Neuroligin 2 Is Required for Synapse Development and Function at the Drosophila Neuromuscular Junction

Targeted inactivation of dnl2 by homologous recombination. A , The structure of the dnl2 locus, targeting vector, and targeted locus are shown. Positions of primers used for PCR screening of targeted mutant flies in B are shown at the bottom of A . B , PCR analysis of genomic DNA from WT and dnl2^KO70 flies. A PCR product with a predicted size of 4.0 kb is generated only with genomic DNA from mutant flies because primer 1 is located outside the 5′ arm of the targeting vector and primer 2 is within the white gene. C , Northern blot analysis of total RNA isolated from the adult head using a dnl2 probe showing the predicted full-length 4.8 kb mRNA in WT but not in dnl2^KO70 flies. D , Western blot analysis of adult head extracts using anti-Dnl2 showing the expression of a 70 kDa protein in WT but not in homozygous dnl2^KO70 or dnl2^KO70/KO17 mutant flies.

Enhanced synaptic transmission at NMJs in dnl2 mutants. A , Representative traces of mEJPs in WT (top row) and dnl2^KO70 (bottom row) mutants recorded from muscles 6/7 in abdominal segment 3. B , Representative traces of stimulus-evoked EJPs in muscles 6/7 in WT (black trace) and dnl2^KO70 (red trace) mutants recorded in 0.8 mm calcium. C , D , No significant differences were observed in either the amplitudes ( C ) or frequencies ( D ) of mEJPs between WT and dnl2^KO70 mutants recorded in either high (0.8 mm) or low (0.2 mm) calcium. E , Stimulus-evoked EJP amplitudes were significantly increased in dnl2^KO70 mutants (*p < 0.05, Student's t test) in both high (0.8 mm) and low (0.2 mm) calcium. F , Pairs of stimuli were delivered at varying interstimulus intervals in 0.2 mm calcium, and the paired-pulse ratio was calculated as the amplitude of the second EJP divided by the amplitude of the first EJP. dnl2^KO70 mutants showed significantly less facilitation than the controls at all interstimulus intervals tested. G , H , Both the rise times ( G ) and the decay times ( H ) of stimulus-evoked EJPs were shorter in dnl2^KO70 mutants in both high (0.8 mm) and low (0.2 mm) calcium. **p < 0.01 versus WT, ***p < 0.001 versus WT, Student's t test.

Reduced synaptic boutons at larval NMJs in dnl2 mutants. A–F′ , NMJs at muscle 6/7 ( A–F ) and muscle 4 ( A′–F′ ) of larval abdominal segment 3 labeled with anti-HRP and anti-Dlg showing reduced NMJ expansion and fewer boutons in dnl2 null mutants dnl2^KO70 ( B , B′ ), dnl2^KO70 /Df ( C , C′ ), and transgenic mutants expressing dnl2 RNAi driven by 24B–Gal4 ( D , D′ ) compared with WT flies ( A , A′ ). The deficits in mutant dnl2^KO70 is rescued by muscle expression of UAS–WT dnl2 cDNA with either 24B–Gal4 ( E , E′ ) or C57–Gal4 ( F , F′ ). Scale bars, 20 μm. G , H , Summary graphs showing significant reduction in the numbers of total boutons at NMJ 6/7 ( G ) and type Ib bouton at NMJ4 ( H ) in both null mutants and dnl2 RNAi transgenic flies compared with WT control. I , Summary graph showing decreased synaptic branching in dnl2 null mutants and dnl2 RNAi transgenic flies. Note the reduced branching and bouton numbers in dnl2 null mutants and dnl2 RNAi transgenic flies are rescued by muscle expression of a UAS–dnl2 CDNA with either 24B–Gal4 (24B–rescue) or C57–Gal4 (C57–rescue). **p < 0.01 versus WT, ***p < 0.001 versus WT, ^## p < 0.01 versus dnl2^KO70 , ^### p < 0.001 versus dnl2^KO70 , Mann–Whitney test.

dnl2 mutants have altered active zones and PSDs. A–C″ , Confocal images of individual NMJ boutons costained with nc82 (anti-Brp) and anti-DPAK showing an increase in the number of BRP and DPAK staining clusters in dnl2 mutants ( B , dnl2^KO70 ; C , dnl2^KO70 /Df) and a reduction in the size of DPAK staining clusters in dnl2 mutants compared with WT control. D , The total number of DPAK clusters per bouton was increased in dnl2 mutants. E , The average staining intensity of Brp spots was also increased in dnl2 mutants. F , Summary graph showing a small but significant decrease in the mean size of DPAK clusters compared with WT control. Box plots depict the 25 and 75% quartiles (top and bottom edges of the box) and the minimum and maximum values (bottom and top bars). *p < 0.05 versus WT, **p < 0.01 versus WT, ***p < 0.001 versus WT, Mann–Whitney test. Scale bar, 1 μm.

Organization of the postsynaptic density is normal in dnl2 mutants. NMJ boutons were costained with anti-DPAK and anti-Fas II antibodies. In both the WT ( A ) and dnl2 mutants ( B–C ), DPAK clusters closely appose Fas II clusters. Although the number and size of DPAK clusters was altered in dnl2 mutants (Fig. 5), the DPAK clusters were still closely apposed to Fas II clusters in both dnl2^KO70 ( B ) and dnl2^KO70 /Df ( C ) mutants. Scale bar, 2 μm.

Ultrastructural analysis of type I synaptic bouton in dnl2 mutants. A–D , Representative transmission electron microscope micrographs showing low-magnification views of synaptic boutons, the presynaptic active zones (arrowheads), and postsynaptic SSR. E , Morphometric analysis of reconstructed synaptic boutons showed a significant increase in the number of T-bars per bouton in dnl2 mutants that was rescued by muscle specific expression of a UAS–dnl2 cDNA with 24B–Gal4. F , The length of PSDs beneath presynaptic T-bars was reduced in dnl2 mutants. This was also rescued by muscle expression of UAS–dnl2 with 24B–Gal4. G , The SSR was thinner in dnl2 mutants compared with WT controls or 24B–rescue flies. All images and analyses were derived from type Ib boutons on muscles 6/7. **p < 0.01 versus WT, ***p < 0.001 versus WT, ^## p < 0.01 versus dnl2^KO70 , ^### p < 0.001 versus dnl2^KO70 , Mann–Whitney test.

dnl2 mutants exhibit changes in DGluRs abundance at the NMJ. A , Subunit composition of two major glutamate receptor complexes at the NMJ: each receptor complex contains GluRIIA or GluRIIB alongside one of each of the other three subunits. B–D , Representative NMJs from WT, dnl2 null mutants (homozygous dnl2^KO70 and dnl2^KO70 /Df), dnl2--RNAi transgenic flies (UAS–dnl2 RNAi driven in muscles by 24B–Gal4), and two rescue strains expressing a UAS–dnl2 cDNA in the dnl2 mutant using either 24B–Gal4 or C57–Gal4 (both muscle drivers). NMJs were costained with anti-DGluRIII and anti-Dlg ( B ), anti-GluRIIA and anti-HRP ( C ), or anti-GluRIIB and anti-HRP ( D ). E–G′ , Summaries of normalized fluorescence intensities for DGluRIII ( E ), DGluRIIA ( F ), and DGluRIIB ( G ) compared with normalized fluorescence intensities for Dlg ( E′ ) or HRP ( F′ , G′ ) in the same boutons. Data are expressed as a percentage of WT fluorescence intensity. Our results show a significant reduction in the levels of DGluRIII ( E ) and GluRIIB ( G ), without any significant change Dlg ( E′ ) or HRP ( G′ ) in both dnl2 null and 24B–dnl2 RNAi transgenic flies compared with WT control. We also saw an increase in GluRIIA ( F ) in dnl2 mutants and 24B–dnl2 RNAi flies without any change in HRP ( F′ ). Expression of a UAS–dnl2 transgene in the muscles was able to partially rescue the defects in GluR expression. Specifically, the decrease in GluRIII expression was rescued by expression of the UAS–dnl2 in the muscle with either 24B–Gal4 or C57–Gal4. The decrease in GluRIIB expression was partially rescued using the C57–Gal4 driver but not using the 24B–Gal4 driver. The increase in GluRIIA expression, however, was not rescued with either of the muscle drivers. ***p < 0.001 versus WT, ^# p < 0.05 versus dnl2^KO70 , ^## p < 0.01 versus dnl2^KO70 , and ^### p < 0.001, Mann–Whitney test.