is a novel mitotic / meiotic cell-cycle regulator and cell fate related gene , during preimplantation mouse 1 embryo development and oogenesis . 2 3

Wwc2 is a novel mitotic/meiotic cell-cycle regulator and cell fate related gene, during preimplantation mouse 1 embryo development and oogenesis. 2 3 Giorgio Virnicchi, Pablo Bora, Lenka Gahurová, Andrej Šušor and Alexander W. Bruce 4 5 1 Laboratory of Early Mammalian Developmental Biology (LEMDB), 6 Department of Molecular Biology & Genetics, 7 Faculty of Science, 8 University of South Bohemia, 9 Branišovská 31, 10 370 05 České Budějovice (Budweis), 11 CZECH REPUBLIC. 12 13 2 Laboratory of Biochemistry and Molecular Biology of Germ Cells, 14 Institute of Animal Physiology and Genetics, 15 Czech Academy of Sciences, Rumburská 89, 16 277 21 Liběchov, 17 CZECH REPUBLIC. 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 correspondence; A.W.B. awbruce@prf.jcu.cz & G.V. giorgio.virnicchi@gmail.com 34

Mouse embryo development begins with the fertilisation of ovulated secondary oocytes, appropriately 57 arrested in the metaphase of the second meiotic division (MII-arrested), resulting in the diploid and developmental 58 competent zygote ( 1 ). Ovulated secondary MII oocytes arise from subpopulations of primary oocytes, that have been 59 arrested in the dictyate stage of prophase of meiosis I (MI) since prenatal development, stimulated by maternal 60 reproductive hormones to re-enter and complete meiosis I (reviewed 2 ). In addition to ensuring both the necessary 61 growth required to support later embryonic development and expression of gene products required to execute meiosis, 62 meiotic maturation ensures replicated chromosomal bivalents are appropriately resolved and faithfully segregated 63 between the first polar body and a developmentally competent secondary MII oocyte (endowed with the capacity to 64 subsequently faithfully segregate sister chromatids between the second polar body and resulting zygote, post-65 fertilisation). Failure to precisely segregate chromosomes, resulting in egg and/or zygotic aneuploidy, has severe and 66 usually terminal consequences for embryonic development, with aneuploidy attributable to the human female 67 germline recorded as the leading single cause of spontaneously aborted pregnancy 3, 4 . An extensive literature 68 covering the varied and integrated molecular mechanisms that underpin the germane segregation of homologous 69 chromosomes in MI exists; ranging across meiotic cell-cycle resumption, germinal vesicle (GV/nuclear envelope) 70 breakdown, meiotic spindle assembly, spindle microtubule-kinetochore attachment, chromosomal congression, 71 functioning of the spindle assembly checkpoint (SAC), regulation of the anaphase promoting complex/cyclosome 72 (APC/C) and regulation of cytokinesis/polar body generation (see comprehensive reviews 2,5,6,7,8 ). In keeping with 73 all mammals, and unlike most mitotic somatic cells, meiotic spindle formation in primary mouse oocytes occurs in 74 the absence of centrioles/centrosomes. Spindle assembly is initiated in the vicinity of condensing chromosomes from 75 coalescing microtubule organising centres (MTOCs) and is further stabilised by chromosome derived RAN-GTP 76 gradients that respectively promote and inhibit microtubule polymerisation and destabilisation 5,6,8,9,10 . Unlike other 77 mammalian developmental systems, the transition from MTOC initiated spindle formation to centrosomal control in 78 mice only occurs by the mid-blastocysts stage, when centrosomes appear de novo ( 11 ); contrasting with other 79 mammals (e.g. humans) were fertilising sperm provide a founder centrosome that ensures the first mitotic spindle is 80 assembled centrosomally 12, 13 . Thus, formation and functioning of the first meiotic, and somewhat uniquely, early 81 mitotic spindles in mice are subject to non-centrosomal based regulation. Amongst known key regulators of spindle 82 dynamics are the conserved Aurora-kinase family proteins (AURKA, AURKB & AURKC, collectively referred to 83 here as AURKs) that all exhibit germ cell and early embryonic expression; although, AURKC is not expressd in other 84 somatic cells 14 . During meiosis, AURKs impart important regulatory roles in spindle formation/ organisation, MTOC 85 clustering, chromosome condensation and alignment, plus correct microtubule-kinetochore attachment, 86 chromosomal cohesion and cytokinesis in mitosis (reviewed in 15 ). Specifically, AURKA protein is essential for MI 87 progression 16 and is continually localised with MTOCs and MI/ MII spindle poles, regulating clustering and initiating 88 microtubule nucleation/dynamics, throughout meiosis 15,17,18,19 ; with similar roles in post-fertilisation zygotes 20 .

89
Despite the lack of centrosomes, AURKA functionally cooperates with the key centrosomal protein Polo-like kinase 90 4 (PLK4), to promote meiotic spindle microtubule nucleation 21  As introduced above, in the context of the central role of Hippo-signalling in mouse preimplantation embryo 167 development 49 , we investigated the potential regulatory role of WWC-domain containing genes (i.e. Kibra and Wwc2 168 in mice 60 ) in blastocyst formation. Accordingly, we first assayed Kibra and Wwc2 mRNA expression in 169 preimplantation mouse embryos microinjected with non-specific control dsRNA or constructs specific for each gene 170 (Fig. 1a). Stable levels of Kibra and Wwc2 transcripts were readily detectable at both 8-(E2.5) and 32-cell (E3.5) 171 stages, with normalised Wwc2 expression twice as abundant as Kibra. Each transcript level could be robustly reduced 172 using dsRNA mediated global knockdown at both assayed stages (note, a lack of available antibodies prevented 173 protein expression assays). We next assayed for overt developmental defects associated with individual or combined 174 Kibra/Wwc2 knockdown by microinjecting dsRNA(s) in one blastomere of 2-cell (E1.5) stage embryos, culturing to 175 the late blastocyst (E4.5) stage and counting total cell number (Fig. 1b). Kibra gene knockdown had no significant 176 effect on total cell number, or embryo morphology (versus control dsRNA), but a severe attenuation in cell number 177 when targeting Kibra and Wwc2 transcripts in combination was observed. Interestingly, this defect was statistically 178 indistinguishable from the group of embryos microinjected with Wwc2 dsRNA alone, indicating sole knockdown of 179 the Wwc2 paralog was sufficient to induce the observed phenotype (note, Kibra mRNA levels were unaffected by 180 Wwc2 dsRNA by the 8-cell/E2.5 stage, whereas Wwc2 transcripts were robustly reduced, confirming Wwc2 dsRNA 181 specificity - Fig. 1a and supplementary Fig. S1). Repetition of the Wwc2 knockdown, in a fluorescently marked clone 182 (whereby rhodamine conjugated dextran beads/ RDBs were co-injected with Wwc2 siRNA) confirmed reduced cell 183 numbers as early as the 16-cell (E3.0) stage, specifically within the marked clone that itself exhibited reduced 184 contribution to the initial inner-cell/ ICM founding population (Fig. 1c). Accordingly, we decided focus our further 185 investigations on the overt phenotype associated with Wwc2 knockdown.

187
Embryo cell number deficits caused by Wwc2 knockdown are associated with defective cell division.

188
Although the utilised Wwc2 dsRNA was carefully designed to avoid off target effects (with specifc respect 189 to Kibra expression, Figs. 1a & S1), we sought to phenocopy our observations using an siRNA mediated approach 190 (targeting an alternative region of the Wwc2 mRNA - Fig. S1); thus, providing added confidence to our initial 191 observations and affording the option of phenotypic rescue, by expressing recombinant siRNA-resistant Wwc2 192 mRNA. Figure 2 confirms the selected Wwc2 siRNA near completely eliminated detectable Wwc2 transcripts, after 193 global knockdown, in 32-cell (E3.5) stage blastocysts. Moreover, subsequent to the microinjected 2-cell (E1.5) stage, 194 an assay of total cell number at all cleavage stages up to the late blastocyst (E4.5) stage revealed an accumulative 195 and statistically robust deficit beginning after the 8-cell (E2.5) stage (Fig. 2c). Indeed, from the equivalent 8-cell 196 (E2.5) to 32-cell (E3.5) stages (judged by control embryo development), total cell number did not significantly 197 increase in Wwc2 knockdown embryos but did start to increase during the blastocyst maturation period (32-to >64-198 cell, as judged in control embryo groups; i.e. E3.5 -E4.5 stages). Significantly, it was not unusual for the Wwc2 199 knockdown embryos to initiate cavitation (not shown). We also assayed if the Wwc2 siRNA phenotype was cell 200 autonomous by creating RDB marked Wwc2 specific siRNA knockdown clones (comprising 50% of the embryo -201 Fig. S2). As expected, we observed a statistically significant Wwc2 siRNA mediated deficit in cell number within the 7 microinjected clone (versus the non-microinjected sister clones and the equivalent microinjected clone in control 203 siRNA embryos) at all stages, from the 8-cell (E2.5) to the late blastocyst (E4.5). However, we did not observe 204 significant differences between control and Wwc2 knockdown groups in the number of cells within the non-205 microinjected clone (with the exception of the 32-cell (E3.5) stage where there were an average of 2.3 fewer cells in 206 the Wwc2 knockdown group). These data support a novel and cell autonomous role for Wwc2 in regulating 207 appropriate cell number during preimplantation mouse embryo development. Whilst assaying cell numbers, we 208 observed a number of nuclear/chromatin morphological abnormalities within Wwc2 knockdown embryos, not evident 209 in control embryos. Illustrative examples, at the equivalent of the 32-cell (E3.5) stage are provided (Fig. 3b) and were 210 categorised as; 'abnormal nuclear morphology' (including that typical of persistent 'association with the mid-body'), 211 associated with 'cytokinesis defects' (typified by bi-nucleated cells) or coincident with 'multiple or micronuclei'. 212 Accordingly, we calculated the frequencies by which each, or a composite of, the stated abnormalities were observed 213 (compared to control siRNA conditions) at each assayed cleavage stage; defined per embryo (in at least one 214 blastomere) or per individual assayed cell (Fig. 3c). Except from a single incidence of an early blastocyst stage cell 215 exhibiting a single micronucleus, no other abnormalities were observed in any control siRNA microinjected embryo, 216 at any developmental stage. Conversely, abnormally shaped nuclei were present in over a quarter of Wwc2 217 knockdown embryos at the 8-cell (E2.5) stage and were found in all assayed embryos by the equivalent late blastocyst 218 (E4.5) stage. A similar trend, from the 16-cell (E3.0) stage, was observed in relation to multiple/micronuclei. Despite 219 being less prevalent, cytokinetic defects affected nearly a quarter of Wwc2 knockdown embryos by the late blastocyst 220 (E4.5) stage. Indeed, all Wwc2 knockdown embryos exhibited one or more defect, in at least one cell, by the mid-221 blastocyst (E4.0) stage, with the collective defects (excluding cytokinesis/bi-nucleation) first arising in just over a 222 quarter of 8-cell stage (E2.5) embryos. When we analysed the same phenotypes on the level of each individually 223 assayed blastomere, we found that just over a quarter of cells were affected by the late blastocyst (E4.5) stage; 224 however, it should be noted that such embryos comprised a much smaller average of overall cells, versus control 225 siRNA groups (i.e. 31.3±3.0 against 87.0±3.8 - Fig. 2c). Hence, our collective interpretation is Wwc2 knockdown in 226 cleavage stage mouse embryos is associated with cell autonomous division defects that contribute to embryos with 227 progressively fewer constituent blastomeres as preimplantation development proceeds past the 8-cell stage; invoking 228 a role for Wwc2 in regulating appropriate cell division/ mitosis in blastomeres of the early mouse embryo. 229 230 siRNA mediated Wwc2 knockdown affects blastocyst cell fate derivation.

231
As described, cell number deficit/defective cell division phenotypes were first evident in global siRNA 232 mediated Wwc2 knockdown embryos around the 8-to 16-cell stage (Figs. 2 & 3). Moreover, clonal dsRNA mediated 233 Wwc2 knockdown analysis was specifically associated with reduced inner-cell clone numbers (Fig. 1c). As the 234 transition from the 8-to 16-cell stage, represents the first developmental point constituent embryonic cells are overtly 235 distinct, both in terms of intra-cellular apical-basolateral polarity and relative spatial allocation, with consequences 236 for ultimate blastocyst cell fate (i.e. comprising polarised outer-cells that can give rise to both TE and further inner-237 cells, plus apolar inner ICM progenitors 66 ), we assayed if clonal Wwc2 knockdown altered TE versus ICM cell fate 238 by the early blastocyst (E3.5) stage. The need to assay the first cell fate decision, in this context, was further reinforced 239 by cited precedents implicating Kibra, and hence by association its paralog Wwc2, in activating Hippo-signalling in 240 flies and mammals 56,57,58,59,60 , itself central to TE /ICM specification in mouse preimplantation embryos 49 .
8 Accordingly, we created stably marked clones of Wwc2 knockdown and control cells (by expressing recombinant 242 GAP43-GFP as an injection membrane marker) and assayed expression of the Hippo-sensitive TE marker protein 243 CDX2 67 (Fig. 4). Control siRNA treated embryos developed appropriately to yield 32-cell (E3.5) stage blastocysts 244 consisting an average of 59% outer (CDX2 positive) and 41% inner (CDX2 negative) cell populations, with a 245 statistically equal contribution from each clone (i.e. either control siRNA microinjected or non-microinjected 246 derived). As expected, Wwc2 siRNA microinjected embryos comprised significantly fewer cells with the 247 microinjected clone also characterised by the previously observed abnormal nuclear morphologies and a robustly 248 significant impaired contribution to the ICM; on average 1.6±0.2 cells versus 6.8±0.3 in the non-microinjected clone 249 (or 7.3±0.4 or 7.4±0.3 in the respective clones of control siRNA microinjected embryos). Clonal TE contribution was 250 also significantly impaired but to a much lesser degree and the potential TE deficit was compensated by an increased 251 contribution from the corresponding non-microinjected clone. We also observed a small outer-cell population of 252 Wwc2 knockdown clones that failed to express CDX2 (~20%; not observed in control embryos) with the remaining 253 outer-cells often exhibiting comparatively reduced CDX2 immunoreactivity, when compared to both the non-254 microinjected clone or either clone in the control siRNA microinjected groups (concurrently immuno-stained and 255 imaged using identical protocols; Fig. 4b). However, no ectopic CDX2 expression within inner-cells of the Wwc2 256 knockdown clone was observed and the non-microinjected clone appropriately only expressed CDX2 in outer-cells.

257
Therefore, consequent to clonal Wwc2 knockdown, the overall percentage make up of outer and inner-cells was 258 skewed in favour of outer TE (largely CDX2 positive -72%) over ICM (exclusively CDX2 negative -28%). These 259 results suggest that the required outer-cell Hippo-pathway suppression (to specify TE) and inner-cell activity (to 260 prevent TE differentiation and promote pluripotency) are predominantly intact within the Wwc2 knockdown clone, 261 although maintenance of TE specification/ differentiation in outer-cells is modestly impaired. Therefore, endogenous

262
Wwc2 is unlikely to function as a centrally critical Hippo-signalling pathway (i.e. LATS1/2) regulator (as implied by  f); indicating a degree of regulative development during blastocyst maturation. However, collectively we noted 281 incidences of fragmented nuclei, indicative of apoptotic cell death, that were significantly more prevalent within the 282 ICM of Wwc2 siRNA, versus control siRNA, microinjected embryo groups. Moreover, the observed apoptosis was 283 significantly enriched within the microinjected clone of Wwc2 knockdown embryos, suggesting such ICM residing 284 cell clones are prone to an increased probability of cell death (Fig. 5b). Consistently, the overall and significantly 285 reduced contributions of Wwc2 siRNA derived clones were more pronounced for ICM versus outer-cell populations 286 (Fig. S4). Focussing, on CDX2 and NANOG immuno-stained groups, we again noted a population of CDX2 negative 287 outer-cells and generally reduced CDX2 expression within the Wwc2 knockdown clone (Fig. 5c). Moreover, the 288 significantly reduced number of ICM cells derived from the Wwc2 knockdown clone did not segregate between 289 NANOG positive (indicative of EPI) and NANOG negative (potentially PrE) cells, as per marked control siRNA 290 clones (i.e. according to a clone's overall percentage contribution within the ICM), but were significantly biased to 291 populate the potential, NANOG negative, PrE (Fig. 5d). Considering the embryos in the whole, the overall reduction 292 in size of the potential PrE (NANOG negative) population that was associated with clonal Wwc2 knockdown, was 293 much less than that observed in the EPI (Fig. S4). This suggests inner Wwc2 siRNA microinjection derived clones 294 are impaired in their contribution to sustain EPI numbers, by the late blastocyst (E4.5) stage, but are able to 295 differentiate to form PrE. This interpretation was supported by data directly assaying the two ICM lineages (i.e.

296
GATA4 in combination with either NANOG or SOX2; Figs. 5e & f, plus Fig. S4), whereby the characteristically low 297 number of inner Wwc2 knockdown cells/clones observed, similarly segregated in a manner favouring the PrE (marked 298 by GATA4) over EPI (marked by either NANOG or SOX2); note, compensatory increases in EPI contribution were 299 also observed in the non-microinjected clone (Fig. S4). We also observed numerous examples of fragmented/ 300 apoptotic inner-cell nuclei displaying distinct immuno-reactivity for SOX2 (Fig.5c), suggesting the increased 301 incidence of apoptosis observed within the ICM residing and microinjected cell clones of Wwc2 knockdown embryos 302 ( Fig. 5b) is centred on specified EPI that is ultimately unable to be maintained (although it was not possible to reliably 303 quantify the number of SOX2 positive apoptotic nuclei). Interestingly, in Wwc2 siRNA microinjected embryo groups, 304 we also observed a small population of marked ICM clones that expressed neither PrE or EPI markers (comprising 305 4.6% of all ICM cells and 18.7% of the clone) that were not present in control siRNA treated groups, potentially 306 indicative of a further impairment in ICM cell fate derivation. In conclusion, our late blastocyst cell fate analyses 307 confirm Wwc2 knockdown clone specific and autonomous reductions in overall cell number, that more robustly affect 308 the ICM versus TE. Such reductions are compensated for by regulation within the non-injected clone, that ultimately 309 preserves the overall TE:ICM ratio (albeit with fewer overall cells). However, the lineage contribution of Wwc2 310 knockdown clones within the ICM is biased against the pluripotent EPI in favour of PrE differentiation, via an 311 implicated mechanism possibly involving clone specific and selective EPI apoptosis; suggesting a role for Wwc2 in 312 contributing to specified EPI maintenance that is additional to that related to cell division, yet inherently difficult to 313 experimentally resolve.

317
Despite obtaining consistent RNAi phenotypes using distinct dsRNA/siRNA constructs and their careful 318 design aimed to avoid potential off-target effects or cross reactivity with paralogous Kibra mRNA (Figs. 1 and S1), 10 we sought to further verify the identified novel role of Wwc2. Accordingly, we derived a recombinant and N-320 terminally HA-epitope tagged Wwc2 mRNA construct (HA-Wwc2) that had been specifically mutated within the construct not only enabled us to verify its translation but also its sub-cellular localisation within the microinjected 338 clone, using an immuno-staining approach. Interestingly, given the clone autonomous cell number/division defects 339 observed after Wwc2 knockdown, we consistently detected anti-HA immuno-reactivity at structures typical of mitotic 340 spindle derived mid-bodies (generated after cytokinesis), that were particularly evident within the HA-Wwc2 341 microinjected clones at the 32-cell stage ( Fig. 6d & Fig. S6a). As a recent study identified a critical role for 342 uncharacteristically persistent interphase mid-bodies (referred to as 'interphase microtubule bridges') as important 343 MTOCs within the blastomeres of cleavage stage embryos 65 , we assayed their number (immuno-staining for a-

344
Tubulin and activated phospho-Aurora-B/pAURKB, as recognised midbody markers -68 ) in control and Wwc2 345 knockdown embryos at the same 32-cell stage, but could not detect any statistically significant variation in their 346 overall incidence (when corrected for the reduced cell number caused by Wwc2 knockdown - Fig

351
We next asked the question whether a similar knockdown of Wwc2 expression in meiosis I arrested germinal 352 vesicle (GV) stage primary mouse oocytes would elicit any defects during maturation to the metaphase arrested stage 353 of meiosis II (MII), particularly because both paradigms of cell division occur in the absence of centrioles 5, 6, 8, 9, 10, 354 11, 22 . We observed abundant Wwc2 transcripts in GV and MII oocytes (plus zygotes) that show evidence of becoming 355 cytoplasmically poly-adenylated, and hence more likely to be translated as required functional proteins 69 , between 356 the GV and MII stages (as revealed using alternative, oligo-dT versus random hexamer, cDNA synthesis priming 357 strategies - Fig. 7a); indeed consultation of our previously published assay of meiotically maturing oocyte polysome 11 associated transcripts 70 supports this interpretation (~40% polysome association; Fig. 7b). We therefore confirmed 359 our Wwc2 siRNA construct could elicit robust knockdown in microinjected GV oocytes, that had been blocked from 360 re-entering meiosis I (using a cAMP/cGMP phosphodiesterase inhibitor, 3-isobutyl-1-methylxanthine/ IMBX) and 361 then permitted them to in vitro mature (IVM) to the MII equivalent stage (Fig. 7c & d). Such oocytes were fixed and 362 immuno-fluorescently stained for a-Tubulin (plus DAPI DNA stain), to assay for potential phenotypes. In our two 363 control conditions (i.e. non-microinjected or control siRNA microinjected oocytes) >90% of GV oocytes matured to 364 the MII arrested stage, typified by an extruded polar body (PB1) and metaphase II arrested spindle. However, in the 365 Wwc2 knockdown group the successful IVM rate was reduced to 8.6% and oocytes presented with various arrested 366 phenotypes (lacking PB1) that we categorised by the presence of metaphase I meiotic spindles (-PB1 +MI spindle; 367 42.9% -note such spindles typically failed to migrate to the oocyte cortex), a spindle-like structures with mis-aligned

379
It is reported that activated/phosphorylated Aurora kinase-A (p-AURKA: p-Thr288) is required to initiate 380 acentriolar/ MTOC mediated spindle assembly in mouse oocytes and that Aurka knockdown phenotypes resemble 381 those observed here for Wwc2 knockdown 21 . We therefore hypothesised the meiotic maturation phenotypes identified 382 in Wwc2 mRNA depleted GV oocytes may be associated with impaired levels of activated p-AURKA. A western blot 383 analysis of staged maturing oocytes confirmed characteristically low levels of p-AURKA in GV oocytes, that robustly 384 increased with the appearance of the first meiotic spindle (during MI) and were maintained with the formation of the 385 second meiotic spindle (in MII arrest; Fig. 8b & 15, 16 ); a trend replicated in control siRNA microinjected GV (assayed 386 after 18 hour incubation in IMBX) and MII (after 18 hours post-IBMX removal) oocytes (note, the specificity of the 387 anti-p-AURKA antibody used was additionally confirmed; Fig. S8). However, in the Wwc2 siRNA microinjected 388 group, we were not able to detect p-AURKA at either the equivalent GV or MII stages. Moreover, we were also 389 unable to detect p-AURKA immuno-reactivity localised to the poles of any spindles or spindle-like structures formed 390 in Wwc2 knockdown oocytes, in stark contrast to that observed on MII arrested spindles in controls (Fig. 8c); 391 confirming a failure to activate AURKA under conditions associated with profound meiotic oocyte maturation 392 defects. We next employed the HA-Wwc2 rescue mRNA construct in an attempt to reverse the observed Wwc2 393 knockdown oocyte maturation phenotypes. Using a similar IVM assay (but assaying at multiple time-points), we 394 found that co-microinjection of HA-Wwc2 mRNA and Wwc2 siRNA was able to robustly rescue the maturation 395 phenotypes caused by Wwc2 siRNA microinjection alone (i.e. MII arrested oocytes at 16 hours post-IBMX; control 396 siRNA -81.8%, Wwc2 siRNA -8.7% & HA-Wwc2 mRNA plus Wwc2 siRNA -76.0%); moreover the progression 12 of the rescue through recognisable maturation stages was in-step with that of control siRNA microinjected oocytes 398 (as measured at 1, 6 , 12 and 16 hours post-IBMX removal; Fig. 8d). Consistently we also found co-microinjection 399 of the rescue HA-Wwc2 mRNA construct was coincident with reappearance of activated p-AURKA protein (Fig. 8e), 400 that was also detected in proximity to forming or matured MI/MII stage meiotic spindles (Fig. S9). Collectively, these 401 data confirm a novel role for Wwc2 in regulating appropriate bivalent chromosome segregation during mouse primary 402 oocyte meiotic maturation that is strongly associated with activation of the acentriolar/MTOC mediated spindle 403 forming meiotic/ mitotic kinase AURKA.

420
Relating to cell fate regulation, the attainment of naïve pluripotency within specified EPI cells of the late 421 (E4.5) mouse blastocyst has recently been shown to be regulated via a heterogeneous mechanism of activated 422 TEAD1-mediated transcription (caused by YAP relocalisation from the cytoplasm to the nucleus), that promotes 423 ICM cell/EPI progenitor competition and elimination (by apoptosis) 54 . This contrasts with earlier specification and 424 segregation of outer TE from founder ICM populations, prior to the blastocyst stage, whereby activated Hippo-425 signalling and YAP nuclear exclusion is required for inner-cells to actively retain pluripotency (e.g. express Sox2) 426 and resist TE differentiation 49 . Hence, active Hippo-signalling seems to both support initial pluripotency in ICM 427 founders (i.e. confirmed -see 49 ) but presumably needs to be suppressed in EPI progenitors to promote naïve 428 pluripotency (i.e. to permit nuclear YAP translocation, inferred from 54 ). Our data, although compounded by the 429 additional cell division phenotypes, suggests Wwc2 (as a paralog of the Hippo-activator Kibra) also contributes to 430 germane Hippo-signalling regulation. For example, relating to late blastocyst (E4.5) stage EPI specification, we find 431 ICM-residing Wwc2 knockdown clones, although fewer than their non-microinjected clone counterparts, do not 432 display the same equal contribution to populate the EPI (NANOG+/SOX2+) and PrE (GATA4+) as observed in the 433 non-specific siRNA microinjected clones of control embryos (Figs. 5 and S4). Indeed, such clones statistically favour 434 PrE over EPI contribution, indicative of compromised pluripotent potential. Whether this is due to compromised 435 Hippo-signalling activation after initial internalisation prior to blastocyst formation, and/or relates to relative fitness 436 deficiencies affecting their ability to successfully compete to colonise specified EPI, is uncertain. However, an 437 increased apoptotic ICM cell number within Wwc2 knockdown clones, often marked by detectable SOX2 protein 438 expression (but never GATA4 immunofluorescence), observed in late (E4.5) blastocysts (

468
Alternatively, it may reflect the developmental point maternally provided WWC2 protein, unaffected by siRNA, is 469 functionally depleted. However, it impossible here to directly probe WWC2 protein expression, due to a lack of 470 available antibodies. Although, we did confirm the abundant presence of poly-adenylated maternal Wwc2 mRNAs 471 in both primary (associated with translating polysomes) and secondary mouse oocytes and zygotes ( Fig. 7a &  after global Wwc2 knockdown, through the 16-/32-cell, E3.0/E3.5, stage equivalents; Fig. 2). This is significant as 484 the mid-blastocyst (E4.0) stage reflects the developmental point at which the atypically acentrosomal/acentriolar cell 485 cleavage divisions of mouse embryos, in comparison with most other mammalian species, reverts back to 486 centrosomal control, after de novo centrosomes synthesis 11 (although it could also reflect the waning effect of Wwc2 487 specific siRNA). Therefore, we propose our data demonstrate a role for WWC2 in regulating acentrosomal cell 488 division in preimplantation stage mouse blastomeres, prior to the mid-blastocyst (E4.0) stage. Moreover, that WWC2 489 is a regulator of mitotic spindle formation/functional dynamics, as reflected in phenotypic nuclear morphologies 490 observed after Wwc2 knockdown (Fig. 3). Supporting this hypothesis, we readily detected recombinant HA-WWC2

505
The first report to implicate WWC-domain containing proteins in mammalian Hippo-signalling, 506 demonstrated human KIBRA (in HEK293T cells) binding to, and stimulating the phosphorylation/activation, of 507 LATS1/2, leading to subsequent phosphorylation of YAP 59 . The same authors later described mitosis specific 508 KIBRA phosphorylation, catalysed by AURKA and/or AURKB at a highly conserved AURK consensus/target motif 509 (centred on Ser539), that could be antagonised by the mitotic phosphatase PP1. Moreover, they reported over-  (Fig. S10). Therefore, it is tempting to speculate the Wwc2 knockdown 520 phenotypes we observe in preimplantation stage mouse embryos may be mechanistically related to those previously 16 described for human KIBRA, although with the caveat they function on the level of MTOC regulation, in the absence 522 of centrosomes/centrioles 11 . For example, targeted genetic ablations of the Lats1 and Lats2 genes in mice are not 523 described in association with the preimplantation embryo cell division/cell number deficits typical of Wwc2 524 knockdown, nor those described phenotypes above in relation to other models of mitosis. Rather, Lats1 -/-/Lats2 -/-525 early blastocysts (E3.5) only present with ectopic nuclear YAP expression within ICM cells 40 . Therefore, not all 526 insights will be directly transferable. However, despite the lack of centrosomes in pre-mid-blastocyst stage mouse 527 embryos, classical key centrosome regulators are expressed and been shown to have functional cell division roles.

546
Interestingly, the identified MTOC mediated spindle formation role of PLK4/AURKA is distinct from the co-existing 547 mechanism of chromosome derived RAN-GTP gradient driven microtubule stabilisation in oocytes (as reviewed 5, 6, 548 8, 9, 10 ). Clearly, in the context of Wwc2 knockdown oocyte maturation phenotypes described here (that include failed 549 AURKA phosphorylation/activation), potential links between described PLK4/AURKA phenotypes and WWC2 550 merit further investigation. They are also consistent with reports implicating AURKA in the regulation of MTOC 551 dynamics and microtubule spindle nucleation 15,17,18,19 . However, importantly the described inhibition of PLK4 was 552 associated with eventual polar body formation 21 , unlike after Wwc2 knockdown (even after an extended IVM Briefly, gene specific PCR primer pairs, incorporating 5'-T7-derived RNA polymerase promoters and spanning 609 designed dsRNA complementary sequence, were used to derive in vitro transcription (IVM) template, using mouse 610 blastocyst (E3.5) cDNA as a template (or plasmid DNA for GFP). After agarose gel verification, the double stranded 611 DNA templates were used in preparatory IVM reactions, incorporating DNaseI and single-stranded RNase treatment 612 (MEGAscript T7; ThermoFisher Scientific, AMB13345), to generate dsRNA. The integrity of derived Kibra-,

613
Wwc2and GFP-dsRNAs was confirmed by non-denaturing gel electrophoresis and quantified (Nanodrop). The PCR 614 primer sequences used are provided in supplementary methods table SM2.