Effect of different cryopreservation media on human nucleus pulposus cells' viability and trilineage potential

Abstract Introduction: Low back pain (LBP) is a significant cause of disability in many countries, affecting more than half a billion people worldwide. In the past, progenitor cells have been found within the nucleus pulposus (NP) of the human intervertebral disc (IVD). However, in the context of cell therapy, little is known about the effect of cryopreservation and expansion on here called “heterogenic” human NP cells (hNPCs), and whether commercially available cryopreservation media are more efficient than “commonly used” media in terms of cell viability. Materials: In this study, hNPCs from four trauma patients (age 40.5 ± 14.3 years) and two patients with degenerated IVDs (age 24 and 46 years), undergoing spinal surgery, were collected. To isolate hNPCs, the tissue was digested with a mild two‐step protocol. After subsequent expansion, hNPCs at passages 2‐5 were separated and either cryo‐preserved for 1 week at −150°C or differentiated into osteogenic, adipogenic, or chondrogenic lineages for 21 days. Cryopreservation was performed with five different media to compare their effect on the cell's viability and differentiation potential. Cell viability was determined with flow cytometry using propidium iodide and the trilineage differentiation potential was assessed by quantitative polymerase chain reaction and histological analysis. Results: After 1 week of cryopreservation, the hNPC's cell viability was comparable for all conditions, that is, independent of the cryopreservation medium used (82.3 ± 0.8% of cell viability). Furthermore, hNPCs from trauma patients showed some evidence for adipogenic and chondrogenic differentiation and at lower levels, this and evidence of osteogenic differentiation could be confirmed with hNPCs from degenerated discs. Moreover, cryopreservation did not affect the cell's differentiation potential in the majority of the cases tested. Conclusion: “Commonly used” cryopreservation media seem to perform just as well as commercially available media in terms of cell viability and the overall maintenance of the hNPCs trilineage differentiation potential.


| Intervertebral disc degeneration and low back pain
A healthy intervertebral disc (IVD) mainly consists of a centrally located, collagen type II rich and highly hydrated nucleus pulposus tissue (NP). The NP is surrounded by the annulus fibrosus (AF), which is composed of 15 to 25 concentric lamellae consisting of collagen type I and II fibers with elastin fibers lying in between. Finally, the IVD is enclosed by two hyaline cartilaginous endplates (CEP) above and below them. [1][2][3] The causes of IVD degeneration (IDD) are manifold. However, in many cases, IDD is initiated by a decrease of the amount of NP cells accompanied by a change of the extracellular matrix (ECM) composition, for example, by loss of the proteoglycan content. 4 Degraded proteoglycans tend to escape from the NP and osmotic imbalance results as a consequence. 5 This reduction of the osmotic pressure in the NP will then lead to its dehydration and destruction, resulting in loss of disc height. 6 Other causes of IDD are changes in collagen repartitions in the ECM. Especially type II collagen can become more and more denatured over time and therefore contribute to IDD. 7,8 While IDD per se may remain clinically silent, some conditions associated with IDD cause symptoms ranging from mild, occasional discomfort to severe, immobilizing back pain and disability. 9 Conservative management, including physical therapy, lifestyle modification, pain-and anti-inflammatory medication is the first-line treatment of back pain related to IDD. Failure of conservative measures may direct towards surgical treatment. 10,11 The spectrum of surgical procedures is broad and relies on distinct pathologies. One of the most frequently performed surgical intervention is spinal fusion with the purpose to induce osseous fusion between at least two vertebrae and thereby eliminating motion and relieving pain. 12 However, the clinical success rate of spinal fusion lies only between 50% and 70%, and more than 25% of the patients require reoperation, which still does not guarantee successful fusion. [13][14][15] Another primary concern of spinal fusion is that these interventions do not have the intention to restore the IVD, but instead leave the patient with a stiff and immobile spine, increasing the risk for adjacent segment disease. Therefore, there is a high demand for new efficient treatment techniques regarding LBP.

| Cell-based therapy for IDD
A promising approach to regenerate the IVD is cell-based therapy. Even if it is still in its infancy, cell-based therapy has received more and more attention over the last two decades. 2 Especially targeting the NP is believed to have considerable potential in this field. 2 A previous study showed the existence of NP derived mesenchymal stromal cells (MSCs) that can differentiate into an osteogenic and chondrogenic lineage. However, they were not able to differentiate into adipocytes. 16 Additionally, over the last decade, progenitor cells positive for Tie2 (aka angiopoietin-1 receptor / TEK receptor tyrosine kinase) and disialoganglioside 2 (GD2 + ) have been found in the NP of humans and other vertebrates. 17,18 It has been shown that these rare cells can form spheroid colonies in methylcellulose-based medium, and they are also able to undergo trilineage differentiation. 17 However, with increasing age and continuous IDD, the amount of Tie2 positive cells in the IVD rapidly decreases. In humans, the frequency of these cells already starts to drop before the age of 20, with more than half of the NP cells usually being positive for Tie2. 17 By the age of 50, the amount of Tie2 positive cells nearly completely vanishes. 17,19 To conclude, the differentiation potential of MSCs and Tie2 positive cells from the NP has been demonstrated in recent years.
However, little is known about the effect of expansion and cryopreservation on here called "heterogenic" human NP cell populations (hNPCs) and their stemness in the context of cell therapy for regeneration of the IVD. 19 Thus, we hypothesized that "heterogenic" hNPCs could undergo trilineage differentiation. Therefore, freshly isolated hNPCs and previously frozen hNPCs (from the same donors) were cultured either in osteogenic inductive medium, adipogenic inductive medium, chondrogenic inductive medium, or control medium. After 21 days of culture, multi-lineage differentiation properties were compared using histology, absorbance measurements, and quantitative polymerase chain reaction (qPCR).

| Cryopreservation
Cryopreservation has become an indispensable technology for biomedical research but also clinical medicine and it is even seen as the gold standard for bio-preservation. 20,21 However, freezing cells at very low temperatures (−150 C) for preservation faces many challenges in terms of cell viability. For example, the freezing rate plays a critical role in successful cryopreservation as it underlies a delicate balance between too fast and too slow cooling. 20 If the cooling process happens too fast, liquid water in the cell's body will undergo a phase change and form ice crystals which potentially damage the cells. 22 In contrast, a too slow cooling rate will first cause extracellular ice formation and as a consequence create an osmotic imbalance between the intracellular and extracellular space. As more and more extracellular ice forms, the concentration of particles in the residual water will further increase causing greater osmotic imbalance and an additional depression of the freezing point. 23,24 As a result, intracellular water will efflux across the cell's membrane leading to dehydration and in the worst case cause cell death. 25 To avoid this ice crystal formation and to preserve osmotic balance, a cryoprotectant is added to the Mammalian cells and then everything is usually cooled at a rate of −1 C per minute. 26 By definition, a cryoprotectant is a solute added to the medium that enables higher cell recovery after thawing compared to its absence. 27 Many different cryoprotectants exist, one of which being dimethyl sulfoxide (DMSO), which is most commonly used for mammalian cells. 28,29 DMSO can cross the cell membrane and act as a solvent for salts. As a consequence, the osmotic pressure diminishes, the freezing point decreases, and intracellular ice crystal formation is depressed. 24,26 In the past, studies have shown that cryopreservation of human adipose-derived mesenchymal stromal cells (ASCs) and bone marrowderived mesenchymal stromal cells (BMSCs) does not affect their viability nor their trilineage differentiation potential. 30,31 However, an optimal way to cryo-preserve hNPCs is yet to be found for IVD cells in general and only very little is known about the influence of cryopreservation on the cell viability and stemness of hNPCs.
Thus, we hypothesized that hNPCs can be cryo-preserved just as well with "commonly used" cryopreservation media, made by simply adding 10% dimethyl sulfoxide (DMSO), as with more complex commercially available media without any significant differences concerning cell viability. Additionally, we wondered if hNPCs can undergo trilineage differentiation and if cryopreservation would affect this ability. Therefore, primary hNPCs were isolated from either trauma patients or patients suffering from degenerated IVDs undergoing spinal surgery. We cryopreserved the isolated hNPCs for 2 days at −80 C followed by 1 week's storage at −150 C with five different cryopreservation media. After testing the cell viability using flow cytometry, the hNPCs' trilineage differentiation potential was assessed.
Cells from the AF the CEP were frozen and the "heterogenous" hNPC population was expanded until a sufficient number of cells were accomplished for future experiments. During cell expansion, the medium was changed twice a week. Cells were cultured at 37 C with normoxic conditions (20% O 2 ) and 5% CO 2 .

| Cell freezing
To find the most suitable cryopreservation medium for hNPCs, two "commonly used" and three commercially available cryopreservation media were tested. For the "commonly used" media we used either a commonly applied cryo-medium for progenitor cells (CMPC) that consisted of 90% FBS and 10% DMSO or a commonly applied cryo-medium for differentiated cells (CMDC) that consisted of 90% LG-DMEM "complete medium" and 10% DMSO. In addition, three commercially available cryopreservation media were tested. One of them was the CellnTec The hNPCs were frozen at a density of 500 000 cells/mL of the respective cryopreservation medium. In brief, the cell suspensions were first stored in a pre-cooled Nalgene Cryo 1 C freezing container (#5100-0001; Thermo Fisher Scientific) filled with 2-propanol for 2 days at −80 C to reduce ice crystal formation. Then, the cells were removed from the freezing container and stored for 7 days at −150 C.

| Cell thawing
To maximize hNPCs' viability, subsequent procedures were followed: First, LG-DMEM "complete medium" was put into a prewarmed water bath (37 C). As soon as the medium reached the same temperature as the water bath, the samples were removed from the −150 C freezer.
Under sterile conditions, the cryovial's caps were twisted a quarterturn to relieve internal pressure and then retightened. Then they were moved into the prewarmed water bath and gently shaken until only a small ice pellet remained in the vials. Afterwards, the cryovials were sprayed with ethanol (80%), and under sterile conditions, the cell suspensions were then diluted 1:10 with warmed LG-DMEM "complete medium" followed by centrifugation (500 g for 5 minutes). After centrifugation, the supernatant was discarded and the cells were resuspended with LG-DMEM "complete medium". To reach a sufficient number of cells for trilineage differentiation, the cells were first expanded in LG-DMEM "complete medium" after the thawing process.

| Cell viability
To assess the most effective cryopreservation medium for NP cells in Medium change was performed twice a week.

| Chondrogenic differentiation
The chondrogenic differentiation potential was tested with hNPCs cultured as 3D pellets. Therefore, between 250 000 and 300 000 hNPCs were centrifuged for 5 minutes at 500 g to a cell pellet and were cultured for 21 days and the medium was changed twice a week.

| Histology
Histological staining was performed for the trilineage differentiation assays after 21 days of culture. Following steps were performed:

| Chondrogenic lineage
At the end of the differentiation assay, the 3D pellets were washed with PBS, fixed overnight in a 4% formalin solution at 4 C. The next day, the formalin was replaced with PBS and the pellets of each condi-

| Quantification of the glycosaminoglycan content in the Supernatant
To study the chondrogenic differentiation of the 3D pellets, the concentration of glycosaminoglycans (GAG) in the supernatant was measured at day 21 of cell culture, with the last exchange of medium being carried out 4 days before the end of the culture. Therefore, a stock solution of dimethylmethylene blue (DMMB) was prepared, consisting of 0.0021% 1.9-dimethyl-methylene blue zinc chloride double salt (#341088, Sigma-Aldrich), 0.2% sodium formate (#71541, Fluka), absolute ethanol, and distilled water. By adding formic acid, the pH was adjusted to 1.5. The DMMB was then added to the supernatant of the respective samples and within 5 minutes, the absorbance could be recorded at a wavelength of 595 nm (SpectraMax M5).

| Statistical analysis
For all data, a non-parametric distribution was assumed, and the results are presented as mean ± standard deviation (SD  This indicates that "commonly used" cryopreservation media and commercially available media both performed the same in terms of cell viability (Figure 1). genes did not differ from the control in any significant way ( Figure 2G).

| Osteogenic differentiation
On the other hand, hNPCs from degenerated IVDs did not show any microscopical evidence of matrix mineralization nor any indication F I G U R E 2 Representative images of the osteogenic differentiation assays. A, The hNPCs from trauma IVDs cultured in control medium or from the CS10 and MesenCult condition and then stained with Alizarin red. B, The hNPCs from trauma IVDs cultured in osteogenic inductive medium and stained with Alizarin red. C, Alizarin red stained single well of a 12-well plate with hNPCs from trauma IVDs that were previously frozen in CMPC and then cultured in osteoinductive medium. D, Alizarin red stained single well of a 12-well plate with hNPCs from trauma IVDs that were previously frozen in CMDC and then cultured in osteogenic inductive medium. E, Alizarin red stained single well of a 12-well plate with hNPCs from trauma IVDs that were previously frozen in CnT and then cultured in osteogenic inductive medium. F, Relative quantification of the amount of solubilized Alizarin red from either fresh or previously frozen trauma hNPCs that were cultured in osteogenic inductive medium relative to cells cultured in control medium, which are set at baseline 100% (mean ± SD, N = 4 for Control, Osteo, CMPC, CMDC and CnT, N = 1 for CS10 and MesenCult). G, Osteogenic gene expression profile of trauma hNPCs relative to the control at baseline 1 (mean + SD, N = 4 for Control, Osteo, and CnT; mean + SD, N = 3 for CMPC and CMDC, N = 1 for CS10, MesenCult, and all osterix (OSX) results). H, Relative quantification of the amount of solubilized Alizarin red from either fresh or previously frozen hNPCs from degenerated IVDs that were cultured in osteogenic inductive medium relative to cells cultured in control medium, which are set at baseline 100% (mean ± SD, N = 2). I, Osteogenic gene expression profile of hNPCs from degenerated IVDs relative to the control at baseline 1 (mean + SD, N = 2 for all samples except OCN is N = 1 due to poor quality results of one donor). Nomenclature: Control: Fresh hNPCS cultured in control medium; Osteo: Fresh hNPCS cultured in osteogenic inductive medium; CMPC: hNPCs cryopreserved in CMPC and subsequently cultured in osteogenic inductive medium; CMDC: hNPCs cryopreserved in CMDC and subsequently cultured in osteogenic inductive medium; CnT: hNPCs cryopreserved in CnT and subsequently cultured in osteogenic inductive medium; CS10: hNPCs cryopreserved in CS10 and subsequently cultured in osteogenic inductive medium; MesenCult: hNPCs cryopreserved in MesenCult and subsequently cultured in osteogenic inductive medium. The following genes were analyzed: Collagen type I (COL1), runt-related transcription factor 2 (RUNX2), osteocalcin (OCN), alkaline phosphatase (ALP), osteopontin (OPN) and osterix (OSX). *P < .05, no indication: P ≥ .05 of osteogenesis when the absorbance of solubilized Alizarin red was measured ( Figure 2H). Furthermore, no osteogenic related gene was significantly up-or downregulated. Nevertheless, osteogenic related genes of degenerated samples were mostly stronger expressed, however not significantly, than trauma samples. This includes the genes COL1, RUNX2, OCN, OPN, and OSX ( Figure 2I).
F I G U R E 3 Legend on next page.  Figure 3G).

The quantification of solubilized Oil-Red-O from degenerated
IVDs, hNPCs cultured in adipogenic inductive medium also revealed no significant differences compared to the control ( Figure 3H). However, in contrast to the trauma cells, no lipid droplets could be found microscopically.  Figure 3I).

| Cryopreservation
This study showed that hNPCs can be cryo-preserved with "commonly used" cryopreservation media just as well as with commercially available media without any significant differences concerning cell viability. Furthermore, no changes in cell proliferation were observed when comparing the different formulations. An explanation for the similar performance of all five cryopreservation media could be that all formulations are DMSO based and consequently have a very comparable mechanism of action.
These results nicely show that researchers, working with hNPCs, can just as well cryo-preserve their cells in "commonly used" cryo-medium, which are potentially more affordable than commercially available media.
Throughout the literature, many studies showed comparable results with different cell types regarding their viability after cryopreservation. 30 To our knowledge, no study analyzed the effect of cryopreservation on the viability and the trilineage potential of hNPCs. However, Hiraishi et al. (2018) transplanted cryo-preserved human NP cells into degenerated canine IVDs. They demonstrated, that thawed cells can arrest further degeneration by maintaining matrix features just as well as precultured cells. 33 Nevertheless, DMSO-based products still face some challenges for successful future cell-based therapy. Despite the beneficial cryoprotective effects of DMSO, it is also known to be chemically toxic for cells, particularly in its liquid phase. 34 In clinical use, DMSO has been shown to cause mild to severe adverse reactions with human tissues. [35][36][37] Hence, it would be essential to properly remove any DMSO residues from frozen-thawed samples before considering regulatory affairs for cell therapy in patients in a good manufacturing practise (GMP)-compliant environment. 38  showed that a mixed NP cell population manages to form some mineralization, but much less than pure Tie2 positive cells and therefore supports our findings. 19

| Study weaknesses and limitations
One limitation of this study is that only the cell viability and their differentiation potential before and after freezing was analyzed. Other parameters that could have been taken into consideration for the freezing assays would have been the cell proliferation rate, the cell's morphology, metabolism, or cell surface markers. However, a recent study involving BMSCs showed that cryopreservation does not affect the cell's surface markers, their proliferation rate, or their morphology. 31 Even though BMSCs are not the same as hNPCs, similar results are likely to be expected with hNPCs. Another limitation concerning the freezing assay is that no long-term data are available. Due to the strict and tight schedule of this study, cells could be frozen at −150 C only for a single week. Nevertheless, as the cellular metabolism almost comes to a standstill with so low temperatures and therefore, it can be assumed that no big changes are expected even after multiple weeks or even months of cryopreservation. 29 An additional limitation of these types of donor studies was that trauma patients cannot be strictly be separated from patients with degenerated IVDs. We cannot fully exclude that trauma patients might have had an age appropriate degenerated IVD as for these patients no pre-operative magnetic resonance image (MRI) was taken.
Another strong limitation of this study was that no Tie2 positive

| Conclusion
In this study, the freezing assay showed that "commonly used" cryopreservation media seem to be just as efficient as the commercially available media in terms of cell viability for hNPCs. Therefore, it is more economical using either a commonly applied cryo-medium for progenitor cells or one for differentiated cells instead of a commercially available cryopreservation medium. Regarding the trilineage differentiation potential of hNPCs, trauma samples showed sporadic mineralization of their ECM but no indication of differentiation based on the results of osteogenic related genes. Concerning the trauma cell's adipogenic differentiation potential, the occasional production of lipid droplets was observed, however, when quantifying the solubilized Oil-Red-O, no significant differences were found compared to the control. Nevertheless, adipogenic related genes like ADPN and CEBPα showed relatively high, but non-significant, upregulations.
Trauma cell pellets cultured in chondrogenic inductive medium and subsequently stained with Alcian blue revealed a higher production of GAG than controls and COL2 to COL1 ratios of up to 1000. However, the amount of GAG released into the supernatant was similar for all samples. Regarding the samples derived from degenerated discs, the adipogenic and chondrogenic differentiation potential was generally less pronounced in hNPCs from degenerated IVDs than from trauma discs. Surprisingly, however, hNPCs from degenerated samples showed mostly a higher, but non-significant, expression of osteogenic related genes than trauma cells, indicating a more successful differentiation towards osteogenesis than trauma samples.
Overall, cryopreservation did not broadly affect the differentiation potential of hNPCs, indicating that hNPCs are perfectly fine if they are being cryo-preserved. However, extensive donor variations could be observed, which were maintained before and after freezing.
Nevertheless, the results are promising and therefore clinical trials considering hNPCs for cell-based IVD regeneration could be taken into consideration.

ACKNOWLEDGMENTS
This study was supported by the iPSpine H2020 project (https://