Response of human oral mucosal epithelial cells to different storage temperatures: A structural and transcriptional study

Purpose Seeking to improve the access to regenerative medicine, this study investigated the structural and transcriptional effects of storage temperature on human oral mucosal epithelial cells (OMECs). Methods Cells were stored at four different temperatures (4°C, 12°C, 24°C and 37°C) for two weeks. Then, the morphology, cell viability and differential gene expression were examined using light and scanning electron microscopy, trypan blue exclusion test and TaqMan gene expression array cards, respectively. Results Cells stored at 4°C had the most similar morphology to non-stored controls with the highest viability rate (58%), whereas the 37°C group was most dissimilar with no living cells. The genes involved in stress-induced growth arrest (GADD45B) and cell proliferation inhibition (TGFB2) were upregulated at 12°C and 24°C. Upregulation was also observed in multifunctional genes responsible for morphology, growth, adhesion and motility such as EFEMP1 (12°C) and EPHA4 (4°C–24°C). Among genes used as differentiation markers, PPARA and TP53 (along with its associated gene CDKN1A) were downregulated in all temperature conditions, whereas KRT1 and KRT10 were either unchanged (4°C) or downregulated (24°C and 12°C; and 24°C, respectively), except for upregulation at 12°C for KRT1. Conclusions Cells stored at 12°C and 24°C were stressed, although the expression levels of some adhesion-, growth- and apoptosis-related genes were favourable. Collectively, this study suggests that 4°C is the optimal storage temperature for maintenance of structure, viability and function of OMECs after two weeks.


Introduction
Limbal stem cell deficiency (LSCD) is a potentially painful and blinding condition caused by damage or loss of the limbal stem cells [1,2]. Cell therapy has gained increasing attention over the last few years in the treatment of LSCD after the first successful transplantation of autologous limbal stem cells in 1997 by Italian scientists for patients with unilateral disease [3]. This therapy is believed to replace or restore remaining functioning limbal stem cells to promote regeneration of the corneal epithelium [4].
For treatment of unilateral LSCD, the use of cultured autologous limbal stem cells from the healthy cornea is often the method of choice. In the event of bilateral LSCD, which is more prevalent, the application of allogeneic limbal stem cells remains an option, but is considered less favourable because it requires systemic immunosuppressive drugs with potential adverse effects and the long-term outcome is not always satisfactory. Therefore, the quest for alternative cell types was initiated. Oral mucosal epithelial cells (OMECs) were the first alternative autologous source to be studied in both rabbits [5] and humans [6]. Later examples of cell sources include epidermal, embryonic, conjunctival epithelial, umbilical cord, hair follicle bulge, immature dental pulp, orbital fat-derived and bone marrow-derived mesenchymal stem cells [7,8]. Of those, OMECs and conjunctival cells are hitherto the only clinically used nonlimbal cell type for the treatment of LSCD [9]. Seeking to improve transplantation success and expand access to regenerative medicine, researchers have also investigated different storage conditions [10][11][12][13][14][15][16] and transportation techniques [17][18][19].
The prepared epithelial stem cells in vitro may not be used immediately for transplantation due to patient readiness and the time required for transport from laboratory units to clinics. Such a gap gives the opportunity for phenotypic investigations [20], sterility control [21] and flexible scheduling of surgery [22]. To establish optimal conditions, several studies have investigated various aspects of storage [10- 16,21,23,24]. These attempts are in line with the European Medicines Agency's approval of stem cell therapy in Europe [25]. In addition, advancements in storage technology has provided the necessary practical methods for worldwide distribution of cultured cells from centralized laboratories to clinics [26] following rather strict regulations for cell therapy [27] and taking into consideration an increasing demand over the last few years [28].
Cryopreservation and refrigeration (4˚C) are two common storage methods, especially for cultured epidermal cells. However, both suffer from the drawback of low cell viability [29][30][31]. Studies on cryopreservation have also shown fragmented cultured cell sheets with rather unsuitable morphology [23], along with complicated procedures requiring specific and costly devices. In search for an alternative and better preservation methods, our group has previously investigated the effects of different temperatures (between 4˚C-37˚C) on the structure and function of cultured human oral keratinocytes [16] and epidermal cells [15] for one and two week periods, respectively.
The present study aimed to examine ten significantly differentially expressed genes among the studied groups by Utheim et al. [32], and 20 genes from various relevant pathways using TaqMan gene expression array cards in cultured OMECs stored at four different temperatures (4˚C, 12˚C, 24˚C and 37˚C) for two weeks. The morphology (light and scanning electron microscopes) and cell viability (trypan blue exclusion test) of cultured epithelial cells were also investigated.

Ethical considerations
Local ethical approval and verbal informed consent were obtained before OMEC harvesting. According to guidelines, the relevant form was completed and signed by the specialist during interview with the next of kin before transferring cadaver to the Department of Pathology, Oslo University Hospital. The research was conducted in accordance with the Declaration of Helsinki. The experimental protocols were approved by The Regional Committee for Medical and Health Research Ethics, Section C, South East Norway (reference: 2017/418).

Cell isolation
Donated lower lip biopsy was harvested from cadaver's oral cavity within 24 h post mortem in a non-refrigerated setting. The cadaver was a 56-year-old male with heart disease and good oral hygiene. The biopsy was transported to tissue-culture laboratory within 1 h in Dulbecco's Modified Eagle Medium/Nutrient F-12 Ham + GlutaMAX™-I (DMEM/F12) supplemented with 100 U/mL penicillin-streptomycin (P/S) at room temperature and then rinsed off using the same medium. It was cut into 1 cm × 0.5 cm pieces and then incubated in Mg 2+ and Ca 2 + -free Hanks' balanced salt solution containing 1.2 U/mL dispase II at 37˚C overnight. The separation of epithelial cell layer from the lamina propria layer was performed under a dissecting microscope by sterile forceps and scalpel. Afterwards, the tissue was rinsed with DMEM/ F12, cut into explants (1-3 mm 2 ), put on plastic inserts and allowed to attach in DMEM/F12 supplemented with 10% fetal bovine serum and P/S. Subsequently, the cultured cells were incubated at 37˚C (5% humidified CO 2 ) for two weeks. Then, the expanded cells were harvested (0.25% trypsin-EDTA solution) as passage 1 (P1) cells.
Trypan blue exclusion test was used for P1 cell counting and viability assessment before being frozen in 5% DMSO at -80˚C overnight followed by liquid nitrogen storage. The OMECs were seeded on culture vessels in serum-free CNT-Prime medium supplemented with P/S on a 48-well-plate. All the cultures were incubated at 37˚C and 5% CO 2 . The culture medium was changed every 2-3 days.

Cell storage
The frozen OMECs were thawed, seeded and expanded on culture flasks in serum-free CNT-Prime medium supplemented with P/S at 37˚C with 5% CO 2 for 5-7 days. The culture medium was changed every 2-3 days. Then, cells were trypsinized and cultured on 24-well plates with (for scanning electron microscopy) or without (for other tests) 8 mm round glass coverslips to obtain a confluent monolayer under the same condition. Before random selection for storage in cabinets at four different temperatures (4˚C, 12˚C, 24˚C and 37˚C with SD 0.4˚C), the storage medium containing MEM with 12.5 mM HEPES, 3.57 mM sodium bicarbonate and 50 mg/ml gentamycin was replaced and the 24-well plates were properly sealed.
Oral mucosa epithelial cells were isolated from a single donor and cultured on uncoated plastic plates to minimize confounding variation, focus on temperature-associated differences and limit variation of scaffolds in the form of biological substrates such as amniotic membrane and fibrin [15]. Three technical replicates were used for each experiment.

Light microscopy
The morphology of individual cells and the integrity of the complete cell layer in all four different temperature conditions were evaluated by a Leica DMIL inverted phase contrast microscope (Leica Microsystems, Wetzlar, Germany) coupled with a Canon EOS 5D mark II camera (Canon, Oslo, Norway). The images were captured at random positions within each well at 4×, 10×, 20× and 40× magnification before (baseline) and after two weeks incubation time (4˚C, 12˚C, 24˚C and 37˚C).

Scanning electron microscopy
Immediately after the incubation period, fixative (PHEM 0.2M, Glutaraldehyde 2% and Paraformaldehyde 8%) pre-adjusted to desired temperature was added in a 1:1 ratio to each well. Afterwards, the plates were returned to their respective storage cabinets for 15 min. Then, fixed cells on each coverslip were dehydrated using a graded ethanol series consisting of a 10-min treatment with 70, 80, 90, and 96% ethanol, followed by 4 x 15-min treatments with 100% ethanol. Samples were then critical point dried (BAL-TEC CPD 030 Critical Point Dryer) and coverslips were mounted on SEM stubs using carbon tape. Finally, samples were sputter coated with 6 nm platinum using a Cressington 308R coating system, then all samples were examined and digital images recorded using a Hitachi S-4800 Field Emission Scanning Electron Microscope operated at 5.0 kV.

Cell viability test
The viability of cultured cells was evaluated by the trypan blue exclusion test [33].

TaqMan gene expression array cards
The cells were harvested directly in RLT buffer (Qiagen, Hilden, Germany) and mixed by pipetting up and down. The lysate was then passed through a needle (0.9-mm diameter) attached to a sterile plastic syringe 10 times for achieving a homogenous lysate. RNeasy micro kit (Qiagen, Hilden, Germany) was used for extraction and purification of total RNA. Concentration and purity of isolated RNA were assessed by spectrophotometry (Nanodrop, Wilmington, Germany). Using the High Capacity cDNA RT Kit (Applied Biosystems, Abingdon, UK), reverse transcription (RT) was performed with 200 ng of total RNA per 20 μL RT reaction based on the manufacturer's instruction.
Comparative relative quantification was performed using the TaqMan Gene Expression Array Card for 30 genes ( Table 1). The choice of genes was partly based on our previous study Utheim et al [32], in which cultured oral keratinocytes were examined for gene expression after one week storage at 4˚C, 12˚C and 37˚C. Ten significantly differentially expressed genes (CCNH, CDKN2B, GADD45B, GAS5, HDAC6, HIST2H4B, HUS1, RPSA, TBRG4 and TGFB2) among the studied groups were selected, accordingly. In addition, 19 genes from various relevant pathways were included. GAPDH was also included to evaluate whether energy consumption originated from oxidative phosphorylation or glycolysis.
The samples were prepared for PCR by mixing 55 μL of cDNA with 55 μL of TaqMan™ Universal PCR Master Mix, no AmpErase™ UNG (Applied Biosystems, Germany). Then, each fill reservoir of a TaqMan Array Card (Applied Biosystems, Germany) were loaded with 100 μL of sample-specific PCR reaction mix. Three wells were allocated per gene in this experiment. The array cards were run in the QuantStudioTM 12K Flex Real-time PCR system (Applied Biosystems, Germany) under the following conditions: 50˚C for 2 min, 95˚C for 10 min, then 40 cycles of 95˚C for 15s and 60˚C for 60s. All samples were run in triplicates.

Real-time quantitative polymerase chain reaction (RT-qPCR)
Comparative relative quantification was performed on prepared cDNA using the StepOne-PlusTM Real-Time polymerase chain reaction (PCR) system (Applied Biosystems) and Taqman Gene Expression assays following protocols from the manufacturer (Applied Biosystems) for 4 genes (ΔNp63α, TJP1, CLND1 and OCLN) ( Table 2). All the samples were run in triplicates (each reaction: 2.0 μL cDNA, total volume 20 μL). The thermo cycling parameters were 95˚C for 10 min followed by 40 cycles of 95˚C for 15 s and 60˚C for 1 min.

Statistical analysis
The analysis of gene expression data was performed using the Relative Expression Software Tool (REST © , Relative Expression Software Tool, Weihenstephan, Germany

Examination of cell morphology
The confluent monolayer culture of OMECs presented typical cobblestone appearance under light microscope before storage (Fig 1). A two-week cultured OMECs at 4˚C resulted in morphological features resembling non-stored baseline control cells. However, uneven cell-cell contact in some spots led to cell separation and deviation from uniform polygonal morphology. Two weeks after the established incubation period, phase contrast micrographs of cells stored at 12˚C has identified mainly cells with rounded morphology. At 24˚C, the mosaic-like growth pattern typical of epithelial cells at confluence had completely disappeared. Most of the cells were detached at 37˚C and formed cell clusters. The detached and/or loosened cells at 4˚C, 12˚C, 24˚C and especially 37˚C were easily washed away during medium change.
The evaluation of cell-cell contacts and cell adhesion of non-stored baseline OMECs using SEM demonstrated a uniform, closely attached and solid monolayer of cells with distinct cell borders (Fig 2). At 4˚C, epithelial cells retained extended filopodia, enlarged flattened appearance and microvilli on the apical surface similar to non-stored baseline control cells, but there was evidence of cell separation between some adjacent cells. The clustered cells, forming uneven islands, at 12˚C and 24˚C resulted in apparent changes in cultured epithelial cells morphology. Blebbing was also seen on some of the cells. At 37˚C, the remaining irregular, shrunken and disintegrated cells were surrounded by debris.

Assessment of cell viability
The viability of OMECs incubated for two weeks was decreased with increasing storage temperature compared to non-stored baseline control cells (Fig 3). The 4˚C experimental group with 58% viability had the highest number of living cells, whereas the 37˚C group recorded no living cells. The 12˚C and 24˚C study groups, presented 14% and 2% viability, respectively. The cell viability in all groups was significantly lower than baseline control (p � 0.05).

Evaluation of gene expression
This study considered a log2-fold change significant at thresholds above 1 and below -1 (i.e., > two-fold change). Moreover, this approach was only applied to 4˚C, 12˚C and 24˚C groups as all cells died upon storage at 37˚C (Supplementary). Thus, altered expression at this temperature was deemed irrelevant. We identified 18 differentially expressed genes upon storage using the TaqMan Gene Expression Array (Fig 4). Four of them (CDKN1A, TBRG4, PPARA and TP53) were downregulated in all four different temperature conditions when compared to non-stored baseline control (p � 0.05). Four other genes were also downregulated at 4˚C and 37˚C, with upregulation at 12˚C (EFEMP1), 24˚C (CCNH and CASP4), or both 12˚C and 24˚C (GADD45B). Similarly, significant downregulation for all 4 temperature conditions, excluding insignificant changes at 4˚C (HIST2H4A; HIST2HB) and 12˚C (RPSA and IKBKE), were also observed. In contrast, EPHA4 showed significant upregulation at 4˚C, 12˚C and 24˚C. Six genes (HIST1H4B, TAF1D, TGFB2, KRT1, KRT10 and GAPDH) presented a combination of significant (both downregulation and upregulation) as well as insignificant changes in expression in studied groups of each different temperature conditions. Results from RT-qPCR (Fig 5) indicated that ΔNp63α, TJP1 and CLND1 were significantly upregulated at 4˚C but OCLN remained unchanged. At 12˚C, three of genes (ΔNp63α, CLND1   and OCLN) were significantly downregulated, except unchanged TJP1. At 24˚C, significant downregulation was observed for ΔNp63α and TJP1, whereas upregulation for CLND1 and OCLN.

Discussion
To establish optimal storage conditions for OMECs to maximize transplantation success and extend access to regenerative medicine, this study aimed to assess the morphology, cell viability and expression of selected genes of cultured OMECs stored at four different temperatures (4˚C, 12˚C, 24˚C and 37˚C) for two weeks. The overall morphology and ultrastructure of cells stored at 4˚C was most similar to non-stored baseline controls, whereas the 37˚C group was the most dissimilar ones with only dead cells. The presence of cell separation between some adjacent cells at 4˚C group and prominent morphological deformation, shrinkage, and membrane blebbing at higher temperatures in our study were also observed by others, Islam et al [16] and Jackson et al [15] for cultured oral keratinocytes and epidermal cells after one and two weeks of storage, respectively. However, the authors found that mid range temperature groups of 12˚C, 16˚C and 20˚C resulted in the best preserved morphology.
Cells stored at 4˚C had the highest viability rate (58%) compared to other three experimental groups with higher temperatures, whereas the 37˚C group presented no living cells. The former is in the viability range of suspended human oral keratinocyte (~63%) in DMEM, assessed using trypan blue exclusion test, after three days storage at 4˚C [37]. Comparably, Islam et al [16] showed that adherent human oral keratinocytes cultured in HEPES-MEM had~45% viability, as measured by calcein-acetoxymethyl ester fluorescence, following one week storage at 4˚C. The highest viability of the latter study, however, belonged to 12˚C group (~85%).
Other cell types have also been studied for the effect of storage temperature. For example, human retinal pigment epithelium [38], epidermal keratinocytes [15] and conjunctival epithelium [39] recorded~4% (one week),~18% (two weeks) and~72% (one week) viable cells at 4˚C, respectively. For those cell types, the viability was greatest at 16˚C (~48%,), 24˚C (~97%,) and 12˚C (~95%), respectively. This variation may be due to the cell types, the culture medium, storage period or the viability assay used. Nevertheless, the survival of donor cells on the ocular surface after limbal stem cells allografts has been found not to correlate with clinical outcomes [40]. Additionally, hypothermia (2-8˚C) is a rather commonly used storage temperature prior to transplantation [41].
Our gene analysis revealed that GAPDH, an important enzyme for energy metabolism, was affected by temperature. Significant GAPDH downregulation was only observed at 4˚C (excluding 37˚C with no viability). Similar response has also been demonstrated for corneal cultures [42] and oral keratinocytes [16]. Based on Arrhenius equation, the principle behind hypothermic storage is that the low temperature reduces chemical reaction rates and hence declines cellular metabolic activities, which lowers the demand for oxygen and substrates [43]. The temperature dependant metabolic status, indicated as lactate/glucose ratio, has been previously shown to be inversely correlated with oxygen tension in epidermal keratinocytes [15]. Additionally, the glycolytic pathway has been suggested to play a major role in keratinocyte energy production, regardless of available oxygen level [15,[44][45][46]. For human oral keratinocytes, Islam et al [16] suggested glycolysis could be responsible for at least partial energy production at 4-37˚C after one week of storage. GAPDH is widely used as a housekeeping gene for RT-qPCR or loading control for western blot. Therefore, researchers should be mindful in selecting appropriate endogenous controls in molecular research associated with temperature. In the present study, ERBB3 was used as reference gene because its expression level was unaffected by experimental factors.
Histones are positively charged basic nuclear proteins that play important structural and functional roles in the transition between active and inactive chromatin states through compacting DNA [47]. Hence, they play a central role in transcription regulation, DNA repair, DNA replication and chromosomal stability [48]. In this study, HIST1H4B, one of the histone cluster genes, was highly upregulated at 4˚C. HIST2H4A/B, another member of the same family, was not significantly different at 4˚C when compared to non-stored baseline control. In accordance with our finding, Raeder et al. [49], have reported the upregulation of HIST1H4B in culture of human limbal epithelial cells stored at 4˚C for 2-7 days. Low temperature may induce expression of this histone H4 encoding gene in order to acclimatize the cells to hypothermia by stabilizing nucleosomes and repressing transcription, even after 2 weeks storage.
The effect of temperature on stored OMECs was also observed in the expression of genes involved in cell adhesion, migration, motility and differentiation. EFEMP1 was upregulated at 12˚C. As a member of the fibulin family of extracellular glycoproteins, it is associated with multiple structural and functional roles including morphology, growth, adhesion and motility [50]. RPSA, another multifunctional gene (involved in e.g., cell adhesion, differentiation and migration), remained unchanged at 12˚C but was downregulated at all other temperature conditions. Finally, EPHA4 was significantly upregulated at 4˚C, 12˚C and 24˚C. Its key functions are the maintenance of the basement membrane and the integrity of the extracellular matrix [51]. Among genes used as differentiation markers, PPARA and TP53 were downregulated at all four temperature conditions. CDKN1A, alias p21, responded in the same fashion. This gene has been previously shown to be under the transcriptional control of p53 [52]. Two other differentiation markers, KRT1 and KRT10, were either unchanged (4˚C) or downregulated (24˚C and 12˚C, respectively), except an upregulation at 12˚C for KRT1. The 37˚C group was not evaluated due to a majority of dead cells. The expression pattern of differentiation markers following storage have been previously reported by our group in other cultured epithelial cells [14,38,53]. Using immunostaining, Islam et al [16] presented similar results for putative stem cell markers in oral keratinocytes stored for one week at 4˚C, 12˚C and 20˚C. These results indicated that the expression levels of some adhesion-, growth-and apoptosis-related genes were favourable for cultured epithelial cells at 4˚C-24˚C.
The expression level of ΔNp63α, a stemness marker for OMECs [54,55], was significantly upregulated at 4˚C compared to other experimental groups. The clinical success in treatment of LSCD has been linked to the high number of cells with ΔNp63α expression in transplantation of cultured LSC sheets [20]. The satisfactory results have been observed by authors in 78% of patients when LSCs cultures expressed more than 3% ΔNp63α. Another important factor in transplantation is the maintenance of barrier function of cultured sheet [56], which is partly defined by cell membrane and cell junctions [57]. For example, Claudin and occludin, members of tight junctions transmembrane proteins, interact with zonula occludens proteins such as TJP1 (alias ZO-1) on cell's plasma membrane face for anchoring them to the actin cytoskeleton [58]. These markers have previously been examined in human and rabbit cultivated OMEC sheets [56]. Our study indicated that genes associated with tight junction were either upregulated (TJP1 and CLND1) or unchanged (OCLN) at 4˚C. Additionally, the culture of OMECs presented upregulation of CLND1 and OCLN at 24˚C. Although 37˚C led to highest expression level of tight junction-associated genes, storage of OMECs for two weeks is disadvantageous due to the dominance of dead cells, downregulation of stemness gene (ΔNp63α) and overexpression of keratinization markers (KRT1 and KRT10).
The upregulation of GADD45B, stress-induced growth arrest gene, is indicative of the high level of stress imposed on OMECs after two weeks of storage at 12˚C and 24˚C. GADD45B is considered a crucial factor in DNA repair, cell survival, growth arrest, apoptosis and DNA demethylation [59][60][61]. Similarly, TGFB2 showed upregulated levels at the same storage temperatures, whereas CCNH was only upregulated at 12˚C. TGFB2 belongs to a group of proteins that regulate several cellular functions, especially inhibition of cell proliferation through cell cycle arrest in G1 phase [62,63]. Its regulatory role is mediated by a series of formation, activation and subsequent inactivation of the G1 cyclin-dependent kinase (CDK) complexes, another group of proteins that are partly controlled by cyclins (e.g., CCNH) [64,65]. The levels and activities of both G1 cyclins and CDKs are directly affected by TGFB2 [66,67].
One unanticipated finding was the upregulation of TAF1D, a histone-relevant transcriptional regulation gene, at 4˚C and 24˚C. It is in contrast to downregulation of TBRG4, also termed cell cycle progression restoration protein 2, in all four experimental groups. TAF1D, the largest subunit of TFIID, is an essential component of transcription machinery which acts through harboring bromodomains and histone acetyltransferase activity [68,69]. It is difficult to explain this result, but it might be related to the association revealed between TAF1D and genotoxic/oxidative stress-induced apoptosis [70]. In this context, we observed upregulated stress-induced growth arrest gene (GADD45B) at 24˚C. Moreover, the apoptosis-associated gene CASP4 was also upregulated at 24˚C, in contrast to other downregulated groups. This was supported by our photomicrographs illustrating cells with membrane blebbing, indicative of apoptosis [71]. However, others have reported cold storage-induced apoptosis and/or necrosis in oral keratinocytes [16] and epithelial cells [71].
The list of 22 regulated genes was subjected to DAVID Functional Annotation Tool (https://david.ncifcrf.gov/) to test for over-representation in functional pathways. Unsurprisingly, Cell Cycle came out as the most significantly affected in the KEGG PATHWAY Database, and the only relevant, sustaining significance after adjustment for multiple testing by the Benjamini-Hochberg method [72] (p = 3.4E-3, and the 5 genes: CCNH, CDKN1A, GADD45B, TGFB2 and TP53). If more experiments had been performed, the statistical power would increase, and we would probably find more affected genes as well as signaling pathways giving further insight in the cellular processes affected upon storage.
Our study had some limitations. We decided not to adjust p-values for multiple testing to avoid false negative findings (type II statistical errors), at the expense of increasing the likelihood of including false positives (type I statistical errors). Another limitation was the use of cells from a single donor, at the possible risk of systemic factors' influence (e.g. donor health and age) on extrapolation of the results [73][74][75].
The results of morphology, cell viability and gene expression assays in the current study showed that 4˚C is the most appropriate temperature to store OMECs for up two weeks. Similarly, Utheim et al [32] found that storage at 4˚C and 12˚C are more suitable for transplantation than 37˚C after one-week culture of oral keratinocytes. However, these results differ from a previous study reporting 12˚C and 16˚C as optimal temperatures for cultured oral keratinocytes to maintain their structure and function after one week of storage [16]. Another similar study also found 12˚C and 24˚C superior for storage of cultured epidermal cells for two weeks [15]. Whilst our findings did not confirm their conclusions, it did support refrigeration (2˚C-8˚C), one of the current methods employed in the storage of cultured epithelial cells [41]. Although this simple technique may be practical for short-term storage, several drawbacks such as limited shelf-life, high cost, risk of contamination and genetic drift have discouraged its application for long-term preservation [76].

Conclusions
This study found that the morphology of cultured human OMECs stored at 4˚C for two weeks was most similar to non-stored baseline controls. Additionally, cells stored at 4˚C had the highest viability rate (58%) when compared to other temperatures experimental groups. Gene analysis also revealed some levels of interruption at 12˚C and/or 24˚C through e.g., stressinduced growth arrest (GADD45B) and cell proliferation inhibition genes (TGFB2). However, some genes involved in cell adhesion, migration, motility and differentiation maintained transcription levels beneficial for cultured epithelial cells at 4˚C-24˚C. Taken together, these results suggest that storing OMECs at 4˚C for two-week is the most appropriate temperature with the highest viability rate and morphology rather similar to pre-stored OME cells. Therefore, it is most likely to be the best storage conditions for clinical transplantation.
Supporting information S1 Table. The list of 30 genes examined using the TaqMan gene expression array in this study. Among 30 studied genes, 12 were omitted (non-bolded genes). This study considered a log2-fold change significant at thresholds over 1 and under -1 (i.e., > two-fold change). Additionally, this approach was only applied to 4˚C, 12˚C and 24˚C groups as all cells died upon storage at 37˚C.