Dynein recruitment to nuclear pores activates apical nuclear migration and mitotic entry in brain progenitor cells

Abstract

Radial glial progenitors (RGPs) are elongated epithelial cells that give rise to neurons, glia, and adult stem cells during brain development. RGP nuclei migrate basally during G1, apically using cytoplasmic dynein during G2, and undergo mitosis at the ventricular surface. By live imaging of in utero electroporated rat brain, we find that two distinct G2-specific mechanisms for dynein nuclear pore recruitment are essential for apical nuclear migration. The "RanBP2-BicD2" and "Nup133-CENP-F" pathways act sequentially, with Nup133 or CENP-F RNAi arresting nuclei close to the ventricular surface in a premitotic state. Forced targeting of dynein to the nuclear envelope rescues nuclear migration and cell-cycle progression, demonstrating that apical nuclear migration is not simply correlated with cell-cycle progression from G2 to mitosis, but rather, is a required event. These results reveal that cell-cycle control of apical nuclear migration occurs by motor protein recruitment and identify a role for nucleus- and centrosome-associated forces in mitotic entry. PAPERCLIP:

Figure 1. Mechanisms for cytoplasmic dynein recruitment to the nuclear envelope

A. Diagram representing G2-specific NE dynein recruitment mechanisms via nucleoporins Nup133 and RanBP2. Dynein is also shown linked to the NE by SUN-nesprin complexes, a mechanism not known to be cell cycle regulated. B. Triple staining with anti-dynein, anti-Cyclin B1, and DAPI (DNA) showing NE dynein localization specifically in cyclin B1 expressing HeLa cells. C. 3D-Structured Illumination Microscopy (3D-SIM) revealing the association of BicD2, dynein, and dynactin with HeLa cell nuclear pores, marked using Mab414 and anti-RanGap1. D. HeLa cells were double-labeled with anti-BicD2 and anti-CENP-F antibodies to test for temporal overlap between the two cell cycle-dependent NE dynein recruitment mechanisms. All cells exhibiting CENP-F-positive NEs were also positive for BicD2, but only a fraction of BicD2-positive cells showed NE CENP-F staining. E. NE labeling in the ventricular zone (VZ) in E19 rat brain sections. Top row, NE dynein staining was seen in a subset of RGP cells (yellow arrowhead), but absent in others (blue asterisks). RGP cells were identified by Pax6 immunostaining. Bottom, dynein and BicD2 colocalize at the NE, with many BicD2-positive cells also positive for dynein. Confocal microscopy was used throughout unless otherwise stated. See also Figure S1, S2, and S4A.

Figure 2. Inhibition of dynein NE recruitment mechanisms affects overall neuronal migration in embryonic rat brain

A. E16 rat embryonic brains were subjected to in utero electroporation with the pRNAT vector expressing shRNAs corresponding to the genes noted along with a fluorescent reporter, or with an RFP-KASH construct (n=3 brains per condition). Brain tissue was fixed and sectioned at E20. Expression of BicD2, Nup133, or CENP-F shRNAs resulted in a marked reduction in distribution of electroporated cells throughout the intermediate zone (IZ) and cortical plate (CP), comparable to the effects of LIS1 shRNA. KASH expression resulted in an intermediate cell redistribution phenotype. No clear effect was observed for BicD1. Scale bar = 50 μm. B. Quantification of transfected cells within the VZ, subventricular zone (SVZ), and IZ+CP show an increase in Nup133, CENP-F, and BicD2 shRNA-expressing cells in the SVZ, and a decrease in the IZ+CP. C. Quantification of Pax6+ (RGP) and NeuN+ (neuronal) cells in brain electroporated with BicD2, Nup133, or CENP-F shRNA shows an increase in RPG cells and a decrease in neurons. D. Quantification of NeuN+ neurons that exhibited a migratory, bipolar morphology supports a loss in migrating neurons. *P<0.05; **P<0.01; ***P<0.001; N.S., Not Significant; Error bars = S.D. (throughout legends). See also Figure S3, S4B, and S5A.

Figure 3. RNAi for Dynein NE Recruitment Factors Inhibits Apical nuclear migration

A. E16 rat embryonic brains were subjected to in utero electroporation to express shRNAs corresponding to BicD2, Nup133, or CENP-F. Brain slices were placed into culture at E20 for live imaging over an 8–15 hr period. Control RGP cell is shown undergoing apical nuclear migration to the ventricular surface of the brain slice, followed by mitosis and basal migration. BicD2, Nup133, and CENP-F shRNAs each caused nuclear arrest. Model of INM depicted on right. Scale bar = 5 μm. B. Tracings of nuclei in Nup133, CENP-F, or BicD2 shRNA-expressing RGP cells show severe impairment of apical migration, with nuclei in BicD2 shRNA cells arresting further from the ventricular surface. C. No clear effect on basal nuclear migration was observed. Velocity is net distance/time. See also Figure S5 and Movie S1–S7.

Figure 4. RNAi-induced arrest of nuclear migration early vs. late during apical INM

To test for general effects of dynein NE recruitment genes on nuclear position, shRNAs for BicD2, Nup133, and CENP-F were expressed in E16 rat embryonic brains, which were then fixed and imaged at E20 (n=3 brains per condition). A. Distribution of RGP nuclei. Control nuclei were distributed at a range of distances from the ventricular surface. Nuclei of BicD2 RNAi expressing cells accumulated relatively far from the ventricular surface. In contrast, nuclei of Nup133 and CENP-F RNAi expressing cells accumulated close to this site. Double knockdown of BicD2 and Nup133 resulted in a nuclear distribution similar to that for BicD2 RNAi alone. Scale bar = 5 μm for all panels. B. Quantification of nuclear distances. RGP cells were identified morphologically and by Pax6 staining, and distance was measured from the ventricular surface to the closest (bottom) edge of the nucleus. Nup133 and CENP-F RNAi caused a marked accumulation of nuclei close to the ventricular surface, which peaked at ~5 μm (panel C).

Figure 5. Nuclei in apically arrested RGP cells fail to enter mitosis

E16 rat embryonic brains were electroporated with vectors expressing shRNAs for Nup133, CENP-F, or BicD2. Sections were stained at E20 with cell cycle markers and scored for the percent of positive RGP nuclei (n=3 brains per condition). A–C. Nup133 and CENP-F RNAi dramatically reduced the percentage of nuclei positive for the G1 marker Cyclin D1 or for the S-phase marker BrdU (following 15 min pulse labeling). D–E. Substantial decreases in the fraction of anti-phosphohistone H3 (PH3) positive nuclei were observed under all RNAi conditions, with an almost completely loss resulting from Nup133 RNAi. Scale bars = 5 μm for all panels.

Figure 6. Pre-mitotic centrosome dynamics

E16 rat embryonic brains were co-transfected by in utero electroporation with cDNAs encoding shRNAs along with DsRed-centrin2 to image centrosomes. Brain slices were placed in culture at E20 for live imaging of RGP cells. Time-lapse images were generated at time intervals indicated at top in min. A. Centrosomes expressing control shRNA vector were retained at the ventricular surface throughout INM until 2:00–2:20 min, at which time the centrosome (arrowhead) departs to meet the soma. Following contact, the centrosome and soma migrate together to the ventricular surface (2:40) and the cell continues through INM (3:20). B. Time lapse recordings of Nup133 shRNA co-electroporated with centrin2 (top) and with centrin2 plus lamin A (bottom). In 83% of Nup133 depleted cells, the centrosome remained at the ventricular surface throughout the recording. In 17% of cases (bottom), centrosomes jumped towards the nucleus as in control cells, but then returned alone to the ventricular surface. The nuclei in both Nup133 examples showed only limited mobility throughout the recording. The lower example shows that the NE remained intact. C. CENP-F shRNA resulted in similar nuclear and centrosome arrest to that seen in the majority of Nup133 RNAi cases. D. Quantification of centrosome behavior shows departure from ventricular surface in all control cells, but in few to no Nup133 and CENP-F shRNA-expressing cells. E. Model for apical INM. Dynein is shown sequentially recruited to RGP cell nuclear pores via the BicD2 and Nup133 pathways. Arrows represent BicD2- (red) and Nup133- (yellow) mediated dynein pulling forces exerted on the NE. The centrosome (green) remains at the ventricular surface of the RGP cell throughout INM, but then moves towards the nucleus just prior to mitosis. Scale bars = 5 μm for all panels. See also Figure S7A–S7B.

Figure 7. Rescue of apically arrested nuclei by targeting of NE dynein

E16 rat embryonic brains were transfected with GFP-BicD2-N-KASH as a general NE dynein targeting approach. Alternatively, full length BicD2 was expressed to enhance dynein targeting to apically migrating RGP nuclei. Brain slices were fixed and sectioned at E20 (n = 3 brains per condition). A. Nuclei of cells expressing GFP-BicD2-N-KASH alone or in combination with BicD2 shRNA showed marked accumulation at the ventricular surface. Note that GFP-BicD2-N-KASH decorates the NE. B. Quantification of transfected RGP nuclei located at the ventricular surface. C. Sections were also stained for PH3 and percentage of PH3+ cells is compared with control and BicD2 RNAi conditions (from Figure 5E). D. RGP cells expressing full length BicD2 alone or in combination with CENP-F or Nup133 shRNA. E. Quantification of nuclear distances from the ventricular surface to the bottom of the nucleus. Top, nuclear distances of cells expressing either full length BicD2, CENP-F shRNA (from Figure 4B), both. Bottom, distances of nuclear located ≤ 10 μm from ventricular surface in cells expressing Nup133 shRNA alone (from Figure 4C) or in combination with full length BicD2. F. Quantification of PH3+ cells among RGP cells transfected with full length BicD2 or, in combination with, Nup133 or CENP-F shRNA. Percentages of PH3+ cells are compared to Nup133 and CENP-F RNAi conditions (from Figure 5E). G. Live recording and tracings of RGP cells co-expressing full length BicD2 and CENP-F shRNA undergoing apical nuclear migration completely to the ventricular surface. Scale bar = 5 μm. See also Figure S4C, S6, and S7C–S7D.