: Biological Overview | References
Gene name - sidestep
Cytological map position - 97F6-97F10
Function - ligand
Keywords - motor axon guidance
Symbol - side
FlyBase ID: FBgn0016061
Genetic map position - 3R:23,202,304..23,244,247 [-]
Cellular location - surface transmembrane
During development of the Drosophila nervous system, migrating motor axons contact and interact with different cell types before reaching their peripheral muscle fields. The axonal attractant Sidestep (Side) is expressed in most of these intermediate targets. This study shows that motor axons recognize and follow Side-expressing cell surfaces from the ventral nerve cord to their target region. Contact of motor axons with Side-expressing cells induces the down-regulation of Side. In the absence of Side, the interaction with intermediate targets is lost. Misexpression of Side in side mutants strongly attracts motor axons to ectopic sites. Evidence is provided that, on motor axons, Beaten path Ia (Beat) functions as a receptor or part of a receptor complex for Side. In beat mutants, motor axons no longer recognize Side-expressing cell surfaces. Furthermore, Beat interacts with Side both genetically and biochemically. These results suggest that the tracing of Side-labeled cell surfaces by Beat-expressing growth cones is a major principle of motor axon guidance in Drosophila (Siebert, 2009).
Wiring of the nervous system is a precisely controlled process that includes axon outgrowth, axon pathfinding, target recognition, and synapse formation. Outgrowing axons of the same subtype often join and migrate collectively toward their targets, a process called selective fasciculation. Upon arrival in the target region, a subset of growth cones must inevitably defasciculate from the main pathway to select a specific synaptic target. Advancing growth cones are believed to express appropriate receptors that detect and evaluate relevant guidance molecules presented by surrounding cells and tissues. Despite the discovery of several conserved regulators of axonal pathfinding, the spatiotemporal sequence of molecular events that steer growth cones is still unclear. How exactly do growth cones recognize and interpret guidance cues to make pathway decisions? Over the years, several models have been proposed to describe general features of axon navigation across species, including the chemoaffinity hypothesis, the guidepost cell hypothesis, the blueprint hypothesis, and the labeled pathways hypothesis (Siebert, 2009).
The labeled pathways hypothesis postulates that a small number of early differentiating neurons pioneer a stereotypic array of differentially labeled axonal pathways. These primary pathways are selectively recognized by subsequently developing growth cones. While follower axons migrate along a homogenous substrate toward their target region, pioneer axons have to interact with many different substrates, actively search for guidance cues, and interpret relevant guidance information. Pathfinding decisions for pioneer axons are greatly simplified if directional cues would be present in the substrate. However, the experimental evidence for the existence of such substrate pathways is scarce. It has been noted that projections of sensory axons in homeotic mutants of Drosophila involved the specific recognition of 'pre-existing trails.' Similarly, axons emanating from transplanted eyes or Mauthner neurons in Xenopus have been observed to project along defined tracks called 'substrate pathways'. Furthermore, it has been reported that chick optic axons are guided along a preformed adhesive pathway. Despite the appealing simplicity of substrate pathways, the molecular tags that provide directional information have not been identified (Siebert, 2009).
One possible molecule that could serve as a directional cue for Drosophila motor axons is the attractant Sidestep (Side), a transmembrane protein of the immunoglobulin superfamily that is dynamically expressed during embryogenesis but prominently enriched in muscles when motor axons arrive in their target areas (Sink, 2001). In side mutant embryos, motor axons fail to defasciculate at key choice points and hence bypass their muscle targets, suggesting that Side functions as a target-derived attractant (Sink, 2001; de Jong, 2005). Interestingly, mutations in beaten path Ia (beat) lead to similar axon guidance phenotypes, and like Side, Beat has been shown to regulate axon defasciculation at choice points (Fambrough, 1996; Holmes, 1999; Sink, 2001). Based on primary structure predictions, Beat is a secreted protein of the immunoglobulin superfamily that has been shown to function as an anti-adhesive factor on motor axons (Fambrough, 1996; Pipes, 2001). Beat consists of two Ig domains and a Cysteine-rich C-terminal domain that shares similarities with cystine knots (Siebert, 2009).
This study shows that side encodes a candidate directional cue that steers Drosophila motor axons from the place of their birth to their peripheral targets. The spatiotemporal expression pattern of Side thus delineates a linear pathway for motor axons. High levels of Side are consistently found ahead of Beat-expressing growth cones. Contact with motor axons induces the down-regulation of Side and eliminates the pathway, preventing other outgrowing nerves from choosing the same route. Furthermore, biochemical and genetic evidence is provided that Beat functions as a membrane-associated receptor or part of a receptor complex for Side. Taken together, these findings support the concept of pre-existing pathways labeled with molecular markers that are recognized by a specialized subset of axons expressing the appropriate receptors (Siebert, 2009).
This work, provides evidence for a simple guidance mechanism that guides Drosophila motor axons to their target regions. During the period of axonal pathfinding, Side is dynamically expressed on different tissues but its temporal shift forms a spatial pattern as such that high levels are located ahead of motor axonal growth cones. Beat-expressing motor axons recognize and follow Side-labeled surfaces. In order to better visualize the complex spatial and temporal dynamics of axon guidance processes, a FasIIGFPMue397-based imaging assay was established that allowed analysis of the activity of growth cones in living wild-type and mutant embryos. These time-lapse observations revealed that in wild-type embryos, the ISN migrates continuously through the lateral body wall until it reaches a choice point near the dorsal trunk. As expected for migration along a substrate pathway, the ISN employs a relatively small growth cone that extends only few filopodia for steering. The observed growth rates are in good agreement with the growth rates of the RP2 axon in filleted, semiviable embryo preparations (Siebert, 2009).
The absence of the attractant Side strongly interferes with axonal growth and results in delayed arrival of the ISN in its dorsal target regions, frequently failing to innervate the dorsal-most muscles. The lack of Side-mediated attraction likely prevents the progression at a normal rate, causing the ISN to develop a complex growth cone that appears to actively search for guidance information. Since Side is a transmembrane protein, it is predicted to function as a contact attractant. The spatiotemporal expression pattern should therefore provide pathway information. The growth cone of the ISN follows Side-positive cell surfaces from its first emergence in the ventral nerve cord to its dorsal target region. Consequently, the tight association of motor axons and their substrates - e.g., sensory axons - is partially lost in side mutants. Since Side is expressed in all sensory clusters (Sink, 2001), it likely also prefigures the SNa and SNc pathways. Based on the positions of the segmental sensory clusters in the body wall, motor axons following Sidestep-labeled sensory axons are guided into the proximity of their target regions. At stage 15, motor axons reach the end of the sensory tracks and begin to defasciculate into the muscle fields that up-regulate Side at this developmental time point. In this respect, it is interesting to note that motor pathways terminate on the cell bodies of sensory neurons when deprived of their target muscles. The opposed migration of efferent motor axons and afferent sensory axons therefore provides a robust mechanism for the establishment of the basic neuromuscular connectivity pattern in Drosophila. In the brain of vertebrates, a similar mechanism controls the wiring of the thalamus and the cortex (Lopez-Bendito, 2003). Corticothalamic and thalamocortical fibers meet at a common intermediate target and continue to grow along each other in opposite directions (Siebert, 2009).
During its journey through the lateral body wall, the ISN completely ignores the nearby ventral and lateral muscle fields. At this stage, muscle fibers are not yet differentiated, and hence do not express endogenous Side. If these muscle precursors, however, are forced to express Side prematurely under control of Mef2-Gal4, the ISN diverts from its normal path and grows straight into the muscle field. The premature attraction drastically slows down the migration toward dorsal targets, leading to a permanent lack of NMJs on dorsal muscles. Thus, both gain and loss of Side cause migratory delays that result in the lack of NMJs on dorsal muscles but for different reasons. Ectopic expression of Side leads to excess attraction in wrong directions, whereas lack of Side leads to reduced attraction along the predestined path (Siebert, 2009).
Since mutations in both beat and side cause similar axon guidance phenotypes in embryos and innervation defects in larvae, it is assumed that the products of both genes might interact functionally. Several pieces of genetic evidence suggest that beat and side function in a common pathway: (1) the double mutant phenotype is similar to the respective single mutant phenotypes; (2) overexpression of Side in muscles of wild-type embryos leads to premature attraction of the ISN into ventral and lateral muscle fields. This gain-of-function phenotype is completely suppressed by coexpression of Beat. (3) Overexpression of Side in muscles of beat mutants renders motor axons unresponsive to ectopic Side; (4) endogenous Side is not down-regulated in beat mutants. In addition, Beat and Side interact in vitro. S2 cells expressing Beat and Side form large cell clusters in aggregation assays when individually transfected cell populations are mixed. Moreover, Beat coimmunoprecipitates with Side, supporting the idea that Beat interacts with Side or a Side-containing complex. The formation of cell aggregates further argues that Beat-Side interactions lead to the formation of heterophilic adhesion complexes (Siebert, 2009).
The spatiotemporal expression of Side appears to be strictly regulated. The levels of Side expression are highest in front of motor axonal growth cones. Side disappears from cell surfaces once these cells have been contacted by motor axons, indicating that motor nerves neutralize attractive surfaces and thereby disguise the path they are following. Motor nerve bundles that exit the CNS at a later time point are thus prevented from choosing the same route. In beat mutants, this regulatory mechanism appears to be nonfunctional. Side is constitutively expressed in peripheral nerves. Expression of exogenous Beat in post-mitotic neurons but not in muscles rescued the regulatory defects, suggesting that Beat induces the down-regulation of Side cell-autonomously. If Beat would be a secreted protein, one would expect that expression in muscles down-regulates Side on sensory axons. The secreted metalloprotease tolloid-related as well as the secreted TGF-β ligand Dawdle have been shown to rescue axonal guidance defects tissue independently. Biochemical data from transiently transfected S2 cells further support a cell-autonomous function for Beat. Beat was not secreted into the medium, and a fraction of it was associated with membranes. However, since Beat is not normally expressed in S2 cells the lack of a coreceptor or chaperone might prevent its secretion. Candidate coreceptors are the remaining members of the beat multigene family (Pipes, 2001). Several family members are expressed in the ventral nerve cord, and might function together with Beat in the regulation of its subcellular localization and/or function (Pipes, 2001). Further experiments will be necessary to determine the composition of Beat-containing complexes, and how they transduce guidance signals into the growth cone (Siebert, 2009).
If Side is an instructive signal for Beat-expressing motor axons it should be possible to redirect their paths in a side mutant background; i.e., in the absence of endogenous Side but in the presence of exogenous Side on a defined tissue. Regardless of whether Side is expressed in trachea, muscles, or hemocytes in a side mutant background, motor axons head toward the ectopic source of Side. The growth cones find and recognize Side-expressing cell surfaces and adopt their route accordingly. In the most extreme case, motor axons strongly interacted with hemocytes, which are highly motile cells. Side therefore potently controls the path of motor axons. Although Side-mediated attraction is likely not the only mechanism to reach the muscle targets, these results, together with the high penetrance of sidestep mutant phenotypes, suggest that it is one of the major mechanisms. In this respect, it is important to note that Side is not required for motor axon outgrowth per se, but rather for the specification of the growth direction. Based on the experimental evidence, a model is proposed for the navigation of motor axons from the ventral nerve cord to their target area in Drosophila. Beat-expressing motor axon fascicles recognize, extend on, and subsequently mask a pre-existing, Side-labeled substrate pathway that determines their growth direction. In the absence of the labeled pathway, in side mutants, axonal migration is delayed or growth cones head into aberrant directions. In either case they will miss their targets. In beat mutants, the pathway is constitutively labeled but cannot be recognized, leading to similar phenotypes. Since Beat and Side are conserved in insects, similar guidance principles might occur in all organisms, in which the peripheral nervous system develops from sensory organ precursors (Siebert, 2009).
Sidestep is a pivotal molecular player in embryonic motor axon pathfinding. But questions about its functional repertoire remain: (1) can Side permanently overturn targeting preferences? (2) does it promote synaptogenesis, and (3) can Side facilitate synaptic stabilization? To address these questions, Side was temporally and spatially misexpressed and the visible consequences for neuromuscular junction<:/A> morphology were assessed. When Side was misexpressed either broadly or selectively in muscles during targeting in a wildtype background motor axon targeting preferences were permanently overturned. However the misexpression of Side in all muscles post-targeting neither changed synapse morphology, nor compensated for a lack of the synapse-stabilizing protein Fasciclin II (FasII). Rather Side appears to be dependent on FasII, instead of on intrinsic ability, for sustaining targeting changes. It is proposed that Side helps to bring motor axons to their correct muscle targets and promotes synaptogenesis, then FasII serves to stabilize the synaptic contacts (de Jong, 2005).
Sidestep (Side) profoundly influences embryonic Drosophila motor axon pathfinding (Sink, 2001). The extreme pathfinding defects observed in the side loss-of-function mutant confounded analysis of post-axon pathfinding responses to the protein. To overcome this technical hurdle Side was misexpressed in the present study in suitable spatio-temporal paradigms, allowing assessment of motor axon responses during targeting, synaptogenesis, and synaptic stabilization. It was found that Side can permanently influence motor axon targeting and synaptogenesis events, but requires FasII for synaptic stabilization (de Jong, 2005).
In addition to expression in muscles, Side protein is observed at low levels in the motor axons and CNS longitudinal connectives. This expression spans embryonic motor axon targeting and early post-targeting periods, and continues across larval stages (Sink, 2001). Strong Side misexpression in all post-mitotic neurons causes motor axon over-fasciculation in the embryo (Sink, 2001). These over-fasciculation defects endure in the larva, with innervation frequently absent on the muscles examined in the present study. By using a weaker neuronal GAL4 driver, the early over-fasciculation defects were avoided. This driver also subsequently failed to significantly alter the morphology of the larval neuromuscular synapse. Hence the role of ongoing Side expression in axons remains undefined, although the robust neuronal misexpression phenotype (Sink, 2001) indicates it could serve to encourage and maintain axon-axon contact and hence aid fascicle organization (de Jong, 2005).
Post-synaptic Side misexpression in the larva did not visibly alter neuromuscular synapse morphology. This suggests that Side is unable to encourage growth of that larval synapse, unlike the BMP signaling pathways. In contrast Side misexpression commencing in muscles during embryogenesis produces targeting changes that remain to the end of larval development. These changes included the attraction to and retention of innervation at unconventional locations on muscles; the incorrect targeting by motor axons native to the muscle region, and the attraction of non-native innervation to Side-misexpressing muscles (de Jong, 2005).
In both vertebrates and Drosophila neuromuscular synapses form at stereotypic locations on somatic muscle fibers. Furthermore, the stereotypic locations arise in response to muscle-derived cue(s) rather than through imposition upon the muscle by the innervating axon (de Jong, 2005).
This study found that misexpression of Side on muscles can overturn stereotypic synapse formation. In both global and select Side misexpression paradigms innervation is observed at novel locations on the muscle fibers. These novel sites include: close to the segmental boundary; diametrically opposite the normal innervation site, and on the external rather than internal surface of the muscle. This shows that the presence of excess Side signal across muscle fiber surfaces can overwhelm the potentially restrictive cues residing at the muscle clefts (normal innervation region), and axonal explorations and contacts are encouraged beyond these sites. As a consequence innervation develops, and becomes stabilized, at non-stereotypic locations. Hence, the current observations lend additional support for the model that intrinsically pre-patterned cues in the muscle fiber determine the innervation site (de Jong, 2005).
Altered expression of several putative targeting molecules readily perturbs the embryonic motor projection arborization pattern. Yet few studies examined whether the changes observed in the embryo are maintained in the larva. In the case of the leucine-rich repeat family protein Capricious the changes are permanent, whereas in the case of the tetraspanin family protein Late bloomer the changes are transient. Evidently, observations in the embryo do not necessarily translate to larval phenotypes. It has now been found that changes induced by Side in the embryo are sustained in the larva (de Jong, 2005).
The global misexpression of Side in muscles with the C142-GAL4 driver permanently altered targeting choices on the muscles of interest. In some larvae this manifested as an excess of innervation, while in others it appeared as a reduction. This suggests that the relative level of Side on the muscle is a deciding factor as to whether innervation was increased or decreased (de Jong, 2005).
The H94-GAL4 line drives expression in two adjacent muscles in the ventral musculature, muscles 6 and 13. When H94-GAL4 drives Side expression in these muscles, at the temporal juncture between pathfinding and targeting, the motor axon growth cones are robustly attracted to the adjoining edges of the two muscles (Sink, 2001). Examination of the subsequent effects on final innervation at muscles 6/7 and 13 showed that excess Side permanently overturned innervation preferences in favor of muscle 13 - often at the expense of muscle 6, and occasionally of muscle 12. This preference for muscle 13 correlates with H94-GAL4 driving more strongly in muscle 13 than in muscle 6 (de Jong, 2005).
A novel feature of this reorganization in targeting was that the loss of native innervation on the Side-expressing 6 muscle was often not significantly compensated for by either 'reach-back' innervation (e.g., Capricious) or ectopic innervation from the transverse nerve (e.g., Late bloomer). Rather, it appears that once the motor axons normally innervating muscle 6 are drawn away to muscle 13, other motor axons in the region do not provide compensatory innervation to muscle 6. Innervation on muscles 6, 7, 13 and 12 derives predominantly from the RP motoneurons. The RP motoneurons have their somata closely grouped at the CNS midline, extend their axons out of the CNS along a shared pathway, and are tightly fasciculated up until the time of targeting. They also express several proteins in common. The Side-sensitiveness of the RP motor axons in misexpression experiments suggests they also share the molecular machinery required to respond to an elevated, muscle-derived Side signal during targeting. This could explain the common attraction to the highest Side-misexpressing muscle (13) on which excess arborizations are evident, and the failure of (RP) motor axons to be drawn to the other lower, Side-misexpressing muscle (6) and its non-misexpressing partner (7) (de Jong, 2005).
Type II innervation is the most common type of novel innervation observed in response to post-synaptic Side misexpression. Type II innervation derives from the VUM neurons; is octopaminergic, and is thought to be have neuromodulatory functions. In wildtype larvae it is not restricted to stereotypic locations on the muscles, nor does it have structurally elaborate postsynaptic partners like the type Ib and Is glutamatergic motor axons. This could, in part, explain its promiscuous innervating capabilities. Ectopic type II innervation can arise in response to decreases in neuronal activity, following the loss of native innervation at a muscle, and in response to elevated FasII levels (de Jong, 2005).
In embryonic post-synaptic Side misexpression experiments it was observed that type II innervation can occur when there is a reduction in type Ib and type Is innervation at muscles 6 and 7. Yet in the same paradigms it was seen the opposite scenario occur, where type II innervation is drawn along with type Ib and Is motor axons onto a Side misexpressing muscle to give an excess of innervation. This would indicate that the presence of type II innervation arises directly in response to the Side signal rather than as a result of altered native innervation quality or quantity (de Jong, 2005).
FasII regulates the pattern of neuromuscular synapse formation, and is also required for synaptic stabilization. To elucidate the potential interplay between Side and FasII in sculpting the neuromuscular synapse, Side was misexpressed in (a) a muscle subset during targeting and (b) all muscles post-targeting in the fasIIe76 mutant background. The fasIIe76 allele is hypomorphic and expresses only 5%-10% of wildtype FasII protein levels (de Jong, 2005).
A previous study shows that elevated FasII expression in muscles 6 and 13 during targeting increases the presence of ectopic innervation on these muscles. This effect is further enhanced if the FasII misexpression is done in a fasIIe76 background. In contrast, this study found that misexpression of Side in muscles 6 and 13 in a wildtype background leads to increased ectopic innervation on muscle 13 (highest expression), and reduced innervation on muscle 6 (lower expression). When the same Side misexpression paradigm was employed in a fasIIe76 background, the attraction of innervation to muscle 13 was still enhanced while muscle 6 again showed a decrease in innervation levels (de Jong, 2005).
These findings suggest that endogenous FasII levels do not impact the responsiveness of motor axon targeting to increased Side levels. There was, however, an approximately 30% increase in the number of bouton on muscle 13 of fasIIe76; UAS-side24; H94-GAL4 larvae compared to muscle 13 in fasIIe76 larvae. This raised the question of whether increased bouton number on muscle 13 in the Side misexpression paradigm was due solely to changes in axonal targeting, or also due to later synaptic stabilization by Side (de Jong, 2005).
It has been found that fasIIe76 bouton number is reduced by ~35% in larval stages whereas early first instar larva bouton numbers are normal. This demonstrates a role for FasII in synapse stabilization. Accordingly bouton number was restored by returning FasII to both the muscle and axon (de Jong, 2005).
In order to assess if Side also has synapse stabilizing ability, it was misexpressed in the fasIIe76 mutant in all muscles from the end of targeting to the end of larval development with MHC82-GAL4. It was found that in spite of the elevated post-synaptic Side levels, the number of boutons was not increased. The fasIIe76 mutant still had 25% fewer boutons than wildtype, irrespective of Side levels. In addition Side misexpression with the H94-GAL4 driver in a wildtype background results in 58 boutons on muscle 13, whereas misexpressing in the fasIIe76 background produced only 43. Again, this represents 25% fewer boutons in the fasIIe76 mutant compared to wildtype. This points to the remaining FasII present in the fasIIe76 allele being able to stabilize some, but not all of the new synaptic contacts attracted by the presence of excess Side. It also suggests that Side has no (or negligible) ability to stabilize synapses (de Jong, 2005).
In summary, this study shows that increasing Side levels on muscle fibers encourages motor axon targeting changes with concomitant synaptogenesis. The sustained retention of these novel contacts, however, in dependent on FasII levels (de Jong, 2005).
At specific choice points in the periphery, subsets of motor axons defasciculate from other axons in the motor nerves and steer into their muscle target regions. Using a large-scale genetic screen in Drosophila, the sidestep (side) gene was identified as essential for motor axons to leave the motor nerves and enter their muscle targets. side encodes a target-derived transmembrane protein (Side) that is a novel member of the immunoglobulin superfamily (IgSF). Side is expressed on embryonic muscles during the period when motor axons leave their nerves and extend onto these muscles. In side mutant embryos, motor axons fail to extend onto muscles and instead continue to extend along their motor nerves. Ectopic expression of Side results in extensive and prolonged motor axon contact with inappropriate tissues expressing Side (Sink, 2001. Full text of article).
Search PubMed for articles about Drosophila Sidestep
de Jong, S., et al. (2005). Target recognition and synaptogenesis by motor axons: Responses to the sidestep protein. Int. J. Dev. Neurosci. 23: 397-410. PubMed ID: 15927764
Fambrough, D. and Goodman, C.S. (1996) The Drosophila beaten path gene encodes a novel secreted protein that regulates defasciculation at motor axon choice points. Cell 87: 1049-1058. PubMed ID: 8978609
Holmes, A. L. and Heilig, J. S. (1999). Fasciclin II and Beaten path modulate intercellular adhesion in Drosophila larval visual organ development. Development 126: 261-272. PubMed ID: 9847240
Lopez-Bendito, G. and Molnar, Z. (2003). Thalamocortical development: How are we going to get there? Nat. Rev. Neurosci. 4: 276-289. PubMed ID: 12671644
Pipes, G. C., Lin, Q., Riley, S. E. and Goodman, C. S. (2001). The Beat generation: A multigene family encoding IgSF proteins related to the Beat axon guidance molecule in Drosophila. Development 128: 4545-4552. PubMed ID: 11714679
Siebert, M., Banovic, D., Goellner, B. and Aberle, H. (2009). Drosophila motor axons recognize and follow a Sidestep-labeled substrate pathway to reach their target fields. Genes Dev. 23(9): 1052-62. PubMed ID: 19369411
Sink, H., Rehm, E. J., Richstone, L., Bulls, Y. M. and Goodman, C. S. (2001). sidestep encodes a target-derived attractant essential for motor axon guidance in Drosophila. Cell 105: 57-67. PubMed ID: 11301002
date revised: 12 October 2009
Home page: The Interactive Fly © 2009 Thomas Brody, Ph.D.
The Interactive Fly resides on the
Society for Developmental Biology's Web server.