Neuroligin 2 governs synaptic morphology and function through RACK1-cofilin signaling in Drosophila

Figures

• 

  Fig. 1. Actin dynamics and pre- and postsynaptic protein apposition are mis-regulated in dnlg2 mutant NMJs. a F-actin is downregulated in dnlg2 and cofilin mutants. Confocal images of WT (n = 19), dnlg2 mutants (dnlg2KO70/KO70) (n = 15), and cofilin mutants (tsrN96A/+) (n = 14) third instar larvae NMJ type Ib boutons at muscles 12/13 labeled with Texas Red phalloidin (red, F-actin) and anti-HRP (blue). b Scatter diagram shows a significant decrease in the relative intensity of F-actin in dnlg2 mutants, and cofilin mutants compared with WT. c G-actin is upregulated in dnlg2 and cofilin mutants. Confocal images of WT (n = 29), dnlg1 (dnlg1ex1.9/ex2.3) mutants (n = 23), dnlg2 mutants (n = 30), and cofilin mutants (n = 9) third instar larvae NMJ type Ib boutons at muscles 12/13 labeled with anti-DNase I (green, G-actin) and anti-HRP (blue). d Scatter diagram shows a significant increase in the relative intensity of G-actin in dnlg1 mutants, dnlg2 mutants, and cofilin mutants compared with WT. e Relative apposed active zones decrease in dnlg2 and cofilin mutants. Confocal images of WT (n = 10), dnlg1 mutants (n = 10), dnlg2 mutants (n = 10) and cofilin mutants (n = 8), third instar larvae NMJ…

• 

  Fig. 2. The decreased F-actin level and bouton number at the NMJ of dnlg2 mutants could be rescued by modulating postsynaptic Cofilin activity. a p-Cofilin is downregulated in dnlg2 mutants. Muscle lysates of the third instar larvae from WT (n = 5), dnlg1 mutants (n = 4), and dnlg2 mutants (n = 4) were subjected to western blots with anti-p-Cofilin antibody. b Scatter diagram shows the relative amount of p-Cofilin in both lines in a. p-Cofilin expression was dramatically inhibited in dnlg1 mutant and dnlg2 mutant. c Muscle lysates of the third instar larvae from WT (n = 23), dnlg1 mutants (n = 9), and dnlg2 mutants (n = 23) were subjected to western blots with anti-Cofilin antibody. d Scatter diagram of the total Cofilin relative amount in c. There was no change in the total Cofilin expression in dnlg1 mutants and dnlg2 mutants. e Confocal images of the third instar larvae NMJs from WT (n = 12), dnlg2 mutants (n = 15), dnlg2; C57 > CofilinWT (n = 17), dnlg2; C57 > CofilinS3A (n = 19) and dnlg2; C57 > CofilinS3E (n = 13). Type Ib boutons at muscles 12/13 were labeled with Texas Red phalloidin (red) and anti-HRP (blue). f Quantitative data of e shows the decreased F-actin relative…

• 

  Fig. 3. The aberrant GluRs recruitment at the NMJ of dnlg2 mutants could be rescued by modulating postsynaptic Cofilin activity. a Confocal images of the third instar larvae NMJs from WT (n = 16), dnlg2 mutants (n = 16), dnlg2; C57 > CofilinWT (n = 12), dnlg2; C57 > CofilinS3A (n = 13) and dnlg2; C57 > CofilinS3E (n = 16). Type Ib boutons at muscle 4 were labeled with anti-GluRIIB (red) and anti-nc82 (green, BRP). b Quantitative data of a shows the decreased apposed active zone in dnlg2 mutants and restoration by muscle expression of CofilinWT, and CofilinS3A. CofilinS3E over-rescued this defect. c Confocal images of the third instar larvae NMJs from WT (n = 8), dnlg2 mutants (n = 8), dnlg2; C57 > CofilinWT (n = 8), dnlg2; C57 > CofilinS3A (n = 8) and dnlg2; C57 > CofilinS3E (n = 7). Type Ib boutons at muscle 4 were labeled with anti-8B4D2 (red, GluRIIA) and anti-HRP (blue). The right lane shows the amplified contents framed in merge images. White arrowheads highlight the orphan boutons. d Quantitative data of c shows the increased percentage of orphan bouton in dnlg2 mutants and restoration by muscle expression of CofilinWT and CofilinS3A. Data are presented as mean ± SEM.

• 

  Fig. 4. The aberrant synaptic transmission and locomotion activity in dnlg2 mutants could be rescued by modulating postsynaptic Cofilin activity. a Representative trace of the mEJP. b Scatter diagram of the mean mEJP amplitude (in mV) from WT (n = 24), dnlg2 mutants (n = 19), dnlg2; C57 > CofilinWT (n = 13), dnlg2; C57 > CofilinS3A (n = 22) and dnlg2; C57 > CofilinS3E (n = 17) third instar larvae. c Scatter diagram of the mean mEJP frequency (in Hz) from WT (n = 26), dnlg2 mutants (n = 23), dnlg2; C57 > CofilinWT (n = 14), dnlg2; C57 > CofilinS3A (n = 22) and dnlg2; C57 > CofilinS3E (n = 17) third instar larvae. d Representative traces of a 3-min crawling of WT (total n = 35), dnlg2 mutants (n = 37), dnlg2; C57 > CofilinWT (n = 63), dnlg2; C57 > CofilinS3A (n = 37) and dnlg2; C57 > CofilinS3E (n = 49) third instar larvae. e Quantification of a 3-min crawling distance (in mm) in d, showing a reduced locomotor activity in dnlg2 mutants and restoration by muscle expression of CofilinWT and CofilinS3A. Data are presented as mean ± SEM.

• 

  Fig. 5. DNlg2 can interact with RACK1 and Cofilin, and RACK1 can potentially interact with active Rac1. a DNlg2, RACK1 and Cofilin can interact with each other in vivo. IP lysates was enriched from da-GAL4 > UAS-DNlg2-HA adult brain extracts with anti-HA antibody. b DNlg2, RACK1 and Cofilin can interact with each other in vivo. IP lysates was enriched from da-GAL4 > UAS-Cofilin-GFP adult brain extracts with anti-GFP antibody. c Schematic diagram of the synthesis N-terminal (Q37-Y684) and C-terminal (L685-V1248) fragments of DNlg2 bound to the Strep tag II. d DNlg2 binds RACK1 through its C-terminal fragment. Only the C-terminal of DNlg2 can directly bind to RACK1, whereas DNlg2 N-terminal cannot. e Western blots show the level of Rac1-GTP in adult heads of WT and dnlg2 mutants. f Scatter diagram of the Rac1-GTP relative amount in e, indicating the decreased Rac1-GTP in dnlg2 mutants. g Western blots show the level of RhoA-GTP and total RhoA in adult heads of WT and dnlg2 mutants. h Scatter diagram of the RhoA-GTP relative amount in g, indicating the decreased RhoA-GTP in dnlg2 mutants. i IP from HEK293T extracts with overexpression of Drosophila RACK1-GFP and Rac1-mcherry, enriched with…

• 

  Fig. 6. Model for Neuroligin2-RACK1-Cofilin signaling in WT and dnlg2 mutants. In WT Drosophila NMJs, postsynaptic DNlg2 interacts with RACK1 which can form a complex with downstream Rac1-Cofilin or RhoA-Cofilin signaling components to activate Cofilin phosphorylation. Only if phosphorylated and non-phosphorylated Cofilin kept in a right balance, can actin polymerization/depolymerization be maintained in a good balance. Otherwise, disruption of DNlg2 will destabilize the big DNlg2-RACK1-Rac1 (or RhoA)-Cofilin-signaling complex. This disruption will affect Rac1 and RhoA activation and therefore inhibit downstream Cofilin phosphorylation. The downregulated p-Cofilin will break F-actin polymerization/depolymerization balance to accelerate its depolymerization.