- Email: [email protected]
Association of mitochondria with spindle poles facilitates spindle alignment
Current Biology Vol 18 No 15 R646
Nadine Krüger and Iva M. Tolić-Nørrelykke In all eukaryotic cells, correct segregation of the genetic material during cell division requires proper positioning and alignment of the mitotic spindle with respect to the cell division plane. In the fission yeast Schizosaccharomyces pombe, the spindle is aligned by interphase microtubules during early mitosis [1], and by astral microtubules and cell shape during late mitosis [2]. However, the mechanisms that control spindle alignment in mid- mitosis are not known. The lack of such a control could lead to extensive spindle rotation and misalignment, which, in turn, would result in chromosome mis- segregation if spindle elongation were impaired [1]. Here, we show that the association of mitochondria with the spindle poles reduces spindle rotation. In wild-type cells, spindles with associated mitochondria did not rotate as much as free spindles. In a mutant lacking the centrosomin-related protein Mto1p, the association between mitochondria and spindles was reduced and the spindles rotated more than those in wild-type cells. We propose that there is a symbiotic relationship between mitochondria and the mitotic spindle, in which close association between the two organelles facilitates the positioning of both: while the spindle helps to segregate mitochondria equally among the nascent daughter cells [3], mitochondria decrease spindle rotation and thus promote spindle alignment. We set out to identify a mechanism that constrains spindle rotation before anaphase and thereby helps the spindle to maintain proper alignment. Based on the observed association of mitochondria with
A
B 0′
Mi
Wild type 1.5′
more than spindles associated with mitochondria. Consistent with our hypothesis, we observed that the free spindles reached a significantly larger angle (α; see inset in Figure 1D) than the associated spindles (Figure 1D). The mitochondrial hypothesis further predicts that, during the time interval when one spindle pole is free while the other one is associated with a mitochondrion, the free pole should move more than the associated one. Indeed, the free pole covered a significantly larger distance than the associated pole (Figure 1E), suggesting that the spindle rotation is mainly induced by the movement of the free pole. These results provide further evidence that the interaction of mitochondria with spindle poles attenuates spindle rotation. Next, we asked whether spindle rotation is driven by random or directed processes. To distinguish between these possibilities, we
Tubulin (atb1–GFP); Mitochondria (Cox4–DsRed) 24′
SPB
2 µm
Mi C 0′
1.5′
24′
E
D 60
α
Associated spindle
40 20 0
18
6
Associated
Free
Free spindle
Pole movement ∆x (µm)
Association of mitochondria with spindle poles facilitates spindle alignment
spindle poles [3,4] (Figure 1A,B), we hypothesized that the constraint on spindle rotation is provided by mitochondria (the mitochondrial hypothesis). To examine the correlation between spindle rotation and the association of mitochondria with the spindle poles, we used a wild-type strain expressing GFP–tubulin and DsRed targeted to mitochondria via a fusion with precytochrome oxidase subunit IV (COX4) [3]. As noted by Yaffe et al. [3], typical wild-type cells showed spindle poles associated with mitochondria throughout mitosis (associated spindles; Figure 1B,D and Figure S1 in Supplemental Data, published with this article online). However, this association was occasionally lost: 20% of the analyzed cells (6/30) had at least one spindle pole free from mitochondria for at least 5 minutes (free spindles; Figure 1C,D). The mitochondrial hypothesis predicts that free spindles should rotate
Max spindle angle α (°)
Correspondences
1.0
∆x
Associated pole ∆x
0.5
0
6
6
Associated
Free
Free pole Current Biology
Figure 1. Association of mitochondria with spindle poles attenuates spindle rotation in wildtype cells. (A) Electron micrograph of a wild-type S. pombe cell in mid-mitosis. A mitochondrion (Mi) is found in close proximity to each spindle pole body (SPB). (B) A typical wild-type cell shows persistent association of mitochondria with the spindle poles throughout mitosis, and only modest rotation of the spindle. (C) A rare wild-type cell, in which mitochondria–spindle pole interactions are lost, shows extensive spindle rotation. Scale bars, 2 µm. (D) Maximum spindle angle (|α|, see inset) is shown as a measure of spindle rotation. Associated wild-type spindles (red) showed smaller maximum angles than free wild-type spindles (green; p = 8·10–4). (E) Distance (∆x, see inset) covered by each spindle pole during the time interval when one pole was associated and the other one was free, for the six free spindles shown in (D). The free pole (green) moved a larger distance than the associated pole (red; p = 6·10–3). The length of the spindles in (D) and (E) was smaller than 4 µm; the numbers of cells are shown in the bars; error bars represent standard error of the mean (s.e.m.).
Magazine R647
0′
mto1∆ 4.5′
Tubulin; Mitochondria 27′
B Max spindle angle α (°)
A
thus facilitating spindle elongation. Mitochondria have been shown to affect spindle alignment in higher eukaryotes [9,10]. The nature of the interaction between spindles and mitochondria would be a good topic to explore.
C 80
mto1∆
60 40 20 0
10
9
Associated
Free
Associated
Free
Current Biology
Figure 2. Dissociation of mitochondria in mto1∆ cells correlates with large spindle rotation. (A) In a typical mto1∆ cell (strain KN002), mitochondria are less associated with spindle poles, and the spindle rotates more than in wild type. Scale bar represents 2 µm. (B) Associated mto1∆ spindles (red) showed smaller maximum angles than free mto1∆ spindles (green; p = 0.003). The numbers of cells are shown in the bars; error bars represent s.e.m. (C) Mitochondria decrease spindle rotation and thus promote spindle alignment. Left: The spindle (green) is initially aligned with the cell axis. Mitochondria (red), which are associated with the spindle poles, decrease the rotation of the spindle and thus help the spindle to remain aligned with the cell axis. Right: Spindles that are not associated with mitochondria, as found occasionally in wild-type cells and often in mto1∆ cells, rotate and lose their initial alignment. The resulting spindle misalignment may be fatal, leading to chromosome mis-segregation if spindle elongation is impaired.
Supplemental data Supplemental data are available at http:// www.current-biology.com/cgi/content/ full/18/15/R646/DC1 Acknowledgments We thank M.P. Yaffe, D. Brunner, and J. Millar for strains; M. Storch for Figure 1A; I. Raabe, S.K. Vogel, Y. Caldarelli, I. Kalinina, H. Ibarra Avila, P. Pitrone, B. Schroth-Diez, M. Weber, D. White, J. Peychl, A. Wende, and J. Nicholls for technical support; and N. Pavin, T. Gross, N. Maghelli, M. Coelho, the members of the Tolić-Nørrelykke group, and U. Anderer for discussions. References
calculated the mean squared angular displacement of spindles (Supplemental data). Random motion is characterized by a linear relationship between the mean squared displacement and time, while a directional bias leads to a nonlinear relationship [5,6]. Our data are consistent with random motion (Figure S1C). Furthermore, the observation that spindle rotation did not decrease when microtubules or actin filaments were depolymerized (Figures S2, S3, S4, and S6) indicates that cytoskeletondependent processes in the spindle are not required for spindle rotation. Together these results suggest that spindle rotation is, to a large extent, random and driven by random intracellular movements. To test the role of the spindle- pole-body structure in spindle–mitochondrion interactions and spindle rotation, we examined a mutant lacking Mto1p (Mod20p/ Mbo1p), a centrosomin- related protein that recruits the gamma- tubulin complex to the cytoplasmic face of the spindle pole body [7,8]. We found that mto1∆ cells were defective in the association of mitochondria with the spindle poles during mitosis (Figure 2A): 23% (5/22) of the mutant cells showed spindles with constant attachment to the mitochondrial network compared with 60% (18/30) in wild-type cells. At least one pole
was free for 5 minutes or more in 41% of the mto1∆ cells (9/22). Free mto1∆ spindles rotated more than mitochondria-associated mto1∆ spindles (Figure 2B), and all mto1∆ spindles, on average, rotated more than wild-type spindles (Figure S6). Moreover, the poles of elongating spindles in mto1∆ cells often passed through a group of mitochondria without becoming permanently associated (Figures 2A and S5). These results suggest that Mto1p is involved in interactions between mitochondria and spindle poles. We propose that mitochondria, through their association with spindle poles, reduce spindle rotation and thereby stabilize spindle alignment in S. pombe (Figure 2C). This association, potentially mediated by a physical link between a component of the spindle pole body and a component of the outer mitochondrial membrane, may guard against mis-segregation of chromosomes [1,7] and of mitochondria [3] during mitosis. Our work suggests a kind of ‘organelle symbiosis’, in which a close association between two organelles is beneficial for both. In the scenario presented here, two types of organelle help the positioning of one another: spindle poles help to segregate mitochondria into the future daughter cells [3], while mitochondria help the spindles to remain parallel to the cell axis,
1. Vogel, S.K., Raabe, I., Dereli, A., Maghelli, N., and Tolic-Norrelykke, I. (2007). Interphase microtubules determine the initial alignment of the mitotic spindle. Curr. Biol. 17, 438–444. 2. Tolic-Norrelykke, I.M., Sacconi, L., Thon, G., and Pavone, F.S. (2004). Positioning and elongation of the fission yeast spindle by microtubule-based pushing. Curr. Biol. 14, 1181–1186. 3. Yaffe, M.P., Stuurman, N., and Vale, R.D. (2003). Mitochondrial positioning in fission yeast is driven by association with dynamic microtubules and mitotic spindle poles. Proc. Natl. Acad. Sci. USA 100, 11424–11428. 4. McCully, E.K., and Robinow, C.F. (1971). Mitosis in the fission yeast Schizosaccharomyces pombe: a comparative study with light and electron microscopy. J. Cell Sci. 9, 475–507. 5. Berg, H.C. (1993). Random Walks in Biology, (Princeton, New Jersey: Princeton University Press). 6. Tolic-Norrelykke, I.M., Munteanu, E.L., Thon, G., Oddershede, L., and Berg-Sorensen, K. (2004). Anomalous diffusion in living yeast cells. Phys. Rev. Lett. 93, 078102. 7. Sawin, K.E., Lourenco, P.C., and Snaith, H.A. (2004). Microtubule nucleation at non-spindle pole body microtubule-organizing centers requires fission yeast centrosomin-related protein mod20p. Curr. Biol. 14, 763–775. 8. Venkatram, S., Tasto, J.J., Feoktistova, A., Jennings, J.L., Link, A.J., and Gould, K.L. (2004). Identification and characterization of two novel proteins affecting fission yeast gamma-tubulin complex function. Mol. Biol. Cell 15, 2287–2301. 9. Kidd, T., Abu-Shumays, R., Katzen, A., Sisson, J.C., Jimenez, G., Pinchin, S., Sullivan, W., and Ish-Horowicz, D. (2005). The epsilonsubunit of mitochondrial ATP synthase is required for normal spindle orientation during the Drosophila embryonic divisions. Genetics 170, 697–708. 10. Dinkelmann, M.V., Zhang, H., Skop, A.R., and White, J.G. (2007). SPD-3 is required for spindle alignment in Caenorhabditis elegans embryos and localizes to mitochondria. Genetics 177, 1609–1620.
Max Planck Institute of Molecular Cell Biology and Genetics, Pfotenhauerstrasse 108, Dresden 01307, Germany. E-mail: [email protected]
Nadine Krüger and Iva M. Tolić-Nørrelykke In all eukaryotic cells, correct segregation of the genetic material during cell division requires proper positioning and alignment of the mitotic spindle with respect to the cell division plane. In the fission yeast Schizosaccharomyces pombe, the spindle is aligned by interphase microtubules during early mitosis [1], and by astral microtubules and cell shape during late mitosis [2]. However, the mechanisms that control spindle alignment in mid- mitosis are not known. The lack of such a control could lead to extensive spindle rotation and misalignment, which, in turn, would result in chromosome mis- segregation if spindle elongation were impaired [1]. Here, we show that the association of mitochondria with the spindle poles reduces spindle rotation. In wild-type cells, spindles with associated mitochondria did not rotate as much as free spindles. In a mutant lacking the centrosomin-related protein Mto1p, the association between mitochondria and spindles was reduced and the spindles rotated more than those in wild-type cells. We propose that there is a symbiotic relationship between mitochondria and the mitotic spindle, in which close association between the two organelles facilitates the positioning of both: while the spindle helps to segregate mitochondria equally among the nascent daughter cells [3], mitochondria decrease spindle rotation and thus promote spindle alignment. We set out to identify a mechanism that constrains spindle rotation before anaphase and thereby helps the spindle to maintain proper alignment. Based on the observed association of mitochondria with
A
B 0′
Mi
Wild type 1.5′
more than spindles associated with mitochondria. Consistent with our hypothesis, we observed that the free spindles reached a significantly larger angle (α; see inset in Figure 1D) than the associated spindles (Figure 1D). The mitochondrial hypothesis further predicts that, during the time interval when one spindle pole is free while the other one is associated with a mitochondrion, the free pole should move more than the associated one. Indeed, the free pole covered a significantly larger distance than the associated pole (Figure 1E), suggesting that the spindle rotation is mainly induced by the movement of the free pole. These results provide further evidence that the interaction of mitochondria with spindle poles attenuates spindle rotation. Next, we asked whether spindle rotation is driven by random or directed processes. To distinguish between these possibilities, we
Tubulin (atb1–GFP); Mitochondria (Cox4–DsRed) 24′
SPB
2 µm
Mi C 0′
1.5′
24′
E
D 60
α
Associated spindle
40 20 0
18
6
Associated
Free
Free spindle
Pole movement ∆x (µm)
Association of mitochondria with spindle poles facilitates spindle alignment
spindle poles [3,4] (Figure 1A,B), we hypothesized that the constraint on spindle rotation is provided by mitochondria (the mitochondrial hypothesis). To examine the correlation between spindle rotation and the association of mitochondria with the spindle poles, we used a wild-type strain expressing GFP–tubulin and DsRed targeted to mitochondria via a fusion with precytochrome oxidase subunit IV (COX4) [3]. As noted by Yaffe et al. [3], typical wild-type cells showed spindle poles associated with mitochondria throughout mitosis (associated spindles; Figure 1B,D and Figure S1 in Supplemental Data, published with this article online). However, this association was occasionally lost: 20% of the analyzed cells (6/30) had at least one spindle pole free from mitochondria for at least 5 minutes (free spindles; Figure 1C,D). The mitochondrial hypothesis predicts that free spindles should rotate
Max spindle angle α (°)
Correspondences
1.0
∆x
Associated pole ∆x
0.5
0
6
6
Associated
Free
Free pole Current Biology
Figure 1. Association of mitochondria with spindle poles attenuates spindle rotation in wildtype cells. (A) Electron micrograph of a wild-type S. pombe cell in mid-mitosis. A mitochondrion (Mi) is found in close proximity to each spindle pole body (SPB). (B) A typical wild-type cell shows persistent association of mitochondria with the spindle poles throughout mitosis, and only modest rotation of the spindle. (C) A rare wild-type cell, in which mitochondria–spindle pole interactions are lost, shows extensive spindle rotation. Scale bars, 2 µm. (D) Maximum spindle angle (|α|, see inset) is shown as a measure of spindle rotation. Associated wild-type spindles (red) showed smaller maximum angles than free wild-type spindles (green; p = 8·10–4). (E) Distance (∆x, see inset) covered by each spindle pole during the time interval when one pole was associated and the other one was free, for the six free spindles shown in (D). The free pole (green) moved a larger distance than the associated pole (red; p = 6·10–3). The length of the spindles in (D) and (E) was smaller than 4 µm; the numbers of cells are shown in the bars; error bars represent standard error of the mean (s.e.m.).
Magazine R647
0′
mto1∆ 4.5′
Tubulin; Mitochondria 27′
B Max spindle angle α (°)
A
thus facilitating spindle elongation. Mitochondria have been shown to affect spindle alignment in higher eukaryotes [9,10]. The nature of the interaction between spindles and mitochondria would be a good topic to explore.
C 80
mto1∆
60 40 20 0
10
9
Associated
Free
Associated
Free
Current Biology
Figure 2. Dissociation of mitochondria in mto1∆ cells correlates with large spindle rotation. (A) In a typical mto1∆ cell (strain KN002), mitochondria are less associated with spindle poles, and the spindle rotates more than in wild type. Scale bar represents 2 µm. (B) Associated mto1∆ spindles (red) showed smaller maximum angles than free mto1∆ spindles (green; p = 0.003). The numbers of cells are shown in the bars; error bars represent s.e.m. (C) Mitochondria decrease spindle rotation and thus promote spindle alignment. Left: The spindle (green) is initially aligned with the cell axis. Mitochondria (red), which are associated with the spindle poles, decrease the rotation of the spindle and thus help the spindle to remain aligned with the cell axis. Right: Spindles that are not associated with mitochondria, as found occasionally in wild-type cells and often in mto1∆ cells, rotate and lose their initial alignment. The resulting spindle misalignment may be fatal, leading to chromosome mis-segregation if spindle elongation is impaired.
Supplemental data Supplemental data are available at http:// www.current-biology.com/cgi/content/ full/18/15/R646/DC1 Acknowledgments We thank M.P. Yaffe, D. Brunner, and J. Millar for strains; M. Storch for Figure 1A; I. Raabe, S.K. Vogel, Y. Caldarelli, I. Kalinina, H. Ibarra Avila, P. Pitrone, B. Schroth-Diez, M. Weber, D. White, J. Peychl, A. Wende, and J. Nicholls for technical support; and N. Pavin, T. Gross, N. Maghelli, M. Coelho, the members of the Tolić-Nørrelykke group, and U. Anderer for discussions. References
calculated the mean squared angular displacement of spindles (Supplemental data). Random motion is characterized by a linear relationship between the mean squared displacement and time, while a directional bias leads to a nonlinear relationship [5,6]. Our data are consistent with random motion (Figure S1C). Furthermore, the observation that spindle rotation did not decrease when microtubules or actin filaments were depolymerized (Figures S2, S3, S4, and S6) indicates that cytoskeletondependent processes in the spindle are not required for spindle rotation. Together these results suggest that spindle rotation is, to a large extent, random and driven by random intracellular movements. To test the role of the spindle- pole-body structure in spindle–mitochondrion interactions and spindle rotation, we examined a mutant lacking Mto1p (Mod20p/ Mbo1p), a centrosomin- related protein that recruits the gamma- tubulin complex to the cytoplasmic face of the spindle pole body [7,8]. We found that mto1∆ cells were defective in the association of mitochondria with the spindle poles during mitosis (Figure 2A): 23% (5/22) of the mutant cells showed spindles with constant attachment to the mitochondrial network compared with 60% (18/30) in wild-type cells. At least one pole
was free for 5 minutes or more in 41% of the mto1∆ cells (9/22). Free mto1∆ spindles rotated more than mitochondria-associated mto1∆ spindles (Figure 2B), and all mto1∆ spindles, on average, rotated more than wild-type spindles (Figure S6). Moreover, the poles of elongating spindles in mto1∆ cells often passed through a group of mitochondria without becoming permanently associated (Figures 2A and S5). These results suggest that Mto1p is involved in interactions between mitochondria and spindle poles. We propose that mitochondria, through their association with spindle poles, reduce spindle rotation and thereby stabilize spindle alignment in S. pombe (Figure 2C). This association, potentially mediated by a physical link between a component of the spindle pole body and a component of the outer mitochondrial membrane, may guard against mis-segregation of chromosomes [1,7] and of mitochondria [3] during mitosis. Our work suggests a kind of ‘organelle symbiosis’, in which a close association between two organelles is beneficial for both. In the scenario presented here, two types of organelle help the positioning of one another: spindle poles help to segregate mitochondria into the future daughter cells [3], while mitochondria help the spindles to remain parallel to the cell axis,
1. Vogel, S.K., Raabe, I., Dereli, A., Maghelli, N., and Tolic-Norrelykke, I. (2007). Interphase microtubules determine the initial alignment of the mitotic spindle. Curr. Biol. 17, 438–444. 2. Tolic-Norrelykke, I.M., Sacconi, L., Thon, G., and Pavone, F.S. (2004). Positioning and elongation of the fission yeast spindle by microtubule-based pushing. Curr. Biol. 14, 1181–1186. 3. Yaffe, M.P., Stuurman, N., and Vale, R.D. (2003). Mitochondrial positioning in fission yeast is driven by association with dynamic microtubules and mitotic spindle poles. Proc. Natl. Acad. Sci. USA 100, 11424–11428. 4. McCully, E.K., and Robinow, C.F. (1971). Mitosis in the fission yeast Schizosaccharomyces pombe: a comparative study with light and electron microscopy. J. Cell Sci. 9, 475–507. 5. Berg, H.C. (1993). Random Walks in Biology, (Princeton, New Jersey: Princeton University Press). 6. Tolic-Norrelykke, I.M., Munteanu, E.L., Thon, G., Oddershede, L., and Berg-Sorensen, K. (2004). Anomalous diffusion in living yeast cells. Phys. Rev. Lett. 93, 078102. 7. Sawin, K.E., Lourenco, P.C., and Snaith, H.A. (2004). Microtubule nucleation at non-spindle pole body microtubule-organizing centers requires fission yeast centrosomin-related protein mod20p. Curr. Biol. 14, 763–775. 8. Venkatram, S., Tasto, J.J., Feoktistova, A., Jennings, J.L., Link, A.J., and Gould, K.L. (2004). Identification and characterization of two novel proteins affecting fission yeast gamma-tubulin complex function. Mol. Biol. Cell 15, 2287–2301. 9. Kidd, T., Abu-Shumays, R., Katzen, A., Sisson, J.C., Jimenez, G., Pinchin, S., Sullivan, W., and Ish-Horowicz, D. (2005). The epsilonsubunit of mitochondrial ATP synthase is required for normal spindle orientation during the Drosophila embryonic divisions. Genetics 170, 697–708. 10. Dinkelmann, M.V., Zhang, H., Skop, A.R., and White, J.G. (2007). SPD-3 is required for spindle alignment in Caenorhabditis elegans embryos and localizes to mitochondria. Genetics 177, 1609–1620.
Max Planck Institute of Molecular Cell Biology and Genetics, Pfotenhauerstrasse 108, Dresden 01307, Germany. E-mail: [email protected]