Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2008 May 3.
Published in final edited form as: Neuron. 2007 May 3;54(3):417–427. doi: 10.1016/j.neuron.2007.04.013

Homophilic Dscam interactions control complex dendrite morphogenesis

Michael E Hughes 1,2, Rachel Bortnick 1,2, Asako Tsubouchi 3, Philipp Bäumer 1, Masahiro Kondo 1, Tadashi Uemura 3, Dietmar Schmucker
PMCID: PMC1963440  NIHMSID: NIHMS22814  PMID: 17481395

Summary

The morphogenesis of complex dendritic fields requires highly specific patterning and dendrite-dendrite recognition mechanisms. Alternative splicing of the Drosophila cell surface receptor Dscam results in up to 38,016 different receptor isoforms and in vitro binding studies suggested that sequence variability in immunoglobulin-like ecto-domains determines the specificity of strictly homophilic interactions. We report that diverse Dscam receptors play an important role in controlling cell-intrinsic aspects of dendrite guidance. We examined the function of Dscam during morphogenesis of dendrite arborization neurons (“da” neurons) and found that loss of Dscam in single neurons causes abnormal dendritic fasciculation and a strong increase in self-crossing of dendritic branches of da neurons. Restriction of dendritic fields of neighboring class III neurons appeared intact in Dscam deficient neurons suggesting that dendritic self-avoidance but not hetero-neuronal tiling may depend on Dscam function. Over-expression of the same Dscam isoforms in two da neurons with normally overlapping dendritic fields forced a spatial segregation of the two dendritic fields. Taken together, our results suggest that dendritic branches of all four classes of da neurons use isoform-specific homophilic interactions of Dscam to ensure minimal overlap of dendrites. The large pool of Dscam’s extracellular recognition domains may allow the same ‘core’ repulsion mechanism to be used in every da neuron without interfering with hetero-neuronal interactions.

Introduction

The morphologies of dendritic arborizations are often complicated and exhibit great cell-type diversity. Importantly, the distinct dendritic branching characteristics, polarity and sensory field sizes directly contribute to circuit specific processing properties and synaptic integration. However, relatively little is known of how the development of distinct dendrite morphologies is genetically or epi-genetically controlled (Jan and Jan, 2003). Studies on multidendritic neurons in the Drosophila peripheral nervous system have established an experimental system which allows for a systematic dissection of genetic and cellular mechanisms underlying dendrite morphogenesis (e.g. Grueber and Jan, 2004; Grueber et al., 2005; Sugimura et al., 2003). The Drosophila embryonic PNS organizes a stereotyped pattern of identified sensory neurons, and the so-called dendritic arborization neurons (“da” neurons) constitute a subfamily of multidendritic neurons. The da neurons grow in a single layer directly underneath the epidermis during late embryonic and larval stages. The 15 described da neurons in each abdominal hemisegment are classified into four categories termed classes I-IV. This classification is based on distinct morphologies of the dendritic trees of the respective sensory neurons (Grueber et al., 2002). Class I da neurons have the smallest number of dendritic branches and their sensory field covers a comparatively small part of each hemisegment. Classes II- IV exhibit increasing degrees of dendritic complexity and larger dendritic fields. Class IV da neurons show the most expansive dendritic arborizations and its large dendritic fields fully extend from the anterior to the posterior segment boundary.

Several studies have established that class III and class IV neurons provide complete non-overlapping “tiling” of the body wall (Grueber et al., 2002; Grueber et al., 2003b; Sugimura et al., 2003). The smaller class I and class II da neurons only partially tile the hemisegment territories but nevertheless occupy non-overlapping territories. A recent study using time-lapse analysis and examining the development of da neurons revealed a series of distinct repulsive interactions as a crucial cellular mechanism for shaping the dendritic fields of each neuron (Grueber et al., 2003b; Sugimura et al., 2003). Based on this and previous studies, it is clear that development of class III & IV neurons depends on contact-mediated repulsion between dendrites of neighboring neurons within the same class (i.e. hetero-neuronal tiling) and persistent repulsion of sister branches of the same neuron (self-avoidance; iso- or intra-neuronal tiling). In contrast, class I and class II da neurons depend on repulsion of dendritic sister branches but heteroneuronal tiling is unlikely to control the restriction of dendritic growth. Although several genes (e.g. Parrish et al., 2006; Emoto et al., 2004) that control development of da neurons have now been identified it is still unknown what proteins control the repulsive dendrite-dendrite interactions and how many different molecular pathways control repulsion (Grueber et al., 2003a; Senti et al., 2003; Sugimura et al., 2004; Sweeney et al., 2006). In addition, given that the characteristic sizes and morphologies of class I & II da neurons do not depend on hetero-neuronal tiling, it seems clear that other mechanisms independent of dendrite-dendrite avoidance must exist.

Several in vitro and in vivo studies suggest that Dscam-Dscam interactions are capable of controlling neurite repulsion (Wang et al., 2002; Zhan et al., 2004). It has been shown that Dscam loss-of-function can disrupt the ability of sister branches from the same axon of mushroom body neurons to segregate along different pathways (Wang et al., 2002). Additionally, over-expression of a single isoform of Dscam in a group of MB neurons significantly disrupted the organization of the axon bundle; in contrast, over-expression of the same isoform in single neurons had no discernible phenotype (Zhan et al., 2004). Homophilic Dscam interactions are also consistent with Dscam’s role in dendrite morphogenesis in the olfactory system. A recent study demonstrated that the loss of Dscam results in clumped dendrites and a reduction in the size of dendritic fields in both projection neurons and local interneurons (Zhu et al., 2006). Although the complex morphology of the antennal lobe neuropile does not allow for an analysis with single dendrite resolution, these studies are nevertheless consistent with a role of Dscam in mediating repulsive signals between single dendrites of projection neurons. The observed complex phenotypes in the olfactory system could be explained by a role of Dscam in either dendrite branching directly, by more complex (potentially heterophilic) interactions between projection and antennal lobe neurites, or – as proposed by Zhu et al. –in intraneuronal tiling of dendrites (Zhu et al., 2006).

Most studies, however, have focused on a role of Dscam in axonal development. Studies of the somatosensory system provided evidence that a very large number of diverse Dscam receptor isoforms are required to ensure precision of neuronal connectivity (Chen et al., 2006) and that the complex afferent projections of mechanosensory neurons are highly sensitive to genetic manipulations of Dscam isoform diversity (Chen et al., 2006). In this context it has been proposed that local isoform-specific interactions instruct axonal branches to connect with their proper targets. PCR-based expression studies have suggested that the Dscam repertoire of each cell is different from those of its neighbors and may be utilized to generate unique cell identities in the nervous system (Neves et al., 2004). Furthermore, in vitro binding studies have shown that Dscam isoforms can interact in a highly selective homophilic manner where even closely related isoforms show little interaction and exhibit almost exclusive isoform-specific binding (Wojtowicz et al., 2004). Although homophilic Dscam interactions first result in the formation of adhesion complexes, it has been proposed that subsequent signaling events may override the adhesive interactions and trigger repulsion of the interacting cell compartments (e.g. dendrites or branched growth cone) (Wang et al., 2002; Wojtowicz et al., 2004). However, the experiments examining homophilic repulsion have been primarily based on over- or mis-expression studies and no loss-of-function phenotypes supporting the notion of homophilic repulsion have been reported. Furthermore, the limited knowledge of Dscam isoform distribution in different (single) neurons has so far not allowed determining where interactions of the same Dscam isoforms (i.e. homophilic interactions) occur within the nervous system.

In this study we examined the role of Dscam in repulsive interactions that shape the dendritic fields of da neurons. Dendrites of da neurons provide an experimental system whereby single dendrite branch resolution is easy to achieve and Dscam-Dscam interactions include homophilic interactions. In addition, da dendrite patterning does not depend on complex interactions with presynaptic targets and dendrite intrinsic interactions can be easily distinguished from interneuronal interactions. We found that Dscam is cell-autonomously required in all four classes of da neurons and that in the absence of Dscam the dendrites of da neurons have strong patterning defects indicating a lack of dendrite-dendrite repulsion. Loss-of-function as well as gain-of-function phenotypes are consistent with the possibility that Dscam specifically controls a da cell-intrinsic ability of dendrites to avoid sister branches. Our studies also suggest that additional Dscam-independent repulsion mechanisms controlling heteroneuronal tiling and spatial restrictions of dendritic field sizes must exist.

Results

Dscam loss-of-function causes excessive self-crossing of dendrites from the same neuron

Dscam is expressed broadly throughout the nervous system including many sensory neurons of the peripheral nervous system (PNS; Figure S1). To examine the role of Dscam in dendrite morphogenesis, we used MARCM analysis (Lee and Luo, 2001) to generate GFP-labeled, Dscam-null clones in the larval PNS. We first examined the branching phenotype of ddaD and ddaE neurons, two class I neurons found in the dorsal cluster of each abdominal hemi-segment (Figure 1& 2). Compared to higher order da neurons, ddaD and ddaE have relatively simple dendritic arborizations. Both neurons send primary branches dorsally and ventrally, with a number of secondary and tertiary branches that project towards the segment boundary; anterior in the case of ddaD, posterior in the case of ddaE (Figure 1 A). Importantly, both neurons have a characteristic receptive field in which individual dendrites fill the available space evenly, with a minimal amount of overlap, which has also been described as a “like-repels-like response” (Emoto et al., 2004) and most likely depends on repulsion between sister dendrites (Figure 1 B,E).

Figure 1. Dscam loss-of function-results in excessive iso-neuronal self-crossing in Class I neurons.

Figure 1

Open in a new tab

(A) Schematic representation of the receptive fields of two class I neurons, ddaD and ddaE. (B-G) MARCM was used to generate wildtype (B, E) and Dscam-null (C-D, F-G) single-cell clones of class I neurons (scale bar, 50μm). (H-K) Ig1-1 Gal4 was used to visualize ddaE neurons in 2nd instar larvae. (H) Dscam-null larvae (Dscam21/Dscam33) have disorganized dendrites compared to wildtype (I). This phenotype is rescued by the expression of either Dscam1.30.30.1 (J) or Dscam11.31.25.1 (K) in an otherwise Dscam-null background (scale bar, 10μm). (L) A single Dscam-null ddaE neuron was selected for quantitative analysis (scale bar, 20μm). (I, J) A 5X zoom of the region of interest marked in panel (M) was used to quantify the number of iso-neuronal self-crossings in both a confocal stack (N) and a single z-section (J) (scale bar, 4μm). Arrows indicate points of self-crossing between sister-branches.

Figure 2. Quantification of Dscam’s null phenotype in ddaE neurons.

Figure 2

Open in a new tab

(A) The average number of iso-neuronal self-crossings for wildtype, Dscam21 and Dscam33 is graphed (P <0.001). (B) The probability distribution of each genotype is plotted as a histogram. (Wildtype N= 21, Dscam21 N = 14, Dscam33 N = 9). Neither the number of dendritic termini per cell (C) nor the total dendritic area (D) was significantly changed by the loss of Dscam, indicating further that Dscam does not significantly alter the cell-fate determination of ‘da’ neurons.

In contrast, the dendritic arborizations of Dscam-null class I neurons show extensive self-crossing between sister branches of the same cell suggesting that they lost the ability to avoid sister (same-cell) branches, resulting in tangled dendritic branches and uneven innervation of the overall dendritic field (Figure 1 C, D, F, G). The overall extent and orientation of the receptive field is largely unchanged in Dscam mutant animals. As in wildtype, Dscam-null ddaD neurons extended dendrites only in the anterior direction (Figure 1 C, D) and ddaE extended dendrites only in the posterior direction (Figure 1 F, G).

The defects in self-crossing were rescued by expressing single Dscam isoforms specifically in class I neurons confirming a cell-autonomous function of Dscam in da neurons (Figure 1 H-K). We used the Gal4221 driver which allows early onset expression selectively in a subset of class I da neurons (Grueber et al., 2003a). Dscam cDNAs of two different isoforms were expressed in Dscam21/Dscam33 homozygous animals via UAS-based transgenes (described previously in Chen et al., 2006; Zhan et al., 2004). Individual isoforms are denoted by the combination of alternative variable Ig domains and transmembrane segments. For instance, the isoform comprising Ig2 alternative 1, Ig3 alternative 30, Ig7 alternative 30, transmembrane segment 17.1, is designated Dscam1.30.30.1 (Schmucker et al., 2000). Due to early lethality of homozygous Dscam animals, morphology of the class I neurons were examined at first and second instar of larval development. We found that both the isoform Dscam1.30.30.1 as well as Dscam11.31.25.1 are capable of rescuing the self-crossing phenotype of class I neurons (Figure 1 J, K).

The crossing of dendrites in homozygous Dscam animals (Figure 1 H) and single mosaic clones of da neurons (Figure 1 C, D, F, G) was observed in single confocal z-sections thereby demonstrating the close proximity of intersecting dendrites (Figure 1 L-N). Therefore, the striking increase of dendritic crossings in mutant neurons is unlikely the result of dendrites now growing in different tissue layers, but rather an abnormal direct contact and intersections of dendrites.

We quantified the defects in self-avoidance in Dscam-null neurons by tracing dendrites of ddaE neurons and counting the number of times each dendrite crosses its sister branches (Figure 2). We used two independent strong loss-of-function Dscam EMS alleles, which have previously been described and found to lack detectable Dscam protein expression (Hummel et al., 2003). Dscam33 and Dscam21 showed a 9-12-fold increase in the number of self-crossing events per cell when compared with wildtype clones (Figure 2 A). When we compared the frequency distributions of self-crossing events per cell, we found that the majority of wildtype class-I neurons exhibited less than 3 self-crossing events per cell and all of the homozygous Dscam21 or Dscam33 neurons displayed at least 4 and as many as 21 self-crossing events per cell (Figure 2 B). In contrast, we found no significant difference in the number of dendritic branches or dendritic termini (Figure 2 C). Additionally, although the dendritic field often appears distorted due to the uneven distribution of dendrites, we found on average no significant change in the overall area of the receptive field (Figure 2 D). Dscam-null class-I neurons were still capable of extending dendritic branches in anterior-posterior or dorsal-ventral direction. Most dendritic branches of mutant neurons were found to reach as far as dendritic branches from wildtype neurons (i.e. reaching the segment boundary). Furthermore, we examined the total dendritic field area by fitting a polygon around the most distant end points of ddaE dendrites and then determining the surface area. Measuring the dendritic field area in 21 wildtype and 23 mutant neurons we found no significant change in the dendritic area of Dscam mutant neurons (Figure 2D).

Dscam controls dendritic self-avoidance in all four classes of da neurons

The abnormal self-crossing phenotype in Dscam mutant neurons was also observed in other types of multi-dendritic neurons (Figures 3, 4, S2). Class II ddaB neurons (Figure 3) normally have a small number of long dendritic branches that project separately in the dorsal direction (Figure 3 B). In contrast, in Dscam-null ddaB neurons abnormal fasciculation and self-crossing was abundantly observed (Figure 3 C, D). The dendritic area decreased minimally (Figure 3 E) and the number of termini was found to increase mildly from ten in wildtype to twelve in mutant ddaB neurons (Figure 3F). However, the most striking defect was the 7-8-fold increase in self-crossing of dendritic branches (Figure 3 G).

Figure 3. Dscam loss-of-function results in excessive iso-neuronal self-crossing in Class II neurons.

Figure 3

Open in a new tab

(A) Schematic representation of the receptive field of ddaB, the only Class II neuron within the dorsal cluster. (B-D) MARCM was used to generate wildtype (B) and Dscam-null (C,D) single cell clones of ddaB neurons (scale bar, 20μm). Arrows mark the cell bodies of ddaB neurons. The area of ddaB receptive fields was unchanged by Dscam loss-of-function (E) although there was a modest increase in the total number of dendritic termini (P < 0.05) (F). Similar to Class I neurons, the frequency of iso-neuronal self-crossing was dramatically increased in Dscam-null neurons (P < 0.001) (G). (Wildtype N = 8, Dscam-null N = 7).

Figure 4. Dscam loss-of-function results in excessive iso-neuronal self-crossing in Class III neurons.

Figure 4

Open in a new tab

(A-F) MARCM was used to generate wildtype (A, B) and Dscam-null (C-F) single-cell clones of class III neurons. Arrows mark the cell bodies of ddaA and ddaF neurons. Neither the area of the receptive field of ddaF neurons (G) nor the dorsal length of ddaF neurons (H) was noticeably altered by Dscam loss-of-function. (Wildtype N = 7, Dscam-null N = 7).

Similarly, class III neurons (ddFA, ddaA, Figure 4) and class IV neurons (ddaC, Figure S2) show considerable disorganization of their dendritic arborizations when Dscam expression is lost. The two class III neurons within the dorsal cluster, ddaF and ddaA, have considerably larger and more complex dendritic fields than class I neurons. However, similar to the class I neurons described above, class III neurons have a characteristic receptive field in which individual dendrites fill the available space with minimal overlap between sister branches of the same cell (Figure 4 A, B). When Dscam expression is lost, individual branches overlap extensively with their neighbors, in some cases projecting together as a bundle for considerable distances (Figure 4 C-F). In wildtype cells, the dendrite branches strictly avoid each other’s territory, projecting in parallel towards the posterior segment boundary. In contrast, in Dscam-null cells the dendrites aberrantly cross or even fasciculate with sister branches and consequently form a disorganized tangle of branches (e.g. Figure 4 C, D).

The length of individual dendrite branches as well as the orientation of the receptive field is mostly unchanged in class III neurons. Only occasionally did we observe more extreme examples in which several dendrite branches are entangled and the dendritic field is clearly altered (Figure 4 C). Importantly, we have used time-lapse analysis of individual class III neurons and found that dendritic growth and developmental adaptation to the growth of the body wall is normal in Dscam-null neurons (Figure S3).

Animals lacking alternative exon 4 sequences show normal dendrite patterning

We also examined da neurons in Dscam mutant animals that still express normal levels of Dscam protein, but lack a subset of alternative exon 4 sequences, and therefore have a significant reduction of isoform diversity. We tested the mutant strain DscamC22-1, lacking exons 4.4-4.12 and therefore some 75% of isoform diversity. DscamC22-1 animals express a maximum number of 9,504 Dscam isoforms (Wang et al., 2004). However, we found that the dendritic self-avoidance of class I ddaE neurons was not impaired in DscamC22-1 mutant flies (Figure 5). Therefore, in contrast to the function of Dscam for axonal targeting of mechanosensory neurons (Chen et al., 2006), even a significant reduction in isoform diversity does not impair normal dendrite patterning. In addition, expression of a randomly chosen Dscam isoform in Dscam-null da neurons is sufficient to restore dendritic self-avoidance (Figure 1 J, K), which suggests that the cell-intrinsic function of Dscam during dendrite development does not require multiple isoforms.

Figure 5. Loss of exon 4 diversity does not significantly disrupt the morphology of Class I neurons.

Figure 5

Open in a new tab

(A-B) C161-Gal4 was used to drive expression of CD8-GFP in ddaE neurons in wildtype (A) and DscamC22-1 (B) larvae. ddaE neurons were traced in red in photoshop to aid visualization. Loss of exon 4 diversity in DscamC22-1 larvae had no noticeable effect on the frequency of iso-neuronal self-crossing (C) the number of dendritic termini (D) or the overall area of ddaE receptive fields (E). (Wildtype N = 10, DscamC22-1 N = 10).

Hetero-neuronal tiling of class III da neurons appeared normal

It has been described that dendrite repulsion is also important for restricting the dendritic field size of neighboring class III and class IV da neurons. The dendritic fields of these complex neurons are relatively large and overlap extensively with da neurons that belong to other classes, but the dendrites avoid neurons of the same da class. Several studies have established that terminal branches of class III neurons avoid other neighboring class III neurons and that dendrites of class IV neurons do not overlap with dendrites of other class IV neurons, thus enabling a non-redundant dendrite coverage of receptive fields, a developmental mechanism referred to as hetero-neuronal tiling (Jan and Jan 2003, Grueber et al., 2003b; Sugimura et al., 2003). Using MARCM analysis we have generated a small number of clones (n=5) that contained neighboring class III neurons, which were both homozygous for a Dscam-null mutation (Figure 6). We have traced the terminal branches of these neighboring Dscam deficient class III neurons, but found no significant overlap of dendritic fields (Figure 6 E-F). The mutant class III neurons showed clear defects in self-crossing (Figure 6 B), however, we found only a single example of a terminal dendrite of a lateral class III neuron crossing a terminal branch of a neighboring dorsal class III neuron (Figure 6D).

Figure 6. Dscam loss-of-function does not disrupt hetero-neuronal tiling between neighboring Class III neurons.

Figure 6

Open in a new tab

(A, B) MARCM was used to generate labeled clones of Class III neurons. In a minority of larvae, two neighboring Class III neurons are labeled, permitting the analysis of hetero-neuronal tiling. In (A) two wildtype Class III neurons strictly avoid each other’s receptive field. Similarly, in Dscam-null neurons (B), the individual dendrites generally respect the boundary of the neighboring cell’s receptive field, although iso-neuronal avoidance is severely disrupted (scale bar, 50μm). Arrows mark the cell bodies of Class III neurons. (C,-F) Traces of the dendritic arborizations of wildtype (C) and Dscam-null (D-F) neurons; in these examples, the neuron projecting from the dorsal side is labeled in red, the ventral neurons are labeled in blue.

We have not been able to obtain clones of neighboring class IV neurons lacking Dscam. Nevertheless, similar to the phenotype of class III neurons (Figure 4 D-F), we found that single Dscam deficient class IV neurons show extensive self-crossing of terminal branches, but exhibit no obvious change in the shape of the dendritic field (Figure S2).

Due to the technical limitations of clonal analysis, we cannot exclude the possibility of a low penetrant phenotype or that Dscam may be involved in avoidance of terminal branches of neighboring class IV neurons. The lack of defects observed in class III neurons, however, suggests that Dscam is unlikely to play a general or essential role in hetero-neuronal tiling.

Mis-expression of single-isoforms of Dscam in class I and class III neurons causes abnormal repulsion between neurons with overlapping dendritic fields

It has been previously proposed that Dscam-Dscam interactions may lead to repulsion (Wang et al., 2002; Zhu et al., 2006). In order to directly test the possibility that homophilic Dscam-Dscam interactions mediate repulsive interactions among dendrites of da neurons, we expressed a single isoform of Dscam simultaneously in class I (ddaE) and class III (ddaF) neurons, which normally have overlapping dendritic fields. For expression of UAS-based Dscam transgenes we used the C161-Gal4 driver which is active in a well-characterized subset of da neurons in the peripheral nervous system (Shepherd and Smith, 1996; Williams and Truman, 2005; Figure S4). Since the dendrites of class III neurons are lined with characteristic protrusions called ‘dendritic spikes’, the morphology of overlapping class I and class III dendrites can be unambiguously distinguished (Figure 7). In wildtype larvae, the dendritic field of ddaE (class I) neurons clearly overlaps with the class III neurons, ddaF and ddaA (Figure 7 A, C, D). In contrast, the mis-expression of a single-isoform of Dscam strongly decreases the number of intersections between the dendrites of class I and class III cells (Figure 7 E-G). In fact, as a consequence of the co-expression of Dscam1.30.30.1, the respective dendritic fields of several ddaE and ddaF neurons were found to be completely non-overlapping. This suggests that homophilic Dscam-Dscam interactions are indeed capable of eliciting repulsion between interacting dendrites (Figure 7 B, E-F). In contrast, overexpression of a Dscam isoform (Dscam 1.30.30.2) containing transmembrane segment 17.2 which is thought to direct localizion to axons or cell bodies (Wang et al., 2004) had no effect on dendrite repulsion (Figure 7G).

Figure 7. Mis-expression of a single isoform of Dscam in neighboring Class I and Class III neurons causes inappropriate repulsion between dendrites.

Figure 7

Open in a new tab

(A-F) C161-Gal4 was used to drive expression of CD8-GFP in Class I (ddaE) and Class III neurons (ddaF, ddaA). (A, C, D) Class I and Class III neurons overlap extensively in wildtype animals (marked by white arrowheads). (B, E, F) In contrast, mis-expression of a single isoform of Dscam (1.30.30.1 or 11.31.25.1) prohibits Class I and Class III neurons from sharing the same region of the hemi-segment (scale bar A, B - 20μm, C, D - 10μm). These data are quantified in panel (G) (P<0.01) (Wildtype N=7, Dscam mis-expression 1.30.30.1, N=5, Dscam mis-expression 11.31.25.1, N = 12). In A-B, ddaE neurons have been pseudocolored and traced in red.

Importantly, we have found that the timing of expression and moderate expression level of the C161-Gal4 driver allowed expression of Dscam isoforms without changing the morphology of ddaE neurons (Figure S4). Nevertheless, dendrites of ddaF neurons that encounter dendrites of ddaE neurons are clearly repelled if both neurons express the same Dscam isoform. In contrast, higher expression of Dscam isoforms in ddaE neurons by using the highly restricted Ig1-1 Gal4 driver line (Sugimura et al., 2003) was found to also enhance repulsion of sister dendrite interactions and a reduction of dendritic complexity (Figure S 5). These results support the possibility that increased expression of Dscam isoforms in da neurons increases Dscam-Dscam mediated signaling and thereby enhances repulsion. This notion is also consistent with our finding that overexpression of a Dscam isoform that lacks the cytoplasmic domain (Zhu et al., 2006) blocks dendrite repulsion and leads to abnormal self-crossing as well as abnormal fasciculation of class I neurons (Figure S5). Importantly, the repulsion between ddaF and ddaE (Figure 7) or between sister dendrites of ddaE does not appear to depend on the type of Dscam isoform, as we observed the same increase in repulsion for two significantly different isoforms Dscam1.30.30.1 and Dscam 11.31.25.1 (Figure 7G).

Dscam function in dendrite repulsion of class I neurons is independent of Pak signaling

Previous studies have shown that Dscam signaling recruits the downstream signaling components Dock and Pak (Schmucker et al., 2000). Loss-of-function studies as well as dosage sensitive genetic interactions between Dscam and Pak suggest that Pak is an important downstream effector of Dscam signaling during axon guidance of Bolwig’s nerve (Schmucker et al., 2000). Furthermore, it has been shown that Dock and Pak can mediate repulsive signaling downstream of the Robo receptor (Fan et al., 2003). To test whether Pak is also required for repulsion and Dscam signaling in da dendrites, we examined loss-of-function as well as gain-of-function phenotypes of Pak in class I da neurons. We found that in homozygous Pak animals at 2nd instar, ddaE morphology is normal and the development of dendrite arborizations is indistinguishable from wildtype (Figure 8 A, B). Pak-null neurons had no observable increase in dendritic self-crossing and no significant alteration in the number of dendritic termini per cell (Figure 8 C). However, expression of a membrane-targeted myristoylated Pak - known to function as a dominant activated version of Pak (Hing et al., 1999) - strongly increased the total number of tertiary branches of ddaE neurons in 3rd instar larvae, suggesting that one of Pak’s downstream substrates may play a role in da neuron branching (Figure 8 E, F). However, endogenous Pak does not appear to be required for dendrite morphogenesis of class I da neurons. Although it is possible that Pak function is required in other da neurons, these results suggest that Pak is not a general factor mediating Dscam’s repellent signaling function, and that other not yet identified components are necessary for Dscam-mediated repulsion.

Figure 8. Dscam does not signal through Pak in Class I dendrites.

Figure 8

Open in a new tab

(A-B, D-E) IG1-1 Gal4 was used to drive expression of CD8-GFP in ddaE neurons. (A-B) The morphology of Class I neurons from 2nd instar larvae was examined in wildtype (A) and Pak-null (B) animals (scale bar, 20μm). (C) Loss of Pak had no significant effect on either iso-neuronal self-crossing or the complexity of dendritic arborizations (Wildtype, N = 12, Pak-null, N=13). Over-expression of myristoylated Pak (E) caused excessive branching of ddaE neurons compared to wildtype cells (D) (scale bar, 20μm), precisely opposite the phenotype of mis-expressing Dscam (Figures 3, 4). These data are quantified in panel (F) (P<0.001) (Wildtype N=48, UAS-Pak N=20).

Discussion

In this study we show that Dscam has an important cell-intrinsic function in dendrite development of da neurons. Dscam is required for correct dendrite morphogenesis by controlling the guidance of sister branches but is not required for other mechanisms of dendrite patterning. Dscam loss-of-function mutations result in strong disruption of dendrite morphogenesis in different classes of da sensory neurons. The phenotypic defects included uneven spacing of dendritic branches, a strong increase in dendritic self-crossing and highly abnormal dendritic fascicles or tangles. All the observed phenotypes are consistent with the possibility that loss of Dscam results in a lack of self-avoidance of sister dendrites. Consistent with a role of Dscam in dendrite-dendrite repulsion, it was found that Dscam over-expression in da neurons, which normally have overlapping dendritic fields, forced the respective dendrites to segregate from each other. In addition, gain-of-function phenotypes resulting from over-expression of single Dscam isoforms or a Dscam isoform lacking the cytoplasmic domain are also consistent with the possibility that repulsion between sister branches is controlled by Dscam signaling. We therefore suggest that in da neurons, direct isoform-specific homophilic Dscam-Dscam interactions result in signal transduction events that lead to repulsion of dendrites expressing identical Dscam isoforms. This model is consistent with previous biochemical studies (Wojtowicz et al., 2004), Dscam’s role in bifurcating mushroom body axons (Wang et al., 2002) as well as Dscam’s function in projection neurons of the olfactory system (Zhu et al., 2006). Previous expression studies have shown that single photoreceptor neurons of the same type express different Dscam isoforms (Neves et al., 2004). Similarly, expression of a large diversity of Dscam isoforms has also been found in olfactory neurons as well as mushroom body neurons (Hummel et al., 2003; Zhan et al., 2004). Based on these findings and our observation that experimentally forced expression of identical isoforms in da neurons causes dominant phenotypes, it is highly likely that different da neurons also express diverse Dscam isoforms. Considering the large diversity of Dscam isoforms, the possibility that dendrites from different neurons present identical Dscam isoforms seems minimal. In contrast, dendritic sister-branches of the same cell, even though they are likely expressing multiple isoforms, will at significant frequency encounter homophilic Dscam-Dscam interactions This model is consistent with our finding that the expression of the same Dscam isoform causes the segregation of normally overlapping dendritic fields. Based on this one might expect that a functionally critical threshold of Dscam diversity must exist. However, we have not detected any obvious morphogenesis defects of class I neurons, such as changes in self-crossing, number of dendritic termini, or dendritic area, in homozygous animals bearing the reduced diversity allele DscamC22-1 (Figure 5). We have also not detected any obvious defects in dendrite morphogenesis of class IV neurons in DscamC22-1 animals (data not shown). Although it is possible that some aspects of dendrite morphogenesis are altered in DscamC22-1 and were not identified in our experiments, we favor the possibility that a few thousands or even significantly fewer Dscam isoforms are sufficient to still ensure non-overlapping expression of identical isoforms in neighboring da neurons. However, we propose that reducing the diversity of Dscam isoforms below a certain threshold would lead to scenarios where different neighboring da neurons express the same isoforms and it would not be possible to limit Dscam-mediated repulsion of dendrites to cell-intrinsic sister dendrite interactions. This would likely lead to strong morphogenesis and functional defects throughout the Drosophila PNS.

Dendrite self-avoidance is molecularly different from hetero-neuronal tiling

Dendrite development of da neurons requires at least four distinguishable patterning mechanisms: First, growth of dendrites and dendritic branches emanating from the same cell has to be controlled such that relatively even spacing between dendrites with minimal overlap is achieved (self-avoidance). Second, for any given class or type of neuron, the dendritic growth has to obey a characteristic polarity and likely limits the extension of the primary dendritic branches (dendrite architecture). Third, the degree of branching has to be adapted to the type of sensory neuron (stereotyped branching). Fourth, inhibitory interactions with nearby neurons are needed to control the size of dendritic fields such that a complete but non-redundant innervation of a receptive area by functionally uniform groups of neurons is achieved (hetero-neuronal tiling).

It has been previously speculated that self-avoidance and tiling might depend on the same molecular mechanism and may not require distinct signals. In such a scenario, isoneuronal dendrites could be developmentally identical to “like” hetero-neuronal dendrites (Grueber et al., 2002). In this study we suggest that this may not be the case. Dscam function is required for correct spacing of dendrites due to self-avoidance of sister branches but is unlikely required for other mechanisms of dendrite patterning. These results suggest that the repulsive mechanism(s) underlying hetero-neuronal tiling are molecularly different from the mechanism controlling repulsive interactions underlying self-avoidance. It seems likely that homophilic Dscam-Dscam interactions represent the major molecular system controlling iso-neuronal dendrite-dendrite repulsion in Drosophila. In this specific context of dendrite morphogenesis, the diversity of Dscam ensures that this repulsive function is restricted to cell-intrinsic interactions, as only dendrites of the same cell are likely to express identical isoforms. As such, the molecular diversity of Dscam is less likely to provide each neuron with a unique “identity” (Neves et al., 2004) but rather provides a molecular buffer for enabling “tolerance” between neurons.

Dscam-mediated repulsive signaling? Switching from homophilic adhesion to repulsion

Several studies have revealed examples consistent with the notion that Dscam signaling can lead to neurite repulsion (Wang et al., 2002; Wojtowicz et al., 2004; Zhan et al., 2004; Zhu et al., 2006). It has been proposed that this repulsive function can be mediated by direct homophilic Dscam-Dscam interactions. For example, it has been shown that the trajectory of interneurons over-expressing a single isoform of Dscam is disrupted upon encountering midline cells that over-express the identical isoform (Wojtowicz et al., 2004). The strongest support for a direct Dscam-Dscam interaction has been provided by a series of impressive biochemical experiments, in which it was shown that from a randomly chosen set of 11 Dscam isoforms each one binds to itself but not to others (Wojtowicz et al., 2004). All three variable Ig domains of Dscam are required for homophilic binding specificity. In addition, recent studies described that over-expression of Dscam in a subset of projection neurons connecting with specific glomeruli (termed DA1 and DC3; see Zhu et al., 2006) resulted in a strong gain-of-function phenotype, again consistent with a repulsive interaction due to homophilic Dscam interactions. This gain-of-function phenotype was found to be dependent on Dscam signaling, as a deletion of the cytoplasmic domain in a Dscam isoform (“Dscam1.30.30.1Δcyto”) blocked this dominant phenotype (Zhu et al., 2006). Similarly, we found that overexpression of Dscam1.30.30.1Δcyto in ddaE neurons blocked the repulsion of sister dendrites and instead lead to abnormal fasciculation (Figure S5 D-E). Although the endogenous physiological function revealed by these experiments is unclear, they nevertheless are consistent with the hypothesis that Dscam can function as a cell-surface receptor mediating neurite repulsion.

How is the homophilic Dscam-Dscam interaction transformed into a repulsive action rather than a stable adhesion? Dscam has been initially identified as a tyrosine phosphorylated receptor functioning upstream of the adaptor molecule Dock (Schmucker et al., 2000). Dock binds to Dscam via SH2 as well as SH3 domains and serves to recruit the effector-kinase Pak to the plasma membrane where it can be activated by Rac or Cdc42 (Hing et al., 1999). Pak has been implicated in several signaling pathways controlling cytoskeletal rearrangement, including pathways underlying neurite repulsion (Bokoch, 2003; Fan et al., 2003). By examining the effect of a constitutively membrane bound form of Pak in da neurons, we found that Pak signaling can influence dendrite morphogenesis. However, loss-of-function analysis provided no evidence for a direct role of Pak in dendrite morphogenesis or self-avoidance of class I neurons. Therefore, at least in class I neurons, Dscam signaling likely bypasses Pak and utilizes alternative downstream components, which have yet to be identified. Although a signaling pathway controlling heteroneuronal tiling and branching in da neurons has been described (Emoto et al., 2004), signaling pathways that control self-avoidance are currently unknown.

Diverse signaling roles of Dscam

It is important to note that Dscam function is not only required for controlling neurite repulsion. For example, it has been proposed that Dscam controls axon guidance of Bolwig’s nerve by signaling through Dock and Pak in response to an as yet unknown guidance cue present at an intermediate target (Schmucker et al., 2000). In early developing mushroom body fibers, Dscam is required for axon bundling and fasciculation, thereby mediating adhesive interactions (Zhan et al., 2004). In addition, one distinct function of Dscam in mechanoreceptor neurons appears to mediate a growth-promoting role of Dscam rather than repulsion (Chen et al., 2006). Importantly, the role of Dscam diversity in the development of mechanosensory neuron projections suggests the possibility that Dscam is not only involved in homophilic interactions of neurites emanating from the same cell. In fact, it has been proposed that in the somatosensory system Dscam isoforms have instructive roles controlling targeting decisions of axonal branches (Chen et al., 2006). Future studies will have to address the molecular differences that allow for such a versatile use of Dscam receptors.

Materials and Methods

Fly stocks and MARCM

To characterize the null phenotype of Dscam in da neurons, virgin females of the stock FRT42D, TubGal80 ; C161-Gal4 , UAS-mCD8-GFP / TM2 were crossed to males from the following three stocks: 1. hs-Flp ; FRT42D, w+ 2. hs-Flp ; FRT42D , dscam21 / CyO and 3. hs-Flp ; FRT42D , dscam33 / CyO. Embryos were collected on grape plates at 2 hour intervals and incubated at 25°C for 3 hours. 3-5 hours AEL, the embryos were heat shocked at 37°C for 1 hour and incubated at 25°C until they were analyzed as 3rd instar larvae. For mis-expression and rescue experiments in ddaE (class I neurons), virgin females from the stock UAS-mCD8-GFP ; Ig1-1 were crossed to males from the following four stocks: 1. Bl / CyO ; UAS-Dscam(4.1, 6.30, 9.30, 17.1) 2. Bl / CyO ; UAS-Dscam(4.1, 6.30, 9.30, 17.2) 3. Bl / CyO ; UAS-Dscam(4.7, 6.27, 9.25, 17.1) 2. Bl / CyO. Mis-expression in Class I and Class III neurons was performed by crossing UAS-Dscam(1.30.30.1) males to C161-Gal4 , UAS-mCD8-GFP / TM2 virgin females. Pak loss-of-function was analyzed at 2nd instar using either Pak2 or Pak7 homozygotes in combination with Ig1-1 Gal4 and UAS-mCD8-GFP. UAS-myr-Pak was crossed to Ig1-1 to examine Pak’s gain-of-function phenotype in 3rd instar larvae.

Microscopy

In both MARCM and mis-expression experiments, 3rd instar larvae were screened for GFP expression using a Zeiss Stemi SV11 fluorescent dissection microscope. Larvae selected for subsequent analysis were dissected in 80% glycerol and 20% PBS. A scalpel was used to remove the posterior end of the larva and through this incision, the gut, fat body, and tracheal tubes were removed. The dissected larvae were gently stretched using forceps and mounted in 80% glycerol using 1mm coverslips as spacers to ensure a stable base for the overlaying coverslip. mCD8-GFP expression was visualized with a Zeiss LSM 410 inverted confocal microscope using a 40X oil emersion objective and a Kr/Ar laser for 488nm excitation. Z-sections of 1 micron depth were taken of the entire dendritic arborization (10-25 total sections) and used to generate maximal z-projections for subsequent analysis. Time-lapse analysis and imaging whole-mount larvae were done essentially as previously described (Sugimura et al., 2003; Figure S2-S4)

Image Analysis

Both individual z-sections and the maximal projections were taken at 512×512 resolution and exported to Adobe Photoshop to normalize the orientation of the neurons and adjust levels and contrast. For MARCM analysis, the identity of any given neuron was determined based on the location of the cell body, the orientation and complexity of the dendritic arborization, and comparison to other neurons in the field of view in cases where more than one neuron was labeled per hemi-segment,. Measurements and analysis were performed using ImageJ software. To quantify dendritic self-crossing, individual neurites were traced from the cell body to individual dendritic termini and the number of times a given branch crosses neighboring dendrites from the same cell was recorded. The total number of dendritic branches and termini were recorded using the same tracing method. In the analysis of Dscam’s null phenotype, approximately 5% of the null clones were so disorganized that it was impossible to accurately trace individual neurites. These clones were not included in the quantitative analysis and likely result in an under-estimation of the degree of dendritic self-crossing present in Dscam-null cells. Total area was computed using the polygon method described previously (Grueber et al., 2002), and dendritic length was calculated as the sum of the lengths of every individual branch or secondary branch projecting to the posterior segment boundary.

Supplementary Material

01
NIHMS22814-supplement-01.pdf (1,023.5KB, pdf)

Acknowledgments

We would like to thank Q. Ma and members of the Schmucker lab for helpful comments on the manuscript. We thank Wes Grueber, Larry Zipursky and Yuh-Nung Jan for communicating data prior to publication. We also thank the Bloomington Drosophila Stock Center, S. L. Zipursky (Dscam transgenes), W. Grueber (Gal4 lines), T. Lee (DscamC22-1) for providing useful fly stocks. This research was supported by the NIH (RO1-NS46747) (D.S.); a Pew Scholars Award (D.S.); a John Merck Fund Award (D.S.); a Leffler Fellowship (M. K.); and a Stuart and Victoria Quan Fund Award (M.H.). Work of T. U. and A. T. was supported by Grants-in-Aid for Scientific Research on Priority Areas-Molecular Brain Science of the MEXT of Japan (17024025 to T. U.). A.T. was supported by a JSPS Research Fellowship for Young Scientists.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Bokoch GM. Biology of the p21-activated kinases. Annu Rev Biochem. 2003;72:743–781. doi: 10.1146/annurev.biochem.72.121801.161742. [DOI] [PubMed] [Google Scholar]
  2. Chen BE, Kondo M, Garnier A, Watson FL, Puettmann-Holgado R, Lamar DR, Schmucker D. The molecular diversity of Dscam is functionally required for neuronal wiring specificity in Drosophila. Cell. 2006;125:607–620. doi: 10.1016/j.cell.2006.03.034. [DOI] [PubMed] [Google Scholar]
  3. Emoto K, He Y, Ye B, Grueber WB, Adler PN, Jan LY, Jan YN. Control of dendritic branching and tiling by the Tricornered-kinase/Furry signaling pathway in Drosophila sensory neurons. Cell. 2004;119:245–256. doi: 10.1016/j.cell.2004.09.036. [DOI] [PubMed] [Google Scholar]
  4. Fan X, Labrador JP, Hing H, Bashaw GJ. Slit stimulation recruits Dock and Pak to the roundabout receptor and increases Rac activity to regulate axon repulsion at the CNS midline. Neuron. 2003;40:113–127. doi: 10.1016/s0896-6273(03)00591-9. [DOI] [PubMed] [Google Scholar]
  5. Grueber WB, Jan LY, Jan YN. Tiling of the Drosophila epidermis by multidendritic sensory neurons. Development. 2002;129:2867–2878. doi: 10.1242/dev.129.12.2867. [DOI] [PubMed] [Google Scholar]
  6. Grueber WB, Jan LY, Jan YN. Different levels of the homeodomain protein cut regulate distinct dendrite branching patterns of Drosophila multidendritic neurons. Cell. 2003a;112:805–818. doi: 10.1016/s0092-8674(03)00160-0. [DOI] [PubMed] [Google Scholar]
  7. Grueber WB, Ye B, Moore AW, Jan LY, Jan YN. Dendrites of distinct classes of Drosophila sensory neurons show different capacities for homotypic repulsion. Curr Biol. 2003b;13:618–626. doi: 10.1016/s0960-9822(03)00207-0. [DOI] [PubMed] [Google Scholar]
  8. Grueber WB, Jan YN. Dendritic development: lessons from Drosophila and related branches. Curr Opin Neurobiol. 2004;14:74–82. doi: 10.1016/j.conb.2004.01.001. [DOI] [PubMed] [Google Scholar]
  9. Grueber WB, Yang CH, Ye B, Jan YN. The development of neuronal morphology in insects. Curr Biol. 2005;15:R730–738. doi: 10.1016/j.cub.2005.08.023. [DOI] [PubMed] [Google Scholar]
  10. Hing H, Xiao J, Harden N, Lim L, Zipursky SL. Pak functions downstream of Dock to regulate photoreceptor axon guidance in Drosophila. Cell. 1999;97:853–863. doi: 10.1016/s0092-8674(00)80798-9. [DOI] [PubMed] [Google Scholar]
  11. Hummel T, Vasconcelos ML, Clemens JC, Fishilevich Y, Vosshall LB, Zipursky SL. Axonal targeting of olfactory receptor neurons in Drosophila is controlled by Dscam. Neuron. 2003;37:221–231. doi: 10.1016/s0896-6273(02)01183-2. [DOI] [PubMed] [Google Scholar]
  12. Jan YN, Jan LY. The control of dendrite development. Neuron. 2003;40:229–242. doi: 10.1016/s0896-6273(03)00631-7. [DOI] [PubMed] [Google Scholar]
  13. Lee T, Luo L. Mosaic analysis with a repressible cell marker (MARCM) for Drosophila neural development. Trends Neurosci. 2001;24:251–254. doi: 10.1016/s0166-2236(00)01791-4. [DOI] [PubMed] [Google Scholar]
  14. Neves G, Zucker J, Daly M, Chess A. Stochastic yet biased expression of multiple Dscam splice variants by individual cells. Nat Genet. 2004;36:240–246. doi: 10.1038/ng1299. [DOI] [PubMed] [Google Scholar]
  15. Parrish JZ, Kim MD, Jan LY, Jan YN. Genome-wide analyses identify transcription factors required for proper morphogenesis of Drosophila sensory neuron dendrites. Genes Dev. 2006;20:820–835. doi: 10.1101/gad.1391006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Schmucker D, Clemens JC, Shu H, Worby CA, Xiao J, Muda M, Dixon JE, Zipursky SL. Drosophila Dscam is an axon guidance receptor exhibiting extraordinary molecular diversity. Cell. 2000;101:671–684. doi: 10.1016/s0092-8674(00)80878-8. [DOI] [PubMed] [Google Scholar]
  17. Senti KA, Usui T, Boucke K, Greber U, Uemura T, Dickson BJ. Flamingo regulates R8 axon-axon and axon-target interactions in the Drosophila visual system. Curr Biol. 2003;13:828–832. doi: 10.1016/s0960-9822(03)00291-4. [DOI] [PubMed] [Google Scholar]
  18. Shepherd D, Smith SA. Central projections of persistent larval sensory neurons prefigure adult sensory pathways in the CNS of Drosophila. Development. 1996;122:2375–2384. doi: 10.1242/dev.122.8.2375. [DOI] [PubMed] [Google Scholar]
  19. Sugimura K, Satoh D, Estes P, Crews S, Uemura T. Development of morphological diversity of dendrites in Drosophila by the BTB-zinc finger protein abrupt. Neuron. 2004;43:809–822. doi: 10.1016/j.neuron.2004.08.016. [DOI] [PubMed] [Google Scholar]
  20. Sugimura K, Yamamoto M, Niwa R, Satoh D, Goto S, Taniguchi M, Hayashi S, Uemura T. Distinct developmental modes and lesion-induced reactions of dendrites of two classes of Drosophila sensory neurons. J Neurosci. 2003;23:3752–3760. doi: 10.1523/JNEUROSCI.23-09-03752.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Sweeney NT, Brenman JE, Jan YN, Gao FB. The coiled-coil protein shrub controls neuronal morphogenesis in Drosophila. Curr Biol. 2006;16:1006–1011. doi: 10.1016/j.cub.2006.03.067. [DOI] [PubMed] [Google Scholar]
  22. Wang J, Ma X, Yang JS, Zheng X, Zugates CT, Lee CH, Lee T. Transmembrane/juxtamembrane domain-dependent Dscam distribution and function during mushroom body neuronal morphogenesis. Neuron. 2004;43:663–672. doi: 10.1016/j.neuron.2004.06.033. [DOI] [PubMed] [Google Scholar]
  23. Wang J, Zugates CT, Liang IH, Lee CH, Lee T. Drosophila Dscam is required for divergent segregation of sister branches and suppresses ectopic bifurcation of axons. Neuron. 2002;33:559–571. doi: 10.1016/s0896-6273(02)00570-6. [DOI] [PubMed] [Google Scholar]
  24. Williams DW, Truman JW. Cellular Mechanisms of dendrite pruning in Drosophila: insights from in vivo time-lapse of remodeling dendritic arborizing sensory neurons. Development. 2005;132:3631–3642. doi: 10.1242/dev.01928. [DOI] [PubMed] [Google Scholar]
  25. Wojtowicz WM, Flanagan JJ, Millard SS, Zipursky SL, Clemens JC. Alternative splicing of Drosophila Dscam generates axon guidance receptors that exhibit isoform-specific homophilic binding. Cell. 2004;118:619–633. doi: 10.1016/j.cell.2004.08.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Zhan XL, Clemens JC, Neves G, Hattori D, Flanagan JJ, Hummel T, Vasconcelos ML, Chess A, Zipursky SL. Analysis of Dscam diversity in regulating axon guidance in Drosophila mushroom bodies. Neuron. 2004;43:673–686. doi: 10.1016/j.neuron.2004.07.020. [DOI] [PubMed] [Google Scholar]
  27. Zhu H, Hummel T, Clemens JC, Berdnik D, Zipursky SL, Luo L. Dendritic patterning by Dscam and synaptic partner matching in the Drosophila antennal lobe. Nat Neurosci. 2006;9:349–355. doi: 10.1038/nn1652. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

01
NIHMS22814-supplement-01.pdf (1,023.5KB, pdf)

RESOURCES