Expression of NDR1, NDR2 and autophosphorylated NDR1/2 proteins in neurons. A. NDR1 and NDR2 proteins are present in the brain during development. Western blots of mouse brain lysates from Postnatal day (P)5, P10, P15, P20 probed by a mouse monoclonal antibody raised against NDR1 and an NDR2 specific polyclonal antibody we raised for this study. Anti-tubulin blot is shown as loading control. B. Top. Immunostaining with NDR1 antibody (green) shows endogeneous NDR1 in CA3 pyramidal cell layer (nuclei are labeled with DAPI shown in blue). Scale bar is 100 μm. Bottom. Immunnostaining with NDR1 antibody (green) labels dendrites and cytoplasm in CA3 hippocampus. Scale bar is 50 μm. C. Cultured hippocampal neurons stained against NDR1 or NDR2 antibodies described above co-stained with MAP2 (microtubule associated protein 2, a dendritic marker) showing NDR1 and NDR2 in dendrites and cytoplasm. Scale bar is 50 μm. D & E. NDR1 mutations used in this study. Red is loss of function, green is gain/ rescue of function, black is analog sensitive mutants. See also Figure S1.

NDR1/2's role on dendritic spine and excitatory postsynaptic development A. Dendritic spines of neurons transfected with NDR1 mutants or siRNA are shown, scale bar is 10 μm. B. Dendritic spine categories MS: mushroom spine, F: Filopodia, A: Aypical, St: Stubby. C. Effect of NDR1 dominant negative and constitutively active expression on different categories of dendritic spine densities. N of cells = 6, 9, 8 for each group, in order. D. Effect of NDR1 and NDR2 knockdown by siRNA on dendritic spines. N = 12, 16 & 7 for each groups, in order. E. Examples of whole-cell patch-clamp recordings of mEPSC from transfected hippocampal neurons. F. Comparison of frequency and G. amplitude of mEPSCs. N = 24, 15, 16, 11 & 8 for GFP, NDR1-KD, NDR1-CA, NDR1si NDR2si and NDR1si NDR2si NDR1* rescue, respectively. NDR1* is wild type NDR1 cDNA which lacks 3’ UTR and is therefore siRNA resistant. See also Figure S3.

Identification of NDR1's phosphorylation targets by chemical genetics. A. Depiction of ATP binding site of wild type Src kinase with ATP (green) (top) and as-Src with Benzyl-ATP-γ-S (yellow) (bottom). Mutation in gatekeeper residue (blue) resulted in an affinity pocket where bulky ATP analog binds. B. NDR1-as mutants (M166A and M166G) in NDR1-CA use Benzyl-ATP-γ-S and their efficiency is increased by two point mutations in the kinase domain M152L and S229A/T. HA tagged kinase was expressed and purified from COS-7 cells using HA tag. Kinase reaction was done using NDR substrate peptide. Thiophosphorylation was detected by anti-thiophosphate ester antibody. C. Structures of ATP and Benzyl-ATP-γ-S are shown. D. Covalent capture method for kinase substrate identification. A protein phosphorylated by endogeneous kinases is depicted in gray. A protein thiophosphorylated by NDR1-as is depicted in red. Blue depicts a protein which contains a cysteine. E. & F. Validation of AAK1 and Rabin8 phosphorylation sites by direct in vitro kinase assays. E. In vitro kinase assay confirmation of AAK1 S635 as the NDR1 specific phosphorylation site. In vitro kinase assays were performed by incubating the indicated NDR1-as-CA with purified wild type AAK1-HA or S635A AAK1-HA protein. Reaction was done using Benzyl-ATP-γ-S which is used by NDR1-as-CA and not AAK1 to prevent the phosphorylation signal caused by AAK1 autophosphorylation when regular ATP is used. Immunoblot with anti-thiophosphate ester specific antibody reveals S635 on AAK1 as the only NDR1 phosphorylation site on AAK1. F. Similar experiment as in E demonstrating Rabin8 as an NDR1 phosphorylation substrate protein. Rabin8 is phosphorylated by NDR1-as-CA in vitro and this phosphorylation is greatly diminished in Rabin8 S240A mutant indicating this site as the major phosphorylation site. When S/T 240-243 are all mutated to Ala (Rabin8-AAAA) all NDR1-as can no longer phosphorylate Rabin8 indicating that these residues may also be phosphorylated when S240 is mutated to Ala. G. Mass spectroscopy identification of AAK1 phosphorylation by HCD (higher energy C-trap dissociation) spectra analysis of AAK1 derived peptide containing phosphorylated S635. See also Figure S4 and Supplemental Table 1.

Rabin8 affects spine morphogenesis in dissociated hippocampal neurons. A. Endogeneous Rabin8 immunostaining in cultured hippocampal neurons at DIV10. Perinuclear Rabin8 (red) colocalizes with Golgi marker GM-130 (green). MAP2 depicts dendrites. Scale bar is 25 μM. B. Dendritic spine morphologies of control, Rabin8 mutants and Rabin8 siRNA expressing neurons. Arrows point to filopodia. Scale bar is 6 μm. C. Quantification of spine morphologies. N = 23, 13, 12 & 17, for GFP, Rabin8-AAAA, Rabin8-EEEE and Rabin8-si, respectively. D. In utero electroporation analysis of Rabin-AAAA co-expressed with GFP. E. Spine head diameter and F. Spine density analysis are shown for Rabin-AAAA expressing neurons in comparison to GFP alone (N = 7 & 11 for GFP and Rabin-AAAA, respectively, * p<0.05). G. Summary depicting NDR1/2's function on dendrite development and spine morphogenesis via two of its phosphorylation targets. See also Figure S6.

NDR1/2's role on dendrite development. A. Hippocampal neurons expressing NDR1 mutants or siRNA together with GFP. Scale bars are 100 μm. B. Sholl graphs of dendrites of neurons transfected with GFP alone or GFP co-transfected with NDR1 mutants. N of neurons = 21, 16, 18 & 11 for GFP, NDR1-KD, NDR1-AA & NDR1-CA, respectively. C. Sholl graphs of neurons expressing GFP plasmid or GFP plasmid which also expresses siRNA. For dual NDR1 and NDR2 siRNA knockdown NDR1 siRNA and NDR2 siRNA plasmids were co-transfected. For rescue with siRNA resistant NDR1 (NDR1*), this plasmid was co-transfected with NDR1si and NDR2si. N = 14, 13, & 9 for GFP, NDR1si NDR2si and NDR1si NDR2si rescue, in order. D & E. Total dendrite branch point analysis for NDR1 mutant expression and siRNA experiments, respectively. F & G Total dendrite length analysis for NDR1 mutant and siRNA experiments, respectively. (N = 10 & 10 for NDR1si & NDR2si.) H. Sholl analysis for chemical genetics inhibition of analog sensitive NDR1-as by 1-Na-PP1. DMSO (solvent) was used as control. N = 7 for each group. I. Total branch points and J. Total length analysis for chemical genetic NDR1-as inhibition experiment. * p < 0.05, ** p < 0.01, *** p < 0.001 in all graphs in all figures assessed by Kruskal Wallis non-parametric test followed by dual test with Dunn's method in comparison with GFP control (unless otherwise indicated). Error bars are standard error of the mean in all graphs. Stars on Sholl graphs statistical comparisons with Kruskal Wallis followed by Dunn's method at 50 μm or 340 μm distance from the soma. See also Figure S2.

In vivo analysis of NDR1 mutants and siRNA by in utero electroporation. A. Projected z-stacks of GFP and NDR1 mutants or siRNA expressing layer 2/3 neurons are shown. Scale bars are 75 μm except NDR1-CA scale bar is 50 μm. B. Drawings of neurons in A. C. Representative images of dendritic spines on labeled layer 2/3 pyramidal neuron basal dendrites. Scale bar is 5 μm. D. Sholl analysis of dendrites of layer 2/3 neurons. Analysis was done for first 150 μm distance from the soma, focusing on basal dendrites and apical oblique dendrites proximal to the soma. N = 15, 10, 6, 9 & 16 for GFP, NDR1-KD, NDR1-CA, Control siRNA & NDR1si NDR2si, respectively. E. Comparison of dendritic branch crossings via sholl analysis at 50 μm from the soma. F. Total dendrite length comparison between groups including dendrites 150 μm from the soma. G. & H. Spine analysis. Spine head diameter (G) and spine density comparison between groups (H). N = 7, 12, 8, 7 & 7 for GFP, NDR1-KD, NDR1-CA, Control-si and NDR1si NDR2si, respectively. See also Figure S3.

AAK1 affects dendrite branching and length in dissociated hippocampal neurons. A. Neurons expressing GFP alone, GFP plus AAK1 mutants (AAK1-KD, AAK1-SA or AAK1-SD) and AAK1siRNA. Scale bar is 100 μm. B. Dendrite branching statistics are done via Sholl analysis at 50 μm distance from the soma. N = 27, 24, 17, 14, 23 & 13 for GFP, AAK1-KD, AAK1-SA, AAK1-SD and AAK1-si, and AAK1si + AAK1 siResistant, respectively. C. Total number of dendrite branches and D. Total dendrite length comparisons. E-H. Experiments showing epistasis between NDR1 and AAK1 in hippocampal neurons. Neurons were transfected with GFP + HA, NDR1si NDR2si +HA, NDR1si NDR2si + AAK1-SD-HA, NDR1-CA-myc + GFP and NDR1-CA-myc + AAK1si (N = 32, 31, 29, 13 & 16, respectively) to test epistasis. E. Sholl analysis of dendrites, F. Number of branch crossings at 40 μm distance from the soma for sholl analysis in E., G. Total number of branch points and H. total dendrite length of all neurons are shown. I. & J In utero electroporation of AAK1 siRNA leads to increased proximal branching. I. A layer 2/3 neuron expressing AAK1 siRNA and its drawing is shown. J. Sholl analysis of AAK1 siRNA expressing neurons in comparison to Control siRNA. Dendrite branching is increased at 40 μm (N = 9 for each group, ** p < 0.01). See also Figure S5.