Ex-vivo recellularisation and stem cell differentiation of a decellularised rat dental pulp matrix

Implementing the principles of tissue engineering within the clinical management of non-vital immature permanent teeth is of clinical interest. However, the ideal scaffold remains elusive. The aim of this work was to assess the feasibility of decellularising rat dental pulp tissue and evaluate the ability of such scaffold to support stem cell repopulation. Rat dental pulps were retrieved and divided into control and decellularised groups. The decellularisation protocol incorporated a low detergent concentration and hypotonic buffers. After decellularisation, the scaffolds were characterised histologically, immunohistochemistry and the residual DNA content quantified. Surface topography was also viewed under scanning electron microscopy. Biocompatibility was evaluated using cytotoxicity assays utilising L-929 cell line. Decellularised scaffolds were recellularised with human dental pulp stem cells up to 14 days in vitro. Cellular viability was assessed using LIVE/DEAD stain kit and the recellularised scaffolds were further assessed histologically and immunolabelled using makers for odontoblastic differentiation, cytoskeleton components and growth factors. Analysis of the decellularised scaffolds revealed an acellular matrix with histological preservation of structural components. Decellularised scaffolds were biocompatible and able to support stem cell survival following recellularisation. Immunolabelling of the recellularised scaffolds demonstrated positive cellular expression against the tested markers in culture. This study has demonstrated the feasibility of developing a biocompatible decellularised dental pulp scaffold, which is able to support dental pulp stem cell repopulation. Clinically, decellularised pulp tissue could possibly be a suitable scaffold for use within regenerative (reparative) endodontic techniques.

a preserved three-dimensional histological architecture and natural structural components 13 . The ECM is a vital and dynamic structure composed of tissue specific structural and functional proteins. This complex integration between cells and the surrounding microenvironment (ECM proteins) regulates cellular activity and may provide a suitable environment for tissue regeneration 14 .
Biological scaffolds derived from tissue decellularisation have been assessed for dental applications including decellularisation of bovine pulp tissue 15 , porcine tooth buds 16,17 , swine pulp tissue 18,19 , human root slices 20 , and human pulp tissue 6 . While successful biocompatible decellularisation have been shown in human pulp tissues 6 , further research is required to show the decellularised scaffolds' ability to support stem cell recellularisation and differentiation into odontoblast-like cells.
Decellularised rat tissues, as experimental models, have been reported in the literature 21,22 , and can be regarded as an alternative research tissue source to human tissues which are becoming increasingly difficult to obtain. Of interest, histological characterisation of rat dental pulp tissues has shown similar structural composition to human pulp tissues 23,24 . Therefore the aim of the present work was to illustrate the ability to decellularise rat pulp tissues using low concentration detergents and evaluate the ability of such decellularised scaffold to support stem cell growth and differentiation.

Materials and methods
Ethical consideration. All experimental protocols and study design was approved by the Research Ethics Committee, University of Leeds, United Kingdom (approval no. 101013/MME/113). The use of human and animal tissue in this study followed relevant guidelines and regulations as set by University of Leeds Research Ethics Committee.
Tissue procurement. Rat jaws were only collected for this study. Wistar male rats (28 ± 2 days old) were humane killed in-advance in the Central Biomedical Services (University of Leeds, Leeds, United Kingdom) as per the institutional animal use and care committee at the University of Leeds and according to guidelines laid down by the Animals (Scientific Procedures) Act 1986. Dental pulp tissues were aseptically retrieved from the rat mandibular incisors and stored in 0.1 M phosphate-buffered saline (PBS, 17-516F, Lonza, Slough, United Kingdom; pH 7.4) at − 80 °C for up to 4 weeks. All reagents and solutions were purchased from Sigma-Aldrich (Poole, United Kingdom) and cell culture plastics were obtained from Corning, Inc. (Amsterdam, The Netherlands) unless stated otherwise.
Contact cytotoxicity assay. Contact cytotoxicity assay was performed in a similar manner to previous studies 13,25 . Samples of decellularised dental pulps (n = 4 per group) were attached to the centre of 6-well tissue culture plates using collagen type I (A10483-01, Gibco, Thermo Fisher Scientific). Cyanoacrylate glue (Z105902-1EA) or collagen type I in triplicate wells served as positive and negative controls, respectively. L-929 cells were seeded into each well at a density that would allow confluency after 48 h in cell standard culture medium. Plates were incubated at 37 °C in 5% (v/v) CO 2 in air for 48 h. Following incubation, the culture medium was carefully aspirated before the wells were washed with PBS (pH 7.4) containing calcium and magnesium and fixed with 10% (v/v) neutral buffered formalin for 10 min. The wells were then stained with Giemsa solution (Merck R66 formulation, 352603R, VWR International, Ltd. Lutterworth, United Kingdom) for 5 min. The wells were rinsed with distilled water until clear and air dried. Changes in cell morphology and confluency were analysed using an inverted microscope under bright field illumination (Cell^B software) with all images captured digitally.
Extract cytotoxicity assay. Segments of decellularised dental pulps (n = 8) were finely minced and left to incubate in DMEM (1 mL/mg) for 72 h at 37 °C with agitation. Following incubation, the mixture was centrifuged at 500 g for 15 min and the supernatant was collected. L-929 cells were seeded onto 96-well plates (Optiplate white plates; PerkinElmer Ltd. Beaconsfield, United Kingdom) at a density that would allow confluency after 24 h in standard culture medium. Plates were incubated at 37 °C in 5% (v/v) CO 2 in air. Following incubation, the culture medium was removed and replenished with 100 µL of fresh culture medium. For the study wells, 100 µL of decellularised extracts was added and further incubated for 24 h. 40% (v/v) dimethyl sulfoxide (DMSO, D8418) or DMEM in triplicate wells served as positive and negative controls, respectively. The Histological and immunohistochemistry assessment of human DPSCs cultured on decellularised pulp tissue. Following recellularisation, the ability of the decellularised scaffolds to support stem cell attachment, growth and differentiation were analysed using H&E and immunohistochemical staining methods (n = 4, each time point). Standard histology and immunohistochemical methods were performed as previously described above. Primary antibodies against odontoblastic markers included mouse monoclonal dentine matrix protein-1 (DMP-1, 73633, 1:100, Santa Cruz Biotechnology, Inc., Heidelberg, Germany), dentine sialophospoprotein ( Statistical analysis. Quantitative data was analysed using either an independent Student's t-tests or oneway analysis of variance (ANOVA) depending on the number of groups. All experiments were performed in triplicates. The statistical analysis was performed using Prism 6 (GraphPad Software, California, USA), p-values < 0.05 were considered significant.

Results
Decellularisation assessment. Decellularisation efficiency of rat dental pulp tissues was performed using both qualitative and quantitative assessment methods. Histology analysis using H&E and DAPI staining revealed a high density of cells within control (native) dental pulps and absence of cells in decellularised samples (Fig. 1a). Immunolabelling with major histocompatibility complex class II antibodies also revealed a positive immunoreactivity in the control pulp tissues, this positive staining was associated with specific cells scattered throughout the ECM (Fig. 1a). In contrast, a negative immunoreactivity was detected in the decellularised samples (Fig. 1a).
No brown staining was seen in isotype stained sections (images not provided). Quantitative analysis using total DNA from both control and decellularised pulp tissues was extracted and quantified. The mean DNA content in the decellularised tissues was 23.00 ± 1.430 ng DNA per mg of dry tissue, in comparison to the 1765 ± 20.66 ng DNA per mg of dry control tissues. This reduction of over 98% was found to be significantly different when compared to the control tissue (Fig. 1b).
Collagen structure assessment. Assessment of the collagen structure of rat dental pulp tissues was performed using both histology and immunohistochemistry analysis. Histological characterisation of picrosirius red stained control tissues viewed under both bright field and polarised light microscopy revealed a collagen matrix mainly composed of thin filaments and fibrils that were weakly birefringent to light (Fig. 2). Following decellularisation, the collagen fibre network was overall well preserved throughout the entire matrix (Fig. 2). Antibody labelling against collagen type I and type III was additionally performed (Fig. 2). In control tissues, a weak positive immunoreactivity to collagen type I and III antigens was observed. These thin fibres and fibrils were distributed throughout the matrix, mainly concentrated around and along the vasculature. Following decellularisation, a positive but less intense staining pattern was seen (Fig. 2). No brown staining was seen in isotype stained sections (images not provided).

Non-collagenous structures assessment.
Assessment of the non-collagenous ECM components of rat dental pulp tissues was performed using histology and immunohistochemistry analysis. Histological characterisation of alcian blue stained control tissues revealed a glycosaminoglycan rich tissue (Fig. 3). Following decellularisation, preservation of glycosaminoglycan molecules was evident with no visible differences between the control and the decellularised tissues (Fig. 3). Antibody labelling against structural proteins and α-gal was performed (Fig. 3). In control tissues, fibronectin and laminin appeared as a fibrous plexus mainly concentrated Scientific Reports | (2020) 10:21553 | https://doi.org/10.1038/s41598-020-78477-x www.nature.com/scientificreports/ around the vasculature. Following decellularisation, a positive but noticeably weaker staining pattern against fibronectin was observed. Although laminin staining was less intense in comparison to the control tissues, the distribution was comparable to control tissues, it was found to be present throughout the entire ECM mainly concentrated around the basement membrane of vasculature structures. No evidence of α-gal was observed within the decellularised ECM (Fig. 3). No brown staining was seen in isotype stained sections (images not provided).
Scanning electron microscopy. Structural assessment of rat dental pulp analysed under SEM revealed a complex irregular fibre mesh structure related to the control tissues (Fig. 4a). Following decellularisation, preservation of complex fibre mesh denuded of some components, mainly cell and cell fragments was evident (Fig. 4b).
Contact cytotoxicity assay. Microscopic analysis of the cell culture plates showed L-929 cell growth up to and in contact with the samples of decellularised dental pulp (Fig. 5a). No obvious morphological change nor Extract cytotoxicity assay. There was no significant reduction in the cellular ATP levels of the L-929 cells following incubation with soluble extracts of decellularised dental pulps compared with that of the DMEM (negative control) (ANOVA, p > 0.05; Fig. 5b). In contrast, DMSO (positive control) caused an almost total loss of ATP and were significantly lower than any other sample group (ANOVA, p < 0.0001; Fig. 5b).

Determination of viability of human DPSCs cultured on decellularised pulp tissue. The viabil-
ity of DPSCs following recellularisation on the decellularised scaffolds was assessed using standard LIVE/DEAD staining kit. Following 7 and 14 days culture, Z-stack constructed confocal laser images clearly demonstrated   (Fig. 6).

Histological assessment of human DPSCs cultured on decellularised pulp tissue. Serial sec-
tions of the recellularised scaffolds were assessed with H&E staining to determine the ability of DPSCs to attach, migrate and differentiate within the decellularised scaffolds. Following 7 and 14 days culture, the DPSCs appeared elongated with a spindle-shaped morphology attached to the outer scaffold surface and between the collagen fibers. As time elapsed, there was an apparent change in the scaffold shape from a longitudinal (day 7) to a contracted circular shape (day 14; Fig. 6).
Immunohistochemistry assessment of human DPSCs cultured on decellularised pulp tissue. Antibody labelling was performed to assess the ability of the seeded DPSCs to express cytoskeleton  www.nature.com/scientificreports/ components, growth factors and odontoblastic markers. Odontoblastic markers, DMP-1 and ALP, were weakly stained at day 7 with an increase in staining intensity following longer culture duration (day 14; Fig. 7). In contrast, DSPP was strongly positive following 7 and 14 days culture (Fig. 7). The seeded DPSCs were stained dark brown against vimentin, nestin, VEGF-A, and VEGFR-2 antigens indicating strong positive cellular immunoreactivity. These antigen positive cells were scattered throughout the  www.nature.com/scientificreports/ scaffold at day 7 and 14 culture (Fig. 8). However, α-SMA was weakly stained at day 7 with an increase in staining intensity following 14 days culture (Fig. 8). No brown staining was seen in isotype stained sections (images not provided).

Discussion
Scaffold development for use in regenerative (reparative) endodontics is currently the subject of intense research in an attempt to improve the clinical outcomes and prognosis of these vulnerable teeth. The ECM is regarded as a vital component of all tissues and, possibly, the ideal scaffold for tissue engineering 14 . Therefore, developing a biocompatible decellularised ECM scaffold is deemed likely to allow the ideal environment for stem cell repopulation and differentiation. The decellularisation protocol performed in this work, involved cell membrane lysis in a hypotonic Tris buffer, solubilisation of cytoplasmic and nuclear components using a single detergent cycle of 0.03% SDS, followed by removal of cellular material with a nuclease treatment.
Rat dental pulp tissues were used in this work due to histological structural similarities to human pulp tissue 23,24 . The use of laboratory rats is also well known as inevitable part of translational biomedical research 26 , and considered an alternative tissue source due to the limited availability of human pulp tissues for experimental studies.
Collagen fibres are known as the most abundant fibrous component of the dental pulp 27 , and largely contribute towards the development and maintenance of the pulp matrix 28 . Analysis of the dried pulp tissue obtained from rat incisors revealed a low collagen content of around 3.5% 23 , with collagen type I and III representing the main bulk of collagen 23,29,30 . This low collagen content was in line in with our results showing immature collagen fibres stained green as viewed under the polarised light and a weak positive immunoreactivity to collagen type I and III antigens within the control tissues. Surface topography assessment of control rat dental pulp viewed under SEM revealed a complex irregular fibre mesh structure. Following decellularisation, preservation of structural collagen was evident with various qualitative methods utilized.
Therefore, in this work decellularised rat dental pulp tissues resulted in a biocompatible acellular scaffold with preservation of the connective tissue architecture, retained vascular channels and ECM proteins similar to those obtained following decellularisation of the human dental pulp tissue 6 .
Xenogenic hyperacute rejection is reported to be mediated by complement activation of antibodies against α-gal epitope 31,32 . Therefore, complete removal of α-gal epitope from the xenogenic tissue is essential. The results of this work showed no evidence of α-gal expression following immunohistochemical assessment of decellularised rat pulp tissues. These results are in line with those reported following decellularisation of porcine medial meniscus 33 , porcine bone cartilage 34 , and porcine pulmonary valve 25 , using comparable decellularisation protocols. A negative immune-reactivity against major histocompatibility complex class II molecules was also evident following decellularisation. These specific class of molecules are expressed by immune cells such as monocytes, macrophages and dendritic cells and serve as initial defence cells 35,36 .
Following recellularisation with DPSCs, the produced ECM scaffolds were able to support stem cell viability, attachment and migration in vitro. As time elapsed, a decrease in the overall size of the seeded scaffold was apparent, changing from original tubular shape observed at day 7 to a circular shape at day 14. Contraction of cell-seeded scaffolds has been reported in the literature following recellularisation of polyglycolic acid matrices with human fibroblastic cells 37 , and collagen-glycosaminoglycan matrices seeded with DPSCs 38 . This matrix contraction was thought to be linked to the expression of α-SMA positive cells 38 . The expression of α-SMA by DPSCs cultured on rat decellularised scaffolds, demonstrated in this study, is likely to have contributed to the contraction of the scaffolds. Further work is needed to assess this hypothesis in addition to the effects of such contraction on cell penetration, survival and differentiation. www.nature.com/scientificreports/ In the present study, immunohistochemical labelling of recellularised rat scaffolds showed a positive expression against several genes and odontoblast markers. DPSCs seeded within the decellularised scaffolds expressed putative markers of odontoblastic differentiation including DSPP, DMP-1 and ALP following 14 days culture in standard medium conditions. These markers, although non-specific, have been recommended in identifying cell differentiation towards odontoblastic-like cell lineage 39,40 . Previous in vivo studies assessing the odontoblastic differentiation of stem cells reported a positive expression to DSPP and DMP-1 following 21 days culture when Puramatrix was seeded with either DPSCs in tooth slice model 41 , or stem cells from human exfoliated deciduous teeth in full length root canals 42 . Furthermore, human recombinant collagen scaffolds seeded with stem cells from human exfoliated deciduous resulted in a positive expression for DSPP and DMP-1 following 14 and 28 days culture, respectively 42 . Interestingly, odontoblastic differentiation of stem cells was not possible in both seeded scaffolds without the surrounding dentine structure 41,42 . Song et al. (2017) also reported positive expression of odontoblastic markers following 14 days cell culture utilizing a decellularised pulp tooth slice model 20 . In contrast, the present study revealed stem cell differentiation solely within the decellularised scaffold. Indeed, tooth specific environmental cues have a significant impact on the differentiation ability of stem cell into odontoblastic lineages [43][44][45] . Hence, it could be hypothesised, that placement of the decellularised scaffold in direct contact with the surrounding tooth structure enhances the tissue engineering process, with new tissue formation in close resemblance to the control tissue.
Successful engineering of a pulp-dentine complex also depends on angiogenesis and neurogenesis. Nestin, an intermediate filament protein, is classified as a neural progenitor marker and has been found to be expressed in newly differentiated and functioning odontoblast cells 46,47 . Vimentin, a major intermediate filament protein, has also been linked to neurogenesis and expressed by neuronal precursor cells 48 . Although vimentin is not considered as a pulp tissue specific protein, it has been suggested as a dental pulp quality standard protein for characterisation of pulp regeneration 49 . Indeed exposure of DPSCs to the correct environmental cues is crucial for their ultimate differentiation into active neurons 50 . Angiogenesis, on the other hand, aids in the transport of oxygen, nutrients and in attracting prevascular stem cells to the regenerated site, thereby preventing tissue necrosis 51,52 . Notably, several studies highlighted the ability of DPSCs in inducing angiogenesis and the expression of proangiogenic factors including VEGF-A and VEGFR-2 53,54 . The immunolabelling results of this study showed positive cellular expression against nestin, vimentin, VEGF-A, and VEGFR-2 antigens following recellularisation indicating, a possibility, of active neurogenesis and angiogenesis.
In conclusion, results of this work indicate successful decellularisation of rat dental pulp tissues utilising a single low concentration detergent based decellularisation protocol. The resulting biocompatible scaffold was able to support dental pulp stem cell repopulation over a 14 day period. The recellularised scaffold also showed a positive cellular expression against odontoblastic markers, cytoskeleton proteins, and growth factors.
Additional scaffold testing utilizing ex vivo culture models; tooth slice 55 and ectopic root transplantation models 56,57 in order to assess scaffolds ability to regenerate within the surrounding environment is of importance for future clinical translation 58 . Following animal testing, all work will be duplicated on human tissue for future clinical trials.