Proliferation, Migration, and ECM Formation Potential of Human Annulus Fibrosus Cells Is Independent of Degeneration Status

Abstract

Objective

The objective was to evaluate the proliferating, migratory and extracellular matrix (ECM) forming potential of annulus fibrosus cells derived from early (edAFC) or advanced (adAFC) degenerative tissue and their usability as a possible cell source for regenerative approaches for AF closure.

Design

EdAFC (n = 5 Pfirrman score of 2-3) and adAFC (n = 5 Pfirrman score of 4-5) were isolated from tissue of patients undergoing spine stabilizing surgery. Cell migration on stimulation with human serum (HS), platelet-rich plasma (PRP), and transforming growth factor β-3 (TGFB3) was assessed by migration assay and proliferation was assessed on stimulation with HS. Induction of ECM synthesis was evaluated by gene expression analysis of AF-related genes in three-dimensional scaffold cultures that have been stimulated with 5% PRP or 10 ng/mL TGFB3 and histologically by collagen type I, type II, alcian blue, and safranin-O staining.

Results

EdAFC and adAFC were significantly attracted by 10% HS and 5% PRP. Additionally, both cell groups proliferated under stimulation with HS. Stimulation with 10 ng/mL TGFB3 showed significant induction of gene expression of collagen type II and aggrecan, while 5% PRP decreased the expression of collagen type I. Both cell groups showed formation of AF-like ECM after stimulation with TGFB3, whereas stimulation with PRP did not.

Conclusions

Our study demonstrated that AF cells retain their potential for proliferation, migration, and ECM formation independent of the degeneration status of the tissue. Proliferation, migration, and ECM synthesis of the endogenous AF cells can be supported by different supplements. Hence, endogenous AF cells might be a suitable cell source for a regenerative repair approaches.

Introduction

Low back pain is a widespread disease often related to degenerative changes in the intervertebral disc (IVD). In a healthy IVD, the proteoglycan rich nucleus pulposus (NP) absorbs the body load transmitted to the spine while being surrounded and held in place by the annulus fibrosus (AF). Over time, the proteoglycan concentration decreases from 70% in juveniles to 20% of dry weight in adults mostly by the loss of aggrecan, the major extracellular matrix component in the NP.¹ The loss of proteoglycan is accompanied by dehydration resulting in a reduced disc height leading to an altered load onto the AF and the neighboring intervertebral bodies.² The additional weight increases the risk for fissure and crack formation in the AF as well as a loosening of the tight lamellae structure resulting in the bulging of the AF.³ The degenerative damage can culminate in the protrusion of the NP into the degenerated AF followed by the prolapse of the NP. The herniated disc often creates low back pain by directly pressuring the nerves of the spinal cord or by the initiation of an inflammatory environment.⁴

Conventional treatment comprises physiotherapy, anti-inflammatory medication as well as analgesics and is sufficient to relief the pain in most patients.^5,6 But in a low percentage of the patients, a surgical procedure is necessary to alleviate the pain. Thus, the herniated disc is usually removed by a lumbar discectomy, the standard procedure annually carried out in more than 300,000 cases in the United States.^6,7 But none of these methods are completely successful. Current research focuses on the regeneration potential of the NP by different scaffold materials, but since the damaged AF is not closed/repaired during this procedure, up to 23% of the patients suffer from recurring herniation of the NP⁸ or the applied implant. Moreover, AF has little self-healing potential,⁴ which emphasizes the major need in the repair of the AF to achieve regeneration as well as to prevent progressive degeneration of the IVD. Nevertheless, sutures for AF closure are not able to provide enough mechanical stability to resist the intradiscal pressure and the tensile forces applied to the AF.⁹ Addition of fibrin in the suture process showed a higher resistance than suture alone, but is still not sufficient enough to attain long-term closure of the AF.¹⁰ Application of solid mechanical barriers to close the AF such as barb rings often demonstrated dislocation as well as destruction of the device.¹¹ A different approach is the usage of biomaterials for AF defect closure that serve as 3-dimensional (3D) structures for cells providing initial stability of the tissue and an initiation point for tissue regeneration.⁶ Several studies investigated cell-based strategies where scaffolds are combined with cells.^12-14 But healthy AF tissue is not always available as cell source without creating a new defect, thus favoring cell-free regeneration approaches. In a cell-free approach a biomaterial is implanted to close/cover the defect and provide endogenous cells with a scaffold for extracellular matrix (ECM) formation. In this regenerative approach, the cell source are body’s own cells from the surrounding tissue that migrate into the defect, proliferate, and synthesize ECM. In daily clinical practice, disc protrusion or prolapse is usually accompanied by tissue degeneration. Thus, remaining annulus and nucleus tissue is usually affected by degenerative changes, which might influence the regenerative potential of cells contained therein.

In observational case studies, polyglycolic acid–hyaluronan (PGA-HA) scaffolds immersed with human serum (HS) or platelet-rich plasma (PRP) have proven to be a suitable implant for cell-free articular cartilage repair.^15,16 First promising results in NP regeneration using cell-free PGA scaffolds immersed with HS were shown in a rabbit animal model.¹⁷ Moreover, it was shown that stimulation with transforming growth factor β-3 (TGFB3), fibroblast growth factor-2 (FGF2), or HS can effectively induce the formation of cartilaginous matrix and enhance the expression of ECM-associated molecules in human AF cells in 3D cultures in vitro.¹⁸

To evaluate if AF cells from early and advanced degenerative tissue are a suitable endogenous cell source for AF repair, we analyzed their proliferation and migration potential. Additionally, ECM formation was evaluated in 3D culture using PGA-HA scaffolds.

Methods

Isolation and Cultivation of Human Annulus Fibrosus Cells

Cells derived from human edAF or adAF tissue were obtained from 10 patients (edAFC n = 5, mean age 48 years, 2 male, 3 female, Pfirrman score 2, n = 3 and 3, n = 2; adAFC n = 5; mean age 63 years, 1 male, 4 female, Pfirrman score 4, n = 2 and 5, n = 3) undergoing spine stabilizing surgery ( Fig. 1 ). Degeneration status of the IVDs were classified after the Pfirrmann score by the surgeon, whereby 0 describes a healthy disc and 5 a totally collapsed disc with hypointense black signal intensity in the magnetic resonance image. Patients with a NP prolapse were excluded, because a clear distinction between AF and NP tissue during surgery is almost impossible. Therefore, only patients with an intact AF undergoing stabilizing surgery due to instability of the lumbar spine were included.

Figure 1.

Representative magnetic resonance images of the spine for a patient with early degenerated annulus fibrosus tissue (A, Pfirrmann score of 2) and a patient with advanced degenerated annulus fibrosus tissue (B, Pfirrmann score of 5) included in this study.

Human AF cells from both groups were isolated individually by enzymatic digestion with 45 U collagenase P (Roche, Switzerland) and 10,000 U collagenase II (Merck-Millipore, Germany) in 30 mL Dulbecco’s modified Eagle medium (DMEM) containing 100 U/mL penicillin, 100 mg/mL streptomycin (all Merck-Millipore), and 10% HS (German Red Cross, Germany) for 18 to 20 hours at 37°C and 5% CO₂. After digestion, the cell suspension was filtered through a nylon mesh (size 100 µm, Becton Dickinson, USA) and centrifuged for 10 minutes at 583 × g. The obtained cell pellet was resuspended in DMEM containing 100 U/mL penicillin, 100 mg/mL streptomycin (all Merck-Millipore), 10% HS (German Red Cross), and 2 ng/mL FGF-basic (PeproTech, Germany) and seeded into cell culture flasks at a density of 10,000 cells/cm². Medium was changed every 2 to 3 days. The study was approved by the ethics committee of the Charité-Universitätsmedizin Berlin (EA2/102/15).

Preparation of Platelet-Rich Plasma

PRP (n = 3 individual batches) was pooled in equal amounts and stored at −20°C. Each batch of PRP was prepared by the German Red Cross from 4 to 6 anonymous blood donors with the same blood type. Number of platelets was 0.7 to 1.8 × 10⁹/mL and leukocyte count was less than 0.5 × 10⁴/mL per batch. To activate platelets and to prevent medium coagulation by residual fibrinogen, 3 freeze-and-thaw cycles were performed as described previously.^19,20 In brief, PRP was thawed at 4°C, centrifuged at 1,600 × g for 10 minutes, residual fibrinogen was discarded, and the supernatant was stored at −20°C.

Growth Kinetics of AF Cells

To determine the growth kinetic, human edAFC or adAFC (n = 5 patients per cell group) were cultivated with DMEM containing 10% HS, 100 U/mL penicillin, 100 µg/mL streptomycin (all Merck-Millipore), and 2 ng/mL FGF2 (PeproTech) at 37°C and 5% CO₂. After reaching confluence, vital cells were detached, counted with trypan blue, and seeded with a cell density of 10,000 cells/cm². All in all, cells were cultivated up to 32 days and medium was changed every 2 to 3 days. For each individual patient the specific growth rate (µ), the cell doubling time (t_d) and the population doubling rate (n) were calculated as follows:

$$\mathrm{\mu }=\frac{\ln numberofcellsattime\left(N_t\right)-\ln initialcellnumber\left(N_0\right)}{timeperiod\left(\mathrm{\Delta }t\right)}$$

$$t_d=\frac{\ln 2}{specificgrowthrate\left(\mathrm{\mu }\right)}$$

$$n=\frac{\log \frac{initialcellnumber}{terminalcellnumber}}{\log 2}$$

Cell Migration Assay

To determine the migration potential of edAFC or adAFC (n = 3 patients for each group), a 96-multiwell migration assay (Neuro Probe, USA) was used. In brief, HS (10%), PRP (5%), and TGFB3 (10 ng/mL) diluted with DMEM containing 0.1% HS, 100 U penicillin, and 100 µg/mL streptomycin was filled into the lower wells of the multiwell plate and covered with a polycarbonate membrane (pore size of 8 μm). DMEM containing 0.1% HS served as control. 30,000 edAFC or adAFC at passage 3 in DMEM containing 0.1% HS were pipetted into the upper compartment of each well. The assay was performed in triplicates and incubated at 37°C and 5% CO₂ for 20 hours. Afterward, the membrane was fixed with methanol/acetone (1:1 v/v) for 5 minutes and the remaining AF cells on top of the membrane were carefully wiped off. Finally, cells underneath the membrane were visualized using a Hemacolor staining kit (Merck-Millipore, Germany). Migrated cells were enumerated microscopically by counting the number of stained cells in 4 representative fields of each well using ImageJ software (National Institutes of Health, USA) and extrapolated to the complete well area. The migration indices (number of migrated cells per tested substance divided by number of migrated cells in the control group; 0.1% HS) of both cell groups are shown.

Three-Dimensional Cultivation of Human AF Cells in PGA Scaffolds

Three-dimensional cultures were prepared as described previously.²¹ Briefly, nonwoven PGA-HA scaffolds (BioTissue AG, Switzerland) with a pore size up to 200 µm were cut into pieces of 10 × 10 × 1.1 mm. Human AF cells (n = 5 for early and advanced degenerative tissue each) were harvested after passage 3 and counted. Cells derived from early or advanced degenerated human AF were pooled in equal amounts (n = 5 for early and advanced degenerated tissue each) to prepare 3D cultures. Therefore, 2.2 × 10⁶ of the pooled cells were resuspended in 33% (v/v) fibrinogen (Tissucol, Baxter, USA) in DMEM and immersed into the PGA-HA scaffold. Polymerization of fibrin was achieved by adding 20 µL thrombin (1:10 in phosphate buffered saline [PBS], Tissucol, Baxter, USA) to each scaffold with subsequent incubation at 37°C for 15 minutes. Each scaffold was transferred into a well of a 6-well plate and cultured with 5 mL DMEM (high glucose) containing 100 U/mL penicillin, 100 mg/mL streptomycin, 1% insulin–transferrin–selenium + 1 (ITS), 1 mM sodium pyruvate, 0.35 mM l-proline, 0.17 mM l-ascorbic acid-2-phosphate, and 0.1 mM dexamethasone (all Sigma-Aldrich) to verify ECM formation potential. ECM formation was induced by addition of 5% PRP or 10 ng/mL TGFB3. Scaffolds without inducer served as control. Medium changes were performed every 2 days and the cell-loaded scaffolds were turned upside down. To verify the cytocompatibility samples for life/dead staining with propidium iodide/fluorescein diacetate (PI/FDA) and cell metabolism assay (MTS) were taken at days 7 and 14 for both cell types. To assess the ECM development of AF cells, samples for gene expression analysis were taken at days 0, 7, 14, and 21. For histology and immunohistochemistry samples were taken at day 21.

Cytocompatibility of PGA Scaffolds Seeded with AF Cells

To verify metabolic activity of AF cells in 3D culture, an MTS assay (CellTiter 96, Promega, Germany) (n = 3 individual measurements at time point per group) was performed according to the manufacturer’s recommendations. In brief, scaffold cultures were transferred in a 12 well-plate and covered with 3 mL DMEM (high glucose) containing 1% ITS, 1 mM sodium pyruvate, 0.35 mM l-proline, 0.17 mM l-ascorbic acid-2-phosphate, and 0.1 mM dexamethasone. AQueous One Solution Reagent (100 µL per well) was added and 3D cultures were incubated for 4 hours at 37°C. Finally, samples were diluted 1:10 in 96-well plates with PBS and adsorption was measured photometrically at 490 nm using a microplate reader (BioTek, USA).

For live/dead staining, 3D cultures of cells derived from early or advanced degenerated AF were incubated with 3 µg/mL FDA solution (Sigma-Aldrich) for 15 minutes at 37°C. In the second step, samples were incubated with 0.1 mg/mL PI solution (Sigma-Aldrich) for 2 minutes at room temperature. The scaffolds were washed twice with PBS and analyzed directly under the fluorescence microscope (Olympus CKX41, Germany).

Histological and Immunohistochemical Staining of 3D Cell Cultures

Scaffolds were embedded in OCT (optimal cutting temperature) compound (Sakura, Netherland), frozen and cryosections (6 µm, n = 3 per sample) were prepared. Proteoglycan accumulation was visualized by alcian blue and safranin-O staining. Therefore, cryosections were stained using alcian blue 8GX solution at pH 2.5 (Roth, Germany) and counterstained with nuclear fast red (Sigma-Aldrich). Additionally, cryo-sections were stained with 0.7% safranin-O staining solution (Sigma-Aldrich) and counterstained with 0.2% fast green solution (Sigma-Aldrich). To analyze collagen type I and type II accumulation cryosections (n = 3 per group) were pretreated with 50 U/mL hyaluronidase (Sigma-Aldrich) for 30 minutes at room temperature. Afterward, sections were stained with rabbit anti-human collagen type I or type II antibodies, respectively (Acris, Germany) for 40 minutes at 37°C followed by colorimetric detection with 3-amino-9-ethylcarbazole (EnVision, Dako, Denmark) and counterstained with hematoxylin (Merck-Millipore).

Gene Expression Analysis for Evaluation of ECM Formation Capacity

To evaluate the gene expression of genes related to IVD cells ( Table 1 ), total RNA from early and advanced degenerative AF scaffold cultures were isolated at days 0, 7, 14, and 21 as described previously.²² Subsequently, 3 µg of total RNA was reversely transcribed with the iScript cDNA Synthesis Kit according to the manufacturer’s instructions (BioRad, Germany). The relative expression level of the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used to normalize the samples. Real-time reverse transcription polymerase chain reaction (RT-PCR) using the i-Cycler PCR System (BioRad) was performed with 1 μL of each cDNA sample in triplicates using the SYBR green PCR Core Kit (Applied Biosystems, USA).

Table 1.

Oligonucleotide Sequences, Concentrations, and Annealing Temperatures.

Gene Name	Accession Number	Temperature (°C)	Oligonucleotide Concentration Up/Down (pM)	Oligonucleotide (5′→3′)(Up/Down)	Base Pairs
Collagen type I	NM_000088	62	5/5	CGA TGG CTG CAC GAG TCA CAC/ CAG GTT GGG ATG GAG GGA GTT TAC	180
Collagen type II	NM_001844	62	1/1	CCG GGC AGA GGG CAA TAG CAG GTT/ CAA TGA TGG GGA GGC GTG AG	128
Collagen type III	NM_000090	52	5/5	GGT TTT GCC CCG TAT TAT GGA/ AGT TTC TAG CGG GGT TTT TAC GAG	142
Aggrecan	NM_001135	58	5/5	GGC TGC TGT CCC CGT AGA AGA/ GGG AGG CCA AGT AGG AAG GAT	163
Biglycan	NM 001711	60	5/5	GCT GCC CCT GCT CTC CCA CCA CA/ GAA ATG CAT GAG GAG GAG GAA CAG AAC	217
Decorin	NM 001920	57	5/5	ACT TCT GCC CAC CTG GAC ACA ACA C/ AAT GGC AGA GCG CAC GTA GAC ACA	128
Glyceraldehyde-3-phosphate dehydrogenase	NM_002046	62	1/6	GGC GAT GCT GGC GCT GAG TAC/ TGG TCC ACA CCC ATG ACG A	149

For presentation of the results, the efficiency corrected fold change²³ is presented separately for the edAFC and adAFC at days 7, 14, and 21 compared with day 0. Differences were considered biologically meaningful at a fold change of <−2 or >2 and statistic significant P values (P < 0.05) compared with day 0 for edAFC or adAFC, respectively.

Statistical Analysis

Statistical analysis was performed using SigmaStat 3.5 software (Systat Software GmbH, Germany). The Anderson-Darling test was used to determine normal distribution of data. Test results confirmed parametric distribution of all data. Subsequently, differences in the same cell group (early or advanced degenerative cells) were analyzed by one-way analysis of variance with the Holm-Sidiak test as post hoc analysis against 0.1% HS for migration assay, noninduced for metabolic activity assay and monolayer expanded cells used for 3D culture (day 0) for gene expression analysis. P values <0.05 were considered statistically significant.

Results

Proliferative Characteristics of Human AF Cells from IVD with Early or Advanced Degenerative Changes

The growth kinetics of edAFC and adAFC were assessed over 32 days. A specific growth rate of 1.42 ± 0.38 per day for edAFC and 1.38 ± 0.44 per day for adAFC, a cell doubling time of 0.52 ± 0.13 days for edAFC and 0.53 ± 0.18 days for adAFC, and a population doubling rate of 6.28 ± 1.61 for edAFC and 5.85 ± 1.98 for adAFC were determined.

HS and PRP, but Not TGFB3 Had a Migratory Effect on Both Cell Groups

The migration potential of both cell groups was examined in the presence of 10% HS, 5% PRP and were compared with 0.1% HS which served as control. Additionally, potential of 10 ng/mL TGFB3 was determined ( Fig. 2 ). An increased cell number was counted in the presence of 10% HS resulting in a migration index of 5.9 ± 0.5 for edAFC and 2.8 ± 0.2 for adAFC. Similar results were observed for 5% PRP leading to a respective migration index of 7.3 ± 0.4 and 2.9 ± 0.2. In contrast, TGFB3 had no influence on the migration of both cell groups (migration index of edAFC 0.8 ± 0.1 and adAFC of 1 ± 0.1).

Figure 2.

Migratory effect of human serum (HS), platelet-rich plasma (PRP), and transforming growth factor β-3 (TGFB3) on annulus fibrosus (AF) cells derived from early or advanced degenerated AF tissue. Medium with 10% HS or 5% PRP significantly stimulated (*P < 0.05) the migration of both cell groups compared with 0.1% HS, while 10 ng/mL TGFB3 did not. The bars show the mean (n = 3, triplicates) and the standard deviation (SD).

Increased Metabolic Activity in AF Cell 3D Cultures in the Presence of PRP and TGFB3

Since in vitro expansion of cells is known to lead to their dedifferentiation, the ECM formation potential was analyzed after cells were expanded up to passage 3, seeded into a PGA-HA scaffold and cultivated in differentiation medium with either 5% PRP or 10 ng/mL TGFB3. Cytocompatibility of the analyzed cell groups with the PGA-HA was verified by measuring an increased relative metabolic activity of both cell groups in the presence of PRP and TGFB3 at day 7 and day 14 compared with the noninduced control ( Fig. 3 ). Additionally, live/dead staining showed viable cells for both cell groups independent from the inducer or investigated time of sampling ( Fig. 3 ).

Figure 3.

Metabolic activity of annulus fibrosus (AF) cells derived from early or advanced degenerated AF tissue seeded into a polyglycolic acid–hyaluronan (PGA-HA) scaffold is enhanced when cultivated in differentiation medium with either 5% platelet-rich plasma (PRP) or 10 ng/mL transforming growth factor β-3 (TGFB3). The bars show the mean (technical triplicates) and the standard deviation (SD). Significantly (P < 0.05) *induced metabolic activity compared with noninduced AF cells. Propidium iodide/fluorescein diacetate (PI/FDA) staining verified cell viability (green staining) in PGA-HA scaffolds cultivated in noninduced, 5% PRP or 10 ng/ml TGFB3 supplemented medium for AF cells derived from early or advanced degenerated AF tissue (For interpretation of the references to colours in this figure legend, refer to the online version of this article).

Influence of PRP and TGFB3 of AF-Related Gene and Protein Expression in 3D Cultures

The expression of collagens ( Fig. 4 ) and proteoglycans ( Fig. 5 ) by edAFC and adAFC cultured three-dimensionally was examined at the gene level at days 0, 7, 14, and 21 and additionally at the protein level at day 21 ( Figs. 4 and 5 ). Gene expression was compared with expanded cells at the end of passage 3 (day 0). In the presence of PRP, edAFC showed a significant (P < 0.05; fold change >2) increased collagen type II gene expression at days 7, 14, and 21, while collagen types I and III were only significantly (P < 0.05; fold change >2) elevated on day 7 and decreased at days 14 and 21 (P < 0.05; fold change <−2). Furthermore, treatment with TGFB3 had no effect on collagen types I and III gene expression, even though it significantly (P < 0.05; fold change >2) increased the gene expression of collagen type II on all examined time points. Three-dimensional cultivation (noninduced) significantly (P < 0.05; fold change >2) increased the gene expression of all analyzed collagens independent of the investigated time point, except for collagen type I at day 21. Immunohistochemical staining of 3D edAFC cultures after 21 days demonstrated the presence of collagen type I in all treatment groups, while collagen type II is only detectable when treated with TGFB3.

Figure 4.

Collagen expression of annulus fibrosus (AF) cells derived from early or advanced degenerated AF tissue seeded in polyglycolic acid–hyaluronan (PGA-HA) scaffolds cultured without inducer (noninduced), 5% platelet-rich plasma (PRP), or 10 ng/mL transforming growth factor β-3 (TGFB3). The bars show the mean (technical triplicates) and the standard deviation (SD). Significantly *induced (fold change >2 and P < 0.05) or ^#decreased (fold change <−2 and P < 0.05) gene expression compared with day 0. Corresponding immunohistochemical staining of collagen types I and II in PGA-HA scaffolds seeded with cells from early and advanced degenerated AF tissue cultured without inducer, 5% PRP or 10 ng/mL TGFB3 at day 21.

Figure 5.

Proteoglycan expression of annulus fibrosus (AF) cells derived from early or advanced degenerated AF tissue seeded in polyglycolic acid–hyaluronan (PGA-HA) scaffolds cultured without inducer (noninduced), 5% platelet-rich plasma (PRP), or 10 ng/mL transforming growth factor β-3 (TGFB3). The bars show the mean (technical triplicates) and the standard deviation (SD). Significantly *induced (fold change >2 and P < 0.05) or ^#decreased (fold change <−2 and P < 0.05) gene expression compared with day 0. Corresponding histological staining of alcian blue and safranin O for evaluation of extracellular matrix formation in PGA-HA scaffolds seeded with cells from early and advanced degenerated AF tissue cultured without inducer, 5% PRP, or 10 ng/mL TGFB3 at day 21.

In adAFC cultures the gene expression of collagen type II is significantly (P < 0.05; fold change >2) increased when treated with TGFB3 at day 7, 14 and 21, whereas collagen type I is only significantly (P < 0.05; fold change >2) increased at day 7. However, collagen type I gene expression is significantly (P < 0.05; fold change <−2) decreased at days 14 and 21 when treated with PRP, while collagen type II is only significantly (P < 0.05; fold change >2) increased at day 7. No significant changes in the gene expression of collagen type III could be detected in adAFC stimulated with PRP or TGFB3 at any time point. Furthermore, 3D culture of adAFC (noninduced) resulted in a significant (P < 0.05; fold change >2) increased gene expression of collagen types I and III at days 7, 14, and 21, whereas no significant changes were detected for collagen type II over time. Analyses of collagen types I and II protein levels in adAFC cultures showed a distinct staining and therefore deposition of collagen types I and II if treated with TGFB3, while only a slight staining was observed for collagen type I when cultured without supplements (noninduced) or PRP ( Fig. 4 ).

Gene expression of aggrecan in edAFC cultures was significantly (P < 0.05; fold change <−2) decreased by PRP at days 14 and 21, while TGFB3 significantly (P < 0.05; fold change >2) increased its expression at all time points. The gene expression of biglycan was significantly (P < 0.05; fold change >2) increased when stimulated with PRP only at day 7. Furthermore, gene expression of biglycan was induced by TGFB3 at days 7, 14, and 21 (P < 0.05; fold change >2), while on the contrary being decreased for decorin (P < 0.05; fold change <−2). The noninduced cultures showed a significant (P < 0.05; fold change <−2) decreased gene expression of biglycan at day 21 and aggrecan at all time points, while decorin was significantly (P < 0.05; fold change >2) increased. Histological staining detected a strong deposition of acidic proteoglycans (alcian blue staining) in edAFC cultures in the presence of TGBF3 with nearly no staining when treated with PRP or left noninduced. Similar results were demonstrated for sulfated proteoglycans (safranin-O staining), where only treatment with TGFB3 resulted in a distinct staining.

Treatment of adAFC cultures with TGFB3 significantly (P < 0.05; fold change >2) elevated the gene expression of aggrecan at day 7 and 14, whereas biglycan was only increased at day 7. AdAFC with PRP and noninduced treatment resulted in a significant (P < 0.05; fold change <−2) decreased expression of aggrecan at all investigated time points, whereas biglycan was significantly (P < 0.05; fold change <−2) reduced for adAFC stimulated with PRP at day 14 and 21 and for noninduced only at day 21. Additionally, the gene expression of decorin was significantly (P < 0.05; fold change >2) upregulated at all investigated time points, when adAFC were cultured three-dimensionally without any induction. Histological staining showed similar results as seen for the edAFC, with a strong deposition of acidic and sulfated proteoglycans in the presence of TGFB3 and with no staining in the other groups ( Fig. 5 ).

Discussion

With the low self-healing capacity of the AF and a recurrent herniation rate of up to 23%⁸ there is a major need in regeneration of the defective/ruptured AF tissue. Sutures alone lacked the mechanical stability needed to resist the loaded forces onto the AF,⁹ which favored the development of new strategies for annulus closure. One option for a regenerative approach is the use of ex vivo expanded autologous cells with or without combination of biomaterials, which showed promising first clinical results.²⁴ The major drawback of this cell-based therapy is that it requires 2 surgeries: first for biopsy harvest used for cell isolation and expansion followed by a second for the implantation of the seeded biomaterial.²⁵ This can be overcome by a 1-step cell-free approach, which uses the potential of residual cells from the adjacent tissue, which are able to migrate into the defect, proliferate, and synthesize tissue-specific repair tissue. The important point for expanding the application of cell-free implants for AF closure is the regenerative potential of endogenous AF cells from early or advanced degenerated tissue as a suitable cell source for AF repair.

This preliminary study showed that early and advanced degenerated AF cells both have the potential to proliferate, migrate and develop ECM. The observed proliferation potential for both cell groups is in line with a study demonstrating the lifelong proliferation capacity of AF cells even in degenerated tissue.²⁶ Moreover, we determined an enhanced metabolic activity of early and advanced degenerated AF cells in the presence of PRP and TGFB3, which might indicate their positive influence on the proliferation of AF cells. This is strengthened by the fact that both PRP and TGFB were able to induce the proliferation of AF in a 3D culture.^27,28 In contrast, in our study, only PRP and HS induced the migration of edAFC and adAFC, while TGFB3 showed no effect. This is in line with a study showing that migration of AF cells into alginate scaffolds was not improved by the addition of TGFB3.²⁹ Moreover, migratory effects of PRP and HS were comparable to other studies that demonstrated their effect on several mesenchymal-derived cells.^30,31

The potential to synthesize ECM was present in both early and advanced degenerated AF cells. Three-dimensional cultivation (noninduced) of both cell groups showed an increased gene expression of decorin as well as a collagen type I, which was even detected at the protein level at day 21. Furthermore, synthesis of ECM could be stimulated in varying degrees by the addition of TGFB3 or PRP in both cell groups. An increased gene and protein expression of collagen type II as well as the deposition of acidic and sulfated proteoglycans in histological stainings was shown for both cell groups when treated with the chondrogenic inductor TGFB3. Collagen type I could be detected at the protein level in both cell groups in the presence of PRP during the 3D cultivation, even though its gene expression was decreased at days 14 and 21. In contrast, a study with porcine AF cells in alginate beads showed an increase in proteoglycan and collagen synthesis when cultured in 10% PRP.²⁷ This supposed contradiction may lie on the fact that PRP is a relatively undefined cocktail of various growth and differentiation factors.³² It is known that different PRP preparation methods lead to different concentrations of platelets, leukocytes, and growth factors. Also, other bioactive compounds may cause variability and unpredictability of PRP efficacy and effects.^19,20,33 This may explain the inconsistency of results examining the effect on proteoglycan and collagen synthesis of AF cells treated with PRP. We hypothesize, that outcomes after PRP stimulation are not necessarily predictable and should be individually adapted to the respective application. In our study, we used only a single dose of 5% PRP to analyze the effect on matrix formation, and we cannot draw any conclusion concerning the effect of higher or lower doses, which is a limitation of the study. Redifferentiation potential of human AF cells cultured in 3D scaffolds with 5% human serum was shown by an increased gene expression of collagen types I and III comparable to the native level after the expression decreased during the in vitro expansion.³⁴ Moreover, pellet cultures of AF cells stimulated with 10% HS resulted in an enhanced deposition of collagen type I and proteoglycans as shown by immunohistochemical and histological stainings.¹⁸ Our results, together with those from the literature, verify the potential of both cell groups to synthesize chondrogenic ECM.

Another limitation of the study was the necessity to pool the available cells to generate a sufficient cell number to conduct the experiments, since this masks the individual patient variability. Further studies need to be performed to analyze if the observed potential varies widely between the individuals.

Our study demonstrated that AF cells retain their potential for proliferation, migration, and ECM formation independent of the degeneration status of the tissue. Furthermore, proliferation, migration and ECM synthesis of the endogenous AF cells, necessary for a successful AF defect repair, can be supported by different supplements. Hence, endogenous cells in the AF might be a suitable cell source for a cell-free AF repair approach independent of the degeneration status of the intervertebral disc.