Visualization of dynein-dependent microtubule gliding at the cell cortex: implications for spindle positioning

Abstract

Dynein motors move along the microtubule (MT) lattice in a processive "walking" manner. In the one-cell Caenorhabditis elegans embryo, dynein is required for spindle-pulling forces during mitosis. Posteriorly directed spindle-pulling forces are higher than anteriorly directed forces, and this imbalance results in posterior spindle displacement during anaphase and an asymmetric division. To address how dynein could be asymmetrically activated to achieve posterior spindle displacement, we developed an assay to measure dynein's activity on individual MTs at the embryo cortex. Our study reveals that cortical dynein motors maintain a basal level of activity that propels MTs along the cortex, even under experimental conditions that drastically reduce anaphase spindle forces. This suggests that dynein-based MT gliding is not sufficient for anaphase spindle-pulling force. Instead, we find that this form of dynein activity is most prominent during spindle centering in early prophase. We propose a model whereby different dynein-MT interactions are used for specific spindle-positioning tasks in the one-cell embryo.

Figure 1.

EVA reveals a dynein-dependent EB1 velocity population at the cortex. (A) EB1 particles are tracked at the cortex of embryos ectopically expressing the MT-severing enzyme MEI-1 (red circles). Bar, 10 µm. (B) MTs conceivably interact with cortical force generators in an end-on or lateral manner. (C and D) The distribution of EB1 velocities >1.0 µm/s in control L4440(RNAi) mei-1(ct46) and dhc-1(RNAi) mei-1(ct46) embryos. The insets are the entire velocity distribution; the boxed areas within the inset are enlarged. The dotted line is the computationally derived Gaussian population with a mean >1.0 µm/s. n = number of cortical EB1 trajectories. Histograms and population analysis used pooled data (four embryos from one L4440(RNAi) trial; 14 embryos from three dhc-1(RNAi) trials).

Figure 2.

EVA analysis in various RNAi treatments. (A–H) The distribution of EB1 velocities >1.0 µm/s in RNAi-treated mei-1(ct46) embryos. The insets are the entire velocity distribution; the boxed areas within the insets are enlarged. The dotted lines are the computationally derived Gaussian population with a mean >1.0 µm/s. n = number of cortical EB1 trajectories. Each histogram and population analysis used pooled data (at least five embryos from at least two independent RNAi trials).

Figure 3.

Relative proportions and mean velocities of the cortical ddMTV populations. (A) The proportion of EB1 velocities in the ddMTV population are shown. All RNAi was performed in a mei-1(ct46) background. dhc-1(RNAi) and lin-5(RNAi) subpopulations >1.0 µm/s were not detected by the computer algorithm and were estimated (Materials and methods). (B) Mean velocity of the ddMTV subpopulation in micrometers per second. For dhc-1(RNAi) and lin-5(RNAi), the mean velocity was not measurable (Materials and methods). Error bars represent SEM. par-2(RNAi), P = 6.5E−4; par-3(RNAi), P = 0.44; goa-1/gpa-16(RNAi), P = 4.6E−3; gpb-1(RNAi), P = 0.11; gpr-1/2(RNAi), P = 0.85; lis-1(RNAi), P = 5.0E−4; nmy-2(RNAi), P = 0.012; dnc-1(RNAi), P = 1.5E−4). The double asterisks indicate significance at α = 0.01. Population analysis used pooled data (at least four embryos from at least two independent RNAi trials, except L4440(RNAi) and dnc-1(RNAi), which were from one RNAi trial).

Figure 4.

dnc-1(RNAi) reduces cortical MT gliding but not centrosome separation. (A) Cortical ddMTV proportions throughout the wild-type (WT) cell cycle. In the one-cell embryo, the proportion of ddMTVs increased during pronuclear migration/centration compared with anaphase (29.5%, 69/234; P = 2.1E−6). mei-1(ct46) embryos were used for pronuclear migration/centration and slow phase stages. L4440(RNAi) mei-1(ct46) embryos from Fig. 1 C were used for anaphase. (B) dnc-1(RNAi) reduces the proportion of ddMTVs during pronuclear migration/centration (5.3%, 21/397; P = 3.2E−4) and anaphase (6.1%, 63/1,040; P = 1.5E−4) compared with control. Significance at α = 0.01 in A and B is indicated by double asterisks. Data in A and B were derived from at least four embryos pooled before population analysis from single RNAi trials (except pronuclear migration/centration dnc-1(RNAi) from two independent RNAi trials). (C) Both dnc-1(RNAi) and dhc-1(RNAi) embryos exhibit defects in centration (Video 10), but only dhc-1(RNAi) significantly reduced centrosome separation rates in anaphase. The box represents 53 ± 2 s before cytokinesis furrow ingression (t = 0). n = 4 embryos. (D) Mean rates of centrosome separation for the interval boxed in C are shown. SEM at 95% confidence is shown.

Figure 5.

A model for two functionally distinct populations of dynein at the embryo cortex. DHC-1, LIN-5, and DNC-1 are necessary for MT gliding behavior and generate force by pulling on MTs that contact the cortex at shallow angles (θ1), contributing to pronuclear centration (left; blue text). DHC-1, LIN-5, and members of the heterotrimeric G protein pathway are necessary for normal anaphase spindle-pulling forces (right), but many of these components did not significantly alter MT gliding behavior (black text). Because most MTs contact the cortex at orthogonal angles in the one-cell metaphase/anaphase embryo (θ2), anaphase force likely occurs primarily through end-on interactions (right). The two mechanisms proposed are shown to reflect relative contributions and are not necessarily mutually exclusive.