Previous studies have identified γ-actin alterations in association with resistance to microtubule targeting agents [Verrills et al., 2003
, b]. This article sought to address whether γ-actin affects microtubule function and the action of microtubule-targeted agents. Herein, we demonstrate for the first time that partial depletion of γ-actin suppresses interphase microtubule dynamics, and delays metaphase–anaphase transition. These findings link γ-actin and microtubule dynamics and potentially influence the action of microtubule-targeted agents.
It is unclear whether suppression of microtubule dynamics due to γ-actin knockdown is via a direct effect of decreased γ-actin expression on microtubule dynamics or via other regulatory proteins that regulate microtubule polymerization and depolymerization. Association between drug-induced suppression of interphase microtubule dynamics and increased microtubule stability and acetylation has been previously demonstrated [Mohan and Panda, 2008
]. However, suppression of microtubule dynamics observed in the γ-actin knockdown cells is not associated with an increase in microtubule stability, since microtubules in the γ-actin knockdown cells were less acetylated and detryrosinated compared to microtubules in the control siRNA cells. In contrast to our study, a correlation between microtubule stability and increased microtubule detryrosination has been observed in other cell types [Gundersen et al., 1987
; Webster et al., 1987
]. It is not clear whether suppression of γ-actin expression inhibits tubulin detyrosination or causes disassembly of detyrosinated microtubules. Of interest, tubulin detyrosination has been shown to inhibit the activity of microtubule depolymerizing kinesins, mitotic centromere-associated kinesin (MCAK) and KIF2A which resulted in stabilization of microtubules [Peris et al., 2009
]. We have not excluded the possibility that MCAK or KIF2A expression is increased in the γ-actin knockdown cells, which in turn could inhibit tubulin detyrosination. The significance of less detyrosination and acetylation of microtubules due to suppression of γ-actin expression is not clearly understood and requires further investigation.
The fact that distance-based frequencies were altered but not time-based frequencies (with constant growth rate) in the γ-actin knockdown cells indicate it is due to an increase in microtubule pauses. In addition, reduced microtubule shortening rates and time spent in microtubule growth following γ-actin knockdown reflects the increase in microtubule pause/attenuation. Microtubule ends pause and tether to the actin cortex which is thought to be aided by several proteins, such as the +TIPS and microtubule-actin cross-linker proteins. Microtubule plus-end tracking proteins, such as EB1 and CLIP170 bind to the distal ends of the microtubule, stabilizing the microtubules and promoting microtubule growth, and preventing catastrophe [Carvalho et al., 2003
; Howard and Hyman, 2003
; van der Vaart et al., 2009
]. Once the microtubule ends reached the cell periphery, +TIPS dissociate from the microtubule distal tips allowing depolymerization of microtubules. As shown in , the microtubule ends in the γ-actin knockdown cells are less dynamic and paused at the cell periphery for a prolonged period indicating that the microtubule ends are highly captured at the cell cortex. Since γ-actin is localized to the cell periphery of SH-EP cells and is depleted from the cell periphery following γ-actin siRNA treatment [Shum et al., 2011
]. It is possible that loss of γ-actin from the cell periphery in the γ-actin knockdown cells perturbs the interaction between +TIPS such as CLIP-associating protein (CLASP) and the microtubule ends at the cell cortex. Therefore, contributing to microtubule prolonged capturing at the cell cortex resulting in enhanced pause/attenuation. CLASP1 and CLASP2 have been shown to accumulate at the microtubule ends near the cell periphery and play a role in regulating microtubule stability [Mimori-Kiyosue et al., 2005
Furthermore knockdown of γ-actin may have altered the function of microtubule-actin cross-linker proteins, such as actin crosslinking family 7, which have been shown to regulate microtubule dynamics and microtubules tethering to the cortical actin sites [Kodama et al., 2003
]. Loss of γ-actin from the cell periphery together with loss of lamella and lamellipodial-like structures [Shum et al., 2011
], may have restricted the microtubules from tethering to the cortical actin sites which is consistent with a reduction in the average microtubule growing length observed in the γ-actin knockdown cells. However, we cannot exclude the possibility that the decrease in microtubule growing length could be a consequence of prolonged microtubule pausing at the cell cortex. Decreases in the period of microtubule growth in γ-actin knockdown cells may lead to EB1 or other +TIPS being retained by the microtubule tips at the cell cortex which in turn results in enhanced microtubule pause/attenuation. Further studies are required to determine the effect of γ-actin knockdown on function of +TIPs. These changes in microtubule dynamic parameters due to depletion of γ-actin may have a combination of effects: fewer microtubules are able to reach the cell cortex; and those microtubules that manage to reach the cell cortex are being captured for a prolonged period.
The observation that distance-based frequencies were increased in the γ-actin knockdown cells appears to correlate with our recently reported reduced cell migration of SH-EP cells [Shum et al., 2011
]. An increase in distance-based catastrophe frequency and reduction in EB1 accumulation at the microtubule plus ends has been linked to the anti-migratory effect of the microtubule-targeted agent, Epothilone B in glioblastoma cells [Pagano et al., 2012
]. The growing microtubules are captured and stabilized at the cell cortex during cell migration and an increase in catastrophe frequency would reduce the number of microtubules reaching the leading edge hence impair cell migration [Pagano et al., 2012
Our finding that there is no correlation between suppression of microtubule dynamics and increased mitotic arrest in the absence or presence of paclitaxel in the γ-actin knockdown cells, suggest that suppression of microtubule dynamics may not always lead to mitotic arrest. This is in contrast with previous published data demonstrating that suppression of interphase microtubule dynamics induced by microtubule-targeted agents leads to mitotic arrest [Honore et al., 2003
; Kamath and Jordan, 2003
; Kamath et al., 2006
; Gan et al., 2010
]. It is possible that the low levels of paclitaxel-induced mitotic arrest in the γ-actin knockdown cells are in part due to reduce cell proliferation decreasing the effectiveness of paclitaxel in blocking cells in mitosis. However, prolonged incubation with paclitaxel still fails to significantly increase the number of γ-actin knockdown cells arrested in mitosis, even though the γ-actin knockdown cells have undergone at least one complete cell division in the presence of paclitaxel. Therefore, inhibition of paclitaxel-induced mitotic arrest is also due to the effect of γ-actin depletion rather than just a consequence of reduced cell proliferation. A limitation with our study is that we are measuring interphase microtubule dynamics and the effects of partial suppression of γ-actin on kinetochore microtubule dynamics are unknown.
Interestingly, the inhibition of paclitaxel induced mitotic arrest observed in the γ-actin knockdown cells is consistent with the resistance phenotype that we previously described in the SH-EP γ-actin knockdown cells against paclitaxel [Verrills et al., 2006b
]. It has been suggested that anti-microtubule agents' induce mitotic arrest by increasing microtubule bundling and polymer levels [Jordan et al., 1992
]. Microtubule bundling was evident in the control cells treated with paclitaxel but not in the γ-actin knockdown cells. This correlates with our previous finding that suppression of γ-actin partially inhibits paclitaxel induced tubulin polymerization [Verrills et al., 2006b
]. Paclitaxel induced mitotic arrest in the control siRNA cells is probably a consequence of (1) suppression of microtubule dynamics and (2) increased microtubule bundling.
In addition to mitotic arrest, previous studies have shown that microtubule targeted agents induce suppression of microtubule dynamics with a concomitant block in metaphase–anaphase transition [Jordan et al., 1992
; Dhamodharan et al., 1995
; Yvon et al., 1999
; Honore et al., 2003
; Kamath and Jordan, 2003
; Kamath et al., 2006
]. Does suppression of interphase microtubule dynamics lead to inhibition of metaphase–anaphase transition? We showed that either knockdown of γ-actin or paclitaxel treatment of γ-actin knockdown cells suppressed microtubule dynamics with no major increase in mitotic arrest and the only common observation was the inhibition of metaphase-anaphase transition. It is possible that inhibition of metaphase–anaphase transition is a consequence of suppression of interphase microtubule dynamics. Importantly, paclitaxel suppression of microtubule dynamics and inhibition of metaphase–anaphase transition may not require γ-actin, however γ-actin is required for paclitaxel induced mitotic arrest.
Furthermore, same parameters of the microtubule dynamics were affected by γ-actin knockdown and paclitaxel treatment of the γ-actin knockdown cells such as a decrease in microtubule shortening rates and increase in length-based catastrophe frequency. Interestingly, paclitaxel treatment did not significantly affect the pause duration and period of microtubule growth in the γ-actin depleted cells compared to the effect of paclitaxel in the control siRNA cells suggesting that knockdown of γ-actin already induced maximal changes in these parameters.
This is the first demonstration that alterations in γ-actin regulate interphase microtubule dynamics, mitotic progression and the ability of paclitaxel to induce mitotic arrest. This study identified that γ-actin is required for proper microtubule function and paclitaxel induced mitotic arrest. Further studies are required to identify how γ-actin modulates interphase microtubule dynamics as this study demonstrated the first link between γ-actin and microtubule dynamics.