Bioengineering trends in female reproduction: a systematic review

To provide the optimal milieu for implantation and fetal development, the female reproductive system must orchestrate uterine dynamics in response to ovarian hormones. Specifically, estradiol and progesterone are produced through the processes of follicle development and luteinization in the ovary, and respectively regulate the proliferative and secretory phases in the endometrium. After ovulation, mature oocytes may be fertilized in the fallopian tubes, and the resulting zygote is transported toward the uterus, where it can implant and continue developing (if the endometrium is in an adequately receptive state). Throughout pregnancy, the cervix acts as a physical barrier to protect the fetus from external microorganisms or foreign objects that may enter through the vagina. Fertility can be compromised by pathologies that affect any of these organs or processes, and therefore, being able to accurately model them or restore their function is of paramount importance in applied and translational research. The study of human reproduction requires multidisciplinary approaches. While animal models provide many opportunities for translational discoveries, there are inherent limitations due to differences compared to human reproductive physiology. Similarly, 2D cell or tissue culture models can provide novel insights on aspects of reproductive biology, but these models are more static and simplified and therefore do not recapitulate the dynamic, complex in vivo biology. These limitations underscore the need for continued research along with development of dynamic and more complex in vitro platforms, ex vivo approaches and in vivo therapies. This need is being filled, in part, by rapid advancements in the field of bioengineering, which applies life science and engineering principles to develop biomaterials for restoring, maintaining and/or improving tissue functions. Indeed, bioengineering is leading the way to a new dimension in the study of female reproduction by providing a wide range of potential applications and approaches for discovery. Proposed bioengineering approaches to repair and/or improve female reproductive potential have evolved in parallel with advances in scientific knowledge and technology. Based on our systematic search, current strategies can be classified into six major categories, and these can be applied synergistically to understand reproductive biology and solve related problems: scaffold-free systems, hydrogels, decellularized extracellular matrix (dECM) or polymer scaffolds, 3D bioprinting, organoids and microfluidic approaches. Scaffold-free approaches make use of cells’ ability to self-organize and synthesize their own matrices, generating structures that can be used as functional units or regenerative blocks ( Hayama et al. , 2014 ; Orabi et al. , 2017 ; Kuramoto et al. , 2018 , 2020 ). Hydrogels (which, for the purposes of this review, are defined by their liquid/injectable original state) can include a variety of natural and synthetic components and offer innumerable options for encapsulating or loading drugs, molecules, cells or reproductive tissues ( Zhu et al. , 2016 ; Tavana et al. , 2016a ; Yang et al. , 2021 ; Zhang et al. , 2021b ). Selecting the most suitable hydrogel requires knowing the necessary mechanical and physicochemical properties for a given application ( Kedem et al. , 2011 ; Shikanov et al. , 2011b ). For example, animal-derived hydrogels include commercial mixtures of extracellular matrix (ECM) components, such as Matrigel and Cultrex, which are purified basement membrane extracts secreted by mouse Engelbreth-Holm-Swarm tumor cells. In contrast, dECM scaffolds derive from tissues and organs that were processed by physical, chemical and/or enzymatic methods ( Hellström et al. , 2014 ; Laronda et al. , 2015 ; Campo et al. , 2017 ; Pors et al. , 2019 ; Li et al. , 2021 ; Sargazi et al. , 2021 ; Pennarossa et al. , 2021a ). These biocompatible scaffolds preserve the structure and biochemical milieu of the tissue of origin (in terms of ECM signaling and migration), minimizing the risk of immune rejection after transplantation ( Raya-Rivera et al. , 2014 ; Daryabari et al. , 2019 ; Yao et al. , 2020b ; Padma et al. , 2021b ). Notably, to facilitate transplantation/implementation, these scaffolds are often solubilized and used in hydrogel format ( López-Martínez et al. , 2021a ). Scaffolds can also be produced from other natural polymers (such as collagen and bacterial cellulose) or synthetic components ( Young et al. , 2003 ; Liu et al. , 2007 ; Edwards et al. , 2015 ). Taking the fabrication of cell-loaded or cell-free scaffolds one step further, 3D bioprinting creates materials with precise shapes, textures and porosities, and offers vast applications in regenerative medicine ( Laronda et al. , 2017 ; Souza et al. , 2017 ; Acién et al. , 2019 ; Tiboni et al. , 2021 ; Wu et al. , 2022 ). Among more recent developments are organoids and microfluidics. Organoids are simplified organs or organ-like structures formed in 3D culture systems, which enable recreation of the architecture and physiology of most female reproductive tissues. Organoids provide models for healthy and diseased tissue phenotypes, making them ideal platforms for personalizing bioengineering and biomedicine through both in vitro and in vivo studies ( Kessler et al. , 2015 ; Turco et al. , 2017 ; Lõhmussaar et al. , 2021 ; Oliver et al. , 2021 ). Microfluidic platforms, increasingly referred to as the ‘organ-on-a-chip’ concept, utilize properties of fluid dynamics in small-channelled platforms to facilitate study of the dynamic hormonal cycles and endocrine interactions that characterize the reproductive organs ( Xiao et al. , 2017 ). The majority of bioengineering studies date from the year 2000. However, innovative works from the 20th century built the foundation of this emerging field ( Fig. 1 ). The groundwork for scaffold-free approaches included the first bone marrow transplant between twins ( Thomas et al. , 1959 ), and the generation of cell-sheets ( Yamada et al. , 1990 ) with regenerative potential ( Pellegrini et al. , 1997 ) ( Fig. 1A1 ). Organoids were described as early as the 1960s, when single-cell suspensions completely reconstituted whole organs ( Weiss and Taylor, 1960 ), retinal organoids self-organized in vitro ( Stefanelli et al. , 1961 ) and later, breast ( Li et al. , 1987 ) and alveolar ( Shannon et al. , 1987 ) epithelial cells aggregated to form 3D structures in Matrigel ( Fig. 1A1 ). Key milestones during the 20th century forging the development of the bioengineering field. ( A ) Evidence. ( A1 ) Advances such as the first bone marrow transplant between twins (1) ( Thomas et al. , 1959 ), the control of attachment and detachment of cultured cells (2) ( Yamada et al. , 1990 ) and the use of cell sheets (3) ( Pellegrini et al. , 1997 ) laid the groundwork for scaffold free-approaches. Concomitantly, in 1960, the reconstitution of a complete organ from single-cell suspensions (4) ( Weiss and Taylor, 1960 ) opened an avenue to the present organoids. The in vitro self-organization of retina (5) (Stefannelli et al. , 1961) and the 3D organization of breast (6) ( Li et al. , 1987 ) and alveolar (7) ( Shannon et al. , 1987 ) epithelial cells after culture with Matrigel moved this path further along. ( A2 ) Some works from the 1980s reported the combination of hydrogels with different biological products such as pancreatic islets (8) ( Lim and Sun, 1980 ), E2 (9) ( Embrey et al. , 1980 ) and epithelial cells (10) ( Yannas et al. , 1989 ), introducing these promising biomaterials for regenerative medicine. In parallel, obtaining ECM from renal glomeruli (11) ( Hjelle et al. , 1979 ), from liver connective tissue (12) ( Rojkind et al. , 1980 ), and a decade later, an intact acellular matrix from intestinal submucosa (13) ( Badylak et al. , 1995 ) and bladder (14) ( Chen et al. , 1999 ) provided the beginnings of the dECM scaffold approaches. ( A3 ) The beginnings of co-culture systems are captured in two main works in which embryos were cultured together with trophoblastic vesicles (15) ( Camous et al. , 1984 ) and ampullary cells (16) ( Bongso et al. , 1989 ). Research that formed the basis of microfluidic systems was reported in the nineties; some examples are the emergence of on-chip capillary electrophoresis (17) ( Harrison et al. , 1993 ) and elastomeric microchannel networks for cell culture (18) ( Folch and Toner, 1998 ). Works from the end of the century paved the way for bioprinting: creation of a tissue-engineered ear (19) ( Cao et al. , 1997 ), use of 3D printed substrates for cell adhesion (20) ( Park et al. , 1998 ) and introduction of soft lithography (21) ( Xia and Whitesides, 1998 ). ( B ) Applications. The establishment of a capillary system for sperm samples (22) ( Ulstein, 1972 ) and the culture of human ovarian epithelial organoids (23) ( Kruk and Auersperg, 1992 ) were the beginnings of the development of in vitro screening platforms. The next generation in vitro platforms are based on studies like those from 1986 and 1988, which established endometrial epithelial cells were co-cultured with an ECM from glandular structures (24, 25) ( Kirk and Alvarez, 1986 ; Rinehart et al. , 1988 ) and a similar system also containing endometrial stromal cells (26) ( Bentin-Ley et al. , 1994 ). Finally, the development of the ESTES technique for dog ovarian transplantation (27) ( Estes, 1909 ) in the early 20th century provided an excellent basis for a later dog uterus replantation (28) ( Eraslan et al. , 1966 ), a rabbit fallopian tube and ovary autograft transplantation (29) ( Winston and Browne, 1974 ) and a primate ovarian transplantation (30) ( Scott et al. , 1981 ). BM, bone marrow; E2, estradiol; ECM, extracellular matrix; dECM, decellularized extracellular matrix. Explorations in the 1980s and 1990s produced different types of in vitro co-culture systems ( Fig. 1A2 and 3 ). In particular, the successful combination of hydrogels with different biological products, such as pancreatic islets ( Lim and Sun, 1980 ), prostaglandins ( Embrey et al. , 1980 ) and epithelial cells ( Yannas et al. , 1989 ), encouraged the use of different biomaterials for regenerative medicine. In this regard, studies in which embryos were cultured together with trophoblastic vesicles ( Camous et al. , 1984 ) or ampullary cells ( Bongso et al. , 1989 ) inspired other co-culture systems. On the other hand, dECM scaffolds appeared after ECM was obtained from murine renal glomeruli ( Hjelle et al. , 1979 ), liver connective tissue ( Rojkind et al. , 1980 ), intact acellular matrix from porcine intestinal submucosa ( Badylak et al. , 1995 ) and bladder ( Chen et al. , 1999 ) ( Fig. 1A2 ). Microfluidic platforms also emerged with micromachining capillary electrophoresis ( Harrison et al. , 1993 ), and microchannel networks for cell culture ( Folch and Toner, 1998 ). Finally, bioprinting gained popularity with the first tissue-engineered ear ( Cao et al. , 1997 ), the use of 3D-printed substrates for cell adhesion ( Park et al. , 1998 ) and introduction of soft lithography ( Xia and Whitesides, 1998 ); the latter encompasses a group of techniques for fabricating or replicating structures, channels or membranes by using soft polymeric material (usually polydimethylsiloxane) stamps or molds ( Kim et al. , 2018 ) ( Fig. 1A3 ). PRISMA flow diagram. Exact terms used for each of the database searches are detailed in Supplementary Table SI . Template adapted from Page et al. (2021) . Created with BioRender.com. Organ-level overview of the bioengineering studies carried out between January 2000 and September 2021 and included in this systematic review . The studies involved the uterus, ovaries, fallopian tubes and cervix/vagina. The numbers reflect the number of studies included in Table I and Supplementary Table SIV . Created with BioRender.com. These six categories of bioengineering strategies promote four main translational and/or clinical applications: the development of next-generation in vitro platforms, or representative in vitro toxicology and drug screening models; the discovery of alternative therapies or new biomarkers; and improvement of tissue/organ regeneration and/or transplantation protocols ( Fig. 1B ). The establishment of a capillary system for sperm samples ( Ulstein, 1972 ) is an excellent example of an innovative platform to improve ART, while the in vitro culture of human endometrial 3D glandular structures ( Kirk and Alvarez, 1986 ; Rinehart et al. , 1988 ), endometrial stromal cells embedded in a collagen matrix [and covered with epithelial cells ( Bentin-Ley et al. , 1994 )] and ovarian epithelial organoids ( Kruk and Auersperg, 1992 ) ensured the initial steps towards personalized in vitro screening platforms. Finally, the early development of the ESTES technique, where a portion of the ovary is transplanted into the uterus ( Estes, 1909 ), provided a foundation for later progress in reproductive organ transplantation ( Eraslan et al. , 1966 ; Winston and Browne, 1974 ; Scott et al. , 1981 ). Since these initial discoveries paved the way, the bioengineering field has undergone rapid growth and expansion. Many engineered reproductive tissues and platforms are currently in different stages of clinical development; most models remain experimental, but others are in pre-clinical trials, and some are already being applied clinically. Given the quantity and heterogeneity of studies published within this specialty, the goal of this review was to systematically summarize the mounting evidence on bioengineering strategies, platforms and therapies, both currently available and under development, in the context of female reproductive medicine, including novel alternatives for fertility restoration.

PubMed and Embase were searched for relevant reports. The search strategy was limited to full-text articles, published in English, involving mammals or material derived therefrom, between January 2000 and September 2021. Combinations of the following keywords were used: bioengineering, reproduction, artificial, biomaterial, microfluidic, bioprinting organoid, hydrogel, scaffold, uterus, endometrium, ovary, fallopian tubes, oviduct, cervix, vagina, endometriosis, adenomyosis, uterine fibroids, chlamydia, Asherman’s syndrome (AS), intrauterine adhesions, uterine polyps, polycystic ovary syndrome and primary ovarian insufficiency. Specific queries used in each database are presented in Supplementary Table SI . Additional studies were identified by manually searching the references of the selected articles and of complementary reviews. Literature search results were exported to an MS Excel spreadsheet and duplicates were identified using electronic and manual methods ( Fig. 2 ). Titles, abstracts and full texts were then screened independently and in duplicate by two authors (E.F.-H. and R.L.) using the following eligibility criteria: original, rigorous and accessible peer-reviewed work published in English, on female reproductive bioengineering techniques in preclinical ( in vitro / in vivo/ex vivo ) and/or clinical testing phases. Studies in which gels were developed for intravaginal delivery of hormones, bactericides, nucleic acids or contraceptive drugs were not considered in this review because of their pharmacological nature. Questions or disagreements were resolved by discussion (E.F.-H., R.L., A.P. and I.C.). The final list of included studies was approved by I.C. Extracted data, including titles, authors, year of publication, reproductive organ (uterus, ovary, fallopian tube, cervix, vagina or full tract), bioengineering strategy, platform/biomaterial used, species, cell/tissue model, study type ( in vitro , in/ex vivo , clinical) and main findings were compiled into a shared Google Sheets spreadsheet and revised by M.H., L.M.-G., S.H. and M.B. Relevant findings extracted from each study are summarized in Table I . Due to the inability to completely detail the many articles comprising this systematic review in Table I , a comparison of in vivo uterine regeneration parameters (e.g. immune tolerance, recovery of thickness and muscle layer, presence of glands, angiogenesis, implantation potential and maintenance of pregnancy) is provided in Supplementary Table SII , while specific outcomes of in vitro follicle growth (IVFG) studies (e.g. follicle survival, initial and final follicle size, steroidogenesis, oocyte maturation rates, developmental competence and/or fertility restoration) are detailed in Supplementary Table SIII . Main findings of bioengineering studies related to female reproductive organs. Description of the main bioengineering findings in the female reproductive system in the last 21 years based on different strategies, platforms/biomaterials, type of study (including in vivo models) and gynaecological related-diseases. To note: model in type of study column only refers to in vivo approach; studies with human cells/tissues are marked in bold, while clinical studies are highlighted in yellow; and pill icons indicate studies carried out with patients, animal models and biological samples with female reproduction-related diseases (established cell lines have not been taken into account). 3βHSD, 3β-Hydroxysteroid dehydrogenase; AD-MSC, adipose-derived mesenchymal stem cell; AMH, anti-Müllerian hormone; AS, Asherman’s syndrome; bFGF, basic fibroblast growth factor; BM-MSC; bone marrow-derived mesenchymal stem cell; BMP, bone morphogenic protein; C. albicans: Candida albicans ; C. trachomatis: Chlamydia trachomatis; CD, cluster of differentiation; COC, cumulus-oocyte complex; Collplant, human recombinant virgin collagen bioengineered in tobacco plant lines; CP, clinical pregnancy; Cx37/43, connexin 37/43; CYR61: cysteine-rich angiogenic inducer 61; DAZL, deleted in azoospermia like; DC, decellularized; dECM, decellularized extracellular matrix; dEMSC, decidualized endometrial stromal cells; DKK, Dickkopf WNT Signaling Pathway Inhibitor 1; E2, estradiol; ECM, extracellular matrix; EEPC, endometrial epithelial progenitor cell; eMSC, endometrial mesenchymal stem cell; EnSC, endometrial stem cell; ESC, embryonic stem cell; ESC-MPC, embryonic stem cell-derived mesenchymal progenitor cell; ESF, endometrial stromal fibroblast; ET, embryo transfer; EVT, extravillous trophoblast; F_/T_, fibrinogen (mg/ml)/thrombin (IU/ml) concentration ratio; FA, fibrin-alginate; FA-IPN, fibrin-alginate interpenetrating network; FOXO1/3a, forkhead box protein O1/3a; FT-MSC, Fallopian tube mesenchymal stem cell; FTE, Fallopian tube epithelium; GC, granulosa cell; GDF9, growth differentiation factor 9; GelMA, gelatin-methacryloyl; GVBD, germinal vesicle breakdown; H9-ESC: human embryonic stem cell-9 line; HA, hyaluronic acid; HBP, heparin binding protein; HP, heparin poloxamer; hrVit, human recombinant vitronectin; HUVEC, human umbilical vein endothelial cell; IGF, insulin-like growth factor; iPSC, induced pluripotent stem cell; IUA, intrauterine adhesion; Jag1, jagged canonical Notch ligand 1; JAR spheroids: human choriocarcinoma (JAR) cells grown as multicellular spheroids; KHDC3, KH domain containing 3 like; LB, live birth; Lhx8, LIM homeobox 8; LIF, leukemia inhibitory factor; MEF, mouse embryonic fibroblast; MenMSC, menstrual blood mesenchymal stem cell; MH, magnesium hydroxide; MII, metaphase II; MRKH, Mayer-Rokitansky-Küster-Hauser; MSC, mesenchymal stem cell; Nanos3, Nanos C2HC-Type Zinc Finger 3; NLRP5, NLR Family pyrin domain containing 5; ORMOCER, Photosensitive organic-inorganic hybrid polymer (ORganically MOdified CERamics); OSE, ovarian surface epithelium; OT, ovarian tissue; P4, progesterone; PD-MSC, placenta-derived mesenchymal stem cell; PDMS, polydimethylsiloxane; PEBP, poly(ethylene glycol)-b-poly( l -phenylalanine); PEG, 4-polymeric poly-(ethylene glycol); PEG-PLA, polymeric poly(ethylene glycol)-block-polylactide methyl ether; PEG-VS, 4-arm poly-(ethylene glycol) tetravinyl sulfone; PGC, primordial germ cell; PLA, poly- l -lactic acid; PLGA, poly( d , l -lactide-co-glycolide); PLO, poly- l -ornithine; POF, premature ovarian failure; POI, premature ovarian insufficiency; PSC-ESF, pluripotent stem cells induced into endometrial stromal fibroblast; PTFE, polytetrafluoroethylene fluoropolymers; PVA/CMC, polyvinyl alcohol-carboxymethylcellulose; QD, quantum dots; RGF-BME, reduced growth factor-basement membrane extract; ROS, reactive oxygen species; SC, subcutaneous; SDS-T-A, sodium dodecyl sulfate-Triton-Ammonium; SIS, small intestine submucosa; SU8, negative photo-resistor based on epoxy components; TC, theca cell; TFAM, transcription factor A (mitochondrial); TGFβR2, transforming growth factor beta receptor 2; tHESC, hTERT-immortalized human endometrial stromal cell; TiO 2 , titanium dioxide; UC-MSC, umbilical cord-derived mesenchymal stem cell; VEGF, vascular endothelial growth factor; VP-MSC, visceral peritoneum mesenchymal stem cell. Studies related to gynecological pathologies, both included in the initial search terms and different ones addressed by the selected articles (such as endometriosis, uterine fibroids, AS, intrauterine adhesions, polycystic ovary syndrome, primary ovarian insufficiency and Mayer-Rokitansky-Küster-Hauser (MRKH) syndrome) were included in Table I , however owing to the extent of relevant studies applying bioengineering techniques to create novel ovarian, uterine and cervical cancer models (published between April 2014 and September 2021), the latter were grouped separately according to their application and organ in Supplementary Table SIV . Finally, throughout the entire review we classify hydrogels as originally softer/injectable materials regardless of whether they gelify afterwards (e.g. collagen solutions), and scaffolds as their more rigid counterparts (e.g. collagen membranes).

The search queries yielded 10 390 results (from a total of 18 748 titles identified) after removal of duplicates. Titles and abstracts were screened for eligibility (based on exclusion criteria presented in Fig. 2 ) and 584 (5.6%) full-text papers were retrieved for detailed assessment. An additional 24 studies were retrieved from manual searching of citations. We classified studies by bioengineering strategy within each organ ( Fig. 3 ), and by their main application(s). Finally, 312 articles were included for systematic review, including 18 (5.8%) studies on scaffold-free approaches, 124 (39.7%) discussing hydrogels, 87 (27.9%) related to dECM and polymer scaffolds, 10 (3.2%) using bioprinting, 45 (14.4%) implementing organoid cultures and 28 (8.9%) exploring microfluidic techniques ( Table I and Supplementary Table SIV ). The majority of the bioengineering studies included in vitro / ex vivo or in vivo work using animal models, with only a few studies reaching clinical stages (seven uterine, two ovarian and seven cervical). Below we summarize studies using the six bioengineering techniques in work related to female reproductive organs. This section mainly contemplates studies based on the non-matrix-assisted self-organizing capacity of cells to generate multicellular entities. Six studies presented scaffold-free approaches applied to bioengineering of the uterus and its tissues, including four in vivo studies based on cell sheets ( Kuramoto et al. , 2015 , 2018 , 2020 ; Sun et al. , 2018 ) and two in vitro investigations based on the generation and study of epithelial and stromal organoids ( Murphy et al. , 2019 ; Wiwatpanit et al. , 2020 ). Of the eight studies involving scaffold-free approaches in ovary, one used a cell-based method involving primordial germ cells to generate ovarian tissue (OT) in vivo ( Hayama et al. , 2014 ), four developed self-assembled spheroids ( Krotz et al. , 2010 ; Chowanadisai et al., 2016 ; Kim et al. , 2018 ; Ward Rashidi et al. , 2019 ), while toxicity of Qdot 655 ITK carboxyl quantum dots ( Xu et al. , 2016 ) and nanoparticles made with chitosan ( Raja et al. , 2021 ), polyethylene glycol (PEG) and polylactic acid or titanium dioxide ( Scsukova et al. , 2020 ), were tested in preclinical and ex vivo models, respectively. Scaffold-free approaches to develop bioengineered vaginal and cervical tissues were included in four studies, which applied self-assembled vaginal constructs in vivo ( Orabi et al. , 2017 ; Jakubowska et al. , 2020 ) or air–liquid interface techniques to generate vaginal and cervical in vitro models ( De Gregorio et al. , 2017 ; Zhu et al. , 2017 ). This review unveiled a plethora of different hydrogel-based research involving uterine cells or tissues. Of the 40 studies compiled, 65%, 30% and 5% used natural, synthetic or hybrid hydrogels, respectively. The most commonly used natural hydrogels were based on collagen (46.2%), Matrigel (24.4%) and hyaluronic acid (HA; 15.4%). Synthetic hydrogels were based predominantly on PEG (38.5%) and poloxamer (30.7%). Studies applied these approaches for in vitro modeling of disease, tissue cross-talk and differentiation ( Schutte et al. , 2015 ; Cook et al. , 2017 ; Pence et al. , 2017 ; Stejskalová et al. , 2021 ), in vivo regeneration of the endometrium and myometrium ( Lin et al. , 2012 ; Yang et al. , 2017 ; Li et al. , 2019 ; Yoon et al. , 2021 ; Lin et al. , 2021a ) and the treatment of intrauterine adhesions ( Müller et al. , 2011 ; Liu et al. , 2020 ). Encapsulation of follicles and tissue fragments is the most exploited hydrogel-based application in ovarian bioengineering. Alginate (alone or in combination with other materials) is the most commonly implemented hydrogel, with use in 47.14% of studies evaluating hydrogels for in vitro cultures of murine ( Xu et al. , 2006 , 2009a ; West et al. , 2007 ; West-Farrell et al. , 2009 ; Jackson et al. , 2009 ; Jin et al. , 2010 ; Parrish et al. , 2011 ; Tagler et al. , 2013 , 2014 ; Skory et al. , 2015 ; Zhang et al. , 2019c ), canine ( Songsasen et al. , 2011 ), caprine ( Brito et al. , 2014 ), ovine ( Mastrorocco et al. , 2020 ), primate ( Xu et al. , 2009c ) and human ( Amorim et al. , 2009 ; Camboni et al. , 2013 ; Laronda et al. , 2014 ; Skory et al. , 2015 ; Xiao et al. , 2015 ) ovarian follicles and tissue. Natural and synthetic hydrogels such as laminin ( Hao et al. , 2020 ), collagen ( Joo et al. , 2016 ), fibrin ( Smith et al. , 2014 ; Paulini et al. , 2016 ), Matrigel ( Xu et al. , 2009b ; Kedem et al. , 2011 ; Ghezelayagh et al. , 2021 ), VitroGel ( Kim et al. , 2020a ), HA ( Desai et al. , 2012 ) or PEG ( Shikanov et al. , 2011a ; Lerer-Serfaty et al. , 2013 ; Tomaszewski et al. , 2021 ) are less frequently reported. Twenty-four studies used hydrogels in preclinical models of IVF ( Xu et al. , 2006 ), allo-/xenotransplantation of ovarian cells, follicles and fragments ( Vanacker et al. , 2012 ) or ovarian function restoration ( Su et al. , 2016 ; Ding et al. , 2018 ; Yang et al. , 2019 ; Yoon et al. , 2021 ) ( Table I ); seven applied hydrogels in oncological modeling and drug testing ( Supplementary Table SIV ). Tissue-specific hydrogels based on rabbit oviductal dECM and alginate resulted, respectively, in in vitro models of embryo culture ( Francés-Herrero et al. , 2021a ) and crosstalk between human epithelium and murine follicles ( Zhu et al. , 2016 ), or ex vivo models of the human fallopian tube fimbriae ( Eddie et al. , 2015 ). Hydrogels based on silk-HA, collagen derivatives and chitosan could recreate an ex vivo pregnant-like human cervical model ( Raia et al. , 2020 ), or treat vaginal atrophy in vivo ( You et al. , 2020 ; Yang et al. , 2021 ). Forty-five uterine-related studies evaluated dECM and polymer scaffolds in cytocompatibility and in vitro modeling experiments, as well as in vivo regeneration and anti-adhesions tests ( Table I ). Those based on decellularized (DC) matrices accounted for 46.6% of reports, while 33.3% and 20% used scaffolds based on purified natural or artificial polymers, respectively. Various studies isolated and evaluated tissue-specific ECM from whole uteri of rats ( Hellström et al. , 2014 , 2016 ; Miyazaki and Maruyama, 2014 ; Santoso et al. , 2014 ; Miki et al. , 2019 ; Padma et al. , 2021a , b ), mice ( Hiraoka et al. , 2016 ), pigs ( Campo et al. , 2017 ), rabbits ( Yao et al. , 2020b ) and sheep ( Daryabari et al. , 2019 ; Tiemann et al. , 2020 ; Padma et al. , 2021c ); from rabbit ( Campo et al. , 2019 ), pig ( López-Martínez et al. , 2021a , b ) and human endometrium ( Olalekan et al. , 2017 ; Sargazi et al. , 2021 ); and from rat and human myometrium ( Young and Goloman, 2013 ). Natural polymer scaffolds included alginate ( Li et al. , 2011b ; Stern-Tal et al. , 2020 ), HA ( Demirbag et al. , 2005 ; Liu et al. , 2005 ) and gelatin-coated ( Su et al. , 2014 ; Edwards et al. , 2015 ; Cai et al. , 2019 ) materials, while artificial polymer scaffolds were created with polyglactin ( Young et al. , 2003 ), polylactide ( Pensabene et al. , 2015 ; Mukherjee et al. , 2019 ), polyglycolic acid ( Magalhaes et al. , 2020 ), emulsion-templated highly porous materials ( Eissa et al. , 2018 ; Richardson et al. , 2019 ) and polytetrafluoroethylene fluoropolymers ( Kuperman et al. , 2020 ). In contrast, only dECM or natural polymers were used for 24 studies involving ovarian research. Several reports described the generation and in vitro / in vivo biocompatibility of DC murine ( Alshaikh et al. , 2019 , 2020 ), porcine ( Liu et al. , 2017 ; Pennarossa et al. , 2020 , 2021a , b ), ovine ( Eivazkhani et al. , 2019 ), bovine ( Laronda et al. , 2015 ; Nikniaz et al. , 2021 ) and human ( Hassanpour et al. , 2018 ; Pors et al. , 2019 ; Sistani et al. , 2021 ) ovaries, as well as porcine small intestine submucosa ( Celik et al. , 2009 ; Abir et al. , 2020 ) or human amniotic membrane ( Motamed et al. , 2017 ). Of the 18 studies that assessed scaffolds for cervicovaginal applications, 33.3% were clinical studies of vaginal reconstruction using oxidized cellulose ( Dadhwal et al. , 2010 ), bovine collagen matrices ( Noguchi et al. , 2004 ; Guerette et al. , 2009 ), human acellular dermis ( Zhang et al. , 2017c ) or porcine small intestine submucosa ( Raya-Rivera et al. , 2014 ; Zhang et al. , 2019b ). The remaining studies consisted of either in vitro cervico-vagina models based on DC porcine vaginas ( Greco et al. , 2018 ), human cervical dECM ( McKinnon et al. , 2020 ) and scaffolds of collagen ( Hjelm et al. , 2010 ), alginate/chitosan ( Tentor et al. , 2020 ), silk ( House et al. , 2018 ), Alvetex ( Arslan et al. , 2015 ) and polylactide ( Roman et al. , 2018 ) or preclinical models of vaginal repair ( Maisel et al. , 2016 ; Zhang et al. , 2017a ; Ye et al. , 2020 ; Ma et al. , 2021 ). The scarcity of bioprinting implementation in reproductive studies reflects the novelty of this technique. Two reports applied 3D bioprinting technology to in vitro studies of uterine contractility ( Souza et al. , 2017 ) and in vivo endometrial regeneration using a gelatin-alginate hydrogel loaded with human induced pluripotent stem cell (iPSC)-derived mesenchymal stem cells (MSCs) ( Ji et al. , 2020 ). Similarly, three studies applied bioprinting with natural polymers for in vitro ovarian modeling ( Ovsianikov et al. , 2007 ), follicle culture ( Wu et al. , 2022 ) or fertility restoration in a preclinical model ( Laronda et al. , 2017 ). Finally, five groups bioprinted bactericidal vaginal rings ( Tiboni et al. , 2021 ), vaginal tissue with acellular bioink ( Hou et al. , 2021 ), PACIENA prosthesis for vaginoplasties ( Acién et al. , 2019 ), cervical implants ( Zhao et al. , 2020 ) or 3D models of cervical cancer ( Gospodinova et al. , 2021 ). We identified 15 studies generating endometrial organoids. Matrigel was primarily used as a supportive matrix, with ( Francés-Herrero et al. , 2021b ) or without (78.5% of studies) endometrial dECM supplementation. Other studies used collagen ( Abbas et al. , 2020 ), or functionalized PEG matrices ( Hernandez-Gordillo et al. , 2020 ). Remarkably, in six studies, epithelial and stromal components were combined to create organoids ( Murphy et al. , 2019 ; Wiwatpanit et al. , 2020 ; Jiang et al. , 2021 ) or assembloids ( Bläuer et al. , 2005 ; Abbas et al. , 2020 ; Rawlings et al. , 2021 ), which simultaneously represent both endometrial populations. Ovarian spheroids or organoids were derived in 18 studies using Matrigel, Cultrex or agarose, together with diverse cell types, including ovarian epithelial cells ( Kwong et al. , 2009 ; Kopper et al. , 2019 ), granulosa and theca cells ( Yoon et al. , 2021 ) or embryonic gonads ( Oliver et al. , 2021 ). Remarkably, 89% of the studies generated ovarian organoids, derived from patients with cancer, that reliably mimicked the original pathology ( Maru et al. , 2019 ). Similarly, in five studies, human organoids were derived from the squamocolumnar junction region of the uterine cervix ( Maru et al. , 2020 ) or ecto- and endocervical tissue using Matrigel or Cultrex ( Lechanteur et al. , 2017 ; Chumduri et al. , 2021 ; Lõhmussaar et al. , 2021 ; Tanaka et al. , 2021 ). Fallopian tube organoids were generated from murine ( Xie et al. , 2018 ) and human ( Kessler et al. , 2015 ; Rose et al. , 2020 ; Zhang et al. , 2021a ) fallopian tube epithelial cells used alone or co-cultured with umbilical vein endothelial cells and fallopian tube-derived MSCs ( Chang et al. , 2020 ), and from human iPSC ( Yucer et al. , 2017 ) or murine fallopian tube epithelial stem cells ( Lin et al. , 2021b ). Culture was supported by Matrigel (in five of these six studies) or Mebiol ( Lin et al. , 2021b ). Organoid studies accounted for 58%, 13% and 13% of bioengineering platforms reported in the fallopian tubes, ovary and uterus, respectively. Microfluidic platforms can be used to model static, passive (created by gravity) or active dynamic (with a determined flow rate) conditions. Six studies applied microfluidics to create advanced in vitro models of the human endometrium ( Astolfi et al. , 2016 ; Gnecco et al. , 2017 , 2019 ; Ahn et al. , 2021 ; De Bem et al. , 2021 ), or co-culture embryos with endometrial cells ( Kimura et al. , 2009 ). Of the 13 microfluidic-based studies in the ovarian field, two applied dynamic follicle culture in alginate and/or collagen matrices ( Choi et al. , 2014 ; Nagashima et al. , 2018 ), while two studies evaluated the mechanical effect of flow on oocyte denudation and maturation ( Sadeghzadeh Oskouei et al. , 2016 ; Weng et al. , 2018 ). The rest of the studies sought to recreate oncological models, evaluate drug effectiveness or elucidate therapeutic targets (as we discuss in detail in the applications sections). Finally, four studies applied microfluidics to model the human cervical epithelial layer ( Lin et al. , 2017 ; Aziz et al. , 2020 ; Yang et al. , 2020 ; Tantengco et al. , 2021 ) and bovine ( Ferraz et al. , 2018 ) and canine ( Ferraz et al. , 2020 ) oviducts, while one study modeled human uterine and ovarian endocrine crosstalk ( Park et al. , 2020 ), and two studies recreated a complete female tract in microfluidic systems, combining endometrial, ovarian, oviductal, cervix and liver tissue to model the hormonal profile of a menstrual cycle ( Xiao et al. , 2017 ) or to manipulate, fertilize and culture embryos in a single device ( Han et al. , 2010 ). Bioengineering approaches can elucidate the normo- and patho-physiology of female reproductive organs by developing next-generation in vitro/ex vivo platforms, creating representative models for toxicology/drug screening, developing alternative therapeutic strategies, discovering new biomarkers and improving tissue/organ regeneration and/or transplantation protocols. The creation of in vitro platforms that faithfully reproduce the physiological and pathological states of higher organisms is of paramount importance in applied and translational research. Bioengineering platforms have provided novel 3D models of follicle culture (see citations in Hydrogels section), human embryo implantation ( Wang et al. , 2012 , 2013 ; Buck et al. , 2015 ; Stern-Tal et al. , 2020 ; Rawlings et al. , 2021 ), a three-layered endometrium that remained functional for 28 days ( Park et al. , 2021 ), endometrial cancer invasion ( Park et al. , 2003 ) and wound healing ( Stavreus-Evers et al. , 2003 ), as well as bidirectional crosstalk between the uterus and the ovaries ( Park et al. , 2020 ). Moreover, collagen scaffolds loaded with human epithelial and endothelial cells ( Pence et al. , 2015 ) or tissue slices ( Muruganandan et al. , 2020 ) respond to ovarian hormones, while collagen-embedded human stromal cells demonstrate decidualization changes ( Schutte and Taylor, 2012 ) and contractile ability ( Kim et al. , 2020b ). Notably, endometrial cells encapsulated in a PEG hydrogel with ECM-binding peptides remodeled the synthetic matrix and displayed hormone-mediated differentiation ( Cook et al. , 2017 ). In vitro follicle growth produced developmentally competent murine oocytes ( Xu et al. , 2006 ; Ahn et al. , 2015 ) that led to live births (LBs) after embryo transfer ( Xu et al. , 2006 ). Most studies (94%) implemented individual follicle culture, while others successfully demonstrated that culturing multiple follicles together substantially improves follicle survival (>80% versus <29% with individual culture) and oocyte maturation ( Hornick et al. , 2013 ; Brito et al. , 2016 ). Interestingly, IVFG benefits from the addition of ECM sequestering peptides ( Tomaszewski et al. , 2021 ), ascorbic acid [which increases expression of ECM and cell adhesion molecules ( Tagler et al. , 2014 )], bone morphogenetic protein 4 ( Felder et al. , 2019 ), mouse embryonic fibroblasts ( Tagler et al. , 2012 ), ovarian cells ( Jamalzaei et al. , 2020 ) and human menstrual blood MSCs ( Rajabi et al. , 2018 ), but not denuded oocytes, oocyte-secreted factors, granulosa cells ( Hornick et al. , 2013 ) or leukaemia inhibitory factor ( Younis et al. , 2017 ). VitroGel, a novel animal-origin free hydrogel, also improves IVFG parameters, outperforming alginate and producing competent oocytes in a recent study by Kim et al. (2020a) . Furthermore, when used exclusively, alginate concentrations ranged between 0.25% and 3% ( Supplementary Table SIII ); however, as a fibrin-alginate interpenetrating network, the concentration of alginate can be reduced below 0.25%, providing a more realistic environment for follicle growth and improving oocyte maturation ( Shikanov et al. , 2009 , 2011b ). Notably, combinations of 25 mg/ml fibrinogen and 4 IU/ml thrombin or 12.5 mg/ml fibrinogen and 1 IU/ml thrombin, are suggested as the best scaffolds for human ovarian stromal cells in vitro ( Luyckx et al. , 2013 ) and for murine follicle development in vivo ( Luyckx et al. , 2014 ), while 50 mg/ml fibrinogen and 50 IU/ml thrombin best mimics the rigidity of the native human ovarian cortex ( Chiti et al. , 2018 ). DC models also offer important advances by maintaining unique tissue-specific ECM milieus, not only providing the most realistic scaffold for each organ’s endogenous cell types, but also remarkably acting as a biocompatible framework for cells from other tissues/species. For example, DC mouse uterine tissue is an adequate natural niche for human menstrual blood MSC differentiation toward uterus-specific cell lineages ( Arezoo et al. , 2021 ), DC sheep uterus stimulates rat fetal dorsal root ganglion regeneration and angiogenesis during chicken embryo development ( Padma et al. , 2021c ) and solubilized porcine endometrial dECM enhances proliferation rates of human endometrial organoids ( Francés-Herrero et al. , 2021b ). The generation and increasing use of organoids are revolutionizing the field of reproductive medicine. Among other limitations, the primarily epithelial nature of these structures is noteworthy. To date, several investigations have already provided models of multicompartment tissues [i.e. endometrial and stromal compartments ( Murphy et al. , 2019 ; Rawlings et al. , 2021 )], ecto- and endo-cervical epithelial regions ( Maru et al. , 2020 ; Chumduri et al. , 2021 ; Lõhmussaar et al. , 2021 ) and heterogeneous tumors (see next paragraph and/or Supplementary Table SIV ), in addition to research models for chlamydia ( Bishop et al. , 2020 ) and herpes ( Zhu et al. , 2017 ) infections. Remarkably, organoids are able to reproduce specific uterine ( Boretto et al. , 2019 ; Bishop et al. , 2020 ; Hernandez-Gordillo et al. , 2020 ; Luddi et al. , 2020 ; Marinić et al. , 2020 ), ovarian (described in detail below) and cervical ( Karolina Zuk et al. , 2017 ; Maru et al. , 2020 ; Lõhmussaar et al. , 2021 ) tissue phenotypes, as well as respond to hormones ( Bläuer et al. , 2005 ; Boretto et al. , 2017 ; Turco et al. , 2017 ; Wiwatpanit et al. , 2020 ; Cheung et al. , 2021 ). These models can be established from patient biopsies ( Maru et al. , 2019 ; Lõhmussaar et al. , 2021 ), biological fluids ( Cindrova-Davies et al. , 2021 ) or cell lines (e.g. SKOV3, H08910 , OVCAR3/4/8 used in oncological studies listed in Supplementary Table SIV ). Further transplantation of spheroids or organoids may restore ovarian function ( Kim et al. , 2018 ) or promote endometrial regeneration ( Jiang et al. , 2021 ). Next-generation platforms for oncological studies include the development of 3D ovarian cancer models using scaffolds of bacterial cellulose with chitosan ( Ul-Islam et al. , 2019 ), collagen ( Zheng et al. , 2015 ), poly- dl -lactide-coglycolide-PEG ( Zhou et al. , 2018 ) or RADA16-I peptide hydrogel ( Song et al. , 2020 ). Similarly, a novel 3D cervical cancer model was created with 3D-printing, using bioinks mixed with sodium alginate ( Gospodinova et al. , 2021 ). Dynamics of cancer progression can be modeled ex vivo in 3D ( Ajeti et al. , 2017 ; Fleszar et al. , 2018 ; Loessner et al. , 2019 ; Flont et al. , 2020 ; Fan et al. , 2021 ), utilizing multilayered microfluidic systems ( Lin et al. , 2017 ; Flont et al. , 2020 ) and ovarian spheroids [to study macromolecular crowding ( Bascetin et al. , 2021 )]. Unique ex vivo and in vitro proof of concept applications include, DC bovine ovarian and uterine ‘tissue papers’ ( Jakus et al. , 2017 ), an in vitro artificial human ovary ( Krotz et al. , 2010 ), a pregnant-like cervix ( Raia et al. , 2020 ), an endocervical model that responds to hormones during a 28-day cycle ( Arslan et al. , 2015 ), automated and reliable oocyte denudation on a chip ( Weng et al. , 2018 ) and the EVATAR platform that models the dynamics of the human menstrual cycle ( Xiao et al. , 2017 ). Bioengineered in vitro platforms enable evaluation of the biocompatibility of biomaterials ( Xu et al. , 2016 ; Scsukova et al. , 2020 ), effects of chemical toxicants [such as doxorubicin ( Zhou et al. , 2015 ; Aziz et al. , 2020 ), or dioxin ( Park et al. , 2020 )] or response to cancer therapies ( Supplementary Table SIV ). For example, 3D tumor models in ring format currently support automated and rapid personalized drug screening ( Phan et al. , 2019 ). Other drug screening models include microdissected tumor tissues in microfluidic culture ( Astolfi et al. , 2016 ) or alginate hydrogels ( Salas et al. , 2020 ), organoids of small cell neuroendocrine carcinoma of the uterine cervix ( Tanaka et al. , 2021 ) and ovarian cancer organoids, which have proven to be excellent models to test chemotherapy drugs ( Maru et al. , 2019 ; de Witte et al. , 2020 ; Maenhoudt et al. , 2020 ). In fact, since endometrial and ovarian organoids can be derived from each patient’s biopsies ( Kopper et al. , 2019 ; Nanki et al. , 2020 ; Bi et al. , 2021 ; Chen et al. , 2021 ; Espedal et al. , 2021 ), they reflect specific tumor heterogeneity and are ideal for drug pre-screening and the development of personalized treatment regimens. Notably, ovarian cancer spheroids exhibited increased tumorigenicity and proportion of cancer stem cells after several passages ( Ward Rashidi et al. , 2019 ); chemoresistant cancer stem cells can also be generated with 3D culture of CD44 + CD117 + cells ( Chen et al. , 2014 ). Fallopian tube organoids are similarly suitable for developing combination therapies for high-grade serous ovarian cancer ( Zhang et al. , 2021a ) while multicellular spheroids derived from these cancer patients’ malignant effusions enable drug screening ( Chen et al. , 2020 ). Further, recent applications of drug-loaded hydrogels ( Jamal et al. , 2018 ; Cabral-Romero et al. , 2020 ) and microfluidic conditions ( Ran et al. , 2019 ; Saha et al. , 2020 ; Yang et al. , 2020 ) evaluated targeted cytotoxicity. HA-carboxymethyl cellulose scaffolds facilitate the study of ovarian cancer persistence ( Picaud et al. , 2014 ), and ovarian constructs enable evaluation of metastatic potential of leukemic cells that could have infiltrated OT ( Soares et al. , 2015 ). The organs and tissues of the female reproductive system are not only subject to pathologies that affect reproductive capacity, but also to those that can be life threatening, such as cancers. Via recent applications, microfluidic platforms and organoid/spheroid cultures are revealing diagnostic ( Wang et al. , 2015 ; Dorayappan et al. , 2019 ; Zhang et al. , 2019a ; Chung et al. , 2021 ), and prognostic ( Chowanadisai et al. , 2016 ; Ward Rashidi et al. , 2019 ; Chung et al. , 2021 ) biomarkers and/or gene signatures, in addition to elucidating drivers of tissue metaplasia ( Chumduri et al. , 2021 ). These approaches are helping to establish new and alternative cancer therapies. For example, endometrial organoids allowed the identification of a menin-mixed lineage leukemia inhibitor for endometrial cancer ( Chen et al. , 2021 ), while fallopian tube organoids provided a platform to test combination therapies for ovarian cancer ( Zhang et al. , 2021a ). Locally injectable hydrogels, such as those made of PEG and polylactic-co-glycolic acid ( Shin and Kwon, 2017 ), PEG and poly(ε-caprolactone) polymeric micelles ( Xu et al. , 2018 ), polypeptide PC10A and silver sulfide quantum dots ( Jin et al. , 2019 ) or light-cured glycol chitosan ( Hyun et al. , 2019 ), successfully sustained delivery of drugs to ovarian or cervical tumor models, while those made of HA-danazol reduced the size of endometriosis cysts ( Nomura et al. , 2006 ). Similarly, 3D cervical models supported testing the efficacy of PEGylated lipoplexes containing silencing RNAs targeting human papillomavirus lesions ( Lechanteur et al. , 2017 ), while gold nanorods can facilitate intracellular drug delivery ( Yan et al. , 2016 ). Another bioengineering strategy that can improve clinical workflow is the encapsulation of follicles in alginate (with or without Matrigel) before cryopreservation, which not only is more time efficient, but also affords a means of improving follicle survival and development ( Camboni et al. , 2013 ; Vanacker et al. , 2013 ). The complete or partial regeneration of damaged tissues and organs is arguably the application for which bioengineering is most recognized. In the reproduction field, many in vivo studies have tested hydrogels and scaffolds for uterine regeneration ( Supplementary Table SII ). Among them, polylactide nanofilm can seal defects smaller than 3 mm in chorion-amnion and uterine membranes ( Pensabene et al. , 2015 ), while degradable polylactic acid- co -poly(ε-caprolactone)-gelatin nanofiber meshes with endometrial MSCs promote tissue integration via an anti-inflammatory response ( Mukherjee et al. , 2019 ). Heparin-poloxamer hydrogels ( Xu et al. , 2017a , b ; Zhang et al. , 2017b , 2020b ), collagen hydrogels or scaffolds loaded with bone marrow MSCs ( Ding et al. , 2014 ), basic fibroblast growth factor [(bFGF; ( Li et al. , 2011a )], embryonic stem cell-derived endometrium-like cells ( Song et al. , 2015 ), vascular endothelial growth factor [VEGF ( Lin et al. , 2012 )] or human umbilical cord-derived MSCs [UC-MSCs ( Xin et al. , 2019 ; Liu et al. , 2020 )] and stromal cell-derived factor-1α-loaded chitosan-heparin hydrogel ( Wenbo et al. , 2020 ) repaired morphology and restored the function of injured rat uteri. Further, improved uterine regeneration, and some restoration of fertility with successful implantations, pregnancies and LBs is achievable via transplantation of DC human amniotic membrane loaded with adipose stem cells ( Han et al. , 2020 ) or oral mucosal epithelial cells ( Chen et al. , 2019 ), DC uterine matrix ( Santoso et al. , 2014 ; Hellström et al. , 2016 ; Hiraoka et al. , 2016 ; Miki et al. , 2019 ; Li et al. , 2021 ), or DC endometrial ECM hydrogel loaded with growth factors ( López-Martínez et al. , 2021b ) ( Supplementary Table SII ). Similarly, gelatin methacrylated and sodium-alginate scaffolds with bFGF ( Cai et al. , 2019 ), MSC-laden Matrigel microspheres ( Xu et al. , 2021 ), hydrogel-encapsulated decidualized endometrial stromal cells ( Kim et al. , 2019 ), HA hydrogels ( Liu et al. , 2019 ), HA-collagen hydrogels with endometrial stem cells, stromal cells and vessel cells ( Park et al. , 2021 ), PEG-based hydrogels ( Wang et al. , 2021 ) or poly(glycerol sebacate) scaffolds seeded with bone marrow-MSCs ( Xiao et al. , 2019 ) also successfully regenerated a damaged endometrium. One reproductive disorder prompting a search for an effective tissue regeneration treatment is AS, an acquired iatrogenic disorder characterized by adhesions within the uterine cavity or cervix. To date, patients have received treatments using collagen hydrogels loaded with UC-MSCs ( Cao et al. , 2018 ; Zhang et al. , 2021b ), bFGF ( Jiang et al. , 2019 ) or bone marrow mononuclear cells ( Zhao et al. , 2017 ) to improve uterine response and function. In vivo studies in rats demonstrated that UC-MSCs facilitate collagen degradation, regenerate uterine wall thickness and restore fertility ( Xu et al. , 2017c ), while organoids derived from human embryonic stem cells regenerate uteri of AS models ( Jiang et al. , 2021 ). Furthermore, HA-based hydrogels ( Liu et al. , 2019 ) and cell sheets made of rat endometrial cells ( Kuramoto et al. , 2018 ), rat oral mucosa epithelial cells ( Kuramoto et al. , 2015 ) or UC-MSCs ( Kuramoto et al. , 2020 ) also demonstrated utility in preventing and/or repairing uterine adhesions. Other biomaterials, such as AdSpray [based on dextrin ( Kai et al. , 2018 )], Carbylan-SX ( Liu et al. , 2007 ), mitomycin C-loaded crosslinked HA films and gels ( Liu et al. , 2005 ), urinary bladder ECM ( Zhang et al. , 2020a ), HA/carboxymethylcellulose membranes ( Demirbag et al. , 2005 ) and polylactic acid-pluronic copolymer ( Yamaoka et al. , 2001 ), also prevent post-operative adhesions, while hydrogels made of PEG with or without poly( l -phenylalanine) ( Wang et al. , 2021 ), aloe poloxamer with estradiol encapsulated in nanoparticulate DC uterus ( Yao et al. , 2020a ) and stromal cell-derived factor-1α-loaded chitosan-heparin ( Wenbo et al. , 2020 ) prevent/reduce uterine fibrosis in preclinical models. Notably, commercial hydrogel-based adhesion barriers, such as PEG-based SprayGel [used for myomectomy patients ( Mettler et al. , 2004 , 2008 )] and Actamax Adhesion Barrier ( Trew et al. , 2017 ), have already proceeded to clinical use. Research over the last two decades also yielded significant strides in reproductive organ transplantation. Although uterine and ovarian transplantation surgeries often are performed without the aid of bioengineering, recent approaches may provide benefit, particularly for some OT transplantation patients. For example, encapsulating human OT in Alloderm allowed two patients to conceive through ART ( Oktay et al. , 2016 ). In mouse models, transplanted HA-encapsulated vitrified ovaries compromises follicles and FSH production ( Taheri et al. , 2016 ), but encapsulating fresh ovaries with a HA-based hydrogel (with/without VEGF and bFGF) protects the follicular reserve and re-establishes endocrine function ( Tavana et al. , 2016a , b ). Similarly, co-culture of human bone marrow- or visceral peritoneal-derived MSC hydrogels with mouse OT can restore endocrine function earlier after transplantation, but delays follicle development ( Mehdinia et al. , 2020 ). On the other hand, culturing OT fragments with laminin components of the native ovarian ECM enhances follicle survival and development to the secondary follicle stage ( Hao et al. , 2020 ), while encapsulating follicles in PEG vinyl-sulfone hydrogels maintains the reserves to day 60 and supports antral development and ovulation ( Kim et al. , 2016 ). In corroboration, encapsulating OT in TheraCyte or Dual-PEG capsules (which has a proteolytically degradable PEG vinyl-sulfone core with a non-degradable shell) restores ovarian function and follicle development after allotransplantation, without evoking an immune response ( Day et al. , 2019 ). Using hydrogels to sustain local release of bFGF decreases fibrosis in human OT, in addition to improving revascularization and follicle density after xenotransplantation ( Tanaka et al. , 2018 ); these findings corroborate prior work using fibrin-bFGF scaffolds, which protect murine follicular reserves and increase revascularization after transplantation ( Gao et al. , 2013 ). Similarly, exogenous mouse endothelial cells engineered to constitutively express anti-Müllerian hormone (AMH) ( Man et al. , 2017 ), or STEMPRO ® adipose-derived MSCs ( Manavella et al. , 2018 ; Cacciottola et al. , 2021 ), can preserve primordial follicles by promoting revascularization of OT encapsulated with fibrinogen-thrombin. A recent report describes improved ovarian cortex xenografting outcomes achieved by embedding OT in fibrin clots and treating mice with simvastatin ( Magen et al. , 2020 ). Furthermore, encapsulation of OT with an alginate hydrogel results in developmentally competent oocytes and protects against metastatic lesions [at least short term ( Rios et al. , 2018 )]. One goal of reproductive bioengineering is to achieve artificial ovaries for alternative fertility preservation strategies. This goal remains somewhat out of reach, but initial work described the encapsulation of ovarian stromal cells in chitosan-silk hydrogels ( Jafari et al. , 2021 ). Further, primordial follicles in murine ovarian fragments encapsulated with fibrin modified with heparin-binding peptide, heparin and VEGF ( Shikanov et al. , 2011c ; Kniazeva et al. , 2015 ) and follicles transplanted in bioprinted scaffolds ( Laronda et al. , 2017 ) have also produced pups after natural mating. Fibrinogen and thrombin, which are clotting factors, are similarly used to encapsulate follicles ( Chiti et al. , 2016 , 2017 ) or OT, with or without addition of stem cells or stromal cells. On the other hand, transplantation of granulosa and theca cell constructs restores hormone function, improving bone and uterine health as well as lowering body fat, compared to pharmacological hormone replacement therapy ( Sittadjody et al. , 2017 ). Hydrogels and scaffolds provide some advantages in models of premature ovarian failure (POF) or premature ovarian insufficiency (POI). For example, human amniotic epithelial cells encapsulated within sodium alginate bioglass protect granulosa cell function and ovarian vascularization in a chemotherapy-induced POF model ( Huang et al. , 2021 ). Similarly transplant of human UC-MSCs embedded in Matrigel promotes granulosa cell proliferation and ovarian vascularization ( Zhou et al. , 2021 ), and adipose-derived stem cells in a collagen scaffold restore ovarian function in POI models ( Su et al. , 2016 ). Notably, local delivery of embryonic stem cell-derived mesenchymal progenitor cells in a HA gel increases the ovarian reserve, and estrogen and AMH levels, ultimately improving the quality of oocytes and embryos in mice that model POI ( Shin et al. , 2021 ). Bioengineered materials and techniques can also be implemented during reconstructive gynecological surgeries. Recently, vaginal reconstruction was successful in a patient with MRKH syndrome, a rare congenital disorder characterized by abnormal uterine and vaginal development despite normal ovarian function and external genitalia; this approach used a DC porcine small intestine submucosa scaffold ( Zhang et al. , 2019b ). Remarkably, this biomaterial achieves structural and functional vaginas for up to 8 years ( Raya-Rivera et al. , 2014 ). Similarly, vaginoplasty with an acellular dermal matrix (called RENOV) is safe and effective, and results in an anatomically correct vagina that provides near-normal sexual function ( Zhang et al. , 2017c ). Neovaginas were also safely constructed using Surgicel (an oxidized cellulose scaffold) for 10 patients ( Dadhwal et al. , 2010 ), or bovine-derived dermis scaffold for another patient ( Noguchi et al. , 2004 ).

Female reproduction is regulated by complex networks of molecular, endocrine and tissue/organ interactions. As such, substituting the entire female reproductive tract will be challenging; however, interdisciplinary work provides novel insight into the physicochemical properties necessary to support and achieve these biological processes. Advances in reproductive bioengineering technologies have redefined the landscape of fertility-restoring strategies and therapeutic options that are, or soon could be, available to patients. These translational endeavors provide substantial promise for effective treatments for a wide range of reproductive system pathologies.

The organs of the female reproductive system—the uterus, ovaries, fallopian tubes, cervix and vagina—work together to provide the hormonal and anatomical support necessary for the generation of offspring. As such, reproductive health is susceptible to a number of negative congenital or acquired factors, restricting fertility and quality of life. These concerns prompt a large field of research into the underlying biology as well as approaches for preventing or treating various pathologies. However, ethical and technical limitations around using and/or transplanting human tissues for research purposes requires that most studies are conducted in vitro or in vivo using animal models. While valuable, these approaches face inherent limitations in translatability, such as the complexity of recreating the anatomy, physiology and interactions of reproductive organs using classical 2D in vitro models, in addition to the differences between species. Thus, bioengineering has become indispensable for creating representative and reliable 3D models (for both in vitro and in vivo uses) as well as providing alternative applications for regenerative medicine. This review’s systematic compilation of the extensive bioengineering advances in the context of the female reproductive system since 2000, provides a global overview of the different techniques, their pre-clinical testing and/or clinical applications and the anticipation of future trends. The uterus, and in particular the endometrium, is fundamental for implantation and maintenance of pregnancy ( Governini et al. , 2021 ). As such, much research is devoted to the creation of functional endometrial models and combining endogenous endometrial cell populations in different formats and biomaterials ( Table I ). Notable among these are paracrine models of epithelial and stromal cell co-culture ( Schutte et al. , 2015 ; Park et al. , 2021 ), as well as models of decidualization ( Schutte and Taylor, 2012 ; Gnecco et al. , 2019 ), implantation ( Park et al. , 2003 ; Wang et al. , 2012 , 2013 ; Buck et al. , 2015 ), vascularization ( Pence et al. , 2017 ), ECM interactions ( Cook et al. , 2017 ) and uterine contractility ( Kim et al. , 2020b ). In recent years, several groups attempted to recreate the complexity of these models with organoids or assembloids ( Boretto et al. , 2017 ; Turco et al. , 2017 ; Murphy et al. , 2019 ; Abbas et al. , 2020 ; Rawlings et al. , 2021 ), which offer an apparently unlimited potential to recreate the physiological and pathological states of the endometrium ( Boretto et al. , 2019 ). In fact, organoid technology is marking a turning point in endometrial-related research. Despite having been described only 5 years ago, more than 13% of the uterus-related articles reported in this study exploit this technology. Remarkably, although most biomaterials attempt to mimic ECM interactions in vitro , only a few studies notably implement native ECMs ( Young and Goloman, 2013 ; Olalekan et al. , 2017 ; Campo et al. , 2019 ; Arezoo et al. , 2021 ; López-Martínez et al. , 2021a ; Francés-Herrero et al. , 2021b ). Absolute uterine factor infertility can be treated with uterine transplantation (UTx). Taking into account scientific ( Brännström et al. , 2021 ) and media reports, as well as personal communications, we currently estimate that more than 40 LBs have been achieved from over 80 UTx procedures that have been performed thus far. The surgical success rate (defined by a viable organ within 3 months, resumption of regular menstruations within a year, successful pregnancy and LB) was 78% and 64% for live and deceased donor UTx procedures, respectively, and the cumulative LB rates in surgically successful UTx procedures were estimated to be above 80%. Despite these promising success rates, this procedure involves an invasive surgery and associated risks. Bioengineering has been used to mitigate these risks by providing alternative clinical applications. Specifically, bioengineering techniques for the uterus focus predominantly on preventing/reducing adhesions, often associated with AS ( Zhao et al. , 2017 ; Cao et al. , 2018 ; Zhang et al. , 2021b ) and related to uterine factor infertility. In these and other cases of endometrial damage, the main therapeutic objectives are to regenerate tissue structure (e.g. recover endometrial thickness, angiogenesis) and consequently restore function, which ultimately allows the uterus to support implantation and carry a pregnancy to term ( Hellström et al. , 2016 ; Kuramoto et al. , 2018 ; Li et al. , 2019 ; Liu et al. , 2019 ; Wang et al. , 2021 ). Toward this end, different hydrogels and scaffolds show potential in vivo , by regenerating injured uteri in rodent models ( Supplementary Table SII ). Emerging technologies, such as 3D bioprinting and microfluidics, remain under-utilized in research applied to uterine health, but promising possibilities exist for both in vitro modeling ( Ahn et al. , 2021 ; De Bem et al. , 2021 ) and in vivo tissue regeneration ( Ji et al. , 2020 ). The ovaries exert two main functions, namely to tightly regulate folliculogenesis so as to avoid premature depletion of oocytes, and to produce sufficient sex hormones (e.g. estrogen and progesterone) to support decidualization, pregnancy, breast development for lactation and even bone health ( Sittadjody et al. , 2017 ). Developing new IVFG platforms opens opportunities for oncological patients who cannot benefit from current fertility preservation strategies (specifically, OT cryopreservation) due to risk of reintroducing malignancy upon autologous re-transplantation. Culturing follicles/OT in vitro ‘bypasses’ this risk and can produce mature oocytes faster than if the OT was xenografted into a murine model [usually in 8–12 days ( Supplementary Table SIII ) versus weeks-months ( Oktay et al. , 2016 )], but does not have the potential to restore endocrine function. Most IVFG studies we included in this review successfully cultured secondary follicles to the antral stage, and some even recovered mature and competent oocytes ( Supplementary Table SIII ). Few groups have ventured into culturing primary follicles because these follicles tend to have lower survival and oocyte maturation rates ( Tagler et al. , 2012 , 2014 ; Smith et al. , 2014 ). The success of IVFG is not only affected by initial follicle size, but also by the saturation of the biomaterial. Physiologically, the rigidity of the ovarian cortex and the ‘sponginess’ of the medulla play important roles in regulating folliculogenesis. In fact, the mechanical forces of the ovarian cortex ECM may maintain reserves of primordial follicles, only releasing a couple of follicles to grow in the medulla every menstrual cycle ( Choi et al. , 2014 ). Nonetheless, although softer/more flexible biomaterials, such as alginate, Matrigel and VitroGel, could facilitate follicle expansion, materials that are too soft (i.e. 1 mg/ml collagen, fibrin alone or HA-Matrigel, rapidly degrading YKNR plasmin substrate) cannot provide the necessary 3D support, causing granulosa cells to erroneously proliferate and migrate into their surroundings ( Shikanov et al. , 2009 , 2011b ; Desai et al. , 2012 ; Joo et al. , 2016 ). In contrast, saturated/rigid matrices [i.e. 1.5% alginate ( West-Farrell et al. , 2009 )] hinder follicle growth. Although OT transplantation has led to more than 200 human LBs so far ( Dolmans et al. , 2021 ), encapsulating OT before transplantation may provide additional benefits by promoting revascularization, decreasing fibrosis, protecting follicles from “burn-out” (ischemia-induced death of follicles during the first couple of days after transplant), and ultimately, providing the best microenvironment for follicle development in vivo . However, in attempts to standardize OT transplantation or replacement and be able to offer these strategies to a broad population (e.g. oncological patients and/or those in need of hormone replacement therapy), the construction of an artificial ovary containing immature stimulable follicles is gaining momentum and could lead the way for the next decade. Another common ovarian bioengineering application with great potential is the development of heterogeneous and/or patient-derived organoid models to evaluate individual drug response and cancer dissemination ( Supplementary Table SIV ). Fallopian tubes (or oviducts) are the anatomical structures that connect the ovaries and the uterus, providing the space and physiological environment for fertilization and early embryo development. Few bioengineering methods exist to date to recapitulate fallopian tubes and their associated functions in vitro , despite their crucial supportive role during early embryo development. Derivation of human fallopian tube organoids from different cell types ( Kessler et al. , 2015 ; Lin et al. , 2021a ) provided an important breakthrough in the creation of functional in vitro models. Among the few other fallopian tube studies in the bioengineering field, some demonstrate the important cross-talk between the ovaries and the fallopian tubes ( Zhu et al. , 2016 ), or the direct effect of oviductal ECM molecules on embryonic metabolism ( Francés-Herrero et al. , 2021a ). Microfluidic platforms, with their small channels, may be the most suitable for modeling the physiology and pathology ( Ferraz et al. , 2020 ) of this tubular organ. Indeed, the implementation of a bovine oviduct-on-a-chip led to improved IVF outcomes ( Ferraz et al. , 2018 ). The cervix and vagina play critical roles in reproduction by serving as an entryway for sperm during ovulation, physical barriers for infectious microorganisms and a pathway during childbirth. Bioengineering these tissues has provided novel multilayered organoid models to study herpes ( Zhu et al. , 2017 ) and cervical cancer ( Tanaka et al. , 2021 ), also enabling testing of their respective treatments. Although a functional vagina can be created by self-dilation of the vaginal dimple in a majority of patients with MRKH syndrome, vaginal scaffolds are used for reconstructive surgeries ( Noguchi et al. , 2004 ; Dadhwal et al. , 2010 ; Zhang et al. , 2017c , 2019b ; Acién et al. , 2019 ). Other bioengineering alternatives may prevent premature rupture of fetal membranes and incontinence ( Roman et al. , 2018 ), or test contractility inhibitors with bioprinted uterine rings ( Souza et al. , 2017 ). Moreover, hydrogels can be used as carriers for antibiotics, antivirals, antifungals, contraceptives and other drugs ( Dos Santos et al. , 2020 ). Female reproductive function is orchestrated by multiple autocrine, paracrine and endocrine dialogues, which so far have only been studied in vivo in model organisms that cannot accurately reproduce the human body. To overcome the limitations of these models, there exists the need to recreate a multiorgan environment that incorporates physical, mechanical and hormonal variables. Microfluidics offers the most promising bioengineering method, having already enabled the development of an organ system-on-a-chip that combines human liver spheroids, mouse ovarian explants, human fallopian tube epithelium, human endometrium and human cervix tissues to physiologically model a 28-day menstrual cycle ( Xiao et al. , 2017 ). Recently, the endocrine crosstalk between the uterus and the ovary has been modeled on-a-chip, to be able to evaluate the effects of reproductive toxicants ( Park et al. , 2020 ). Another application rarely exploited to date is the possibility of combining, in a single microfluidic platform, a major portion of the workflow in assisted reproduction clinics, thereby minimally altering the environmental conditions to which gametes and embryos are exposed ( Han et al. , 2010 ). New 3D in vitro models representing multiple cell types and/or tissue layers are not only helping to elucidate the physiological dynamics of complex biological processes within the reproductive tract (e.g. those that regulate folliculogenesis, ovulation, decidualization and cancer progression), but also improving personalized medicine ( Stejskalová et al. , 2021 ). In particular, organoids generated in 3D culture can adequately mimic healthy and diseased cell-cell and cell-ECM native tissue interactions, making them ideal models for evaluating individual drug response (for cancer, endometriosis, dysmenorrhea, hormone disorders or other related issues, bacterial/viral/fungal infection, etc.) or implantation potential ( Wei et al. , 2021 ). However, organoid models, especially endometrial ones, have unresolved issues, which the scientific community has started, and should continue, to investigate. Among others, the main limitations are: the lack of expandable organoid lines with stromal and immunological components; the inaccessibility to the organoid lumen; the lack of interactions with native ECM components; and the variability associated with patient tissue origin and culture handling. Automated ‘lab-on-a-chip’ technologies that can rapidly screen various bodily fluids (e.g. blood, ascites or pleural fluid, urine) for specific biomarkers, cancer cells, drugs or oocytes may also efficiently and reliably refine future clinical/therapeutic decisions. Since body-on-a-chip platforms have the potential to model hormone dynamics and systemic disease (e.g. PCOS, diabetes, cancer), combining them with organoid or organ culture and ECM-based environments may provide more robust 3D models for genetic/epigenetic and pharmacokinetics testing. Much remains to be achieved for the field to create (and eventually offer) a completely artificial female reproductive system. Nevertheless, recent advances in the creation of an artificial ovary ( Krotz et al. , 2010 ; Chiti et al. , 2016 ; Sittadjody et al. , 2017 ; Jafari et al. , 2021 ; Yoon et al. , 2021 ; Wu et al. , 2022 ), uterus ( Souza et al. , 2017 ; Ji et al. , 2020 ; Li et al. , 2021 ; Park et al. , 2021 ), cervix ( Arslan et al. , 2015 ; De Gregorio et al. , 2017 ; Zhao et al. , 2020 ) and vagina ( Orabi et al. , 2017 ; Hou et al. , 2021 ) have made promising headway toward this incredible goal. For example, the development of alternative, more natural, options for hormone replacement therapies offers promise for mitigating menopause-associated problems ( Sittadjody et al. , 2017 ; Yoon et al. , 2021 ). In the race to manufacture transplantable tissues and organs, 3D bioprinting has played a discreet role so far, accounting for only 3% of the studies included in this review. Specifically, its relative novelty, limited accessibility among research groups worldwide and lack of standardized protocols and technology could be slowing down its take-off, making it an attractive and necessary niche for investment. Studies focused on bioengineering of the fallopian tubes are scarce, since their functions are bypassed in assisted reproduction clinics. However, recent work demonstrates that an artificial oviduct-on-a-chip may substantially improve IVF and early embryo culture systems by providing a more realistic microenvironment ( Ferraz et al. , 2018 ). Moreover, these anatomical structures are the target of numerous studies to develop alternative contraceptive methods. Among these, artificial hydrogels based on styrene maleic anhydride ( Subramanian et al. , 2019 ) and PEG ( McLemore et al. , 2005 ) offer promise as contraceptive approaches through successful testing in the fallopian tubes of rats and rabbits. Finally, we note the need for greater clinical translation in reproductive bioengineering. Despite the large number of proposals described at the preclinical level, only 5% of the studies compiled in this review are clinical. Advances at the legislative level, meta-analyses to establish optimal procedures, and stronger networks of collaboration between laboratories and medical centers, could be of value. This systematic review identified a wealth of bioengineering-related studies in the context of female reproduction. Nonetheless, it is possible that relevant studies were not found or were excluded because of the keyword selection, subjective nature of the filtering process or reference limit. We compiled the 312 articles that we considered the most significant and representative of the current state of the field. There is an additional limitation in terms of classification of the articles by biomaterial, since the literature lacks consensus in delineating certain hydrogels and scaffolds (e.g. collagen was reported as a hydrogel and scaffold), and some studies combined bioengineering techniques (e.g. organoid or culture with hydrogel/scaffold within a microfluidic system). Therefore, we classified articles, on a case-by-case basis, in a way we deemed most appropriate. Since the original Embase search identified numerous oncology-related studies, additional searches with keywords representing reproductive diseases were conducted to ensure appropriate coverage of the latter. Finally, due to different organs under consideration and divergences in study objectives and designs, the included studies exhibit wide heterogeneity that precluded meta-analysis of the results.

Click here for additional data file.