Fission yeast cells mix parental mitochondria in a progressive manner during meiosis

ABSTRACT

Mitochondria in many fungi are inherited uniparentally during meiosis. It has remained unclear whether parental mitochondria in the fission yeast Schizosaccharomyces pombe are inherited uniparentally or biparentally. Here, we assessed the mixing of parental mitochondria carefully by live-cell microscopy and developed an algorithm to determine the degree of mitochondrial mixing in a quantitative manner. We found that parental mitochondria in fission yeast cells were mixed progressively as meiosis progressed. Moreover, we established that mitochondrial fission and the size of the conjugation neck are the limiting factors in restricting the mixing of parental mitochondria. We further employed a combination of quantitative polymerase chain reaction, fluorescent live-cell microscopy, and transmission electron microscopy approaches to examine the mitochondrial inheritance of progeny cells derived from a cross between wild-type and Rho0 (mitochondrial DNA absent) cells. The results show that all progeny cells of the cross carry mitochondrial DNA. Hence, our data support the model in which parental mitochondria in the fission yeast S. pombe are inherited biparentally during meiosis.

Introduction

Mitochondria cannot be generated de novo, and thus proper inheritance of mitochondria is crucial for cell growth and proliferation (Basse, 2010). Similar to the maternal inheritance of mitochondria during fertilization in mammals, the inheritance of mitochondria in many fungi during meiosis is uniparental, i.e. progeny cells obtain mitochondria from only one of the parental cells. It was hypothesized that uniparental inheritance of mitochondria plays an important role in avoiding the spread of heterogenous mitochondrial genomes that may be detrimental to the host cell (Basse, 2010; Greiner et al., 2015). Uniparental inheritance of mitochondria is regulated by different mechanisms in fungi undergoing sexual reproduction. For example, in unicellular fungi undergoing isogamy, mitochondria are degraded from one gamete in a mating-type dependent manner, thus ensuring homoplasy in the zygote (Yan et al., 2007; Fedler et al., 2009). In filamentous fungi, after two size-compatible mycelia cross each other, the fused nuclei comigrate to one side of the meiotic cell, and as a consequence, the cytoplasmic contents, including mitochondria, on the other side of the meiotic cell are left behind. The comigration of nuclei to one side of the cell ensures uniparental inheritance of mitochondria (Hintz et al., 1988; May and Taylor, 1988).

In the budding yeast Saccharomyces cerevisiae, mitochondria from two parental cells (referred to as parental mitochondria) are mixed rapidly once two opposite mating types of cells fuse (Nunnari et al., 1997). Nonetheless, mitochondrial DNA (mtDNA) is not distributed freely within the continuous mitochondrial reticulum in zygotic cells, leading to a possible combination of uniparental and biparental inheritance of parental mitochondria in progeny cells (Nunnari et al., 1997). In the fission yeast Schizosaccharomyces pombe, the interaction between parental mitochondria throughout meiosis has not been examined carefully. Moreover, it has remained unclear whether mitochondria in the fission yeast S. pombe are inherited biparentally or uniparentally. On the one hand, tetrad-dissection analysis and next-generation sequencing of colonies formed from spores suggested that mitochondria in the fission yeast S. pombe appear to be inherited biparentally during meiosis (Kamrad et al., 2021). On the other hand, it has been shown that mitochondria are tethered to the cell cortex in a Mcp5-dependent manner during meiosis in the fission yeast S. pombe and that the Mcp5-mediated cortical tethering ensures uniparental inheritance of mitochondria (Chacko et al., 2019).

The fission yeast S. pombe is a genetically tractable ‘micromammal’ model organism (Vyas et al., 2021), which has proven to be an excellent model organism for studying the cell cycle, the cytoskeleton, cell polarity, mitosis, meiosis, and mitochondria (Egel et al., 1980; Hayles and Nurse, 1992; Zheng et al., 2019; Rasul et al., 2021; Vyas et al., 2021). Fission yeast meiosis can be conveniently induced by nitrogen starvation, and two parental cells of opposite mating types conjugate and form a diploid zygote that undergoes two rounds of chromosome segregation (i.e. meiosis I and meiosis II) to form four haploid spores encapsulated with an ascus coat during meiosis (Yamashita et al., 2017). Thus, fission yeast meiosis consists of four discernible meiotic stages, i.e. prophase, meiosis I, meiosis II, and sporulation (i.e. ascus) (Vyas et al., 2021). Notably, during meiotic prophase, two fused nuclei form a horsetail shape and oscillate between the two poles of the zygote (referred to as horsetail movements) (Ding et al., 2004; Ananthanarayanan et al., 2013).

Fission yeast mitochondria are dynamic and are in a constant balance of fission and fusion (Dong et al., 2022). Therefore, the characteristics of mitochondrial dynamics may allow the spreading of mitochondria within fission yeast zygotes, leading to biparental inheritance of mitochondria. In this study, we demonstrate that mitochondria are mixed in a progressive manner during meiosis in fission yeast zygotes and provide evidence to support the model in which mitochondria in fission yeast spores are inherited biparentally during meiosis.

Results

Parental mitochondria in fission yeast zygotes are mixed progressively during meiosis

It has remained unclear whether mitochondria in the fission yeast S. pombe are inherited in a uniparental or biparental manner during meiosis (Chacko et al., 2019; Kamrad et al., 2021). Using fluorescently marked Cox4 expressed from the thiamine-suppressible promoter nmt1 as a proxy for microscopic analysis, Chacko et al. (2019) reported that parental mitochondria in fission yeast zygotes are mainly segregated from each other and that the segregation depends on the tethering of mitochondria to the cell cortex by Mcp5. The thiamine-suppressible promoter nmt1 may cause variable expression of Cox4 in cells even from the same batch of culture, making it difficult to analyze mitochondrial mixing in zygotes in a quantitative manner. In addition, fission yeast meiosis is a lengthy process. Therefore, precise analysis of the mixing and inheritance of meiotic mitochondria requires both a fluorescent marker that is expressed stably and continual examination of mitochondrial mixing throughout meiosis. To meet these requirements, we tagged Sdh2, a component of complex II in the mitochondrial respiratory chain (Ito et al., 2014), at its own locus with green fluorescent protein (GFP) or mCherry in wild-type (WT) cells carrying two opposite mating types (i.e. h+ and h–), respectively. WT cells of two opposite mating types expressing Sdh2-GFP and Sdh2-mCherry, respectively, were crossed to generate zygotes, and nuclear DNA was stained with Hoechst 34580 for distinguishing meiotic stages. We then performed live-cell microscopy to analyze mitochondrial mixing in zygotic cells. As shown in Figure 1A, Hoechst 34580 staining enabled us to distinguish the four characteristic stages of meiosis: meiotic prophase (the characteristic horsetail movement was observed), meiosis I (two nuclei were separated during this stage), meiosis II (four nuclei formed during this stage), and ascus (DNA in the four spores was not stainable during this stage). The merged fluorescent images in Figure 1A clearly show that parental mitochondria were mixed progressively as meiosis progressed.

Figure 1.

Parental mitochondria are mixed in a progressive manner during meiosis in fission yeast zygotes. (A) Maximum projection images of WT cells at the indicated meiotic stages, i.e. prophase, meiosis I, meiosis II, and ascus. Cells carrying opposite mating types (i.e. h+ and h–) and expressing Sdh2-mCherry and Sdh2-GFP, respectively, were crossed to generate meiotic cells. Nuclei were stained with Hoechst 34580, and the nucleus undergoing horsetail movement during prophase is indicated by red arrow. DIC, differential interference contrast. Scale bar, 10 μm. (B) Schematic diagram illustrating the calculation of the mixing ratio of mitochondria. Note that cells were rotated to ensure that the side on the left contains more Sdh2-GFP signals, while the side on the right contains more Sdh2-mCherry signals. The mitochondrial mixing ratio was determined by dividing the percent of mCherry signals by the percent of GFP signals on the left side of the cell (referred to as L_{mixing ratio}) or dividing the percent of GFP signals by the percent of mCherry signals on the right side of the cell (referred to as R_{mixing ratio}). (C and D) The L_{mixing ratio} and R_{mixing ratio} of WT cells at the four indicated meiotic stages. The number of cells analyzed is 10 for each stage. Statistical analysis was performed by one-way ANOVA with the Tukey honest significant difference test (*P < 0.001).

To assess mitochondrial mixing in a quantitative manner, we developed an algorithm to calculate the ratio of mitochondrial mixing (Figure 1B). First, zygotic cells were rotated horizontally to allow the side of cells displaying more Sdh2-mCherry signals to position on the right and the side of cells displaying more Sdh2-GFP signals to position on the left. Second, zygotic cells were divided into left and right compartments from the middle of a conjugation neck, and the percentage of GFP and mCherry signals in the two separated compartments was calculated (referred to as L_{GFP percent}, R_{GFP percent}, L_{mCh percent}, and R_{mCh percent}). Finally, the ratio of mitochondrial mixing in the left compartment was determined by dividing L_{mCh percent} by L_{GFP percent} (referred to as L_{mixing ratio}), and similarly, the ratio of mitochondrial mixing in the right compartment was determined by dividing R_{GFP percent} by R_{mCh percent} (referred to as R_{mixing ratio}). According to the algorithm, both L_{mixing ratio} and R_{mixing ratio} should be smaller than 1 but close to 1 when parental mitochondria are mixed almost completely.

We used the above algorithm to calculate the ratio of mitochondrial mixing in WT zygotes at the prophase, meiosis I, meiosis II, and ascus stages, as indicated in Figure 1A. Consistent with direct microscopic observation, L_{mixing ratio} (Figure 1C) and R_{mixing ratio} (Figure 1D) increased as meiosis progressed. These results suggest that fission yeast zygotes mix parental mitochondria progressively during meiosis.

We further tested whether translation during meiosis affected the analysis of mitochondrial mixing by the method stated above. We treated zygotic cells in malt extract (ME) medium using cycloheximide, a chemical used for inhibiting protein synthesis, for 3 h and analyzed mitochondrial mixing during meiosis by live-cell microscopy. As shown in Supplementary Figure S1, inhibiting protein synthesis for 3 h did not significantly affect the analysis of mitochondrial mixing by our method.

Nuclear oscillation and Mcp5 are not involved in regulating mitochondrial mixing during meiosis

Fused nuclei oscillate between the two poles of the zygotic cell during meiotic prophase in fission yeast (Ding et al., 2004), and both the cortical protein Mcp5 and the motor protein dynein (Dhc1) are required to orchestrate the oscillation of the fused nuclei (Ananthanarayanan et al., 2013). It is conceivable that the oscillation may promote the mixing of parental mitochondria. To test this possibility, we observed mitochondria in dhc1-deleted (dhc1∆) cells at the four meiotic stages: prophase, meiosis I, meiosis II, and ascus (Figure 2A). As reported previously (Hiraoka et al., 2000; Ananthanarayanan et al., 2013), the fused nuclei were immobile and resided at the conjugation neck during meiotic prophase in dhc1∆ zygotes. If nuclear oscillation promotes mitochondrial mixing during meiosis, we would expect that the ratio of mitochondrial mixing in dhc1∆ zygotes becomes smaller. However, calculations showed that L_{mixing ratio} and R_{mixing ratio} of dhc1∆ and WT zygotes were comparable (Figure 2B and C), suggesting that nuclear oscillation does not contribute to mitochondrial mixing during meiosis.

Figure 2.

Mitochondrial mixing during meiosis in cells lacking dynein or Mcp5. (A) Maximum projection images of dhc1Δ cells at the indicated meiotic stages. dhc1Δ cells (h+ and h–) expressing Sdh2-mCherry and Sdh2-GFP, respectively, were crossed to analyze mitochondrial mixing during meiosis. DNA was stained with Hoechst 34580. Note that the nucleus marked by red arrow was immobile and resided in the middle of the dhc1Δ cell during prophase. Scale bar, 10 μm. (B and C) Comparison of L_{mixing ratio} and R_{mixing ratio} of dhc1Δ and WT cells at the four indicated meiotic stages. The number of cells analyzed is 10 for each stage. For convenient comparison, the WT data in Figure 1C and D were used here. P-values were calculated by two-way ANOVA. (D) Maximum projection images of mcp5Δ cells at the indicated meiotic stages. mcp5Δ cells (h+ and h–) expressing Sdh2-mCherry and Sdh2-GFP, respectively, were crossed to analyze mitochondrial mixing during meiosis. The DNA was stained with Hoechst 34580. Note that the nucleus marked by red arrow was immobile and resided in the middle of the mcp5Δ cell during prophase. Scale bar, 10 μm. (E and F) Comparison of L_{mixing ratio} and R_{mixing ratio} of mcp5Δ and WT cells at the four indicated meiosis stages. The number of cells analyzed is 10 for each stage. For convenient comparison, the WT data in Figure 1C and D were used here. P-values were calculated by two-way ANOVA.

In addition to its role in directing nuclear oscillation, Mcp5 was reported to function as a tether that anchors mitochondria to the cell cortex to promote uniparental inheritance of mitochondria during meiosis in fission yeast (Chacko et al., 2019). This tethering model predicts that mitochondrial mixing would be greatly enhanced after mcp5 is deleted. To test this prediction, mitochondrial mixing in mcp5∆ cells was examined by live-cell microscopy at the four meiotic stages: prophase, meiosis I, meiosis II, and ascus (Figure 2D). Calculations showed that L_{mixing ratio} and R_{mixing ratio} of mcp5∆ and WT zygotes were also comparable (Figure 2E and F). This result suggests that Mcp5 does not function to restrict mitochondrial mixing during meiosis. Hence, Mcp5 and Dhc1 are unlikely to play a crucial role in regulating the mixing of parental mitochondria during meiosis in fission yeast.

Mitochondrial fission plays a crucial role in restricting mitochondrial mixing during meiosis in fission yeast zygotes

The progressive nature of mitochondrial mixing in fission yeast zygotes makes it possible to visualize the initial fusion of mitochondria that are originally from two parental cells. As expected, when examining the mitochondrial dynamics of two parental cells that had just fused (the two nuclei still separated from each other), we found that a separated mitochondrion at one pole of the zygote extended and moved toward and fused with the mitochondrial mass at the other pole of the zygote (Supplementary Figure S2). Frequent events of mitochondrial fusion were also found in the zygotic cell undergoing horsetail movements during prophase (Figure 3A). This evidence supports the conclusion that parental mitochondria can fuse with each other during early meiosis.

Figure 3.

Mitochondrial mixing in cells lacking the mitochondrial fission factor Dnm1 during meiosis. (A) Kymograph and montage images of a WT meiotic prophase cell. The stage of horsetail movement is indicated in the DNA kymograph graph by white arrow, and multiple events of mitochondrial fusion are indicated by white arrows in the Sdh2-GFP kymograph graph. Montage images showing the mitochondrial fusion events are shown on the right. Dashed lines mark the cell edge. Scale bar, 10 μm. (B) Schematic diagram illustrating the method of counting the number of mitochondria and mitochondrial branches in meiotic and vegetative cells. Images were skeletonized, and the number of mitochondria and mitochondrial branches was determined by the mitochondrial network analysis plugin in ImageJ. (C) Comparison of the numbers of mitochondria and mitochondrial branches in meiotic and vegetative cells. Statistical analysis was performed by Student's t-test. (D) Maximum projection images of dnm1Δ cells at the indicated meiotic stages. dnm1Δ cells (h+ and h–) expressing Sdh2-mCherry and Sdh2-GFP, respectively, were crossed, and the cells were stained with Hoechst 34580. Red arrows mark the nucleus undergoing horsetail movement during prophase. Scale bar, 10 μm. (E and F) Comparison of L_{mixing ratio} and R_{mixing ratio} of dnm1Δ and WT cells at the four indicated meiosis stages. The number of cells analyzed is 10 for each stage. For convenient comparison, the WT data in Figure 1C and D were used here. P-values were calculated by two-way ANOVA.

Interestingly, quantification showed that the number of mitochondria and mitochondrial branches in zygotic cells was greater than that in vegetative cells. This result indicates that mitochondrial fission may be a limiting factor for preventing the mixing of parental mitochondria (Figure 3B and C). To further test this idea, we analyzed mitochondrial mixing in zygotic cells lacking Dnm1, which is the master regulatory protein of mitochondrial fission (Otsuga et al., 1998; Dong et al., 2022), at the four meiotic stages: prophase, meiosis I, meiosis II, and ascus by live-cell microscopy (Figure 3D). It was remarkable that parental mitochondria were mixed rapidly during meiosis, and in some zygotic cells, parental mitochondria were mixed almost completely even at meiotic prophase or meiosis I (Figure 3D). Quantification confirmed our microscopic observation because L_{mixing ratio} (Figure 3E) and R_{mixing ratio} (Figure 3F) of dnm1∆ cells were significantly higher than those of WT cells. Taken together, these results suggest that mitochondrial fission functions to prevent mitochondrial mixing during meiosis in fission yeast zygotes.

We further created strains lacking Fzo1, the master regulatory protein promoting fusion of the mitochondrial outer membrane (Rapaport et al., 1998; Dong et al., 2022), and then examined mitochondrial mixing in the cross of strains lacking Fzo1. As shown in Supplementary Figure S3, the absence of mitochondrial fusion significantly affected mitochondrial mixing during meiosis. Even during the late meiotic stage, i.e. the ascus stage, the signals of Sdh2-mCherry and Sdh2-GFP were quite separable, indicative of a low degree of mitochondrial mixing. Since the fragmented mitochondria in cells lacking Fzo1 were still able to be distributed into the left and right compartments of the zygotic cell in a fusion-independent manner, our algorithm was not applicable in this case for quantitatively determining the degree of mitochondrial mixing.

The size of the conjugation neck restricts mitochondrial mixing during meiosis in fission yeast

We noticed that zygotic mitochondria from two parental cells generally flanked the conjugation neck during early meiosis (Figures 1A and 2A, D) and that the size of the conjugation neck was generally smaller than the width of the two parental cells. We then asked whether the size of the conjugation neck could restrict the mixing of mitochondria in zygotes. To address this question, we searched for mutant cells that have a small conjugation neck. We found that the absence of Spn1 significantly decreased the size of the conjugation neck (Figure 4A). Spn1 is a cytoskeleton protein localized to the cell cortex and is involved in regulating cytokinesis and cell septation (Zheng et al., 2018). Moreover, it has not been reported that Spn1 is involved in regulating mitochondrial dynamics in fission yeast. Therefore, direct involvement of Spn1 in mitochondrial mixing during meiosis is unlikely. Microscopic analysis showed that parental mitochondria in spn1∆ zygotes were also mixed progressively (Figure 4B). However, quantification revealed that the L_{mixing ratio} and R_{mixing ratio} of spn1∆ zygotes were significantly lower than those of WT zygotes (Figure 4C), suggesting that mitochondrial mixing was prevented in spn1∆ zygotic cells. Hence, this result supports the model that the size of the conjugation neck is a limiting factor restricting parental mitochondrial mixing during meiosis in fission yeast zygotes.

Figure 4.

Mitochondrial mixing in cells with a small conjugation neck. (A) Comparison of the conjugation neck sizes between WT and spn1Δ cells. The number of cells analyzed is indicated, and statistical analysis was performed by Student's t-test. Shown on the left are DIC images of WT and spn1Δ cells. Scale bar, 10 μm. (B) Maximum projection images of spn1Δ cells at the indicated meiotic stages. spn1Δ cells (h+ and h–) expressing Sdh2-mCherry and Sdh2-GFP, respectively, were crossed to generate meiotic cells, and DNA was stained with Hoechst 34580. Red arrows mark the nucleus undergoing horsetail movement. Note that DNA appeared to fail to migrate to the other side within the zygote. Scale bar, 10 μm. (C) Comparison of L_{mixing ratio} and R_{mixing ratio} of spn1Δ and WT cells at the four indicated meiotic stages. The number of cells analyzed is 10 for each stage. For convenient comparison, the WT data in Figure 1C and D were used here. P-values were calculated by two-way ANOVA.

Forcing mitochondria to bind actin filaments promotes mitochondrial mixing during meiosis

We have previously developed three chimeric molecules for the targeted localization of mitochondria to microtubules and actin filaments (Li et al., 2015): (i) Ase1-chimera, the microtubule-binding domain of Ase1 and the mitochondria outer membrane protein Tom22 are fused to TagRFP at the N- and C-termini, respectively; (ii) Klp3-chimera, the motor domain of Klp3 and Tom22 are fused to TagRFP at the N and C-termini, respectively; and (iii) CHD-chimera, the actin binding domain of Rng2 and Tom22 are fused to TagRFP at the N and C-termini, respectively. In the present study, we replaced TagRFP with the nonfluorescent tag Myc and used these chimeric proteins to force mitochondria to bind either microtubules or actin filaments. We reasoned that the forced association of mitochondria with microtubules may promote mitochondrial mixing during the stage of nuclear oscillation (i.e. meiotic prophase) since thick microtubule bundles form during meiotic prophase to direct the pole-to-pole oscillation of the fused nuclei. Surprisingly, we did not observe enhanced mixing of mitochondria at meiotic prophase in zygotic cells expressing Ase1-chimera or Klp3-chimera (Figure 5A–C). This may be due to the short life and dynamic nature of the microtubule bundles that direct nuclear oscillation. Intriguingly, enhanced mixing of mitochondria at meiotic prophase was found in zygotic cells expressing CHD-chimera (Figure 5A–C), which was similar to the phenotype observed in dnm1∆ zygotes. Actin filaments have been shown to be highly enriched within the shmoos to promote cell‒cell fusion when two parental cells cross each other (Dudin et al., 2015). Therefore, it is conceivable that CHD-chimera enabled mitochondria from the two parental cells to position closer to each other upon cell‒cell fusion, leading to enhanced mitochondrial mixing. Interestingly, treatment of zygotic cells with latrunculin A, a chemical used for depolymerizing actin filaments, for 3 h did not appear to affect mitochondrial mixing during meiosis (Supplementary Figure S1). This result indicates that mitochondria are mixed in an actin filament-independent manner after cell‒cell fusion.

Figure 5.

Mitochondrial mixing in cells expressing chimera proteins that tether mitochondria to microtubules and actin filaments. (A) Maximum projection images of meiotic prophase cells expressing Ase1-13Myc-Tom22 (Ase1-chimera, tethering mitochondria to microtubules), Klp3-13Myc-Tom22 (Klp3-chimera, tethering mitochondria to microtubules), or Rng2-13Myc-Tom22 (CHD-chimera, tethering mitochondria to actin filaments). The cells (h+ and h–) expressing Sdh2-mCherry and Sdh2-GFP, respectively, were crossed to generate meiotic cells, and DNA was stained with Hoechst 34580. Red arrows mark the nuclei undergoing horsetail movement. Scale bar, 10 μm. (B and C) Comparison of L_{mixing ratio} and R_{mixing ratio} of the indicated cells at meiotic prophase. The number of cells analyzed in each group is 10. For convenient comparison, the WT data in Figure 1C and D and the data for dnm1Δ cells in Figure 3 were used here. P-values were calculated by Student's t-test, and no statistically significant difference was found between CHD-chimera and dnm1Δ cells and among WT, Klp3-chimera, and Ase1-chimera cells (indicated as n.s.). (D and E) Viability test for WT, spn1Δ, dnm1Δ, and CHD-chimera progeny cells grown on YE5S and YE5S plus glycerol media. Red dashed circles mark the places where spores failed to grow. Quantification is shown in E, and four independent experiments were carried out. P-values were calculated by Student's t-test.

What is the consequence of the premature rapid mixing of mitochondria in dnm1∆ zygotes and zygotic cells expressing CHD-chimera? To address this question, we assessed the viability of spores on plates containing fermentable (YE5S-glucose) or nonfermentable (YE5S-glycerol) medium by tetrad-dissection analysis (Figure 5D). Quantification revealed that the percentage of viable spores derived from dnm1∆ zygotic cells or zygotic cells expressing CHD-chimera decreased significantly if the spores were grown on plates containing nonfermentable medium (Figure 5E). Note that CHD-chimera functions specifically to tether mitochondria to actin filaments, leading to premature mitochondrial mixing. Therefore, these results indicate that premature mixing of parental mitochondria during early meiosis may contribute to the decreased environmental fitness of fission yeast progeny. Alternatively, the decreased spore viability of dnm1∆ and CHD-chimera cells could be caused by the effects of the absence of Dnm1 and CHD-chimera on mitochondrial functions.

The absence of Spn1 prevents mitochondrial mixing (Figure 4). We further assessed the viability of spn1∆ spores by the approach stated above. Consistently, the percentage of viable spores derived from spn1∆ zygotic cells on fermentable and nonfermentable media was comparable (Figure 5D and E). Thus, progressive mixing of parental mitochondria may be an approach for fission yeast progeny to resist unfavorable conditions.

Mitochondrial inheritance during meiosis is biparental in fission yeast

To further test whether mitochondria are inherited uniparentally or biparentally, we first assessed the distribution of mitochondria in zygotes generated by crossing Sdh2-GFP-expressing WT cells with Rho0 cells, whose mtDNA is absent (Haffter and Fox, 1992). As shown in Figure 6A and B, although Shd2-GFP-marked mitochondria were positioned on one side of the zygotic cell during meiotic prophase, Sdh2-GFP signals were clearly detectable in all four spores at later stages of meiosis. These findings support the model in which mitochondria are inherited biparentally in fission yeast zygotes. We then grew spores derived from the cross between WT and Rho0 cells on YE5S plates (Figure 6C). Interestingly, each tetrad displayed two large and two small colonies (Figure 6C). We then performed quantitative real-time polymerase chain reaction (qPCR) to quantify mtDNA for progeny cells from three tetrads of the cross between WT and Rho0 cells. Consistently, mtDNA was detected in all tested colonies (Figure 6D), suggesting that all progeny cells from the cross between WT and Rho0 cells inherited mtDNA from the WT parental cell. In addition, we stained progeny cells from one tetrad with MitoTracker Green and found that all progeny cells contained mitochondria (Figure 6E). Nonetheless, we noticed that cells from one colony (i.e. Tetra-4 in Figure 6E) displayed very weak fluorescent signals (Figure 6E). The weak mitochondrial signals may be due to mitochondrial malfunction or reduced mitochondrial mass caused by the mutation inherited from Rho0 parental cells.

Figure 6.

Mitochondria exist in progeny cells generated by crossing WT and Rho0 strains. (A) Maximum projection images of meiotic cells. WT cells expressing Sdh2-GFP were crossed with Rho0 cells to generate meiotic cells, and mitochondrial distribution at the prophase and ascus stages is shown. DNA was stained with Hoechst 34580. Red arrows mark the parental Rho0 cell. Scale bar, 10 μm. (B) Quantification of the ratio of mitochondrial fluorescent signals in Rho0 and WT parental cells at the indicated meiotic stages. P-values were calculated by Student's t-test. (C) Tetrad-dissection analysis of asci generated by crossing WT and Rho0 cells. Colonies from the four spores of one tetrad were positioned vertically. Red squares mark the two small colonies from one tetrad. (D) Relative mtDNA copy number of colonies from one tetrad. Three tetrads were analyzed by qPCR. WT and Rho0 strains were used as controls, and measurement data were normalized to the WT data. (E) Representative maximum projection images of mitochondrial staining for the four colonies from one tetrad. WT and Rho0 cells were used as positive and negative controls, respectively. Mitochondria were stained with MitoTracker Green. Note that Tetra-4 cells displayed weak mitochondrial staining. Scale bar, 10 μm. (F) Mitochondrial and nuclear DNA staining images of cells from the four colonies indicated above. WT and Rho0 cells were used as positive and negative controls, respectively. DNA was stained with DAPI. Red arrows indicate mtDNA. Scale bar, 10 μm. (G) Transmission electron microscopic images of cells from the four colonies indicated above. WT and Rho0 cells were used as positive and negative controls, respectively. Magnified images of the indicated regions (dashed squares) are shown on the right. Mitochondria are indicated by red arrows. Scale bar, 2 μm.

To visualize mtDNA directly, we stained cells of the same tetrad shown in Figure 6E with 4′,6-diamidino-2-phenylindole (DAPI) (Figure 6F). Consistently, Rho0 parental cells did not display staining signals in the cytoplasm except in the nucleus, and similar to WT parental cells, all progeny cells displayed staining signals both in the cytoplasm and in the nucleus (Figure 6F). We noticed that the progeny cells showing weak staining of mitochondria (Tetra-4 in Figure 6E) also displayed a reduced number of DAPI-stained foci within the cytoplasm (Tetra-4 in Figure 6F). To further confirm the above findings, we employed transmission electronic microscopy to examine cells of the same tetrad shown in Figure 6E and F. As shown in Figure 6G, mitochondria were present in WT cells and progeny cells from the tetrad, whereas typical mitochondrial structures were not discernible in Rho0 cells. Collectively, these results support our claim that mitochondrial inheritance is biparental in fission yeast.

Discussion

Uniparental inheritance of mitochondria is prevalent and has been observed in numerous ascomycetous and basidiomycetous species, including but not limited to Aspergillus nidulans, Candida albicans, and Ustilago maydis (Basse, 2010; Mendoza et al., 2020). By contrast, in the budding yeast S. cerevisiae, mitochondria are mixed rapidly after fusion of the isogamous mating partners but mtDNA in progeny cells could be inherited from both parents or one parent during meiosis (Nunnari et al., 1997). How parental mitochondria in the fission yeast S. pombe, another popular model organism, interact with each other during meiosis remains to be examined carefully. Moreover, it has remained unclear whether meiotic mitochondria are inherited biparentally or uniparentally (Chacko et al., 2019; Kamrad et al., 2021). In the present study, we demonstrate that mitochondria are mixed progressively as meiosis progresses and are inherited biparentally in fission yeast zygotes (Figure 7).

Figure 7.

Working model for mitochondrial mixing during meiosis in fission yeast. Two fission yeast cells with opposite mating types are crossed to form a zygote undergoing meiosis. During meiosis, karyogamy enables nuclear fusion, while mitochondria are mixed progressively. Progressive mitochondrial mixing is partly due to mitochondrial fission and the restriction of the conjugation neck and has important physiological significance, since premature complete mixing of mitochondria causes low viability of progeny cells in nonfermentable medium.

Kamrad et al. (2021) found evidence that supports the biparental inheritance of mitochondria during meiosis in the fission yeast S. pombe. However, Chacko et al. (2019) suggested otherwise. In this study, we have provided evidence to support the former model that mitochondria are biparentally inherited in fission yeast. We employed a similar approach, as reported previously (Chacko et al., 2019), to examine mitochondrial distribution in a cross between WT cells expressing Sdh2-GFP and Rho0 cells, which do not carry mtDNA (Haffter and Fox, 1992), and to assess the inheritance of mtDNA in progeny cells derived from the cross (Figure 6). We reasoned that if meiotic mitochondria inheritance is biparental, we would expect to see that all progeny cells contain Sdh2-GFP-marked mitochondria. Similar to the result presented in a previous study (Chacko et al., 2019), Sdh2-GFP-marked mitochondria localized mainly within only one side of the zygote during meiotic prophase (Figure 6A and B). Nonetheless, when observing Sdh2-GFP-marked mitochondria at later stages of meiosis, we found that all four spores in one zygotic cell displayed GFP signals (Figure 6A and B). This result clearly suggests that meiotic mitochondrial inheritance is biparental. We further rigorously tested this conclusion by qPCR quantification of mtDNA (Figure 6D), fluorescence microscopic analysis of mitochondria and mtDNA in progeny cells (Figure 6E and F), and electron microscopic analysis of mitochondria in progeny cells (Figure 6G). All results from these different experiments support the model of biparental inheritance of meiotic mitochondria in fission yeast.

Unlike mitochondria in budding yeast zygotes (Nunnari et al., 1997), mitochondria in fission yeast zygotes were mixed slowly during meiosis in a progressive manner (Figure 1). By contrast, it was reported previously that parental mitochondria in ∼60% and ∼40% of zygotes at meiotic prophase and the ascus stage, respectively, did not mix (Chacko et al., 2019). In this previous work, Cox4 expressed from the thiamine-suppressible promoter nmt1 was used to mark mitochondria in parental cells. Given the suppressible nature of nmt1, the expression of Cox4 may vary from cell to cell, affecting the precision of the analysis of mitochondrial mixing in zygotes. In addition, experimental conditions and the method for analyzing mitochondrial mixing could affect the precision of the analysis of mitochondrial mixing in zygotes. Therefore, in the present study, we chose to mark parental mitochondria with Sdh2, which was tagged endogenously. Moreover, we made two improvements for precisely analyzing the mixing of mitochondria in zygotes: (i) nuclei were stained with Hoechst to distinguish meiotic stages, i.e. prophase, meiosis I, meiosis II, and ascus (Figure 1A); and (ii) an algorithm was developed to determine the degree of mitochondrial mixing at the four meiotic stages in a quantitative manner (Figure 1B–D). Our data revealed a new pattern of mitochondrial mixing, i.e. progressive mixing of mitochondria during meiosis in fission yeast. In contrast to the finding reported previously that parental mitochondria are separated from each other upon cell‒cell fusion (Chacko et al., 2019), our live-cell microscopic data showed that the mixing of parental mitochondria takes place during even early meiosis soon after karyogamy (Figure 3A; Supplementary Figure S2).

In the present work, we also provided evidence to support the claim that mitochondrial fission and the size of the conjugation neck play crucial roles in restricting the mixing of parental mitochondria. First, we found that mitochondria are more fragmented in zygotes than in vegetative cells (Figure 3C). Second, inhibition of mitochondrial fission by deletion of dnm1, the master regulatory protein of mitochondrial fission, significantly enhanced the mixing of parental mitochondria (Figure 3D–F). Parental mitochondria in some dnm1∆ zygotes mixed almost completely during even early meiotic stages (i.e. meiotic prophase and meiosis I). Last, reduction of the size of the conjugation neck by deletion of spn1 caused significantly slower mixing of parental mitochondria as meiosis progressed (Figure 4). It is possible that nuclear oscillation during meiosis may promote the mixing of parental mitochondria. However, our data, along with the findings reported previously (Chacko et al., 2019), showed that the absence of nuclear oscillation does not affect the mixing of parental mitochondria. Therefore, nuclear oscillation may play a negligible role in mitochondrial mixing during meiosis in fission yeast. Tethering of parental mitochondria to the cell cortex by Mcp5 was found to be a crucial factor in restricting mitochondrial mixing during meiosis in fission yeast (Chacko et al., 2019). However, our quantitative analysis showed that the progressive mixing of parental mitochondria in WT and mcp5∆ zygotes was comparable (Figure 2D–F), suggesting that Mcp5 also plays a negligible role in regulating the mixing of parental mitochondria. Thus, both mitochondrial fragmentation and the physical barrier of the conjugation neck, but not nuclear oscillation and Mcp5-mediated tethering, function to prevent the mixing of parental mitochondria during meiosis in fission yeast. Our findings predict that parental mitochondria would mix rapidly if they were positioned closely upon cell‒cell fusion. Consistently, positioning parental mitochondria in the shmoo, where actin filaments are enriched, by using the chimera protein CHD-chimera (Figure 5A–C) significantly enhanced the mixing of parental mitochondria.

Why do then parental mitochondria mix progressively in fission yeast? Interestingly, the viability of the spores derived from WT, spn1∆, dnm1∆, and CHD-chimera zygotes was comparable if the spores were grown on fermentable medium (i.e. YE5S plus glucose) (Figure 5D and E). By contrast, the viability of the spores derived from dnm1∆ and CHD-chimera zygotes was reduced significantly when the spores were grown on nonfermentable medium (i.e. YE5S plus glycerol) (Figure 5D and E). Hence, progressive mixing of parental mitochondria may contribute to the increased environmental fitness of fission yeast progeny. The underlying mechanism of the increased environmental fitness by the progressive mixing of parental mitochondria awaits further investigation.

Materials and methods

Plasmids and yeast strains

Fission yeast strains were created by tetrad-dissection or random spore digestion. Plasmids were constructed by homologous recombination using the ClonExpress II One Step Cloning Kit (www.vazyme.com). The strains and plasmids used in this study are listed in Supplementary Tables S1 and S2, respectively. For visualization of mitochondria, Sdh2 was tagged at its own locus with GFP or mCherry, and no drug inhibiting protein translation was used when cells were crossed unless otherwise specified.

Microscopy

Imaging was performed with a PerkinElmer UltraVIEW Vox spinning-disk microscope equipped with a Hamamatsu C9100-23B EMCCD camera and a CFI Apochromat TIRF 100× objective (NA = 1.49). Imaging slides were prepared by sandwiching cells/zygotes between an Edinburgh minimal medium (EMM)-agarose pad and a coverslip (Shen et al., 2019). To generate zygotic cells, cells of opposite mating types were first streaked on YE5S (yeast extract plus five supplements: adenine, leucine, uracil, histidine, and lysine, with the concentration of each supplement at 0.225 g/L) plates and grown overnight at 30°C. Fresh cells carrying opposite mating types from YE5S plates were then mixed on ME plates for 10 h before the cells were collected for analysis by live-cell microscopy. Nuclei were stained with 10 μg/ml Hoechst 34580 (Thermo Fisher Scientific; H21486). For imaging vegetative cells, cells were grown in EMM5S (EMM plus five supplements: adenine, leucine, uracil, histidine, and lysine, with the concentration of each supplement at 0.225 g/L) and collected for imaging until the OD₆₀₀ reached 0.5–1.0. All media used for culturing fission yeast cells were purchased from Formedium (www.formedium.com). Stack images containing 11 planes at 0.5 μm spacing were acquired every 2 min. Time-lapse maximum projection images were created using MetaMorph 7.7 (www.moleculardevices.com).

Drug treatment of meiotic cells

Fresh WT cells carrying opposite mating types from YE5S plates were mixed on ME plates for 12 h. The newly prepared meiotic cells were then added to 250 μl of ME liquid in a 1.5-ml Eppendorf tube. Then, DMSO and cycloheximide (at a concentration of 4.32 mM) or latrunculin A (at a concentration of 0.48 mM) were added to the tube to treat the cells for 3 h. After treatment, zygotic cells were collected for nuclear staining before analysis by live-cell microscopy.

MitoTracker staining

For staining mitochondria in Figure 6E, cells were washed once with deionized distilled water and stained with 200 nM MitoTracker Green FM (Biosharp Life Science; M7514) dissolved in EMM5S, followed by shaking at 200 rpm for 10 min. The cells were then washed three times with EMM5S and collected for imaging.

DAPI staining

Staining of mtDNA in live cells (Figure 6F) was performed using DAPI as described previously (Baruffini et al., 2010). Cells were washed twice with deionized distilled water and were then resuspended in 200 μl 50 mM Tris–HCl, followed by the addition of 20 μl 10 μg/ml DAPI (Biosharp; BL105A). The cells were then kept on ice for 60 min before collection by centrifugation at 3000× g for 1 min and imaged with an Olympus IX73 fluorescence microscope equipped with a DP73 CCD camera and a UPI 100× objective (NA = 1.40).

Transmission electron microscopy

Transmission electron microscopy was performed as described previously (Zhao et al., 2020). Cells were cultured in EMM5S for 12 h, harvested and washed once with deionized distilled water, fixed with freshly prepared 1% glutaraldehyde and 4% KMnO₄, dehydrated through a series of ethanol solutions at concentrations of 50%, 70%, 90%, and 100%, and embedded in Spurr's resin (Kaiser and Schekman, 1990). Thin sections of 70 nm were examined using a Hitachi HT7700 transmission electron microscope equipped with an AMT Model XR81-B-M1-FX camera.

Measurement of relative mtDNA copy number

Yeast cells were grown in 10 ml of EMM5S medium at 30°C to mid-log phase. Genomic DNA was isolated with a TIANamp Yeast DNA kit (TIANGEN; DP307-02) and quantified using a NanoDrop One spectrophotometer (Thermo Scientific). The relative copy number of mtDNA was determined by qPCR with the primers (for primer sequences, see Supplementary Table S3) used previously (Wang et al., 2017). The median cycle threshold (C_t) values for mtDNA-encoded genes (cob1, cox2, and atp6) and the median C_t values for nuclear genes (histone, actin, and gpm1) were used to estimate mtDNA and nuclear DNA levels, respectively. The C_t value was calculated by subtracting the median C_t value for mtDNA from the median C_t value for nuclear DNA. The fold change in the mtDNA copy number of each mutant strain compared to that of the WT strain (whose value was set to 1) was determined by the equation 2^−ΔΔCt of the ΔΔC_t method.

Tetrad dissection

Tetrad dissection was performed with a dissection microscope (SporePlay+; Singer Instruments). To test spore viability, the dissected tetrad spores were then allowed to grow on YE5S agar plates containing 3% glucose or YE5S glycerol agar plates containing 0.1% glucose and 3% glycerol at 30°C for 5–7 days.

Data analysis

The intensity of mitochondria marked with GFP and mCherry was measured by MetaMorph 7.7 (Molecular Devices). Montage and kymograph graphs were constructed using MetaMorph 7.7. Plots were created using KaleidaGraph 4.5 (Synergy). Imaging data were processed with MetaMorph 7.7 and ImageJ (NIH). P-values were calculated by Student's t-test or analysis of variance (ANOVA) in KaleidaGraph 4.5.

Funding

This work is supported by grants from the National Natural Science Foundation of China (91754106, 31621002, 31871350, and 32070707) and the Natural Science Key Project of Anhui Institution of Higher Education (KJ2020A0441).

Conflict of interest: none declared.