The presence of blastocyst within the uteri facilitates lumenal epithelium transformation for implantation via upregulating lysosome proteostasis activity

ABSTRACT The mammalian endometrium is covered by the lumenal epithelium (Le), which directly interacts with the blastocyst and plays an important role in the establishment of reciprocal crosstalk between the embryo and receptive uterus during implantation. However, the effect of the blastocyst on uterine differentiation during the window of receptivity is far from well understood. Through transcriptomic profiling of the uterine Le isolated by laser capture microdissection (LCM), it was demonstrated that global gene expression changes occurred in Le between pseudopregnant mice without embryos and pregnant mice with embryos. Some differentially expressed genes, including upregulated Areg (amphiregulin), Ihh (Indian hedgehog), Lifr (leukemia inhibitory factor receptor) and downregulated Msx1 (msh homeobox 1), Pgr (progesterone receptor), and Gata2 (GATA binding protein 2) in pregnant mice, have been reported to regulate the establishment of uterine receptivity. Besides, we found that blastocysts induced an increase in both the number and acidification of lysosome, consistent with enhanced lysosomal hydrolase activity in uterine Le. Further exploration uncovered that blastocyst-derived IGF2 was involved into the activation of epithelial STAT3 to induce lysosomal hydrolase expression, and inhibition of lysosomal function derails both uterine receptive maker gene expressions and embryo implantation. Finally, based on the proteomic data of both epithelia and the separated lysosome, it was revealed that CLDN1 (claudin 1) and MUC1 (mucin 1, transmembrane), two well-known downregulated molecules for successful implantation, are degraded by epithelial lysosome. In brief, our data demonstrated that blastocysts induced normal epithelium differentiation with lysosome activation to promote the uterine epithelial differentiation for embryo implantation. Abbreviations ACTB: actin beta; AREG: amphiregulin; ATP6V0A4: ATPase, H+ transporting, lysosomal V0 subunit A4; Baf A1: bafilomycin A1; BSA: bovine serum albumin; CLDN1: claudin 1; CTSB: cathepsin B; DEGs: differentially expressed genes; E2: 17β-estradiol; ESR: estrogen receptor; GATA2: GATA binding protein 2; GLA: galactosidase, alpha; GO: gene ontology; HBEGF: heparin-binding EGF-like growth factor; IGF1R: insulin-like growth factor 1 receptor; Ihh: Indian hedgehog; ISH: in situ hybridization; LAMP1: lysosomal-associated membrane protein 1; LCM: laser capture microdissection; Le: lumenal epithelium; LGMN: legumain; LIF: leukemia inhibitory factor; LIFR: LIF receptor alpha; MSX1: msh homeobox 1; MUC1: mucin 1, transmembrane; P4: progesterone; PBS: phosphate-buffered saline; PCA: principal component analysis; PPT1: palmitoyl-protein thioesterase 1; PGR: progesterone receptor; PSP: pseudopregnancy; PTGS2/COX2: prostaglandin-endoperoxide synthase 2; qPCR: quantitative real-time polymerase chain reaction; SP: pregnancy; TFEB: transcription factor EB.


Introduction
The reciprocal interaction between the blastocyst and receptive uterus is essential for embryo implantation [1].Uterine receptivity is a specific state in which the endometrium can receive the blastocyst for implanting.In mice, the major ovarian hormones that specify uterine receptivity are estrogen (17β-estradiol, E 2 ) and progesterone (P 4 ), whose functions are primarily executed by the nuclear ESR (estrogen receptor) and PGR (progesterone receptor) proteins, respectively [2].The entry into the uterine receptivity requires endometrium priming with P 4 , superimposed with estrogen on day 4 morning (with day 1 being the presence of vaginal plug).Blastocysts enter into uterine cavity on the early morning of day 4 (D4) and attach to the lumenal epithelium (Le) by midnight on D4 [3].The uterine Le is the first site in the maternal side contacting an implanting embryo [4], and blastocyst produced HBEGF (heparin-binding EGF-like growth factor) has been reported to mediate the embryo-uteri interaction for the attachment [5][6][7].However, whether the molecular crosstalk between the blastocysts and uterus influencing the E 2 and P 4 directed uterine receptivity before the embryo attachment still remains elusive.
The lysosome has been known as a membrane-enclosed cytoplasmic organelle responsible for the degradation of a variety of biological macromolecules, including proteins, lipids, carbohydrates and nucleic acids [35,36].Most of its functions depend on protease activities found in the lysosomal lumen.Lysosomal proteolytic capacity relies mainly on three different enzyme families: the pepsin-related aspartyl proteases (CTSD [cathepsin D] and CTSE), the papain-like cysteine proteases (e.g., CTSB and CTSL) and a distinct cysteine proteases known as LGMN/AEP (legumain) [37][38][39][40].Ultrastructural studies of the uterine epithelium have revealed changes of lysosome during the peri-implantation stage [4].The lysosomes appeared more active in the Le on D6 of pregnancy compared to the non-pregnant state in rat [41].Transmission electron microscopic study of bovine and mouse endometrium also detected lysosomes in Le [42,43].In addition, uterine epithelial acidification during embryo implantation implied important role of lysosomes in mouse uterine Le [44,45].However, the molecular mechanisms regulating the changes of lysosomes, and the physiological significance of lysosomes in mouse uterine Le have not been well identified.
In this study, pregnant and pseudopregnant mouse uterine Le on D4 at 10:00 and 17:00 were isolated through laser capture microdissection (LCM) and subjected to RNA-seq.
We found that blastocyst imparted a previously unnoticeable influence on uterine Le differentiation, and further exploration uncovered that blastocyst-derived IGF2 upregulated the expression of lysosomal hydrolase via STAT3.Finally, lysosome inhibition by V-ATPase inhibitor bafilomycin A 1 (Baf A1) led to failed embryo implantation, and combined proteome analysis demonstrated that CLDN1 and MUC1 were two critical substrates for degradation by the lysosome in uterine Le.Our study provides a valuable resource for deciphering the communication between blastocyst and uterine Le for uterine epithelial differentiation, which is prerequired for successful implantation in mice.

Blastocyst induced uterine lumenal epithelial differentiation on D4 of pregnancy
To explore the potential influence of blastocysts upon uterine Le, LCM was applied to collect pregnant and pseudopregnant mouse uterine lumenal epithelia for RNA sequencing (RNAseq) (Figure 1A).The epithelia were isolated at two time points on D4 at 10:00 and 17:00 respectively, which represent the early and late phase since the blastocysts enter into the uteri.Our RNA-seq data indicated that the uterine Le at 17:00 on D4 of pregnancy (D4 17:00) was significantly different from the other 3 groups (Figure 1B).The principal component analysis (PCA) analysis showed that pregnant and pseudopregnant mouse uterine Le on D4 at 10:00 (SP D4 10:00 and PSP D4 10:00) were similar, while D4 17:00 showed dramatic differences compared with others (Figure 1C).As shown in the volcano plot (Figure 1D), a total of 1973 differentially expressed genes (DEGs) were found (fold change ≥ 2, P < 0.05) in D4 17:00 Le compared to D4 10:00, with 1077 down-regulated genes and 896 up-regulated genes.To further explore the effects of communication between the blastocysts and Le, we compared the gene expression between the PSP D4 17:00 group and D4 17:00, and found that 456 genes were downregulate and 528 genes were upregulated (fold change ≥ 2, P < 0.05) (Figure 1E).Gene ontology (GO) analysis revealed that upregulated genes in the D4 17:00 Le in comparison to PSP D4 17:00 mainly enriched in the terms of embryonic organ development, regulation of intrinsic apoptotic signaling pathway in response to DNA damage, ERK1 and ERK2 cascade, regulation of pH and response to pH, etc. (Figure 1F).The downregulated genes were enriched in calcium ion transport, actin filament organization, retinol metabolic process, regulation of actin cytoskeleton organization, etc. (Figure 1G).Interestingly, we noticed that several implantation related genes, such as Areg, Ihh, Lifr, Msx1, Pgr, Gata2, were significantly changed, which is confirmed by both the in situ hybridization (ISH) and quantitative real-time polymerase chain reaction (qPCR) (Figure 1H, Figure S1B).For protein expression, we noticed that the PR and GATA2 were obviously downregulated in the epithelia until on D5 (Figure S1C).Areg, Ihh, Lifr are reported highly expressed in the Le on D4 morning, and our observation revealed that these genes were upregulated on D4 afternoon in the presence of blastocysts.Msx1, Pgr and Gata2 were reported to be down-regulated for implantation, and our data suggested that their dynamic expression is related with the blastocyst, since this downregulation was not observed in the pseudopregnant mice on D4 afternoon.

STAT3 regulated the lysosomal hydrolase expression in uterine lumenal epithelia
The transcriptional factor TFEB was the master regulator for lysosome genesis through transcription regulation of various lysosomal hydrolases [59,60].Accordingly, we detected the expression of Tfeb in uterine Le.RNA-seq results showed that the mRNA level was very low (Figure S3C), consistent with the undetectable level of TFEB in the uterine Le through IHC assay (Figure S3D).Recently, STAT3 was reported as a new regulator for lysosomal hydrolase expression under certain cellular context [40,61,62].Moreover, the Lifr, which activate the downstream STAT3 was significantly upregulated in the lumenal epithelia and uterine Lif ligand expression was also induced when the blastocysts were in the uteri (Figure 1H, Figure S4).These motivated us to examined the dynamic expression of phosphorylated STAT3 in the uterus on D4 10:00/ 17:00 and PSP D4 17:00.As shown in Figure 3A, on D4 10:00, the active STAT3 was low in the epithelia, and then, STAT3 was activated in the whole layer of epithelia at D4 17:00, and no obvious difference was observed for p-STAT3 in lumenal epithelial cells between lumenal cells nearby or far from the embryo on D4 14:00/17:00 (Figure S2B).Interesting, the active STAT3 displayed an increase in the epithelium on D4 17:00 compared with PSP D4 17:00.Immunoblotting assay for isolated epithelia further confirmed the increased STAT3 phosphorylation on D4 17:00 (Figure 3B).Next, to explore whether STAT3 was critical for the lysosome genes' expression, mice were treated with STAT3 inhibitor HO-3867.After confirmation of the blockage of STAT3 activation (Figure 3C), we found that several lysosome-associated genes, including Atp6v0a4, Lgmn and Ctsb, exhibited reduced expression at RNA level (Figure 3D,E).Accordingly, the proteins of CTSB and lumenal epithelial cells on D4 17:00 of pregnant and pseudopregnant mouse.(H) ISH of Areg, Ihh, Lifr, Msx1, Pgr and Gata2 mRNA in the pregnant and pseudopregnant mouse uteri on D4 10:00 and 17:00.Scale bar: 100 μm.Le: lumenal epithelium; Ge: glandular epithelium; S: stroma; SP: pregnancy; PSP: pseudopregnancy.LGMN were also reduced in the lumenal epithelia in STAT3 inhibitor treatment group compared to the control group (Figure 3F).In addition, in the delayed implantation model, when the P 4 treated ovariectomized mice were supplied with E 2 to initiate the implantation, the increased phosphorylated STAT3 activation was observed in epithelia and consistently, the lysosome was also activated with upregulated hydrolase expression and activity (Figure S5A-E).

Blastocyst-derived IGF2 induced the increase of lysosomal hydrolase expression in mouse uterine lumenal epithelia
Next, we intended to explore how the uterine lumenal epithelial STAT3 was enhanced by the blastocysts to induce the expression of lysosomal hydrolase.It is conceivable that the secretory proteins orchestrate communication between the blastocyst and endometrium, and in our previous study, we uncovered that the epithelial expressed IGF1R (insulin-like growth factor 1 receptor) response to blastocyst produced IGF2 [25].Thus, we speculated that the IGF2 protein secreted by the blastocyst may trigger the increase of lysosomal hydrolase through STAT3 in uterine Le.Initially, ISH data showed that Igf2 was barely expressed in the blastocyst on D4 10:00 and was later dramatically increased in the trophectoderm on D4 17:00 (Figure 4A).The increased mRNA expression of Igf2 in blastocysts was further confirmed by qPCR (Figure 4B).We then transferred agarose beads coated with recombinant IGF2 into the pseudopregnant uteri on D4.The IGF2-coated beads could upregulate phosphorylated STAT3 in the epithelia compared with the bovine serum albumin (BSA)-coated beads as revealed by the western blotting (Figure 4C).The IGF2 treatment also induced Atp6v0a4 expression, and upregulation of Lgmn and Ctsb expression in uterine Le on D4 17:00, while the uterine Le around the BSA-coated beads did not show obvious elevation in Atp6v0a4, Lgmn or Ctsb level (Figure 4D).Accordingly, the activities of CTSB and LGMN were higher in uterine Le with the IGF2-coated beads on D4 17:00 (Figure 4E).IGF2 is a peptide that bind to three different types of receptors, including IR (insulin receptor), IGF1R, IGF2R (insulin-like growth factor 2 receptor) [63].Embryonic derived IGF2 interacts with IGF1R causes the downstream activation of STAT3 and MAPK1/ERK2-MAPK3/ERK1 signaling pathways [25].To block the interaction between the embryo and Le through the IGF2, the uterine IGF1R specific deletion mouse model was utilized.The phosphorylated STAT3 staining was observed in the Igf1r f/f epithelium but not in the Igf1r d/d epithelium on D4 17:00 (Figure 4F).Next, the uterine epithelia were isolated and prepared for the RNAseq.The results exhibited that Atp6v0a4, Lgmn and Ctsb were all aberrantly expressed in Igf1r d/d uterine Le compared with the Igf1r f/f control (Figure 4G).qPCR data also confirmed abnormal decrease of Atp6v0a4, Lgmn and Ctsb in uterine Le on D4 17:00 knockout mouse (Figure 4H).To more clearly demonstrate the regulation of IGF2-STAT3 signal cascade for lysosome activation, the in vitro culture epithelial organoid, which could be induced to differentiation by estrogen and progesterone [64,65], were treated by recombinant IGF2.our data suggested that recombinant IGF2 could induce an increase of STAT3 phosphorylation, and both the expression and activity of CTSB and LGMN in epithelial organoids (Figure 4I,J and Figure S6).These results implied that IGF2-IGF1R signaling was vital for STAT3 activation to regulate lysosomal hydrolase expression in uterine Le.

V-ATPase inhibitor Baf A1 resulted in embryo implantation failure
To explore whether the increased lumenal epithelial acidification was important for the establishment of uterine receptivity, a specific V-ATPase inhibitor Baf A1 was injected via uterine fat pad to inhibit the lysosome function [45].Local uterine fat pad injection of 2.5 μg/kg Baf A1 could efficiently suppress acidic pH in the uterus (Figure 5A,B).Local uterine fat pad injection of Baf A1 suppressed blastocysts implantation on D5 (Figure 5C), consistent with previous reports [45].The failed implantation was further revealed by the aberrantly expressed MUC1 and disappearance of PTGS2/COX2 (prostaglandin-endoperoxide synthase 2) in both epithelia and stroma surrounding the embryos (Figure 5D).We also examined the receptive state of uteri on D4 afternoon, and uncovered that epithelial Msx1 exhibited aberrant high level in Baf A1-treated mice, and other genes, such as Lifr, Ihh and Areg showed a low expression (Figure 5E,F).These data indicated that uterine epithelial acidification is important for uterine differentiation and embryo implantation.

Lysosomes in epithelia degraded the proteins of CLDN1 and MUC1
A lumen pH of ~ 4.5-5.0 in lysosome is optimal for > 60 different hydrolytic enzymes to break down biomolecules within intracellular and extracellular origins [44].Therefore, we wonder how the lysosomes modulate the proteome homeostasis in uterine Le through their degradation function.Protein extract for both the isolated lumenal epithelial lysosome and whole lumenal epithelia were used for mass spectrum-based proteomics analysis (Figure 6A,B).We performed biological function distribution analysis using the Metascape database on the lysosomal proteome.Besides the constituent component of lysosome, our proteomic analysis uncovered multiple processes, including localization within membrane, actin cytoskeleton organization, regulation of vesiclemediated transport (Figure 6C).Next, we intersected the differentially changed proteins of quantitative proteome data on D4 10:00 and 17:00 of the pregnant and pseudopregnant mouse uterine lumenal epithelia.ACTB/β-actin was used as a loading control.(G) Western blotting analysis of CTSB and LGMN protein on D4 10:00 and 17:00 of the pregnant and pseudopregnant mouse uterine lumenal epithelia.ACTB was used as a loading control.Le: lumenal epithelium; SP: pregnancy; PSP: pseudopregnancy.Error bars, mean ± SEM, n ≥ 3. The data were analyzed with a one-way ANOVA.n.S. ***P < 0.001, ****P < 0.0001.
in lumenal epithelia on D4 10:00/17:00 with lysosomal proteome and found that 133 proteins were significantly increased, including ATP6V0A4, LGMN and CTSB (Figure 6D,F).Interesting, 90 proteins exhibited significantly downregulation (e.g., CLDN1, CLDN23, MUC1) (Figure 6E,  F).Reduced epithelial cell polarity is prerequisite for prereceptive uterus differentiation into a receptive state [8,66].The tight junction molecules CLDN1 was significantly downregulated from D4 morning to evening [25].We speculated that the protein of CLDN1 was degraded by lysosome in uterine lumenal epithelia.Co-localization of LAMP1 and CLDN1 was observed in the apical pole of epithelia and the signal intensity of CLDN1 in the cell membrane was significantly lower on D4 17:00 than D4 10:00 (Figure 6G,H).The CLDN1 protein was significantly higher and showed more colocalization with LAMP1 in the Baf A1-treated uterus on D4 17:00 compared with control (Figure 6I), which was further supported by the western blotting (Figure 6J).Immunohistochemical staining and double immunostaining of MUC1 and LAMP1 showed that MUC1 was also downregulated from D4 10:00 to 17:00 and was degraded by lysosome during the establishment of uterine receptivity (Figure 6K,L).Moreover, in the pseudopregnant mice in D4 17:00, the MUC1 and CLDN1 are still highly expressed in the lumenal epithelia (Figure S7).These results indicated that the critical molecule in epithelia for implantation, exemplified as CLDN1 and MUC1, can be downregulated by protein degradation in lysosome.

Discussion
Reciprocal interactions between blastocyst and uterine Le are important for early pregnancy success.Previously, several groups have reported comprehensive analyses of genes expressed in early pregnant uteri [67][68][69][70][71].However, these studies focused on the different expression in uteri between the pre-receptivity (D3) and receptivity (D4) or compared the whole uteri before or after implantation.In this study, we focused on uterine lumenal epithelial cells that directly interacted with blastocyst during the establishment of uterine receptivity governed by D4 nidatory estrogen before blastocyst attachment.Impressively, our data showed that the dynamics of DEGs is not only due to the estrogen signal, but also related with the blastocyst derived signal to induce lumenal epithelial cells differentiation.Combined the previous reports and this study, a new scenery was displayed as that based on the P 4 priming, the D4 nidatory estrogen prepared the uteri and the blastocysts further fine-tuned the epithelia to become more receptive for the blastocyst.Therefore, the blastocyst can induce lumenal epithelial cells differentiation facilitating the embryo implantation.Indeed, based on the observation in IVT-ET (In Vitro Fertilization and Embryo Transfer) practice, it was uncovered that implantation of each embryo facilitates the chance of remaining embryo [72,73], which suggested the embryo's influence on endometrium [74].Previous reports about the embryonic derived signal, such as HBEGF and S100A9 [5,75], mainly focused on the interaction between the embryo and surrounding uterine epithelia in the implantation chamber, which mediated the crosstalk during the attachment.In this study, this influence of blastocyst on the epithelia occurred more earlier and more broadly to the whole epithelial layer.
Maternal LIF is indispensable for embryo implantation, since its absence lead to defective uterine receptivity and complete implantation failure [76,77].Under the nidatory surge of estrogen on the morning of D4 of pregnancy and pseudopregnancy in mouse, the mRNA level of LIF was expressed most strongly in glandular epithelium.LIF executed its function though LIFR receptor [18][19][20][21] and is primarily responsible for STAT3 activation in the Le [24].Our data showed dramatic upregulation (>4 folds) of Lifr in D4 17:00 uterine Le compared to PSP D4 17:00.Consistent with the increase of Lifr, the active STAT3 phosphorylation in uterine Le on D4 17:00 was also higher than that on PSP D4 17:00.The rapid Lif induction by estrogen is responsible for downregulating Msx1 expression [10], and Msx1 expression persists in Lif -/-and Lifr conditional knockout mouse uterus [18,78].Our data show that Msx1 expression persists in D4 17:00 pseudopregnant mouse uteri compared to D4 17:00, suggesting that LIF signal cannot transduct properly to downregulate Msx1 on D4 17:00 uterine lumenal epithelia without elevated expression of Lifr in the absence of blastocysts.
Though the Lif-STAT3 signal cascade is considered as a critical regulator for the epithelial differentiation to establish the uterine receptivity, the downstream effector still remains elusive.Previous papers have reported that conditionally inactivation of STAT3 in uterine epithelia caused dysregulated protein expression of MUC1 and CLDN1 within the implantation window [27,28].In addition, phosphorylated STAT3 positive signals in lumenal epithelia was markedly reduced and CLDN1, MUC1 were significantly higher in mice with uterine-specific knockout of IGF1R [25].In our study, we found that the activation status of STAT3 and lysosome activation are tightly associated in the presence of blastocyst, and more importantly, the acute inhibition of STAT3 through the pharmacological inhibitor just before the embryo implantation led to the downregulation of lysosome proteases.This suggested that one of downstream effects of STAT3 in the epithelia was to regulate the lysosome genesis, which have been reported in another cell context [40,61,62].However, as a transcriptional factor, whether STAT3 directly binds and regulates these genes' expression deserve further exploration.The integrating of ChIPseq data the RNA-Seq data would provide a more comprehensive scenery to understand the molecular mechanism for STAT3 in the uterine epithelia.
Based on our novel finding that the embryo induced the epithelial lysosome activation, we further explore the function of lysosome in the aspect of protein degradation, which is the classical function of this organelle.Combining the protein profile of both the whole epithelia and fractioned epithelial lysosome, several target proteins were found localized in the lysosome and downregulated in the epithelia.These targets are thought to be downregulated during the establishment of uterine receptivity via the lysosome mediated protein degradation.MUC1 and CLDN1, which are well-known to behave as a physical barrier for implantation, have been reported to be downregulated.Our finding provided the first evidence that how this physical barrier was removed at protein level.Additionally, Mucins are a family of high molecular weight, highly glycosylated glycoproteins [79], and are delivered to lysosomes for catabolism either by endocytosis from outside the cell or by macroautophagy/ autophagy within the cell [80].Once inside the lysosome, glycoproteins are broken down by a combination of   proteases and glycosidases.Indeed, lysosomal proteases CTSD can mediates endogenous mucin degradation in mammals [81].CLDN1 was also reported to be degraded in the lysosomal by the CTSL in the intestines [82].Whether the similar mechanism for two molecules existed in the uterine epithelia remain unclear at present.Therefore, how MUC1 and CLDN1 are selected to enter the lysosome, and what hydrolases are responsible for the degradation of MUC1 and CLDN1 needs further exploration.
Besides the protein degradation function, lysosomes are also involved in endocytosis in the mouse and rat uterine epithelium during the early pregnancy [83,84].Early studies also revealed increased lysosomal activities in endometrial glandular epithelia during secretory phase, which may contribute to the enhanced secretion from glandular epithelia and imply a role of lysosomes in exocytosis [44].Extracellular vehicles (EVs) are present in the uterine fluid and the corresponding mucus in woman [85], and lysosomes can affect the fate of the EVs, e.g., sequestered, degraded, or excreted via exocytosis [86].In addition to the functions of lysosomes described above, in recent years they have been found to participate in many other cellular processes, including plasma membrane repair, cell adhesion and migration, apoptotic cell death and metabolic signaling [56].Therefore, the other function of epithelial lysosomes during the establishment of uterine receptivity deserves further exploration.
In summary, our study demonstrated that blastocysts facilitate embryo implantation via protein degradation in the uterine epithelia.We found that blastocyst-derived signal attended into the activation of epithelial STAT3, accompanied with upregulation of epithelial Lifr expression, to induce lysosomal hydrolase expression (Figure 7), and the activated lysosomes degraded CLDN1 and MUC1.Our findings will remarkably advance our knowledge on current study of uterine preparation for implantation and inspire future studies for human fertility improvement.

Mice
Adult females ICR mice were purchased from the laboratory Animal Center (Xiamen University).Female mice carrying floxed Igf1r (Igf1r f/f ) were obtained from Jackson Lab [87], and crossed with PR Cre/+ mice.Female mice were mated with fertile males, and the mating was confirmed the vaginal plug was denoted as D1.The feeding conditions according to the guidelines of the Animal Welfare Committee of Research Organization (X200811) of Xiamen University with a 12/12 h light/dark cycle (lights off at 1900 h) with access to chow and water at libitum.All experiments were conducted according to the approved guidelines of the Animal Welfare Committee of Research Organization (X200811) of Xiamen University.

Pregnancy and pseudopregnancy
Female mice were mated with fertile or vasectomized males of the same strain to induce pregnancy or pseudopregnancy by co-caging, respectively.On D4, pregnancy was confirmed by recovering blastocysts from the uterus.The implantation sites on D5 were identified by intravenous injection of 0.1 ml of 1% Chicago Blue (Sigma, C8679) in 0.85% sodium chloride and implantation sites were demarcated by discrete blue bands [88].Uterine tissues were collected from animals at indicated time points on D4 or D5.

Laser capture microdissection
LCM was performed as previously described [89,90].To isolate the lumenal epithelia, frozen sections of pregnant mouse and pseudopregnant uteri (20 µm) were mounted on polyethylene naphthalate slides (Leica Microsystems, 11600289).The target regions in the sections were dissected using the Leica LMD7 (Leica Microsystems, Wetzlar, Germany).The uterine tissues from 3 mice for each group.Total RNA was extracted from collected cells using RNeasy Micro Kit (Qiagen, 74004) and eluted with 14 μl of RNase-free water.

Isolation of mouse uterine epithelial cells
Uterine horns were excised, trimmed of fat, and cut into 50mm-long sections and placed into 25 g/L Trypsin (Sigma, T8003) for 6-8 min at 37°C.Then slowly pushed from one side of the uterus to the other, squeezed out lumenal epithelial cells.The cells were washed twice in Dulbecco's phosphatebuffered saline (Sigma, D8537) and repelleted.The pellets were lysed with TRIzol reagent (Invitrogen,15596018) for RNA extraction or with radioimmunoprecipitation assay RIPA lysis buffer (Beyotime, P0013B) for the preparation of whole-cell protein lysates.

qPCR analysis
Total RNA was isolated from lumenal epithelial cells of mouse with TRIzol (Invitrogen,15596018) and converted to cDNA according to the manufacturer's instructions (Takara, RRO36A).The cDNA was amplified by qPCR to quantify gene expression using gene-specific primers and SYBR Green (Takara, RR820B).Analysis of variance single-factor analysis was conducted on the grouped means to determine statistical significance at a significance level of P < 0.05.All primers for qPCR are listed in Table S1.The relative expression of the mRNA was calculated by the value of 2 −ΔCt .Relative mRNA expression levels were normalized to Gapdh.Assays were performed at least three times with each in duplicate.

RNA-seq and data analysis
Total RNA was extracted from mouse uterine Le by TRIzol (enzymatic isolated epithelia in Igf1r f/f and Igf1r d/d mice) or RNeasy Micro Kit (mouse uterine Le collected by LCM on D4 10:00/17:00) according to the respective protocol.Poly A + -RNA was prepared and subjected to RNA sequencing at BGI Genomics Co., Ltd.(Shenzhen, China).DEGs were normalized to fragments per kilobase of exon model per million mapped reads (FPKM) using the EdgeR3.9package in R with the criteria of fold change significantly greater than 2 and P < 0.05.The visualization of RNA-Seq data was done by ggplot2 package in R.

In situ hybridization (ISH)
ISH was performed as previously described [7].In brief, frozen sections (10 µm) of mouse uterine tissues were fixed in 4% paraformaldehyde in PBS solution at 4°C for 40 min.Digoxin (DIG) labeled mouse-specific probes for Atp6v0a4, Ctsb, Lgmn, Gata2, Pgr, Msx1, Lifr, Areg, and Ihh, were used for hybridization and incubated overnight at 55°C.After a series of washed, blocking sections for 1 h, then incubated overnight with anti-digoxin antibodies (1:2500; Roche, 11277073910) at 4°C.Using BCIP/NBT (Beyotime Biotechnology, C3206) as the substrate, the staining signal was visualized according to the manufacturer's instructions.Subsequently, the sections were stained with 0.1% nuclear fast red solution (Solarbio, G1321) and sealed.All primers for ISH probe production are listed in Table S2.

Transmission electron microscopy
Tissues were cut into to 3-mm pieces, and then were fixed in 3% glutaraldehyde and then embedded in epoxy resin.Ultrathin sections were shaved by ultramicrotome (Leica EMUC7) and stained by both uranyl acetate and lead citrate.Ultrastructure of lysosomes were observed under a transmission electron microscope (JEM-2100HC).

LysoSensor Green DND-189 staining
To investigate uterine epithelial acidification, LysoSensor Green DND-189 was used as previously described [45].LysoSensor Green DND-189 (YEASEN, 40767ES50) is an acidotropic fluorescence probe that becomes fluorescent only when it is inside acidic compartments (pKa˜5.2) and its fluorescence intensity correlates with intracellular acidity thus serves as an intracellular pH indicator [57,[93][94][95].Two µl of LysoSensor Green DND-189 (1 µM) was injected into both uterus horns of pregnant and pseudopregnant mice, ~30 min before dissection, then frozen sections of uterine sections were heated at 55°C for 20 min and detected at 488 nm and emission at 505 nm.At least 3 mice were examined.The lumenal epithelial cells were maintained at 37°C in an atmosphere of 5% CO 2 :95% air in DMEM-F12 medium (Gibco, 11039-021) supplemented with 10% (v:v) FBS.According to the manufacturer's instruction by incubating cells with LysoSensor Green DND-189 (1 µM) and Hoechst 33,342 for 1 h at 37°C, cells were then washed in PBS buffer.

Uterine fat pad injection
This method was established in rats by Dr. Koji Yoshinaga [96][97][98] and was adapted in mice to deliver drugs into the uterus without disturbing the intrauterine environment [99].Yong virgin ICR females were mated naturally with ICR males.Plugged females were randomly distributed into control group and V-ATPase inhibitor Baf A1treated groups.Baf A1 (MedChemExpress, HY-100558) specifically inhibits V-ATPase via binding to V0 domain [100].On D3 at 18:00 h, a small incision was made in the dorsolateral region on the right side of flank under anesthesia with isoflurane inhalation.Five microliter vehicle (20% of ethanol in 1 × PBS with blue dye to monitor the injection) or Baf A1 (2.5 μg/kg in vehicle) was injected at 1-2 spots on the adipose tissue next to the right uterine horn only.Implantation sites were detected using blue dye injection on D5 [101].Uterine tissues were snap-frozen and kept in −80°C for in situ hybridization and some were also processed for LysoSensor Green DND-189 staining as described above.

Delayed implantation model
Pregnant mice of ICR were ovariectomized at 8:00-10:00 on D4 of pregnancy to induce delayed implantation.Delayed implantation was maintained by injecting P 4 (2 mg per mouse; Sigma) from D5 to 7. E 2 (25 ng per mouse) was given to P 4 -primed delayed-implantation mice to terminate delayed implantation.The mice were killed to collect uteri and Le 6 h after E 2 treatment.The mice of delayed implantation were killed to collect uteri and Le 6 h after P 4 treatment on D7.

Beads transfer into the pseudopregnant mice
The preparation of IGF2-carrying beads and transferring beads into uteri lumen were performed as previously described [6].Affi-Gel Blue Gel (Bio-Rad, 153-7302; 100-200 mesh) beads about the size of a blastocyst were washed six times with sterile PBS and then incubated with IGF2 recombinant protein (100 ng/µl; R&D systems, 792) or similar concentrations of BSA (Sigma, V900933), in 20 µl PBS several times and used immediately.Siliconized pipette tips and dished were used during the transfer.Mice were sacrificed on D4 17:00.

Organoid cultures
For mice uterine epithelial organoid derivation, epithelia were dissected from estrous mouse uteri.The epithelia are carefully separated and digested with 6 mg/ml dispase, (Gibco,17105) and 25 mg/ml trypsin (Sigma, T8003) at 37°C for 30 min, inactivation with 1% FBS in DMEM/F12.The digested cells were passed through a 70-μm filter to obtain the epithelia.Dispersed epithelia cells were mixed with 70% Matrigel (BD Bioscience, 354234) in DMEM/F12, seeded, and maintained in culture as described previously [65].Mouse epithelial organoids were either treated for 2 days with E 2 (1 nM) followed by another 2 days with P4 (50 ng/ml) or for 2 days with E 2 followed by another 2 days with P 4 (50 ng/ml) and IGF2 recombinant protein (100 ng/ml).

Qualitative and quantitative proteomics
Protein lysis from lysosomes and lumenal epithelial cells on D4 10:00 and/or 17:00.Proteins were digested, and subjected to liquid chromatography-MS in the school of Life Sciences, Xiamen University.

Figure 1 .
Figure 1.Blastocyst-induced lumenal epithelial differentiation during the establishment of uterine receptivity.(A) Workflow of the establishment of LCM.After lumenal epithelium collection and RNA extraction, RNA was identified by RNA sequence.LCM: Laser capture microdissection.(B) Heatmap of differentially expressed genes (DEGs) in RNA-seq for pregnant and pseudopregnant mouse uterine lumenal epithelia on D4 10:00 and 17:00.FPKM of each gene from different samples was normalized to Z-score.(C) score plot of PCA of 4 groups samples.PC1 and PC2: principal component 1 and principal components 2. Black and orange dots represent Le cells on D4 at 10:00 and 17:00 of pregnant mouse uterus, respectively.blue and pink dots represent Le on D4 at 10:00 and 17:00 of pseudopregnant mouse uterus, respectively.(D, E) Volcano plot of DEGs in pregnant and pseudopregnant mouse uterine lumenal epithelial cells on D4 10:00 and 17:00 based on RNA-seq analysis.Blue dots represent downregulated DEGs and red dots represent upregulated DEGs (fold change ≥ 2, P < 0.05).(F, G) GO Biological Process terms for DEGs in

Figure 2 .
Figure 2. Blastocyst induced an increase of lysosomal hydrolase expression in uterine lumenal epithelia.(A) Heatmap of DEGs for pregnant and pseudopregnant mouse lumenal epithelial cells on D4 at 10:00 and 17:00.FPKM of each gene from different samples was normalized to Z-score.(B) ISH of Atp6v0a4, Lgmn, and Ctsb mRNA in the pregnant and pseudopregnant mouse uteri on D4 at 10:00 and 17:00.Scale bar: 200 μm.(C) qPCR analysis for mRNA level of Atp6v0a4, Lgmn, and Ctsb in uterine lumenal epithelial cells on D4 10:00 and 17:00 of pregnant and pseudopregnant mouse.The values are normalized to Gapdh.(D) Ultrastructure of uterine lumenal epithelia in the pregnant and pseudopregnant mouse on D4 10:00 and 17:00 (Scale bars: 5 μm) and detailed images of acidic organelles in Le.Red arrows indicate lysosome in Le. (E) Immunostaining of LAMP1 (red) and LysoSensor staining (Green) of pregnant and pseudopregnant mouse uteri on D4 at 10:00 and 17:00.Nuclei were stained with DAPI (blue).White arrow indicates the LAMP1 expression in mouse uteri.Scale bar: 50 μm.(F) Western blotting analysis of LAMP1 protein

Figure 4 .
Figure 4. Blastocyst-derived IGF2 induced the increase of lysosomal hydrolase expression in mouse uterine lumenal epithelia.(A) ISH of Igf2 mRNA expression in the pregnant mouse uteri on D4 at 10:00 and 17:00.Scale bar: 50 μm.Arrows indicate signal in blastocyst.Bl: Blastocyst.(B) qPCR analysis for Igf2 mRNA expression in the blastocysts on D4 at 10:00 and 17:00.The values are normalized to Gapdh.(C) Western blotting analysis of p-STAT3 and STAT3 proteins induced by IGF2-coated beads or BSA-treated beads in uterine lumenal epithelial cells on D4 17:00.ACTB was used as a loading control.(D) qPCR analysis for Atp6v0a4, Lgmn, and Ctsb mRNA expression induced by IGF2-coated beads or BSA-treated beads in uterine lumenal epithelial cells on D4 17:00.The values are normalized to Gapdh.(E) Western blotting analysis of CTSB and LGMN proteins induced by IGF2-coated beads or BSA-treated beads in uterine lumenal epithelial cells on D4 17:00.ACTB was used as a loading control.(F) Immunostaining of p-STAT3 in Igf1r d/d and Igf1r f/f uteri on D4 17:00.Scare bar: 50 μm.(G) Heatmap of DEGs in Igf1r d/d and Igf1r f/f uterine lumenal epithelium on D4 17:00.FPKM of each gene from different samples was normalized to Z-score.(H) qPCR analysis for Atp6v0a4, Lgmn, and Ctsb mRNA expression in Igf1r d/d and Igf1r f/f uterine lumenal epithelia on D4 17:00.The values are normalized to Gapdh.(I) Primary lumenal epithelial cells from the uteri were embedded in Matrigel to form epithelial organoids and immunostaining of p-STAT3 in epithelial organoids with or without IGF2 recombinant protein treatment.Scale bars: 50 μm.(J) Western blotting analysis of CTSB and LGMN proteins with or without IGF2 recombinant protein (200 ng/ml) in epithelial organoids.ACTB was used as a loading control.Le: lumenal epithelium.Error bars, mean ± SEM.The data were analyzed by Student's t-test, n ≥ 3, *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 7 .
Figure 7. Schematic roles of blastocyst derived signal facilitating lumenal epithelium transformation for implantation via upregulating lysosome proteostasis activity.In the presence of blastocyst, blastocyst-derived signal attended into the activation of epithelial p-STAT3, accompanied with upregulation of epithelial Lifr expression, to induce lysosomal hydrolase expression.The activated lysosomes in epithelia degraded CLDN1 and MUC1 proteins to facilitate the epithelial differentiation for the successful embryo implantation.