Abstract

Absolute uterine factor infertility, or the absence of a functional uterus, has a prevalence of 3%–5% in the general population. Despite the great strides being made in reproductive medicine, patients diagnosed with absolute uterine factor infertility remain untreatable. The only available solution has been gestational surrogacy, but recently the Brannström group presented a viable alternative by reporting the first successful live birth after uterus transplantation. Similar to other transplantations, this approach has inherent limitations such as the paucity of donor organs and the need for long-term immunosuppression. Whole organ de- and recellularization, a novel tissue engineering approach within the field of regenerative medicine, could eventually provide another solution. Several groups have described animal models in which they have performed decellularization of whole uteri, while maintaining the extracellular matrix to enable recellularization attempts. Our work offers a new perspective; in decellularizing the porcine uterus, this constitutes the first pilot study using large whole reproductive organs. We demonstrated the preservation of a reusable/functional extracellular matrix while maintaining its vascular network. Furthermore, we report the first use of human side population stem cells in the successful recellularization of small acellular disk scaffolds procured from the decellularized organs. To conclude, this research opens new avenues in whole uterus bioengineering, opening the way towards the transplantation of functional bioengineered uteri into humans.

Introduction

Infertility is defined as the inability to achieve clinical pregnancy after one year of regular unprotected intercourse and has been recognized as a worldwide public health issue by the World Health Organization [1]. This medical condition affects up to 14% of couples at reproductive age, and it is estimated that the cause is predominantly female origin in 38% of the cases [2,3] (data from World Health Organization). The percentage tends to be higher in developed countries, mainly due to a higher percentage of toxic habits, uterine pathologies, and the delaying of motherhood [25]. The most common identifiable female causes are ovulatory disorders and endometriosis; great progress has been made in reproductive medicine to treat these pathologies. Despite important advances, there remains a group of women suffering from absolute uterine factor infertility (AUFI), a condition for which there has been no available treatment, until now. Absolute uterine factor infertility can be described as the absence of a functional uterus and has a prevalence of 3%–5% in the general population [6]. The congenital types of AUFI include Mayer-Rokitansky-Küster-Hauser syndrome, uterine hypoplasia, and uterine malformation. Acquired AUFI is caused by hysterectomy due to malignant uterine tumor, benign diseases (including leiomyoma and adenomyosis), postpartum hemorrhage, and loss of fertility due to intrauterine adhesions (Asherman's syndrome).

Currently, the only solution available is gestational surrogacy, which comes with its own set of economic, legal, and ethical challenges/problems [7]. However, uterine transplantation may provide an alternative to enable these women to deliver a child. Recently, the Brannström group published the first successful live birth after uterus transplantation, providing the proof-of-concept for uterus transplantation as a possible treatment [8]. Although it is a milestone in reproductive medicine, the transplantation of the uterus is inherently limited, as with other transplantations, by several problems, such as the lack of donor organs and the need for long-term immunosuppression after the transplantation.

In the last decade, rapid progress has been achieved with biomaterial development and tissue engineering. The once-limited materials (completely lacking a functional vascular network) to generate solid organs are now easily created from scratch using novel biomaterials and techniques complemented by the recipient's cells or with the replacement of the cellular compartment of existing donor organs (and as such their cellular antigens) to avoid immune rejection through the technique of decellularization (DC) followed by recellularization (RC) with several types of cells (autologous stem cells, allogeneic stem cells, etc.). The use of decellularized matrix has been proved in several tissues like esophagus [9], trachea [10], bladder [11], arteries [12], skin [13], heart [14], liver [15], lung [16], and vagina [17].

Badylack et al. defined the procedure of DC as “to efficiently remove all cellular and nuclear material while minimizing any adverse effect on the composition, biological activity, and mechanical integrity of the remaining extracellular matrix (ECM)” [18]. Through DC biologic scaffolds are created. These scaffolds maintain their native ultrastructure and composition with the presence of several growth factors, microvasculature, and structural components, which contributes to an adequate cell mitogenesis, chemotaxis, cell differentiation, and constructive host tissue remodeling response, among others [19]. Decellularization could be useful in a variety of reconstructive surgical applications and is increasingly used in regenerative medicine strategies for tissue and organ replacement [19].

Several recent reports described the potential use of whole-organ bioengineering in the uterus [2022]. Maruyama et al. developed a DC protocol for the rat uterus, and investigated the potential of the resulting matrix for regeneration and reconstruction of the uterine tissues in vivo and in vitro [20]. By perfusing with Sodium dodecyl sulfate (SDS) and Triton X-100, they produced a macroscopically translucent acellular organ lacking nuclei and typical endometrial intracellular proteins such as vimentin, cytokeratin, and α-smooth muscle actin. Microvasculature (injection with dye) and ECM (presence of collagen type I and laminin) remained intact. The functionalization of the decellularized organ was achieved by the injection of uterine and mesenchymal stem cells and the subsequent maintenance in vitro. After 3 days, the matrix presented endometrium-like tissue. The regenerative ability in vivo was demonstrated by using excision/replacement technique with a drop of pregnancy rates to 75%. In parallel, Hellström et al. compared three different protocols for whole uterus DC in the rat. They demonstrated that a freeze-thaw (F/T) step, with repeated cycles of Triton X-100 and Dimethyl sulfoxide (DMSO), gave the best results. Also, the use of sodium deoxycholate produced a good candidate scaffold for future uterus engineering. The potential use of these scaffolds for RC and future in vivo applications was demonstrated by successful removal of major histocompatibility complex class I and II antigens [21].

Given these advances, bioengineering based on DC and RC could become an interesting subfield of tissue engineering, especially for future applications in the regeneration of whole or partial uterus tissue giving rise to novel therapeutic solutions to treat AUFI.

The main objective of this study was to scale up previous research by developing for the first time a whole-organ DC protocol usable for large reproductive organs. Moreover, we compared and assessed the effect of F/T step vs. untreated fresh (F) organs. The major components of the generated ECM were tested to demonstrate the possible functional use of these matrices. The created scaffolds were also analyzed for its ultrastructural integrity and vascular network patency using established and novel techniques. The in vitro biocompatibility of the scaffolds was confirmed in RC experiments using human endometrial stem cell lines. One of the main challenges remaining is to repopulate uterine pig decellularized organs partially with human endometrial stem cells to generate tissue-specific cells in the new scaffolds.

Material and methods

Porcine uterus and cannulation

The uteri were extracted from pigs weighing approximately 220 kg; all organs were donated for scientific research by the Mercavalencia slaughterhouse (complying with the ISO 9001 quality management system). They were sacrificed by exposure to increasing concentrations of CO2, which led to metabolic acidosis, reduction of intracellular pH, and resulted in gradual loss of consciousness and sensibility. Exsanguination was performed by transection of the jugular vein. The uteri were collected early in the morning from the local slaughterhouse (Mercavalencia, Spain) and preserved at 4°C during transport (15 min). In total six uteri were selected, three for each protocol, based on their (similar) size and general aspect with much attention given to their vascular system (

describes the workflow for each sample). We cannulated the uterine artery of a single horn (decellularized horn), leaving the other not perfused. After cleaning the artery from surrounding tissue, 20-G cannulas (BD, Apósitos Navarro, S.L.) were inserted whilst pushing phosphate-buffered saline by syringe (PBS, Sigma-Aldrich; pH 7.4) to dilate the artery. After guaranteeing that the cannula was properly inserted and positioned, and blood (clots) was exiting the uterine vein when infusing the PBS remaining in the syringe, the cannula was fixed into place using braided silk sutures (Laboratorio Aragó, Barcelona, Spain).

Decellularization of the porcine uterus

The now cannulated uterine horn was carefully attached to a peristaltic pump (Cole-Parmer instruments, Chicago, USA) using L/S 16 tubing (Masterflex, Fisher scientific) for DC. An initial perfusion of PBS for one hour was done to remove the remaining blood and cell debris. The perfusion speed for all protocols was set at a physiological flow rate of 15 mL/min [23]. Two DC protocols, either in freeze/thawed (F/T, n = 3) or fresh (F, n = 3) pig uteri using chemical and/or physical agents, were compared (see Table 1 for full protocol).

Table 1.

Decellularization protocol of freeze/thawed and fresh samples. Starting with the preparation of the organ the protocol is unchanged. A perfusion speed of 15 mL/min was used throughout.

 Protocol Duration 
Sample preparation Freeze/thawed uterus (at −20°C, one week max) (F/T) or fresh uterus (F)  
Day 1 Cycle 1 PBS 1 hour 
  0.1% SDS 18 hours 
  dH230 min 
  1% Triton X-100 30 min 
  PBS 5 hours 
Day 2 Cycle 2 0.1% SDS 18 hours 
  dH230 min 
  1% Triton X-100 30 min 
  PBS 5 hours 
 Protocol Duration 
Sample preparation Freeze/thawed uterus (at −20°C, one week max) (F/T) or fresh uterus (F)  
Day 1 Cycle 1 PBS 1 hour 
  0.1% SDS 18 hours 
  dH230 min 
  1% Triton X-100 30 min 
  PBS 5 hours 
Day 2 Cycle 2 0.1% SDS 18 hours 
  dH230 min 
  1% Triton X-100 30 min 
  PBS 5 hours 

Protocol 1: Freeze/thawed samples (P1:F/T). The uterus (n = 3) was frozen after cannulation, thawed, and underwent two identical cycles after the initial perfusion with PBS. Each cycle took 24 hours and consisted of four perfusion steps: first we perfused for 18 hours (overnight) with a 0.1% SDS solution (Sigma-Aldrich), followed by 30 min of dH2O, 30 min of 1% Triton X-100 (Sigma-Aldrich), and finally five hours of PBS.

Protocol 2: Fresh samples (P2:F). The second protocol also used the same cycles twice, consisting of perfusion steps with SDS, dH2O, Triton X-100, and PBS, but in contrast to the first protocol, we did not use an F/T step immediately after cannulation (n = 3).

Histological analysis

Ring-shaped or circular segments with a thickness of about 5 mm were taken from four locations for each uterus subjected to the protocols. These samples came from the perfused horn at equidistant positions between the uterine body and the oviduct to assure homogeneity in the resulting tissue (

). These segments of decellularized samples and segments of fresh control were fixed for 24 hours in 4% paraformaldehyde (PFA) at 4°C for histological staining and immunofluorescence. Subsequently, they were transferred to 70% ethanol overnight at 4°C and dehydrated in graded alcohol. This was followed by immersion in xylenes and embedding in paraffin wax. The embedded samples were serially sectioned (4 μm) on a microtome (HM 310, Microm) and mounted onto 21 glass slides (Superfrost plus, Thermo Scientific). For each histological analysis, three slides of the serial cuts were selected. To assess the efficacy of the DC protocols and the integrity of the ECM, we rehydrated the sections and stained the slides with hematoxylin and eosin (H&E), Masson's Trichrome, and Alcian blue using standard protocols. This was to demonstrate the presence of cellular components, collagen, and sulphated glycosaminoglycans, respectively. For the detection of nuclear DNA, mounting media containing 6-diamidino-2-phenylindole (DAPI, Thermo-fisher Scientific) was used. All images, both for bright-field and fluorescence microscopy, were taken with a Nikon eclipse 80i microscope.

Immunofluorescence analysis

Immunofluorescence of five highly constitutive proteins of the ECM was performed on decellularized samples procured from both protocols and compared to an untreated, fresh uterus as control to evaluate the integrity of the ECM. From each uterus (a total of seven slides, three from each protocol plus control), one slide of the serial cuts was deparaffinized and rehydrated as described above. Mouse monoclonal primary antibodies for fibronectin and elastin (final concentration for both: 1/300) and rabbit polyclonal primary antibodies for collagens type I and IV and laminin (final concentration for all: 1/200) were used (Abcam antibodies: #ab6328, #ab9519, #ab82504, #ab6586, and #ab11575, respectively). Enzymatic antigen retrieval was performed for fibronectin and elastin: slides were incubated for 10 min at 37°C with a 12 units/mL proteinase K solution in a humid chamber, followed by 10 min at room temperature and rinsing with TE buffer (pH = 8, Tris and Ethylenediaminetetraacetic acid (EDTA) buffer). Antigen retrieval of the remaining proteins was done by incubation in citric acid solution (10 mM sodium citrate, 0.05% Tween-20, pH 6.0) at 95°C for 20 min and a cooling step on ice for another 20 min, after which we rinsed with PBS (TE buffer in case of Laminin). Blocking against nonspecific binding was done by using a solution consisting of 5% bovine serum albumin (BSA), 5% normal goat serum (NGS), and 0.4% Triton X-100 in PBS for one hour at room temperature for collagen IV and at 37°C for collagen I. A 5% BSA/5% NGS/0.4% Triton X-100 in TE buffer mixture was used for one hour at room temperature in the case of elastin and fibronectin, where for laminin this was performed at 37°C. Primary antibodies were diluted to the required concentrations in their respective blocking buffer at 1%. After rinsing, the slides were incubated with secondary antibodies with Alexa Fluor 488 green-fluorescent dye at 1/500 (Invitrogen) in PBS or TE buffer; for fibronectin and elastin, the antibody used was goat antimouse (A21121, Invitrogen). In the other cases, a goat antirabbit secondary antibody was used (A11034, Life Sciences). Slides were visualized under a fluorescence microscope and Olympus FV1000 confocal microscope. Three-dimensional stacks of the samples were also made.

DNA quantification and fragment-size determination

The circular segments were put on filter paper to remove the excess liquid, cut longitudinally in one side to lay it flat. From the obtained slab, cubic pieces weighing less than 25 mg were procured. These were cut up by hand and DNA was extracted by using a commercial kit (DNeasy Blood & Tissue, Qiagen) following the protocol provided. The measurements of the DNA concentration were done on a Nanodrop (Thermo Scientific). The sample from the decellularized horn with the highest amount of DNA from each uterus (P1, n = 3; P2, n = 3) was loaded on a 2% agarose gel containing Gel Red nucleic acid stain (Biotium #41003) for a total runtime of 90 min at 100 V. 1000 ng DNA of each sample were loaded; a 1 kb plus DNA ladder (Invitrogen) was used for comparison.

Protein extraction and quantification

Similar to the DNA quantification, circular fragments were taken, dried on filter paper, weighed, and minced. Protein extraction was done by using a modified Laemli buffer (0.125M Tris HCl, 4% SDS 10% β-mercaptoethanol) for 48 hours at 37°C under agitation. The Pierce BCA protein assay kit (Thermo-Fisher Scientific) was used for total protein quantification following the manufacturer's protocol and using a spectramax 190 (Molecular Devices).

Scanning electron microscopy and transmission electron microscopy

Ultrastructure was investigated by SEM and TEM. The decellularized samples from each uterus were fixed in 4% PFA overnight and stored in 50% EtOH after another overnight step with 70% EtOH. The selected samples (only one location) were subjected to critical point drying and Au-Pd sputtering and ultimately put under an FEG Hitachi S-4100 scanning electron microscope. For TEM samples were postfixed in 1% osmium tetroxide in 100 mM phosphate buffer and embedded in resin. After ultrasectioning, a JEOL JEM1010 transmission electron microscope was used for TEM.

Vascular corrosion cast

To assess the integrity of the vascular tree Batson's No.17 plastic replica and corrosion kit (Polysciences, Inc., ImmunoStep SL) was used on one uterus per protocol and on one control. The perfusion mixture was made as described by the protocol provided, two different perfusion speeds (2.5 and 15 mL/min) were used for the corrosion cast of the control. The casts of the F/T Protocol (P1) and F Protocol (P2) samples were perfused at speeds of 2.5 and 15 mL/min, respectively. The 2.5 mL/min perfusion speed was selected as the lower speed after comparing perfusion speeds at previous experiments done with the liver [24]. After perfusion, all the veins and arteries of the uteri were clamped and left overnight at 4°C for polymerization. The tissue was macerated using a 10% NaOH mixture overnight; this step was repeated until the mixture remained clear, indicating that all the tissue was successfully removed. Circular cuts were made of the vascular mold and opened; these samples were photographed under a stereomicroscope and submitted to Au-Pd sputtering for SEM.

Recellularization of uterine extracellular matrix disks

Biopsies from decellularized organs (P2) oriented with the endometrium facing up were embedded in Tissue-Tek O.C.T. Compound (OCT, Sakura Finetek USA Inc., VWR). A maximum of four cuts at 100 μm were made with the cryotome. Punch biopsies with a diameter of 5 mm were made and put in a cooled chamber slide (μ-Slide 8 Well, Ibidi). Sterilization was performed by exposure to UV light for two hours, while ensure the scaffolds did not dry out. These circular scaffolds were seeded with a viscous pellet consisting of a mixture of stromal and epithelial side population (SP) stem cells from human endometrial origin (4/5 ICE6 and 1/5 ICE7, respectively, corresponding to their physiological ratio in the endometrium). These endometrial stem cells, named ICE, were isolated in 2010 by flow cytometry using the well-known efflux Hoechst method, and later characterized for stem cell features by, among others, the expression of mesenchymal markers CD73, CD90, and CD105. [25]. Next, 0.5 million of ICE mixed stem cells from the 11th passage were seeded and cultured on the scaffolds under hypoxic conditions. Endometrial culture medium (DMEM, 10% fetal bovine serum, 2mM L-glutamine, 1/1000 Streptomycin/Penicillin/Amphotericin) was changed after 3, 6, and 9 days of culture. After this the organoid-like structure was embedded in paraffin and serial cuts (4 μm) were made. Successful RC and correct cell differentiation was assessed by H&E staining and indirect immunofluorescence against human vimentin [1:200, antivimentin antibody (ab8069), Abcam] and human anticytokeratin [1:500, anticytokeratin 18 antibody (ab52948), Abcam]. Goat antimouse Alexa-fluor 488 (A21121, Thermo-Fisher) and goat antirabbit Alexa-fluor 555 (A21428, Thermo-Fisher) were used as secondary antibodies. Positive (human endometrium) and negative (mouse kidney) controls were used in this methodology (data not shown). Standard blocking solution and antigen retrieval were performed along with this protocol. Slides were examined under a Nikon Eclipse 80i fluorescence microscope.

Statistical analysis

Statistical analysis was performed to identify any significant differences between the protocols and the fresh control at DNA and protein levels. This was done first by using ANOVA; if the means of the three populations were not considered as equal, we followed up with a posthoc test (Bonferroni method). A P-value obtained in a two-tailed test ≤0.05 was considered statistically significant.

Results

Visual, histological, and quantitative assessment of successful decellularization

Upon visual inspection of the uterine horn, a color change from pink to opaque white was observed, indicating a successful DC after 49 hours for both protocols (Figure 1A and B). No difference in opacity or color was seen between protocols (not shown). Hematoxylin and eosin and DAPI staining demonstrated the absence of nuclei and, as such, indicated the efficient removal of DNA while preserving tissue structure (Figure 1C and D). Further, there was no noticeable difference between the replications (n = 3) along the four particular sample locations. Hence, in further figures only one representative replicate is shown. Histology showed that the ECM structure appeared more uniform after the fresh protocol. The successful removal of cellular material was further confirmed by the quantification of residual DNA; a significant reduction from the normal DNA content of a fresh uterus (1571.41 ± 427.17 ng DNA per mg wet tissue) was observed. The remaining DNA for F/T and F protocol was measured as 139.604 ± 89.561 and 40.792 ± 16.462 ng DNA per mg wet tissue, respectively, which corresponds to a decrease to 9.06% and 2.65% compared to the unprocessed uterus (Figure 1E). The Bonferroni method showed no significant difference between the two protocols with a P-value <0.05. The quantification of the total ECM protein fraction displayed a similar trend: a statistically significant drop from 8.35 ± 1.8 μg protein/mg wet tissue for the control to 3.26 ± 1.98 and 2.5 ± 1.64 μg protein/mg wet tissue, for both protocols respectively was observed (Figure 1F).

Figure 1.

Visual, histological, and quantitative assessment of successful decellularization (DC). Macroscopic inspection of uterine horns before DC protocols (A) and after DC protocols (B). Pictures showing the hematoxylin and eosin (C) and DAPI (D) stainings of n = 3 repeats/replications for both protocols [Freeze/thaw (P1): F/T and Fresh (P2): F]. Quantification of residual DNA showing a drop to 9.06% and 2.65% for the F/T and F protocol, respectively (E). Quantification of remaining protein fraction showing a drop to 38.98% and 29.97% (F). Scale bars: 100 μm (C–D). Significance levels = *P > 0.05.

Figure 1.

Visual, histological, and quantitative assessment of successful decellularization (DC). Macroscopic inspection of uterine horns before DC protocols (A) and after DC protocols (B). Pictures showing the hematoxylin and eosin (C) and DAPI (D) stainings of n = 3 repeats/replications for both protocols [Freeze/thaw (P1): F/T and Fresh (P2): F]. Quantification of residual DNA showing a drop to 9.06% and 2.65% for the F/T and F protocol, respectively (E). Quantification of remaining protein fraction showing a drop to 38.98% and 29.97% (F). Scale bars: 100 μm (C–D). Significance levels = *P > 0.05.

Electron microscopy and histological analysis of (ultra) structure of the decellularized uterine extracellular matrix

Scanning electron micrographs at low magnification showed the surface topography of both the epithelium and upper part of the stromal fraction of the endometrium. The epithelial layer in tissue from the fresh protocol appeared more damaged, but this was due to manipulation of the sample (Figure 2A, top row). At higher magnification, it was clear that the fibers and surface architecture retained their appearance in both protocols and vascular conduits kept their conformation (Figure 2A, bottom row). Transmission electron micrographs showed the collagen fibrils, which maintained their striated patterns and were abundant in both orientations throughout the tissue (Figure 2B). No noticeable differences at ultrastructural level between the protocols were observed. Masson's trichrome staining demonstrated the preservation of collagens following DC. The red staining of the cytoplasm was completely removed after both protocols, while the blue collagen fibers remained. The qualitative analysis of sulphated glycosaminoglycan proteins by Alcian blue staining showed a widespread distribution with a higher signal at the epithelial layer of the endometrium and secretory glands. After DC, only the widespread distribution remained in both protocols with a noticeably lower signal at the epithelium.

Figure 2.

(Ultra-) structural and histological comparison after DC protocol with both conditions (P1 and P2). Scanning electron micrographs at lower magnification (300× magnification) showing epithelial (E) and stromal (S) fractions (upper panel A) and at higher magnification to demonstrate the condition of the fibers (800× magnification, lower panel A); arrowheads indicate blood vessels. Transmission electron micrographs showing collagens and elastin fiber bundles (arrowheads), (3000× magnification, scale bars: 1 μm, B). Masson's trichrome (C) and Alcian blue (D) staining's for collagen and sGAG detection, respectively. In all cases, the positive control corresponds to the pig uterus before DC. Scale bars: 100 μm (C–D).

Figure 2.

(Ultra-) structural and histological comparison after DC protocol with both conditions (P1 and P2). Scanning electron micrographs at lower magnification (300× magnification) showing epithelial (E) and stromal (S) fractions (upper panel A) and at higher magnification to demonstrate the condition of the fibers (800× magnification, lower panel A); arrowheads indicate blood vessels. Transmission electron micrographs showing collagens and elastin fiber bundles (arrowheads), (3000× magnification, scale bars: 1 μm, B). Masson's trichrome (C) and Alcian blue (D) staining's for collagen and sGAG detection, respectively. In all cases, the positive control corresponds to the pig uterus before DC. Scale bars: 100 μm (C–D).

Immunofluorescence of major components of the extracellular matrix

The results showed the presence of major ECM components before and after DC (Figure 3). Collagen type I, the most abundant component in collagen fibers, gives structural support to resident cells; it remained evenly dispersed in ECM. Elastin, another important structural constituent of the ECM, was mainly found in the myometrium, basal endometrium, and around arteries. After DC there was a noticeably weaker signal in the endometrium (not shown), while myometrium and arteries remained unaffected. Two of the prominent components of the basement membrane, collagen IV and laminin (important in cellular processes like cell differentiation, migration, and adhesion) were detected before and after DC in endothelial layers around blood vessels and glandular structures in both layers. Collagen IV was also observed in the myometrium. Fibronectin, sometimes referred to as the “master organizer” for the biogenesis of the ECM [26], was unaffected and evenly dispersed in ECM, before and after both protocols.

Figure 3.

Immunofluorescence staining of the major ECM components after DC. Images showing blue signal for nucleus (DAPI) and green signal for collagen I (A), collagen IV (B), elastin (C), fibronectin (D), and laminin (E) proteins. Positive control used in all the cases corresponds to the pig uterus before DC. Intensity projection over Z-axis. (40× magnification, scale bars: 5000 μm (C–D).

Figure 3.

Immunofluorescence staining of the major ECM components after DC. Images showing blue signal for nucleus (DAPI) and green signal for collagen I (A), collagen IV (B), elastin (C), fibronectin (D), and laminin (E) proteins. Positive control used in all the cases corresponds to the pig uterus before DC. Intensity projection over Z-axis. (40× magnification, scale bars: 5000 μm (C–D).

Vascular tree network cast of the decellularized organ

We could successfully perfuse four uteri at two different perfusion speeds with the monomer solution to make a vascular corrosion cast (Figure 4). Macroscopic analysis showed a lack of rigidity at the uterine horn after both DC protocols. At many locations the tubular uterine horn did not form a closed circle, in contrast to the control. Both controls were extensively perfused with PBS for one hour to remove all possible clots and blood, but this was not done for the acellular horns. The vascular deformation was observed more clearly when sections were placed under a stereomicroscope. The control also displayed a noticeable white region corresponding to saponified tissue not completely removed at the subepithelial capillary plexus region of the endometrium, which was also present to a lesser extent in the decellularized samples. The diameter of these capillaries are similar to those found in the rat uterus, which average around 7.5 ± 0.4 microns (and up to 18.5 ± 2.5 microns closest to the site of embryo implantation) [27]. When this white zone was investigated at higher magnification, we could appreciate this to be in fact the relatively intact subepithelial capillary plexus, with capillaries at the correct range of thickness in both the control and decellularized uteri.

Figure 4.

Corrosion cast of vascular network of the uterine horn. After maceration of tissue the integrity of vasculature was assessed macroscopically (left column). Sections of the cast were used for stereoscopic close-up (middle column, ext and int referring to external border of perimetrium and inner lumen of the porcine horn) and consequently for scanning electron micrograph of capillaries at the epithelial layer (right column). Pictures from the uterus control (A), F/T DC protocol (B), and F DC protocol (C) representing macroscopic, stereomicroscopic, and SEM, respectively.

Figure 4.

Corrosion cast of vascular network of the uterine horn. After maceration of tissue the integrity of vasculature was assessed macroscopically (left column). Sections of the cast were used for stereoscopic close-up (middle column, ext and int referring to external border of perimetrium and inner lumen of the porcine horn) and consequently for scanning electron micrograph of capillaries at the epithelial layer (right column). Pictures from the uterus control (A), F/T DC protocol (B), and F DC protocol (C) representing macroscopic, stereomicroscopic, and SEM, respectively.

Preliminary recellularization experiments with the generated three-dimensional scaffolds

To demonstrate the feasibility of whole organ RC and the bio-inductive capacities of the generated scaffold we used ECM disks corresponding to P2 as it provided the most uniform ECM (Figure 5A). After seeding human SP cells we saw that the coated scaffolds rolled up and contracted, appearing to form an organoid-like structure (Figure 5B). Hematoxylin and eosin staining showed that the cells were encapsulated in the decellularized scaffold, demonstrating that an interaction between the cells and ECM occurred (Figure 5C). Immunofluorescence signal for vimentin was present in the entire structure and several cytokeratin-positive cells were observed in the organoid-like structure.

Figure 5.

Recellularization process in uterine decellularized disks from whole organ. Workflow for creation of scaffold disks. The decellularized tissue was embedded in OCT, and 100 μm sections of the endometrial fraction (E) punch biopsies were taken and seeded with human stem cells (A). Formation of organoid-like structure after 3–4 days under hypoxic culture conditions (scale bar = 50 μm) (B). Hematoxylin and eosin staining showing reorganization of the human endometrial SP cells (scale bar = 100 and 20 μm, respectively). (C) Immunofluorescence staining of vimentin (green) and cytokeratin (red) after cell seeding of human SP cell lines for 9–12 days on scaffold, nuclear staining appears blue (DAPI) (D).

Figure 5.

Recellularization process in uterine decellularized disks from whole organ. Workflow for creation of scaffold disks. The decellularized tissue was embedded in OCT, and 100 μm sections of the endometrial fraction (E) punch biopsies were taken and seeded with human stem cells (A). Formation of organoid-like structure after 3–4 days under hypoxic culture conditions (scale bar = 50 μm) (B). Hematoxylin and eosin staining showing reorganization of the human endometrial SP cells (scale bar = 100 and 20 μm, respectively). (C) Immunofluorescence staining of vimentin (green) and cytokeratin (red) after cell seeding of human SP cell lines for 9–12 days on scaffold, nuclear staining appears blue (DAPI) (D).

Discussion

Many techniques and technologies are being developed to face one of the biggest challenges concerning modern medicine in the western world, namely end-stage organ failure, where currently the only definitive solution is allogeneic transplantation. One promising line of investigation, the DC and RC of whole organs, circumvents the problem of low availability of donor organs and need for long-term immunosuppression by removing xeno- and allogeneic cellular antigens of organs. This could potentially be a powerful option for the upcoming creation of implantable bioengineered organs. Successful DC has been recently achieved for a multitude of organs and tissues, creating bioscaffolds that are re-implantable in future recipients [914,16,17,28,29]. Here, we present the first study describing the comparison between a fresh method and the effect of an F/T cycle on an efficient DC protocol of a large reproductive organ and followed by RC using human stem cells.

Successful DC has been achieved recently in the field of bioengineering for a multitude of organs and tissues [917]. The main objective of this type of procedure is to create three-dimensional biological acellular matrices also named bioscaffolds that are re-implantable in future recipients [28]. The pig uterus is a bicornuate organ comprised of two horns fused into a uterine body; its size ranges from 30 to 35 cm (for pigs weighing approximately 220 kg). The horns contain the typical histological elements of the uterus: endometrium (epithelial, stromal, and vascular cells), myometrium, and perimetrium. The DC procedure is based on the removal of the native cells while preserving the ECM. Its success depends on tissue density, thickness, and cellularity. In our study, a DC perfusion setup was custom-made, enabling reproducible experiments. Different DC protocols were tested by perfusing one of the uterine arteries at a flowrate of 15 mL/min [23]. Based on these results, two different protocols, focused on the use of freeze/thawed (F/T, P1) vs. fresh (F, P2) organs, were compared. Two identical cycles using SDS and Triton X-100 (see Material and methods) produced a macroscopically acellular matrix. Uterine horns from both protocols displayed a semitransparent and white appearance after perfusion in comparison with controls. As demonstrated in previous studies [19], by using SDS, an ionic detergent, we were also able to effectively remove cell residues with minimal damage to the ECM [30]. Furthermore, the F/T step, a physical DC method, induced cellular lysis and did not noticeably affect the ECM architecture in our experience. Moreover, a previous study reported that a single F/T cycle can reduce adverse immune responses in vascular ECM scaffolds [31], suggesting that this initial step could be an interesting and very convenient option during the transplantation stage. Through our work, we demonstrated that both DC protocols worked satisfactorily and could be used in standardized conditions. The described whole-organ detergent perfusions in this study demonstrated that we can safely step up the size of the decellularized scaffolds, while maintaining the natural three-dimensional structure of the native organs, allowing scientists and clinicians to eventually create a functional uterus.

Using the aforementioned perfusion setup, we were able to further investigate the effects of DC (n = 3 for each protocol). This was done first by demonstrating how our protocols significantly affected the total amount of DNA and proteins after DC. The absence of these vital cellular constituents shows that sufficient amounts of native cell material (and, thus, its epitopes) were removed to consider the organ as a “decellularized scaffold”.

In this investigation, it was of the utmost importance to demonstrate the preservation of the native structure of the ECM including the vasculature after DC. The major components of the ECM, such as collagens I and IV, elastin, laminin, and fibronectin, were preserved with our methodology in spite of the lack of specific quantification of these proteins. The results imply that the basal membrane remained intact under both DC conditions. This is an important observation because these key components act in the microenvironment to provoke changes like cell migration, proliferation, and differentiation [32]. The maintenance of the protein composition in RC processes contributes to support cell homing, cell attachment, tissue integration, development, and normal cellular phenotypic function [33,34]. Additionally, the physical and possible mechanical properties of ECM were demonstrated by the presence of native collagen fibers visualized with Masson's trichrome and Alcian blue staining [35,36]. Altogether, it is apparent that the destructive structural changes to the ECM are reduced to a minimum, offering an excellent/viable bioscaffold for future RC experiments.

Preservation of the vascular network is also critical, first for the efficient perfusion of detergents while avoiding degeneration of veins, arteries, and capillaries; and second for the future use of the circulatory system to repopulate the organ [37]. By making a vascular corrosion cast and combining it with SEM we were able to detect perfusable vascular conduits up to the capillary level, showing proper structural and functional features. The vascular system remained intact after both protocols, which opens up the possibility to recellularize the whole organ by adding cells via perfusion or injection, and creates a system capable of nutrient and gas exchange in the neo organ [15]. The generated scaffolds can then be used to build an organized tissue that mimics the original organ using the native vascular tree.

The final step was to verify the in vitro biocompatibility of the scaffolds. This was done by performing preliminary experiments concerning its bio-inductive properties with the presence and the perseverance of human cells in the generated xenogeneic scaffolds. After a punch biopsy of the endometrial fraction of our decellularized organs and using human stem cells, we were able to identify endometrial (epithelial and stromal fractions) cells in the disks. For this purpose, we seeded human endometrial SP stem cells (cell lines generated in our laboratory [25]) onto disks of decellularized endometrial tissue. After 3–4 days of cell culture in hypoxic conditions the seeded scaffolds rolled up and contracted, after 9–12 days we observed three-dimensional structures, similar to an organoid-like form. The self-organizing aspect of this structure was apparent when the flat disks formed a three-dimensional structure on their own. This structure was further investigated to determine the presence of human cells recapitulating the in vivo tissue architecture. In this context, we demonstrated in vitro the reorganization of the seeded cells onto the scaffold and the presence of vimentin and cytokeratin positive cells after 9–12 days [38].

The main limitation of this study is its preliminary nature in the setting of whole organ DC in large animals, and the challenge is in the optimization of the process for applicability in humans.

Conclusions

This report describes two methods to efficiently obtain whole-uterus scaffolds. Nevertheless, further studies are now required to optimize the functionality of the regenerated organs. To our knowledge, this is the first study to use the entire porcine uterus and to demonstrate the preservation of a reusable ECM, with the maintenance of a vascular network and the feasibility of RC with human stem cells of small decellularized ECM disks. Previously, uterine transplantation was the only available treatment for AUFI. However, could tissue engineering now be a potential tool in reproductive medicine by the promise of creating implantable bioengineered organs?

Supplementary data

Supplementary data are available at BIOLRE online.

Supplemental Figure S1: Diagram describing the workflow followed for each sample, from initial manipulations and application of DC protocols until final sample extraction and characterization. Here, a balance was made between the successful DC and the impact of the protocol on the ECM (NP: not perfused horn, D1-4: decellularized horn 1 to 4).

Acknowledgment

We thank Mercavalencia for providing the samples necessary to perform this study.

References

1.
Zegers-Hochschild
F
,
Adamson
GD
,
de
Mouzon J
,
Ishihara
O
,
Mansour
R
,
Nygren
K
,
Sullivan
E
,
Vanderpoel
S
.
International Committee for Monitoring Assisted Reproductive Technology (ICMART) and the World Health Organization (WHO) revised glossary of ART terminology, 2009
.
Fertil Steril
 
2009
;
92
:
1520
1524
.
2.
Gurunath
S
,
Pandian
Z
,
Anderson
RA
,
Bhattacharya
S
.
Defining infertility–a systematic review of prevalence studies
.
Hum Reprod Update
 
2011
;
17
:
575
588
.
3.
Matzuk
MM
,
Lamb
DJ
.
The biology of infertility: research advances and clinical challenges
.
Nat Med
 
2008
;
14
:
1197
1213
.
4.
Nicoletti
C
,
Tanturri
ML
.
Differences in Delaying Motherhood Across European Countries: Empirical Evidence From the ECHP
 .
Institute for Social and Economic Research, University of Essex
;
ISER Working Papers Number 2005–4
.
5.
Healy
DL
,
Trounson
AO
,
Andersen
AN
.
Female infertility: causes and treatment
.
Lancet
 
1994
;
343
:
1539
1544
.
6.
Kisu
I
,
Banno
K
,
Mihara
M
,
Hara
H
,
Kato
Y
,
Suganuma
N
,
Aoki
D
.
Uterine transplantation in primates: a mini-review of the literature
.
Transplant P
 
2014
;
46
:
1212
1216
.
7.
Shenfield
F
,
Pennings
G
,
Cohen
J
,
Devroey
P
,
de Wert
G
,
Tarlatzis
B
ESHRE Task Force on Ethics and Law
.
ESHRE task force on ethics and law 10: surrogacy
.
Hum Reprod
 
2005
;
20
:
2705
2707
.
8.
Brännström
M
,
Johannesson
L
,
Bokström
H
,
Kvarnström
N
,
Mölne
J
,
Dahm-Kähler
P
,
Enskog
A
,
Milenkovic
M
,
Ekberg
J
,
Diaz-Garcia
C
,
Gäbel
M
,
Hanafy
A
et al
Livebirth after uterus transplantation
.
Lancet
 
2015
;
385
:
607
616
.
9.
Nieponice
A
,
Gilbert
TW
,
Badylak
SF
.
Reinforcement of esophageal anastomoses with an extracellular matrix scaffold in a canine model
.
Ann Thorac Surg
 
2006
;
82
:
2050
2058
.
10.
Hung
S-H
,
Su
C-H
,
Lin
S-E
,
Tseng
H
.
Preliminary experiences in trachea scaffold tissue engineering with segmental organ decellularization
.
Laryngoscope
 
2016
;
126
(
11
):
2520
2527
.
11.
Yoo
JJ
,
Meng
J
,
Oberpenning
F
,
Atala
A
.
Bladder augmentation using allogenic bladder submucosa seeded with cells
.
Urology
 
1998
;
51
:
221
225
.
12.
Dahl
SLM
,
Koh
J
,
Prabhakar
V
,
Niklason
LE
.
Decellularized native and engineered arterial scaffolds for transplantation
.
Cell Transplant
 
2003
;
12
:
659
666
.
13.
Schechner
JS
,
Crane
SK
,
Wang
F
,
Szeglin
AM
,
Tellides
G
,
Lorber
MI
,
Bothwell
ALM
,
Pober
JS
.
Engraftment of a vascularized human skin equivalent
.
FASEB J
 
2003
;
17
:
2250
2256
.
14.
Ott
HC
,
Matthiesen
TS
,
Goh
S-K
,
Black
LD
,
Kren
SM
,
Netoff
TI
,
Taylor
DA
.
Perfusion-decellularized matrix: using nature's platform to engineer a bioartificial heart
.
Nat Med
 
2008
;
14
:
213
221
.
15.
Baptista
PM
,
Siddiqui
MM
,
Lozier
G
,
Rodriguez
SR
,
Atala
A
,
Soker
S
.
The use of whole organ decellularization for the generation of a vascularized liver organoid
.
Hepatology
 
2011
;
53
:
604
617
.
16.
Ott
HC
,
Clippinger
B
,
Conrad
C
,
Schuetz
C
,
Pomerantseva
I
,
Ikonomou
L
,
Kotton
D
,
Vacanti
JP
.
Regeneration and orthotopic transplantation of a bioartificial lung
.
Nat Med
 
2010
;
16
:
927
933
.
17.
Raya-Rivera
AM
,
Esquiliano
D
,
Fierro-Pastrana
R
,
López-Bayghen
E
,
Valencia
P
,
Ordorica-Flores
R
,
Soker
S
,
Yoo
JJ
,
Atala
A
.
Tissue-engineered autologous vaginal organs in patients: a pilot cohort study
.
Lancet
 
2014
;
384
:
329
336
.
18.
Gilbert
T
,
Sellaro
T
,
Badylak
S
.
Decellularization of tissues and organs
.
Biomaterials
 
2006
;
19
:
3675
3683
.
19.
Crapo
PM
,
Gilbert
TW
,
Badylak
SF
.
An overview of tissue and whole organ decellularization processes
.
Biomaterials
 
2011
;
32
:
3233
3243
.
20.
Miyazaki
K
,
Maruyama
T
.
Partial regeneration and reconstruction of the rat uterus through recellularization of a decellularized uterine matrix
.
Biomaterials
 
2014
;
35
:
8791
8800
.
21.
Hellström
M
,
El-Akouri
RR
,
Sihlbom
C
,
Olsson
BM
,
Lengqvist
J
,
Bäckdahl
H
,
Johansson
BR
,
Olausson
M
,
Sumitran-Holgersson
S
,
Brännström
M
.
Towards the development of a bioengineered uterus: comparison of different protocols for rat uterus decellularization
.
Acta Biomater
 
2014
;
10
:
5034
5042
.
22.
Hellström
M
,
Moreno-Moya
JM
,
Bandstein
S
,
Bom
E
,
Akouri
RR
,
Miyazaki
K
,
Maruyama
T
,
Brännström
M
.
Bioengineered uterine tissue supports pregnancy in a rat model
.
Fertil Steril
 
2016
;
106
:
487
496
.
23.
Geisler
K
,
Künzel
J
,
Grundtner
P
,
Müller
A
,
Beckmann
MW
,
Dittrich
R
.
The perfused swine uterus model: long-term perfusion
.
Reprod Biol Endocrin
 
2012
;
10
:
110
120
.
24.
Debbaut
C
,
Monbaliu
D
,
Casteleyn
C
,
Cornillie
P
,
Loo
DV
,
Masschaele
B
,
Pirenne
J
,
Simoens
P
,
Hoorebeke
LV
,
Segers
P
.
From vascular corrosion cast to electrical analog model for the study of human liver hemodynamics and perfusion
.
IEEE T Biomed Eng
 
2011
;
58
:
25
35
.
25.
I
Cervelló
,
Mas
A
,
Gil-Sanchis
C
,
Peris
L
,
Faus
A
,
Saunders
PTK
,
Critchley
HOD
,
Simón
C
.
Reconstruction of endometrium from human endometrial side population cell lines
.
PLoS One
 
2011
;
6
:
e21221
.
26.
Sabatier
L
,
Chen
D
,
Fagotto-Kaufmann
C
,
Hubmacher
D
,
McKee
MD
,
Annis
DS
,
Mosher
DF
,
Reinhardt
DP
.
Fibrillin assembly requires fibronectin
.
Mol Biol Cell
 
2009
;
20
:
846
858
.
27.
Tawia
SA
,
Rogers
P a. W
.
In vivo microscopy of the subepithelial capillary plexus of the endometrium of rats during embryo implantation
.
J Reprod Fertil
 
1992
;
96
:
673
680
.
28.
Peloso
A
,
Dhal
A
,
Zambon
JP
,
Li
P
,
Orlando
G
,
Atala
A
,
Soker
S
.
Current achievements and future perspectives in whole-organ bioengineering
.
Stem Cell Res Ther
 
2015
;
6
:
107
118
.
29.
Uygun
BE
,
Soto-Gutierrez
A
,
Yagi
H
,
Izamis
M-L
,
Guzzardi
MA
,
Shulman
C
,
Milwid
J
,
Kobayashi
N
,
Tilles
A
,
Berthiaume
F
,
Hertl
M
,
Nahmias
Y
et al
Organ reengineering through development of a transplantable recellularized liver graft using decellularized liver matrix
.
Nat Med
 
2010
;
16
:
814
820
.
30.
Shupe
T
,
Williams
M
,
Brown
A
,
Willenberg
B
,
Petersen
BE
.
Method for the decellularization of intact rat liver
.
Organogenesis
 
2010
;
6
:
134
136
.
31.
Lehr
EJ
,
Rayat
GR
,
Chiu
B
,
Churchill
T
,
McGann
LE
,
Coe
JY
,
Ross
DB
.
Decellularization reduces immunogenicity of sheep pulmonary artery vascular patches
.
J Thorac Cardiov Sur
 
2011
;
141
:
1056
1062
.
32.
Halper
J
,
Kjaer
M
.
Basic components of connective tissues and extracellular matrix: elastin, fibrillin, fibulins, fibrinogen, fibronectin, laminin, tenascins and thrombospondins
. In:
Halper
J
(ed).
Progress in Heritable Soft Connective Tissue Diseases
 , vol. 802.
Dordrecht
:
Springer Netherlands
,
2014
:
31
47
.
33.
Hynes
RO
.
Extracellular matrix: not just pretty fibrils
.
Science
 
2009
;
326
:
1216
1219
.
34.
Watt
FM
,
Huck
WTS
.
Role of the extracellular matrix in regulating stem cell fate
.
Nat Rev Mol Cell Biol
 
2013
;
14
:
467
473
.
35.
Youngstrom
DW
,
Barrett
JG
,
Jose
RR
,
Kaplan
DL
.
Functional characterization of detergent-decellularized equine tendon extracellular matrix for tissue engineering applications
.
PLoS One
 
2013
;
8
:
e64151
.
36.
Wang
L
,
Johnson
JA
,
Zhang
Q
,
Beahm
EK
.
Combining decellularized human adipose tissue extracellular matrix and adipose-derived stem cells for adipose tissue engineering
.
Acta Biomater
 
2013
;
9
:
8921
8931
.
37.
Sánchez
PL
,
Fernández-Santos
ME
,
Costanza
S
,
Climent
AM
,
Moscoso
I
,
Gonzalez-Nicolas
MA
,
Sanz-Ruiz
R
,
Rodríguez
H
,
Kren
SM
,
Garrido
G
,
Escalante
JL
,
Bermejo
J
et al
Acellular human heart matrix: a critical step toward whole heart grafts
.
Biomaterials
 
2015
;
61
:
279
289
.
38.
Fatehullah
A
,
Tan
SH
,
Barker
N
.
Organoids as an in vitro model of human development and disease
.
Nat Cell Biol
 
2016
;
18
:
246
254
.

Author notes

Grant Support: This work was supported by SAF/2012-31017 (PI: CS), PROMETEOII/2013/018 (PI: CS), and by GRISOLIA/2015/002 (HC). PMB was supported by a H2020-MSCA-IF-2014 grant.
Both authors contributed equally.
Conference presentation: Presented in part at the Congrès annuel Saint-Luc UCL Bruxelles: L’endomètre dans tous ses états, 26–28 November 2015, Cliniques universitaires Saint-Luc, Brussels and as a poster at SRI’s 63rd Annual Scientific Meeting, 16–19 March 2016, Montreal, Canada.

Supplementary data