arl8 is necessary for the rapid remodeling of AZ structure during homeostatic signaling and for the chronic expression of PHP. (A) Representative images of NMJs immunostained with anti-BRP, anti-vGlut, and anti-HRP of wild type and arl8 mutants at baseline and following PhTx application, as well as GluRIIA mutants (w;GluRIIA^sp16) and GluRIIA;arl8 double mutants (w;GluRIIA^sp16;arl8^e00336). (B) Quantification of BRP and vGlut puncta sum intensity in the indicated conditions normalized to baseline values (no PhTx application for wild type and arl8 mutants, wild type for GluRIIA mutants, arl8 for GluRIIA;arl8 mutants). Note that while both BRP and vGlut intensity is enhanced following PhTx application to wild type and in GluRIIA mutants, no change in intensity is observed in arl8+PhTx nor in GluRIIA;arl8 mutants compared with their baseline values. (C) Quantification of total sum intensity of BRP and vGlut per NMJ demonstrates a failure to increase AZ and SV components at presynaptic terminals in arl8+PhTx and GluRIIA;arl8 double mutants. (D) Schematic and representative traces of recordings from the indicated genotypes and conditions. Note that while mEPSP amplitudes are reduced following PhTx application or by loss of GluRIIA, as expected, rapid PHP is robustly expressed following acute application of PhTx. However, PHP fails to be expressed over chronic timescales in GluRIIA;arl8 double mutants. (E) Quantification of EPSP amplitude in the indicated genotypes. (F) mEPSP and quantal content values in the indicated genotypes normalized to baseline values demonstrating that while PHP can be rapidly induced by PhTx application to wild-type and arl8 NMJs, PHP fails to be chronically expressed in GluRIIA;arl8. Error bars indicate ± SEM (n ≥ 7; t test; Table S1). **, P < 0.01; ****, P < 0.0001; ns, not significant.