Abstract Study question What happens to the sperm tails that are no longer needed after the goal of fertilization has been achieved? Summary answer The sperm tail lost its fibrous sheath and changed to a coiled morphology with a whisker root-like structure at the tip during the early post-fertilization. What is known already The sperm tail has been thought to play only a role in allowing the head and neck to reach the oocyte. The flagellum is found in the mammalian sperm tail and consists of a middle, main, and terminal part. It also contains an axon composed of two central microtubules surrounded by nine peripheral microtubule doublets (9d + 2s microtubule pattern) surrounded by an accessory structure called the fibrous sheath (FS), but few studies have been conducted on the post-fertilization fate of the mammalian sperm tail. Study design, size, duration To investigate sperm tail structure from the pronuclear stage to the second cleavage, we used abnormally fertilized oocytes derived from either conventional in vitro fertilization (cIVF, n = 41) or intracytoplasmic sperm injection (ICSI, n = 11). In this study, abnormally fertilized oocytes were collected between March 2022 and November 2023 and subjected to immunofluorescence analysis and field emission scanning electron microscopy (FE-SEM) analysis. Participants/materials, setting, methods Under ethical approval, abnormally fertilized oocytes were donated by infertile couples undergoing cIVF or ICSI cycles at the Yamashita Shonan Yume Clinic in Japan. Informed consent was obtained from all participants. Immunofluorescence assay was performed to analyze sperm tail morphology using antibodies against H3K27me3, α-tubulin, and AKAP4. FE-SEM analysis was performed to detect the 9d + 2s structure characteristic of the sperm tail axon. Main results and the role of chance Tri-pronuclear zygotes derived from conventional in vitro fertilization (cIVF-derived 3PN) had two male pronuclei, whereas those derived from intracytoplasmic sperm injection (ICSI-derived 3PN) had one. Each zygote had the same number of coiled microtubule structures as the male pronucleus, and the 9d + 2s structure characteristic of the axon was identified, suggesting a sperm tail. No difference in sperm tail shape was observed between ICSI-derived 3PN and cIVF-derived 3PN. AKAP4, which is specifically localized in the FS associated with the principal piece of the flagellum, was observed in sperm tails attached to the zona pellucida but not in those in the oocyte, suggesting that the FS of the flagellum is lost before the pronuclear stage, leaving the microtubules of the axonemes exposed. The sperm tail was attached to one of the spindle poles during the first cleavage of abnormally fertilized oocytes, but no differences in the structure of the spindle pole with or without the sperm tail were observed. The tip of the sperm tail was branched like a “whisker root”, suggesting that the sperm tail was degraded from the end, but the degree of tip branching did not differ from the pronuclear stage to the second cleavage. Limitations, reasons for caution Normal fertilized oocytes were not used in this study because they are used for clinical treatment. The cell biological significance of sperm tail branching is currently unknown and requires further experimentation. Wider implications of the findings These findings shed light on the processes that occur during human fertilization, suggest new roles for the sperm tail, and may have implications for assisted fertilization approaches in the future. Trial registration number not applicable

Abstract Study question Is it possible to apply RNA-sequencing to assisted reproductive technology (ART) to predict blastocyst quality? Summary answer We identified 14 commonly identified differentially expressed genes by grouping blastocysts according to three distinct parameters such as developmental potential, maternal age, and Gardner score What is known already In ART, while the selection of a suitable embryo for transfer is critical for a successful pregnancy, neither preimplantation genetic testing nor morphological and chronological evaluation of the embryo can fully guarantee a successful pregnancy. Recently, transcriptional events in early human embryonic development have been analyzed using RNA-sequencing (RNA-seq) and researchers are attempting to apply this information to ART. We have reported that 96 differentially expressed genes (DEGs) were identified using RNA-seq of each inner cell mass (ICM) and trophectoderm (TE) in blastocysts classified according to the developmental potential which correlates with pregnancy rate at the ESHRE 37th annual meeting. Study design, size, duration After retrospectively analyzing 1,890 cases undergoing freeze-thaw blastocyst transfer from March 2018 to December 2020 to examine the correlation between blastocyst developmental potential and pregnancy rate, a total 13 blastocysts cryopreserved between February 2011 and September 2018, then scheduled for disposal and with consented, were subjected to RNA-seq to identify genes associated with pregnancy expectation. RNA-seq data were then examined whether common DEGs could be found when classified by maternal age and Gardner score, respectively. Participants/materials, setting, methods Blastocysts were donated by infertile couples undergoing c-IVF or ICSI cycles at the Yamashita Shonan Yume Clinic with informed consent under ethical approval. TE cells and ICM cells were collected from blastocysts classified by developmental potential and subjected to RNA-seq to identify DEGs. In addition, RNA-seq data were regrouped by maternal age and Gardner score to find common DEGs. DEGs (q-value < 0.01) were identified using the R package “DESeq2” (version 1.32.0). Main results and the role of chance When the RNA-seq data obtained from blastocysts classified according to pregnancy expectation were re-grouped by maternal age and re-analyzed, we identified 7 and 17 genes that were down- and up-regulated in the elder group, respectively, in ICM. In TE, 2 and 12 genes were down- and up-regulated in the elder group, respectively. On the other hand, when re-grouped by Gardner score ( Limitations, reasons for caution Although we established an expected pregnancy rate concerning the degree of blastocyst development from retrospective clinical outcomes and used it as a surrogate marker for assigning biopsied blastocysts to different analysis groups, it remains unknown whether the gene expression profiles accurately reflect the pregnancy outcomes. Wider implications of the findings UCHL1 expression was commonly increased with lower blastocyst developmental potential, higher maternal age, and lower Gardner score. Since UCHL1 has been reported to be essential for blastocyst development in mice, our results suggest that UCHL1 may also be a marker of blastocyst quality in humans. Trial registration number not applicable

Abstract Study question Is it possible to identify the molecular factors that contribute to the implantation potential of blastocysts? Summary answer Genes correlated with expected pregnancy rate in trophectoderm (TE) and inner cell mass (ICM) respectively were identified, and aneuploidy alone couldn’t predict the pregnancy expectation. What is known already The selection of suitable embryos for transfer is critical for achieving successful pregnancy outcomes in assisted reproductive technology (ART). Although pre-implantation genetic testing for aneuploidy (PGT-A) as well as morphological and chronological evaluation of embryos, have been conducted in clinical practice, they do not fully guarantee successful pregnancy. Recently, transcriptional events in early human embryonic development have been analyzed using RNA-sequencing (RNA-seq) and researchers are attempting to apply this information to ART. Study design, size, duration To determine the correlation between blastocyst evaluation and pregnancy rate, we retrospectively analyzed 1,890 cases underwent frozen-thawed blastocyst transfer from March 2018 to December 2020. A total of 13 blastocysts that were cryopreserved for clinical use between February 2011 and September 2018, then scheduled for disposal and with consented for research, were subjected to RNA-seq without distinguishing between conventional in vitro fertilization (c-IVF) and intracytoplasmic sperm injection (ICSI). Participants/materials, setting, methods Blastocysts were donated by infertile couples undergoing c-IVF or ICSI cycles at the Yamashita Shonan Yume Clinic with informed consent under ethical approval. TE and ICM cells were collected from blastocysts by using a micromanipulator and then subjected to RNA-seq. Gene expression analysis and digital karyotyping using RNA-seq were performed simultaneously for TE and ICM cells, respectively. One-way analysis of variance, chi-square test and Tukey's multiple comparison test were used for this study. Main results and the role of chance Blastocysts were classified into three groups to correlate with pregnancy rates based on the diameter of the blastocyst and the time to reach this size: those taking less than 130 h to reach a diameter of > 170 μm (Group 1, n = 676), those taking more than 140 h to reach a diameter of < 180 μm (Group 2, n = 158), and the rest (Group 3, n = 1,056). The pregnancy rates of Groups 1, 2 and 3 were 59.0%, 16.5%, and 34.2%, respectively (p < 0.01). Assessing the differences in overall transcripts correlated between Group 1 (n = 5), Group 2 (n = 4), and Group 3 (n = 4), 26 and 67 differentially expressed genes (DEGs) were identified in ICM and TE cells, respectively. Importantly, downregulated genes in TE of blastocysts with lower expectation of pregnancy included tight junction-related genes, such as CXADR, CLDN10, and ATP1B1, which were implicated in peri-implantation development. Digital karyotyping revealed karyotypic abnormalities and mosaicism in all groups with no common abnormalities observed, suggesting that aneuploidy alone cannot predict the pregnancy expectation. Limitations, reasons for caution Although 93 genes potentially related to implantation have been identified, it is still unclear how these genes are involved in implantation. In vitro implantation models using human embryos and artificial embryos currently under development are expected to contribute to the elucidation of the functions of these genes. Wider implications of the findings Our results provide reliable candidates for genes that could allow for non-invasive selection of high-quality blastocysts for ART and add to the knowledge base of transcriptional events in human peri-implantation development. Trial registration number not applicable

The aim of this study is to analyse the association between blastocyst diameter and TVs development, and to examine the efficacy of AH. Blastocysts with a diameter of more than 170 μm leads to high incidence of TVs and AH applied from the incidence should be effective. TVs are protrusion of trophectoderm cells often observed in expanding blastocyst stages. TVs can be observed in expanding blastocysts regardless of Intracytoplasmic sperm injection (ICSI) and Conventional-IVF (C-IVF), when the internal pressure of blastocysts increase. The rate of TVs incidence in blastocysts inseminated by ICSI is higher than that by C-IVF, due to penetration of the needle into the zona pellucida. Moreover, it has been reported that TVs may inhibit blastocyst hatching. However, the developmental timing of TVs is still unclear, and there is no study that has analysed the association between blastocyst diameter and the incidence of TVs. 1) Diameters and TVs incidence of blastocysts by ICSI and C-IVF were measured, and the cut-off value and the area under the curve (AUC) of the receiver operating characteristic (ROC) curve were calculated to estimate the timing of TV incidence. 2) We analysed the clinical pregnancy rates of blastocysts with TVs treated by AH compared to those of blastocysts by C-IVF not subjected to AH. This study included 821 transferred frozen blastocysts ranging from March 2018 to November 2019. The embryos were cultured in a dry incubator after insemination by ICSI or C-IVF. Blastocyst freezing conditions were set at day5 to day7 with a diameter of more than 150 μm in inner diameter of zona pellucida, and this was measured before freezing. The ROC curve was performed using EZR statistical analysis software. 1) The incidence of TVs in blastocysts by ICSI and C-IVF was 27.5% (117/424) and 14.6% (58/397) respectively. The rate of the incidence of TVs in blastocysts inseminated by ICSI and C-IVF; 8.6% (12/140) and 0.95% (1/105) in 150–159 μm, 12.7% (14/110) and 8.2% (6/73) in 160–169 μm, 40.6% (28/69) and 10.5% (6/57) in 170–179 μm, 55.6% (30/54) and 25.5% (13/51) in 180–189 μm, 66.7% (20/30) and 35.7% (10/28) in 190–199 μm, and 68.4% (13/19) and 26.8% (22/82) in the diameter of more than 200 μm. The cut-off value of the ROC curve was respectively 170 μm (sensitivity 78.6% and specificity 73.0%) and 176 μm (sensitivity 84.5% and specificity 59.6%) in the diameter; the AUC was 0.8 [95%CI:0.752–0.848] and 0.74 [95%CI:0.687–0.793] respectively. 2) The clinical pregnancy rate of TVs blastocyst vs C-IVF blastocyst was 52.7% (88/167) vs 57.8% (37/64) respectively. There is no significant difference between the two clinical pregnancy rates (P = 0.556). The findings of this study have to be seen in light of some limitations. Since this study aimed to analyse the incidence of TVs based on blastocyst size, we did not take into account the grade according to the Gardner classification and the number of trophectoderm cells. Wider implications of the findings: Blastocysts inseminated by ICSI and C-IVF were highly likely to have TVs above 170 μm and 176 μm respectively. The clinical pregnancy rates of the blastocyst with TV treated by AH was similar to those of the C-IVF blastocyst. Not applicable

Abstract Study question Why do multinucleated blastomeres appear at high frequency in two-cell-stage embryos in humans? Summary answer Failure in microtubule assembly during the first mitotic spindle body formation by sperm centrosome-dependent microtubule organising centres (MTOCs) may lead to chromosomal instability. What is known already Unlike that in mice, multinucleated blastomeres appear at high frequency in two-cell-stage embryos in humans. However, the underlying mechanism remains elusive. In mice, multiple acentriolar MTOCs appear around the male and female pronuclei after pronuclear disappearance and contribute to dual-spindle formation, engulfing each parental chromosome. This spindle formation may ensure an error-free division, keeping the chromosomes stable during the first cleavage, as observed in mice, but it is unclear whether a similar mechanism exists in humans. Study design, size, duration To examine how sperm centrosomes contribute to MTOC formation in humans, two types of 3PN zygotes derived fromeither conventional in vitro fertilization (c-IVF, n = 30) or intracytoplasmic sperm injection (ICSI, n = 10) were used. The zygotes were collected from October 2018 to January 2020. MTOC and mitotic spindle formation at consecutive stages of development during the first cleavage were analysed under static and dynamic conditions using immunofluorescence assay and fluorescent live-cell imaging. Participants/materials, setting, methods Under ethics approval, 3PN zygotes were donated by infertile couples undergoing c-IVF or ICSI cycles at the Yamashita Shonan Yume Clinic in Japan. All participants provided informed consent. Immunofluorescence assay was performed using antibodies against α-tubulin, pericentrin, and H3K9me3 after fixation with MTSB-XF solution. Fluorescent live-cell imaging was performed using TagGFP2-H2B mRNA (chromosome marker) and FusionRed-MAP4 mRNA (microtubule marker). Main results and the role of chance Immunofluorescence revealed that while 3PN zygotes derived from c-IVF showed four pericentrin dots, those derived from ICSI exhibited two pericentrin dots. In pro-metaphase, an independent group of chromosomes derived from each pronucleus and MTOCs were formed by the sperm centrosome at the core. Microtubules from each MTOC extended toward the chromosomes in the early metaphase; a quadrupolar spindle was formed in the c-IVF-derived zygotes, and a bipolar spindle was formed in the ICSI-derived zygotes by the MTOCs at the zygote apex after chromosome alignment. In pro-metaphase, the microtubules extended from the MTOCs to the nearest chromosome. Since microtubule assembly was found on oocyte-derived chromosomes, we hypothesised that whether a chromosome is surrounded by microtubules depends on the location of the MTOCs, irrespective of its origin. Live-cell imaging of histone H2B and MAP4 revealed that four MTOCs appeared around the three pronuclei just before the disappearance of the pronuclear membrane; microtubules then extended from the MTOCs toward the chromosomes, beginning to form a mitotic spindle as the chromosomes moved to the centre of the oocyte. Interestingly, one of the three assembled chromosome groups showed no microtubule assembly in the pro-metaphase. Similar results were obtained in all six 3PN zygotes subjected. Limitations, reasons for caution We demonstrated the high risk of developing bare chromosomes not surrounded by microtubules during the formation of the first mitotic spindle, using human tripronuclear zygotes. However, owing to unavailability of normal fertilized oocytes for this study because of the clinical use, we were unable to confirm this in normal zygotes. Wider implications of the findings Although two sperm centrosome-dependent MTOCs are expected to be formed in normal fertilized oocytes, these MTOCs are not sufficient to completely enclose physically separated female and male chromosomes with the microtubules. This explains the high frequency of zygotic division errors that lead to unstable human chromosomes. Trial registration number not applicable

The radiotherapy position (RP) (neck flexion posture) could be the burden for patients in acquisition of positron emission tomography/computed tomography (PET/CT) images, which are utilized for more accurate delineation of gross tumor volumes (GTVs) of nasopharyngeal carcinoma (NPC). The PET/CT imaging in diagnostic position (DP) may be beneficial for many patients as well as radiation oncological staffs if they are applicable to delineation of GTVs. The purpose of this study was to evaluate the delineation accuracy on DP-PET/CT images for NPC, compared with RP images, based on an observer study. Fourteen patients with NPC without distant metastasis, who underwent radical chemoradiotherapy were selected for this study. 18F-fluorodeoxyglucose (18F-FDG) PET/CT images were acquired with two-phase scans at 60 (early scan) and 120 minutes (delayed scan) after FDG injection. The early scan was performed in DP on a curved diagnostic table, and the delayed scan was carried out in RP with patient fixation using a thermoplastic shell on a flat tabletop, which was the same positioning as the planning CT (pCT) images. PET/CT and pCT images in DP and RP were registered together using rigid image registration (RIR) and deformable image registration (DIR), respectively, in an oncology imaging informatics system based on the consensus between a radiation oncologist and a medical physicist. Since the neck flexion angles in DP and RP were different, a volume of interest (VOI) in RIR was set the head area in order to match the primary tumor regions between PET/CT and pCT images. In DIR, VOI was set from the top of the head to the tracheal bifurcation. Four experienced radiation oncologists individually delineated GTVs (primary tumors and lymph node metastases), i.e., GTVRIR and GTVDIR, on the pCT images registered with DP-PET/CT images. The interval terms between the first and second delineations of GTVRIR and GTVDIR contours were more than two weeks. Then, more than two weeks later after the two delineation, each reference GTV (GTVREF) was independently delineated by the four oncologists on pCT images registered with RP-PET/CT images using RIR. Dice’s similarity coefficients (DSCs) of all combinations between the GTVRIR vs. GTVREF, and the GTVDIR vs. GTVREF were compared with taking into account the interobserver variability in the GTVREF contours. The average DSCs for GTVRIR and GTVDIR were 0.771 (p>0.05; paired t-test), which may be acceptable for their delineations. DP-PET/CT images using RIR could be comparable with DP images using DIR, and thus they may be applicable to the delineation of GTVs for NPC. Therefore, the burden for patients may be mitigated in treatment planning.