CK2 controls synaptic stability via Ank2-L. (A) A stable wild-type NMJ on muscles 6/7. (B) The hypomorphic ank2^P2 mutation causes synaptic retractions (arrows). (C) Expression of wild-type CK2α in ank2^P2/P2 mutants significantly alleviated the synaptic retraction phenotype. (D) Expression of kinase-dead CK2α in ank2^P2/P2 mutants further enhanced the synaptic retraction phenotype. Bars, 10 µm. (E and F) Quantification of synaptic retraction frequency (E) and severity (F) for the genotypes described in A–D. (G and H) Quantification of synaptic retraction frequency (G) and severity (H) of second instar ank2^P1/P1-null mutant animals and ank2^P1/P1-null mutant animals expressing wild-type or kinase-dead CK2α. The ank2^P1/P1 phenotype was not significantly altered by CK2α expression. n.s., not significant. (I and J) Analysis of the genetic interaction between CK2α and ank2-L. Removal of one copy of ank2-L in CK2α^P1/P1 mutant animals or removal of one copy of CK2α in ank2^P2/P2 mutant animals significantly increased the frequency (I) and severity (J) of synaptic retractions compared with CK2α^P1/P1 and ank2^P2/P2 mutants, respectively (**, P ≤ 0.01; ***, P ≤ 0.001; n = 12–15 animals for E and F). (K) In vitro phosphorylation assays. Human CK2 phosphorylates purified Ank2-L protein domains (Ank2-L^1682–2089 = Ank2^MT; Ank2-L^3134–3728 = Ank2-L^C). Phosphorylation was inhibited in the presence of the CK2-specific inhibitor TBB. (L) Analysis of CK2-dependent phosphorylation of Ank2-L domains by mass spectrometry identified phosphorylated serine/threonine residues (red, underlined) in the Ank2^MT and Ank2-L^C domain. Examples of phosphorylation sites corresponding to the CK2 consensus motifs S/TXXE/D or S/TE/D are depicted. Acid residues are marked in blue. Error bars represent SEM.