Abstract
Ineffective endometrial matrix remodeling, a key factor in infertility, impedes embryo implantation in the uterine wall. Our study reveals the cellular and molecular impact of human collagenase-1 administration in mouse uteri, demonstrating enhanced embryo implantation rates. Collagenase-1 promotes remodeling of the endometrial ECM, degrading collagen fibers and proteoglycans. This process releases matrix-bound bioactive factors (e.g., VEGF, decorin), facilitating vascular permeability and angiogenesis. Collagenase-1 elevates embryo implantation regulators, including NK cell infiltration and the key cytokine LIF. Remarkably, uterine tissue maintains structural integrity despite reduced endometrial collagen fiber tension. In-utero collagenase-1 application rescues implantation in heat stress and embryo transfer models, known for low implantation rates. Importantly, ex vivo exposure of human uterine tissue to collagenase-1 induces collagen de-tensioning and VEGF release, mirroring remodeling observed in mice. Our research highlights the potential of collagenases to induce and orchestrate cellular and molecular processes enhancing uterine receptivity for effective embryo implantation. This innovative approach underscores ECM remodeling mechanisms critical for embryo implantation.
Introduction
Embryo implantation is a complex process involving dynamic genetic, biophysical, and biochemical rearrangements in the uterine endometrium layer. A prerequisite for a successful implantation is the establishment of an appropriate endometrial microenvironment, influenced by many parameters such as hormone production, ECM remodeling, local immune-system adjustment, and angiogenesis (Quenby et al, 2002; Brosens et al, 2014; Macklon & Brosens, 2014). Proper ECM remodeling is particularly crucial for creating optimal receptive conditions and synchronization of functionally appropriate embryo-endometrium crosstalk. This process takes place during each reproductive cycle to prime the endometrium for implantation and continues during the window of implantation (WOI) and the decidualization phase (Fata et al, 2000).
During embryo implantation, the spatiotemporal expression of abundant ECM components in the endometrium is altered (Spiess et al, 2007; Diao et al, 2011; Oefner et al, 2015). In mice, collagen types I and III are locally reduced at the implantation sites due to extensive remodeling (Spiess et al, 2007; Diao et al, 2011). Similarly, the ultra-structure of fibrillar collagens in the human uterus changes during the first trimester of pregnancy. These observations suggest that fibrillar collagens play a role in trophoblast cell adhesion and invasion (Sinai Talaulikar et al, 2014), and impaired remodeling of endometrial extracellular proteins correlates with implantation failure and pregnancy loss. Endometrial remodeling during embryo implantation is mediated by various enzymes, particularly matrix metalloproteinases (MMPs) (Benkhalifa et al, 2018), which are widely expressed by both the endometrium and trophoblast cells (Chen et al, 2007; Anacker et al, 2011). MMPs regulate trophoblast invasion and contribute to endometrial vascular remodeling and inflammatory processes by proteolysis of ECM molecules and cell receptors (Fata et al, 2004; Chen et al, 2013; Nissinen & Kahari, 2014). Studies on MMPs during embryo implantation show several MMPs, including MMP-2, MMP-9, MMP-1, MMP-3, MMP-7, and MT1-MMP, are likely involved in this process, but their exact mechanisms are not fully understood (Chen et al, 2007; Anacker et al, 2011; Chen & Khalil, 2017; Benkhalifa et al, 2018; Latifi et al, 2018). MMP-2 and MMP-9 are notably expressed in invading human trophoblast cells, influencing their invasiveness (Isaka et al, 2003; Staun-Ram et al, 2004). However, excessive MMP-2 and -9 activities are linked to an unfavorable uterine environment, similar to endometrial inflammation in women with recurrent implantation failure (Yoshii et al, 2013). MT1-MMP (MMP-14) is crucial at the fetal-maternal interface, aiding MUC-1 release at the implantation site (Meseguer et al, 2001; Thathiah & Carson, 2004; Latifi et al, 2018). Collagenase-1 (MMP-1), secreted by trophoblast cells, may regulate trophoblast invasion (Vettraino et al, 1996, Huppertz et al, 1998). In addition, peri-implantation mouse embryos have been shown to express genes analogous to collagenase-1 (murine collagenase-like A [McolA/Mmp1a] and murine collagenase-like B [McolB/Mmp1b]) (Chen et al, 2007). However, beyond its expression, the specific role of this enzyme in embryo implantation remains unclear.
Exposing trophoblasts ex vivo to gelatinases (MMP-2/9) did not increase pregnancy rate but slightly improved pups’ viability, implying an unelucidated mechanism involving collagen reduction in the blastocysts (Zhang et al, 2020). However, the potential of MMP-mediated collagen/ECM remodeling mechanisms in promoting normal uterine tissue physiology, such as implantation, has been largely unexplored. Addressing the challenge of improving conditions for successful implantation through an optimized endometrial ECM receptive state may be facilitated by the precise regulation and understanding of endometrial remodeling. This approach preserves trophoblasts from ex vivo interventions.
Previously, we reported that collagenase-1 (MMP-1)-mediated proteolysis of collagen-rich ECM leads to unique changes in collagen morphology, ECM viscoelastic properties, molecular composition and cell adhesion capacity. These alterations in the ECM have profound impacts on neighboring cells, affecting their morphology, signaling patterns and gene-expression profiles. Notably, we observed enhanced adhesion and invasion of fibroblasts in collagenase-1-treated collagen-rich ECM in ex vivo settings (Solomonov et al, 2016). These findings motivated us to investigate the effects of mild ECM remodeling facilitated by collagenase-1 treatment on embryo implantation rates.
Here, we demonstrate that a single topical in-uterus administration of collagenase-1 significantly improves embryo implantation rates in mice, regardless of genetic background. Collagenase-1 treatment induces mild extracellular remodeling by reducing the tension (de-tensioning) of endometrial collagen fibers, releasing matrix-bound active factors, promoting neo-angiogenesis, and initiating a pro-inflammatory environment, all necessary for successful embryo implantation. Our study highlights the potential use of in-uterus collagenase-1-specific proteolysis for increasing embryo implantation rates and provides new mechanistic insights into endometrial receptivity.
Results
Endometrial collagen remodeling boosts embryo implantation
We previously reported that collagenase-1-induced proteolysis of natural collagen fascicles from rat tendons results in specific and unique ECM changes (Solomonov et al, 2016). Further examination of the collagen network by scanning electron microscopy (SEM), revealed a less densely packed, relaxed and disorganized arrangement of collagen fibrils within each fiber. These observations signify that collagen fibers undergo de-tension when treated with collagenase-1 (Fig 1A). In addition, we showed that fibroblasts exhibit enhanced adhesion to collagen-rich ECM that has been pretreated with collagenases compared with untreated collagen (Solomonov et al, 2016). Based on these results and recognizing the crucial role of endometrial ECM remodeling by MMPs for successful embryo implantation (Benkhalifa et al, 2018), we hypothesized that inducing moderate remodeling of collagen fibers through exogenous collagenase-1 treatment would enhance embryo-endometrium adherence and invasion, thereby improving the implantation process. To test this hypothesis, we incubated five mouse blastocysts ex vivo with native collagen-rich ECM pretreated with either vehicle or collagenase-1. After 4 h of incubation, more blastocysts interacted with the collagenase-1-treated collagen-rich ECM compared with the vehicle-treated matrix (Fig S1A). This observation that collagen/ECM proteolysis enhances blastocyst interactions ex vivo, prompted us to investigate the effect of in-uterus collagenase-1 treatment on embryo implantation in vivo. For this purpose, we used a spontaneous pregnancy model in which copulated ICR female mice were topically administered with a low dose of recombinant collagenase-1 (1.25 μg) or vehicle at day E2.5 (2.5 d post-coitum) (Fig 1B). The number of embryo implantation sites, recorded in the harvested uteri 4 d later at E6.5, showed a significant increase in ∼60% in collagenase-1-treated mice (Fig 1C). This effect was also observed in C57BL/6 mated females, demonstrating a significant improvement of ∼50% in embryo implantation (Fig S1B). Notably, the litter of treated animals was healthy and viable, exhibiting normal development and behavior. In addition, the pup’s weight measured 3 wk after birth was within the expected range of average weight reported for healthy mice of the same age and strain (12 ± 3 g) (Croy et al, 2013).
(A) Representative SEM images of native rat tail collagen fiber pretreated with either collagenase-1 or vehicle, demonstrating alterations in fibrils alignment within the fiber because of collagenase-1–treatment (scale 2 µm). (B) Schematic representation of the spontaneous pregnancy model. (C) Quantification of implantation sites and representative images of the uteri at E6.5. Implantation sites marked by arrowheads (n = 30). Data were analyzed by an unpaired, two-tailed t test. Results are presented as mean ± SEM with significance: ***P < 0.001.
(A) Images of blastocyst interactions with fascicle-derived native collagen-rich ECM pretreated with collagenase-1 or vehicle. Blastocysts that interacted with the collagen were counted (n = 4). (B) Number of implantation sites in the spontaneous pregnancy model using the C57BL/6 mouse strain (n ≥ 10). (C) Images of uteri administered with 1.5 μl of Fas Green using the NSET kit to study the distribution of collagenase-1. Arrows indicate locations where the dye is visible along the uterus. Data were analyzed by an unpaired, two-tailed t test. Results are presented as mean ± SEM with significance: ***P < 0.001.
Collagenase-1 treatment maintains uterine tissue integrity
To elucidate the molecular mechanisms underlying collagenase-1’s enhancement of embryo-ECM interactions, we examined the overall structure of the endometrial tissue at E4.5, a stage at which the embryo is fully attached and decidualization has commenced. No discernible differences were observed in the uterus tissue layers comparing collagenase-1–treated and control groups. The uteri myometrium and endometrium layers appeared intact in both groups with clearly demarcated borders (Fig 2A). Pathological examination further confirmed that collagenase-1 treatment did not induce any cellular abnormalities or distinct morphological changes in epithelia, stroma, and myometrial smooth muscle cells of the uterus. To exclude the possibility that collagenase-1 treatment induces pathological trophoblast invasion or excessive embryo invasiveness into the myometrium, we imaged the implantation site of GFP-expressing embryos within the uterine fibrillar collagens using two-photon microscopy with a second-harmonic generation (SHG) modality. This examination was performed at E6.5 when the invasion process was completed and gastrulation began. Representative images in Fig 2B depict GFP-expressing embryos interacting solely with endometrial collagens. Thus, using a low dose of collagenase-1 did not lead to abnormal invasion depth of the implanted embryos nor did it interfere with overall uterine tissue integrity.
(A) Schematic and representative H&E images of uteri at E4.5 (n = 3). The endometrium (E), myometrium (Myo), and perimetrium (Peri) layers of the uteri appear intact in both groups. The treatment did not affect the luminal epithelium (L) morphology, which is organized as a simple-columnar monolayer with basal nuclei. The glandular epithelium (GE) remained intact in the treated uteri and appeared similar to the vehicle-treated uteri. Enlarged insets are shown in the right panels. (B) Representative SHG images of transverse uteri sections of vehicle and collagenase-1–treated mice at E6.5 show that collagenase-1 treatment did not affect embryo invasiveness. Enlarged insets are shown on the right panels. green: venus embryo, red: collagen fibers, scale bar = 50 μm, n = 1.
De-tensioning of endometrial collagen fibers
Next, we profiled collagen organization before and during the WOI, that is, E2.5 and E4.5, respectively. First, we characterized the effect of collagenase-1 treatment on the organization of endometrial collagen fibers by applying SHG modality, a fast and convenient optical method that allows visualization of collagen fiber assembly within the tissue. At E2.5 (1 h after the treatment) and at E4.5, the treated uteri presented loosely packed and wavy fibers. In contrast, the vehicle-treated uteri maintained dense morphology with well-aligned and tightly packed collagen fibers (Fig 3A and B). To assess the degree of fiber misalignment, we applied an unbiased orientation analysis developed in-house and calculated the probability of collagen fiber orientation expressed in entropy terms (see the Materials and Methods section). This analysis revealed a slight, but significant increase in orientation entropy in treated compared with control uteri, both at 1 h after collagenase treatment (E2.5) and E4.5 (Fig 3C, top), demonstrating de-tensioning of collagen fibers. Nevertheless, analysis of the endometrial area covered by collagen showed a slight, but not significant reduction between the samples, confirming the absence of substantial fiber degradation (Fig 3C, bottom). Finally, these analyses revealed that collagenase-1 treatment did not affect the structural organization of the myometrial collagen fibers (Fig S2A and B). High-resolution SEM images of E2.5 uteri taken 1 h after collagenase treatment revealed distinct topological features of relaxed endometrium collagen fibers (Fig 3D). Their surface was covered by non-aligned wavy fibrils, indicating mild collagenolysis, in contrast to the well-aligned and densely packed fibrils in the control group.
(A, B) Representative SHG images of longitudinal sections vehicle- and collagenase-1–treated uterine samples at E2.5 (1 h after treatment) and E4.5, respectively. Only the endometrium layer is imaged, scale bar = 50 μM, (n ≥ 3). Enlarged insets are shown on the lower panels. (C) Measurements of the entropy of fibers orientation (top), and area covered by collagen (bottom). Analysis was performed using ImageJ software. (D) Representative SEM images of endometrial layers vehicle and collagenase-1 treated, 1 h after treatment. The images display well-aligned dense fibril packing on fiber surfaces of vehicle samples vs misaligned wavy fibrils on the fiber surfaces of treated samples, scale bar = 1 µm. Enlarged insets are shown on the right panels. (E) Representative 3D reconstructions from serial block-face SEM analysis of endometrium fibrils of vehicle- and collagenase-1-treated uteri. (F) Analyses of volume and surface area of fibrils from vehicle- and collagenase-1-treated uteri (n = 8 fibrils). Data were analyzed by an unpaired, two-tailed t test. Results are presented as mean ± SEM with significance: *P < 0.05, **P < 0.01.
(A) Representative SHG images of cross sections of vehicle- and collagenase-1–treated uterine samples at E4.5. Yellow dashed lines represent the myometrium layer (scale bar = 50 μM). (B) Measurements of the entropy of fiber orientation (upper), and area covered by collagen (lower). Analysis was performed using ImageJ software. Resulting orientation data were analyzed using Matlab code to assess fiber entropy (n = 2).
Next, we analyzed the three-dimensional microstructure of endometrium collagen fibrils using a high-resolution serial block-face SEM. This volume electron microscopy modality provides sufficient resolution to study individual collagen fibrils by generating a representative reconstructed 3D model (Starborg et al, 2013). Analysis of randomly selected reconstructed collagen fibrils from endometrium samples taken 1 h after treatment did not show significant changes in fibril volume fraction when compared with vehicle samples (Fig 3E and F), further indicating that collagenase-1 did not induce gross fibril degradation. Taken together, structural analyses at different resolutions provide evidence that mild proteolysis by collagenase-1 treatment mostly manifests in de-tensioning of collagen fibers affecting their spatial organization, rather than intensive fibril degradation. Such collagen proteolysis supports trophoblast invasion and angiogenesis in vivo (Seandel et al, 2001; Sinai Talaulikar et al, 2014).
COL1A1 and DCN: collagenase-1 substrates in pregnant uteri
To gain a deeper understanding of the molecular effects of collagenase-1 treatment, we investigated the substrates released by collagenase-1 from utrus of pregnant mice. E2.5 uteri were subjected to a 4-h incubation with collagenase-1, and the resulting supernatants were subjected to LC–MS/MS proteomics analysis (Fig 4A). The analysis revealed 41 substrates that pass the threshold of log2(collagenases-1/vehicle) > 0.4 and q < 0.1 (Fig 4B). Among these substrates, three proteins were identified as matrisome proteins: collagen alpha-1 (COL1A1), decorin (DCN), and annexin A6 (ANXA6) (Fig 4C). To enrich for collagenase-1 matrisome substrates, we replicated the experiment using de-cellularized E2.5 uteri, which were incubated with collagenase-1 for 24 h (Fig 4D). In this de-cellularized set-up, we identified 38 substrates of collagenase-1 that pass the same threshold of log2(collagenases-1/vehicle) > 0.4 and q < 0.1 (Fig 4E). The most significantly up-regulated substrates (log2[collagenases-1/vehicle] > 2.5) included ECM structural proteins and proteoglycans such as dermatopontin, lumican, decorin, and collagen alpha-1,2 (Fig 4F). We then conducted a more in-depth analysis of the protein decorin, which was identified as a collagenase-1 substrate in both experimental set-ups in pregnant uteri. Our analysis revealed that of 32 peptides that were cleaved, nine were semi-/non-tryptic peptides, and among them, three exhibited favorable cleavage motifs of collagenase-1, specifically leucine and isoleucine, as indicated by MEROPS (https://www.ebi.ac.uk/merops/cgi-bin/pepsum?id=M10.001) (Fig 4G). Based on these findings, it is likely that collagenase-1 breaks down decorin proteoglycan in the matrix of pregnant uterus. Decorin is a crucial stabilizer of the collagen fibrillar network, and its proteolysis can contribute to fiber de-tensioning by inducing collagen misalignment. In addition, decorin binds a variety of signaling molecules, and its degradation can modulate the bio-activity of many important factors, including VEGF and others (Jarvelainen et al, 2015). To investigate it further, we analyzed the supernatant from the de-cellularized uteri in a Western blot using VEGF-A antibody revealing that VEGF-A was released after incubation with collagenase-1 (Figs 4H and S4A).
(A) Schematic representation of the experimental set-up: E2.5 uteri were incubated ex vivo with collagenase-1 for 4 h. The supernatant was subjected to mass spectrometry to reveal potential substrates of collagenase-1 within a pregnant uterus. (B) Volcano plot of proteins identified in the supernatant of E2.5 uteri incubated ex vivo with collagenase-1. Red and blue dots are proteins that were up-regulated or down-regulated (respectively) in the supernatant after incubation with collagenase-1 and pass the threshold of log2(collagenases-1/vehicle)≥0.4 or ≤ −0.4; q ≤ 0.1. (C) Bar graph of the matrisome proteins among the up-regulated substrates that passed the threshold: collagen alpha-1 (COL1A1, q = 0.057, P = 0.0041), decorin (DCN, q = 0.055, P = 0.0038), and annexin A6 (ANXA6 q = 0.062, P = 0.0057) **P < 0.01. (D) Schematic representation of the experimental set-up: E2.5 uteri were de-cellularized and incubated ex vivo with collagenase-1 for 24 h. The supernatant was subjected to mass spectrometry to reveal potential substrates of collagenase-1 within the matrix of a pregnant uterus. (E) Volcano plot of proteins identified in the supernatant of de-cellularized E2.5 uteri incubated ex vivo with collagenase-1. Red and blue dots are proteins that were up-regulated or down-regulated (respectively) in the supernatant after incubation with collagenase-1and pass the threshold of log2(collagenases-1/vehicle)≥0.4 or ≤ −0.4; q ≤ 0.1. (F) Bar chart showing the most abundant ECM structural proteins and proteoglycan substrates (log2[collagenases-1/vehicle] > 2.5) found in (E) ***P < 0.001. (G) Schematic summarizing of semi-tryptic decorin peptides detected in the supernatant of E2.5 de-cellularized uteri after incubation with collagenase-1, at respective amino acid positions. (H) Representative Western blot analysis shows protein expression of VEGF-A in supernatants from de-cellularized uterus samples that were incubated with collagenase-1 or vehicle for 24 h.
Mass spectrometry results of the ex vivo experiment in which E2.5 uteri were subjected to a 4 h incubation with collagenase-1 or vehicle, also revealed the release of several proteins associated with epithelial integrity after collagenase-1 administration (Fig S3A). These proteins included the adherence junction proteins catenin alpha-3 (CTNNA3) and catenin delta-1 (CTNND1), desmosomes like desmoplakin (DSP) and plakoglobin (JUP), and the basement membrane protein laminin subunit alpha-5 (LAMA5). The release of these proteins indicates that collagenase-1 can slightly disrupt the intact luminal epithelial barrier, presumably allowing the diffusion of the enzyme into the stroma. In addition, careful analysis of the mass spectrometry results also revealed the release of two integrins: integrin α6 and integrin β1 (Fig S3B). Of note, we did not detect the release of other integrins, such as integrin αv and integrin β3, which are reported in the literature as mediators of embryo attachment in the mouse (Cai et al, 2000; Illera et al, 2000). Staining E4.5 uteri for integrin αv and integrin β3 showed no differences between the vehicle-treated and the collagenase-1-treated uteri (Fig S3C). Given that the in vivo administration involves a single dose of 2 d before implantation, it is reasonable to assume that the release of the observed integrins altered the integrin landscape without significantly impairing embryo attachment. These experiments indicate that collagenase-1 not only remodels collagens but also plays a role in breaking down proteoglycans, like decorin, thereby releasing factors from the matrix and making them bioavailable in the tissue.
E2.5 uteri were incubated ex vivo with collagenase-1 for 4 h. The supernatant was subjected to mass spectrometry to reveal potential substrates of collagenase-1 within a pregnant uterus. (A) Bar graph of the proteins associated with epithelial integrity. Catenin alpha-3 (CTNNA1, q = 0.079, P = 0.009), catenin delta-1 (CTNND1, q = 0.22, P = 0.07), desmoplakin (DSP q = 0.15, P = 0.03), plakoglobin (JUP, q = 0.16, P = 0.04), and laminin subunit alpha-5 (LAMA5, q = 0.18, P = 0.05). (n = 6). (B) Bar graph of integrins: integrin beta-1 (ITGB1, q = 0.15, P = 0.04) and integrin alpha-6 (ITGA6, q = 0.17, P = 0.04). (n = 6). (C) IHC staining of ITGB3 (integrin β3) and ITGAV (integrin αv) in E4.5 uteri treated with either vehicle or collagenase-1 show no differences between the groups. **P < 0.01, *P < 0.05 (n = 6).
Vascular permeability and angiogenesis-related processes
Embryo implantation is closely associated with angiogenesis and enhanced uterine vascular permeability, resulting from vessel leakage and the development of new blood vessels (Plaks et al, 2006; Fournier et al, 2021). Building upon our discovery that VEGF-A is released from a pregnant uterus-ECM after ex vivo exposure to collagenase-1, we delved further to investigate the in vivo state of angiogenesis in the tissue after collagenase-1 treatment. We examined various angiogenic makers at two distinct time points after the treatment, E4.5 and E6.5. Remarkably, we observed an increase in the expression of VEGF-R2 at E6.5 and VEGF-A at E4.5 (Figs 5A and S4B and E). Angiogenesis is endogenously regulated, among other mechanisms, by the balance between the pro-angiogenic VEGF and the anti-angiogenic PEDF factors. A higher VEGF/PEDF ratio promotes angiogenesis and embryo implantation, whereas increased PEDF levels impair it (Sun et al, 2012; Chuderland et al, 2014; Wu et al, 2021). In line, we found that collagenase-1 treatment reduced PEDF levels, concurrent with an increase in VEGF-A (Figs 5A and S4C–E). Furthermore, to investigate the direct effect of collagenase-1 on PEDF, we conducted an in vitro incubation of PEDF protein with collagenase-1 for 24 h, which resulted in a decrease in PEDF levels, indicating its degradation by the enzyme (Fig 5B). Functional assessment to validate the outcome of elevated angiogenic factors was conducted, by staining with CD34, a neovascularization marker (Koo et al, 2021). Imaging of the implantation site endometrium revealed a higher number of newly formed blood vessels at E6.5 after the WOI in collagenase-1–treated mice (Fig 5C).
(A) Western blot analysis shows the increase of VEGF-R2 and VEGF-A levels in E4.5-E6.5 samples treated with collagenase-1 as well as a decrease in PEDF levels (left). Quantification of WB using ImageJ analysis tool (right) (n = 3). (B) Representative WB images of incubation of PEDF protein with collagenase-1 in vitro for 24 h. Initial PEDF levels at T0 are similar. However, after a 24-h incubation with collagenase-1, there is a noticeable reduction in PEDF, pointing to its degradation. (n = 5). (C) Representative images of immunofluorescent staining against CD34 of E6.5 samples (scale 200 µm). The percentage area of CD34 was measured in the endometrium layer, excluding the myometrium edges (the quantified area is marked by a yellow dashed line). Covered area quantification was performed using the ImageJ analysis tool. Avarage per animal (n ≥ 5). Avarage per implantation site (n ≥ 6). Data were analyzed by an unpaired, two-tailed t test. Results are presented as mean ± SEM, with significance: *P < 0.05.
Further analysis of angiogenesis at the implantation site after collagenase-1 treatment was performed using magnetic resonance imaging (MRI) at E4.5 to document the dispersion of i.v. injected contrast agents (Fig S5A). We calculated two key parameters that characterize vascular remodeling and function: permeability surface area (PS) and fractional blood volume (fBV) (Plaks et al, 2006). These analyses showed that implantation sites in utero treated with collagenase-1 exhibited slightly elevated values of PS and fBV, although non-significantly when compared with vehicle-treated uteri (Fig S5B). These results indicate that collagenase-1 induces a modest increase in uterine vascular permeability without evidence of pathological hypervascularity. Remarkably, immunofluorescent imaging demonstrated that collagenase-1 treatment enhanced penetration and accumulation of the injected contrast agents within the implantation site (Fig S5C).
(A) Representative MRI images at E4.5 of vehicle- and collagenase-1-treated uteri. Arrows indicate implantation sites. (B) Quantification of the permeability surface (PS) and fractional blood volume (fbv) at implantation sites, taken from the MRI data (n = 7; 9.4T, 3D-GE; IV biotin-BSA-GdDTPA). (C) Immunofluorescence imaging of the contrast agents: biotin-BSA-GdDTPA (green) and BSA-Rox (red), demonstrating their accumulation in the implantation sites during MRI imaging for 30 and 2.5 min, respectively. IS, implantation site; P, perimeter.
Taken together, these results indicate that mild proteolysis of the endometrium promotes local vascular remodeling associated with permeability and angiogenesis, without inducing early pathological hyper-vascularization that could negatively impact pregnancy outcomes.
NK cell infiltration and up-regulation of the cytokine LIF
Intensive collagen degradation has been associated with immune cell migration and infiltration (Goda et al, 2006; Kuczek et al, 2019). In addition, embryo implantation involves unique immune activities and pro-inflammatory modulation (Zhao et al, 2021). Therefore, we examined whether collagen de-tensioning in the uterus affects the infiltration of immune cells into the treated endometrium. Flow cytometry analysis of implantation sites at E4.5 showed a specific increase in ∼40% in NK cells with no changes in T cells, dendritic cells, or macrophages (Figs 6A and B and S6A and B). Furthermore, Western blot analysis of the NK cell marker (NKp46) at E4.5 confirmed the increase in NK cell infiltration to the implantation site after treatment (Figs 6C and S6C). These changes induced by collagenase-1 are crucial for improved implantation rates since uterine NK cells are pivotal for successful early pregnancy (Kanter et al, 2021), and their absence characterizes recurrent implantation failure (Lai et al, 2022).
(A) A representative flow cytometry image showing the gating strategy used to identify NK cells (CD45+CD3ε−NK1.1+NKp46+). (B) Quantification of flow cytometry results showing the percent of NK cells of live cells at E4.5 (n ≥ 12). (C) Representative Western blot (left) and quantification analysis (right) showing expression of LIF and NKp46 in E4.5 implantation sites. Western blot quantification performed using ImageJ analysis tool (n = 3). NKs, natural killer cells. (D) Real-time PCR analysis of cytokines’ transcripts at E4.5 (n = 8). Data were analyzed by an unpaired, two-tailed t test. Results are presented as mean ± SEM with significance: *P < 0.05.
(A) Representative flow cytometry images showing the gating strategy used to identify macrophages (CD45+CD11b+F4/80+CD64+) and dendritic cells (CD45+CD11c+F4/80−MHC-II+). Macs, macrophages (upper panel). DCs, dendritic cells (lower panel). (B) Quantification of flow cytometry results showing the percent of macrophages, dendritic cells, and T cells among live cells at E4.5 (n ≥ 12). Data were analyzed by an unpaired, two-tailed t test. Results are presented as mean ± SEM. (C) Western blot full membranes of indicated proteins presented in Fig 6C. Arrow indicates the specific bend for LIF.
Because numerous cytokines and chemokines are involved in the embryo-endometrium crosstalk (van Mourik et al, 2009; Dekel et al, 2014), we compared the expression levels of several such molecules at E4.5. We found that collagenase-1 treatment significantly increased the transcript and protein levels of leukemia inhibitor factor (Lif) (Figs 6C and D and S6C), one of the most important pro-inflammatory cytokines involved in successful implantation (Kimber, 2005). Interestingly, there was no significant change in the expression of other tested factors (Fig 6D), indicating that collagenase-1 treatment did not induce a robust and unbalanced immune response. Of note, establishing a pseudopregnant model (Fig S7A), we documented the same tissue response to collagenase-1 remodeling, indicated by increased NK cell, LIF, and VEGF-A proteins (Fig S7B–D), demonstrating an embryo-independent process. Taken together, our results indicate that collagenase-1 treatment promotes a mild yet specific pro-inflammatory signaling in the endometrium, which is recognized as a prerequisite for successful embryo implantation.
(A) Schematic representation of the pseudo-pregnancy experimental model. (B) Western blot analysis of E4.5 samples treated with either collagenase-1 or vehicle. (C) Quantification of Western blots using ImageJ analysis tool (n = 3). (D) Full membranes of indicated proteins presented in Fig 6. Arrow indicates the protein’s specific bend. Data were analyzed by an unpaired, two-tailed t test. Results are presented as mean ± SEM with significance: *P < 0.05, **P < 0.01.
Improved implantation in heat stress and embryo transfer
Next, we investigated whether collagenase-1 treatment could also be effective in models known for exhibiting low embryo implantation rates, that is, heat stress and embryo transfer (ET). Prolonged exposure to elevated environmental temperature reduces successful pregnancy rates (Simon & Laufer, 2012; Wani et al, 2021; Lian et al, 2022). In line, with our experimental model, keeping mice under heat stress conditions of 38°C, we observed reduced embryo implantation rates with an average of only five implantation sites per female (Fig 7A and B). Notably, these females were in a normal and healthy physiological state (Fig S8A). We hypothesized that the in utero structural and signaling changes mediated by collagenase-1 treatment could compensate for the harmful effect of the elevated temperature. Indeed, treatment with collagenase-1 rescued the heat stress effect as we observed an average of 10.6 implantation sites per female (Fig 7B).
(A) A schematic representation of the heat stress model. (B) Quantification of implantation sites and representative images of the uteri at E6.5 in the heat stress model. Arrowheads mark implantation sites (n = 9). (C) A schematic representation of the embryo transfer model. (D) Quantification of implantation sites and representative images of the uteri at E6.5 in the embryo transfer model. Implantation sites marked by arrowheads (n ≥ 15). (E) Representative SHG images of human uterine tissue sections incubated with either 50 nM of collagenase-1 or vehicle for 24 h (scale bar = 50 μm). Enlarged insets presented at the lower panel show misaligned and loosely packed collagen fibers, in the collagenase-1-treated sample. (F) Representative Western blot of supernatants from de-cellularized human uterus samples that were incubated with collagenase-1 or vehicle for 24 h (n = 3). Three samples from three different women were compared, and for each sample, the fold change between collagenase-1 and vehicle was measured. The graph shows the three ratios (P-value = 0.09). Quantification of Western blot analysis using ImageJ analysis tool. Data were analyzed by an unpaired, two-tailed t test. Results are presented as mean ± SEM with significance: *P < 0.05.
(A) Monitoring of mice body temperature during the heat stress protocol, ranged between 36°C–38°C throughout the experiment, indicating that females were in a normal and healthy physiological state. (B) Number of implantation sites in the cross-strain embryo transfer model in which cbcF1 embryos were transferred to pseudo-pregnant ICR females with or without collagenase-1 treatment (n ≥ 6). Data were analyzed by an unpaired, two-tailed t test. Results are presented as mean ± SEM with significance: **P < 0.01.
Another well-known scenario with typical low implantation rates is the embryo transfer model. In humans, in vitro fertilization (IVF) and ET show relatively low pregnancy success mainly because of impaired implantation (Cozzolino et al, 2018, Ozer et al, 2023). Thus, we investigated whether collagenase-1 treatment can improve implantation rates in a mouse embryo transfer model, in which blastocysts are transferred to a pseudo-pregnant female, which is mated with a vasectomized male (Fig 7C). A significant twofold increase in the number of embryo implantations was achieved upon topical administration of recombinant collagenase-1 (Fig 7D). Furthermore, a similar increase in implantation rates induced by the treatment was observed in a cross-strain embryo transfer protocol, in which cbcF1 embryos were transferred into ICR pseudo-pregnant females (Fig S8B).
Altogether, these results show a general phenomenon in which collagenase-1-mediated mild remodeling of the endometrium increases the number of implanted embryos regardless of their genetic background. Moreover, this mild and controlled intervention overrides environmental stressors that impair the tightly coordinated embryo implantation process.
Collagenase-1 similarly affects human endometrial tissue
Finally, a major intriguing question was the feasibility of our approach to create similar effects in assisted reproduction treatment protocols used in human females. To obtain molecular mechanistic indications for the response of the human endometrium to collagenase-1 treatment, we de-cellularized fresh endometrial biopsies from healthy women. Tissue samples were incubated ex vivo with collagenase-1 or vehicle and imaged under two-photon microscopy. Using the SHG imaging modality, we observed collagen fiber de-tensioning in collagenase-1-treated endometrium layers, similar to what was observed in mice (Fig 7E). Furthermore, we examined the effect of collagenase-1 on the release of matrix-bound VEGF-A. Consistent with the results observed in murine samples, we observed an increased release of ∼50 kD dimers of VEGF-A, specifically from the collagenase-1–treated matrices (Fig 7F).
Fig 8 provides a schematic illustration that summarizes the key factors and molecular mechanisms involved in the processes by which collagenase-1 treatment improves embryo implantation. Overall, our results demonstrate that mild proteolysis by collagenase-1 generates similar changes in human and mouse tissues in terms of altering the collagen morphology and the release of pro-angiogenic factors. These data indicate the potential use of a specific remodeling enzyme in human reproductive medicine.
Collagenase-1 administration leads to the de-tensioning of the dense collagen network through mild degradation of collagen and decorin. This process leads to a greater accumulation of uterine NK cells secreting LIF. In addition, collagenase-1 releases VEGF-A from the uterine matrix via decorin proteolysis and up-regulates VEGF-R2 expression. These enhancements in inflammatory and angiogenic pathways optimize the receptive environment for embryo implantation.
Discussion
Embryo implantation is a pivotal event in mammalian reproduction, representing a critical determinant for the establishment of natural early pregnancies and the success of assisted reproductive techniques such as IVF-embryo transfer. Given the challenges and limited success rates associated with assisted reproductive technology, extensive research efforts have been dedicated to exploring strategies involving ECM remodeling to improve embryo implantation. Among these strategies, endometrial scratching has emerged as one of the most widely recognized approaches (Barash et al, 2003). Application of this method before the IVF procedure dramatically increases the chances of implantation by provoking inflammation and endometrial angiogenesis, processes recommended by many clinicians before IVF (Gnainsky et al, 2010; Siristatidis et al, 2014; Yang et al, 2019). However, the efficiency of this procedure has been questioned by several studies, probably because of reproducibility problems (Santamaria et al, 2016; Lensen et al, 2019; van Hoogenhuijze et al, 2019). LIF administration has emerged as an alternative approach to enhance embryo implantation because this factor is dysregulated in infertile women experiencing recurrent implantation failure (Hambartsoumian, 1998). Unfortunately, LIF treatment did not improve implantation rates and pregnancy outcomes for women with recurrent unexplained implantation failure (Brinsden et al, 2009). Another approach focused on enhancing vascularization. VEGF levels are significantly reduced in uterine fluid during the receptive phase in women with unexplained infertility compared with fertile women (Hannan et al, 2011). Indeed, mouse embryos cultured with recombinant human VEGF had significantly higher implantation rates after blastocyst transfer (Binder et al, 2014). However, the net effect of recombinant human VEGF on the uterus environment was not described. Taken together, these studies strongly indicate that factors associated with ECM remodeling have the potential to improve embryo implantation. However, a deeper understanding of ECM remodeling mechanisms and their role in embryo implantation is required to better harness this strategy for medical reproductive interventions.
Our findings unequivocally demonstrate that the judicious application of collagenase-1 induces a mild, yet highly targeted ECM remodeling process, which orchestrates crucial physiological changes within the uterine microenvironment to foster optimal conditions for successful embryo implantation. In particular, collagenase-1 actively participates in the intrinsic and dynamic remodeling of vital ECM components essential for successful implantation, namely fibrillar collagens and decorin. Notably, the degradation of these pivotal proteins mirrors the natural course of events during peri-implantation and the early stages of implantation (Hjelm et al, 2002; San Martin et al, 2003; San Martin et al, 2004), indicating that collagenase-1 administration must be tightly synchronized before embryo attachment to the uterus wall. The release of proteins associated with epithelial integrity by collagenase-1 also indicates a slight disruption of the intact luminal epithelial barrier, likely allowing the enzyme to diffuse into the stroma. Physiological de-tensioning of endometrial collagen fibers associates with trophoblast proliferation and invasion (Sinai Talaulikar et al, 2014) as well as angiogenesis (Seandel et al, 2001). We found that facilitating this process exerts a slight increase in the permeability surface area of blood vessels and fBV. Enhanced angiogenesis is evident through the up-regulation of VEGF-A, VEGF-R2, and CD34 expression and reduced levels of the angiogenic suppressor PEDF. Collagenase-1 has been previously shown to promote VEGF-R2 expression through the stimulation of protease-activated receptor-1 (Mazor et al, 2013). Here, we show that collagenase-1 enhances angiogenesis also via the release of matrix-bound VEGF-A from the pregnant uterus. The release of matrix-bound VEGF-A primarily occurs through the degradation of proteoglycans, with our proteomics results pointing to decorin degradation as its main source. Annexin A6 was also identified as one of the matrisome proteins released from the uterus after ex vivo incubation with collagenase-1. Annexins are cytoplasmic proteins that attach to phospholipid membranes and are highly conserved, Ca2+-dependent membrane-binding proteins involved in various physiological and pathological processes (Hu et al, 2024). They are also considered ECM-affiliated proteins and are part of the matrisome (Hynes & Naba, 2012). In one study, ANXA6 was found to be up-regulated during decidualization in women experiencing recurrent implantation failure (Dhaenens et al, 2019). In another study, ANXA6 was down-regulated in non-receptive endometrium compared with receptive endometrium (Garrido-Gomez et al, 2014). These findings indicate that further research is needed to explore the potential link between collagenase-1 and annexin A6 in embryo implantation.
Remarkably, collagen de-tensioning creates space for increased infiltration of NK cells. NK cells play a significant role in trophoblast invasion, vascularization, and the initiation of decidualization during embryo implantation (Hanna et al, 2006). Interestingly, among tested embryo implantation-related cytokines, only LIF showed elevated gene expression and protein levels. The increase in both, LIF and NK cell infiltration at E4.5 indicates that these cells could be a source of LIF. LIF affects trophoblast invasion, adhesion, and trophoblast cell proliferation and may stimulate their adhesion to the ECM via different signaling pathways (Poehlmann et al, 2005; Tapia et al, 2008; Prakash et al, 2011).
Using a pseudo-pregnancy model at E4.5, we observed that collagenase-1 induces pro-inflammatory and angiogenesis pathways independently of the presence of embryos (Fig S7). These data further confirm that collagenase-1 primes the uterine environment preparing the embryo-endometrium interface for successful implantation. Clinically significant, collagenase-1 administration was able to rescue low implantation rates in the heat stress and embryo transfer models, without compromising endometrial integrity or normal offspring development. Crucially, a feasibility test of collagenase-1 application on human endometrial samples ex vivo showed induction of spatial re-organization of collagen fibers and the release of VEGF-A, mechanisms similar to those characterized in mice. A schematic model illustrating the mechanisms by which enforced collagenase-1 remodeling facilitates embryo implantation is presented in Fig 8. In summary, our study reveals that the administration of collagenase-1 directly into the uterine environment represents a powerful strategy to augment endometrial receptivity, consequently enhancing the chances of successful embryo implantation. The key mechanisms underlying this enhancement lie in collagen de-tensioning, a process triggered by collagenase-1, and the concurrent induction of angiogenesis within the endometrial tissue. Accordingly, our study not only sheds light on the fundamental biology of embryo implantation but also offers compelling implications for the field of biomedical engineering.
The application of collagenase-1 as a localized treatment holds immense promise across diverse domains and in the realm of livestock breeding, where efficient reproductive outcomes are crucial for agricultural productivity and environmental sustainability. Moreover, the implications of our study extend to the realm of assisted human reproductive protocols, encompassing both natural conception and IVF-based approaches. For future applications in humans, it is preferable to use collagenase-1 purified from mammalian cells, as it shares similar properties with the enzyme purified from bacteria. Another option is to use bacterial collagenases for embryo implantation; however, further characterization and study are needed. Collagenase-1 emerges as a biomedical tool, capable of fine-tuning the uterine microenvironment to create a receptive milieu for embryonic attachment. This innovation has the potential to revolutionize the success rates of IVF procedures. This convergence of biological insights and biomedical engineering applications represents a significant step towards addressing pressing challenges in reproduction, with far-reaching implications for both agriculture and healthcare.
Materials and Methods
Collagenase-1 preparation, activation, and enzymatic assays
Collagenase-1 was overexpressed and purified as described previously (Solomonov et al, 2016). Before each experiment, collagenase-1 was activated with 1-mM 4-aminophenylmercuric acetate (APMA) in TNC buffer (50 mM Tris [pH 7.5], 150 mM NaCl, 10 mM CaCl2) at 37°C for 60 min. The solution was then centrifuged (5 min, 400g) to allow the APMA, a heavy salt, to settle at the bottom of the tube. We then collected most of the liquid, intentionally leaving a few microliters at the bottom to prevent the APMA from being collected with the activated enzyme. The enzymatic activity of collagenase-1 was measured at 37°C by monitoring the hydrolysis of the fluorogenic peptide Mca-Pro-Leu-Gly-LeuDpa-Ala-Arg-NH2 at λex = 340 nm and λem = 390 nm as previously described (Knight et al, 1992).
Seeding blastocysts on fascicle-derived ECM
Fascicle-derived ECM was prepared from the tails of adult Norwegian rats (6 mo). Rat tails were dissected, and tendon fascicles (diameter ∼0.6 mm, length ∼7 mm) were gently extracted and were washed extensively in TNC buffer (50 mM Tris [pH 7.5], 150 mM NaCl, 10 mM CaCl2) to remove the macroscopic tissue debris and proteases remains. ECM samples were incubated with 500 nM collagenase-1 or vehicle (TNC) at 30°C for 24 h (Solomonov et al, 2016). After incubation, the tails were extensively washed with PBS. Five mouse blastocysts were seeded on top of the ECMs placed in 24-well cell culture plate for 4 h in 37°C and then imaged by EVOS M5000.
SEM
Fascicle-derived native collagen-rich ECM or E2.5 de-cellularized mouse uterus tissues (longitudinal cut) were fixed in a 0.1-M cacodylate buffer solution (pH 7.4) containing 2.5% PFA and 2.5% glutaraldehyde (pH 7.2) for 60 min at room temperature and then were washed three times in the same buffer. The samples then were stained with 4% (wt/vol) sodium silicotungstate (Agar Scientific) (pH 7.0) for 45 min followed by dehydration through an ascending series of ethanol concentrations up to 100% ethanol. Next, the samples were dried in a critical point dryer and attached to a carbon sticker. Finally, the samples were coated with a thin gold/palladium (Au/Pd) layer. The samples were observed under a Zeiss FEG Ultra55 SEM operating at 2 kV.
Mice
All animals were obtained from Envigo Laboratories. On arrival, male mice were housed individually, and female mice were housed 3–5 per cage in animal rooms maintained at 20–22°C with an average relative humidity of 35% under a 12:12 h light-dark cycle and were housed in standardized ventilated microisolation caging. All experiments and procedures were approved by the Weizmann Institute of Science Animal Care and Use Committee (IACUC; approval no. 27170516-2).
Topical administration of collagenase-1 or vehicle in a spontaneous pregnancy protocol
Female ICR or C57BL/6J mice (8–10 wk) were copulated with fertile ICR or C57B/6J male mice (8–12 wk), respectively. Females who presented a vaginal plug on E0.5 were administered with 1.5 μl of 15.5 μM activated recombinant collagenase-1 (total amount 1.25 μg) or 1.5 μl of the vehicle as control (TNC buffer) at E2.5 using the NSET Device of ParaTechs: mice were placed on a wire-top cage and allowed to grip the bars. The small and large specula were placed sequentially into the vagina to open and expose the cervix. Then, the NSET catheter was inserted through the large speculum, past the cervical opening, and into the uterine horn allowing the topical administration of recombinant collagenase-1 or vehicle. Uteri were excised 1 h, 2 or 4 d after treatment (E2.5, E4.5, and E6.5 respectively). On E4.5 and E6.5 the number of implantation sites was counted. On E6.5 implantation sites were visible and on E4.5 Evens-blue dye was injected intravenously 10 min before euthanizing which allows the visualization of the implantation sites. For generating fluorescent venus embryos, C57BL/6J female mice were mated with Myr-Venus homozygote males (Rhee et al, 2006).
Administration of collagenase-1 in the heat stress protocol
After treatment at E2.5, mice were transferred to preheated housing cages at 38°C for 4 d and were euthanized at E6.5. Selected females were injected with a thermal microchip in the back of the mice a week before the experiment. The microchip injection was performed under an isoflurane sedation coupled with carprofen (5 mg/kg) as an analgesic.
Administration of Fas Green
To study the distribution of collagenase-1 in the uterus, 1.5 μl of Fas Green dye was administered to E2.5 pseudo-pregnant females using the NSET kit. Mice were euthanized 5 min later, and their uteri were examined. Our findings show that the dye can disperse along either one or both uterine horns (Fig S1C). This suggests that both scenarios are possible and may be influenced by statistical factors and environmental conditions, such as handling and laminar flow within the uterus.
Administration of collagenase-1 or vehicle in a non-surgical embryo transfer protocol
ICR mice (8–10 wk) were mated with vasectomized males to achieve a pseudo-pregnant state. Females who presented a vaginal plug on E0.5 were administered at E2.5 with 1.25 μg activated recombinant collagenase-1 or vehicle as control using the NSET device as described above. After 15 min, embryos at the blastocyst stage were transferred also through the NSET catheter into the uterus (10 embryos per mouse). Then, the device and specula were removed, and the mouse was returned to its home cage. The number of implanted embryos was counted and recorded on E6.5.
H&E staining of frozen sections
E4.5 uteri samples treated with collagenase-1 or vehicle embedded in OCT were cross sectioned (12 μm) on glass microscope slides. Slides were washed with PBS before staining. Then, slides were incubated in Hematoxylin for 3 min and washed with tap water. After, slides were dipped in 95% ethanol and incubated in Eosin for 45 s. Next, samples were dipped in 95% ethanol and incubated 2 min in 100% ethanol. Finally, samples were incubated in xylene for 2 min and were mounted in a mounting medium.
Two-photon microscopy and SHG
Snap-frozen murine uterine samples (excised at E2.5 and E4.5) were thawed in PBS, cut longitudinally, placed on a slide as the endometrium facing up, and covered with cover slip. Then samples were imaged using a two-photon microscope in an SHG mode (2PM: Zeiss LSM 510 META NLO; equipped with a broadband Mai Tai-HP-femtosecond single box tunable Ti-sapphire oscillator, with automated broadband wavelength tuning 700–1020 nm from Spectra-Physics, for two-photon excitation). For second-harmonics imaging of collagen, a wavelength of 800–820 nm was used (detection at 390–450 nm). For imaging the myometrium layer and venus embryo uteri samples embedded in OCT were cross sectioned (50 μm) on glass microscope slides.
Entropy analysis of the distribution of fiber orientations
Detection of collagen fibrils of vehicle- and collagenase-1–treated endometrium samples was performed using ImageJ (Schindelin et al, 2012). A mask highlighting collagen fibers was generated using the Tubeness plugin and was followed by orientation analysis using the Directionality plugin (Liu, 1991). This gave rise to plots and spreadsheets of distribution of fiber orientations throughout the stacked images (code will be provided upon request). Next, the orientation data were analyzed using Matlab. The degree of orientation diversity, that is, distribution dispersion was measured by calculating the edge-orientation entropy that refers to the probability of encountering particular orientations in an image (Grebenkina et al, 2018), and according to the Shannon index (Shannon, 1948). For every orientation i, let pi be the probability of the occurrence of orientation i in the image. Entropy or Shannon index (H’) of the probability distribution is defined as
Fiber density analysis
Fiber density analysis was performed by ImageJ software: For endometrium analysis a z-projection of max intensity was applied for 13 μm of the endometrium layer and only fibrillar areas were chosen. Next, auto threshold on the default setting was applied, and analyze particles plugin was used to detect the collagen-covered area. Size of particles was chosen as 0.172-infinity. On E2.5, we analyzed three vehicle-treated females (4–12 images per each) and five collageanse-1–treated females (3–10 images per each). On E4.5, we analyzed three vehicle-treated females (4–16 images per each) and three collageanse-1–treated females (2–15 images per each). For myometrium analysis, a single section was taken and the myometrium area was chosen to auto threshold and analyze particles.
De-cellularization of uterine tissue
Mice (E2.5) or woman biopsy uteri samples treated with collagenase-1 or vehicle were incubated in a de-cell solution containing 3% Triton-100 (6 h, 4°C) and then in a de-cell solution with 0.4% Triton-100 overnight at 4°C (de-cell solution: 1.5 M NaCl, 50 mM Tris pH 8, 50 mM EDTA, protease inhibitor cocktail [Roche]). Samples were washed three times in ddH2O and then incubated with 0.5% sodium deoxycholate (60 min, 25°C) to remove lipid remaining. Samples were washed again three times in ddH2O and stored at 4°C until use.
Serial block-face SEM
Samples were prepared based on methods described previously (Starborg et al, 2013). In brief, de-cellularized E2.5 uteri samples treated with collagenase-1 or vehicle were fixed in situ by using 2% (wt/vol) glutaraldehyde in 0.1-M phosphate buffer (pH 7), en-bloc stained in 2% (wt/vol) osmium tetroxide, 1.5% (wt/vol) potassium ferrocyanide in 0.1-M cacodylate buffer (pH 7.2), for 1 h at 25°C. Specimens were then washed again 5 × 3 min in ddH2O and then specimens were placed in 2% osmium tetroxide in ddH20 for 40 min at 25°C. Samples were then infiltrated and embedded in TAAB 812 HARD after sectioning using a Gatan 3View microtome within an FEI Quanta 250 SEM. For endometrium samples, a 41 × 41 μm field of view was chosen and imaged by using a 4096 × 4096 scan; section thickness was set to 100 nm in the Z (cutting) direction. Z volumes datasets comprised 700 images (100 μm z depth). The IMOD suite of image analysis software was used to build image stacks, reduce imaging noise, and generate 3D reconstructions. Fibrils were contoured using functions in the IMOD image analysis suite.
Proteomic analysis
Mice uteri samples (E2.5) were excised and a vertical incision was made to gain maximum exposure of the endometrium. The samples were incubated in 100 μl of DMEM supplemented with TNC (vehicle) or 100 nM collagenase-1 for 4 h at 37°C. For matrisome proteins enrichment, the E2.5 were de-cellularized, then a vertical incision was made and the samples were incubated in 100 μl of TNC buffer supplemented with 100 nM collagenase-1 for 24 h at 37°C (for control no enzyme was added). At the end of incubation, the supernatants were collected. 10 μg total protein from each sample was mixed 1:1 with 8 M urea, 100 mM Tris–HCL (pH 8.5) to a final concentration of 4 M urea. TCEP was added to a final concentration of 10 mM and incubated in shaking at RT for 30 min. Then, CAA was added to a final concentration of 40 mM and incubated in shaking at RT for 30 min. The samples were then diluted 1:3 with 50 mM ammonium bicarbonate buffer. In-solution digestion was performed with LysC-Trypsin mix (1:100 enzyme: protein ratio) and trypsin (1:50 enzyme: protein ratio; Promega). Peptides were desalted on C18 stage tips, vacuum dried, and resuspended in 0.1% TFA.
Peptides were introduced into the mass spectrometer by means of Waters nanoAcquity HPLC system connected to a Symmetry trap column (180 µm × 20 mm) and Analytical column HSS T3 75 µm × 250 mm, both from Waters. Data were acquired on Q exactive HF mass spectrometer (Thermo Fisher Scientific) with the Top 15 method. Raw files were searched against the Mus musculus (Taxonomy ID: 10090) database compiled from the UniProt reference proteome using Sequest from Proteome Discoverer 2.4 software (Thermo Fisher Scientific). The following parameters were selected for database searches: semi-Trypsin for enzyme specificity with tolerance of one missed cleavage; carbamidomethyl(C) as fixed modifications, and acetyl (N-term), pyroQ (N-term), oxidation (M), deamidation (NQ), as variable modifications. Precursor mass error tolerance of 10 ppm and fragment mass error at 0.02 D. Percolator was used for decoy control and FDR estimation (0.01 high confidence peptides, 0.05 medium confidence).
Cell extraction and Western blotting
Frozen uterus tissues of 1 cm length were washed in PBS, homogenized in 0.5 ml RIPA buffer (EMD Millipore) with a protease inhibitor (Roche) using a hand homogenizer and centrifuged (14,000g, 15 min, 4°C). Supernatants were resuspended in sample buffer (200 mM Tris pH 6.8, 40% glycerol, 8% SDS, 100 mM DTT, 0.2% bromophenol blue), and boiled for 5 min. Tissue extracts were then subjected to SDS PAGE and transferred onto nitrocellulose membranes (Whatman) by electro-blotting. Membranes were blocked in Tris-buffered saline with Tween 20 (TBST) buffer (200 mM Tris pH 7.5, 1.5 M NaCl, 0.5% Tween 20) and 5% BSA (60 min, 25°C) and then incubated with the corresponding primary Ab (60 min, 25°C), washed three times with TBST and incubated with HRP-conjugated secondary antibody (60 min, 25°C). Quantification of the band intensities was performed using the ImageJ analysis tool.
Antibodies used in this study: VEGF-A (A-20) (sc-152; Santa Cruz Biotechnology), LIF (AF449; R&D systems), VEGF-R2 (55B11; cell signaling), NKp46 (AF2225; R&D systems), β-tubulin (sc-9104; Santa cruz). Polyclonal anti-PEDF was kindly provided by Dr. Galia Maik-Rachline (Prof. Roni Seger lab, Weizmann institute, Rehovot, Israel). Secondary antibodies (both anti-rabbit and mouse) conjugated to HRP were purchased from Jackson ImmunoResearch (cat No.111-001-003 and 115-001-003, respectively). Antibodies were used at the manufacturer’s recommended dilution.
Immunofluorescence staining
E6.5 uterine samples were fixed with PBS 4% PFA, paraffin embedded and sectioned (4 μM). Sections were deparaffinized and epitope retrieval was performed in citric acid buffer (pH 6). Samples were blocked in PBS, 20% normal horse serum, and 0.2% Triton X-100 and then incubated with primary Ab in PBS, containing 2% normal horse serum and 0.2% Triton X-100 (overnight, 25°C). Next, samples were washed three times in PBS and incubated with a secondary antibody (60 min, 25°C) and mounted in a mounting medium. Primary antibodies: CD34 (CL8927PE; Cedarlane). Secondary antibody conjugated to a fluorophore was purchased from Jackson ImmunoResearch. For each image, a suitable threshold was applied for CD34 channel, and Analyze Particles plugin was used to detect the covered area.
Immunohistochemical stain for integrins
E4.5 uteri were fixed in 4% PFA, incubated in sucrose 30% and embedded in OCT. serial 12 mm sections were prepared using cryostat. Samples were blocked in PBS, 20% normal horse serum, and 0.2% Triton X-100 and then incubated with primary Ab diluted 1:100 in PBS, containing 2% normal horse serum and 0.2% Triton X-100 (overnight, 25°C). The corresponding secondary HRP antibodies were used at manufacturer’s recommended dilutions (Jackson ImmunoResearch). A peroxidase substrate kit (SK-4100; Vector Labs) was used as a chromogen and hematoxylin as a counterstain. Tissue exposed only to the secondary antibody was used as negative control. primary antibodies: anti-integrin alpha V antibody (ab179475; Abcam), anti-integrin beta 3 (SJ19-09; Invitrogen).
MRI imaging
MRI experiments were performed at 9.4 T on a horizontal-bore Biospec spectrometer (Bruker) using a linear coil for excitation and detection (Bruker) as reported previously (Plaks et al, 2006). The animals were anesthetized with isoflurane (3% for induction, 1–2% for maintenance; Abbott Laboratories) in 1 liter/min oxygen, delivered through a muzzle mask. Respiration was monitored, and body temperature was maintained using a heated bed. The pregnant mice were serially scanned at E4.5. Three-dimensional gradient echo (3D-GE) images of the implantation sites were acquired before, and sequentially, for 30 min after i.v. administration of the contrast agent. A series of variable flip angle, precontrast T1-weighted 3D-GE images were acquired to determine the precontrast R1 (repetition time [TR]: 10 msec; echo time [TE]: 2.8 msec; flip angles 5°, 15°, 30°, 50°, 70°; 2 averages; matrix, 256 ׳ 256 ׳ 64; field of view, 35 ׳ 35 ׳ 35 mm3). Postcontrast images were obtained with a single flip angle (15°). During MRI experiments, the macromolecular contrast agent biotin-BSA-GdDTPA (80 kD; Symo-Chem), 10 mg/mouse in 0.2 ml of PBS, was injected i.v. through a preplaced silicone catheter inserted into the tail vein. The MRI scans allowed quantification of the fBV and the permeability surface area product (PS) of embryo implantation sites, as previously reported (Plaks et al, 2006). In brief, the change in the concentration of the administered biotin-BSA-GdDTPA over time (Ct), in the region of interest, was divided by its concentration in the blood (Cblood); calculated in the region of interest depicting the vena cava, also acquired during MRI, and extrapolated to time 0. Linear regression of these temporal changes in Ct/Cblood yielded 2 parameters that characterize vascular development and function: (a) fBV (fBV = C0/Cblood), which describes blood-vessel density and is derived from the extrapolated concentration of the contrast agent in implantation sites, at time zero, divided by the measured concentration in the vena cava, ∼5 min after i.v administration, and (b) PS = ([Ct–C0]/[Cblood ׳ t]), which represents the rate of contrast agent extravasation from blood vessels and its accumulation in the interstitial space and which is derived from the slope of the linear regression of the first 15 min after contrast agent administration (t = 15). Mean fBV and PS were calculated separately for single implantation sites, considering homogeneity of variances between mice. At the end of the MRI session, embryo implantation sites were harvested and immediately placed in 4% PFA after euthanizing the pregnant mice by cervical dislocation. Slides were then washed in PBS and incubated in Cy3 or Cy2-conjugated streptAvidin (Jackson Immunoresearch Laboratories), diluted 1:150 in PBS for 45 min.
Quantitative real-time PCR (qRT-PCR)
E4.5 uteri samples treated with collagenase-1 or vehicle were homogenized using a hand homogenizer. Total RNA was isolated using PerfectPure RNA Tissue Kit (5 Prime GmbH, Deutschland). 1 μg of total RNA was reverse transcribed using High Capacity cDNA Kit (Applied Biosystems Inc.). qRT-PCR was performed using specific primers with SYBR Green PCR Master Mix (Applied Biosystems Inc.) on ABI 7300 instrument (Applied Biosystems) readouts were normalized to a B2M housekeeping. The cDNA quantity was 8 ng and the primer concentration was 0.15 μM in a total reaction volume of 20 μl. Primer sequences are listed in Table 1 below. Data are presented as mean fold change using the 2-∆∆CT method (Schmittgen & Livak, 2008). The standard error of the mean (SEM) was calculated on the 2-∆∆CT data, as was the statistical analysis.
Primer sequences used for qRT-PCR.
Cell isolation from uterus tissue and flow cytometry analysis
Single implantation sites were stained using i.v. injection of Evan’s blue dye (Sigma-Aldrich), excised and isolated on E4.5. Tissues were minced into small fragments and incubated in shaking for 40 min at 37°C with PBS +/+ containing 0.5 mg/ml collagenase type IV (Sigma-Aldrich, Rehovot, Israel) and 0.1 mg/ml DNase I (Roche). Digested tissue was filtered and smashed with a syringe plunger through a 250 μm nylon sieve in FACS buffer (PBS, 2% FCS, 2 mM EDTA) to mechanically dissociate the remaining tissue. The supernatant cell pellet was then centrifuged at 390g, and the pellet cells were lysed for erythrocytes using a red blood cell lysis buffer (Sigma-Aldrich) (2 min, 25°C). Then, the cells were incubated with antibodies for 30 min in FACS buffer (dark, 4°C) and then washed once with FACS buffer. Cells were analyzed with BD LSR II, special order system (BD Biosciences). Flow cytometry analysis was performed using FlowJo software (TreeStar). The following anti-mouse antibodies were used, diluted 1/100 resulting in concentration ranging from 2 to 5 μg/ml): CD45 (clone 30-F11), CD3ε (clone 145-2C11), NKp46 (clone 29A1.4), NK1.1 (clone PK136), CD64 (clone 10.1), CD11b (clone M1/70), CD11c (clone N418), IAb (clone AF6-120.1)—all purchased from BioLegend (San Diego, USA). Anti-mouse F4/80 (clone A3-1) was purchased from BIO-RAD.
Human uterine samples
Fresh uterine samples from healthy women biopsies were obtained by Dr. Eitan Ram, Gynecologic Oncology Division, Helen Schneider Hospital for Women, Rabin Medical Center; Petah-Tikva. Written informed consent was obtained from the patients providing samples in compliance with the specified Helsinki approval (no. 0450-16-RMC). The tissues were de-cellularized using the same procedure already described. The samples were treated by vehicle or 50 nM collagenase-1 in a volume of 50 μl at 37°C for 8 h.
Statistical analysis
Statistical analyses were carried out using GraphPad Prism software (VIII; GraphPad Software Inc.). Data were analyzed by unpaired, two-tailed t test to compare between two groups. Multiple comparisons were analyzed by one-way analysis of variance (ANOVA). After the null hypothesis was rejected (P < 0.05), Tukey’s Honestly Significant Difference or Dunnett tests were used for follow-up pairwise comparison of groups in the one-way ANOVA. Data are presented as mean ± SEM in the figures; values of P < 0.05 were considered statistically significant (∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001).
Data Availability
The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE (Perez-Riverol et al, 2022) partner repository with the dataset identifier PXD054389.
Acknowledgements
We thank Dr. Alina Berkovitz for embryo preparations and fruitful discussions, Ms. Anna Aloshin for collagenase-1 purification, and Ms. Yinhui Lu at the University of Manchester for collecting SBF-SEM images. We thank Dr. Ori Brenner for the assistance in histopathological characterization of mice uteri. The images in this article were acquired at the Optical Imaging & Translational Bioengineering Unit, Department of Veterinary Resources, and the Advanced Optic Imaging Unit, de Picciotto-Lesser Cell Observatory In memory of Wolfgang and Ruth Lesser at the Moross Integrated Cancer Center Life Science Core Facilities, Weizmann Institute of Science. This research was supported by the Israeli Science Foundation (1226/13. I Sagi).
Author Contributions
E Zehorai: conceptualization, resources, software, formal analysis, validation, investigation, visualization, methodology, and writing—original draft, review, and editing.
T Gross Lev: software, formal analysis, validation, investigation, visualization, methodology, and writing—original draft, review, and editing.
E Shimshoni: software, methodology, and writing—review and editing.
R Hadas: formal analysis, validation, methodology, and writing—review and editing.
I Adir: formal analysis, validation, investigation, and writing—review and editing.
O Golani: software, formal analysis, methodology, and writing—review and editing.
G Molodij: software, formal analysis, methodology, and writing—review and editing.
R Eitan: resources.
KE Kadler: resources, investigation, methodology, and writing—review and editing.
O Kollet: validation and writing—review and editing.
M Neeman: resources and writing—review and editing.
N Dekel: methodology and writing—review and editing.
I Solomonov: conceptualization, formal analysis, supervision, validation, investigation, visualization, and writing—original draft, review, and editing.
I Sagi: conceptualization, supervision, funding acquisition, validation, investigation, visualization, and writing—original draft, review, and editing.
Conflict of Interest Statement
“Compositions for remodeling extracellular matrix and methods of use thereof” (US patent no. 10722560). Assignee: NanoCell Ltd, Investors: I Sagi, I Solomonov, E Zehorai. All other authors declare that they have no competing interests.
- Received February 13, 2024.
- Revision received July 25, 2024.
- Accepted July 25, 2024.
- © 2024 Zehorai et al.
https://creativecommons.org/licenses/by/4.0/This article is available under a Creative Commons License (Attribution 4.0 International, as described at https://creativecommons.org/licenses/by/4.0/).
References
