila model. Axon regeneration was similarly impaired in neurons when HSP proteins atlastin, seipin and spichthyin were reduced. Impaired regeneration was dependent on genetic background, and was observed when partial reduction of HSP proteins was combined with expression of dominant-negative microtubule regulators, suggesting HSP proteins work with microtubules to promote regeneration. Microtubule rearrangements triggered by axon injury were, however, normal in all genotypes. Other markers were examined to identify additional changes associated with regeneration. While mitochondria, endosomes and ribosomes did not exhibit dramatic repatterning during regeneration, the endoplasmic reticulum (ER) was frequently concentrated near the tip of the growing axon. In atlastin RNAi and spastin mutant animals, ER accumulation near single growing axon tips was impaired. ER tip concentration was only observed during axon regeneration, and not during dendrite regeneration. In addition, dendrite regeneration was unaffected by reduction of spastin or atlastin. It is proposed that the HSP proteins Spastin and Atlastin promote axon regeneration by coordinating concentration of the ER and microtubules at the growing axon tip.
Hereditary Spastic Paraplegia (HSP) is a devastating neurological disease causing spastic weakness of the lower extremities and eventual axonal degeneration. Over 20 genes have been linked to HSP in humans; however, mutations in one gene, spastin (SPG4), are the cause of >40% of all cases. Spastin is a member of the ATPases Associated with diverse cellular Activities (AAA) protein family, and contains a microtubule interacting and organelle transport (MIT) domain. Previous work in cell culture has proposed a role for Spastin in regulating microtubules. Employing Drosophila transgenic methods for overexpression and RNA interference (RNAi), the role of Spastin in vivo was investigated in vitro. Drosophila Spastin (DSpastin) is enriched in axons and synaptic connections. At neuromuscular junctions (NMJ), Dspastin RNAi causes morphological undergrowth and reduced synaptic area. Moreover, Dspastin overexpression reduces synaptic strength, whereas Dspastin RNAi elevates synaptic currents. By using antibodies against posttranslationally modified alpha-Tubulin, it has been found that Dspastin regulates microtubule stability. Functional synaptic defects caused by Dspastin RNAi and overexpression were pharmacologically alleviated by agents that destabilize and stabilize microtubules, respectively. It is concluded that loss of Dspastin in Drosophila causes an aberrantly stabilized microtubule cytoskeleton in neurons and defects in synaptic growth and neurotransmission. These in vivo data strongly support previous reports, providing a probable cause for the neuronal dysfunction in spastin-linked HSP disease. The role of Spastin in regulating neuronal microtubule stability suggests therapeutic targets for HSP treatment and may provide insight into neurological disorders linked to microtubule dysfunction (Trotta, 2004).
Hereditary Spastic Paraplegia disease causes dysfunction in the corticospinal tract, causing progressive weakness and spasticity in the lower extremities (Reid, 2003). The age of HSP onset ranges from the first to the sixth decade, and symptoms also range widely in HSP patients, making clinical data difficult to interpret. HSP inheritance has been linked to at least 20 genetic loci (Reid, 2003), but less than half of these genes have been molecularly identified. The characterized HSP-linked genes encode a diverse range of products, including cell adhesion molecules [L1-CAM, Jouet, 1994], myelination proteins [DM20, PLP (Suagier-Veber, 1994)], and microtubule motor proteins [KIF5A; (Reid, 2002)]. However, approximately 40% of HSP disease cases are linked to mutations of the SPG4 locus, which encodes Spastin (Fonknechten, 2000; Hazan, 1999). Spastin is an ATPase that contains a microtubule-interacting domain (Ciccarelli, 2003), suggesting an active role in cytoskeleton interactions. Interestingly, a number of other HSP-linked genes have similar molecular features, including Spartin and Paraplegin (Zhao, 2001; DeMichele, 1998; Casari, 1998. In vitro experiments have confirmed that Spastin binds polymerized microtubules through its N terminus and demonstrate a requirement for Spastin ATPase activity to release it from microtubules (Patel, 2002). These data suggest that Spastin normally plays some role involving the microtubule cytoskeleton in neurons. Loss of this Spastin function leads to progressive neuronal dysfunction, culminating in axonal degeneration (Trotta, 2004 and references therein).
Microtubules affect multiple facets of neuronal development and function, with pronounced roles in growth cone motility and synaptic elaboration of both pre- and post-synaptic structures being documented in a range of vertebrate and invertebrate systems. The Drosophila neuromuscular junction (NMJ) synapse has been a particularly useful genetic system in which to assay the role of the microtubule cytoskeleton and model its role in inherited neurological disease. For example, mutations in microtubule-interacting Futsch (MAP1b) and VAP-33A alter synaptic morphology, thereby decreasing bouton number and increasing bouton size. These data have produced the hypothesis that decreased microtubule stability leads to growth and elaboration of the presynaptic terminal. In addition, the delivery of proteins and organelles required for neurotransmission, both pre- and post-synaptically, depends on microtubules and associated motor proteins. Acute pharmacological perturbation of microtubules at the Drosophila NMJ has not revealed major defects in synaptic vesicle cycling. In contrast, loss or gain of expression studies of the Drosophila fragile X-related (dfxr) gene have shown that it regulates Futsch to cooperatively modulate both synaptic architecture and neurotransmission strength. These studies indicate that microtubules play key functions in the control of synaptic structure and function and that mutations perturbing microtubule dynamics are causatively linked with inherited neurological disorders (Trotta, 2004 and references therein).
Drosophila contains a highly conserved spastin homolog (Dspastin) (Kammermeier, 2003). The Drosophila NMJ system therefore provides an attractive opportunity to model HSP and to test the in vivo function of D-Spastin. Drosophila Spastin is shown to be localized at the NMJ and specifically enriched in presynaptic boutons. The dosage of Dspastin was altered in a targeted fashion in the nervous system with transgenic RNA interference (RNAi) and overexpression techniques, and the consequence was assayed with electrophysiology and confocal imaging. Knockdown of D-Spastin (i.e., reduction of Dspastin mRNA through RNAi) in neurons results in synaptic undergrowth. Conversely, loss of D-Spastin in neurons causes much stronger neurotransmission, whereas D-Spastin overexpression weakens synaptic function. Microtubule stability was assayed locally within the synapse by using antibodies specific to modified Tubulin, which label stable and long-lived microtubules. Knockdown of D-Spastin results in accumulation of stabilized Tubulin, whereas D-Spastin overexpression conversely reduces stable microtubules. Direct pharmacological manipulation of microtubule stability, to counter the observed defects, rescues the functional synaptic defects in both classes of Dspastin mutant. Together, these data provide the first in vivo evidence that Spastin regulates microtubule stability and show that this function of Spastin strongly modulates both synaptic architecture and neurotransmission strength (Trotta, 2004).
Mutations in the spastin (SPG4) gene are the leading cause of Hereditary Spastic Paraplegia (HSP), a progressively debilitating neurological disease characterized by dysfunction of corticospinal tract neurons, followed by eventual axonal degeneration (Reid, 2003). The Spastin protein is an ATPase that associates with microtubules in an ATP-dependent manner, leading to the hypothesis that loss of Spastin may cause microtubule-related defects in neuronal function required for axonal maintenance. Drosophila has a single, highly conserved spastin homolog (Dspastin) containing both ATPase and microtubule-interacting domains (Kammermeier, 2003). The Drosophila neuromusculature has previously proven important for assaying microtubule mechanisms relevant to neuronal function and neurological disease, making it a logical choice for modeling the neurological role of Spastin in vivo. A targeted transgenic approach for RNAi knockdown and overexpression of Spastin in the nervous system was used to assay the in vivo requirement for Spastin specifically in neurons (Trotta, 2004).
Drosophila Spastin is widely expressed in many cell types, including both muscles and neurons, where it is present throughout the cell but undetectable in the nucleus. These findings support the conclusion that Spastin is a cytoplasmic protein, in contrast to some previous reports that mammalian Spastin is a nuclear protein (Charvin, 2003). Consistent with recent RNA in situ data in the Drosophila embryo (Kammermeier, 2003), D-Spastin protein is highly enriched in neurons, and the protein is particularly abundant in the presynaptic bouton compartments in synaptic connections. Strongly driven Dspastin RNAi results in pupal lethality, showing that the gene is essential for viability. Strongly driven Dspastin overexpression causes cell death and neurodegeneration, showing that Dspastin gene dosage must be carefully controlled. Up- or down-regulation of Dspastin dosage results in striking alterations in synaptic morphology and function. Targeted removal of D-Spastin from neurons causes the loss of synaptic area but, paradoxically, strengthened neurotransmission. In contrast, overexpression of D-Spastin in neurons reduces synaptic efficacy. As predicted by previous in vitro studies in mammals (Errico, 2002), Drosophila Spastin closely regulates the stability of the neuronal microtubule cytoskeleton in vivo. Loss of D-Spastin leads to a dramatic increase in stabilized microtubules, which, in the synaptic terminal, inappropriately ramify beyond their normal axonal compartments into the synaptic boutons. Overexpression of D-Spastin causes an erosion of the stabilized microtubule network. These results indicate that D-Spastin acts as a negative regulator of microtubule stability. The relationship between neuronal function and microtubule stability was investigated in an effort to identify a causative mechanism to account for spastin-mediated HSP disease. Most importantly, misregulation of synaptic function in both the loss and overexpression of Dspastin is rescued by pharmacologically modifying microtubule stability, as appropriate, to counter the observed changes in microtubule stability in both mutant conditions. Taken together, these results strongly suggest that D-Spastin acts locally in the synapse to control microtubule stability and the cytoskeletal compartmentalization of functional synaptic domains, and that this regulation plays a critical role both in maintaining the synapse and in modulating the synaptic efficacy dictating neurotransmission strength (Trotta, 2004).
The elaboration of synaptic architecture, determining both the maintenance and extent of synaptic area, is well documented to be microtubule dependent. At the Drosophila NMJ, multiple known regulators of microtubule stability have been shown to modulate the size and complexity of the terminal arbor. The current study shows that D-Spastin has a positive role in maintaining the synapse by encouraging growth through increasing the dynamic instability of microtubules. This conclusion is consistent with previous results showing that synaptic terminals with stabilized microtubules are less prone to branching and budding. The hypothesis is that proteins that stabilize microtubules should inhibit synaptic growth, whereas proteins that destabilize microtubules should facilitate synaptic growth. D-Spastin would therefore be predicted to promote synaptic growth by destabilizing microtubules. The biochemical action of Spastin in destabilizing microtubules, which has been shown previously in mammalian cell culture (Errico, 2002), and the results here that D-Spastin destabilizes microtubules within individual Drosophila synaptic terminals in situ, strongly corroborates this hypothesis (Trotta, 2004).
The effects on synaptic growth may have a compounding influence on synaptic function. However, an inverse relationship is observed between changes in synaptic area and synaptic transmission strength after manipulation of D-Spastin dosage. Knockdown of D-Spastin results in much smaller (area reduced 50%) but much stronger (EJC amplitude increased 50%) NMJ synapses. The increase in synaptic function may be an overcompensation for the synaptic size, or these two synaptic parameters may be misregulated completely independently; however, the interdependency of these data is unclear. Independent regulation of synaptic growth and efficacy has been shown repeatedly during the course of analyzing Drosophila mutations in several genes. The fact that opposing defects in synaptic transmission caused by either Dspastin knockdown or overexpression can be rescued with acute pharmacological treatments to counter the misregulation of microtubule dynamics suggests that D-Spastin plays a direct role in modulating synaptic efficacy, independent from a separable role in modulating synaptic growth and architecture (Trotta, 2004).
Synaptic function appears sensitive to the stability of the microtubule cytoskeleton. The exact nature of this dependency is unclear, since stabilized microtubules are normally tightly restricted to axons and interbouton connectives and excluded from presynaptic boutons. One obvious possibility is that alteration of microtubule stability influences the microtubule-dependent transport of proteins or organelles required for maintained synaptic vesicle cycling and neurotransmitter release (Kammermeier, 2003). D-Spastin protein is normally tightly restricted to synaptic boutons, which lack stabilized microtubules, and absent from interbouton connectives, which contain stabilized microtubules. It is tempting to speculate that the D-Spastin domains determine the domains of stabilized versus unstabilized Tubulin within the synaptic arbor. Changes in D-Spastin levels not only regulate the abundance of stabilized microtubules but also determine their synaptic localization; loss of D-Spastin causes stabilized microtubules to inappropriately invade synaptic boutons. The defects in neurotransmission in both loss and gain of D-Spastin conditions were strikingly rescued by pharmacologically shifting microtubule dynamics away from the aberrant condition. Interestingly, the microtubule-modulating drugs that effectively restore aberrant neurotransmission in both Dspastin loss and overexpression mutants do not significantly alter basal synaptic function in wild-type synapses. This shows that alteration in D-Spastin levels is first required to prime the system for responsiveness. It appears that alterations in Dspastin dosage lead to a constitutive shifts in the microtubule equilibrium, which makes the synapse acutely sensitive to additional changes in microtubule stability. Why this should be the case is presently unclear (Trotta, 2004).
Hereditary Spastic Paraplegias are linked to multiple genetic loci-encoding proteins apparently involved in disparate cellular activities. However, one prominently affected cellular mechanism appears to be alteration in microtubule cytoskeleton formation, stability, and/or function. Specifically, well over half of all HSP disease cases are linked to microtubule motor proteins (e.g., KIF5A) or other proteins with microtubule-interacting domains (e.g., Spartin, Paraplegin, and Spastin). The most commonly mutated gene to cause HSP is Spastin, making it the obvious choice to model the mechanistic defects caused by loss of function of this suite of related microtubule-interacting proteins. This study shows for the first time that Spastin is enriched at the synapse, controlling both functional processes driving synaptic transmission, as well as developmental pathways determining synaptic growth and maintenance. Alteration in microtubule stability locally at the synapse appears to be a primary consequence that follows changes in Spastin dosage, consistent with the hypothesis that Spastin is a negative regulator of the microtubule cytoskeleton stability. By extension, it is likely that misregulation of microtubule stability is the primary cause of HSP disease, both in the 40% of disease cases directly associated with mutations in the spastin (SPG4) gene as well as cases linked to Atlastin, Spartin, Paraplegin, and related proteins. This mechanistic insight suggests that therapeutic targets aimed at the mechanism of microtubule dynamic instability are likely to prove effective in the treatment of Hereditary Spastic Paraplegia disease (Trotta, 2004).
Spastin, the most common locus for mutations in hereditary spastic paraplegias, and katanin are related microtubule-severing AAA ATPases involved in constructing neuronal and non-centrosomal microtubule arrays and in segregating chromosomes. The mechanism by which spastin and katanin break and destabilize microtubules is unknown, in part owing to the lack of structural information on these enzymes. This study reports the X-ray crystal structure of the Drosophila spastin AAA domain and provides a model for the active spastin hexamer generated using small-angle X-ray scattering combined with atomic docking. The spastin hexamer forms a ring with a prominent central pore and six radiating arms that may dock onto the microtubule. Helices unique to the microtubule-severing AAA ATPases surround the entrances to the pore on either side of the ring, and three highly conserved loops line the pore lumen. Mutagenesis reveals essential roles for these structural elements in the severing reaction. Peptide and antibody inhibition experiments further show that spastin may dismantle microtubules by recognizing specific features in the carboxy-terminal tail of tubulin. Collectively, these data support a model in which spastin pulls the C terminus of tubulin through its central pore, generating a mechanical force that destabilizes tubulin-tubulin interactions within the microtubule lattice. This work also provides insights into the structural defects in spastin that arise from mutations identified in hereditary spastic paraplegia patients (Roll-Mecak, 2008).
The combination of X-ray crystallography, SAXS ab initio reconstructions and structure-guided mutagenesis provides the first structural information on microtubule-severing proteins and allows the proposal of a molecular model for spastin-mediated severing. Owing to their similar domain organization and high sequence similarity, this model probably pertains to katanin as well. It is proposed that face A of the spastin AAA ring docks onto the microtubule, placing the positively charged N-terminal pore entrance in contact with the negatively charged C terminus of tubulin. The translocation from face A to face B would correspond to the direction of substrate translocation proposed for the distantly related AAA ATPases ClpX, ClpA and ClpB. The linker and MIT domains extending from the ring would make additional contacts with the microtubule, thus increasing microtubule avidity and potentially stabilizing the hexamer on the microtubule. On the basis of affinity measurements, only a subset of the six arms is likely to make strong binding interactions with the microtubule (Roll-Mecak, 2008).
It is proposed that the tubulin polypeptide is threaded through the pore, perhaps driven by nucleotide-driven conformational changes of the pore loops. However, spastin may not need to completely translocate the tubulin polypeptide substrate, but instead just grip the C-terminal tubulin tail and exert mechanical 'tugs' that might partially unfold tubulin or locally destabilize protomer-protomer interactions, leading to catastrophic breakdown of the microtubule lattice. It also remains possible that the MIT domains could participate in this nucleotide-driven process by binding and 'feeding' the C-terminal tails to the pore. Further biophysical characterization will be needed to decipher the structural details of substrate recognition and mechanical force production. The data also suggest that spastin may selectively recognize post-translationally modified tubulins ('Glu' tubulins) that are part of stable microtubules. Consistent with this idea, loss of spastin in Drosophila results in the accumulation of stabilized polyglutamylated tubulin in neurons and spastin knockout mice show axonal swellings enriched in detyrosinated, stable microtubules. The structure also provides the first glimpse into how spastin disease mutations contribute to spastin dysfunction and disease, most of which are likely to be involved in destabilizing protomer-protomer interactions, microtubule- or ATP-binding; in such cases, spastin-linked HSP is probably caused by haploinsufficiency and not a dominant negative effect. Further elucidation of the mechanistic details of how spastin interacts with particular tubulin isoforms and post-translational modifications and leads to microtubule destabilization may provide insight into the origin of spastin paraplegias and potential treatments for this disease (Roll-Mecak, 2008).
The human SPG4 locus encodes the spastin gene, which is responsible for the most prevalent form of autosomal dominant hereditary spastic paraplegia (AD-HSP), a neurodegenerative disorder. The predicted gene product CG5977 is identified in this study as the Drosophila homolog of the human spastin gene, with much higher sequence similarities than any other related AAA domain protein in the fly. Furthermore a new potential transmembrane domain is reported in the N-terminus of the two homologous proteins. During embryogenesis, the expression pattern of Drosophila spastin becomes restricted primarily to the central nervous system, in contrast to the ubiquitous expression of the vertebrate spastin genes. Given this nervous system-specific expression, it will be important to determine if Drosophila spastin loss-of function mutations also lead to neurodegeneration (Kammermeier, 2003).
Close inspection of the Drosophila sequence and of the alignment of Drosophila and human spastin reveals, besides the two known highly conserved domains [namely the AAA domain (70% identity) of 300 residues, and the MIT (microtubule interacting and trafficking molecules) domain (55% identity) of about 100 residues], an additional conserved region of 33 residues near the N-terminus. The three domains are also recognized with SMART domain analysis as AAA, MIT and as a putative transmembrane domain, respectively. Since the similarity in the N-terminal domain extends beyond the hydrophobic core that could be a transmembrane domain, this domain is called a TM+ domain. This domain appears to be specific for spastin proteins. MIT domains, and to a comparable degree AAA domains, also appear in unrelated proteins; one was discovered in spartin, the gene responsible for SPG20 (Patel, 2002). The association of MIT and AAA domains, however, appears to be evolutionarily old since even the highly conserved vacuolar protein sorting 4 (VPS4) subfamily shows this domain structure. The members of the AAA family most closely related to spastin, however, are not VPS4 but members of the fidgetin subfamily, which may be of neurological interest because of the shaking phenotype in the fidget mouse (Kammermeier, 2003 and references therein).
Paraplegin, the gene responsible for SPG7, however, is only distantly related to spastin, and the paraplegin protein is localized to the mitochondria and contains a peptidase domain. Since spastin has already been shown not to be targeted to the mitochondria, but appears to be localized to the nucleus (Charvin, 2003), it will be important to test whether the potential transmembrane domain with the highly conserved neighboring residues could be involved in localization of spastin to the nuclear membrane system (Kammermeier, 2003 and references therein).
Human spastin is comprised of several functional domains, all of which are conserved in the Drosophila gene (Kammermeier, 2003). Overall, Dspastin exhibits 48% identity and 60% similarity at the amino acid level with the human homolog. Both human and Drosophila proteins contain a predicted transmembrane domain (40% identity), a microtubule interacting and organelle transport (MIT) domain (55% identity) and an AAA domain (70% identity). The high degree of conservation of these motifs suggests that the function of Spastin has been tightly maintained throughout evolution (Trotta, 2004).
date revised: 2 October 2004
Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.
The Interactive Fly resides on the
Society for Developmental Biology's Web server.