Abstract Study question Does the presence of breast cancer alter the ovarian response to hormonal stimulation and the cholesterol homeostasis in ovarian follicles? Summary answer The deleterious impact of breast cancer on the ovarian response might be link to an alteration of the cholesterol biosynthesis in cumulus cells. What is known already Breast cancer is the most frequent cancer in reproductive-aged women. The cancer itself seems to exert a deleterious impact on the ovarian functions. Indeed, cancer patients who undergo fertility preservation before cancer therapy initiation show a poorer response to ovarian stimulation than healthy women. Nevertheless, this impact remains controversial, and the mechanisms by which the cancer impairs the ovarian functions are poorly understood. To our knowledge, no study has yet evaluated the impact of breast cancer on the cholesterol biosynthesis pathway in human cumulus cells, a pathway essential for oocyte competence to maturation and fertilization. Study design, size, duration 30 patients with breast cancer (10 triple negative (TNBC), 10 hormone-receptor positive (HR+) and 10 HER2 positive (HER2+)) and 30 women oocytes donors undergoing oocyte cryopreservation will be included from January 2019. We assess ovarian response to stimulation (number, quality of retrieved oocytes; total FSH dose). We quantify the level of enzymes and regulators of the cholesterol biosynthesis pathway in cumulus cells (CC) by RT-qPCR. We plan to quantify cholesterol levels in follicular fluids (FF). Participants/materials, setting, methods Since January 2019, 25 breast cancer patients (9 TNBC, 8 HR+ and 8 HER2+) and 30 oocytes donors (control group) under 37 years old have been included. Ovarian stimulation was performed with antagonist protocol. Following ovarian pick up and denudation, CC and FF were collected. After total RNAs extraction from CC, the expression levels of enzymes and regulators of the cholesterol biosynthesis pathway were quantified by RT-qPCR. Cholesterol levels will be quantified in FF. Main results and the role of chance The clinico-biological characteristics (age, BMI) are comparable between cancer patients and the control group (p > 0.05). Serum AMH levels are lower in the TNBC subgroup (2.3±1.6 ng/mL) than in the control group (4.7±2.8 ng/mL, p = 0.04). A poorer ovarian response to ovarian stimulation is observed in the breast cancer group compared with the control group (number of mature oocytes: 9.4±5.2 vs. 14.5±7.7, p = 0.007). Among the cancer patients group, the TNBC subgroup shows the poorest ovarian response to stimulation (number of mature oocytes: 7.6±4.3 vs. 14.5±7.7 in control group, p = 0.006) despite the highest total administered FSH dose (2480±880 IU vs. 1831±460 IU in control group, p = 0.02). Dysregulation of the cholesterol biosynthesis pathway is observed in the breast cancer group. For TNBC and HR+ cancer patients, we observed an increase in the transcripts levels of CYP51 enzyme (respectively 3.28±1.69 vs. 1.84±2.33, p = 0.0098 and 1.77±1.78 vs. 1.84±2.33, p = 0.03) and SREBP2 activator (respectively 2.04±0.95 vs. 1.27±1.13, p = 0.003 and 1.12±0.47 vs. 1.27±1.13, p = 0.02) of the cholesterol biosynthesis pathway compared to the control group. In HER2+ cancer patients, a significant decrease in the expression levels of SQLE (0.39±0.34 vs 1.81±3.99, p = 0.001) and DHRC7 enzymes (0.33±0.32 vs 1.21±0.88, p = 0.0002) is observed compared to the control group. Limitations, reasons for caution Our preliminary results should be confirmed with the analysis of a larger number of participants. We should confirm that the observed modulations in the expression levels of enzymes and regulators of the cholesterol biosynthesis pathway led to a loss of cholesterol homeostasis by quantifying cholesterol in FF. Wider implications of the findings Our findings strengthen the hypothesis that cancer exerts a deleterious impact on the ovarian functions, as we observed a lower ovarian response in patients with breast cancer. This project will have clinical implications, such as to adapt the fertility preservation approach according to the breast cancer subtype. Trial registration number Not applicable

The presence of supernumerary centrosomes is a hallmark of cancer and is frequently observed in aggressive tumors. Cancer cells with centrosome amplification achieve pseudo-bipolar spindles through specific coping mechanisms in order to survive. However, their distribution and prevalence in cancer remain largely unknown. Here, using the NCI60 panel of cancer cell lines, we show that the presence of coping strategies correlates with centrosome amplification, with the clustering of extra-centrosomes within the two spindle poles being the most widespread mechanism. Moreover, we report an association between centrosome clustering ability and the epithelial-to-mesenchymal transition (EMT) and observe that the induction of mesenchymal characteristics in breast cancer cells with centrosome amplification promotes clustering. Furthermore, we unveil hematological malignancies, which lack epithelial characteristics, as the most proficient in centrosome clustering. Finally, we show that acute lymphoblastic leukemia is particularly sensitive to targeting clustering through inhibition of the spindle assembly checkpoint. Our study reveals how centrosome clustering and EMT collaborate to promote carcinogenesis, suggesting new possibilities to treat tumors with low epithelial characteristics, in particular leukemias.

Abstract Study question Did the use of universal vitrification/warming media have an impact on the clinical pregnancy rates in oocyte donation cycles compared with oocyte-specific vitrification/warming media? Summary answer Using universal media for oocyte vitrification and warming yields similar clinical pregnancy rates but higher oocyte survival rates than oocyte-specific media. What is known already Vitrification is now the gold standard for oocyte and embryo cryopreservation. Several commercial kits are available, some are designed for specific cell types (e.g., oocytes, zygotes, cleavage-stage embryos or blastocysts) and others are suitable for several stages, therefore termed “universal”. While the composition and the exposition protocol of stage specific media are optimized for specific cell types, “universal” media display a single composition, therefore exposition protocols need to be adapted to each specific developmental stage to ensure optimal cell survival rates. The potential use of the universal media allows to optimize the management of media in ART laboratories. Study design, size, duration This is a retrospective monocentric study. We compared the results obtained in our center for oocyte vitrification in donor cycles with the successive use of oocyte-specific vitrification/warming media (RapidVit/Warm™ Oocyte, Vitrolife) and afterwards universal vitrification/warming media (RapidVit/Warm™ Omni, Vitrolife). We analyzed 111 oocyte recipient ICSI cycles performed between March 2016 and July 2020 (86 recipient couples) and their 81 corresponding oocyte donation cycles (79 donors) with these media. Participants/materials, setting, methods Oocyte recipient ICSI cycles were divided in three groups depending on the combination of vitrification and warming media used for the oocyte vitrification and warming procedures: “specific/specific” (S/S), “specific/universal” (S/U) and “universal/universal” (U/U). The primary outcome was the clinical pregnancy rate per embryo transfer. Secondary outcomes were the oocyte survival rates, fertilization rates, cleavage and blastocyst rates, live birth and miscarriage rates. Main results and the role of chance The age of the donors (31.7+/-3.6 yo) and the recipients (35.1+/-4.5) was similar for the three groups (p > 0.05). Total FSH dose was similar for the 3 groups (1814.2+/-505UI) (p > 0.05) Clinical pregnancy rates were similar when universal vitrification/warming media were used (25.6% for S/U; 25.8% for U/U) compared with oocyte-specific vitrification/warming media (12.2% for S/S, p = 0.25). We observed higher oocyte survival rates when universal media were used (93.7% for U/U, 85.5% for S/U) compared with the use of the oocyte-specific media (75.6% for S/S, p < 0.0001). Fertilization (74.9% for U/U; 68% for S/U; 68.1% for S/S), cleavage (97.9% for U/U; 94.6% for S/U; 89.3% for S/S), and blastulation rates (46.6% for U/U; 50.3% for S/U; 41.7% for S/S) were not different between the three groups (p > 0.05). There was no difference in miscarriage (6.5% for U/U; 10.2% for S/U; 4.9% for S/S) and live birth rates (19.4% for U/U; 15.4% for S/U; 7.3% for S/S) between the 3 groups (p > 0.05). Limitations, reasons for caution The preliminary results of this retrospective study need to be confirmed by a larger prospective study. Wider implications of the findings The use of vitrification/warming universal media, which improves the management of media used for vitrification of oocytes and embryos in ART laboratory, allows to obtain similar clinical pregnancy rates for oocyte recipient donors. Trial registration number not applicable

Abstract Study question What are the optimal vitrification method (semi-automated vs. manual) and oocyte stage for in vitro -matured oocyte cryopreservation using meiosis kinetics and chromosome segregation as readouts? Summary answer The semi-automated vitrification method does not impact oocyte nuclear maturation quality compared with the manual method. Immature oocyte cryopreservation should be performed after IVM. What is known already Fertility preservation using oocyte vitrification should be performed before oncological treatments. The reference protocol consists in collecting mature oocytes after ovarian stimulation. Nevertheless, ovarian stimulation sometimes yields immature oocytes or cannot be performed (e.g., emergency oncological treatment). An IVM step is therefore required but it is not clearly demonstrated whether IVM should be performed before or after vitrification. Oocyte vitrification is usually performed with manual methods. A semi-automated vitrification device (Gavi®, Genea Biomedx) showing high performances for embryo was recently released. To our knowledge, no study has analysed its efficiency on oocyte vitrification. Study design, size, duration 200 immature oocytes collected from ICSI cycles from January 2020 will be used. Oocytes will be divided in five groups (40 oocytes/group): freshly matured oocytes (group 1 control), oocytes vitrified after IVM by a manual technique (group 2a) or by Gavi® (group 2b) and oocytes vitrified prior IVM (groups 3a and 3b). We assess oocyte nuclear maturation quality by evaluating IVM kinetics by time-lapse (Geri®) and the accuracy of homologous chromosomes segregation by CGH array. Participants/materials, setting, methods Since January 2020, 124 out of 200 immature oocytes have been included for this study. These oocytes provide from women under 37 years old without ovulatory disorder after signing an informed consent. The kinetics of meiotic resumption (germinal vesicle breakdown (GVBD) and polar body extrusion (PBE) timings), is determined by time-lapse technology (Geri®, Genea Biomedx). The accuracy of the homologous chromosome segregation during the first meiotic division will be assessed by CGH-Array. Main results and the role of chance The clinico-biological characteristics (age, BMI, smoking and total FSH dose) are comparable between the five groups (p > 0.05). No significant difference in post-thawing oocyte survival rate is observed between the two vitrification methods (semi-automated 58% vs. manual 64%). A significant difference in the overall oocyte survival rate is observed according to the stage of vitrification with a significantly higher survival rate if oocytes are vitrified at the mature stage (93% (2a+2b) vs. 61% (3a+3b), p = 0.02). The IVM rate is significantly higher if oocytes are matured freshly (86% in group 1 (fresh IVM, n = 7) and 93% in group 2a+2b (IVM before vitrification, n = 27)) compared to post vitrification (71% in group 3a+3b, n = 28), p = 0.03. The vitrification technique does not seem to impact IVM rate since it reached 64% (group 3a, n = 11) and 76% (group 3b, n = 16) (p = 0.4). GVBD and PBE timings are not significantly different between the 5 groups, suggesting that neither the oocyte stage of vitrification nor the vitrification technique affects maturation kinetics. Similarly, our preliminary results of polar bodies and oocytes chromosomal profiles assessed by CGH-Array demonstrate a similar rate of aneuploidy (monosomy or trisomy) between the groups. Limitations, reasons for caution Our preliminary results of CGH array should be confirmed with the analysis of a larger number of IVM oocytes. Wider implications of the findings Vitrification of immature oocytes should be performed by semi-automated or manual methods after IVM. Tour knowledge, this is the first study comparing the efficiency of both semi-automated and manual vitrification methods on immature and in vitro-matured oocytes. Trial registration number NCT03680937

Aim: To understand the mechanisms of centrosome amplification and their therapeutic value in cancer. Introduction: Centrosomes are the major microtubule-organising centres of animal cells. Centrosome amplification (CA) – the presence of more than two centrosomes in a cell – is a common feature in cancer1 and was recently shown to be sufficient to drive tumourigenesis.2 Recent work from the Bettencourt-Dias Lab has identified a new recurrent feature of cancer cells: centriole over-elongation, which also promotes CA. However, origins of those abnormalities and their therapeutic value remain poorly understood. Methods: We have screened the NCI-60 panel of human cancer cell lines3 for centriole number and individual length to test their frequency and interdependence. We have thereby also generated a metric capturing each abnormality level per cell line that we then correlated with the publicly available molecular (e.g. genomic, transcriptomic and proteomic) and drug-sensitivity quantitative profiles for that panel. Results: Our single-centriole analyses showed that longer centrioles are more common in cells with CA and that cells do not control their overall centriolar mass when the centriole number increases. Moreover, cancer cell lines with longer centrioles proliferate slower due to an accumulation of cells in G1 phase, suggesting that centriole length defects could lead to a cell cycle delay in G1. In addition, our original genome-wide approach highlighted putative mechanisms associated with susceptibility to both abnormalities, such as the proteasome protecting cells from CA. Correlation with drug activity identified some compounds as potential therapeutic options to selectively target cells with higher incidence of centriole abnormalities. Conclusion: This work provides the first single-centriole-level portrait of centriole abnormalities in cancer and contributes to the understanding of their molecular origins, namely by revealing novel molecular mechanisms in cell cycle biology. Given the cancer-specificity of these abnormalities, the identified compounds will inspire the development of drugs to selectively target cancer cells. Acknowledgements: This work is supported by an EMBO Installation Grant to NLBM.