QUICK SEARCH:   [advanced] Author: Keyword(s):
Home     Help     Feedback     Subscriptions     Archive     Search     Table of Contents

First published online 26 January 2005
doi: 10.1242/dev.01661

This Article Summary Full Text Full Text (PDF) Supplementary Material Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Ting, C.-Y. Articles by Lee, C.-H. Search for Related Content PubMed PubMed Citation Articles by Ting, C.-Y. Articles by Lee, C.-H.

Drosophila N-cadherin functions in the first stage of the two-stage layer-selection process of R7 photoreceptor afferents

Chun-Yuan Ting^1,*, Shinichi Yonekura^1,*, Phoung Chung¹, Shu-ning Hsu², Hugh M. Robertson², Akira Chiba² and Chi-Hon Lee^1,

¹ Unit on Neuronal Connectivity, Laboratory of Gene Regulation and Development, National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, MD 20892, USA
² Department of Cell and Structure Biology, University of Illinois, Urbana, IL 61801, USA

View larger version (25K): [in a new window]   Fig. 5. Modular exon organization of the Ncad gene. The Ncad gene contains 21 exons that span approximately 100 Kb of genomic DNA. Mutually exclusive alternative splicing occurs in exons 7, 13, and 18. As deduced from cDNA sequences, mature mRNA contains exon 7a or 7b, exon 13a or 13b, and exon 18a or 18b. In addition, exon 7a' is found in some exon 7a-containing transcripts. Constant exons are shown as red boxes and alternative exons as green (7a,13a,18a) or blue (7b,13b,18b) boxes. Exons 7a and 7b each encodes the C-terminal half of CA8 and the N-terminal half of CA9; exons 13a and 13b each encode the C-terminal half of CA11 and the N-terminal half of CA12; exons 18a and 18b each encode the C-terminal half of EGF-CA3, the entire EGF-CA4, and the N-terminal half of the TM. The exon 7a' encodes an insertion of four amino-acid residues in CA8. The amino-acid sequence identity between the regions encoded by the alternative exons is indicated. SP: signal peptide; CA: cadherin domain; EGF: EGF-like calcium-binding repeat; LamG: Laminin-G-like domain; TM: transmembrane region; CP: cytoplasmic domain containing ß-catenin binding site.

View larger version (82K): [in a new window]   Fig. 1. R8, R7, and L1-L5 axons sequentially project into distinct layers in the medulla during the first layer-selection stage. (A,A',B,B',B'') 17% APF; (C,C',D,D') 35% APF. (A,A') The initial projections of R7 and R8 axons were assessed at 17% APF, using PM181-Gal4, UAS-mCD8-GFP (green) and Mab24B10 (red), respectively. (A') A high magnification view of medulla of A. The developing medulla is viewed in cross section to reveal the developmental sequence of R7 and R8 innervation at 17% APF. In this view, the younger R7 and R8 growth cones are to the left and the older ones to the right. The newly arriving R7 growth cones (arrow) project past the R8 growth cones and then expand (arrowhead) to terminate at the layer below the R8 growth cones. (B,B',B'') The L1-L5 projections viewed in the same direction and at the same stage as in A,A'. The L1-L5 axons expressed Gcm-Gal4, UAS-lacZ, and were visualized with anti-lacZ antibody (blue). At this stage, a subset of young medulla neurons (MN) also expresses Gcm although their axons have yet to enter the medulla. Both R7 and R8 axons were labeled using a pan-R-cell marker, GMR-GFP (green). R8 axons and older R7 axons are visualized with Mab24B10 (red). (B') A high magnification view of the medulla of B. The newly arriving R7 growth cones (arrow) express GFP but not Mab24B10. (B'') The green channel (GFP) is removed from B' to facilitate the visualization of the L1-L5 projections (blue). Based on the R7 and L1-L5 markers, we estimated that the newly differentiating L1-L5 growth cone (arrow) reaches the medulla 1-2 rows behind R7s. The older L1-L5 growth cones (arrowhead) terminate between R7 and R8 growth cones. (C,C') R7 and R8 growth cones (red) form two separate medulla layers at 35% APF. R7 axons were labeled using PM181-Gal4, UAS-mCD8-GFP marker, which was prolonged with the UAS-Gal4 method. This method also labels subsets of medulla (MN) and lobular (lo) neurons, which normally express PM181 at a very low level. (D,D') L1-L5 growth cones form two separate layers between R7 and R8 growth cones at 35% APF. The L1-L5 neurons were labeled as in B. At this stage, medulla glia (MG), which wrap around external medulla, express Gcm. The presumptive R7-temporary layers are indicated by dotted lines in B',B''. (A'-D') High magnification views of A-D, respectively. ed, eye disc; la, lamina; me, medulla: lo, lobula; LN, L1-L5 neurons; MN, medulla neurons; MG, medulla glia. Scale bars: in A, 30 µm for A-D; in A' 10 µm for A'-D'. (E) A schematic diagram illustrating the order of R-cell and L1-L5 afferents innervating the medulla. The developmental sequences in the eye disc, lamina, and medulla are indicated by arrows. R8 (red), R7 (green), and L1-L5 (blue) axons sequentially innervate the medulla (gray shaded box) to reach their temporary layers (as indicated).

View larger version (85K): [in a new window]   Fig. 2. R8 and R7 growth cones reach their destined layers during the second layer-selection stage. (A-D') R7 and R8 layer-specific targeting assessed at 40% (A,A'), 50% (B,B'), 60% (C,C'), and 70% APF (D,D'). Approximately 70% of R7 axons were labeled using a flipase-based method (see Materials and methods) and visualized with the anti-GFP antibody (green). R8 axons (red) were marked with RFP but not GFP. The neuropils were stained with the anti-HRP antibody (purple). (A') At 40% APF, R7 (yellow, large arrow) and R8 (red, large arrowhead) growth cones terminate at their temporary layers. (B') At 50% APF, R7 growth cones (yellow, large arrow) extend 3 µm further with filapodia (small arrow) leaving behind in the R7-temporary layer. R8 growth cones (red) extend filapodia (small arrowhead) deeper into the medulla while leaving their growth cone propers (large arrowhead) in the R8-temporary layer. (C') At 60% APF, both R8 (large arrowhead) and R7 (large arrow) growth cones have reached their final target layers. Note that R7 axon shafts expanded in the R7-temporary layer (small arrow). (D') At 70% APF, both R8 (large arrowhead) and R7 growth cones retract their filapodia and consolidate into axonal terminals. Two types of R7 terminals, straight (large arrow) and horseshoe (double-arrow) are noticeable (as seen in adult flies). (E-G') Layer-specific connections of R7 (C,C'), R8 (D,D') and L2 (E,E') assessed at the adult stage. (E,E') R7 axons (green) were labeled with Rh3 Rh4-Gal4, UAS-synb-GFP. Glass-lacZ marker labeled both R7 and R8 axons (red). (F,F') R8 axons (green) were labeled with Rh5 Rh6-Gal4, UAS-synb-GFP. (G,G') L2 axons (green) were labeled with L2-Gal4 UAS-mCD8-GFP. R7 and R8 axons were visualized with Mab24B10. Note that L2 axons terminate at the M2 layer, just above the R8-recipient layer (M3). (A'-G') High magnification views of A-G, respectively. Scale bars: in A, 30 µm for A-D; in A' 5 µm for A'-D'; in E, 20 µm for E-G; in E', 10 µm for E'-G'. (H) A schematic diagram showing the progression of R7 and R8 axonal targeting during the second layer-selection stage. For clarity, most L1-L5 neurons are omitted in the figure for the pupal stages. Abbreviations as in Fig. 1.

View larger version (71K): [in a new window]   Fig. 3. R8, R7, and L1-L5 axons target to distinct medulla layers independently at the first layer-selection stage. (A-D') Layer-specific targeting of R8, R7, and L1-L5 growth cones was assessed in various mutant backgrounds at 17% (A,A',C-D') and 40% APF (B,B'). The R7 axons (green) were labeled using PM181-Gal4, UAS-mCD8-GFP (A,A') or PM181-lacZ marker (C,C'), and visualized with anti-GFP or anti-LacZ antibodies, respectively. The L1-L5 axons (blue) were labeled using Gcm-Gal4, UAS-lacZ and visualized with anti-LacZ antibodies (C-D'). R8 and older R7 axons were visualized with Mab24B10 (red). Glia were visualized with anti-Repo antibody (green in B,B'). (A-B') In hh1 mutants, L1-L5 fail to differentiate while 11-13 rows of R-cells still develop (Huang and Kunes, 1996). In the absence of L1-L5, the R7 axons project past R8 growth cones and terminate at the layer below (A', arrowheads). However, the R7 growth cones do not separate from R8 growth cones until 40% APF (A,B'). The presumptive R7 and R8 layers were marked with dotted lines. The R-cell growth cones appear disorganized at 40% APF. (C,C') Expressing EGFRDN in the young lamina blocks the development of L1-L5 in this region and recapitulates the hh1 phenotypes in the younger part of medulla (bracket). (D,D') In sevenless mutants, R7 neurons fail to develop. However, the R8 (red) and L1-L5 (blue) axons (arrow and arrowhead) target correctly to their temporary layers. Scale bars: in A, 30 µm for A-D; in A',B', 10 µm for A'-D'. (E) A schematic diagram summarizing A,C,D.

View larger version (84K): [in a new window]   Fig. 4. N-cadherin functions in the first layer-selection stage to promote R7 axons reaching and remaining in the R7-temporary layer. R7 targeting was assessed at 17% APF (A-D'), at 35% APF (E-H'), and in adult flies (I-L'). Single mutant R7 cells were generated using GMR-Flp-mediated mitotic recombination (see text), and labeled using the MARCM system and mCD8-GFP (A-H') or synb-GFP (I-L'). R7 and R8 axons were visualized with Mab24B10 (red). (A,A',E,E',I,I') Wild-type; (B,B',F,F',J,J') Ncad405 mutant; (C,C',G,G',K,K') LAR2127 mutant; (D,D',H,H',L,L') Ncad405 LAR2127 double mutant. (A,A') At 17% APF, single R7 cells, homozygous for a wild-type FRT40 chromosome arm, project axons into the medulla. The wild-type axons first project past the R8 layer (arrow), then expand their growth cones (double arrow) just below the R8 growth cones, and separate (arrowhead) from the R8 growth cones. (B,B') At 17% APF, single Ncad405 mutant R7 axons project into the medulla. Approximately 21% of these mutant axons expand their growth cones incorrectly at the R8-temporary layer or between the R7- and R8-temporary layers (arrows). Over half of these mutant growth cones show severe morphological defects (arrowheads). (C,C') Single LAR2127 mutant R7 axons project into the medulla as in wild-type at 17% APF. (D,D') At 17% APF, single Ncad405 LAR2127 double mutant R7 axons exhibit targeting (arrows) and morphological defects (arrowhead) as seen in Ncad405 mutants. (E,E') Single wild-type R7 axons (arrow) terminate at the R7-temporary layer at 35% APF. (F,F') At 35% APF, 55% of the Ncad mutant R7 axons (arrow) terminate at the R8-temporary layer or between the R7- and R8-temporary layers. Some of them leave a small filapodium (arrowhead) connecting to the R7-temporary layer. (G,G') At 35% APF, most single LAR2127 mutant R7 axons terminate correctly at the R7-temporary layer in the younger part of the medulla. Approximately 9% of the mutant R7 growth cones exhibit abnormal morphology (arrowhead). Assessment of older R7 axons at this stage is limited because Elav-Gal4 driver is expressed at a low level in the older R7 cells. (H,H') At 35% APF, single Ncad405 LAR2127 double mutant R7 axons targeted to incorrect layers (arrows), as seen in the Ncad405 mutant. (I,I') In adult flies, single wild-type R7 axons terminate at the R7-recipient layer. (J-L') At the adult stage, single Ncad405 or LAR2127 or Ncad405 LAR2127 double mutant R7 axons terminate incorrectly at the R8-recipient layer. Note that in the region where mutant R7 axons mistarget to the R8-recipient layer, the corresponding R8 axons target correctly to the R8-recipient layer, leaving the R7-recipient layer uninnervated by any R-cell afferent. The presumptive R7- and R8-temporary layers are indicated by dotted lines in A'-D',I'-L'). (A'-L') High-magnification views of A-L, respectively. Scale bars: in A, 30 µm for A-L; in A', 10 µm for A'-L'. Abbreviations as in Fig. 1.

View larger version (78K): [in a new window]   Fig. 6. Expressing a single Ncad isoform in R7s is sufficient to rescue Ncad mutant phenotypes. Single Ncad mutant R7 cells were generated using GMR-Flp-mediated mitotic recombination, and a single Ncad isoforms 7b-13a-18a (A,A',C,C') or 7b-13a-18b (B,B',D,D') was expressed in these mutant R7s. These R7 axons (green) were labeled using the MARCM system and mCD8-GFP to assess layer-selection at 17% APF (A-B'), and 35% APF (C-D'). As in the wild type, the MACRM-rescued R7 growth cones (arrows) project past the R8-temporary layer and expand their growth cones (double arrows) to terminate (arrowheads) at the R7-temporary layer (A',B') where they still remain at 35% APF (arrows in C',D'). R7 and R8 axons were visualized with Mab24B10 (red). The presumptive R7 and R8-temporary layers are indicated by dotted lines in (A',B'). (A'-D') High-magnification views of A-D, respectively. Scale bars: in A, 30 µm for A-D; in A', 10 µm for A'-D'. Abbreviations as in Fig. 1. (E) Table summarizing the observed R7 targeting defects in wild type, various mutant backgrounds, and transgene-rescues.

View larger version (55K): [in a new window]   Fig. 7. Ncad isoforms mediate promiscuous heterophilic interactions. S2 cells expressing different cadherins and GFP or Ds-red marker are assessed for their ability to induce cell aggregations. (A-D) Cross-adhesion between cells expressing different Ncad isoforms or E-cadherin. Two S2 cell populations expressing different cadherins are separately labeled with GFP (green) or Ds-red (red) and mixed to form aggregates. (A) E-cadherin and Ncad 7b-13a-18b-expressing S2 cells segregate into two types of clusters. (B-D) S2 cells expressing different Ncad isoforms (as indicated) coaggregated to form mixed cell clusters. (E) A table summarizing the cross-adhesion results. +: intermixed; -: segregated.