Extracellular matrix-derived scaffolds in constructing artificial ovaries for ovarian failure: a systematic methodological review

Abstract STUDY QUESTION What is the current state-of-the-art methodology assessing decellularized extracellular matrix (dECM)-based artificial ovaries for treating ovarian failure? SUMMARY ANSWER Preclinical studies have demonstrated that decellularized scaffolds support the growth of ovarian somatic cells and follicles both in vitro and in vivo. WHAT IS KNOWN ALREADY Artificial ovaries are a promising approach for rescuing ovarian function. Decellularization has been applied in bioengineering female reproductive tract tissues. However, decellularization targeting the ovary lacks a comprehensive and in-depth understanding. STUDY DESIGN, SIZE, DURATION PubMed, Embase, Web of Science, and the Cochrane Central Register of Controlled Trials were searched from inception until 20 October 2022 to systematically review all studies in which artificial ovaries were constructed using decellularized extracellular matrix scaffolds. The review was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) protocol. PARTICIPANTS/MATERIALS, SETTING, METHODS Two authors selected studies independently based on the eligibility criteria. Studies were included if decellularized scaffolds, regardless of their species origin, were seeded with ovarian cells or follicles. Review articles and meeting papers were removed from the search results, as were articles without decellularized scaffolds or recellularization or decellularization protocols, or control groups or ovarian cells. MAIN RESULTS AND THE ROLE OF CHANCE The search returned a total of 754 publications, and 12 papers were eligible for final analysis. The papers were published between 2015 and 2022 and were most frequently reported as coming from Iran. Detailed information on the decellularization procedure, evaluation method, and preclinical study design was extracted. In particular, we concentrated on the type and duration of detergent reagent, DNA and extracellular matrix detection methods, and the main findings on ovarian function. Decellularized tissues derived from humans and experimental animals were reported. Scaffolds loaded with ovarian cells have produced estrogen and progesterone, though with high variability, and have supported the growth of various follicles. Serious complications have not been reported. LIMITATIONS, REASONS FOR CAUTION A meta-analysis could not be performed. Therefore, only data pooling was conducted. Additionally, the quality of some studies was limited mainly due to incomplete description of methods, which impeded specific data extraction and quality analysis. Several studies that used dECM scaffolds were performed or authored by the same research group with a few modifications, which might have biased our evaluation. WIDER IMPLICATIONS OF THE FINDINGS Overall, the decellularization-based artificial ovary is a promising but experimental choice for substituting insufficient ovaries. A generic and comparable standard should be established for the decellularization protocols, quality implementation, and cytotoxicity controls. Currently, decellularized materials are far from being clinically applicable to artificial ovaries. STUDY FUNDING/COMPETING INTEREST(S) This study was funded by the National Natural Science Foundation of China (Nos. 82001498 and 81701438). The authors have no conflicts of interest to declare. TRIAL REGISTRATION NUMBER This systematic review is registered with the International Prospective Register of Systematic Reviews (PROSPERO, ID CRD42022338449).


Introduction
Ovarian failure is characterized by the disruption of both endocrine and reproductive ovarian function and has gained increased interest in reproductive medicine, oncofertility, and organ aging (Mauri et al., 2020). Ovarian failure can be attributed to chemotherapy, radiotherapy, natural aging, or genetic predisposition (Sü kü r et al., 2014;Qin et al., 2015;Chemaitilly et al., 2017;Cui et al., 2018). An exhausted ovary will not only lead to fertility loss but can also increase the risks of cardiovascular disease, osteoporosis, and urogenital atrophy (Proserpio et al., 2020). As the average lifespan of women has exceeded 80 years, a naturally menopausal woman will spend almost 40% of her lifetime in the post-menopausal phase (Takahashi and Johnson, 2015). However, young women with primary ovarian insufficiency experience menopause even earlier. Ovarian failure can considerably affect a woman's overall health, work productivity and quality of life. Accordingly, minimizing the adverse effects of ovarian failure is important and urgent.
Common treatments for ovarian failure include pharmacological medication, which mainly involves the supplementation of estrogen alone or estrogen-progestogen combinations. However, hormone replacement therapy should be implemented with particular caution, as the dosage, frequency, and time frame require individualization (Kling et al., 2019;Flores et al., 2021). Novel strategies such as ovarian tissue cryopreservation (OTC) and ovarian tissue transplantation (OTT) have arisen over the last two decades (Donnez et al., 2004), particularly for restoration of fertility after cancer treatments. To date, OTC and OTT have led to 189 deliveries and have been shown to restore ovarian function for many years Dolmans, 2017, 2018;Khattak et al., 2022). However, some groups and countries, such as the American Society of Clinical Oncology, consider OTC an experimental technique (Oktay et al., 2018). The method is also hampered by the risk of tumor reoccurrence due to hidden malignant cells within tissue grafts (Stern et al., 2014;Fajau-Prevot et al., 2017).
An artificial ovary is also a promising approach for rescuing ovarian function (Cho et al., 2019). It is constructed by encapsulating healthy ovarian cells or follicles in scaffolds to replace failed ovaries (Amorim and Shikanov, 2016). Various polymers have been utilized to create scaffolds to fabricate biomimetic functional ovaries. Synthetic polymers, such as gelatin-methacryloyl and polyethylene glycol, are easily manufactured and show enhanced mechanical properties but have limited cell adhesion sites (Mendez et al., 2018;Wu et al., 2022a). In contrast, alginate and fibrin are the two most commonly used natural materials as they have a large diversity of integrin-binding motifs and are more biocompatible. Decellularized ovaries are also based on these individual extracellular matrix (ECM) components and have the advantage of retaining the tissue inner spatial distribution of the tissue as well as its vascularization channels and mechanical properties. Decellularized ovaries are therefore attracting much attention in the field of organ regeneration (Kim et al., 2021b).
Decellularization refers to the removal of the cellular compartments while preserving the natural ECM with optimal porosity, stiffness and elasticity, thus yielding decellularized ECM (dECM) constructs ( Fig. 1) (Saldin et al., 2017;Wu et al., 2022b). Decellularization has been widely studied in bone, heart, dermal tissues, and small intestinal submucosa, in both basic research and clinical practice (Bejleri and Davis, 2019;Xu et al., 2020;Datta et al., 2021). For example, commercial dECM products have been approved for pericardial reconstruction and have demonstrated good performance (van Rijswijk et al., 2020;Mohamed et al., 2021). Acellular dermal matrices are also available to promote breast reconstruction following mastectomy for breast cancer (Folli et al., 2018). Similar to the aforementioned dECM materials, decellularized ovarian scaffolds can also provide tissue-specific biomechanical cues to facilitate cell growth and can therefore be an ideal platform to support follicular development and restoration of ovarian function (Hoshiba, 2021). This encouraging method has been commonly discussed in female reproduction bioengineering topics (Gandolfi et al., 2020;Kim et al., 2021b;Francé s-Herrero et al., 2022). However, a comprehensive and in-depth understanding of bioengineered ovaries is currently lacking. Since the first successful decellularization of human and bovine ovarian tissues in 2012, an increasing number of relevant papers has been published in recent years (Laronda et al., 2015). As many more different strategies have been proposed, there is an urgent need to compile previous work and to guide future studies.
This systematic review summarizes the recent progress in constructing artificial ovaries based on dECM scaffolds, elucidates its application in restoring ovarian function, and provides a theoretical basis for future optimizations and improvements.

Protocol and registration
This systematic review was registered with the International Prospective Register of Systematic Reviews (PROSPERO, ID: CRD42022338449) and was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) protocol (Moher et al., 2009). This systematic review aims to answer the question 'What is the current state-of-the-art

WHAT DOES THIS MEAN FOR PATIENTS?
Ovarian failure will not only lead to fertility loss but can also affect a woman's overall health and quality of life. One of the strategies under development to treat ovarian failure is the construction of artificial ovaries by encapsulating healthy ovarian cells or follicles into ovarian scaffolds. Decellularized scaffolds are created by the removal of the cells while preserving the natural tissue matrix; they have been widely studied in tissue regeneration research and have been applied in clinical practice. This systematic review was conducted to evaluate whether the decellularization-based artificial ovary can be used to restore ovarian function. Specifically, we provide detailed information on the ovarian decellularization procedure, evaluation method, and preclinical study design. The decellularized scaffolds loaded with ovarian cells can produce estrogen and progesterone, though with high variability, and support the growth of follicles, without reports of serious complications. Thus, the decellularization-based artificial ovary may be a promising choice for restoring ovarian function and improving female fertility in the future. methodology to assess dECM-based artificial ovaries for treating ovarian failure?' The search terms were based on a PICO (population, intervention, comparison and outcome) framework: animals and humans (P) with dECM-based artificial ovaries (I) as compared with controls (C) to support follicular growth or restore ovarian function (O) (Schardt et al., 2007).

Literature search
A systematic search was conducted in the PubMed, Embase, Web of Science, and the Cochrane Central Register of Controlled Trials electronic medical databases from inception until 20 October 2022. We used the following search keywords to maximally cover the relevant literature: 'decellularization', 'acellular', 'recellularization', 'ovary', 'ovarian tissue', and 'follicle' (Supplementary Table S1). Articles were identified using MeSH headings and keywords combined with Boolean operators. There was no restriction on the date or publication status.

Eligibility criteria
Studies were included if the decellularized scaffolds were seeded with ovarian cells or follicles. For example, scaffolds derived from amniotic membranes and loaded with ovarian cells were included. Studies were excluded if: (i) the ovarian dECM scaffolds were not loaded with ovarian cells, (ii) the decellularization protocol was not described, (iii) the scaffold was not obtained by decellularization, (iv) no control group was established, (v) it was a repeated record, or (vi) it was a non-English language paper. Studies describing the changes before and after decellularization without control groups were excluded because the alterations might be confounded by placebo effects, or research bias or other changes in the experiment site (Grimes and Schulz, 2002). Different original articles by the same group using the same decellularization method were defined as repeated records and only the most recent version was retained.

Study selection and data collection
Two reviewers (T.W. and K.-C.H.) independently searched the electronic medical databases and selected studies based on the eligibility criteria (Fig. 2). Discrepancies between the selected studies by both authors were discussed in a consensus meeting with the senior authors (J.-J.Z. and S.-X.W.) providing a binding verdict.

Data extraction
Data were independently extracted by two reviewers (T.W. and J.-F.Y.). The outcomes of interest covered most ovary decellularization details and were classified into three groups: decellularization procedure, evaluation method, and preclinical study design. Specifically, the following information was recorded: author, publication year, country, animal species, preprocessing program, type and duration of detergent reagent and enzyme, biocompatibility, DNA and ECM detection methods, seeding cells, experimental grouping, and main findings.
The risk of bias was assessed by two independent blinded reviewers (T.W. and J.-F.Y.) using the SYstematic Review Centre for Laboratory animal Experimentation (SYRCLE) tool (Hooijmans et al., 2014). The SYRCLE tool addresses selection bias, performance bias, attrition bias, detection bias and reporting bias. Each item of a study is assigned 'yes' (low risk of bias), 'no' (high risk of bias), or 'unclear' (insufficient details).

Included articles
The initial electronic database search returned 754 papers, of which 398 remained after duplicates had been removed. There were 46 full-text studies remaining for assessment of the eligibility criteria. After full-text screening, 34 studies were excluded as they did not meet the inclusion criteria and 12 studies were included ( Fig. 2, Supplementary Table S2). All 12 studies performed decellularization and recellularized dECM scaffolds with cells and compared them with the control groups. The studies were appraised for risk of bias (Supplementary Table S3). The allocation concealment was unclear in all studies, and the lack of information on housing of animals and observer blinding to the interventions resulted in a risk of performance bias. The attrition dECM-based artificial ovary | 3 bias was either low risk (n ¼ 5) or unclear (n ¼ 7) in the included studies. The risks of selective reporting and other biases were low for most studies. The studies were published between 2015 and 2022 and were conducted in five countries, most frequently in Iran (n ¼ 7), followed by China (n ¼ 2), Belgium (n ¼ 1), the USA (n ¼ 1), and Italy (n ¼ 1).

Discussion
This systematic review suggests that dECM materials hold promise for constructing artificial ovaries and counteracting ovarian failure. A total of 12 studies were included and the decellularization methods and evaluation parameters varied greatly. An optimal reproducible and standardized procedure is a prerequisite for future clinical application (Fig. 3) (Naso and Gandaglia, 2022).
Both preclinical and clinical trials should comply with the quality control and research pipelines. For these reasons, the evaluations of the following characteristics of ovarian-specific dECM scaffolds are proposed: (i) DNA removal efficiency; (ii) ECM preservation; (iii) cell debris residues; (iv) biocompatibility; and (v) restoration of ovarian function. Effective decellularization is reflected by the adequate removal of cellular components and good preservation of ECM proteins (Luo et al., 2019). The tissue type, animal species, chemical reagents, and exposure time all affect the efficiency and should be balanced (Eivazkhani et al., 2019). For example, minimizing the detergent concentration contributes to more retention of ECM, but it might also cause insufficient cell removal, risking a relevant immune response after in vivo implantation (Luo et al., 2019;Chakraborty et al., 2020). Regarding the DNA evaluation methods, it was interesting that only one study performed electrophoresis, while all studies conducted H&E/DAPI staining and DNA extraction. We speculated that this might be due to the complex manipulation processes compared with histological staining and DNA quantification (Lee et al., 2012). However, neither DNA staining nor quantification can substitute for the evaluation of DNA size, as the two methods occasionally cannot detect the minimal virus DNA contents, which may nevertheless elicit an immune response (Chakraborty et al., 2020). Furthermore, peracetic acid/ ethanol treatment eliminates small DNA debris (Hodde and Hiles, 2002). Altogether, there should be more focus on comprehensive assessment of DNA residues, especially DNA fragments.
ECM is composed of structural and soluble components. The former includes collagen, laminin, fibronectin and elastin (Naba et al., 2017;Yuzhalin et al., 2018). The Masson trichrome and Heidenhain's Azan staining method mainly detects fibrillar collagens (Polat et al., 2007). Mallory staining can distinguish collagens from elastin (Chambrone et al., 2015), while the Alcian blue staining is specific for glycosaminoglycans (Mead, 2020). Immunological methods can reveal the ECM protein distribution via antigen-antibody binding. However, the fact that various methods serve similar purposes and produce repeated results  Higher rates of antral cavity formation and maturation in › than ‹. Higher levels of E 2 and P 4 in › than ‹. Survival rate and follicular diameter in ‹›, nsd. Higher levels of Zp2, Gdf9, Bmp6 and Bmp15 in fi than ‹›fl. Chiti et al. (2022) ‹ Alginate › 25% Alginate þ 75% dECM fi 10% Alginate þ 90% dECM fl dECM Follicle recovery rate rose with increased alginate content.
No follicles were recovered in fl.
More first polar body in the 3D groups containing dECM than other groups. More MII oocytes of 2D culture containing dECM than the 2D control groups. More nuclear maturation in 3D culture groups containing dECM than the 3D control group. Higher antral follicles and lower follicle degeneration rate in ‹› than fifl. Follicle diameter and E 2 in ‹›, nsd.
The presence of immune cells and neovascularization in fifl.
The distribution of INHa, ER and PR in fl, and DAZL in fi.
Higher levels of E 2 and P 4 in fl than ›fi.  remains an issue. Most studies neglect other equally important molecules such as soluble hormones and growth factors, and properties such as stiffness (Fazelian-Dehkordi et al., 2022;Wu et al., 2022b). To address these problems, it is suggested that dECM undergo enzyme-linked immunosorbent assay (ELISA), stress relaxation testing and atomic force microscopy. Ovaries are important endocrine glands that secrete estrogen, progesterone, and anti-Mü llerian hormone (Leong, 2018). Nevertheless, the amounts of these components within dECM scaffolds remain obscure although they are expected to facilitate the growth of seeding cells. Recently, ECM mechanical cues were demonstrated to underlie the development of polycystic ovary syndrome and in vitro activation (Lunde et al., 2001;Wood et al., 2015;Kawamura et al., 2016). ECM accumulation along with aging leads to a fibrotic ovary and worsens follicle development (Shah et al., 2018), which underlines the fact that dECM rigidity should be tested and modified for artificial ovary construction (Chiti et al., 2018). The dECM-derived artificial ovaries demonstrate great potential for restoring ovarian function. In most of the included studies, preantral follicles formed the antral cavity, underwent maturation, and produced estradiol after reseeding on the dECM scaffolds. Many groups achieved healthy follicle development both in vitro and in vivo. Nevertheless, further progress concerning any pregnancy or live birth of experimental animals with the dECM scaffolds has never been reported, which would be a milestone in the field of decellularized materials. Additionally, more in-depth studies concerning safety are needed.

Future perspectives
In addition to direct application as solid platforms, dECM-based materials can also be produced via 3D printing and microfluidic chips. Such bioengineering approaches have recently been applied in the vagina and for ovary regeneration (Hou et al., 2021;Zheng et al., 2022). After a series of lyophilization, pulverization, digestion, and solubilization, the dECM powder is formed into a hydrogel, serving as a printable bioink . Meanwhile, the rheological, flow and gelatin behaviors of the bioinks are characterized to aid parameter optimization during printing (Das et al., 2019). The bioink extrusion speed, nozzle routing, strut distance, and layer height should be coordinated. Finally, the dECM hydrogel is shaped via 3D printing and a biomimetic ovary mimicking the actual cell arrangement is precisely fabricated (Laronda et al., 2017). Mixing dECM hydrogel with collagens is expected to create a more rigid environment similar to that of ovary fibrosis and the cortex of polycystic ovary syndrome (Filatov et al., 2015;Ouni et al., 2020;Kim et al., 2021a). The combination of dECM and hyaluronic acid is suitable for the survival of the cumulus cell-oocyte complex, which is essential for ovulation and fertilization (Briggs et al., 2015;Serati et al., 2015). When integrating with microfluidic chips, the decellularization and recellularization manipulation can be reproduced directly in the devices or used as medium to fill the chips (Hong et al., 2017;Palikuqi et al., 2020;Bhatt et al., 2022). The latter provides a dynamic stimulus to the cumulus cell-oocyte complex or denuded oocytes, which improves the outcomes of oocyte maturation and IVF (Nagashima et al., 2018;Podwin et al., 2020;Healy et al., 2021;Sadeghzadeh Oskouei et al., 2021). Altogether, the integration of these advanced culture systems will facilitate the development of ovarian regeneration and drug-testing models.
Even though decellularization eliminates most immunogenic substances, adverse effects that include inflammatory reactions, fibrosis and calcification have been recorded (Padalino et al., 2016;Woo et al., 2016;Hofmann et al., 2017). Most studies did not focus adequately on evaluating the residual antigenicity. a galactosidase (aGal) is the major xenogeneic antigen that causes rejection-related responses, but no included studies examined the aGal level . It is also suggested that dECM is elevated for the release of interleukins, chemokines and other cytokines; therefore, blockade intervention is recommended in xenotransplantation (Islam et al., 2021). Antigenicity is also associated with the decellularization protocols. Prolonged operation times and increased detergent concentrations are used to reinforce decellularization efficiency. However, these adjustments may further expose the hidden antigenic motifs in ECM components, such as laminin, aggrecan, versican, collagen types I and IV, and hyaluronan (Chakraborty et al., 2020). The 7S domain constitutes the amino-terminal end of type IV collagen; when exposed after decellularization, it will cause a neutrophil chemotactic response (Senior et al., 1989). Hyaluronan is an abundant ECM component enriched in follicular fluid and ovarian stroma. However, it also has diverse roles in chronic inflammation and immune cell activation and can even cause autoimmune diseases (Johnson et al., 2018;Nagy et al., 2019). The matrikines refer to a group of ECM fragments and are inactive in most cases. However, structural and conformational alterations in ECM proteins may result in matrikine release by proteolysis, which contributes to fibrosis, cancer, and aging (Abdul Roda et al., 2015;Jariwala et al., 2022). In some circumstances, these minor alterations occur in ECM ultrastructure and cannot be observed via histological staining or electrical microscopy. In such cases, thermal analysis characterization using differential scanning calorimetry can be used (Samouillan et al., 1999;Wu et al., 2022b). To summarize, appropriate selection criteria are a prerequisite to identify the antigenic motif or matrikine on xenogeneic decellularized tissue to avoid the possibility of interspecies reaction upon clinical application.

Conclusion
To our knowledge, this is the first systematic review to provide a broad overview on the current state-of-the-art of dECMbased-artificial ovaries. Artificial ovaries provide promising opportunities to restore ovarian function, yet animal studies and preclinical applications of dECM-derived artificial ovaries have only just begun. It is important to comprehensively assess the decellularized scaffolds and demonstrate their effectiveness. Standardizing decellularization protocols and the implementation of quality controls and cytotoxicity measurements will enable the construction of generic and comparable dECM-based artificial ovaries in the future.

Supplementary data
Supplementary data are available at Human Reproduction Open online.

Data availability
All data used for the study have been included in the article and Supplementary Material.

Funding
This study was funded by the National Natural Science Foundation of China (Nos. 82001498 and 81701438).