Mesenchymal stem cells (MSCs) are a popular stem cell type commonly used in regenerative medicine and tissue engineering. Biologically, MSCs are stromal cells that exhibit a diversity of properties such as self-renewal, multi-lineage differentiation, immunomodulation and mediating cell-matrix interactions (26, 29–31). The use of MSCs is particularly relevant to vascular regeneration in VCA through their capacity to enhance angiogenesis and neovascularization by mechanisms that include direct trans-differentiation into vascular cell types and by the secretion of paracrine factors such as vascular endothelial growth factor and basic fibroblast growth factor (29, 32, 33). The immunomodulatory function of MSCs is also well known and can help promote constructive tissue remodeling and tissue survival following in vivo transplantation by suppression of host immune responses (34). Altogether, these properties, combined with their ease of acquisition in clinical settings make MSCs an attractive cell source for tissue regeneration and recellularization.

Recently, there has also been interest in the use of induced pluripotent stem cells (iPSC), which can be derived from the reprogramming of somatic cells with specific pluripotent factors that can reverse cell phenotype back to an “embryonic-like” state (35). Likewise, iPSCs are relatively easy to generate and have comparatively fewer ethical hurdles compared to embryonic stem cells, making them a good candidate for tissue engineering strategies (36). To date, the majority of work with iPSC scaffold recellularization has been performed in whole organ engineering, with examples of iPSC-derived endothelial cells seen in lung vascular (37, 38) and kidney vascular (39) engineering. However, there have been no reported attempts that utilize iPSCs in VCA recellularization, but such an approach may be an exciting future direction with great research and clinical potential.

Zhang et al. explored perfusion decellularization and recellularization of skeletal muscle obtained from porcine rectus abdominus. Decellularization was conducted through both antegrade and retrograde perfusion using 0.02% trypsin/0.05% EGTA, 0.1% sodium dodecyl sulfate (SDS) and 1% Triton X-100 each through the inferior epigastric vessels, with more perfusion time in retrograde perfusion than antegrade. For recellularization, C2C12 mouse myoblasts were statically seeded onto pepsin-digested decellularized matrices in 24-well plates. Seeded myoblasts showed higher expression of myosin-heavy chain (marker of matured muscle) after 1 day post-differentiation, higher fiber diameter, and higher fusion index relative to the control tissue, small intestinal submucosa ECM. While perfusion recellularization was not pursued, the seeded cells indicated an ability for differentiation and showed evidence for a preserved ECM to allow cell growth and maturation. However, the stages of myoblast cell proliferation and differentiation were not monitored long-term nor were details of how differentiation was induced given. Further, no muscle stimulation nor characterization of muscle contraction of the differentiated cells was conducted. Separately, Zhang et al. attempted to use decellularized grafts to treat partial thickness abdominal wall defects in rats. Eight weeks after surgery, long ovoid nuclei in nerve bundle membranes could be detected, similar to Schwann cell nuclei morphology. As well, some organized neo-muscle fibers could also be observed along the endomysium (3).

For nerve injuries, multiple surgical reconstructive options are available such as autologous nerve grafting, bone excisions to shorten large nerve gaps, and nerve transfers. However, many limitations exist such as limited graft availability, scarring, donor site morbidity, challenges in procuring relevant nerve size and morphologies, and delayed or impaired reinnervation post-surgery. Using nerve allografts to fill nerve gaps between the proximal and distal ends of the defected nerve is an alternative approach. Many natural and synthetic-based biomaterials have been used for preparing nerve conduits to bridge nerve gaps. Of these, some nerve conduits are commercially available for nerve gap repair although its application in long nerve gap (>3 cm) repair has had mixed results (67, 74). Further, an additional challenge is ensuring conduits contain regenerative capacity. Some studies have attempted applying methods of de- and recellularization in nerve grafts.

Using neural-derived ADSCs, Zhang et al. proposed recellularization methods in a rat intercostal nerve model. Segments of intercostal nerves procured from pigs were subjected to a multi-step decellularization process involving freeze-thaw cycles, 0.25% trypsin exposure, 0.1% SDS/3% Triton X-100 incubation, with agitation. Notably, authors used 1-ethyl-3 (3-dimethylaminopropyl) carbodiimide (EDC) to facilitate crosslinking. This is a method employed in various solid organ decellularization studies, used to improve mechanical properties of decellularized scaffolds (69). The ECM structure and porosity, and some growth factors were preserved upon decellularization. Recellularization involved mixing neural differentiated ADSCs with collagen gel and injecting the cell suspension on proximal and distal ends of the nerve. The nerves were incubated for coagulation before being cultured in medium for 4 days. Some recellularized nerves were used to bridge sciatic nerve gaps in rats. Cells were labeled with PKH-26 and authors traced the neural cells in vivo post-transplantation, revealing longitudinally aligned neural cells and expression of neural tropic factors. Seven days post-transplantation, authors also reported glial cell characteristics in the neural differentiated ADSCs (74).

Wang et al. proposed methods of recellularization with Schwann cells in Wistar and Sprague-Dawley rats. Autologous Schwann cells are difficult to expand and obtain hence authors identified the need for alternative sources of Schwann cells for recellularization. BMSC and ADSCs isolated from rat femur, tibias, and subcutaneous back fat tissues were differentiated into Schwann cells. Sciatic nerves measuring 20 mm were decellularized using a previously established protocol with 3% Triton X-100 overnight followed by 4% sodium deoxycholate (SDC) with agitation for an additional 24 h (75, 76). For recellularization, 5 treatment groups were tested. Left sciatic nerves of all rats were exposed and used to create a 15 mm gap. While control group animals had an immediate nerve repair using the transected nerves, the other 4 treatment groups involved nerve repair using acellular nerve grafts whereby 5 × 10⁵ Schwann cells were injected through both proximal and distal ends of the nerve grafts. Authentic Schwann cells, BMSC-derived and ADSC-derived Schwann cells as well as cells suspended in DMEM were used in the remaining 4 treatment groups. Analyses were conducted at 2-week and 12-week intervals. BMSC and ADSC Schwann cells independently showed regeneration and functional recovery of the sciatic nerves whereby recellularization with either cell type was comparable across autografting with Schwann cells alone and better than acellular nerve autografting. Authors identified alternative Schwann cell sources and the enhanced regenerative capabilities these cell types offer upon recellularization into acellular nerve grafts.

It was further speculated by Wüthrich et al. whether vascularization could help overcome limitations with bioengineering nerve conduits given that vascularization may promote nerve regeneration in facilitating and promoting neurotropic growth factor secretion, cell survival and growth, and provision of nutrients and oxygen (67, 68). Porcine sciatic nerves containing vascular pedicles from porcine hindlimbs were decellularized using 1% SDS and 1% Triton X-100. Upon decellularization, a patent vasculature was visualized, and nerve tissues retained most of their architecture and bundle-like structures histologically. Absence of myelinated fibers and nuclei was observed. Basal lamina of individual nerve fibers was collapsed and disorganized upon decellularization whereas the perineurium remained intact. A panel of neurotropic, inflammatory and complement factors were assessed in decellularized samples of which most were detectable and some showed reduced expression. Re-endothelialization was conducted using fluorescently-labeled wtPAEC. Following an overnight pre-conditioning of the scaffold in growth medium, wtPAEC cells were manually injected into the vasculature across four consecutive injections every half hour for 2 h total. Perfusion was started following cell injection and increased every half hour from 0.5 to 2 mL/min. Nerves were analyzed on Day 0, 1, and 7 whereby wtPAEC were observably settled in the vessels and using the endothelial marker, CD31, cells were observed closer to the pedicle and near large caliber vessels only (67). These initial results suggest that this vascularized nerve graft has biocompatibility and can support endothelial cell engraftment. However, further recellularization efforts examining recellularization from both arterial and venous conduits may be an alternative to address the lack of cells in distal parts of the pedicle. Notably, authors focused on re-endothelialization given the identified need to develop a functional vasculature to promote nerve regeneration. Co-cultures with cell types, for example vascular smooth muscle cells and pericytes, can be considered as a future step to promote vascular reconstitution.

An engineered fasciocutaneous flap offers versatile and clinically-relevant applications in reconstructive surgery, particularly in the repair of large soft tissue defects. Additionally, the composite nature of such flaps with the epidermal cutaneous layer, the fibroblastic dermal layer, and the adipose subcutaneous layer make these tissues an attractive investigational model in VCA. To date, three reports in rat and pig models are notable for their attempts to recellularize fasciocutaneous flap tissues. The report by Chang et al. to recellularize the rat fasciocutaneous flap is one of the earliest. While the tissue flap used was not decellularized prior to recellularization the authors showed that the flap could be maintained ex vivo in a perfusion bioreactor and that adult stem cell populations such as bone derived and adipose derived human MSCs could be recellularized intravascularly and supported for up to 24 h ex vivo and up to 7 days in their transplantation studies in nude rat recipients (77).

Later work by Zhang et al. used a decellularized rat inguinal fasciocutaneous flap as an acellular scaffold for cell seeding (3). These surgically explanted free flaps were based on the superior epigastric vessels and contained skin, subcutaneous fat, and vascular tissues. Flaps were then treated with a protocol of freeze-thaw at −80 °C, and subsequently, perfused with NaCl (0.5 M/1 M), 0.25% trypsin/EDTA, isopropanol, and 1% Triton to generate acellular flaps. Recellularization relied on both vascular perfusion of HUVECs and injections of human ADSC-HUVEC co-culture with HUVECs into the dermal and subcutaneous fat layers of the flap. A similar approach of recellularizing fasciocutaneous flaps was undertaken by Jank et al. in the pig inguinal fasciocutaneous flap (7). Explanted flaps based on the superior epigastric artery were perfusion decellularized with SDS/Triton and then recellularized using 4 × 10⁷ HUVEC by gravity perfusion at 110 mmHg perfusion pressure. The recellularized flap was maintained under perfusion bioreactor conditions for five days. However, in the recellularized flaps by Zhang et al. and Jank et al., the regenerated vasculature was incomplete as evidenced by thrombosis noted after in vivo transplantation in both cases. In Zhang et al., an occlusion of the main feeding arterial inlet was noted at 7 days post-transplantation. In Jank et al., an absence of venous outflow on gross inspection and evidence of only partial endothelial coverage only to vessel diameter of 300 µm by histological examination similarly demonstrated incomplete re-vascularization in the reseeded scaffold (7). These issues of incomplete recellularization and poor post-transplant vascularization is hypothesized to be due to shear stress leading to myointimal hyperplasia and luminal occlusion or activation of the coagulation cascade caused by areas of exposed ECM created by incomplete coverage of reseeded endothelial cells. At present, the vascularization in a recellularized fasciocutaneous VCA tissue remains a key obstacle to be addressed in future work.

Engineered facial allografts introduce added complexity to VCA tissue engineering compared to vascularized skin/adipose grafts. This is attributable to the diverse population of constituent tissues, including skin, adipose, mucosa, cartilage, and muscle. To date, recellularized facial scaffolds have been attempted with both rodent and human scaffolds.

Duisit et al. studied recellularization on full-thickness hemifacial grafts in a rodent model. An initial comparative analysis of two decellularization protocols based on either perfusion or agitation of 1% SDS detergent determined that both perfusion and agitation were equivalent in decellularization efficacy although the perfusion method was noted to have an advantage of increased decellularization of cartilage. Later seeding of rat hemifacial scaffolds was done with human ADSCs seeded statically onto 1 cm² discs on both the dermal and internal surfaces and were observed to be viable using Live/Dead vital staining after 7 days of in vitro culture. No perfusion seeding was performed in this study (47).

In the human face, Duisit et al. utilized both segmental lower-facial and full facial allografts procured with pedicled access with the facial artery and the external jugular vein from cadaveric donors and decellularized with a protocol using sequential perfusion of 1% SDS, 1% Triton, isopropanol, and DNase. Characterization of the decellularized scaffolds demonstrated preservation of the ECM microstructure and a patent vascular tree on CT angiography. Recellularization was selectively performed on the lip subunit using both static disc and whole perfusion bioreactor formats. For static disc seeding, NIH-3T3 dermal fibroblasts and C2C12 progenitor myoblast cells were seeded atop 1 cm² acellular lip discs and cultured in DMEM for 14 days. In the perfusion bioreactor, either human aortic endothelial cells (HAECs) or C2C12 myoblasts were used; HAEC were perfused into the facial artery cannula and cultured for 48 h afterwards whereas C2C12 were seeded by direct syringe injections into the lip parenchyma and cultured for 2 weeks. In both static disc and whole bioreactor lip seeding, examination of the reseeded tissue showed viable cells with evidence of attachment to the ECM. In the case of HAEC, homogenous distribution was seen in the vascular lumen with CD31 positive staining supporting the endothelial phenotype (2). While this study demonstrated encouraging results with regards to cell survival and distribution using their whole bioreactor system with an acellular lip subunit, it remains to be seen whether a perfusion bioreactor reseeding approach could be applied to the large whole face graft scaffold. Duisit et al. did not focus on functionalization of human face graft however such a composite tissue has a unique functional and anatomical tissue composition which warrants different types of recellularization strategies to mobilize different components of the face to a functional extent (2). This requires considerations for the types of tissues present in such a graft. For example, the nose and ear are cartilage-rich tissues whereas the cheeks are predominant with adipose and muscle tissues. Given the diversity of tissue composition of such a graft, recellularization of the graft as a whole may be challenging whereas targeted, partial recellularization focusing on different anatomic locations may be feasible.

Ear recellularization has been documented in the pig and human ear model, with both work performed by Duisit et al. (48, 50). With the porcine ear grafts, grafts were obtained from cadaveric sources and perfusion decellularized with 1% SDS and 1% Triton-X to obtain an acellular scaffold with retained ECM biochemical components and perfusable vascular tree. As with the facial scaffolds, two different formats were used for reseeding: static disc and whole bioreactor seeding. With static discs, fibroblastic Cos-7 cells and GFP-tagged MSCs were used separately used to demonstrate that the acellular scaffold could support viable and proliferating cells. For the whole-ear bioreactor seeding strategy, acellular ear scaffolds were cannulated via the artery and seeded with either MSCs or NIH-3T3 fibroblasts, each using four sequential intra-arterial injections. The system was maintained up to 2 weeks after which the recellularized ear showed homogenous distribution of cells both extra- and intra-vascularly on histology. Ear grafts seeded with MSC were also found to show engrafted cells intravascularly in both large vessels and capillaries as well as an increase in lipid content in the recellularized parenchyma using Sudan Red stain (48). Despite these encouraging results, the overall cell density in the seeded scaffold was poor, owing potentially to inadequate cell numbers for seeding.

Follow-up work in the human ear used an acellular human ear with vascular access via the external carotid artery which was decellularized with a protocol similar to the porcine ear above. Using the whole organ perfusion bioreactor system, Duisit et al. selected HAECs for intra-arterial recellularization. After ex vivo maturation for 48 h, intra-vascular distribution of viable HAEC was seen in both superficial and deep aspects of the tissue. However, cell coverage was inconsistent and segmented through the graft with no cells seen in the larger caliber vessels suggesting an incomplete recellularization of the human ear (50). While the recellularization of both the pig and human ear, particularly the vascular compartment, still require additional work with respect to cell coverage and engraftment, recellularization of ear scaffolds is a technological advancement that has brought attention to whole-tissue reseeding as a promising strategy in VCA engineering.

The larynx is a complex composite tissue that is composed of various components including muscle, cartilage and mucosa. The first tissue engineered larynx study can be attributed to Hou et al. using a decellularized rabbit larynx with SDS/Triton-X. Subsequent in vitro whole-scaffold recellularization with MSCs and myogenesis-induced MSCs demonstrated the formation of a muscle-like structure after 8 weeks of heterotopic transplantation within recipient rabbit omentum (78).

A more recent study by Moser et al. attempted canine larynx recellularization with multiple cell populations (8). Procured larynxes were decellularized and recellularized all within a custom-made perfusion bioreactor. Notable in this study is their strategy of recellularization using three distinct cell populations: human airway basal cells, primary human myoblasts, and HUVECs. To facilitate their approach, the investigators validated a tailored culture media, termed “multi-cell growth medium,” that permitted concurrent culture of all three cell populations. Additional experiments with whole larynx HUVEC cell seeding supported evidence of endothelialization to different vascular calibers after perfusion culture of the scaffold for three days (8). In both the two mentioned cases of Hou et al., and Moser et al. however, no orthotopic transplantation of the larynx was performed thus leaving open the question of whether functional recapitulation of a tissue engineered larynx is achievable.

A tissue engineered trachea has received much attention as a clinically viable solution in long-segment tracheal replacement. However, attempts to regenerate a functional trachea using naturally derived scaffolds and exogenous cells have seen mixed success and recent clinical studies have cast doubt on its feasibility (70). Nonetheless, work with trachea regeneration offers pertinent insights that can be applied to multi-tissue recellularization needed in VCA engineering. Using the pig as a clinically relevant animal model, Haykal et al. applied a novel double chamber perfusion bioreactor to seed long-segment decellularized tracheal scaffolds with separate luminal and external compartments for seeding of tracheal epithelial cells and bone marrow-derived MSCs, respectively. Over an ex vivo maturation period of 72 h, recellularization of the trachea within this double-chamber perfusion bioreactor showed improved cell attachment and distribution compared to traditional static-seeding methods without perfusion (9). Additional advancement of porcine trachea regeneration by Aoki et al. described a chimeric trachea recellularization approach characterized by partial decellularization of the donor tracheal epithelium while preserving an immune-privileged cartilage compartment (10). Subsequent re-epithelialization of the trachea scaffold using exogenous human bronchial epithelial cells matured in an ex vivo perfusion bioreactor over 7 days facilitated cell attachment, viability and growth. Altogether, these two descriptions of tracheal engineering that utilize dynamic perfusion seeding within a double chamber bioreactor and a recellularization approach based on donor-recipient tissue chimerism are potential concepts that may be adopted to the broader scope of recellularization of other complex VCA tissues.

Jank et al. conducted successful perfusion decellularization of a rat forelimb and primate forelimb using 1% SDS and 1% Triton X-100. Preservation of vessels, nerve, bone, muscle, and skin was observed histologically. Biomechanical properties of the composite forelimb scaffold showed decellularization did not affect joint flexibility nor mechanical, geometric, and mineral properties of the bone. Perfusion recellularization was conducted in several phases using co-cultures and tri-cultures of HUVECs, C2C12 mouse myoblasts, and mouse embryonic fibroblasts to accommodate regeneration of different tissue compartments. The grafts were perfused and matured in a biomimetic bioreactor system with electrical stimulation for 21 days total. Perfusion and injection seeding strategies were tested separately with myoblasts. Perfusion seeding resulted in marginal distribution of myoblasts in the muscle ECM with most cells retained in the vasculature whereas injection seeding allowed cell engraftment and formation of muscle-like tissue. While this allowed targeted cell delivery, it also resulted in apoptosis at injection sites and disruption to the ECM. This highlights important considerations in selecting appropriate seeding strategies for recellularizing a physiologically dense tissue such as skeletal muscle. Functional testing using isometric force contraction of recellularized muscle showed some tetanic contractile force response relative to native. Upon orthotopic limb transplantation using recellularized forelimbs, the arterial network filled with blood however venous return was not reported. Flexion was observed at the wrist and metacarpophalangeal joints upon electrical stimulation of the transplanted recellularized grafts. Electrical stimulation can evidently allow enhanced myofiber alignment and muscle maturation and while authors suggest electrical stimulation enhanced myofiber alignment in the endomysium during cell growth and differentiation phases, recellularization without electrical stimulation was not conducted to show the extent of muscle regeneration with or without it (51, 65).

Penile tissues are a recent addition to VCA tissue engineering and present unique compositions and functional needs for regeneration. The penile scaffold features corpora cavernosa required for erectile function, spongy corpus spongiosum tissue which surrounds the urethra, and additional fibrous and fascial layers surrounding these structures along with a skin/adipose tissue enveloped exteriorly. This penile graft houses heterogenous and unique tissue compartments.

Tan et al. pursued perfusion decellularization and recellularization of human penile scaffolds. Decellularization was conducted in a hybrid fashion via micro-arterial perfusion and urethral perfusion using 1% SDS with slow mechanical agitation simultaneously. An intact tissue ECM architecture was maintained post-decellularization along with low residual DNA levels (less than 50 ng/mg), maintenance of collagen, laminin, fibronectin, and growth factors: VEGF, endothelial growth factor, and transforming growth factor beta-1. A perfusable vascular tree was also observed. Human adipose-derived stromal vascular fraction (SVF) cells were statically seeded on sectioned slices of the decellularized scaffolds and cultured for 28 days. Cells showed adherence, viability, and proliferative capabilities in the cavernosa, tunica, and urethra. Cells showed lining along luminal structures in the cavernosa as well (53). Evidence of early formation of vascular microtubes and increased expression of endothelial marker, von Willebrand factor, in the reseeded cavernosa after 28 days of in vitro scaffold culture was reported (53). Derived from discarded human-derived adipose tissues, the SVF contains a heterogenous cell combination such as adipose derived stem cells (ADSCs), endothelial primary cells, endothelial progenitor cells, immune cells, smooth muscle cells, and pericytes. The varied cell composition of SVF have been shown to exhibit beneficial immunomodulatory and angiogenic properties, which can contribute to their regenerative and therapeutic potential (79, 80). Thus far, Tan et al.’s study is the only study that has used SVF for recellularization.

While Tan et al. focused on the composite scaffold, some authors have previously focused on tissue-engineering corpora cavernosum tissue given the lack of a suitable substitute for penile reconstruction (53, 81). Ji et al. evaluated the use of muscle-derived stem cells for recellularizing acellular corpora cavernosa matrices. Tissues were obtained from male rabbit penises and subjected to decellularization with 2% Triton X-100 and 0.1% ammonium hydroxide. Tissues were simultaneously agitated and decellularized for 10 days. Tissue structure and organization was maintained post-decellularization although cavernosal sinus lacunae were irregular and larger post-decellularization relative to native tissue. Muscle-derived stem cells were isolated and statically seeded for recellularization. The cells became aggregated in the corpus cavernosum 5 days after recellularization. While cell proliferation was observed, some cell death was noted. Authors used the expression of α-smooth muscle actin (α-SMA) as an indication for muscle-derived stem cell differentiation. Decellularized tissues were also transplanted in normal rabbits and evaluated 2-, 4-, and 6-months post-transplantation. Neovascularization was noted at every timepoint. After 4 months, cell proliferation and cell distribution was observed however no differences could be observed after 6 months (81). Ji et al.’s study was extended by An et al. where authors used overexpression of VEGF to enhance contractility of recellularized corpora cavernosa tissues which were obtained using the same protocol (71). Muscle-derived stem cells were transfected with VEGF lentiviral gene vector and seeded onto decellularized cavernosa tissues. VEGF-expressing cells elucidated better growth and engraftment relative to non-transfected cells. α-SMA for cell differentiation and endothelial markers, CD31 and vWF, were increased in scaffolds recellularized with VEGF-expressing cells (71). Along with establishing successful de- and recellularization methods for the corpus cavernosum, these authors also highlighted the potential use of growth factor supplementation to enhance recellularization.

While the urethra has not been discussed for VCA, it is worth consideration given its vascularity and unique dual blood supply that allows it to be mobilized and divided to construct a fasciocutaneous onlay flap for urethral reconstruction (72). Further, it contains an internal mucosal layer, a middle smooth muscle layer, and an external urethral sphincter comprised of skeletal muscle myofibers, altogether introducing tissue complexity that requires customized recellularization approaches.

Simões et al. conducted the first perfusion decellularization and recellularization study of whole porcine urethras. Authors identified that there is a lack of whole urethral decellularization efficiency which hinders the ability to bioengineer physiologically relevant urethral tissues. Porcine urethras were cannulated at their most distal region and perfused with 0.5% and 1% SDS. Cell removal was observed and ECM characterization showed retention of various proteins albeit with a noted ∼50% reduction in elastin content. For recellularization, cross-sections of acellular urethra scaffolds containing the mucosa, smooth and skeletal muscle ECM were produced using a cryostat and conditioned in culture medium prior to cell seeding. Static seeding was performed using highly concentrated cell suspensions of the following cell types: human muscle progenitor cells, human bone marrow MSCs, L929 fibroblasts, and SVF. All cell types were seeded separately in individual experiments and were selected due to their role in muscle and sphincter formation, and tissue regeneration. Bioscaffolds showed cell engraftment, viability, and proliferation up to 2 weeks across all cell types. Muscle progenitor cells showed fusion and ability to form multinucleated fibers, expressing markers of skeletal muscle. Spontaneous contraction could also be observed upon myogenic induction (52). While these recellularization studies were conducted in vitro, they substantiate the potential of regenerating different components of the urethral tissue that can help support regeneration of larger tissues.

The uterus comprises of an endometrium with endothelial cell lining, surrounded by muscle-cell rich myometrium and a perimetrium comprising of epithelial cells. Miyazaki & Maruyama conducted decellularization of rat uteri via aortic perfusion using an SDS protocol from previous liver decellularization studies involving progressively increasing SDS concentrations from 0.1%–1%. A perfusable and maintained tissue ECM architecture was observed. For recellularization, authors injected a combination of primary neonatal and adult uterine cells, and rat MSCs throughout the uterine wall. Collagen gel was used to cover the graft post-injections to prevent cell leakage. The graft was then perfused in a bioreactor system. Endometrial-like tissue formation was observed and partial cellular distribution in the matrix by Day 3. Vessel-like structures positive for CD31 and smooth muscle actin were observed. The graft became increasing atrophic by Day 6 however was sustained until Day 10 (82).

Campo et al. extended uterine decellularization and recellularization to a porcine model, using both fresh and freeze-thawed uteri. Decellularization entailed comparison of two SDS and Triton X-100 protocols whereby no difference was found in terms of ECM characterization and vascular patency. Acellular disk scaffolds were created by taking punch biopsies of the decellularized uteri tissue, with the endometrial tissue facing upwards. Human endometrial stem cells were statically seeded and cultured for a total of 12 days under hypoxic conditions. Contraction of the recellularized tissue was observed 3–4 days post-seeding. The recellularized disks arranged into 3D structures 9–12 days post-seeding. Endometrial intracellular proteins, vimentin and cytokeratin, were detected (49).

Upscaling into an ovine model, Tiemann et al. employed a sodium-deoxycholate (SDC) detergent protocol for decellularization of ovine uteri after trialing three different decellularization protocols. Vascular patency and ECM architectural maintenance was observably superior in the SDC approach. Interestingly, authors froze isolated uteri at −20 °C before beginning recellularization. Rings of the decellularized tissue were made using a cryostat and recellularized using sheep fetal bone marrow stem cells. Cells were injected into the ring of tissue, either with 1 × 10⁶ or 1 × 10⁷ cells. Rings were cultured until Day 3 and Day 14. Regardless of seeding density, cells remained in the superficial layers, at the injection site. Cells showed expression for proliferation marker, Ki67, and with additional endometrial surface protein markers. No differences were seen in expression between Day 3 and Day 14 (55). Despite the lack of cell migration into deeper parts of the ovine tissue, cells remained viable for a two-week period suggesting a non-toxic culturing environment. Further, the use of an ovine model is evidently suitable for extrapolating translatable findings as the uteri anatomy and size is similar to that of humans relative to cow and porcine uteri.