Isolation of multipotent progenitor cells from pleura and pericardium for tracheal tissue engineering purposes

Abstract Tissue engineering (TE) of long tracheal segments is conceptually appealing for patients with inoperable tracheal pathology. In tracheal TE, stem cells isolated from bone marrow or adipose tissue have been employed, but the ideal cell source has yet to be determined. When considering the origin of stem cells, cells isolated from a source embryonically related to the trachea may be more similar. In this study, we investigated the feasibility of isolating progenitor cells from pleura and pericard as an alternative cells source for tracheal tissue engineering. Porcine progenitor cells were isolated from pleura, pericard, trachea and adipose tissue and expanded in culture. Isolated cells were characterized by PCR, RNA sequencing, differentiation assays and cell survival assays and were compared to trachea and adipose‐derived progenitor cells. Progenitor‐like cells were successfully isolated and expanded from pericard and pleura as indicated by gene expression and functional analyses. Gene expression analysis and RNA sequencing showed a stem cell signature indicating multipotency, albeit that subtle differences between different cell sources were visible. Functional analysis revealed that these cells were able to differentiate towards chondrogenic, osteogenic and adipogenic lineages. Isolation of progenitor cells from pericard and pleura with stem cell features is feasible. Although functional differences with adipose‐derived stem cells were limited, based on their gene expression, pericard‐ and pleura‐derived stem cells may represent a superior autologous cell source for cell seeding in tracheal tissue engineering.


Abstract
Tissue engineering (TE) of long tracheal segments is conceptually appealing for patients with inoperable tracheal pathology. In tracheal TE, stem cells isolated from bone marrow or adipose tissue have been employed, but the ideal cell source has yet to be determined. When considering the origin of stem cells, cells isolated from a source embryonically related to the trachea may be more similar. In this study, we investigated the feasibility of isolating progenitor cells from pleura and pericard as an alternative cells source for tracheal tissue engineering. Porcine progenitor cells were isolated from pleura, pericard, trachea and adipose tissue and expanded in culture.
Isolated cells were characterized by PCR, RNA sequencing, differentiation assays and cell survival assays and were compared to trachea and adipose-derived progenitor cells. Progenitor-like cells were successfully isolated and expanded from pericard and pleura as indicated by gene expression and functional analyses. Gene expression analysis and RNA sequencing showed a stem cell signature indicating multipotency, albeit that subtle differences between different cell sources were visible. Functional analysis revealed that these cells were able to differentiate towards chondrogenic, osteogenic and adipogenic lineages. Isolation of progenitor cells from pericard and pleura with stem cell features is feasible. Although functional differences with adipose-derived stem cells were limited, based on their gene expression, pericard-and pleura-derived stem cells may represent a superior autologous cell source for cell seeding in tracheal tissue engineering.

K E Y W O R D S
airway reconstruction, stem cells, tissue engineering, tracheal replacement A recent development is the use of aortic allografts, but long-term follow-up and validation is needed to show their value in clinical practice and find out whether complications such as calcification and degeneration will occur. 8,9 Allogenic transplantation of tracheas is limited by the lack of donors, as well as the need for immunosuppressants, undesirable in oncological patients. 10,11 Although matrices of decellularized tracheas do support cell adherence and ingrowth, 12,13 these constructs fail long term due to inadequate mechanical support and insufficient revascularization. Moreover, immunological responses caused by incomplete removal of cell remnants may occur. 14,15 Attempts to create a long-term solution using tissue engineering (TE) techniques have not been successful so far.
Failure to successfully create a functional trachea may be due to both cell choice and matrix choice. For the development of a sustainable construct, the right cell source and cell type, proper differentiation stimuli and a suitable scaffold are essential. [16][17][18] Thus far, the ideal source for cartilage progenitor cells is still unknown.
In tissue engineering, the cellular origin subdivision is based on the embryonic layers: ectoderm, endoderm and mesoderm. 19 This subdivision was generated as it is pivotal to fundamentally understand tissue origin prior to attempts to imitate the natural process of tissue development. 20,21 Mesenchymal stem cells, especially derived from bone marrow and adipose tissue, are often used in tissue engineering. 22,23 However, though these cells share a mesodermal origin, they are not highly related to those of the trachea.
The respiratory epithelium arises from the endodermal part of the respiratory diverticulum, while tracheal cartilage and smooth muscle cells, essential for a sustainable construct, grow from the lateral mesodermal layer. Since pericardium and pleura also arise from this layer, progenitor cells from these tissues may provide a better source for tracheal tissue engineering.
In the present study, we investigated the presence and relative value for tracheal tissue engineering of progenitor cells isolated from different tissues that originate from mesoderm, like the native trachea. Progenitor cells from these tissues may be used as a new and possibly superior cell source in tracheal tissue engineering.

| Animal information and related guidelines
Tissues were harvested from six landrace pigs (±50kg) according to the institutional guidelines of Laboratory Animal Research. This study was approved by the Ethical Committee on Animal Research of the RadboudUMC, the Netherlands. Tissues were harvested from animals with planned termination for non-related studies.

| Isolation and culture of progenitor cells
Immediately after sacrifice, trachea, pericard, pleura and adipose tissue were harvested and placed in sterile phosphate buffered saline (PBS) with 1% penicillin/streptomycin (P/S). Tracheal explants were placed in culture flasks. Pericard and pleura were digested with collagenase type I for 30 min at 37°C, and adipose tissue was digested with collagenase type II for 60 min at 37°C. Suspensions were washed with PBS, and cells were collected by centrifugation. Isolation was performed using plastic adherence within the culture flasks. Cells were cultured in complete αMEM culture medium (Gibco). Cell culture medium was changed every 3 days. When cultures reached 80% confluence, cells were harvested by trypsin digestion and seeded 1:3. Cells in passage 3 were harvested for further analysis.

| RNA isolation and RT-qPCR
Isolation of total RNA and synthesis of cDNA was performed according to manufacturer's manual (Invitrogen). Gene expression was evaluated using SYBR Green qPCR analysis with the LightCycler LC480 (Roche). HPRT and GAPDH mRNA expression was used for normalization. Relative gene expression of several stem cell-related genes such as CD73, CD90, CD115, CD117 and SOX9 was calculated using the ΔΔCt-method and compared to the control group of isolated adipose-derived stem cells (ADSCs). 24 Primer sequences are listed in Table S1.
RNA used for RNA sequencing was purified using RNA Clean and Concentrator column (Zymo), clean-up, and purified RNA was run on a tricine/triethanolamine electrophoresis gel to assess the integrity. comparisons between all combinations. The differential gene list was filtered for a log2-fold change >1 and p-adjusted value <0.05, filtered for genes annotated in the Sus scrofa genome (a list of 520 significant differential genes) and visualized with pheatmap v1.0.12.
The R package clusterProfiler (v3.10.1) was used to perform Gene Ontology enrichment on the gene clusters.

| Differentiation assays
Pluripotency was investigated by RT-PCR and histological evaluation. Cells were chemically induced to differentiate and compared to a control group of cells cultured in standard αMEM.  Table S2.
Osteogenic induction was assessed using Alizarin-Red (Sigma-Aldrich) staining to objectify the presence of calcium deposits. Oil-Red-O (Sigma-Aldrich) staining was used to detect lipid deposits after adipogenic induction. All staining's were performed according to the manufacturer's manual.

| Scaffolds
Collagen matrices (type I collagen) were prepared from bovine achilleas tendon (Southern Lights Biomaterials) as previously described. 27,28 In brief, a 0,5% (w/v) collagen suspension was made by swelling and subsequently homogenization in 0.25 M acetic acid at 4°C. The suspension was deaerated by centrifugation (at

| Cell seeding and survival assays
Because cells should adhere and proliferate on a template, isolated progenitor cells were seeded on scaffolds at a density of approximately 1 × 10 6 cells/cm 2 for 24 h at 37°C in a polyHEMA (SantaCruz Biotechnologies) coated 6-well plate to prevent adherence of cells to the wells. Seeded scaffolds were cultured for 7 days. Samples were snap frozen in TissueTek (Firma), sectioned and stained with haematoxylin and eosin (H&E) according to manufacturer's manual. Cell viability was assed using the WST-1 assay (Sigma-Aldrich) according to manufacturer's manual. Absorbance was measured at OD 450 nm using a microplate reader (Wallac 1420 Victor).

| Statistical analysis
Statistical analysis was performed using GraphPad Prism software.
Expressed values are all shown as mean ± SD. Minimum requirement for individual experiments was N = 3. To compare different groups, analysis of variance ANOVA with Bonferroni post hoc test was used, with a significance threshold of p < 0.05.

| Characterization of isolated progenitor cells
Isolation of cells from porcine pericard, pleura, adipose tissue and trachea was achieved using established protocols. In culture, cells formed a heterogeneous monolayer of adherent, spindle-shaped cells ( Figure 1A). After expansion, cells maintained this morphology for over five passages.
RT-PCR analysis of the isolated cell populations for mesenchymal stem cell-related markers showed expression of CD73 and CD90 and to a lesser extend of CD105 and CD117 in all isolated cell populations. SOX9, associated with cartilage formation and tracheal patterning, was also expressed ( Figure 1B). Although differences in the relative expression of mesenchymal stem cell markers were observed, these were statistically insignificant.

| Gene expression
To test whether the expression profiles of pleura and pericard-

| Multilineage differentiation
To confirm the multipotency of the cells, they were exposed to different chemical stimuli to induce chondrogenic, osteogenic and

F I G U R E 2
Heat map of RNAsequencing analysis for differentially expressed genes. Total RNA was extracted from isolated progenitor populations of pericard, pleura, adipose and tracheal cells, sequenced and analysed as described previously. The differential gene list was filtered for a log2-fold change >1 and a p-adjusted <0.05. N = 3. Expression of lineage-specific molecules was examined by RT-PCR ( Figure 5). Pleura-and adipose-derived cells, subjected to F I G U R E 3 Gene ontology term analysis. Gene Ontology enrichment was performed on the gene clusters from isolated progenitor population of pericard, pleura, adipose and tracheal cells, sequenced and analysed as described previously. N = 3. Control groups did not show signs of differentiation. These findings indicated that isolated cell types contained subpopulations that have potential to differentiate in vitro towards mesenchymal phenotypes of bone, cartilage and/or adipose tissue, confirming stemcellness. 29

| Cell viability on scaffolds
An important variable in tissue engineering is the ability of cells to survive and proliferate on a scaffold. Cell survival was assessed at day 0 and day 7 after seeding using the WST-1 assay. In essence, cells were able to survive on the collagen scaffolds regardless of the cell type used. Cells grew into the scaffolds and proliferated. Slight differences in viability and proliferation were observed, but these were not significant. (Figure 6).

| DISCUSS ION
Creation of a sustainable tracheal substitute for tracheal repair after surgery remains a challenge. Most studies aimed at the development F I G U R E 5 Normalized mRNA expression of chondrogenic, osteogenic and adipogenic associated genes after induction. Comparison of transcript levels of related genes in porcine isolated cells before and after differentiation for 14 days presented here in fold change (logarithmic scale). Significant differences are indicated with *p < 0.05. **p < 0.005. ***p < 0.001. of tracheal substitutes, focus on new biomaterials with novel techniques. 30 Although it is obvious that the designed matrix is of utmost importance, the importance of cell type and cell source seems to be underestimated. Evidence supports the benefits of graft cell seeding, particularly in the case of mesenchymal stem cells, but it remains unclear which specific cell types provide the most optimal regeneration. 31 First, our data show that it is possible to isolate progenitor-like cells from pericard, pleura and trachea as judged by expressed surface markers, gene expression and the possibility to differentiate along multiple lines depending on the chemical stimuli. The rather high variation in the relative expression of mesenchymal stem cell markers is most likely due to donor variation. To the best of our knowledge, this is the first report showing that progenitor cells can be isolated from these tissues. Mesenchymal stem cells or progenitor cells have been isolated from many sources, for example kidney, liver, amniotic fluid, synovium and umbilical cord, confirming the idea that these cells reside in the connective tissue of most organs. 32 In general, these isolated cells were heterogeneous containing undifferentiated progenitors as well as lineage restricted precursors, and the potential to differentiate towards an osteogenic, adipogenic and chondrogenic lineage varied. 33 Furthermore, the stem cell-like character of the isolated cells is confirmed by their multilineage differentiation potential as confirmed by staining and marker expression. The morphological appearance of the isolated cells and glycosaminoglycan production, calcium deposition and lipid vacuoles accumulation was similar, regardless of the cell source. However, gene expression analysis after induction did reveal slight differences between the cells. Pericardderived cells failed to induce chondrocyte-related genes, and induction of osteoblast-related was limited. In contrast, pleura-derived cells showed an inverse pattern: chondrocyte-related genes were induced, and induction of osteoblast-related genes was limited.
Remarkably, all cells showed strong upregulation of adipose-related genes.
To further delineate differences between the isolated cells and their possible preference to differentiate towards a particular cell type, we performed RNA-sequencing and gene enrichment analy- Mesenchymal stem cells are a well-accepted choice for cell seeding in tissue-engineered organs due to their availability, capacity to expand in culture and multilineage differentiation capacity. 31,34,35 Both bone marrow-and adipose-derived stem cells are the most common used sources due to their easy accessibility, isolation potential and production of immunomodulatory factors. 18,22,23 Unfortunately, when used in tracheal TE issues with mechanical failure, stenosis of grafts and anastomosis seem recurrent. [36][37][38][39][40] Thus, even though multipotent cells have been isolated from many different sources and showed the capacity for multilineage differentiation, their therapeutic potential might be different. 33 In our case, differences between several progenitor cell sources mainly consisted of variation in gene signatures, whereas the functional analyses did not reveal outstanding differences.
We characterized new cell sources with a stem cell gene expres-

| S TATEMENT OF S I G NIFIC AN CE
The ideal cell source in tracheal tissue engineering (TE) has yet to be determined. An alternative, possibly superior autologous cell source for cell seeding purposes was found in pleura-and pericard-derived stem cells, based on their gene expression. These cells may be valuable for tracheal TE approaches as they may be more easily committed to differentiate into the cells of interest (chondrocytes), leading to better functional outcome of engineered constructs.

ACK N OWLED G EM ENTS
None.

CO N FLI C T S O F I NTE R E S T
The authors indicate no potential conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.