To investigate if anaphase spindle midzone affects interphase progression in daughter cells, we cut NRK52E cells at early anaphase, as soon as sister chromosomes have separated completely, with a fine glass fiber to form daughter cells with or without spindle midzone structures. Staining of severed daughter cells with antibodies against aurora B (Figure ), which is known to relocate from centromeres to the spindle midzone during early anaphase [12
], confirmed that the spindle midzone was removed from one of the daughter cells. However, both daughter cells showed typical interphase microtubules organization (Figure ), and normal cellular and nuclear morphology (Figure ), indicating that cells recovered fully from the microsurgery and exited mitosis successfully.
Figure 1 Microsurgery of dividing NRK cells. The site of severing is indicated by dotted lines (A, a; B, a). Immunofluorescence indicates that aurora B, a protein associated with midzone microtubules, is partitioned predominantly to the cell with midzone microtubules (more ...)
Figure 2 Inhibition of the interphase progression following surgical removal of the spindle midzone structures. An NRK cell was cut at anaphase to form daughter cells with (top) and without (bottom) the spindle midzone structures (cutting site indicated by the (more ...)
A total of eight manipulated cells were followed by extended time-lapse imaging through the following interphase. All the daughter cell with spindle midzone showed cytokinesis-like cortical contraction activities, while the daughter cell without spindle midzone showed no cortical contraction (Figure , see Additional file 1
). The daughter cell with spindle midzone subsequently progressed with normal timing through the following interphase, entering mitosis at a time similar to that of adjacent control cells (Figure , Table ). However, in 5 out of 8 cases, the cell without midzone showed a duration of interphase more than twice that of the sister cell with spindle midzone (Table ). In the example shown in Figure , interphase lasted for 22 h, as compared to the normal 11 h for the sister cell with the spindle midzone. In two cases, cell cycle appeared to be arrested in interphase since no mitosis was observed over a period of 34 – 46 h.
Duration of interphase for daughter cells with and without the spindle midzone or midbody#
The inhibition of interphase progression was affected by the position of severing. When cells were severed near the equator to divide midzone structures between the two daughter cells, both cells showed cytokinesis-like cortical contractions and no effect on the progression of interphase. This observation was repeated with 4 cells. To test the possibility that the duration of interphase was affected by the size of the daughter cell, we took pairs of normally divided daughter cells and cut away a lateral region near the pole from one of the daughter cells. In all 5 cases, both daughter cells entered the subsequent mitosis with a timing similar to that of neighboring unmanipulated cells (Figure ). Together, these results indicate that the progression of interphase was affected directly or indirectly by the presence of the anaphase spindle midzone.
Figure 3 Normal interphase progression following manipulation of the size of a daughter cell. A lateral region near the pole was cut away from one of the daughter cells at the end of cytokinesis (cutting site indicated by the dotted line, a). Subsequent time lapse (more ...)
To test if cortical contraction is the primary determining factor of the rate of interphase progression, we treated cells with cytochalasin D at early anaphase for 10 min then severed them at various positions in the spindle midzone, at a time when midbody started to form and ingression started to appear in control cells. Cytochalasin D was removed 20 min later to ensure that there was no lingering cytokinesis activity upon removal of the drug [[14
] see Methods]. In all 5 cases, both daughter cells entered the subsequent mitosis with a timing similar to that of neighboring cells, despite the complete inhibition of cortical contraction (Table ). In addition, unmanipulated, binucleated cells, which failed cytokinesis spontaneously, also entered subsequent mitosis with a similar timing. Thus, the present observation is distinct from the "tetraploidy checkpoint", which was identified with dihydrocytochalasin B-treated REF52 cells [11
], but remained controversial with regard to its universal existence [15
Duration of interpahse for cells severed in the presence of cytoskeletal inhibitors#
We then asked if midzone microtubules, or proteins associated with the anaphase spindle midzone, are involved in interphase progression. Treatment of cells with nocodazole for 6–13 min at early anaphase before severing, with or without additional incubation with the drug, caused a significant delay (~4.5 h; p = 0.002) in the progression of interphase for both daughter cells, as compared to neighboring cells (Table ). Aurora B in these cells showed a complete dispersal from the midzone (Figure ) [12
], with a similar amount distributed in the two severed cells. These observations suggest the midzone microtubules may provide a scaffold for the activation/deactivation of some components to allow normal progression of the cell cycle. We suspect that the weaker effect compared to that caused by severing may be due to the known resistance of some midzone microtubules to nocodazole [6
]. Alternatively, a brief association of the components with the midzone microtubules before and during nocodazole treatment may be sufficient for partial activation/deactivation.
Figure 4 Presence of aurora B in both daughter cells following microsurgery of nocodazole-treated cells. An NRK cell at early anaphase was treated with 10 μM nocodazole for ~10 min before microsurgical cut between segregated chromosomes (cutting site indicated (more ...)
We also tested if telophase midbody is required for interphase progression. Cells were severed at mid- or late cytokinesis to form daughter cells with and without the midbody (Figure , see Additional file 2
). Using cells expressing aurora B-GFP, which is known to associate with the midbody [12
], we confirmed that the midbody was completely segregated from one of the daughter cells (Figure ). All nine manipulated pairs showed a normal progression of interphase indistinguishable from that of unmanipulated control cells, irrespective of the presence of the midbody (Table ). We conclude that the presence or absence of telophase midbody no longer affects the progression of subsequent interphase, despite the similarity of molecular components to those in the earlier spindle midzone. Thus, most likely it is transient catalytic reactions in the anaphase midzone that are crucial for the progression of cell cycle.
Figure 5 Normal interphase progression following removal of the midbody. A cell was cut at the end of cytokinesis to form daughter cells with or without the midbody (A, b, arrow). Subsequent long-term time-lapse imaging indicated a normal phase morphology for (more ...)
Some of major mitotic regulators are degraded by anaphase promoting complex/cyclosome (APC) during anaphase [17
]. Degradation of the polo-like kinase (Plk1) and aurora A by APC occurred while they were localized along midzone microtubules [17
], raising the possibility that interphase progression may require the degradation of some mitotic/cytokinetic proteins, which may then cause activation of downstream components crucial for cell cycle. In addition, some molecules associated with midzone microtubules may be directly involved in cell cycle events such as DNA synthesis, as suggested by the chromosomal passenger protein-like dynamics of a DNA replication initiating factor, Orc6 during cell division [7