The DSCAMs (Down syndrome cell adhesion molecules) belong to the Immunoglobulin-superfamily (IgSF) of cell adhesion molecules that are involved in synaptic pairing, tiling, and mosaic patterning – all of which are critical in the organization of retinal circuit development (Yamakawa et al., 1998). The majority of focus has been in the Drosophila, where there are four types of DSCAMs that experience incredible diversity due to alternative splicing (Schmucker et al., 2000). Recently, however, a lot of interest and controversy has been generated by the vertebrate DSCAM and DSCAML, which seem to play divergent roles in chick and mice.
Perhaps the most prominent study in chicks is the one carried out by Yamagata and Sanes (2008), where through a series of separate experiments, it was demonstrated that DSCAMs are necessary to promote synaptic specificity, and that this may be consistent with the notion that they mediate homophilic adhesion or attractive cues. The first of these showed that each DSCAM molecule not only mediates homophilic adhesion, but that these adhesions are not heterophilic (Yamagata and Sanes 2008). Cells were transfected with the complementary DNA appropriate to the different molecules, and were then labelled with florescent dyes (Yamagata and Sanes 2008). As shown by figure 1, cells expressing the same IgSF member aggregated with each other, whereas those expressing different adhesion molecules did not, indicating that DSCAMS do mediate homophilic adhesion and not heterophilic adhesion.
Figure 1 | ‘Homophilic adhesions (DSCAM-DSCAM and DSCAML-DSCAML), and not heterophilic adhesions (DSCAM-DSCAML, DSCAM-SDK2 or DSCAML-SDK2). No aggregation observed with untransfected cells.’ (Adapted from Yamagata & Sanes, 2008).
This seemed to concur with studies carried out by Agarwala et al. (2000 and 2001) in mice, which also showed that DSCAMs mediate homophilic adhesion, which is quite a surprising finding as we shall see later. It is also worth mentioning that although Yamagata et al. (2008) claim that untransfected cell aggregations did not occur, mixed aggregates between different adhesion molecules did occur minimally (they quote that this ‘was rare’). Taken together with the results of Agarwala et al. (2000 and 2001), one has to question the integrity of the adhesion assays in that they may not be entirely reliable.
Yamagata and Sanes, however, did not stop there; they also carried out two further experiments (within the same study) to further substantiate their claims. The first of these exploited the fact that within the vertebrate retina, the inner plexiform layer (IPL) is divided into sublaminae where certain retinal ganglionic cells (RCGs) and interneurons arborize and form synaptic connections situated in particular sublaminae (Wässle 2004). Each DSCAM also seems to be concentrated in distinct sublaminae, such that when the IPL is divided into the conventional ‘5-parallel-slabs’ (S1-S5), DSCAM is predominantly observed in S5 and DSCAML in S1 (Yamagata and Sanes 2008). Based on the studies of Wohrn et al. (1998) and their own previous work (Yamagata et al., 2006), independent markers co-expressed by RGCs that also express DSCAM were identified and so could be used to assess lamina specific connectivity upon depletion of DSCAMs. This marker for DSCAM-positive RGCs was R-Cadherin, which was used (along with other markers) to immunostain the IPL that was infected with retroviral vectors carrying control and DSCAM-interfering RNA, figure 2, (Yamagata and Sanes 2008).
Figure 2 | ‘IPL in regions where retroviral vectors carried control and Dscam-interfering RNA; then immunostained with markers selectively coexpressed by RGCs that express Dscam (Rcadherin), Sidekick-1 (cadherin-7) and Sidekick-2 (calbindin). Depletion of Dscam, leads to selective mislocalization of R-cadpositive processes in S5, only.’ (Adapted from Yamagata & Sanes, 2008).
As shown by figure 2, DSCAM depletion selectively perturbed ‘R-cad-positive’ processes in the S5 sublaminar layer. The fact that this perturbation was not observed with other markers (calbindin and Cad-7) suggests that the lamina-specific influences of DSCAMs are necessary to properly confer neurite arborization, and without them these arborizations would occur haphazardly in the IPL; indicative of an instructive role played by DSCAM, reflecting adhesion-mediated properties.
Figure 3 | ‘The GFP-labelled processes were distributed in all sublaminae when GFP alone was overexpressed (control). Ectopic expression of Dscam shifted processes to Dscam-positive S5 (brackets in top panels mark regions magnified in lower panels).’ (Adapted from Yamagata & Sanes, 2008).
Ectopic expression of DSCAMs using an adapted insect-derived system (where transposons were generated and combined with GFP, green-fluorescent-protein, and then introduced into the retina) was carried out to show whether DSCAMs are sufficient for this lamina-specifying function (Yamagata and Sanes 2008).
Figure 3 shows that when only GFP was expressed as a control, IPL processes seemed to be distributed in all of the sublaminae, whereas, ectopic DSCAM expression seemed to influence the processes such that they diverted to the ‘Dscam-positive’ S5 region. This shows that DSCAMs are sufficient to influence the laminar choices of retinal neurons and that without their adhesive cues, the various subsets of interneurons and RGCs would not discriminate between synaptic partner choices, and thus, not arborize and form the appropriate connections required to wire up the retina in the chick.
Although the integrity of the adhesion assays were questionable, it is really difficult to dispute the depletion experiments, especially since full knockouts were carried out and not dominant-negatives.
Collectively, these findings do suggest that DSCAMs, in the chick at least, do mediate homophilic cell adhesions. The depletion and ectopic expression of DSCAMs show that they are necessary and sufficient, respectively, for lamina-specific arborizations – consistent with the notion of them interacting as adhesion/attractive cues that can direct synaptic specificity.
While Yamagata and colleagues were carrying out their experiments in chick, a different group were occupied with investigations regarding DSCAM function in mice. Initially, through the processes of positional cloning Fuerst et al. (2008) identified a spontaneous mutation in mice which, through complementary DNA sequencing, could allow them to carry out loss-of-function experiments for DSCAM.
They subsequently carried out in situ hybridisations using markers for the various amacrine cell subtypes, which demonstrated that DSCAM expression occurs in two subsets of amacrine cells in the retina: ‘dopaminergic amacrine cells (tyrosine hydroxylase (TH)-positive) and bNOS-positive (Nosl-expressing) cells’; and not in choline acetyltransferase (Chat)-positive starburst amacrine cells (SACs) and disabled 1-positive (Dab1) AII amacrine cells, which did not display the defects observed in bNOS-positive and TH-positive cells (Fuerst et al., 2008).
When wild-type dopaminergic amacrine cells were first labelled by anti-TH staining (figure 4, left panel), their spacing was observed to be normal with formation of a broad and uniform plexus. The Dscam -/- dopaminergic amacrine cells, however, exhibited an abnormal morphology, where there seemed to be arborization defects; fasciculation of neurites, and defective spacing of cell bodies, which were clumped (figure 4, right panel). The same finding was observed in bNOS amacrine cells (figure 5).
Figure 4 | ‘On the left; a wild-type P6 retina stained with anti-TH; dopaminergic amacrine cell processes have little proximal contact and rarely self-cross. On the right; a P6 Dscam -/- retina stained with anti-TH; individual dopaminergic amacrine cells frequently self-cross.’ (Adapted from Fuerst et al., 2008).
Figure 5 | ‘The cell bodies of bNOS-positive cells, which also express DSCAM in control retinas, are spaced and arborized in a manner similar to dopaminergic amacrine neurons, (left panel). On the right panel, the bNOS-positive amacrine cells show fasciculated processes and randomly spaced cell bodies in Dscam -/- retinas.’ (Adapted from Fuerst et al., 2008).
Moreover, in a different study, Fuerst et al. (2009) also found that the loss of the other vertebrate DSCAM, DSCAML, resulted in the same disruptive defects as those observed in their 2008 study (Fuerst et al., 2009). The combined findings of both studies suggests that the two types of DSCAMS are expressed in non-overlapping retinal neuron sets, where they are involved in antagonizing adhesions between processes of homotypic cells, as well as interactions between ‘sister neurites’, thus regulating mosaic spacing (Fuerst et al., 2009). This is consistent with the notion that, in mice at least, DSCAMs mediate homophilic repulsive interactions – and not homophilic adhesion.
It is fair to mention that the evidence produced from the mice studies is as convincing, if not more, than the chick study, in that as in the chick, full knockouts were implicated – making the results reasonably objective and difficult to dispute.
The findings in the mice concur with findings in the Drosophila fly (an invertebrate), where the DSCAM genes also mediate self-avoidance, and hence, homophilic repulsion. Such repulsion is demonstrated very well by Matthews et al. (2007), who showed that the broadly expressed DSCAM1 (which exhibits unique identities depending on the neuron it is expressed from) is necessary to allow selective recognition and repulsion of sister neurites during arborization. Through the same mechanism of homophilic repulsion, DSCAM2 prevents the formation of connections between adjacent cells to promote tiling (Millard et al., 2007). So it would be reasonable to suggest that the mouse DSCAM seems to consist of a blend of the Drosophila DSCAM1/DSCAM2 functions.
The roles of DSCAMs in dendritic self-avoidance and tiling mediated by homophilic repulsion are shared by both vertebrates (mice) and invertebrates (Drosophila), although, as discussed earlier, homophilic attractions are also mediated by DSCAMS (in chicks). Studies in the Drosophila and mouse emphasise that homophilic repulsion is crucial for the regulation of neural circuit assembly, while in chick they accentuate that DSCAM proteins act ‘context-dependently’ to mediate different recognition events during retinal wiring.
Having said that, the divergent roles of DSCAMs in vertebrates seem quite puzzling, and this is in no way helped by the fact that the mechanisms underlying the discrepancy remain elusive. A potential explanation may be that DSCAM function in different species is mediated by different signalling as observed with netrin (an axon guidance molecule) which mediates attraction or repulsion, based on the receptors it binds to (Hong et al., 1999). This would be an interesting concept to test; if DSCAM from mice was incorporated into the chick and vice versa, for example. This may further allude to whether different signalling cascades are responsible for the divergent roles of DSCAMs. What these signalling mechanisms are and how they mediate homophilic repulsion or attraction is yet to be elucidated. While such impending issues regarding underlying mechanisms are yet to be cleared, speculation that DSCAMs mediate only one type of interaction seem futile.
More complex and in-depth studies investigating DSCAM function in cellular, molecular and functional levels will undoubtedly be more beneficial; providing a comprehensive insight into the mechanisms underlying neural development in the retina. However, for now only, it would be appropriate to propose that DSCAMs undertake roles of attraction and repulsion in retinal development; and that this is dependent on the species in question, the type of cells involved and also context.
- Agarwala, K. L., Ganesh, S., Tsutsumi, Y., Suzuki, T., Amano, K., Yamakawa, K., (2001), ‘Cloning and functional characterization of DSCAML1, a novel DSCAM-like cell adhesion molecule that mediates homophilic intercellular adhesion’, Biochem. Biophys. Res. Commun. 285: 760–72
- Agarwala, K. L., Nakamura, S., Tsutsumi, Y., Yamakawa, K., (2000), ‘Down syndrome cell adhesion molecule DSCAM mediates homophilic intercellular adhesion’, Brain Res. Mol. Brain Res. 79: 118–126
- Fuerst, P. G., Bruce, F., Tian, M., Wei, W., Elstrott, J., Feller, M. B., Erskine, L., Singer, J. H., Burgess, R. W., (2009), ‘DSCAM and DSCAML1 function in self-avoidance in multiple cell types in the developing mouse retina’, Neuron 64: 484–497
- Fuerst, P. G., Koizumi, A., Masland, R. H., Burgess, R. W., (2008), ‘Neurite arborization and mosaic spacing in the mouse retina require DSCAM’, Nature 451: 470–474
- Hong, K., Hinck, L., Nishiyama, M., Poo, M. M., Tessier-Lavigne, M., Stein, E., (1999), ‘A ligand-gated association between cytoplasmic domains of UNC5 and DCC family receptors converts netrin-induced growth cone attraction to repulsion’, Cell 97: 927–941
- Matthews, B. J., Kim, M. E., Flanagan, J. J., Hattori, D., Clemens, J. C., Zipursky, S. L., Grueber, W. B., (2007), ‘Dendrite self-avoidance is controlled by Dscam’, Cell 129: 593-604
- Millard, S. S., Flanagan, J. A, Pappu, K. S., Wu, W., Zipursky, S. L., (2007), ‘Dscam2 mediates axonal tiling in the Drosophila visual system’, Nature 447: 720–24
- Schmucker, D., Clemens, J. C., Shu, H., Worby, C. A., Xiao, J., Muda, M., Dixon, J. E., Zipursky, S. L., (2000), ‘Drosophila Dscam is an axon guidance receptor exhibiting extraordinary molecular diversity’, Cell 101: 671-684
- Wässle, H., (2004), ‘Parallel processing in the mammalian retina’, Nat. Rev. Neurosci. 5: 747–757
- Wohrn, J. C., Puelles, L., Nakagawa, S., Takeichi, M., Redies, C., (1998), ‘Cadherin expression in the retina and retinofugal pathways of the chicken embryo. J. Comp. Neurol. 396: 20–38
- Yamagata, M., Sanes, J. R., (2008), ‘Dscam and Sidekick proteins direct lamina-specific synaptic connections in vertebrate retina’, Nature 451: 465–469
- Yamagata, M., weiner, J. A., Dulac, C., Roth, K. A., Sanes, J. R., (2006), ‘Labeled lines in the retinotectal system: markers for retinorecipient sublaminae and the retinal ganglion cell subsets that innervate them’, Mol. Cell. Neurosci. 33: 296–310
- Yamakawa, K., Huot, Y. K., Haendelt, M. A., Hubert, R., Chen, X. N., Lyons, G. E., Korenberg, J. R., (1998), ‘DSCAM: A novel member of the immunoglobulin superfamily maps in a Down syndrome region and is involved in the development of the nervous system’, Hum. Mol. Genet. 7: 227–37