Vertical transmission of maternal DNA through extracellular 1 vesicles modulates embryo bioenergetics during the 2 periconceptional period 3

30 The transmission of DNA through extracellular vesicles (EVs) represents a novel genetic 31 material transfer mechanism that may impact genome evolution and tumorigenesis. We 32 aimed to investigate the potential for vertical DNA transmission within maternal endometrial 33 EVs to the pre-implantation embryo and describe any effect on embryo bioenergetics. 34 We discovered that the human endometrium secretes all three general subtypes of EV - 35 apoptotic bodies (ABs), microvesicles (MVs), and exosomes (EXOs) - into the human 36 endometrial fluid (EF) within the uterine cavity. EVs become uniformly secreted into the EF 37 during the menstrual cycle, with the proportion of different EV populations remaining 38 constant; however, MVs contain significantly higher levels of mitochondrial (mt)DNA than 39 ABs or EXOs. During the window of implantation, MVs contain an eleven-fold higher level 40 of mtDNA when compared to cells-of-origin within the receptive endometrium, which 41 possesses a lower mtDNA content and displays the upregulated expression of mitophagy-42 related genes. Furthermore, we demonstrate the internalization of EV-derived nuclear-43 encoded (n)DNA/mtDNA by trophoblast cells of murine embryos, which associates with a 44 reduction in mitochondrial respiration and ATP production. 45 These findings suggest that the maternal endometrium suffers a reduction in mtDNA content 46 during the preconceptional period, that nDNA/mtDNA become packaged into secreted EVs 47 that the embryo uptakes, and that the transfer of DNA to the embryo within EVs occurs 48 alongside the modulation of bioenergetics during implantation. 49 50 51 52

). Overall, we demonstrate that DNA-containing EVs generated by human endometrial cells 2 4 2 become internalized by the trophectoderm of murine embryos in vitro. To assess the functional relevance of vertical endometrial EV-derived DNA transmission and 2 4 5 the impact on embryo bioenergetics, we analyzed ATP concentrations in cells of the embryo.

4 6
We co-cultured overnight hatched murine blastocysts with different human EF-derived EV absence of endometrial EVs (Control, Cnt). We found a significant reduction in ATP levels 2 5 0 following embryo co-culture with "All EVs" or EXOs compared to control embryos (both 2 5 1 p<0.001) ( Figure 5A); however, murine embryos incubated with only ABs or MVs from 2 5 2 human EF maintained ATP levels similar to control (p>0.05). These results suggest that EVs 2 5 3 (EXOs in particular) significantly impact ATP consumption/production in the embryo. The 2 5 4 observed reduction in ATP concentration after exposure to "All EVs" and EXOs indicates an 2 5 5 increase in cellular metabolic rate; therefore, embryos display ATP turnover after 2 5 6 internalizing endometrial EVs, a situation that imitates the physiological state at 2 5 7 periconception when the embryo comes into contact with the EF and EVs secreted by the 2 5 8 endometrium.

5 9
To further understand the metabolic changes produced by EVs on the embryo, we studied the 2 6 0 oxygen consumption rate (OCR) of embryos, as the manipulation of microenvironmental 2 6 1 parameters to help uterine implantation represents a well-understood first signal from the with EF-derived MVs and EXOs ( Figure 5B). We observed a reduction in the OCR of  These results suggest that EVs play a role in altering the bioenergetics of the embryo by 2 7 4 altering the metabolic rate and oxygen consumption during periconception when a move 2 7 5 towards increased glycolytic flux to support implantation is required. These results suggest 2 7 6 that the endometrium may exchange signals with the embryo to aid the creation of conditions 2 7 7 conducive to implantation. We demonstrate the existence of endometrial EVs as potential vectors for the transport of 2 8 0 DNA from the endometrium into the pre-implantation embryo; furthermore, our results 2 8 1 suggest the vertical transmission of maternal DNA to the embryo as a mechanism to 2 8 2 modulate embryo bioenergetics during the preconception period. These findings pave the way 2 8 3 for more extensive, more detailed studies that aim to decipher the exact role of EV contents 2 8 4 on embryo energetics, discover the relevance of packaged endometrial cell-derived DNA to 2 8 5 this mechanism, and further expand our knowledge regarding materno-fetal communication.

8 6
We discovered that human endometrial cells secreted the three major EV subtypes into the 2 8 7 EF during the receptive phase (phase VI) and characterized ABs, MVs, and EXOs 2 8 8 morphologically using electron microscopy, molecularly using Western blotting, and by 2 8 9 concentration and size distribution. In the case of MVs and EXOs, we identified similar 2 9 0 concentrations in the EF throughout the menstrual cycle. While other studies focused on the miRNA and protein content of EVs, we focused on the 2 9 2 DNA cargo. We sequenced the DNA content of ABs, MVs, and EXOs isolated from the EF 2 9 3 and found that all vesicles contained nDNA and mtDNA; however, MVs contained an ~11-2 9 4 fold enrichment in thirteen known mitochondrial genes (all coding for protein subunits that 2 9 5 constitute the different complexes of the electron transport chain (Chinnery and Hudson, 2 9 6 2013; Taanman, 1999)). To understand the increased encapsulation of mtDNA within EVs, 2 9 7 we quantified the mtDNA content of cells from endometrial biopsies, focusing on the 2 9 8 possible differences between the pre-receptive, receptive, and post-receptive phases of the 2 9 9 menstrual cycle. We observed a decrease in mtDNA content in endometrial cells during the 3 0 0 receptive and post-receptive phases, which occurred at the same time as the upregulation of ABs in the horizontal transfer of DNA by phagocytosis, which the authors found to integrate 3 0 5 into receiving cells and remain functional, as they can be rescued from the ABs and used by 3 0 6 somatic cells (Holmgren et al., 1999). ABs support the formation of DNA molecular hybrids 3 0 7 or hybrid chromosomes (Bergsmedh et al., 2001;Waterhouse et al., 2011) and transfer 3 0 8 oncogenes that become internalized to increase target cell tumorigenic potential in vivo 3 0 9 (Bergsmedh et al., 2001;Ehnfors et al., 2009). MVs also represent critical regulators of 3 1 0 cancer pathogenesis (Antonyak et al., 2011;Clancy et al., 2015) in addition to their 3 1 1 involvement in materno-fetal crosstalk (Tong and Chamley, 2015) and embryo self-  EVs containing DNA must first become internalized by trophectodermal cells to influence the  Considering the DNA content of EVs and their uptake by the embryo, we evaluated a 3 2 9 potential effect on embryo energetics. We hypothesized that maternal signals might regulate 3 3 0 embryo bioenergetics to aid implantation (Gardner, 2015;Gardner and Harvey, 2015). Embryo co-culture with "All EVs" (ABs, MVs, and EXOs) induced a lower ATP content 3 3 2 than unstimulated controls, suggesting that the internalization of DNA (which includes the 3 3 3 prevalent mtDNA content) may impact ATP consumption, which directly associates with the 3 3 4 metabolic programming of embryos (Gardner and Harvey, 2015). We primarily observed the 3 3 5 substantial reduction of ATP content when co-culturing murine embryos with isolated EXOs, suggesting that a combination of contents (e.g., DNA, RNA, proteins, and lipids) controls the 3 3 7 metabolic state of embryos. Additional studies will aim to fully explore each EV cargo's 3 3 8 potential contribution. We next studied the OCR of the embryos treated with EXOs and MVs 3 3 9 from the pre-receptive and receptive phases (we removed ABs from this analysis due to their 3 4 0 heterogeneity and methodological difficulties). We observed that EXOs obtained from 3 4 1 receptive but not pre-receptive endometria appeared to reduce basal respiration (while not 3 4 2 reaching significance), similar to the changes to ATP levels when treating embryos with "All 3 4 3 EVs" or EXOs. Again, future studies with more samples may reveal the real significance of  We suggest EVs as vectors for transporting nDNA and mtDNA into the pre-implantation during the periconceptional period, suggesting that the mother sends signals to the embryo to 3 4 8 aid the implantation process. menstrual cycle were collected from each donor (one from each) and classified according to 3 5 5 the menstrual cycle as phase I (early proliferative, days 0-8 (n =6)), phase II (late 3 5 6 proliferative, days 9-14 ( n = 7)), phase III (early secretory, days 15-18; (n = 7)), phase IV 3 5 7 (mid-secretory or preconception period, days 19-24; n = 36), and phase V (late secretory  The number of samples used in every experiment is specified throughout the methods section. Inclusion criteria were patients with regular menstrual cycles, body mass index (BMI) of 18- were resuspended in 50 µL of Karnovsky's fixative solution (2.5% glutaraldehyde/2% 3 9 8 formaldehyde in 0.1 M phosphate buffer, pH = 7.4), and a drop of the resulting mix was laid 3 9 9 onto a 300-mesh grid. Samples were then incubated with 2% uranyl acetate in Reynold's lead    The protein content of EV samples, supernatants obtained during EV isolation, and tissue cell   The following day, membranes were washed and incubated with secondary antibodies: goat were detected using the SuperSignal West Femto Chemiluminescent kit (Thermo Fisher 4 3 0 Scientific, Waltham, MA, USA) and a LAS-3000 imaging system (Fujifilm, Japan). (1-6,000 nm). For analysis, pellets from serial differential centrifugation steps were  EF samples for the natural menstrual cycle, phase I (early proliferative, days 0-8 (n =6)), 4 4 0 phase II (late proliferative, days 9-14 (n = 7)), phase III (early secretory, days 15-18; (n = 4 4 1 7)), phase IV (mid-secretory or preconception period, days 19-24; n = 6), and phase V (late and introduced into the NanoSight 300. To analyze the DNA cargo of the different EV populations within the human EF, EVs were 4 5 2 isolated from single EF samples in phase IV (n=10) of the menstrual cycle as described DNase treatment accounted for 56% and 57% of the total variability for ABs and EXOs, 4 6 1 respectively, in PC1 (Supplementary Figures 4A and 4B). Associated volcano plots 4 6 2 revealed abundant enrichment in coding sequences over whole genomic DNA after DNase     Raw data from pair-ended Illumina sequencing was converted into FASTQ files using 4 9 0 bcl2fastq (version 2.16.0.10). Raw counts were normalized using the TMM method from the 4 9 1 edgeR R package, and each sample was aligned to the reference genome (GRCh37) using 4 9 2 BWA (version 0.7.10). Reads with mapping quality >90% were filtered using Samtools 4 9 3 (version 1.1), and duplicates were removed with PICARD software. Insert size was retrieved 4 9 4 from filtered reads using PICARD software, and feature coverage was obtained with out. The approach used for differential DNA enrichment analysis was based on the edgeR 4 9 8 methodology. PCA graphs were obtained from log2 normalized CPM using the prcomp R were carried out to analyze the effect of reducing initial DNA concentration. All patients underwent an HRT cycle for endometrial preparation (n=70). Endometrial day of the endometrial biopsy is calculated after estrogen priming, leading to a trilaminar 5 0 5 endometrium measuring ≥ 6.5 mm after five full days of progesterone impregnation (∼120 h).

0 6
After the biopsy, the endometrial tissue was transferred to a cryotube containing 1.5 mL 5 0 7 RNAlater (Qiagen), vigorously shaken for a few seconds, and kept at 4°C or on ice for ≥ 4 h.

0 8
Determining the receptivity window used the ERA test, as previously described (Díaz- housekeeping gene for normalization. mtDNA was quantified using an mtDNA fragment 5 1 5 within the ATP8 gene. The ratio of mtDNA/nDNA was used to indicate the mitochondrial 5 1 6 copy number per cell. 100 ng total RNA was reverse transcribed using the Ion AmpliSeq Transcriptome Human 5 1 9 Gene Expression Kit following the manufacturer's protocol (Thermo Fisher Scientific).

4 8
Briefly, female mice aged 6-8 weeks were stimulated to ovulate by 10 IU intraperitoneal with 60% reaching the hatching state with excellent quality at E3.5.

8
Tagged-DNA production and EV internalization by murine embryos The human Ishikawa were grown in flasks in Modified Eagle's Medium (MEM, Gibco, 70% confluence.

6 8
Scientific) and incubated overnight to allow EdU incorporation into the DNA. The next day, the different EVs populations were isolated as described above.

7 2
Pellets containing Ishikawa EdU DNA-tagged EVs were added to good-quality E3.5 hatching 5 7 3 embryos and co-incubated overnight. Twenty embryos were used for each condition (ABs, MVs, EXOs, and Neg), for a total of eighty embryos. The supernatant of the isolation process were used as negative controls. After embryo co-culture with EVs, transferred EV-DNA was the manufacturers with some modifications. Embryos were fixed with 3.7% formaldehyde in Olympus confocal microscope using a 60X oil-immersion lens.

8 5
The Click-iT™ EdU Alexa Fluor™ 488 flow cytometry assay Kit (Thermo Fisher Scientific, following the manufacturer's instructions. After a washing step, EVs were analyzed in a Hatched mouse embryos (3.5 days) were co-culture with an ATP8 sequence (Forward:

Analysis of embryo ATP level modulation after co-incubation with human EVs
To analyze the impact of EVs in embryo bioenergetics modulation, EF samples (n=5) in 6 0 4 phase IV of the menstrual cycle were pooled, and their EV populations were isolated as 6 0 5 described above. Embryos from ten mice were obtained and incubated until day E3.5, when 6 0 6 only hatching embryos were co-incubated with isolated EF-derived EVs overnight. The EXOs and all EVs conditions, and three for fresh G2 condition media. 1 µL PBS was incubated without EVs were used as a control group (n=6; total embryos: 160), while blocked 6 3 1 embryos were used as negative control (n=3; total embryos: 120). A total of 720 murine 6 3 2 embryos were used for this experiment. Embryos were seeded at ten embryos/well with XF 6 3 3 DMEM medium containing 10 mM glucose, 2 mM L-glutamine, and 1 mM pyruvate and 6 3 4 stimulated sequentially with 1 µM oligomycin, 1.5 µM carbonyl cyanide-p- replicates per condition were used in each experiment. ZetaSizer Nano corresponding to particle concentrations for different EV populations were 6 3 9 obtained as an average value from five independent measures of random NTA fields. The  To evaluate variation in vesicle concentration throughout the menstrual cycle, concentration   The authors declare no conflicts of interest.  Youle RJ, Narendra DP. 2011. Mechanisms of mitophagy. Nat Rev Mol Cell Bio 12:9-14. Western blotting (C, F, and I). TEM images obtained using two different protocols for an protein markers analyzed by Western blotting for (C) ABs, (F) MVs, and (I) EXOs. Analyzed markers (and associated molecular mass) were calnexin (90-100 kDa), calreticulin and Kruskal-Wallis rank sum tests performed to compare the distinct menstrual cycle phases 8 3 5 -no significant differences were observed. proteins (upper panel) and for genes coding for proteins related to mitochondrial function 8 4 8