The spindle pole-body localization of activated cytoplasmic dynein is cell cycle-dependent in Aspergillus nidulans

Abstract

Cytoplasmic dynein is a minus end-directed microtubule motor that can be activated by cargo adapters. In Aspergillus nidulans, overexpression of ΔC-HookA, the early endosomal adapter HookA missing its cargo-binding site, causes activated dynein to accumulate at septa and spindle pole bodies (SPBs) where the microtubule-organizing centers are located. Intriguingly, only some interphase nuclei show SPB signals of dynein. Here we present data demonstrating that localization of the activated dynein at SPBs is cell cycle-dependent: SPB dynein signals are seen to associate with nuclei at early G1 but disappear at about the G1-S boundary.

Introduction

Cytoplasmic dynein-1 (called dynein thereafter for simplicity) is a minus-end-directed microtubule motor that transports many cargoes. Dynein motility is activated by dynactin and cargo adapters that link dynein to the cargoes (Reck-Peterson et al., 2018). In filamentous fungi, dynein is required for the distribution of nuclei and many cargoes, including early endosomes, away from the hyphal tip (Steinberg et al., 2017; Xiang, 2018). The dynein-early endosome interaction is mediated by the dynactin complex as well as the early endosomal adapter Fhip-Hook-Fts (FHF) complex (Bielska et al., 2014; Yao et al., 2014; Zhang et al., 2014; Zhang et al., 2011). Within the FHF complex, the Hook protein (HookA in A. nidulans and Hok1 in Ustilago maydis) interacts with dynein-dynactin using the N-terminal domain while the C-terminal domain binds the early endosome cargo via Fts and Fhip proteins (Bielska et al., 2014; Yao et al., 2014; Zhang et al., 2014). In A. nidulans, overexpression of ΔC-HookA activates dynein in vivo (Qiu et al., 2019). While dynein normally accumulates at the microtubule plus ends near hyphal tips (Xiang, 2018), overexpression of ΔC-HookA via the gpdA promoter causes activated dynein to accumulate at septa and the spindle pole bodies (SPBs) (Qiu et al., 2019), where active microtubule-organizing centers (MTOCs) are located (Oakley et al., 1990; Zhang et al., 2017). Interestingly, only a fraction of nuclei showed SPB signals of dynein. Here we investigated whether dynein localization at the SPB is cell cycle-dependent.

Results and Discussion

For visualizing interphase nuclei and dynein’s SPB signals, we used a strain containing both the Nuclear Localization Signal (NLS)-DsRed fusion (Shen and Osmani, 2013; Suelmann et al., 1997) and GFP-labeled dynein heavy chain (Zhuang et al., 2007). In A. nidulans, nuclear pore complexes (NPCs) are partially disassembled during mitosis, and thus, nuclei in mitosis do not contain the NLS-DsRed signals (De Souza et al., 2004). Upon completion of mitosis and reassembly of the NPCs, NLS-DsRed is reimported into the G1 nuclei. This has allowed us to follow dynein’s SPB signals during the interphase-to-mitosis and the mitosis-to-interphase transitions. We found that dynein’s SPB signals are bright on the post-mitotic G1 nuclei (Figure 1A – E and Video 1). Some dynein signals could be observed even before NLS-DsRed reenters the nucleus (Video 1), reminiscent of wild-type dynein localization at the mitotic spindle poles during anaphase (Li et al., 2005). Importantly, the SPB dynein signals disappear during the interphase before the next mitosis (Figure 1A – E and Video 1). The disappearance of dynein signals happens even before any apparent increase in nuclear volume (Figure 1C for an example), suggesting that it most likely occurs at late G1 or S. In a previous study, the SPB localization of CdhA, an activator of the anaphase-promoting factor activator, disappears at a point corresponding to the beginning of the S phase (Edgerton-Morgan and Oakley, 2012). Given the documented length of G1 at 25°C (Edgerton-Morgan and Oakley, 2012), our current data (Figure 1C–E) suggest that the SPB dynein signals disappear at about the G1-S boundary.

Figure 1.

A and B. Time-lapse sequences (from Video 1) showing 3D-rendered surfaces representing NLS-DsRed-labelled nuclei in the gpdA-ΔC-hookA-S strain superimposed upon GFP-dynein fluorescence micrographs. (A) shows the first round of nuclear division and (B) shows the second. Rendered nuclei are color-coded such that daughter nuclei retain the color of their mother nuclei (either blue, cyan or magenta). C – E. Plots showing the volume of each blue- (C), cyan- (D) and magenta- (E) colored nucleus, based on NLS-DsRed fluorescence, over the course of a 400-minute time lapse experiment (Video 1). Green lines indicate the period of time in which GFP-dynein could be detected, either intermittently or continuously, at the spindle pole body (SPB) of that nucleus. (F) Time-lapse sequence showing the localization of GFP-dynein and NLS-dsRed in a wildtype hypha undergoing a single round of mitosis. Dashed white lines outline the edge of the hypha (G) Time-lapse sequence highlighting the organization of cytoplasmic and spindle microtubules (GFP-tubA) in a hypha undergoing a single round of mitosis. Fluorescence images represent maximum intensity projections of Z series acquired at 0.2 μm intervals spanning the entire depth of the hypha. Scale bar = 5 μm.

Video 1.

Download video file ^{(6MB, mp4)}

Time-lapse sequence of a gpdA-ΔC-hookA-S Aspergillus nidulans strain, expressing NLS-DsRed and GFP-dynein, undergoing two rounds of mitosis. GFP-dynein localizes at the SPBs of G1 nuclei but disappear during interphase prior to the next mitotic phase. Interphase nuclei are labelled with NLS-DsRed, which disperses from nuclei during mitosis. The sequence plays through twice, the second time with 3D-rendered nuclei, generated from the NLS-DsRed fluorescence signal, superimposed upon the GFP-dynein channel. The rendered nuclei are colored either blue, cyan or magenta, and their daughter nuclei retain these colors after nuclear division. Images represent maximum intensity projections of Z-series acquired at 0.2 μm intervals spanning the entire depth of the hypha and every 2 minutes. Video plays at a rate of 10 frames per second and the time stamp shows hours and minutes.

This cell cycle-dependent SPB accumulation of activated dynein has not been previously reported. In mammalian cells, dynein at the centrosome can be detected at S and G2 but not at G1 (Quintyne and Schroer, 2002). In wild-type A. nidulans, although the localization of dynein at the mitotic spindle poles could be detected at anaphase (Li et al., 2005), SPB dynein signals were never observed at interphase (Figure 1F and Video 2). Thus, the G1 SPB signals observed in this study are caused by ΔC-HookA-driven dynein activation. While activated dynein can move from the plus ends of cytoplasmic microtubules during G1, some SPB signals may also come from dynein moving on astral- or spindle microtubules during anaphase. In comparison, budding yeast dynein signals are visible at the SPBs during G1, pre-anaphase and anaphase (Sheeman et al., 2003), and although the SPB localization is enhanced by dynein’s cortical anchor Num1, it happens regardless of whether the cells are at G1, pre-anaphase or anaphase (Lammers and Markus, 2015; Sheeman et al., 2003).

Video 2.

Download video file ^{(1.8MB, mp4)}

Time lapse sequence of a “wildtype” Aspergillus nidulans hyphae, expressing NLS-DsRed and GFP-dynein, undergoing a para-synchronous wave of mitosis. GFP-dynein comets can be observed at the plus-ends of cytoplasmic microtubules localized in the hyphal tips but was undetected from SPBs throughout the cell cycle. Images represent maximum intensity projections of Z-series acquired at 0.2 μm intervals spanning the entire depth of the hypha and every 2 minutes. Video plays at a rate of 10 frames per second and the time stamp shows hours and minutes.

In A. nidulans, the disappearance of SPB accumulation of activated dynein after G1 is not due to a lack of cytoplasmic microtubules at later stages of interphase (Figure 1G). However, several other possibilities may explain this cell cycle stage-dependent accumulation. For example, it could be caused by a lack of microtubules terminating at the SPB beyond G1. Alternatively, changes in the SPB or in dynein-dynactin may occur during G1, which may reduce the binding affinity of dynein for the SPB. Or, dynein may be removed from microtubules at the SPB by a process that becomes active after G1. Thus, this intriguing phenomenon we observed in this study will lead to important future questions of general interest to the fungal biology field: Could the SPB MTOC be inactivated during a later stage of G1? Or, does the SPB or dynein-dynactin undergo any modification during late G1? These questions will need to be addressed in the future.

In this study, we have also observed nuclear behaviors that have not been documented previously. Overexpression of ΔC-HookA causes a partial clustering of nuclei observed in different hyphal regions. This seems to be caused by SPBs being pulled toward each other (Video 1), which may be due to activated dynein anchored at one SPB walking towards the minus ends of MTs from an adjacent SPB. It has been suggested that cargo adapter-activated mammalian dynein (within the dynein-dynactin-cargo adapter complex) accumulated and anchored at the minus end of one microtubule can walk along another microtubule, thereby clustering the minus ends of different microtubules together in vitro (Tan et al., 2018). Since Hook3 (HookA homolog) tends to form a dynein-dynactin-cargo-adapter complex with two dynein dimers with four motor heads (Urnavicius et al., 2018), it is possible that one dynein dimer could be anchored at one SPB while the motor heads of the other dimer could walk toward the other SPB. In A. niulans, post-mitotic G1 nuclei undergo distinctive transient microtubule-dependent back-and-forth movement, which is negatively regulated by Nup2, an NPC protein (Suresh et al., 2018). Future work will be needed to address whether Nup2 has any effect on the SPB localization of dynein.

Methods

A. nidulans strains

The A. nidulans strains used in this study are XX506 (gpdA-ΔC-hookA-S-AfpyrG; GFP-nudA^HC; ΔyA::NLS-DsRed; possibly pyrG89; possibly ΔnkuA::argB), XX514 (GFP-nudA; ΔyA::NLS-DsRed; possibly nkuA::argB) and XX516 (gpdA-ΔC-hookA-S-AfpyrG; ΔyA::NLS-DsRed; GFP-tubA; pyrG89; possibly nkuA::argB). Spores were stab inoculated onto glucose minimal media from a master culture using a disposable platic micropippete tip and grown for 16 h at 30 °C. Colonies were allowed to acclimate to room temperature (21–23°C) for 2–3 hours before being excised, along with a small block of agar media (~1 cm²), and inverted onto 35 mm glass-bottomed FluoroDishes (World Precison Instruments).

Live Cell Imaging and Image Analysis

Fluorescence imaging of A. nidulans hyphae was performed on a Nikon Ti-E Eclipse inverted epifluorescence microscope, equipped with a Perfect Focus System (Nikon) and a motorized Piezo stage, using a 100× 1.49 N.A. oil immersion Apo TIRF Nikon objective. GFP and DsRed were excited using an AURA II triggered illuminator with 485nm and 560nm LEDs, respectively, and detected using an iXon Ultra 897 electron multiplier CCD Camera (Andor Technology). All hardware was controlled by NIS-Elements Advanced Research (version 4.60). Live-cell imaging was performed at room temperature (21–23°C). Strains XX506 and XX514 were imaged using identical acquisition parameters (488nm LED at 15% intensity and 560 nm LED at 5% intensity, 80 ms exposure and EM gain set to 139). Two-color 3D time lapse data sets were deconvolved, with spherical aberration correction and background subtraction, using the “Automatic” 3D deconvolution option in NIS-Elements Advanced Research.

Three-dimensional rendering and volume analysis of NLS-DsRed-labelled nuclei was performed in Imaris 9.5.1 (Bitplane) using the “surfaces” function. Surfaces were added to individual nuclei, and their daughter nuclei, using the “Marching Cubes” tool and were tracked using the “manual tracking” function and “Auto-connect to selected Surface” option. All statistics generated in Imaris (Bitplane) were exported as Microsoft Excel files and volume data were plotted in Prism 8 (version 8.2.1; Graphpad).

Video editing and figure preparation

Video editing was performed in Adobe Premiere Pro 2020 version 14.0.0. Panels in Figure 1 were generated by extracting TIFs, representing maximum intensity projections, from 4D data sets using the “Snapshot” tool in Imaris (Bitplane). Images were orientated and cropped in Adobe Photoshop CC version 2017.1.4 and assembled in Illustrator CC 2018 22.1.

Highlights.

The SPB signals of activated dynein are observed during G1 but disappear at about the G1-S boundary.