Fig 2: Identification of a dynein binding domain in the N terminus of Nudel. A, schematic diagram of Xenopus Nudel with deletion series from either N- or C terminus is shown. DBD, dynein binding domain. LBD, Lis1 binding domain. All proteins were tagged by GST (black diamonds). B, the purified GST-tagged Nudel fragments and their ability to interact with either dynein or Lis1 in Xenopus egg extracts is shown. Purified GST was used as controls. The minimal dynein binding region was mapped to the first 80 amino acids (1–80 as indicated by a red dashed square). As expected the minimal Lis1 binding domain is mapped to amino acids 100–153 of Nudel.

Fig 3: Nudel functions to stimulate spindle pole focusing through Lis1. A, immuno-depletion of Nudel from egg extracts is shown. One round of immuno-depletion is sufficient to remove most of Nudel/NudE in Xenopus egg extracts, whereas two rounds of immuno-depletions (?Lis1.1 and ?Lis1.2) routinely remove ~90% of Lis1 in egg extracts. Neither Nudel/NudE nor Lis1 depletion reduced the level of dynein as judged by Western blotting probing for the DIC. Control IgG-depleted egg extracts (Mock) were loaded at the indicated amount to assess the degree of immuno-depletion. B, spindle pole defects caused by Nudel/NudE depletion is fully rescued by adding purified Nudel protein. Immuno-depletion of Nudel resulted in the formation of mostly fence-like microtubules aligned perpendicularly across the sperm chromatin and a small fraction of spindles with multiple or partially focused poles (see the images in C), which can be fully rescued by addition of 50 nm concentrations of purified His-Nudel full-length protein. C, shown are representative images of spindles that are defined as bipolar, multipolar, partially focused, or parallel array of microtubules appearing as fence-like structure (Fence) perpendicular to the condensed chromatin. D, depletion of Lis1 caused a similar spindle pole defects as Nudel/NudE depletion. These defects could be rescued by adding 1 µm purified His-Lis1 but not by 1 µm His-Nudel. E, purified His-Lis1 fully rescues Nudel/NudE depletion. The addition of increasing concentrations of purified His-Lis1 shows that equal or greater than 0.8 µm His-Lis1 (the endogenous Lis1 is estimated at ~1 µm, see supplemental Fig. S1) is sufficient to fully rescue the spindle pole defects caused by Nudel/NudE depletion. Error bar, S.E. Scale bars, 20 µm.

Fig 4: Reduced level of Lis1 specifically activates RhoC in brain. (A) Abnormal increase in RhoC NDU in Lis1fl/- cortex. Ubiquitin binding domain pull down from Lis1fl/+ or Lis1fl/- brain is increased in mono-, di- and tri-ubiquitinated RhoC when Lis1 protein is reduced. Graph quantifies results from n = 5 biologically independent experiments, **P = 0.0037. Data are shown as the mean ± s.d. (B) Elevated RhoC NDU disrupts its interaction of total RhoC with RhoGDIa in vivo. RhoGDIa immunoprecipitation from Lis1fl/+ or Lis1fl/- brain shows decreased RhoC binding, while RhoA and Rac3 remain unchanged. Graph quantifies results from n = 3 biologically independent experiments done in duplicate, *P < 0.0312, ns = not significant with P = 0.4375 (RhoA) and P < 0.9999 (Rac3). Data are shown as the mean ± s.d. (C) RhoC, but not RhoA knock down by shRNAi rescues cell spreading impairment in Lis1fl/- neonatal astrocytes. Immunostaining of Lis1fl/+ or Lis1fl/- astrocytes transduced with a GFP-expressing lentivirus to knock down RhoA or RhoC for tGFP (anti-tGFP, orange). Golgi apparatus (GM-130, white) and DNA (DAPI, blue). GFP staining is used to define cell area (shown in the black and white mask of transduced cells) measured in arbitrary units (AU) in Image J. Control = scrambled shRNA transduced cells. Scale bar, 10 µm. Graph represents cell area measurements from three independent experiments. Data are presented as individual data points from all three biological replicates in dark gray (Lis1fl/+) or light gray (Lis1fl/–) circles and their mean ± s.d. as triangles color coded by experiment. *P < 0.05, **P < 0.01, ***P < 0.001. Where indicated, statistical significances were obtained using a two-tailed paired t-test for amount of ubiquitinated RhoC (A) and cell spreading assay (C), or non-parametric, two-tailed Wilcoxon matched-rank signed test for the RhoGDIa IP (B).

Fig 5: Lis1 binds to LNX1 and inhibits RhoC monoubiquitination. (A) Lis1 co-immunoprecipitates with LNX1. Anti-FLAG pull-down from HEK293T cells co-transfected with pEGFP-LIS1 and pPafah1b3-MycFLAG (positive control). EGFP-LIS1 co-immunoprecipitated with pLNX1-MycFLAG, but not pFLAG-ARHGDIA or pFLAG-RHOC. EGFP alone (negative control) does not co-immunoprecipitate with LNX1-MycFLAG. Therefore, we confirm Lis1-Pafah1b3 binding, identify novel Lis1-LNX1 interaction and show lack of direct interaction between Lis1 and RhoGDIa or RhoC. (B) Lis1 inhibits LNX1 mediated RhoC monoubiquitination in a dose-dependent manner. HA co-immunoprecipitation from HEK293T cells co-transfected with pcDNA3.1(+)-3xHA-RHOC, pLNX1-MycFLAG and varying amounts of pEGFP-LIS1(WT). Graph quantifies results from three independent experiments. Data are expressed as % control, where control is set at 100% and mean values are presented. P values were obtained using a two-tailed paired t-test.