A large-scale RNAi screen reveals that mitochondrial function is important for meiotic chromosome organization in oocytes

A large-scale screen identified 106 genes required for the integrity of the karyosome

In Drosophila melanogaster oocytes, the meiotic chromosomes together form a compact spherical cluster called the karyosome in prophase of the first meiotic division (Fig. 1A). To identify genes important for karyosome formation in Drosophila melanogaster oocytes, we carried out a genome-wide RNAi-based screen. By using RNAseq data from ovaries of mature 4-day-old females (Mortazavi et al. 2008; Celniker et al. 2009), we excluded genes expressed in ovaries at no/extremely low levels (0 reads per million; bin 0). Initially, we included genes expressing at a very low level (1–3 reads per million; bin 1), but due to a low frequency of hits during an early phase of the screen, these genes were also excluded from the screen. Among 13,969 genes in the Drosophila melanogaster genome, 6501 genes are expressed in ovaries at a low level or higher (≥ 4 reads per million; bin 2–7). Among these 6501 genes, 3916 genes had at least one available transgenic line suitable for RNAi in female germlines when we started the screen (Fig. 1B; Table S1).

Fig. 1figure 1

A large-scale screen identified 106 genes required for the integrity of the karyosome. A A stage-5 egg chamber from a control ovary stained for Lamin and DNA. The arrowhead indicates the karyosome, a compact spherical cluster of meiotic chromosomes, in the oocyte nucleus. Other nuclei belong to nurse cells and follicle cells. Bar = 10 μm. B The number of genes selected for the screen. C The karyosome screen workflow. D Summary results of the screen. Among 3916 genes screened, 106 genes showed reproducible karyosome abnormalities when they are silenced by RNAi, while 569 genes showed severe oogenesis defects which prevented examination of the karyosome. E The frequencies of genes with severe oogenesis defects and karyosome abnormalities when silenced, according to expression levels in ovaries. Bins 2, 3, 4, 5, 6, and 7 of the ovary expression level represent genes with 4–10, 11–25, 26–50, 51–100, 101–1000, > 1000 kb−1 million−1 from RNAseq (Mortazavi et al. 2008; Celniker et al. 2009). ***Significant differences (p < 0.001) in the frequency of genes with severe oogenesis defects in comparison to bin 2. F The frequencies of genes with karyosome abnormalities, excluding genes with severe oogenesis defects that prevented examination of karyosomes, according to different expression levels in ovaries. *Significant differences (p < 0.05) in the frequencies of genes with severe oogenesis defects in comparison to bin 2. ns no significant differences (p > 0.05). G The frequencies of genes resulting in different ovary sizes when silenced, in relation to female fertility. ***Significant differences (p < 0.001) in the frequencies of genes resulting in no or tiny ovaries in comparison to genes showing female fertility. H The frequencies of genes with karyosome abnormalities in relation to female fertility. The two graphs show the frequencies including or excluding genes resulting in no/tiny ovaries, which prevented examination of karyosomes. ***Significant differences (p < 0.001) in the frequencies of genes with abnormal karyosomes in comparison to genes showing female fertility. I The frequencies of genes with karyosome abnormalities according to their ovary size. No/tiny ovaries prevented examination of karyosomes. ***Significant differences (p < 0.001) in the frequencies of genes with abnormal karyosomes in comparison to genes with normal ovaries

For each of the 3916 genes, we tested one transgenic RNAi line from the TRiP collection (Ni et al. 2011). In these lines, expression of a short hairpin (sh) RNA against a target gene is controlled by the Gal4-responsive upstream activation sequence (UAS). Each RNAi line was crossed with flies expressing Gal4 in the germline under the nanos regulatory elements. This enables expression of shRNA starting from premeiotic stages in the female germline of the progeny. Dissected ovaries were fixed and stained for DNA (Fig. 1C). In the first round of the screen, we examined a small number of oocytes between oogenesis stages 3 and 9, which roughly correspond to zygotene to pachytene stages. In wild type or a negative control (RNAi of the white gene), the karyosome was spherical in most oocytes and slightly deformed in a small proportion of oocytes. If some karyosomes showed abnormal morphology, the shRNA lines were crossed with the same driver again to re-examine more karyosomes.

Among the 3916 genes, RNAi of 106 genes (2.7%) showed frequent karyosome defects (≥ 25% of 20 karyosomes examined in the second examination) (Fig. 1D). We could not examine the karyosome morphologies for 569 genes (14.5%) upon RNAi, as they showed severe oogenesis defects (no/tiny ovaries) that prevent examination. Our screen is likely to have missed many genes required for karyosome integrity due to unavailability of RNAi lines, inefficiency of RNAi, or severe oogenesis defects. Nevertheless, this is the first systematic, unbiased, large-scale identification of genes required for the karyosome integrity in oocytes. We focused our further studies on these 106 genes that showed frequent and reproducible karyosome defects.

Genes important for fertility are enriched among the 106 genes required for the karyosome

To test a correlation between expression levels in ovaries and the karyosome or oogenesis defects, genes were grouped according to estimated amounts of mRNA from RNAseq (Mortazavi et al. 2008; Celniker et al. 2009). We found a strong correlation between the expression level in ovaries and severe oogenesis defects (Fig. 1E). The higher the expression level in ovaries is, the more likely RNAi showed severe oogenesis defects. Next, after excluding genes with severe oogenesis defects that prevent observation of karyosomes, we calculated the frequencies of genes with the abnormal karyosomes in relation to the expression levels in ovaries. Genes expressed at a higher level in ovaries are more likely to show abnormal karyosomes (Fig. 1F), although the correlation is not as strong as between expression level and severe oogenesis defects.

During the first round of the screen, we recorded the size of ovaries and tested fertility for most (~ 80%) of the genes. When ovaries are too small to examine the karyosome, they were recorded as “no or tiny ovaries.” When ovaries are substantially smaller than a control but large enough to examine the karyosome, they were recorded as “small ovaries.” Fertility was judged by production of larvae. When no larvae or a few larvae were observed, it was recorded as “sterile” or “low/reduced fertility,” respectively. Otherwise, it was recorded as “fertile.”

As expected, sterility is strongly correlated with no or very small ovary size (Fig. 1G). After excluding genes with severe oogenesis defects that prevent observation of karyosomes, we tested whether the karyosome defects are correlated with fertility or ovary size. Twenty-three percent of sterile lines and 20% of lines with reduced fertility showed abnormal karyosomes, while only 4% of fertile lines showed abnormal karyosomes (Fig. 1H). In contrast, 9% of lines with small ovaries showed abnormal karyosomes, while 5% of lines with normal sized ovaries showed abnormal karyosomes (Fig. 1I). Therefore, the integrity of the karyosome morphology is much more strongly correlated with fertility than ovary size which may represent overall ovary growth/health. Our results from a large-scale screen further highlights the importance of the karyosome for reproduction.

The 106 genes are highly interconnected and include genes regulating chromatin, nuclear envelope, and actin

We concentrated our further studies on these 106 genes with reproducible and penetrant karyosome defects. We wondered how these 106 genes are related to each other and how they function together. Using STRING database (Szklarczyk et al. 2021), we found 198 known physical and/or functional interactions among the 106 genes (Fig. 2A). To test whether these interactions are more frequent than expected, we randomly selected 106 genes from the 3356 genes we have examined in the screen and counted how many known interactions are found among the random 106 genes. By repeating this 1000 times, we obtained a distribution of the number of interactions among the random 106 genes. This gave 63 interactions on average with a maximum of 128 interactions (Fig. 2B). Our 106 genes with karyosome abnormalities have > 3 times interactions than random sets of 106 genes. Therefore, the 106 genes identified in our screen are highly enriched in interactions. Among our 106 genes, only 15 genes are without known links to other genes, while all the others are directly or indirectly linked. This showed that we have identified a set of genes that are highly interconnected with each other.

Fig. 2figure 2

The 106 genes required for the karyosome are highly interconnected and include genes regulating chromatin, nuclear envelope, and actin. A The physical and/or functional interaction network among the 106 hits. Each node represents a gene identified in this screen and are coloured according to an associated key word indicated in the box. Each line represents a physical and/or functional interaction between two genes in the STRING database. B The numbers of interactions among random 106 genes. One thousand random sets of 106 genes were selected from 3356 genes examined in the screen, and the numbers of the physical and/or functional interactions were plotted. 198 interactions found among the 106 genes identified in the screen is much higher than expected from a random set of 106 genes

Next, to find out what kind of proteins are encoded by these 106 genes, we compared gene ontology between these 106 genes with karyosome defects and the 3356 genes expressed in ovaries that we have screened (excluding genes with severe oogenesis defects). Gene ontologies over-represented among the 106 genes with karyosome defects include chromosome organisation, female gamete generation, nucleic acid metabolic process, and mitochondria-related terms. Over-representation of these terms are expected, with notable exceptions of mitochondria-related terms. This suggests that the screen was successful in identifying genes with potential roles in the meiotic chromosome organisation in oocytes. It also suggests that the substantial proportion of RNAi targeted the intended genes, which is consistent with a previous study using the same shRNA collection and driver that showed a very low frequency of off-target effects induced by shRNA (Sopko et al. 2014).

Most genes (> 90%) identified in this screen are conserved in humans. Using information available on the genes and their orthologs, we manually grouped these 106 genes into categories (Fig. 3; Table S2). Some categories are related to functions or properties previously known to have links to the karyosomes, such as chromatin-related function (Zhaunova et al. 2016), nuclear envelope proteins (Breuer and Ohkura 2015), and actin regulators (Ilicheva et al. 2019). Others are related to functions or properties not previously implicated in the karyosome, such as mitochondrial proteins. Importantly, a significant proportion of the genes (14 out of 106 genes) have not been previously characterised in Drosophila and are only referred to as “CG” numbers. Therefore, our screen provides the first functional insights into these genes in Drosophila.

Fig. 3figure 3

A list of the 106 genes important for the karyosome identified in the screen. The names of the 106 genes identified in the screen and their orthologue are shown with short summaries. They were grouped under common key words. Twenty-four genes are separately listed, as their karyosome defects caused by gene silencing are dependent on the meiotic checkpoint

As expected, our screen identified chromatin proteins, chromatin-modifying enzymes, or proteins involved in DNA metabolism. It was previously reported that the histone demethylase Kdm5/Lid controls chromatin architecture in meiotic prophase I oocytes, although it does so independently of its catalytic activity (Zhaunova et al. 2016). Our screen results revealed that silencing of additional histone-modifying enzymes triggered karyosome defects. These genes include the histone methyltransferase Set2 (Stabell et al. 2007) and the histone acetyltransferases Nejire/CREBBP and Atac2 (Ogryzko et al. 1996; Suganuma et al. 2008). In addition, our screen also identified BEAF-32 (boundary element-associated factor of 32kD), which is known to regulate gene expression by modulating a higher-order chromatin structure (Gilbert et al. 2006).

Furthermore, our screen identified nuclear envelope proteins as expected. We have previously shown that the karyosome formation requires release of chromatin from the nuclear envelope and also the nuclear pores (Cullen et al. 2005; Lancaster et al. 2007; Breuer and Ohkura 2015). Attachment of chromatin to the nuclear envelope or pore complex is mediated by BAF or Nup155 and released by NHK-1 kinase or Nup62, respectively. Our screen identified Lamin and three further nuclear pore complex components, which could advance our understanding of how untethering of chromatin to the nuclear envelope or pore complex is regulated.

Interestingly, our screen also identified 7 proteins potentially regulating actin. Involvement of actin has previously been suggested for karyosome/karyosphere formation (Ilicheva et al. 2019; Maslova and Krasikova 2012). Actin is one of the main protein constituents of the karyosphere capsule in grass frogs (Ilicheva et al. 2019). In Drosophila melanogaster, karyosome formation was frequently delayed in mutants of Scr64, which encodes a tyrosine kinase regulating actin reorganisation during oogenesis (Djagaeva et al. 2005). Our results provide further evidence for the involvement of actin in karyosome formation and could serve a vital starting point for future mechanistic studies.

Silencing of 24 genes results in karyosome defects mediated by meiotic recombination checkpoint activation

During meiotic prophase I, DSBs (DNA double-strand breaks) are formed and repaired through recombination. In the face of persistent DSBs, the meiotic recombination checkpoint is activated and triggers defects in both karyosome formation and oocyte polarity (Morris and Lehmann 1999; Fig. 4A). Persistent DSBs can be generated by two alternative situations: a failure of DSB repair or a failure of suppressing retrotransposition by a piRNA-mediated mechanism. Therefore, the karyosome abnormalities of some of our hits may be induced by the meiotic checkpoint due to a defect in either DSB repair or piRNA processing.

Fig. 4figure 4

The karyosome defects of 24 genes are mediated by the meiotic recombination checkpoint. A A diagram of the meiotic recombination checkpoint that induces the karyosome defects in response to persistent DNA double strand breaks (DSBs). Persistent DSBs can be caused by failure of DNA repair or retrotransposition due to failure of piRNA-mediated silencing. B A summary result of rescue experiments by mnk/chk2. C The karyosome morphology in an oocyte in which a gene implicated in DNA repair was silenced by RNAi in the presence (+ mnk) or absence of a heterozygous mnk/chk2 mutation. Bar = 2 μm. D The graph represents the frequencies of the karyosome morphologies in oocytes in which each gene involved in DNA double strand break (DSB) repair was silenced by RNAi in the presence (+ mnk) or absence of a heterozygous mnk/chk2 mutation. E The karyosome morphology in an oocyte in which a gene implicated in piRNA processing was silenced by RNAi. Bar = 2 μm. F The graph represents the frequencies of the karyosome morphologies in an oocyte in which each gene involved in piRNA processing was silenced by RNAi. G Immunostaining of stage-4 oocyte using an antibody against γH2Av which marks DSBs. Bar = 2 μm. H The frequencies of stage 3–6 oocytes with γH2Av foci. ***Significant differences (p < 0.001) in the frequencies of oocytes with γH2Av foci in comparison to the control RNAi

To test whether the karyosome defects depend on the meiotic checkpoint, we suppressed the checkpoint by using a mutation in mnk/chk2 encoding a key kinase essential for the functional checkpoint (Klattenhoff et al. 2007). We combined the mnk/chk2 mutation with RNAi of 106 genes with reproducible and penetrant karyosome defects to see whether the mnk/chk2 mutation rescues the karyosome defects. The mnk/chk2 mutation, even when heterozygous over the wild-type allele, can suppress the karyosome defect in spnA/rad51 RNAi, which is known to be essential for DSB repair (Fig. 4C, D). We found that, among the 106 genes with reproducible and penetrant karyosome defects, the defects of 24 genes (23%) were rescued by the mnk/chk2 mutation fully or partially, and the defects of 53 genes (50%) were not rescued. The remaining 29 genes (27%) could not be determined mainly due to severe oogenesis defects when gene silencing and the mnk/chk2 mutation were combined (Fig. 4B).

As expected, these 24 genes with checkpoint-dependent karyosome defects include genes (or their orthologues) already known to be required for DSB repair (spnA, mus301/spnC, mre11, top3α; Fig. 4C, D) or piRNA processing (armi, aub, rhi, ars2; Fig. 4E, F) in some experimental systems. As neither top3α nor mre11 have been previously reported to be involved in karyosome formation, meiotic DSB repair or checkpoint dependency in Drosophila oocytes, we aimed to confirm that silencing of these genes results in persistent DSBs in Drosophila oocytes. In wild type, DSBs were formed in region 2a and fully repaired by region 3 in the germarium. We immunostained the ovaries expressing shRNA against a gene with an antibody recognizing a phosphorylated H2A variant (γH2Av) that marks DSBs (Fig. 4G, H). In control RNAi oocytes, γH2Av foci were only observed at early stages of meiosis in the germarium, but not in later-stage oocytes (stage 3 or later, which roughly corresponds to zygotene to pachytene stages), showing timely DSB repair. In contrast, γH2Av foci were still observed in oocytes at later stages when top3α or mre11 was silenced, similarly to spnA or mus301/spnC. These results suggest that top3α and mre11 are important for efficient repair of meiotic DSBs in Drosophila oocytes.

Gene silencing of mitochondrial proteins leads to distinct karyosome defects

The 106 genes with reproducible and penetrant karyosome defects were further analyzed for the karyosome morphology. We classified abnormal karyosome morphologies and distribution into two categories (Fig. 5A). The normal karyosome morphology is largely spherical, and chromosomes are in one mass commonly away from the nuclear periphery, or appear to contact the nuclear periphery only at small areas. The first category of abnormal morphologies (called “distortion”) includes a chromosome mass (or masses) whose overall shape is far from spherical, but largely away from the nuclear periphery. The second one (called “attachment”) includes karyosomes in which the overall shape of the chromosome mass (or masses) is far from spherical and is located very close to the nuclear periphery.

Fig. 5figure 5

Mitochondrial dysfunction leads to distinct karyosome defects. A The numbers of genes showing predominantly attached, predominantly distorted, and only distorted karyosome morphologies upon silencing. B Immunostained nuclei of stage-6 oocytes expressing control and ND-B22 shRNAs using a Lamin antibody and the DNA probe DAPI. RNAi of ND-B22 gene encoding a mitochondria protein predominantly shows an “attached” karyosome morphology, in which meiotic chromosomes, often three masses, are located in proximity to the nuclear envelope. Bar = 2 μm. C Genes showing predominantly attached karyosome morphologies upon silencing. Twelve out of 14 genes in this category encode proteins with roles in mitochondria. D Rescue of the karyosome defects of ND-B22 or l(2)37Bb RNAi by an RNAi-resistant wild-type ND-B22 or l(2)37Bb transgene, respectively. ***Significant differences (p < 0.001) in the frequencies of oocytes with abnormal karyosomes in comparison to the control RNAi. ns; no significant differences (p > 0.05)

Among the 106 genes with reproducible and penetrant karyosome defects, silencing of 14 genes (13%) predominantly showed attachment morphology more than the distortion morphology. Silencing of 42 genes (40%) showed both abnormal morphologies but the distortion morphology is predominant. Silencing of the remaining 50 genes (47%) showed only distortion morphology (Fig. 5A).

Interestingly, 12 out of 14 genes predominantly showing attachment morphology encode proteins with a known or a likely role in mitochondria (Fig. 5B, C). In addition to the similarity of the karyosome morphologies, the karyosome defects in all of the examined genes that predominantly show attachment morphology are independent from the meiotic checkpoint (Figure S1; Table S2).

To confirm that the karyosome defects are due to depletion of the mitochondrial proteins and not off-target effects, we selected two of the genes (ND-B22 and l(2)37Bb) encoding mitochondrial proteins that are part of the electron transport chain for rescue experiments. An RNAi-resistant wild-type ND-B22 or l(2)37Bb transgene was generated and expressed in ovaries together with shRNA against ND-B22 or l(2)37Bb, respectively. The karyosome defects were fully rescued by expression of the wild-type transgenes (Fig. 5D). This demonstrated that silencing of ND-B22 and l(2)37Bb is responsible for the observed karyosome defects.

Gene silencing of a mitochondrial protein has multiple phenotypic consequences in female meiosis

We further characterized the karyosome and other defects in meiotic processes, induced by silencing of ND-B22, which encodes a component of the electron transport chain. In the wild-type karyosome, sister-centromeres are cohesed together and homologous centromeres then paired (Dernburg et al. 1996). To test whether centromere cohesion and pairing are affected by knockdown of mitochondrial proteins, we visualized centromeres by fluorescence in situ hybridization probed with pericentromeric satellites from chromosomes 2 and 3 (Fig. 6A). In over 80% of control RNAi oocytes, one signal or a pair of closely located signals (< 0.7 μm) were observed for each probe, showing cohesion of sister-centromeres and tight pairing of homologous centromeres (Fig. 6B). In ND-B22 RNAi, one signal or a pair of closely located signals were still observed for each probe in most oocytes, but loosely paired signals (0.7–1.5 μm) were more often observed than in a control RNAi (Fig. 6B). These results suggest that ND-B22 RNAi does not affect sister-centromere cohesion but slightly loosens pairing of homologous centromeres. To test whether relative locations of non-homologous centromeres are affected, distances between signals from peri-centromeric satellites of chromosome 2 and chromosome 3 were measured (Fig. 6C). The measurement revealed that the distance between these two signals was significantly larger than in the control RNAi. This is the case even in ND-B22 RNAi oocytes that still maintain a spherical karyosome morphology. It suggests that the internal organization of the karyosome is disrupted even within the spherical karyosome morphology in ND-B22 RNAi oocytes.

Fig. 6figure 6

Silencing of ND-B22 encoding a mitochondrial protein has multiple phenotypic consequences in female meiosis. A Fluorescence in situ hybridisation of the karyosome using peri-centromeric satellites of chromosomes 2 and 3 (cen2 and cen3) in oocytes with control RNAi and ND-B22 RNAi. B The distances between cen2 signals or between cen3 signals. They are categorized into three groups based on the distances. C The distance between cen2 and cen3 signals. The error bars represent the standard errors of the means. ***p < 0.001. **p < 0.01. D The karyosome morphologies in different stages of oocytes expressing shRNA against ND-B22 using an early or late driver. Ovaries were immunostained using a Lamin antibody and DAPI. Bar = 2 μm. E Frequencies of the karyosome morphologies in different stages of oocytes expressing shRNA against ND-B22 using an “early” (nos-Gal4) or “late” (mat-α-tubulin) driver. F The frequency of meiotic cells containing foci of γH2Av, a DSB marker, in various stages of ND-B22 RNAi ovaries. G A progenitor cell undergoes four mitotic divisions to generate 16 interconnected cells. The synaptonemal complex is formed in four cells and completes disassembling in two of the cells first and then in the third cells. The remaining cell (the oocyte) gradually disassembles the synaptonemal complex in later stages. H Morphologies of the synaptonemal complex (C(3)G) in four meiotic nuclei in various stages of oogenesis. The nucleus with the most well formed synaptonemal complex is assigned as nucleus 1, and the nucleus with second, third, and fourth most well formed are assigned as nucleus 2, 3, and 4, respectively

To test at which stage the karyosome becomes abnormal, the karyosome morphology was observed stage by stage. When shRNA was expressed using the GAL4-induced promoter and a routinely used GAL4 driver (driven by nanos regulatory elements), the karyosome defects were observed from stage 3, at which point the karyosome is formed in the wild type (Fig. 6D, E). This indicates that normal mitochondria function is important for karyosome formation.

To test whether mitochondria function is important also for maintenance of the karyosome, we first allowed the karyosome to form before depleting the mitochondria protein sufficiently by delaying the start of expression of shRNA. We used Gal4 driven by the maternal α-tubulin promoter that starts expressing Gal4 later than nanos regulatory elements we routinely used (Matthews et al. 1993). At stages 3 and 4, spherical karyosomes were formed and maintained in most of oocytes. However, by stage 7, meiotic chromosomes were attached to the nuclear envelope in most oocytes (Fig. 6D, E). This showed that mitochondria function is important for maintenance of the karyosomes as well as formation.

During the recombination process in meiotic prophase, DSBs are formed and repaired. To test whether DSBs are formed and repaired at the right timing, ovaries were immunostained with an antibody against phosphorylated H2Av (γH2Av) that marks DSBs (Fig. 6F). In control, DSBs were formed in region 2a and fully repaired by region 3. Next, we examined RNAi of spnA/rad51 that is essential for DSB repair. As reported before in a spnA/rad51 mutant, DSBs in RNAi oocytes fully persisted until late stages of the oogenesis (at least stage 6/7). In contrast, ND-B22 RNAi oocytes showed that DSBs were formed and started disappearing at the expected timings, but the disappearance was significantly slower than control RNAi. Eighty to 90% of oocytes successfully repaired DSBs by stage 2 or 3, while 10–20% still retained DSBs. However, eventually nearly all the DSBs were repaired. This delay of DSB repair potentially activates the meiotic checkpoint pathway to prevent the formation of spherical karyosomes. However, the data described above (Table S2; Figure S1) showed that inactivation of the checkpoint failed to rescue the karyosome defects caused by ND-B22 RNAi. Therefore, there must be another cause of the karyosome defects other than just the delay of DSB repair.

Next, we examined the dynamics of the synaptonemal complex during meiotic progression. In wild type (Fig. 6G), one progenitor cell undergoes four mitotic divisions to form 16 interconnected cells. The synaptonemal complex is first assembled as filamentous structures in the nuclei of four cells in region 2a of each germarium. In two of them, including the future oocyte, the synaptonemal complex is fully assembled, while it is only partially assembled in the other two cells. Three of the cells, excluding the future oocyte, start disassembling the synaptonemal complex, and the disassembly is first completed in two of the cells by region 2b and then in the third cell by region 3. This leaves the synaptonemal complex only in the oocyte. Disassembly in the oocyte starts at stage 3 and gradually progresses (Page and Hawley 2001).

To test whether dynamics of the synaptonemal complex is affected, immunostaining was carried out using an antibody against the transverse filament protein C(3)G. In control RNAi, we found similar dynamics as observed in wild type. In ND-B22 RNAi, more cells contained the synaptonemal complex in the germarium in comparison to the same region in the control, although the number of cells with the synaptonemal complex did not exceed 4 in each region. To estimate the dynamics of the synaptonemal complex for each cell, the C(3)G pattern of each cell in a region (each cluster) was separately recorded (Fig. 6H). It showed that the dynamics of the synaptonemal complex in the most persistent cell was similar to wild-type oocytes. Disassembly of the synaptonemal complex in the other three meiotic cells was slower than the equivalent cells in wild type. This demonstrated that the dynamics of the synaptonemal complex showed little change in oocytes, although the disassembly was delayed in other cells that initially form the synaptonemal complex.

Karyosome defects related to mitochondrial dysfunction are not due to apoptosis

To gain insights into which mitochondrial function is important for the integrity of the karyosome, we analyzed the results of our screen to determine the proportion of genes with each mitochondrial function that gave karyosome defects when silenced (Fig. 7A). A significantly higher proportion of genes with electron transport chain function gave karyosome defects when silenced, in comparison to the other genes with mitochondria functions. In addition, genes involved in gene expression of the mitochondrial genome are significantly more likely to give karyosome defects. This also points to the electron transport chain, as all 13 proteins encoded by the mitochondrial genome have roles in the electron transport chain (Wolstenholme and Clary 1985). In contrast, genes involved in various metabolic processes were not significantly more likely to give karyosome defects. Although it is impossible to make a firm conclusion, this may suggest that dysfunction of the electron transport chain might cause the karyosome defects.

Fig. 7figure 7

Karyosome defects caused by mitochondria dysfunction is not mediated by apoptosis. A Frequencies of genes with different mitochondrial functions that show karyosome defects upon silencing. ** and *Significant differences (p < 0.01 and 0.05, respectively) in the frequency of genes with abnormal karyosomes, in comparison to the other genes with mitochondrial functions. B Immunostaining of apoptotic stage-8 egg chambers in wild-type oocytes after 24 h of starvation, along with healthy stage-8 egg chambers without starvation. All chromatin in ND-B22 RNAi and wild type starved for 24 h shows abnormal morphology associated with apoptosis in this figure, except in the follicle cells. Some examples are indicated by the arrowheads. Bar = 10 μm. C Frequencies of apoptosis and karyosome morphologies in stage-8 egg chambers in wild-type oocytes after different lengths of starvation, along with in ND-B22 RNAi

留言 (0)

沒有登入
gif