Delattre, Marie et al. (2010) C. elegans: Development and Gene Expression, EMBL, Heidelberg, Germany "Evolution of mechanisms controlling asymmetric spindle positioning in Caenorhabditis species"

Trajkovic-Bodennec, Selena, Haraghi, Aimen, Bozonnet, Noemie, Delattre, Marie

[C. elegans: Development and Gene Expression, EMBL, Heidelberg, Germany, 2010]

To further understand the mechanisms controlling spindle positioning during asymmetric cell division, we are comparing the one-cell stage embryo of different Caenorhabditis species. Despite constancy in the final phenotype (an asymmetric division with a small posterior cell), we found that nuclei and mitotic spindle movements prior to division differ between species. We are currently concentrating on C. elegans and C. briggsae comparisons in the hope to identify different (and hopefully new) equivalent solutions for the control of spindle positioning. In C. elegans, the spindle sets up in the center of the cell and is further displaced towards the posterior pole, due to an unbalanced of pulling forces acting on the asters. As the posterior pole is strongly pulled, it undergoes large tran sverse oscillations. Importantly, the mitotic spindle sets up in the anterior part of C. briggsae embryos and asters undergo limited transverse oscillations during anaphase, suggesting that a larger elongation of the spindle is taking place despite weaker pulling forces compared to C. elegans. We are trying to solve this paradox by a combination of actual measurement of forces using laser microdissection of microtubules, mathematical modeling of spindle movements and analysis of the C. briggsae orthologs of known C. elegans molecular player. We performed systematic RNAi inactivation of the genes and established several lines of GFP fused proteins to follow their dynamic localization in the early embryo of C. briggsae.

Marie Delattre et al. (2005) International Worm Meeting "SAS-1 and centriole biogenesis"

Pierre Gonczy, Marie Delattre

[International Worm Meeting, 2005]

Centrosomes are made by a pair of centrioles and a PCM (pericentriolar material). The regulation of centrosome number remains a poorly understood process. As in many species, C. elegans oocytes are devoid of centrioles, but contains the constituents of the PCM. At fertilization, the sperm provides a pair of centrioles into the oocyte. Therefore, the first centrosome is reconstituted from both maternal and paternal contributions. This centrosome will duplicate during the first cell cycle and assemble a bipolar spindle, each of the parental centriole giving rise to a new daugther centriole. Among a set of parental-effect embryonic lethal mutations on chromosome III, we identified two alleles of sas-1 which result in defective bipolar spindle assembly in early embryos. Our genetic analysis shows that this phenotype is mostly the consequence of defects in sperm cells. Serial section electron microscopy reveals that the pair of centrioles normally present in wild type cells are degraded during spematogenesis in sas-1 mutants. Consequently, mutant sperm cells provide abnormal centrioles to embryos at fertilization; these centrioles cannot duplicate normally. Therefore, sas-1 is required for centriole integrity during spermatogenesis. The molecular characterization of sas-1 is under way. Interestingly, centrosome duplication eventually resume, one or two cell cycles later, in embryos fertilized by sas-1 mutant sperm cells. This shows that functional centrioles can form, even though parental centrioles were abnormal to start with.We are currently using markers of centrioles (i.e. SAS-4, SAS-5, SAS-6) to test three scenarii: parental centrioles were repaired before templating the formation of new ones, new centrioles were formed de novo independently of old ones, abnormal centrioles could still template the formation of new centrioles, but less rapidly. We expect that the analysis of sas-1 will provide key insights into the mechanisms of centriole biogenesis.

Marie Delattre et al. (2006) European Worm Meeting "Towards A Pathway for Centriole Duplication"

Marie Delattre, Pierre Gonczy

[European Worm Meeting, 2006]

C. elegans embryos are well-suited for deciphering the process of centrosome duplication in metazoan organisms. We and others have identified five molecules as being crucial for centriole formation in the early C. elegans embryo: the kinase ZYG-1 and the coiled-coil proteins SPD-2, SAS-4, SAS-5 and SAS-6. All five proteins are enriched at centrioles and found diffusely in the cytoplasm. Moreover, SPD-2 accumulates at the PeriCentriolar Material (PCM). We set out to investigate the relationships between these five proteins to better understand how they contribute towards building centrioles. First, we addressed which protein is recruited first to newly formed centrioles. We devised conditions where oocytes express in their cytoplasm one of the five proteins fused to GFP and are fertilized by sperm containing unlabelled centrioles. In this manner, we can assess the timing of centriolar recruitment of each fusion protein. We found that SPD-2 and ZYG-1 are recruited to centrioles shortly after fertilization, during meiosis I, whereas SAS-5 and SAS-6 start to be enriched at centrioles at the end of meiosis II. By contrast, gradual SAS-4 incorporation occurs during the subsequent S phase. We next tested whether this differential timing of recruitment reflects related epistatic relationships. Accordingly, we found that in the absence of SPD-2, the centriolar localization of the four other proteins is abolished, while SPD-2 localization remains unchanged in the absence of either ZYG-1, SAS-4, SAS-5 or SAS-6. Our results also indicate that even though ZYG-1 controls centriolar localization of SAS-5 and SAS-6, there is an unsuspected feedback loop from SAS-5 and SAS-6 that modulates ZYG-1 distribution. Together with previous observations, these findings lead us to propose the following working model for a pathway of centriole duplication in C. elegans. Initially, SPD-2 is loaded to centrioles at the very onset of the duplication cycle. This leads to centriolar recruitement of ZYG-1, which in turns allows SAS-5 to deposit SAS-6 onto the mother centriole or a closely associated structure. The presence of SAS-6 in turn triggers SAS-4 recruitment and formation of a new centriole.

Marie Delattre et al. (2007) International Worm Meeting "Evolution and Robustness of the First Asymmetric Division in Nematode Embryos."

Marie-Anne FELIX, Marie Delattre

[International Worm Meeting, 2007]

The first embryonic division of most nematode species is asymmetric. In C. elegans , the sperm centrosome breaks the symmetry of the unpolarized oocyte cortex at fertilization. However, parthenogenetic species, as well as some gonochoristic (male/female) species do not rely on the sperm to polarize their oocytes (Goldstein 98). This shows that during nematode evolution, a variety of mechanisms have thus appeared, all leading to an apparently similar first embryonic division. We propose to search for the modifications of this process at different evolutionary distances. 1) Between distantly related species: We are looking for the polarity cue in two parthenogenetic species of the Panagrolaimidae family. We are also investigating the role of the centrosome in polarity establishment for a pseudogamous species, in which the sperm contributes centrosomes but no genetic material to the zygote. 2) Amoung Caenorhabditis species: Our preliminary observations suggest that some early embryonic events vary among species of the Caenorhabditis genus. Even though all these species rely on the sperm to polarize the zygote cortex, the mechanisms translating this polarity cue into a first asymmetric division may have evolved. Some of the species show phenotypes partly resembling known C. elegans mutant phenotypes. We are currently testing the role of a few candidate genes, by an RNAi approach and by immunolocalisation studies in these species. In parallel, bioinformatic analyses are under way to estimate the conservation or divergence of key proteins of the polarization process. 3) Within C. elegans: No differences in embryonic polarity are visible between wild C. elegans strains. To reveal cryptic genetic variations between the different genetic backgrounds we need to abolish the buffering of the system. First, embryos will be subjected to extreme environmental conditions such as high or low temperatures and will be collected from starved or old mothers. Second, mutations of candidate genes will be introduced, from the N2 background, by repeated backcrosses into several genetic backgrounds. This should reveal the relative quantitative contribution of the proteins. We are particularly interested in testing the evolution of the role of PAR-2, since PAR-2 can be experimentally dispensable in N2 as long as the activity of the anterior PAR complex is reduced (Watts, 96). Ultimately, this will help to assess the robustness of the polarization process against environmental and genetic perturbations in C. elegans .

Marie Delattre et al. (2002) European Worm Meeting "sas-1, a gene required for centrosome duplication in one-cell stage C.elegans embryos"

Pierre Goenczy, Marie Delattre

[European Worm Meeting, 2002]

Centrosome duplication is essential for the proper division of animal cells, yet remains a poorly understood process. We have started a cellular and molecular dissection of centrosome duplication in C. elegans. In wild-type one-cell stage C. elegans embryo, the sperm contributes the only pair of centrioles; the single centrosome then duplicates shortly after fertilization and forms a bipolar spindle. Among a large set of parental-effect embryonic lethal mutations on chromosome III (Gnczy & al. JCB 1999), we identified mutations in two loci, sas-1 and sas-2, that result in defective bipolar spindle assembly in early embryos. Here, we focus on our analysis of two alleles of sas-1. We performed immunofluorescence on fixed specimens using antibodies against a-tubulin, the pericentriolar material component ZYG-9 and the presumptive centriolar component PLK-1, to further characterize the sas-1 mutant phenotype. Strinkingly, this analysis revealed that one-cell stage embryos possess a single centrosome, which nucleates microtubules correctly but does not duplicate. As a result, a bipolar spindle does not assemble, leading to failure in chromosome segregation and cytokinesis. These observations were confirmed using time-lapse fluorescence microscopy with transgenic animals carrying -tubulin::GFP and -tubulin::GFP fusion proteins. Moreover, we showed that progression through S phase is not affected in sas-1 mutants, suggesting that the failure of centrosome duplication is not a consequence of a more general cell cycle defect. Ultrastructural analysis is under way to understand which precise step of the centrosome duplication cycle is affected in sas-1 mutant embryos, Using a combination of time-lapse DIC and fluorescence microscopy, we found that the centrosome never duplicates in 20% of sas-1 mutant embryos. In contrast, while the centrosome fails to duplicate during the first cell cycle, it does duplicate at the onset of the second or the third cell cycle in the remaining 80% mutant embryos. These observations suggest that sas-1 function may be required strictly early during development. These two phenotypes are reminiscent of those that can be observed in zyg-1 mutant embryos (O'Connell & al. Cell 2001). Interestingly, the authors showed that the first duplication of centrosomes in the C. elegans embryo is under the control of paternally contributed zyg-1, whereas subsequent rounds of duplication are under the control of maternally contributed zyg-1. Taken together, these findings raised the possibility that sas-1 is a paternal gene. To test this hypothesis we crossed wild-type males with homozygous mutant sas-1 hermaphrodites and showed that the cross-progeny is 100% viable for both alleles of sas-1. These results establish that the two alleles studied strictly affect a paternal contribution of sas-1. We have initiated the cloning of sas-1 using SNP mapping and have narrowed the search down to 36 candidate genes. Because RNAi does not appear to work in sperm, and because sas-1 is most likely a paternal gene, we have initiated rescue experiments using cosmids and YACs covering the region of interest to identify the molecular nature of sas-1.

Marie Delattre et al. (2001) International C. elegans Meeting "Microevolution of vulval cell lineages within two nematode genera Caenorhabditis and Oscheius."

Marie-Anne FELIX, Marie Delattre

[International C. elegans Meeting, 2001]

The cell lineage of Caenorhabditis elegans and some other nematodes is mostly invariant. Given this invariance, one can wonder how a cell lineage can evolve. We have started a microevolutionary approach by observing lineage variations of vulva precursor cells (VPCs) in different natural nematode populations : 22 strains of the C. elegans species and 5 different species in the Caenorhabditis genus, as well as 38 strains belonging to 9 different species of the Oscheius genus. Our results show that, within both genera, lineage variations between species and even between strains of the same species occur mostly for VPCs that do not participate to the vulval invagination (VPCs with a 3deg non-vulval fate). These cells are probably not under strong selection pressure and this could explain the large variations observed at a small evolutionary timescale. In the reference strain C. elegans N2, the lineage of the P3.p cell is not invariant (Sulston and Horvitz, 1977). In 50% of N2 animals, P3.p is a true VPC : it is competent to form the vulva and divides once before fusing with the surrounding hypodermis. In the other 50%, P3.p is not competent, does not divide and fuses with the hypodermis. The percentage of occurrence of P3.p division varies between strains of the C. elegans species (from 10% to 60%) and between species of the Caenorhabditis genus (from 0% to 100%). In the C. elegans N2 strain, P3.p division is correlated with the competence of the cell. However, the correlation does not hold in the C. elegans strain CB4857, where P3.p divides in only 15% of the animals whereas it is competent in about 60% (as determined by ablation of P(4-8).p in the L1 stage). P3.p competence varies between Caenorhabditis species, suggesting that the size of the competence group has evolved between closely related species, but that the program of cell division can vary even within a species. Species of the Oscheius genus can be easily found in soil samples from around the world (much more easily than Caenorhabditis species). Variations mostly concern P4.p and P8.p lineages (which adopt a 3deg non-vulval fate in these species) : these cells divide twice, once, or not at all, but are always competent to replace the vulval cells P(5-7).p. In this genus too, intra-species polyporphism in cell lineages is amplified between closely related species. Thus, cell lineage variations observed between species could be due to fixation of variants that segregate within species. Variations observed between strains are larger within the species Oscheius sp.1 than in C. elegans . Therefore, we have undertaken a genetic analysis between two pairs of strains of Oscheius sp. 1 that show distinct P4.p and P8.p lineages. For each pair, at least three loci are involved in the phenotypic differences. We also find that variation at one locus has a relatively strong effect on the phenotype (instead of small additive effects of many genes). We have also detected "errors" at a very low frequency (0.1-1%) in the fate and lineage of cells which do form the vulva. These frequencies give the level of precision of vulval patterning mechanisms and point to a selection pressure for maintenance of a large vulval equivalence group. Such "errors" are more frequent in the Oscheius group than in Caenorhabditis species ; thus the robustness in vulval patterning may be stronger in the latter genus.

Valfort, Aurore-Cecile et al. (2014) Evolutionary Biology of Caenorhabditis and Other Nematodes "Evolutionary comparisons of mitotic spindle dynamics"

Valfort, Aurore-Cecile, Delattre, Marie, Needleman, Dan, Farhadifar, Reza

[Evolutionary Biology of Caenorhabditis and Other Nematodes, 2014]

Cellular mechanisms are essential to translate molecular information into phenotypes. Although genomic, molecular, development and phenotypic evolution are largely explored; there is currently little understanding on the evolutionary dynamics of cell biological processes. We chose to analyze the mitotic spindle in an evolutionary context. The mitotic spindle is a microtubule-based structure that generates mechanical forces to segregate chromosomes apart and find its position within the cell during anaphase. In the one-cell C. elegans embryo, the spindle undergoes very reproducible movements leading to its elongation and asymmetric positioning towards the posterior side of the cell. Consequently, the first embryonic division is asymmetric in size. We recorded the first asymmetric embryonic division of 42 different species of the Rhabditidae family (corresponding to 111 different strains). We quantified spindle dynamics in details by measuring ~25 different cellular parameters, such as spindle size, cell size, centrosome motion, etc. We found very different spindle movements depending on the species observed, despite a conserved final asymmetric position. We also found some clear intra- species variations. Therefore, we reveal cryptic evolutionary changes in the regulation of mechanical forces behind a conserved cellular process. We found that transverse oscillations of the spindle, as found in C. elegans, were observed only in species from the Caenorhabditis genus, except C. sp. 1. Otherwise, the majority of parameters varied randomly on the phylogeny and did not show any directionality of changes. We will present our final results that allow us to 1) describe the diversity of cryptic changes of an fundamental cellular structure and 2) identify the mechanisms that are either flexible or highly constrained. At the same time, such comparative approach allows us to uncover essential properties of the spindle mechanics shared between species 1. Ultimately, we want to identify the compensatory mechanisms allowing the system to maintain its final output despite the accumulation of cryptic changes.1. Riche & al. "Evolutionary comparisons reveal a positional switch for spindle pole oscillations in Caenorhabditis embryos". Journal of Cell Biology. 2013: 201, 653-662.

Blanc, Caroline et al. (2021) International Worm Meeting "Meiosis modifications at the origin of asexuality in Mesorhabditis pseudogamous nematodes"

Saclier, Nathanaelle, Galtier, Nicolas, Rey, Carine, Delattre, Marie, Blanc, Caroline

[International Worm Meeting, 2021]

Sexuality is widespread and asexuality is a derived character. Despite the recognized costs associated with sex, asexuals remain rare, which constitutes one of the most intriguing paradox in evolutionary biology. To understand the adaptative value of sex, there is clearly a need for further exploration of the asexual world, and in particular of the transition to asexuality. We recently characterized the nematode genus Mesorhabditis, in which regular sexual species are found, as well as species featuring progressive loss of males and sexuality. These pseudo-sexual species are composed of 90% females and only 10% males. Males and sperm are needed for most eggs to develop by gynogenesis (the sperm is needed to activate the oocytes but its DNA is not used, i.e pseudogamy). In the few cases where the sperm DNA is incorporated, the eggs develop as males, because only the Y-bearing sperm of males are competent to fertilize (Grosmaire & al. Science 2019; Launay & al. BMC Evol Bio 2020). In this intriguing system, females are thus produced asexually whereas males are produced sexually. Mesorhabditis nematodes represent an ideal system to study the evolutionary consequences of transition to asexuality in closely related species and to explore the molecular origins of asexuality. There are many ways to lose sex. Here we asked which modifications to meiosis led to the production of diploid oocytes in asexuals. Using cytological descriptions, we found that meiosis of asexual females is similar to sexual species up to anaphase I. Chromosome pairing and crossing-over occur and bivalent forming chiasmata are present in diakinesis. Next, the first meiotic spindle forms and bivalent chromosomes initiate their segregation. However, chromosome segregation eventually stops and all univalents realign at metaphase of meiosis II. Meiosis II then proceeds normally with the segregation of sister chromatids and formation of the single polar body. Hence, meiosis in these asexuals is characterized by the maintenance of recombination and the assortment of non-sister chromatids in the diploid oocyte. In parallel, we are comparing the level of recombination in the sexual and asexual species of this genus. For that, we have sequenced the genome of a dozen strain for one sexual and one asexual species, to measure linkage desequilibrium. The theoretical consequence of such modified meiosis is genome-wide homozygosity, except maintenance of heterozygosity on chromosome centers. This will be contrasted with the actual measure of heterozygosity.

Riche, Soizic et al. (2013) International Worm Meeting "Evolutionary comparisons reveal a positional switch for spindle pole oscillation, and divergent regulation of GPR in Caenorhabditis embryos."

Delattre, Marie, Arneodo, Alain, Pecreaux, Jacques, Zouak, Melissa, Riche, Soizic, Argoul, Francoise

[International Worm Meeting, 2013]

Mitotic spindle positioning is essential for oriented cell division. During the first embryonic division in C. elegans, the mitotic spindle is pulled towards the posterior pole of the cell and undergoes vigorous transverse oscillations. We identified variations in spindle trajectories by analyzing the outwardly similar one-cell stage embryo of its close relative C. briggsae. Compared to C. elegans, C. briggsae embryos exhibit an anterior shifting of nuclei in prophase and reduced anaphase spindle oscillations. By combining physical perturbations and mutant analysis in both species, we show that differences can be explained by inter-species changes in the regulation of the cortical Ga/GPR/LIN-5 complex. However, we uncover that in both species 1) a conserved positional switch controls the onset of spindle oscillations, 2) GPR posterior localization may set this positional switch, and 3) the maximum amplitude of spindle oscillations is determined by the time spent in the oscillating phase. By investigating microevolution of a subcellular process, we therefore identify new mechanisms that are instrumental to decipher spindle positioning (1). Interestingly, GPR is poorly conserved at the amino acid level between these species. We are currently analyzing this unexpected divergence. To this end, we are performing protein replacement between species, as well as analysis of protein chimeras. By correlating sequence divergence, GPR localisation and phenotypes of nuclei and spindle positioning we hope to identify new regulatory modules of GPR, an essential but poorly caracterized player in the control of spindle positioning. (1) Riche, S. & al. Evolutionary comparisons reveal a positional switch for spindle pole oscillation in Caenorhabditis embryos. Journal of Cell Biology, 2013 In press.

Laure BONNAUD et al. (2003) International Worm Meeting "Evolutionary variations in P3.p division and competence within the Caenorhabditis genus and the C. elegans species"

M-Anne FELIX, Laure BONNAUD

[International Worm Meeting, 2003]

In order to study how a cell lineage varies during evolution, we use a microevolutionary approach with the nematode vulva as a model system. The vulva is formed by precursors in the ventral epithelium, called the Pn.p cells, and each Pn.p cell has a specific fate. The same vulva lineage characters that diverge between closely related species of Caenorhabditis are also polymorphic between strains (Delattre & Felix, 2001). The frequency of P3.p division varies within the Caenorhabditis genus (between species) and within C. elegans (between strains). In order to understand the processes that are responsible for variation in P3.p division frequency, our aims are to define the number of genes involved and to understand the molecular and cellular mechanisms that underlie the vulva lineage variations: 1)We performed a genetic analysis of P3.p division frequency between divergent strains of C. elegans. Recombinant inbred lines between the N2 and CB4857 strains and between the N2 and CB4856 strains (the latter set a kind gift of M. Hammarlund & E. Jorgensen) were analyzed for P3.p division frequency. Non-parental (intermediate or transgressive) levels for the lineage character were observed, showing that the P3.p division polymorphism between two strains is due to variations at several loci. 2)Using Single Nucleotide Polymorphisms in relation to the phenotype of the N2xCB4857 lines, we could not detect a major locus that could be at the origin of the difference; we could rule out candidate loci such as lin-39.3)P3.p is competent to form vulval tissue after P(4-8).p ablation in C. elegans, but not in C. briggsae (Delattre & Felix, 2001). The frequency of P3.p division is not fully correlated with its competence. We are analyzing more strains and species in order to draw phylogenetic hypotheses on the polarity of changes in P3.p division and competence.4) The frequency of P3.p division may be regulated by the (non)-fusion of the cell in the L2 stage.Therefore, we are studying the frequency and timing of P3.p fusion using the AJM-1::GFP marker, and will compare it to the frequency of its division. The fusion process could have consequences on the observed level of P3.p competence. The understanding of the relationships between the processes of division, fusion and competence could help understand the evolution of the size of the competence group from the intraspecific to the supraspecific level.