YAP1 and WWTR1 are required for murine pregnancy initiation

DNA isolated from uteri was interrogated to determine Cre mediated recombination utilizing primers that flank the 5’ and 3’ regions of the loxP sites inserted around exon 3 in the Yap1 gene( Xin, et al. 2011 ). The first set of reactions probed upstream of the first loxP site and downstream of the first loxP site (primers YF + YR1) indicating whether uteri retained this loxP site. DNA gel results shown in lane 1 indicate loxP, lane 2 is inconclusive, and lane 3 indicates a wild-type band ( Supplemental Figure 1A ). The triple allele mutants that followed, all show a wild-type band in lanes 4–6 ( Supplemental Figure 1A ). The second set of reactions probed upstream of the first loxP site and downstream of the second loxP site (primers YF + YR2). Control uteri in lanes 1–2 did not produce product for the second reaction indicating that Cre mediated recombination did not occur ( Supplemental Figure 1B ). Lanes 4–6 show one band of the lowest molecular weight indicating that recombination occurred producing product ( Supplemental Figure 1B ). In addition to DNA assessments, mRNA of whole uteri was evaluated for expression of Yap1 , Wwtr1, Ccn1 , and Ccn2 at 3.5-, 5.5-, and 7.5-days post coitus (dpc) ( Figure 1 ). Despite primers designed specifically to flank the excised exon 3 of Yap1 and Wwtr1 , depletion of mRNA for these genes and their targets, Ccn1 and Ccn2 , was not detected at any of the timepoints investigated (multiple unpaired t-tests FDR >0.01, Figure 1C – E ). Immunohistochemical analysis indicated epithelial and stromal compartment-specific expression of YAP1 and WWTR1 in uteri at 3.5dpc, however expression levels were not significantly different between controls and YWTs (2-way ANOVA p >0.05, Figure 1 F–G). In addition, due to the widespread expression of Pgr in the female reproductive tract, we investigated ovarian expression of Yap1 and Wwtr1 ( Soyal, et al. 2005 ). Evidence of Yap1 and Wwtr1 and target gene depletion was not evident in whole ovaries (multiple unpaired t-tests p >0.05, Supplemental Figure 1C ). In addition, both YAP1 and WWTR1 protein expression was not significantly different in granulosa cells of 3.5dpc control and YWT ovaries based on immunohistochemistry (multiple unpaired t-tests p >0.05, Supplemental Figure 1D ). In addition to these molecular analyses, no structural differences were observed in the ovaries, oviducts, and uteri of virgin females at proestrus ( Supplemental Figure 2A ) or nulliparous females at 3.5dpc ( Supplemental Figure 2B ). Female individual mutants ( Y, W ), heterozygous mutants ( YWH ), and triple allele mutants ( YWT ) were assessed for fertility in a 6-month breeding trial with proven fertile wild-type males to determine if Yap1 , Wwtr1 , or a combination of alleles from both genes were required for proper fertility. Y mutants were subfertile and had fewer pups per litter compared to floxed controls (n=4/genotype, unpaired t-test p= 0.0007, Figure 2A , Table 1 ). W individual mutants were classified as subfertile and produced fewer average pups per litter compared to floxed controls (n=4/genotype, unpaired t-test p =0.0105, Figure 2B , Table 1 ). YW heterozygous mutants in a shortened 3-month breeding trial also produced fewer average pups per litter compared to floxed controls (n=4, unpaired t-test p =0.0331, Figure 2C , Table 1 ). Triple allele mutants were also subfertile producing significantly fewer average pups per litter compared to floxed controls (n=4/genotype, unpaired t-test p =0.0007, Figure 2C , Table 1 ). The YWT mice bore approximately 50% fewer pups and litters than the floxed controls throughout the breeding trial ( Table 1 ). At the termination of the 6-month breeding trial, n=3 YWT mice were collected for qualitative analyses. These mice did not display any significant differences in body, uterine, or ovarian weight at the time of collection (multiple t-tests FDR >0.01, Figure 2D ) and no alterations in reproductive tract structure were observed grossly nor with histological analyses ( Supplemental Figure 2C ). Given the increasing severity of phenotype observed as an increasing number of Yap1 and Wwtr1 alleles were deleted, all further investigation focused on triple allele mutants. Pregnancy success was assessed in control and YWT nulligravid mice at various time points including implantation (4.5dpc, n=6–8), post-implantation (5.5dpc, n=6–10), early pregnancy (7.5dpc, n=5–7), early placentation (9.5dpc, n=5–7), and post-placentation (12.5dpc, n=8–10) ( Figure 3A and B ). The gross morphology of the uteri was normal at all time points except at 12.5dpc when YWT uteri exhibited resorption sites ( Figure 3A ). Comparisons of implantation site quantity within each time point (4.5, 5.5, 7.5, 9.5, 12.5dpc) revealed comparable number of normal implantation sites in YWT and floxed control uteri (multiple unpaired t-tests FDR >0.01, Figure 3B ). However, YWT females exhibited a higher likelihood of being not pregnant across all time points compared to controls indicating significant pregnancy failure (33% of individuals not pregnant, Fisher’s exact test p <0.0001, Figure 3C ). To determine the cause of pregnancy failure exhibited in the YWT nulligravid mice, ovulation and fertilization rates and embryo transport were assessed. Post-fertilization at 1.5dpc, when all ovulated oocytes should be in the oviducts, we noted comparable numbers of unfertilized oocytes (UFO) and 2-cell embryos in YWT and controls indicating fertilization and ovulation rates were not affected in YWTs ( Figure 4A ). However, at 3.5dpc, the time of embryo entry into the uterus, we noted that some YWT females had embryos in their oviducts but not in their uteri, however the differences were not significant ( Figure 4B – C ). The total number of visual corpora lutea did not differ between YWT and controls, nor did the total number of embryos flushed at 3.5dpc ( Figure 4D ). In addition, visualization of oviducts indicated appropriate folding patterns ( Supplemental Figure 3A ) ( Harwalkar, et al. 2021 ). Individual uteri were imaged utilizing tissue clearing to allow visualization of embryos using the whole-mount immunofluorescent protocol. Embryos were visualized within all but one YWT uterine horn at 3.5dpc ( Supplemental Figure 3B and C ). To disseminate the repercussions of partially penetrant delayed embryo entry into the maternal endometrium combined, we investigated endometrial receptivity at 3.5dpc ( Figure 5 ). The progesterone receptor expression by both mRNA and protein was not altered in YWT uteri at 3.5dpc ( Figure 5A and B ). However, we noted significantly decreased target gene expression of Ihh , and Nr2f2 in 3.5dpc uteri (unpaired t-test, p <0.01, Figure 5A ). In addition, luminal epithelial expression of proliferation marker Ki67 was maintained in YWT uteri ( Figure 5C ). Despite these noted alterations in progesterone receptor signaling, we did not observe alterations between YWT and floxed control serum hormone levels at 3.5dpc ( Figure 5D ). Interestingly, estrogen receptor signaling decreased in YWT compared to controls at 3.5dpc (Unpaired t-test, p <0.03, Supplemental Figure 4 ). Artificial decidualization was maintained for 5 days, and uteri were then collected for analysis. Representative images show unstimulated (left) and stimulated (right) horns for floxed controls and YWT ( Figure 6A ). The uterine wet weight ratio of stimulated to unstimulated horn was not different in the YWT group compared to controls ( Figure 6B ). However, mRNA expression of decidualization response genes, Wnt4 and Bmp2 , indicated a blunted decidualization response in the stimulated uterine horns of YWT compared to controls (Two-way ANOVA p =0.04 Wnt4 , and p =0.006 Bmp2 , Figure 6C ). To assess how partially penetrant delayed embryo entry, blunted endometrial receptivity, and compromised decidualization response affect embryonic development in YWT primigravid dams we assessed implantation chamber morphology as well as embryonic morphological development. Implantation chamber morphology was notably normal when visualized with whole uterine imaging at 5.5dpc ( Figure 7A ). However, YWT uteri contained embryos that morphologically appeared delayed with decreased elongation compared to controls at 5.5dpc ( Figure 7B ). Most embryos visualized (59%) within YWT uteri exhibited a morphological delay ( Figure 7C ). Overall, implantation chamber length was not affected despite morphologically delayed embryos being found in most implantation sites visualized in YWT uteri ( Figure 7D ). Morphologically delayed embryos with decreased elongation had 3D surface volumes significantly lower compared to normal elongated embryos found within both control and YWT implantation sites (unpaired t-test, p= 0.01 compared to normal control embryos and p =0.001 compared to normal YWT embryos, Figure 7E ). Additionally, the ratio of embryo volume to implantation chamber length of those embryos that were identified as delayed was significantly decreased (unpaired t-test, p <0.01, Figure 7F ). These findings in conjunction with delayed embryo entry, decreased endometrial receptivity, and decreased decidualization response indicate that loss of Yap1 and Wwtr1 in the uterus leads to a significant decline in fertility and pregnancy success by delaying early embryonic development. Whole uterine cross-sections at 7.5dpc from YWT (n=3) and floxed controls (n=4) were subjected to bulk mRNA sequencing to elucidate the transcriptional repercussions associated with the loss of Yap1 and Wwtr1 in the pregnant uterus. Samples were separated on an MDS plot based on genotype ( Figure 8A ). We identified 16,674 total genes after filtering and 1,785 differentially expressed genes (DEG; FDR<0.05), with 884 genes upregulated and 901 downregulated in the knockouts compared to controls ( Figure 8B ). Differentially expressed genes (DEG) included 8 genes from the Hippo signaling pathway including upstream regulators ( Amotl1 and Amotl2 ), kinases ( Sav1 and Lats2 ), transcription factor ( Tead4 ), transcriptional effectors (Yap1 and Wwtr1 ), and target genes ( Ccn2 and Birc5 ) ( Figure 8C , left panel). Decreased DEG also included 4 decidualization response genes ( Figure 8C , right panel). DEG contributed to Gene Ontology Cellular Component terms like extracellular matrix, collagen trimer, and collagen-containing extracellular matrix, indicating that loss of Yap1 and Wwtr1 affected genes involved in cellular structure ( Figure 8D ). In addition, upregulated genes were enriched for Rat Genome Database (RGD) disease terms like pre-eclampsia, endometriosis, liver cirrhosis, and pulmonary fibrosis, indicating that YWT uteri were enriched for genes connected with reproductive diseases associated with pregnancy loss and fibrosis, which has classically been associated with Hippo signaling ( Figure 8E ). To better understand the molecular complexity resulting from the Yap1 and Wwtr1 mutation, DEG were compared with the transcriptional analyses done in other genetically engineered mouse models including Progesterone receptor knockout (PRKO), Estrogen receptor alpha Exon 3 knockout (ERKO), and Epidermal growth factor receptor knockout (EGFRKO) ( Hewitt, et al. 2010 , Jeong, et al. 2005 , Large, et al. 2014 ). We compared the DEG identified in the ovariectomized PRKO vs. floxed controls after 4 hours of vehicle treatment to the 7.5dpc YWT DEG. Unsurprisingly, overlap of DEG in the same direction as PRKO DEG were very limited (n=17) and did not point toward any specific mechanism ( Supplemental Table 7 ) ( Jeong, et al. 2005 ). The lack of DEG overlap between the PRKO and YWT models is likely attributed to a lack of circulating progesterone in the ovariectomized PRKO females and very little progesterone receptor signaling active while at 7.5dpc the YWT uteri it would be expected to have strong progesterone receptor signaling at the sites of implantation therefore the DEG identified in each model would be expected to be mostly unique to their respective conditions. Given that Yap1 and Wwtr1 are known to be estrogen and growth factor receptor regulated, we compared these knockout DEG to YWT mutant DEG( Ando, et al. 2021 , Moon, et al. 2022 , Zhu, et al. 2019 ). We identified many overlapping DEG between both of these comparisons. Ovariectomized vehicle treated ERKO vs floxed control DEG overlapped with YWT DEG to include pathways such as integrin signaling (e.g. Col8a1, Col4a6, Itgb8 ), Oxytocin receptor signaling ( Plcl1 and Oxtr ), and Gonadotropin-releasing hormone receptor signaling ( Nos1, Mmp14, Syt4, Inhbb ), all of which are critical in the establishment of pregnancy and maintenance and hormonal regulation ( Supplemental Table 8 )( Hewitt, et al. 2010 ). In addition, comparison of overlapping genes in EGFRKO DEG at day 1 of decidualization with 7.5dpc YWT DEG overlap included genes belonging to cell surface receptors (e.g. Itga1, Itgb7, Slc43a2 ), extracellular structure (e.g. Cldn3, Col24a1, Col4a6, Col4a5, Col5a3) , and EGF signaling (e.g. Rhoj, Egfr, Nrg2, Shc1, Ereg ) ( Supplemental Table 9 ). Together these results suggest that Yap/Wwtr1 regulate genes similar to ESR1 and EGFR and potentially play critical regulatory roles in female reproductive function.

Progesterone receptor (Pgr) Cre ( Pgr Cre/+ )( Soyal, et al. 2005 ) expressing males were crossed to Yap fl/fl Wwtr1 fl/fl females ( Xin, et al. 2013 , Xin, et al. 2011 ) to generate conditional mutation of a combination of Yap1 and Wwtr1 alleles. Investigation into the Mouse Genome Informatics database revealed that Progesterone receptor and Yap1 are less than 1 centimorgan apart, therefore it was impossible to generate a quadruple allele mutant model utilizing the Pgr Cre . Therefore, we generated and investigated each allele in combination: Yap1 mutants ( Pgr Cre/+ Yap fl/+ , Y ), Wwtr1 mutants ( Pgr Cre/+ Wwtr1 fl/fl , W ), Yap1 and Wwtr1 heterozygous mutants ( Pgr Cre/+ Yap fl/+ Wwtr1 fl/+ , YWH ), Yap1 and Wwtr1 triple allele mutants ( Pgr Cre/+ Yap fl/+ Wwtr1 fl/fl , YWT) . The Pgr Cre/+ mice are a mixed background of 129Sv × C57BL/6, the Yap fl/fl Wwtr1 fl/fl mice are 129SvEv. Animals were housed and maintained in a designated animal care facility at Michigan State University on a 12-hour light/dark cycle with free access to food and water. All animal procedures were approved by the Institutional Animal Care and Use Committee of Michigan State University. Nulliparous females were co-housed with proven fertile wild-type males for a 3- or 6-month breeding trial. Males were rotated in or out of cages if females did not produce a live born litter within one-month from time of set up. For timed mating experiments, proven fertile wild-type males were placed in nulligravida and nulliparous female cages in the evening. Seminal plugs were checked each morning with day of plug designated at 0.5 days post coitus (dpc). Tail vein injection with Chicago blue dye served as a positive identifier for implantation sites at all time points. Following the blue dye injection, female mice were sacrificed for collection at different timepoints including 1.5, 3.5, 4.5, 5.5, 7.5, 9.5, and 12.5dpc. Body weight, uterine wet weight, ovarian wet weight, implantation site number, and gross morphology were catalogued. Uterine, oviductal, and ovarian tissues were divided and flash frozen or stored in RNAlater for downstream RNA and protein analyses or fixed in 4% paraformaldehyde for histological analysis. A subset of tissues were fixed in 10% DMSO in methanol for tissue clearing and advanced light sheet microscopy. Sexually mature (8 weeks of age or older) female mice were ovariectomized followed by two weeks of rest. Animals were treated with three daily subcutaneous injections of 100ng 17β-estradiol followed by two days of rest then three daily injections of 1mg progesterone plus 6.7ng 17β-estradiol. Six hours following injection on the third day, an intraluminal scratch was performed surgically on the anti-mesometrial side of the endometrium of one uterine horn utilizing a blunted 25G needle. The unscratched uterine horn served as an unstimulated hormonal control. The decidualization reaction was maintained with daily subcutaneous injections of 1mg progesterone plus 6.7ng 17β-estradiol for at total of 5 days followed by euthanasia (n=5/genotype). Uterine tissues were collected as described above. Decidual reaction was measured by uterine wet weight ratio of stimulated/unstimulated horn and molecular markers Wnt4 and Bmp2 by qPCR of stimulated uterine horn compared to unstimulated horn from the same animal. Sexually mature female mice (8 weeks or older) were mated to proven fertile wild-type males. Oviductal flushes were performed with phosphate buffered saline by inserting a 30G needle into the infundibulum of both oviducts at 1.5dpc (n=6 per genotype). Flushed products were collected, counted, categorized, and imaged. Uterine and oviductal flushes were also performed at 3.5dpc with uterine flushes performed by inserting a 30G needle into the uterotubal junction and flushing toward the cervix utilizing PBS. Flushed products were similarly collected, counted, categorized, and imaged. Total DNA was isolated from snap frozen uterine tissues at 3.5dpc. Reverse transcriptase PCR (RT-PCR) was performed utilizing genotyping primers as follows: YF 5’-ACATGTAGGTCTGCATGCCAGAGGAGG-3’ and either YR1 5′-AGGCTGAGACAGG AGGATCTCTGTGAG-3′, or YR2 5′-TGGTTGAGACAGCGTGCACTATGGAGC-3′ ( Xin, et al. 2013 , Xin, et al. 2011 ). YF is upstream of the first loxP site; YR1 is downstream of the first loxP site; YR2 is downstream of the second loxP site. Primer scheme and bands produced by RT-PCR and electrophoresis are provided in the supplemental information of the publication that generated the Yap mutant mouse model ( Xin, et al. 2011 ). Tissues were fixed in 4% paraformaldehyde, dehydrated in ethanol and xylene, and embedded in paraffin. Sections (6μm) were deparaffinized and rehydrated in a graded alcohol series followed by antigen retrieval (Vector Laboratories, Burlingame, CA) and hydrogen peroxide treatment. Next, sections were blocked and incubated with antibodies against YAP, WWTR1, ER-alpha, PGR, Ki67, SUSD2 or ASMA overnight at 4◦C (see Supplemental Table 1 for complete antibody information). On the following day, sections were incubated with biotinylated secondary antibodies followed by incubation with horseradish peroxidase conjugated streptavidin. Immunoreactivity was detected using the DAB substrate kit (Vector Laboratories) and visualized as brown staining by light microscopy. Incubation with secondary antibody only served as a negative control. Alternatively, after dehydration, slides were stained with Masson’s Trichrome followed by rehydration in a graded ethanol series then cover slipped and visualized by light microscopy. ImageJ image analysis software (NIH, v2.14.0), was utilized to determine a digital HSCORE for staining intensity of luminal epithelium, glandular epithelium, and stromal compartments of each uterine section. Total RNA was isolated from frozen mouse tissue using TRIzol reagent (Invitrogen, Waltham, MA). About 1μg of RNA was reverse transcribed to cDNA using a High- Capacity cDNA Reverse Transcription kit according to the manufacturer’s instructions (Applied Biosystems, Foster City, CA). Quantitative real-time PCR (qPCR) was performed with SYBR Green PCR Master Mix (Applied Biosystems) using the BioRad CFX Opus 384 qPCR system utilizing primers targeting genes of interest ( Supplemental Table 2 ). Expression was normalized to the average of 36b4 and 18s per sample and fold change determined compared to time matched floxed controls ( Pgr +/+ Yap fl/fl Wwtr1 fl/fl ). Total RNA was isolated from flash frozen mouse uteri utilizing TRIzol reagent as described above. Following isolation, RNA was treated with a RNA Clean and Concentrator kit (Zymo) followed by DNAse treatment (TURBO DNA-free kit, Invitrogen) and stored in nuclease-free water at −80°C. Concentration was determined utilizing a Qubit RNA BR Assay Kit (Invitrogen). Samples were sent for RNA integrity analysis then subsequently sent to a sequencing facility for library preparation and sequencing (Michigan State University Genomics Core). Libraries were prepped with Illumina Stranded mRNA Library Prep kit (Illumina) and sequenced (paired end 150bp) on a NovaSeq 6000 Instrument (Illumina) to an average depth of 40 million read pairs per sample. Reads were quality trimmed, adapters were removed using TrimGalore (version 0.6.10), and quality-trimmed reads were assessed with MultiQC (version 1.7). Trimmed reads were mapped to Mus musculus GRCm39.110 and gene counts quantified using STAR (version 2.6.0c). Model-based differential expression analysis was performed using edgeR-robust method (version 3.42.4) ( Zhou, et al. 2014 ) in R. Genes with low counts per million (CPM) were removed using the filterByExpr function from edgeR. Multidimensional scaling plots, generated with the plotMDS function of edgeR, were used to verify group separation prior to statistical analysis. DEG were identified as FDR P value less than 0.05. Visualization of DEG was performed utilizing EnhancedVolcano (version 1.18.0) to generate a volcano plot. Counts per million of selected DEG were plotted using the box plot function of ggplot2 (version 3.4.4). Gene set enrichment and visualization was completed using the ShinyGO online tool (version 0.80) ( Ge, et al. 2020 ). Uteri were dissected from 3.5 (n=6–7) and 5.5dpc (n=8) YWT and control females. Whole mount immunofluorescent staining was performed as previously described( Arora, et al. 2016 ). Briefly, following animal euthanasia, uterine samples were subsequently fixed in DMSO:methanol (1:4) and stored at −20°C. Then, samples were rehydrated in 1:1, Methanol:PBST (PBS + 1% Triton X-100) for 15 minutes then washed for 15 minutes in 100% PBST. Tissues were then incubated in a blocking solution of PBS, 1% Triton X-100, and 2% powdered milk for 2 hours at room temperature. Samples were stained with primary antibodies for rat anti-CDH1 (M108, Takara Biosciences), rabbit anti-cytokeratin 8 (MA5–14476, Invitrogen) and rabbit anti-FOXA2 (Abcam, ab108422) diluted at 1:500 in blocking solution for 7 nights at 4°C. Uterine samples were then washed in PBST for 2X15 minutes and 4X45 minutes each then incubated with fluorescently conjugated Alexa Fluor 555 Donkey anti-Rabbit IgG secondary antibody (A31572, Invitrogen), and 647 Goat anti-Rat secondary antibody (A21247, Invitrogen) and Hoechst (B2261, Sigma Aldrich) at 1:500 for 2 nights at 4°C. Samples were then washed in PBST for 2X15 and 4X45 minutes each, dehydrated in methanol then incubated overnight in 3% H2O2 diluted in Tissues were then washed in 100% methanol for 2X15 minutes and 1X60 minutes each and cleared overnight using BABB (benzyl alcohol:benzyl benzoate, 1:2). Stained tissue samples were imaged with a Leica TCS SP8 X Confocal Laser Scanning Microscope System (Leica Microsystems) with a white-light laser, using a 10x air objective. For each uterine horn, z-stacks were generated with a 7.0 μm increment, and image analysis was carried out using commercial software Imaris v9.2.1 (Bitplane, Zurich, Switzerland). Briefly, confocal LIF files were imported into the Surpass mode of Imaris, and Surface module 3D renderings were used to create structures for the oviductal–uterine junctions, embryos, implantation chambers, and horns as described previously( Valero-Pacheco, et al. 2022 ). We used the Contour modules for Hoechst and FOXA2 fluorescent signal (embryo), and CDH1 signal (oviduct and implantation chamber). Embryo 3D volume was assessed using the surface statistics function, while embryo implantation chamber length was calculated using the measurement points module of Imaris software. Western blot densitometric data were tested for normality then analyzed using a one-way ANOVA with Dunnett multiple comparison correction. qPCR data were tested for normality then log transformed in the case of non-normal data and analyzed by one-way ANOVA. Data that were not normal were analyzed utilizing Kruskal-Wallis nonparametric tests. Comparisons of two groups were assessed by individual t-tests. Statistical analyses were performed utilizing GraphPad Prism 10 (GraphPad Software) and values were considered significant if p <0.05.

The deficient uterine receptivity and dysregulation of hormonal signals within the YWT females is indeed interesting. However, the interplay between the Hippo signaling pathway and the hormonal response has been observed before. The functional kinases of the Hippo kinase cascade, serine/threonine kinase 3 and 4 (Stk3/4), have been shown to be differentially expressed throughout the murine estrous cycle and, indeed, primarily under the control of estrogen signaling( Moon, et al. 2019 ). Indeed, Yap expression is also differentially regulated throughout the murine estrous cycle, with its level being highest during estrous and diestrus( Moon, et al. 2022 ). Importantly, estrus in female mice is when ovulation occurs, and 3 days later is when the window of receptivity begins. In addition to the cyclical presence of Yap1 as well as activation of the Hippo signaling pathway, the function of this pathway has also been associated with hormonal action. In particular disease states like breast cancer and endometriosis, YAP1 bound to TEAD transcription factors can induce transcription of ESR1 , or alternatively, in endometriosis, YAP1 can regulate upstream factors that influence PGR expression ( Lin, et al. 2023 , Ma, et al. 2022 ). These reports support an indirect method of hormonal control and interpretation of the signal, but it is possible within the scope of YAP1/WWTR1 function as transcriptional coactivators that they could play a direct role within hormone receptor function. In one study, YAP/TEAD complexes bound to ESR1 enhancer regions and estrogen response elements ( Zhu, et al. 2019 ). We propose that within our model, it is possible that the loss of Yap1/Wwtr1 and subsequent binding to TEADs, prevents the acquisition and activation of distal estrogen and progesterone receptor response elements and enhancer recruiters. This would explain why the loss of Yap1/Wwtr1 during the receptive window leads to a failure of Pgr and Esr1-regulated genes to be activated despite appropriate expression of receptor levels during this time ( Figure 9 ). The mechanistic aspect of this hypothesis remains to be tested but it suggests novel roles for YAP1/WWTR1 function within the murine uterus. We showed that depletion of Yap1 and Wwtr1 led to a lack of decidualization response and maternal remodeling. YAP1 and WWTR1 are classically thought to respond to extracellular changes in stiffness transduced through the Hippo signaling pathway. Additionally, decidualization and processes during early pregnancy rely on appropriate extracellular changes to alter the maternal endometrial environment( Totaro, et al. 2018 ). They also classically regulate the transcription of extracellular and structural component genes in addition to cell cycle regulatory genes( Totaro, et al. 2018 ). Unsurprisingly, the loss of Yap1/Wwtr1 led to the enrichment of DEG in GO Cellular Component terms like plasma membrane, collagen-containing matrix, and extracellular matrix ( Supplemental Table 5 ). Amongst the upregulated and downregulated DEG of the bulk mRNA sequencing, specific classes of genes included extracellular matrix genes such as collagens, CCNs, laminins, and extracellular remodeling genes such as matrix metalloproteases ( Supplemental Tables 5 and 6 ). However, differential gene expression analysis of whole tissue makes it difficult to determine whether altered wound healing or fibrosis occurs in the YWT females. Given the phenotypic data that show a progressive loss of fertility in the YWT females, it is likely that either aberrant wound repair mechanisms or increased scarring and associated fibrosis are occurring. Both wound healing and fibrosis-associated markers were found with the DEG at 7.5dpc, together with the decrease of angiogenic markers Vegfc , Vegfb , and an increased expression of Pdgfrb , Il13ra1 and Il4ra1 . Angiogenesis is associated with wound healing but is negatively associated with fibrosis( Wynn 2008 ). These results suggest that loss of Yap1 / Wwtr1 leads to decreased wound healing during pregnancy through the downregulation of proangiogenic factors. In addition, IL13RA1 forms a heteromeric complex with IL4RA and activation of this receptor complex is widely considered pro-fibrotic, suggesting that the YWT uteri may be exhibiting increased fibrosis ( Figure 8F ) ( Wynn 2008 ). It is likely that loss of Yap1/Wwtr1 contributes to both a lack of wound healing and increased fibrosis within the murine uterus during pregnancy and this contributes to the repetitive pregnancy loss seen within the YWT females. The incomplete penetrance of phenotype in the YWT females can be attributed to a lack of recombination of one Yap1 allele. Both Progesterone receptor and Yap1 are located on chromosome 9 in mice with Pgr being located at 8899834–8968612 bp on the sense strand and Yap1 being located at 7932000–8004597 bp on the antisense strand (Mgi). Both Pgr and Yap1 are located at 2.46cM and due to the basics of recombination efficiency, the likelihood of recombination for these two genes is 0%. The unfortunate proximity of these genes led to the partiality of the double knockout and the lack of complete phenotype. In addition, this partially explains the lack of recombination and no differences in target mRNA expression ( Figure 1A – C ). Despite the incomplete penetrance of phenotype, we observed significant differences in YWT females compared to controls, which compromised fertility. It is likely that complete conditional uterine knockout of Yap1 combined with Wwtr1 would lead to a much more severe effect, including fully compromised decidualization and fertility due to a lack of maternal remodeling. Currently, we are working to generate cell type-specific knockouts to determine the compartmental contributions of total Yap1 and Wwtr1 knockout.

Successful reproduction is an impeccably complex process. The coordination of intricate molecular signals with large-scale physiological changes imparts many opportunities for error. An unfortunate statistic is that 20% of couples in the United States attempting to reproduce are infertile from either male or female contributions( Chandra, et al. 2013 ). Following fertilization and early embryonic development, an embryo implants into a receptive maternal endometrium that is significantly remodeled throughout pregnancy to support gestation( Cha, et al. 2012 ). In humans and mice, successful reproduction requires the coordinated efforts of many molecular and cellular signals and the appropriate interplay between multiple biological systems, including the neuroendocrine, immune, vascular, and female reproductive tract. Humans undergo spontaneous endometrial stromal cell decidualization, a terminal differentiation of the underlying stroma within the endometrium, which is induced by increasing levels of progesterone during the secretory phase of the menstrual cycle( Cha, et al. 2012 ). This process is critical to control trophoblast invasion and allow appropriate embryo invasion and occurs spontaneously each cycle in the absence of a conceptus( Gellersen, et al. 2010 , Sternberg, et al. 2021 ). In mice, decidualization only occurs when an embryo physically attaches to the maternal endometrial epithelium( Ramathal, et al. 2010 ). Despite these different initiating stimuli, decidualization in both humans and mice is critical for pregnancy success. Therefore, understanding the molecular mechanisms that underlie these complex coordinated events is essential to identifying mechanisms that may go awry and contribute to infertility. Many molecular pathways have been implicated in decidualization and implantation, including WNT, NOTCH, and HIPPO signaling( Gellersen and Brosens 2014 ). Of particular interest is the mechanosensing HIPPO pathway. This pathway was first identified in Drosophila and so named due to mutations in this pathway leading to increased size( Meng, et al. 2016 ). The Hippo signaling pathway in mammals controls organ size, growth, and proliferation and does so by sensing and responding to changes in the extracellular environment, like the presence and tension of nearby cells, as well as growth factor availability( Totaro, et al. 2018 ). The Hippo pathway is a kinase cascade whereby external signals, like those mentioned above, induce the phosphorylation of MST1/2 through a variety of upstream signals. MST1/2 phosphorylates LATS1/2, which then phosphorylates YAP1 and WWTR1, two transcriptional cofactors. YAP1 and WWTR1 are inactive in the cytoplasm and bound or degraded when the HIPPO signaling pathway is activated. Conversely, when the HIPPO signaling pathway is deactivated, YAP1 and WWTR1 translocate to the nucleus and bind their canonical partner transcription factors, the TEADs, to regulate gene transcription of extracellular and cytoplasmic matrix components, cell cycle genes, and act as distal enhancer recruiters( Totaro, et al. 2018 ). This pathway is well conserved among mammals and has been implicated to play a role in female reproductive function. In the murine uterus, YAP1 exhibits dynamic expression with the highest levels of mRNA and protein expression at estrus, an estrogen-dominated stage( Moon, et al. 2022 ). Indeed, YAP1 is significantly increased and phosphorylated in response to 17b-estradiol treatment of ovariectomized mice, suggesting its importance in specific phases of the estrus cycle( Moon, et al. 2022 ). Beyond the estrus cycle, YAP1 is expressed throughout the murine endometrium in early pregnancy prior to embryo implantation at embryonic day 0.5 and beyond( Zhang, et al. 2021 ). Post-implantation, YAP and its targets, Ctgf and Ankrd1 , are significantly increased. A similar response is observed in oil-induced decidualization, suggesting a potential role for YAP1 in maternal preparation of pregnancy( Zhang, et al. 2021 ). The expression of Wwtr1 in the pregnant and nonpregnant murine uterus has not been investigated. In addition, conditional deletion of Yap1 and its homolog, Wwtr1 , under Anti-Mullerian hormone receptor type 2 (Amhr2) driven Cre expression, results in degradation of the oviductal isthmus myosalpinx complicating embryo transport, suggesting an essential role for YAP1 and WWTR1 in maintaining the structural integrity of the female reproductive tract( Godin, et al. 2020 ). However, the authors noted that Yap1 and Wwtr1 were not sufficiently deleted in the murine uterus under Amhr2 Cre and therefore they were unable to determine the contributions of YAP1/WWTR1 during decidualization and pregnancy( Godin, et al. 2020 ). Beyond what is known in the mouse, YAP1 and WWTR1 have been independently identified as being critical for endometrial stromal cell decidualization in vitro . YAP1 expression increases in the first 48 hours of in vitro endometrial stromal cell decidualization( Chen, et al. 2017 ). Targeted knockdown of Yap by short hairpin RNA (shYAP) prior to the induction of in vitro decidualization results in a compromised decidualization response( Chen, et al. 2017 ). In addition, WWTR1 (formerly known as TAZ) increases during in vitro decidualization at day 6, and the knockdown of WWTR1 also compromises the expression of decidualization markers IGFBP1 and dPRL (K Morris et al (unpublished) shared with permission)( Strakova, et al. 2010 ). These studies show the roles of YAP and WWTR during in vitro decidualization, but this potential has yet to be explored in vivo in the murine uterus. In this study, we utilized Progesterone receptor Cre to conditionally target Yap1 and Wwtr1 throughout the female murine reproductive tract to explore the potential roles and regulation of these Hippo homologs in pregnancy establishment and maintenance.

Supplemental Table 1. Antibodies used for western blot and immunohistochemistry analyses. Supplemental Table 2. Primer sequences. Supplemental Table 3. Gene Ontology Enrichment for Biological Process for upregulated DEG. Supplemental Table 4. Gene Ontology Enrichment for Biological Process for downregulated DEG. Supplemental Table 5. Gene Ontology Enrichment for Cellular Component for upregulated DEG. Supplemental Table 6. Gene Ontology Enrichment for Cellular Component for downregulated DEG. Supplemental Table 7. DEG overlap with PRKO mice. Supplemental Table 8. DEG overlap with ERKO mice. Supplemental Table 9. DEG overlap with EGFRKO mice. Supplemental Figure 1. There is not evidence of recombination in Yap1-Wwtr1 triple allele mutant ovaries. A. RT-PCR upstream and B. downstream of exon 3 of Yap in control and YWT uterine DNA indicating heterozygous mutation of Yap . C. mRNA analysis of Yap1, Wwtr1, Ccn1, and Ccn2 of whole ovaries collected at 3.5dpc indicates no difference in expression between floxed controls and YWTs . Bar represents geometric mean ± geometric standard deviation. D. Immunohistochemical analysis of YAP1 and WWTR1 in ovaries at 3.5dpc and semiquantitative HSCORE reveals no difference in protein expression. Bar represents mean. Supplemental Figure 2. Histoarchitecture of Yap1 Wwtr1 triple allele mutants is comparable to control in nulligravid and primigravid females. A. Representative micrographs of trichrome staining of ovaries, oviducts, and uteri of virgin females at proestrus (8 weeks of age). B. Representative trichrome micrographs of ovaries, oviducts, and uteri at 3.5dpc in primigravid females (8–12 weeks of age). C. Representative trichrome micrographs of ovaries, oviducts, and uteri at 3.5dpc in multigravida females (end of breeding trial). Supplemental Figure 3. 3D imaging reveals normal oviductal fold patterning in YWTs. A. Representative micrographs of fluorescent whole mount imaging of oviducts of floxed controls (top panel) and YWT s (bottom panel) at 3.5dpc in primigravid females. Images are whole oviduct (left), longitudinal folds (left middle), transverse folds (right middle), and continuous longitudinal folds (right) (white=Hoechst, blue=E-Cadherin, n=4). B. Representative micrographs of fluorescent imaging of embryos (marked with adjacent asterisk) within uteri at 3.5dpc (white=Hoechst, E-cadherin=blue). C. Embryos counted in uterine horns at 3.5dpc with whole mount fluorescent uterine imaging (n=6–7 females/genotype). Supplemental Figure 4. Estrogen receptor signaling is not aberrant in YWT s. A. Esr1 and target genes, Muc1, C3, Muc4 mRNA expression in 3.5dpc uteri. B. Immunohistochemical analysis of Estrogen receptor expression and semiquantitative HSCORE of positive DAB stain in 3.5dpc uteri.