Presynaptic CK2β is essential for synapse stability. (A) A stable wild-type NMJ on muscle 4. (B) Loss of presynaptic CK2β caused severe synaptic retractions. (C) Knockdown of muscle CK2β did not impair synapse stability. (D–F) MARCM-based analyses of control and CK2β-null mutant motoneurons (arrows). MARCM motoneurons are marked by the expression of membrane-bound GFP (mCD8-GFP, green). (D) Stable control NMJ. (E and F) CK2β^26-2 mutant motoneurons. Synaptic retractions were evident at distal areas of the mutant nerve terminals, characterized by a loss of presynaptic Brp and fragmentation of the MARCM membrane marker opposite postsynaptic glutamate receptor clusters (arrows). Bars: (main panels) 10 µm; (enlarged panels below) 5 µm. (G) Quantification of synaptic retraction frequency. n.s., not significant. (H) Quantification of synaptic retraction severity (**, P ≤ 0.01; ***, P ≤ 0.001; n = 9–13 animals, muscles 1/9 and 2/10 for G and H). (I and J) Quantification of synaptic retraction frequency (I) and severity (J) of MARCM control and CK2β^26-2 mutant motoneurons (n = 17–33 animals). (K) Western blot analysis of CK2β, CK2α, and tubulin protein levels in larval brain extracts of the genotypes analyzed in G. Neuronal knockdown of CK2β led to a severe reduction in CK2β and a modest reduction in CK2α protein levels. CK2β and CK2α levels were rescued by neuronal coexpression of wild-type CK2β. Brain extracts of CK2β^P1/26-2 mutants showed slightly reduced CK2β and normal CK2α protein levels. CK2β^26-2 mutant animals rescued by ubiquitous expression of CK2β had wild type–like CK2α and CK2β protein levels. Error bars represent SEM.