Biomimetic Janus Collagen Patch with Antideformation, Antiadhesion, and Prohealing Properties for Rehabilitation of Full‐Layer Uterine Injuries, Toward Efficient Live Births

Uterine abnormalities infertility can be treated with implantable patches for tension‐free healing. However, the existing clinical patches are still limited in effective uterus recovery, resulting in low pregnancy and live birth rates. Herein, a novel engineered, and biocompatible Janus collagen patch (JCOP) is developed to achieve excellent uterine defect recovery and increase effective live births. The JCOP can highly match the structure, micromorphology, and mechanical properties of a natural uterus due to homologous‐like tissue homogeneity. JCOP's rough surface and loose‐extracellular‐matrix‐like porosity promote fibroblast adhesion and tissue regeneration, whereas its smooth surface reduces fibroblast adhesion and prevents visceral adhesions. Additionally, it exhibits exceptional resistance to swelling and deformation along with stable mechanical strength for long‐term preservation in the body and can withstand the highest uterine pressure in a humid interior environment attributed to its superior structural layout. Moreover, JCOP@bF, a functional JCOP linked to basic fibroblast growth factor (bFGF) via a specific collagen‐binding domain, significantly promotes repair of the injured uterus, restores the ability of the endometrium to accept embryos, and supports their development to a viable stage in a rat model of uterine damage. This study shows great promise for treating infertility caused by uterine damage.


Introduction
In modern society, infertility has emerged as an increasingly severe public health issue, with a soaring number of women suffering from primary or secondary infertility as a result of uterine trauma. [1] Total or partial uterine damage and the resulting scar formation are ubiquitous in women of childbearing age after cesarean delivery, [2] myomectomy, [3] or inappropriate curettage. [4] After natural repair, the endometrial layer in the damaged region is usually too thin to accommodate placental implantation, and the absence of myometrium in this area may also make it impossible to sustain another pregnancy. [3,5] Several synthetic meshes that are lightweight and exhibit high tensile strength and good deformation resistance, such as polypropylene (PP) and polyester meshes, have been employed extensively for the tension-free repair of the injured uterus. [6] However, owing to foreign body reactions, these meshes can cause significant endometrial Uterine abnormalities infertility can be treated with implantable patches for tension-free healing. However, the existing clinical patches are still limited in effective uterus recovery, resulting in low pregnancy and live birth rates. Herein, a novel engineered, and biocompatible Janus collagen patch (JCOP) is developed to achieve excellent uterine defect recovery and increase effective live births. The JCOP can highly match the structure, micromorphology, and mechanical properties of a natural uterus due to homologous-like tissue homogeneity. JCOP's rough surface and looseextracellular-matrix-like porosity promote fibroblast adhesion and tissue regeneration, whereas its smooth surface reduces fibroblast adhesion and prevents visceral adhesions. Additionally, it exhibits exceptional resistance to swelling and deformation along with stable mechanical strength for long-term preservation in the body and can withstand the highest uterine pressure in a humid interior environment attributed to its superior structural layout. Moreover, JCOP@bF, a functional JCOP linked to basic fibroblast growth factor (bFGF) via a specific collagen-binding domain, significantly promotes repair of the injured uterus, restores the ability of the endometrium to accept embryos, and supports their development to a viable stage in a rat model of uterine damage. This study shows great promise for treating infertility caused by uterine damage.
adhesions, resulting in inadequate wound healing. [7] Even though natural-biopolymer-based materials are used widely for repairing endometrial abnormalities, these materials cannot always prevent detrimental endometrial adhesions owing to a lack of differences in the functionalities of their positive and negative surfaces. [8] Constructing composite patches with antiadhesive barriers is one of the most common methods for treating uterine adhesions. [9] Parietex composite (PCO) nets, for instance, are made from polyester coated with an antiadhesive collagen-based barrier and are designed to be effective against visceral adhesions. [10] However, polyester or PP-based composite nets can result in poor wound healing owing to their inherently inadequate inflammatory response and the unsuitability of their microstructure for cell migration and proliferation. [11] In addition, patches constructed from natural polymers, such as hyaluronic acid, may swell in the moist uterine environment, leading to the distortion of the composite mesh. [12] These problems are the primary reason tension-free mending techniques do not yield the desired results. Therefore, the repair of uterine defects requires overcoming the scientific challenge of creating a patch with antideformation, antiadhesion, and prohealing characteristics.
There has been significant progress in the synthesis of antiadhesion barriers based on tissue-adhering surfaces, and these barriers have been formed using different methods. [13] These include creating dense physical surfaces, [6,14] superhydrophobic surfaces, [15] and negatively charged surfaces. [16] Hydrophilic surfaces, such as those that mimic a loose extracellular matrix (ECM) and positively charged surfaces, have been realized in bioengineered patches to improve adhesion and tissue formation. [17] Given their remarkable tunability when used in multifunctional platforms, hydrogels with an ECM-like structure are extremely promising bioengineered patches. [18] However, many antiadhesion hydrogels have the propensity to swell considerably and show significant macroscopic deformation in both dry and humid environments, which is not ideal for the tensionfree repair of soft tissues. [19] Therefore, realizing antideformation, antiadhesion, and healing characteristics simultaneously in a single hydrogel is the key to the successful use of patches for uterine defect repair. Concerning materials that show both antiadhesion and prohealing properties, asymmetric hydrogels are a promising option.
ECM-like materials with biochemical and biophysical properties comparable to those of the uterine ECM have been employed to promote tissue repair and regeneration. [20] The mechanical characteristics of ECM-like hydrogels can be modified such that they are similar to the mechanical and biophysical properties of native tissues to improve stem cell proliferation and differentiation. [21] Collagen, which is a primary structural component of the ECM, is used widely for tissue regeneration and wound healing because of its exceptional biodegradability and biocompatibility. [22] In addition to being a structural support system for tissues, collagen also plays a role in regulating cellular activities, such as adhesion, migration, and differentiation. [23] Previous studies have highlighted the efficacy of using stem cells encapsulated within collagen materials to heal the damaged uterus. [23][24] Here, we have successfully developed a novel Janus collagen patch (JCOP) with tunable shapes, which are based on the Janus structure of the intestine synthesized using engineering technology ( Figure 1A). The bFGF can activate the mitogenactivated protein kinase (MAPK) pathway, resulting in potent angiogenic action. [25] In our work, the preparation of fused bFGF by genetic engineering method, which convergence the collagen-binding domain (CBD) with peptide sequence TKK-TLRT to create a fusion protein. [26] The CBD-bFGF can specifically bind to the collagen scaffold, subsequently, functionalized JCOP@bF can be obtained by simple fusion incubation with JCOP. These poly-collagen-based patches with a Janus porous structure were synthesized using an engineered decellularized plasm fabrication approach, lyophilization, and rehydration, as shown in Figure 1A. To aid fibroblast attachment and soft tissue formation, the asymmetric porous JCOP has a porous surface similar to an ECM, while its thick and smooth surface prevents the adhesion of other abdominal contents. In addition, unlike other patches, JCOP has natural-uterus-mimicking characteristics. Specifically, it is similar to the natural uterus in terms of its structure, microscopic morphology, and mechanical properties. The excellent antiswelling and mechanical properties of JCOP allow it to exhibit long-term antideformation characteristics and maintain its outstanding compressive strength such that it can withstand the maximum intrauterine pressure in the internal moist environment for the effective tension-free repair of uterine defects. Establishing a full-layer incision model in the rat uterus allowed for the implantation of JCOP@bF at the surgical site ( Figure 1B). The results showed that JCOP@ bF effectively boosted the proliferative capacity of endometrial and muscle cells, accelerated microvascular regeneration, and restored the potential of the endometrium to accept and support the development of viable embryos ( Figure 1C). Thus, JCOP@bF is a rosy therapeutic agent for the therapy of uterine damage.

Synthesis of Biomimetic JCOP
The approximate steps for synthesizing the biomimetic JCOP, are shown in Figure 1A and the Experimental Section. The final product was a customizable centimeter-sized aerogel membrane ( Figure S1, Supporting Information). Because of the cuttable and water-absorbing nature of the membrane, we were able to obtain JCOP samples of different shapes (hearts, circles, squares, and rectangles) and forms (in the dry and wet states) ( Figure S2, Supporting Information). The structural integrity of the collagen is important for linking CBD-bFGF, so the 3D helical structure of collagen comprising the JCOP remained intact and was verified using circular dichroism measurements ( Figure S3, Supporting Information). A negative peak was observed at ≈190 nm and a positive peak at ≈220 nm, these were indicative of typical α-helix and β-fold structures, thus confirming that the three-stranded helix structure of collagen was intact and that JCOP exhibited excellent biological activity. The water contact angle of JCOP, which is closely correlated with surface roughness and hydrophilicity, was measured. As shown in Figure S4 in the Supporting Information, the reverse side of JCOP had a contact angle of 106.3° at 1 s, and remains constant www.advmatinterfaces.de for the next 5 s, suggesting that the surface was smooth and dense. The front side of JCOP exhibited optimal surface wettability and a contact angle of less than 30° at 5 s. In our study, one pair of bovine intestines could produce a meter-sized collagen membrane patch that could treat over 200 rats with severe uterine defects, with great potential for large-scale production.

Characterization of Natural Uterus-Like Basic
Properties of JCOP Figure 2A shows a photograph of JCOP in the dry state, while Figure 2B shows JCOP in the wet state, in the wet state, it could be stretched to 1.5 times its original length. Scanning electron microscopy (SEM) images ( Figure 2C) of the lyophilized JCOP hydrogel confirmed that it had a Janus structure similar to that of the natural uterus ( Figure S5, Supporting Information), exhibiting a highly porous fibrous network-like structure on its top surface, a layered structure in its middle, and a dense structure on its bottom surface. Additional images showing the top-surface morphologies of JCOP, and the natural uterus are shown in Figure 2D,E, respectively. We measured the porosities of JCOP, JCOP@bF, and the natural uterus, which were found to be 89.41 ± 1.2%, 81.5 ± 0.59%, and 85.76 ± 1.7%, respectively ( Figure 2F), while the mean pore sizes for their top surfaces were 18.72 ± 6.32, 21.19 ± 8.98, and 14.29 ± 2.08 µm, respectively ( Figure 2G). Interestingly, the loose and porous structure of the rough top surface was similar to that of the uterine ECM in terms of the porosity and mean pore size, and the pore size was of the same order of magnitude as that of human fibroblasts, this will facilitate the attachment and infiltration of fibroblasts. Thus, it was assumed that the porous top surface of JCOP would support cell adhesion and proliferation to accelerate tissue repair, if JCOP were to be used as a patch to treat Figure 1. Janus Collagen Patch (JCOP@bF) for efficient live birth uterine recovery in rats. A) Construction of JCOP@bF for uterine healing. bFGF was obtained from Escherichia coli by genetic engineering and subsequently attached to the patch to form JCOP@bF. JCOP@bF can be designed into shapes (e.g., rectangular, triangular prismatic, and cubic) according to requirements. B) JCOP@bF for therapy of rats with grievous uterine injury. According to the structural characteristics of the uterus, the porous side of the extracellular matrix-like surface on JCOP@bF faces toward the uterine cavity, promoting endometrial regeneration, and muscle collagen reconstruction and accelerating wound healing. The dense and smooth surface faces outward, effectively preventing abdominal adhesions. C) JCOP@bF not only has a physical barrier to block adhesions but also has a slow-release bFGF factor to regulate the uterine microenvironment and promote uterine recovery. It can be seen that JCOP@bF can effectively promote the recovery of the uterus and support live birth in rats.
www.advmatinterfaces.de uterine abnormalities. Conversely, it was also assumed that the bottom surface of JCOP would inhibit cellular and visceral attachment owing to its dense and smooth topology.
Matching the mechanical properties of patch materials to those of the target tissue to be repaired is essential as this facilitates accelerated wound healing. [27] We tested the tensile properties of JCOP and JCOP@bF as well as those of the natural uterus ( Figure 2H). Both JCOP and JCOP@bF fracture stresses reached about 0.6 MPa, slightly greater than the 0.51 MPa of the natural uterus. As a visual indication, the JCOP with a width of 3 mm can lift a weight of 100 g. The fracture strain natural uterus (75.93 ± 1.56%) is slightly greater than that of JCOP (60.78 ± 1.82%) and JCOP@bF (62.05 ± 1.97%) ( Figure S6, Supporting Information). In addition, the rheological properties of the samples were analyzed to evaluate their hydrodynamic characteristics. The time-scan curves ( Figure 2J) showed that the elastic moduli of JCOP, JCOP@bF, and natural uterus were of the same order of magnitude. These results indicated that JCOP and JCOP@bF have a Janus microscopic morphology and mechanical properties similar to those of natural uterine tissue, making JCOP (and JCOP@bF) suitable for use for uterus repair.

Evaluation of In Vitro and In Vivo Dynamic Stability of JCOP
Owing to the humid and complex in vivo environment, uterine repair materials are inevitably subjected to internal and external tissue extrusion, friction, and wetting, which may cause significant swelling and the deterioration of their mechanical properties, leading to poor repair results. [19] The ideal antiadhesion patch hydrogel should show high resistance to deformation. In addition, it should be able to tolerate the pressure in the uterus and withstand internal and external compression and friction K) The elastic modulus of rat uterus, JCOP, and JCOP@bF. Data are mean ± SD (n = 3; ns : no significance; *p < 0.05, ***p < 0.001).

www.advmatinterfaces.de
during the repair of tissue defects. The repair outcomes for soft uterus tissue defects are significantly affected by the mechanical and antiswelling properties of the material used. As shown in Figure 3A, JCOP hydrogel is flexible and easy to handle and can be easily sutured to porcine muscle. It can withstand bending, distorting, phosphate-buffered solution (PBS) immersion, and stretching and still return to its original wrinkle-free state. The patch hydrogel is also stably sutured to the natural uterus and stays in place in a dynamic environment with excellent recoverability ( Figure S7, Supporting Information). The uterine leakage model ( Figure 3B) was used to evaluate the resistance of the sutured JCOP to fluid leakage. [28] Leakage of dye was observed in the untreated incision after injection of 2 mL of blue dye. However, in the JCOP group, no significant fluid leakage was observed ( Figure 3B). We also measured the sealing pressure using a burst pressure test device ( Figure 3C). The results showed that the sealing pressure in the case of the JCOP hydrogel group was 352. 7 ± 37 kPa and significantly greater than that for the suture group (239.03 ± 33.09 kPa; Figure 3D). The tearing force of the JCOP hydrogel was also measured to assess its fracture under stress deformation ( Figure S8, Supporting Information), the results show that the existence of a force exceeding 0.6 N will cause the obvious cracks of JCOP hydrogels.
The cyclic stability of JCOP was evaluated next. At 30% tensile strain, JCOP was able to recover its initial shape after 200 loading/unloading cycles ( Figure 3E). The stress loss of only 10.5% after 200 cycles of stretching ( Figure S9, Supporting Information). The tensile stress (0.15-0.2 MPa) during cyclic loading was noticeably higher than the maximum pressure www.advmatinterfaces.de within the rat uterus at 30% tensile strain ( Figures 2H and 3E), indicating that the JCOP hydrogel revealed extraordinary fatigue and deformation resistance.
The long-term stability of the JCOP hydrogel was also assessed. The tensile strength of JCOP was measured by placing it in PBS solution for 1, 7, and 14 days, respectively. The results showed that the tensile strength did not change significantly even after immersion in the PBS solution for 7 and 14 days ( Figure 3F). Moreover, the average swelling ratios of the JCOP hydrogel after immersion in the PBS solution for 1, 3, 7, and 14 days were 24.1%, 25.1%, 24.3%, and 24.5%, respectively ( Figure 3I) and considerably lower than those of most documented antiadhesion patches. The low swelling ratios can be attributed to the high number density of hydrogen-bonded crosslinks.
The JCOP hydrogels are subjected to friction and compression in all directions in the body, and to simulate this dynamic effect, rheological characterization was performed to evaluate the stability of the JCOP hydrogel during frictional extrusion deformation in vivo. Specifically, frequency ( Figure 3G) and strain scans ( Figure 3H) were performed to simulate high-frequency friction and extrusion in vivo. During high-frequency rotation (1-10 Hz) and strain extrusion (1-10%), the trend in the modulus of the JCOP hydrogel was similar to that of gels, indicating that JCOP remained structurally stable at both high frequencies and strains. Finally, the rate of bFGF release from JCOP@bF in vitro was evaluated. The temporal assay data ( Figure 3J) showed that the bFGF loaded on the JCOP hydrogel was released continuously for 14 days. Similar to the results of our previous work, [29] the factor can be released slowly over 14 days to achieve long-term retention of the treatment. These results suggest that JCOP hydrogels with excellent in vitro deformation resistance may meet the mechanical requirements for prolonged uterus implantation in vivo.

Evaluation of Biocompatibility of JCOP
Cytotoxicity tests were performed before JCOP was used for in vivo repair. The biocompatibility of the hydrogel was tested with NIH 3T3 cells using the LIVE/DEAD and Cell Counting Kit-8 assays. The cells were also cultivated on sterile tissue culture dishes as a control. The status of the fibroblast cultures was monitored for 1, 3, 5, and 7 days. Most fibroblasts grew well after 7 days of coculturing with the hydrogel, and only a few dead cells were observed in the LIVE/DEAD assay during the culturing period ( Figure S10, Supporting Information). The high cytocompatibility and bioactivity of collagen account for the low cytotoxicity of JCOP. The significantly enhanced activity of the JCOP@bF group was owing to the release of bFGF, which improved fibroblast viability. In addition, all the groups exhibited excellent cell proliferation propensity. The relative cell activity was counted at 1, 3, 5, and 7 days. After 7 days, the cell survival rates of the two treatment groups (JCOP and JCOP@ bF) were 122.3.2 ± 1.2 and 137.3 ± 2.5, respectively, which were statistically greater than that of the control group ( Figure S11, Supporting Information). Thus, the data revealed that both JCOP and JCOP@bF showed high bioactivity and cell proliferation ability because they form a microenvironment suitable for cell survival. This is especially true for the latter owing to the slow release of bFGF. The JCOP@bF hydrogel showed remarkably high cell proliferation and survival abilities. In summary, JCOP@bF is highly cytocompatible and stimulates fibroblast growth and thus can be applied as an effective trauma patch in vivo.
Next, SEM was performed to analyze the adherence and diffusion patterns of fibroblasts on the surface of the porous JCOP@bF hydrogel ( Figure S12, Supporting Information). The fibroblasts attached and proliferated well in the JCOP@ bF hydrogel, as several normal spherical and elliptical fibroblasts were observed sticking to the hydrogel surface after the first day, while a few fibroblasts had penetrated the material. Intriguingly, fibroblasts eventually covered almost the entire hydrogel surface as the culturing time increased. These data suggest that the highly porous ECM-like structure of JCOP and the characteristic three-stranded helical structure of collagen provide strong support for fibroblast growth and proliferation.

Evaluation of In Vivo Antiadhesion Properties of JCOP
The bioadhesive behavior of JCOP was investigated both in vitro and in vivo to determine how its Janus porous structure affects adherence. [11] Since fibroblasts are essential for the formation of fresh connective tissue throughout the healing process, NIH3T3 cells were seeded on the JCOP hydrogel. After 1 day of culture, more NIH3T3 cells adhered and grew on the porous surface of the JCOP hydrogel (Figure 4A), which resembled an ECM. In contrast, only a few NIH3T3 fibroblasts were detected on the dense and smooth surface ( Figure 4A). Here we refer to the reverse smooth surface placement as RJCOP. The relative areas with adhering cells after 1 day of culturing were 14.26 ± 2.39% and 1.0 ± 0.38%, respectively ( Figure S13, Supporting Information). SEM further confirmed the adhesion of fibroblasts to the different surfaces of JCOP. More spherical and elliptical fibroblasts were attached to the porous top surface of JCOP after 1 day of incubation ( Figure 4B), while only a few fibroblasts were found on its dense and smooth bottom surface ( Figure 4B); the relative numbers of adhering cells were 7.33 ± 2.08 and 43 ± 6.24, respectively (Figure S14, Supporting Information). This result again confirmed the distinct bioadhesive properties of the top and bottom surfaces of JCOP, in that the top surface has proadhesion properties, while the bottom surface has antiadhesion properties. The anisotropic bioadhesive properties of JCOP are modulated by its Janus porous structure.
To further evaluate the bioadhesive properties of JCOP in vivo, a rat model of a total uterine defect was constructed, and the uterine defect was repaired using different treatment methods. [6] To verify the antiadhesion properties of the hydrogel, its porous surface was sutured to the uterine cavity side, while its dense and smooth surface was oriented toward the abdominal cavity. An untreated group, a suture group, and a group with JCOP placed in the opposite direction (RJCOP) were used as the control groups. The rats were sacrificed 14 days after the implantation patch to check for postoperative abdominal visceral adhesions. In addition, the clinical grading of the degree of adhesion, type of adhesion, and adhesion toughness was performed to quantify the level of adhesion www.advmatinterfaces.de (Table S1, Supporting Information). Figure 4C shows photographs of the visceral adhesions of the various treatment groups, while Figure 4D,E; and Figure S15 (Supporting Information) show the corresponding quantitative scores. When left untreated, the visceral adhesions reached the maximum clinical adhesion score of 9.25 ± 0.96. Specifically, the degree of complete adhesion in the case of the untreated group was 4.0 ± 0, which meant that the endometrium had adhered completely to the intestine and omentum, resulting in a wide range of complications, including uterine effusion and intestinal obstruction. The adhesion extent score for the suture group was 3.0 ± 0.82, and the thickness of the adhering visceral tissue was considerably higher (nearly 60%); the adhesion tenacity score was 2.0 ± 0.82 because of the difficulty in separating the adhering organs from the horns during blunt dissection. In contrast to the case for the untreated and suture groups, the JCOP hydrogel resulted in essentially no visceral adhesion and had a very low clinical score (0.5 ± 1.0). These findings confirmed that the dense and smooth bottom surface of the JCOP hydrogel can reliably prevent the formation of visceral adhesions. To prove that the dense and smooth bottom surface of the JCOP hydrogel is responsible for its superior antiadhesion www.advmatinterfaces.de characteristics, the porous top surface was reverse fixed in a control group (referred to as RJCOP). As predicted, the viscera almost completely covered the porous top surface of RJCOP, with the clinical adhesion score being 5.75 ± 0.96. These findings imply that, despite having the same chemical and mechanical properties as JCOP, the bioadhesive surfaces of RJCOP exhibited distinct functionalities in vivo. Hence, the antiadhesion surface design of JCOP is an effective one.

Efficacy of JCOP for the Repair of Severe Uterine Defects
To verify the efficacy of JCOP@bF in promoting the recovery of the entire uterus, we constructed a rat model of severe uterine defects ( Figure 1B; and Figure S16, Supporting Information). The entire animal experimental procedure we have demonstrated in Figure 5A. Five groups of rats were generated at random: group 1, normal rats without any intervention; group 2, rats with severe uterine defect surgery without therapy; group 3, rats with severe uterine defect surgery with sutures only; group 4, rats with severe uterine defect surgery with the transplantation of JCOP; and group 5, rats with severe uterine defect surgery with the transplantation of JCOP@bF.
As shown in Figure S17 (Supporting Information), in the case of group 5 (JCOP@bF), the uterine structures regenerated almost completely by day 90, similar to the case of group 1 (Sham). On the 90th postoperative day, the uterine wound had healed completely, and the patch (JCOP@bF) used to cover the defect had been absorbed. In the case of group 4 (JCOP), the JCOP hydrogel sutured to the defect was not completely absorbed; however, most rats exhibited uterine wound healing by day 90. In the case of group 3 (Suture), approximately half of the rats exhibited uterine deformation, indicating poor recovery of uterine function. Moreover, severe deformation and wound adhesions were present on day 30 after uterine defect surgery in the case of group 2 (Untreated), resulting in serious intrauterine stenosis and intrauterine adhesions. The recovery of the damaged uterus was evaluated further using H&E and Masson staining. Group 5 showed effective wound healing and endometrial regeneration ( Figure 5B; and Figure S18, Supporting Information). In addition, the myometrium of group 5 had regenerated almost completely by day 90 (Figure 5B,E). By day 90, the endometrial thickness as well as the number of endometrial glands were quite close to normal. ( Figure 5D; and Figure S19, Supporting Information). In addition, groups 3 and 4 showed good endometrial regeneration ( Figure 5B,D); however, the number of endometrial glands present was smaller than that for group 5 ( Figure S19, Supporting Information). In the case of group 2, there was a significant decrease in the number of glands. Adhesions cause the uterine wall to thin, which substantially impedes uterine healing. Based on these findings, we propose that the JCOP@bF hydrogel contributes to uterus recovery by fostering endometrial regeneration.

Effects of JCOP on Myometrium Regeneration and In Vivo and Ex Vivo Angiogenesis
We used α-smooth muscle actin (α-SMA) immunofluorescence staining to examine myometrial generation ( Figure 5C). [1c,30] Figure 5E,F shows that the JCOP@bF hydrogel group had a notably higher α-SMA expression level and uterine myometrium thickness than those of the other groups. As angiogenesis is one of the main indicators of wound healing, we used CD31 immunohistochemical staining to study angiogenesis in the regenerating endometrium ( Figure 6A). [31] The expression level of CD31 was higher in the JCOP@bF hydrogel group than those in the other three groups ( Figure 6C). To further investigate the efficiency of angiogenesis, we performed an in vitro angioplasty assay ( Figure 6B) and counted the number of new blood vessels and their total length ( Figure 6D,E). The JCOP@ bF hydrogel group performed better than the other two groups. These histological examination results provide strong evidence regarding the healing benefits of JCOP@bF. Thus, JCOP@bF, which has superior antiadhesion and prohealing qualities along with antideformation characteristics, is highly suited for rehabilitating the uterus to its predisruption state.

Effects of JCOP on Recovery of Gestational Function and Live Birth in Rats
Fertility restoration is a benchmark for measuring uterine function. We evaluated ovum implantation, pregnancy, embryo growth, and live birth to determine whether the uterus could recover from serious damage using JCOP. [20] As shown in Figure 5A, we established a flow chart for the overall repair of severe uterine defects and allowed the rats to mate 90 days after surgery. We assessed the gestational age based on the time of appearance of the vaginal pessary and abdominal palpation. The rats were euthanized 16-19 days after conception. Following that, the uterine horns were retrieved for analysis of the reproductive results, including the endometrium's ability to accept embryos (i.e., Conception time, mean number of fetuses, and mean fetal weight) and the fertility of the uterine horns (i.e., the number of live births). As shown in Figure 7A-E; and Table S2 (Supporting Information), group 5 achieved live births, similar to group 1 (Sham). For group 4, the pregnancy rate, mean gestation time, and the number of fetuses were similar to those of group 1 and much higher than those of the other groups, indicating that the JCOP@bF hydrogel can significantly improve the live birth rate.

Conclusions
We fabricated and evaluated a Janus collagen scaffold, which mimics the natural uterus structure and allows for almost www.advmatinterfaces.de complete uterine recovery as well as successful live delivery in rats with extensive uterine abnormalities. Compared with current uterine recovery systems that rely on biomaterial scaffold mediation, JCOP@bF has several distinct advantages: i) its microscopic morphology and mechanical properties are similar to those of the natural uterus and can be adapted to the structural and mechanical microenvironment of the uterus; ii) the unique Janus structure of JCOP@bF effectively prevents adhesion between the uterine and abdominal tissues and accelerates wound healing, and iii) JCOP@bF allows for live births in rats with serious uterine defects. Specifically, similar to the case for the control group, JCOP@bF allows for normal pregnancy in the regenerating uteri and promotes fetal growth as well as live birth.

Experimental Section
The Suzhou Institute of Nano-Tech and Nano-Bionics, Chinese Academy of Sciences, Review Committee, and Ethics Committee reviewed and approved the research protocol (approval number: SINANO/EC/2022-070). The DB Wuderegen Biomedical Technologies Co., Ltd (Zhenjiang, China) provided the materials for the experiment, and the core patents for material synthesis are owned by the company. The experimental materials, the core of material synthesis, physical performance characterization procedures, animal experimental methodologies, and histological analysis were all included in the Supporting Information.
Synthesis of Biomimetic JCOP: The fresh mammalian cavity tissue was soaked in a 10 wt% solution of acetylcysteine acid for 2 h and then the plasma membrane layer was removed; soaked in a 2 wt% trypsin solution Figure 6. Effects of in vivo and in vitro angiogenesis. A) Immunohistochemical staining images of CD31 at 30 and 90 days, red arrows point to neovascularization, scale bar: 300 µm. B) Tube formation assay. Tube formation after treatment with basal medium, JCOP leachate, and JCOP@bF leachate alone was compared microscopically. Images taken with 4x(a,b,c) and 10x objective lenses(a',b',c'). C) Quantitative statistics of neovascularization in A. D,E) Observation of tube Total length and number were calculated in three wells with 3-5 random areas in each well. Data are expressed as mean ± SD (n = 3), ns: no significance *p < 0.05, **p < 0.01, and ***p < 0.001.

www.advmatinterfaces.de
containing a mixture of 1 mol L −1 sodium hydroxide and 3% by mass of acetylcysteine acid for 4 and 2 h in turn; soaked in a mixture of 2 wt% Tween-20 and 2 wt% acetylcysteine acid in turn and sucrose ester and isopropanol and 2 wt% acetylcysteine for 16 h each; soaking in a second enzyme solution for 12 h; wherein said second enzyme solution comprises: 2% Aggrecanase, 0.5% DNAase, 0.5% RNAase, and 2% papain and water; biofilm obtained by supercritical carbon dioxide cleaning; freeze-drying and irradiation sterilization to obtain the finished product.
Expression and Purification of CBD-bFGF: CBD-bFGF recombinant was produced as described previously. [26a] In brief, the CBD substrate peptide's coding sequence was ligated to bFGF154 complementary DNA via polymerase chain reaction. Afterward, the CBD-bFGF gene was inserted into the pET-28a (+) vector, and the corresponding recombinant protein was expressed in BL21 (DE-3) E. coli. Purified bFGF protein was obtained by nickel column chromatography and desalting.
Synthesis of Biomimetic JCOP@bF: Take the prepared dry JCOP and cut it into any desired shape, the CBD-bFGF was configured with PBS to a certain concentration of protein solution. Place the cut JCOP rough side up flat on a sterile table and add the configured protein solution to the collagen patch in slow drops, The JCOP was immersed in deionized water for 2 h after 30 min of standing. Finally, the wet JCOP is freezedried to obtain JCOP@bF. Sealed and stored before use.
Circular Dichroism Characterization: A certain amount of JCOP-derived collagen was dissolved into 1 mg mL −1 with deionized water to determine the secondary structure of the protein using circular dichroism chromatography, and the sample was diluted in equal gradients to keep the absorption light intensity within a stable range. Deionized water as a reference.
Structural Characterization of JCOP: The samples were described in terms of their morphology and structure. The collagen patch and natural uterus's shape and microstructure were examined using field emission scanning electron microscopy (Quanta 400 FEG).
Porosity Characterization: The measurement was carried out to evaluate the pore structure of dry gels by using a Micromeritics Autopore IV 9500 porosimeter.
Pore Size Analysis: Randomly select a specific size area of the sample, use the SEM software that comes with the range tool to measure the pore size of the selected area, and do homogenization statistics, calculate the average diameter pore size.
Rheology Test: With an 8 mm parallel-plate design, a rheometer was utilized to analyze the rheological characteristics of all collagen hydrogels as prepared. Water was added to the platform surrounding the sample stage during the rheological tests to prevent the gels from drying up and to reduce evaporation. Time-sweep tests were carried out at a frequency of 1 Hz and a temperature of 37 °C over a time range from 0 to 3 min. Frequency-sweep tests were carried out at a stain of 1% and a temperature of 37 °C over a frequency range from 1 to 100 Hz. Stainsweep tests were carried out at a frequency of 1 Hz and a temperature of 37 °C over a stain range from 1% to 100%. Three independent tests for each type of sample (n = 3).
Strength Test of JCOP: Universal material testing machine (Instron, 3400) is used to measure the tensile properties of JCOP, JCOP@bF, and rat uterus. All samples are subjected to 5 parallel experiments and use a standard 10N sensor.
Swelling Ratio: The swelling ratio of the JCOP was investigated using a gravimetric method. Briefly, A certain mass of JCOP (m 0 ) was immersed in PBS (2 mL, pH 7.4) and incubated at 37 °C in an incubator. At indicated time points (1, 3, 5, 7, and 14 days), the hydrogel was taken out and weighed after wiping off all surface water with filter paper (recorded as m t ) The Swelling Ratio 1 00% Release Test of bFGF: An enzyme-linked immunosorbent assay (ELISA) was used to evaluate the kinetics of bFGF release. For this purpose, JCOP@bF was first immersed in 4 mL of PBS at pH 7.4 and placed on a shaker at a constant temperature of 37 °C. The obtained 40 µL of PBS retardant was stored at specific time points, while 40 µL of fresh PBS was added to the system. bFGF release kinetics from JCOP@bF were determined by the ELISA kit (Mutisciences).

www.advmatinterfaces.de
Cytocompatibility: The cytocompatibility of collagen membranes was evaluated by direct cytotoxicity assay. For the research, NIH3T3 cells were obtained from the Chinese Academy of Sciences cell bank and cultivated in DMEM/F-12 media (Gibco, USA) with 10% fetal bovine serum (Gibco, Australia). The material was spread over the bottom of confocal well plates, NH3T3 were incubated on the surface of the scaffolds, and after the cells were plastered, they were cultured for 1, 3, 5, and 7 d. The cells were then stained with Calcein-AM and propidium sodium (PI) (Sigma, USA) for 15 min in the dark, laser confocal scanning fluorescence microscopy collected fluorescence pictures (Nikon A1Rsi, Nikon, Tokyo, Japan). CCK-8 was employed to qualitatively evaluate the cell proliferation rate of the JCOP and JCOP@bF hydrogel extract liquid. Control group: 96-well plate without extract liquid. Microplate readers measured 450 nm absorbance. (Model 550, Bio-Rad, Hercules, CA). All experiments were carried out three times (n = 3).
Tube Formation Assay: Matrix gel (150 µL) was added in advance to each well of a 24-well plate and cured in an incubator at 37 °C for 30 min. 200 µL of leachate was mixed with 100 µL (≈500 000 cells) of resuspended HUVEC cells and added on top of the matrix gel. Plain DMEM-F12 was used as a control. 12 h later the tubes were observed under an electron microscope and photographed to compare the number and length of HUVECs tube junctions.
Animals: Sixty-eight female Sprague Dawley rats weighing 220-250 g were purchased from Shanghai SLAC Laboratory Animal Co., Ltd and kept in standard laboratory conditions with adequate food and water with a 12 h light/dark cycle. The rats were acclimatized for 1 week and vaginal smears were performed daily to determine the stage of the estrous cycle until all rats entered oestrus to begin the experiment.
Uterine Sealing Test: Rats receive anesthesia, the lower abdomen is dissected and the uterus is exposed. To test the sealing ability, two 2-3 mm surgical gaps are drilled with a needle in the horn of the rat's uterus. One side was closed with a JCOP suture and one side was left untreated as a control. To observe the leakage of fluid, blue food dye was infused into the uterus. Observe and take pictures to document the specifics of the leak.
In Vitro Burst Pressure Test: To evaluate the effectiveness of the tissuesealing process, a burst test was conducted using a dynamic flow model. A 2 mm hole was drilled into the uterus. An incision 10 mm in diameter was securely stitched using a JCOP hydrogel. A measuring device was linked to a PBS-filled syringe pump (LSP01-1A, China). Finally, PBS (5 mL min −1 ) was injected into the apparatus, and the highest peak pressure was recorded (MD-S280, MEOKON, China). In all cases, the tests were conducted three times.
Construction of a Rat Uterine Whole-Layer Injury Model: The SD rats were anesthetized with pentobarbital at 30 mg kg −1 body weight, and under aseptic conditions, a longitudinal incision of ≈3 cm was made in the mid-lower abdomen of each rat, and a notch of ≈1.5 cm in length and 0.5 cm in width was made in the middle and lower third of the uterine horns on each side of the rat to form a total defect. The notch was aligned and sutured in the suture group, and JCOP and JCOP@bF groups were each grafted with collagen membrane material of equal size in the original location, and the material was fixed to both sides of the incision with sutures. In the untreated group, the notch was hemostatic and then left untreated. In the sham group, the rats only opened the abdomen without doing any treatment to the uterine horn. The operation was completed by closing the abdominal wall layer by layer, and all animals were given an intraperitoneal injection of penicillin for 3 consecutive days after the operation, and the rats were continued to be kept for subsequent experiments.
Histological Analysis: The rats in each of the five groups (n = 3) were randomly selected and the uterine horn samples were removed from the graft area by suture identification at 30 and 90 days after injury, respectively, followed by gross observation and histological analysis. The damaged area of each uterine horn was fixed with 4% paraformaldehyde for 24 h and paraffin embedded. Paraffin sections were stained with hematoxylin-eosin (H&E) and immunohistochemical staining with anti-CD31 antibody (ab182981,1:2000, Abcam), and the thickness of the myometrium and endometrium was measured using ImageJ software.
The number of CD31-positive neovascularization under at least three high-magnification microscopes was counted, and comparative analysis was performed between different treatment groups.
Immunofluorescent Staining: The rats in each of the five groups (n = 3) were randomly selected and euthanized, the thoracic cavity was opened, perfused with PBS to flush out blood from the whole body, and all the surgical areas of the uterine horns were collected and fixed in 4% paraformaldehyde for 1 day. The specimens were sequentially dehydrated in a gradient of 20% and 30% sucrose solution, embedded in OCT, and made into frozen slices. The primary antibody was an antismooth muscle actin antibody (α-SMA,1;500, Cell Signaling), which was incubated at 4 °C overnight. The secondary antibody was chosen as monkey antirabbit IgG AlexaFluor 488 (A21206,1:500, Invitrogen). Cell nuclei were stained using DAPI (Beyotime) and photographed with laser confocal (Olympus, Japan). The mean immunofluorescence intensity of α-SMA-positive cells in the damaged area and the number of endometrial cells were counted using ImageJ for comparison between groups.
Functional Testing: The function of the repaired uterine injury area was assessed by determining whether it could accept the implantation of blastocysts and whether it could support embryos until late gestation. At 90 days after injury, female rats (n = 16) and males were caged together in a 2:1 ratio, and mating was judged according to the time of appearance of the vaginal plug, which was day 0 of gestation, and the rats were euthanized at mid-to-late gestation (days [16][17][18][19] to check whether embryos were still present in the uterine horns, count the pregnancy rate and the location of embryo implantation were observed.
Statistical Analysis: The majority of data are expressed as mean ± standard deviation. Rates are expressed as percentages. One-way ANOVA was used to analyze the mean among multiple groups. and statistical significance was set at p < 0.05.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.