Effect of the CCL5-Releasing Fibrin Gel for Intervertebral Disc Regeneration

Abstract

Objective

To explore if chemokine (C-C motif) ligand 5 (CCL5) delivery could recruit annulus fibrosus (AF) cells to the injury sites and facilitate the repair of ruptured AF.

Design

The effects of CCL5 on bovine AF cells in vitro were tested by transwell assay and quantitative real-time polymerase chain reaction. Fibrin gel containing CCL5 was used to treat annulotomized bovine caudal discs cultured under dynamic loading conditions. After 14 days of loading, the samples were collected for histological examination. A pilot animal study was performed using sheep cervical discs to investigate the effect of fibrin gel encapsulated with CCL5 for the treatment of ruptured AF. After 14 weeks, the animals were sacrificed, and the discs were scanned with magnetic resonance imaging before histopathological examination.

Results

CCL5 showed a chemotactic effect on AF cells in a dose-dependent manner. AF cells cultured with CCL5 in vitro did not show any change of the gene expression of CCL5 receptors, catabolic and proinflammatory markers. In vitro release study showed that CCL5 exhibited sustained release from the fibrin gel into the culture media; however, in the organ culture study CCL5 did not stimulate homing of AF cells toward the defect sites. The pilot animal study did not show any repair effect of CCL5.

Conclusions

CCL5 has a chemotactic effect on AF cells in vitro, but no ex vivo or in vivo regenerative effect when delivered within fibrin gel. Further study with a stronger chemotactic agent and/or an alternate biomaterial that is more conductive of cell migration is warranted.

Introduction

Degenerative disc disease is a major cause of low back pain in modern society.¹ Diverse etiological factors are known to serve as primary events that lead to abnormal production of cytokines and catabolic molecules from the disc cell. These factors include genetic predisposition, smoking, infection, abnormal biomechanical loading, decreased nutrient transport across the endplate and aging,^2-8 which endow this disease with an elevation in levels of inflammatory cytokines and chemokines, such as tumor necrosis factor (TNF)-α, interleukin (IL)-1 α/β, IL-6, IL-17, chemokine (C-C motif) ligand 5 (CCL5), and chemokine (C-X-C motif) ligand 6 (CXCL6). These cytokines and chemokines promote matrix degradation and changes in cell phenotype. The resulting imbalance between catabolic and anabolic responses leads to degeneration, as well as herniation and radicular pain. However, some of these cytokines and chemokines have also been shown to be involved in the disc repair process by activation of endogenous disc cells and chemokine mediated cell homing.⁹

CCL5 is an 8-kDa protein classified as a chemotactic cytokine or chemokine,¹⁰ which is also known as RANTES (regulated on activation, normal T cell expressed and secreted). CCL5 is chemotactic for T cells, eosinophils, and basophils, and plays an active role in recruiting leukocytes into inflammatory sites. Lee et al.¹¹ suggested that CCL5 may play a role in the migration of human periodontal-ligament stem cells into inflammatory periodontal lesions. Stalman et al.¹² also demonstrated that CCL5 together with CCL2, CCL3, and CXCL10 may exhibit vital targets for biological modulation of tendon repair. Hirotaka et al.¹³ first investigated the expression of CCL5 in herniated nucleus pulposus in 1996 and did not identify any expression of CCL5 using immunohistochemical analysis. In 2002, Sang et al.¹⁴ found that only 4 (17%) of 23 herniated disc specimens showed gene expression of CCL5. However, a recent study¹⁵ verified that the CCL5 gene was highly expressed in painful intervertebral disc specimens. Gruber et al.¹⁶ also reported that CCL5 expression was significantly upregulated in more degenerated discs (Thompson grades IV and V) versus healthier discs (Thompson grades I, II, and III).

Using an in vitro organ culture model, we have shown that CCL5 was released into conditioned medium of induced degenerative discs in organ culture, and the released chemokine promoted the migration of mesenchymal stem cells into the degenerative discs.¹⁷ We also found that elevated systemic blood plasma levels of CCL5 were associated with moderate/severe lumbar disc degeneration in a clinical case-control study.¹⁸ Furthermore, Liu et al.¹⁹ demonstrated that CCL5 could induce the migration of annulus fibrosus (AF) cells in vitro with a high expression of C-C chemokine receptor 5. Therefore, in this study, we investigated whether CCL5 delivery could attract AF cells in vitro and ex vivo, and whether it facilitates the repair of ruptured AF ex vivo and in vivo.

Methods

Study Design

First, an in vitro transwell study was performed to test the homing effect of CCL5 on AF cells. Then an intervertebral disc (IVD) organ culture model was used to investigate if CCL5 delivery could attract AF cells to migrate toward the injury site, and whether it facilitates the repair of ruptured AF. Finally, a pilot sheep animal study was performed to evaluate the feasibility and efficacy of CCL5 treatment for AF rupture repair in an in vivo situation.

Effect of CCL5 on Annulus Fibrosus Cells In Vitro

IVDs were excised from 6- to 11-month-old bovine tails obtained from a local abattoir. AF tissue was dissected and carefully separated from the nucleus pulposus (NP) tissue without using the transition zone. After enzymatic digestion using 0.2% Pronase (Roche Diagnostics GmbH, Rotkreuz, Switzerland) at 37°C for 90 minutes and 400 U/mL collagenase type II (Worthington, Lakewood, NJ, USA) for 13 to 17 hours at 37°C, all cells were collected after filtering with sequential 100 μm and 40 μm filters.

For the transwell migration assay, primary AF cells were seeded onto the top chamber of the 96-well transwell plates (membrane pore size 8 µm, Corning, Lowell, MA, USA) at a concentration of 50,000 cells/well and cultured in α-minimal essential medium (α-MEM) (Gibco, Grand Island, NY, USA) supplemented with 100 U/mL penicillin (Gibco, Grand Island, NY, USA) and 100 µg/mL streptomycin (Gibco, Grand Island, NY, USA) (1% Pen/Strep). The bottom chamber contained α-MEM with different concentrations of human CCL5 (0, 25, 75, and 100 ng/mL, Miltenyi Biotec, Auburn, CA, USA). For chemotaxis tests, the top chambers contained 0 ng/mL of CCL5. For chemokinesis tests, the top chambers contained equal concentration of CCL5 as the bottom chambers (0, 25, 75, 100 ng/mL). Human stromal cell–derived factor 1 (SDF1, Miltenyi Biotec, Gladbach, Germany) was used at the same concentrations as a control, since it has shown cell homing effect in cartilage tissue.²⁰ Chemotaxis effect means a dose dependent cell migration behavior toward the chemokine, which is tested by no chemokine in the top chamber, and increased dose of chemokine in the bottom chamber. Chemokinesis effect means a random cell migration behavior when cells are exposed to the chemokine, which is tested by adding increased dose of chemokine in both the top and bottom chambers at the same concentration. After incubation at 37°C for 17 hours, migrated primary AF cells were dissociated from the bottom side of the membrane using trypsin/ethylene diamine tetraacetic acid (Gibco, Grand Island, NY, USA), stained with 2 µg/mL calcein (Sigma, Buchs, Switzerland) and assessed with a Victor plate reader (ex485 em535 PerkinElmer).

To investigate the effect of CCL5 on the AF cells’ gene expression, primary AF cells were seeded into 6-well plates at a concentration of 400,000 cells/well and cultured in α-MEM supplemented with 1% Pen/Strep, and 100 ng/mL CCL5. Cells cultured without chemokine served as negative control. After incubation for 17 hours, the cells were collected for gene expression measurement via quantitative real-time polymerase chain reaction (qRT-PCR).

Quantitative Real-Time Polymerase Chain Reaction

RNA isolation was carried out with TRI Reagent (Molecular Research Centre Inc, Cincinnati, OH, USA) according to the manufacturer’s protocol. Reverse transcription was performed using SuperScript VILO cDNA Synthesis Kit (Invitrogen) and 400 ng of total RNA according to the manufacturer’s protocol. QuantStudio 6 System (Applied Biosystems) was used to conduct qRT-PCR. Gene expression of bovine collagen type I (COL1), a disintegrin-like and metallopeptidase with thrombospondin type 1 motif, 4 (ADAMTS4), TNFα, matrix metalloproteinase 1 (MMP1), and MMP13 in disc cells was analyzed using custom-designed primers and TaqMan probes (Microsynth, Balgach, Switzerland, Table 1 ). For amplification of RPLP0 (Bt03218086_m1), chemokine (C-C motif) receptor 3 (CCR3, Bt04317048_g1), chemokine (C-C motif) receptor 5 (CCR5, Bt03237589_m1), and TIMP metallopeptidase inhibitor 1 (TIMP1, Bt03223720_m1), gene expression assays from Applied Biosystems (Life Technologies) were used. Comparative Ct method was performed for relative quantification of target mRNA with RPLP0 as endogenous control.

Table 1.

Oligonucleotide Primers and Probes (Bovine) Used for Quantitative Real-Time Polymerase Chain Reaction.

Gene	Primer/Probe Type	Sequence
COL1	Primer forward (5′-3′)	TGC AGT AAC TTC GTG CCT AGC A
	Primer reverse (5′-3′)	CGC GTG GTC CTC TAT CTC CA
	Probe (5′FAM/3′TAMRA)	CAT GCC AAT CCT TAC AAG AGG CAA CTG C
MMP1	Primer forward (5′-3′)	TTC AGC TTT CTC AGG ACG ACA TT
	Primer reverse (5′-3′)	CGA CTG GCT GAG TGG GAT TT
	Probe (5′FAM/3′TAMRA)	TCC AGG CCA TCT ACG GAC CTT CCC
MMP13	Primer forward (5′-3′)	CCA TCT ACA CCT ACA CTG GCA AAA
	Primer reverse (5′-3′)	GTC TGG CGT TTT GGG ATG TT
	Probe (5′FAM/3′TAMRA)	TCT CTC TAT GGT CCA GGA GAT GAA GAC CCC
ADAMTS4	Primer forward (5′-3′)	CCC CAT GTG CAA CGT CAA G
	Primer reverse (5′-3′)	AGT CTC CAC AAA TCT GCT CAG TGA
	Probe (5′FAM/3′TAMRA)	AGC CCC CGA AGG GCT AAG CGC
TNFα	Primer forward (5′-3′)	CCT CTT CTC AAG CCT CAA GTA ACA A
	Primer reverse (5′-3′)	GAG CTG CCC CGG AGA GTT
	Probe (5′FAM/3′TAMRA)	ATG TCG GCT ACA ACG TGG GCT ACC G

CCL5 In Vitro Release from Fibrin Gel

Since CCL5 delivery in IVD organ culture model or animal model had not been tested before, a wide range of concentrations were selected for the release study and organ culture study to reveal the dose dependent effect. Fibrin gel (30 μL per gel, fibrinogen 17 mg/mL + thrombin 0.5 U/mL, Baxter Biosurgery, Vienna, Austria) containing different doses of human CCL5 (0, 50, 500, or 2500 ng, Miltenyi Biotec, Auburn, CA, USA) was incubated in 1.5 mL α-MEM at 37°C, 85% humidity, and 5% CO₂. At the time points of 0 hour, 2 hours, 4 hours, 6 hours, 1 day, 2 days, 5 days, 7 days, and 14 days, 750 μL medium was collected and replaced with the same volume of fresh medium. The released CCL5 was analyzed by enzyme-linked immunosorbent assay (R&D Systems, Minneapolis, MN, USA) according to the manufacturer’s protocol.

AF Defect Model of Organ-Cultured IVDs and Repair with Fibrin Gel Containing CCL5

An organ culture study was performed with the aim to test the effect of CCL5 at different doses (0, 50, 500, or 2500 ng), within a microenvironment that is closer to the in vivo condition. IVDs were excised from 6- to 11-month-old bovine tails obtained from a local abattoir. The endplates of each IVD were cleaned with a Pulsavac jet-lavage system (Zimmer, Warsaw, IN, USA). Initial disc height and diameter were recorded before IVDs were washed with phosphate buffered saline (pH 7.4) containing 10% Pen/Strep for 15 min. Then IVDs were transferred to a 6-well plate containing IVD culture medium and incubated overnight at 37°C, 85% humidity, and 5% CO₂. The IVD culture medium was composed of Dulbecco’s modified Eagle medium (DMEM; Gibco, Grand Island, NY, USA) with 4.5 g/L glucose, 2% fetal calf serum, 1% Pen/Strep, 1% ITS + Premix (Discovery Labware, Inc, Bedford, MA, USA), 50 mg/mL ascorbate-2-phosphate (Sigma Aldrich, St. Louis, MO, USA) and 0.1% Primocin. IVDs from each tail were randomly assigned among experimental groups to obtain similar average disc size and distribution of disc levels for each group (4 discs per group).

After a 0.5 × 0.5 cm operation window was cut on the surface layer of AF tissue, a thin (~1 mm) layer of AF tissue was preserved for closure of the defect. A defect was created under the operation window through the AF tissue and partial NP tissue using a biopsy punch (diameter 2 mm, depth 7 mm; Kai Medical Europe GmbH, Solingen, Germany). The defect was then refilled with 30 μL of fibrin gel (same composition as in the release study) containing different doses of human CCL5 (0, 50, 500, or 2500 ng). The IVD was incubated at 37°C for 60 minutes to let the fibrin gellify before the operation window was sutured (4 point-suture) with the adjacent AF tissue using a 5-0 Prolene (Ethicon, Somerville, NJ, USA) ( Fig. 1 ). All discs were cultured at 37°C, 85% humidity and 5% CO₂ with one hour of physiological load every day (0.02-0.4 MPa, 1 Hz) applied with a bioreactor system.^21-23

Figure 1.

Process of building an annulus fibrosus (AF) defect organ culture model. (A and B) A 0.5 × 0.5 cm operation window was cut; (C) a defect was created under the operation window through the AF tissue and partial nucleus pulposus (NP) tissue using a biopsy punch; (D) the defect was refilled with 30 μL of fibrin gel; (E and F) the operation window was sutured after gelation of the fibrin gel.

Histological Staining

After 14 days of culture, IVDs were snap-frozen, embedded in cryocompound, and cryosectioned. Transverse sections of IVDs were made at a thickness of 10 μm. Sections were stained with 0.1% Safranin-O and 0.02% Fast Green to reveal proteoglycan and collagen deposition, respectively, and counterstained with Weigert’s hematoxylin to reveal cell distribution. A semiquantitative scoring scheme adapted from Shu et al²⁴ was used to evaluate the Safranin-O/Fast Green staining sections ( Table 2 ). The score numbers for Safranin-O/Fast Green proteoglycan staining, IVD structure/lesion characteristics, and formation of clefts in vicinity to lesion were summed up to assess the degree of IVD degeneration.

Table 2.

Histopathological Scoring of Normal and Pathological IVDs.

Grade	Histopathological Features
	A. Safranin-O/Fast Green proteoglycan staining
0	Fast green staining only of OAF, intermediate Safranin-O staining of IAF, intense Safranin-O staining in NP. Alternate AF lamellae discernable due to differing Fast Green staining intensities of adjacent lamellae
1	Slightly reduced Safranin-O staining of MAF/IAF in vicinity of lesion, Fast Green staining of OAF only, normal Safranin-O staining of NP
2	Moderately reduced Safranin-O staining of MAF/IAF in vicinity of lesion, Fast Green staining of OAF only, normal Safranin-O staining of NP
3	Reduced patchy Safranin O-staining around lesion, Fast Green staining in OAF (no Safranin-O staining)
4	Reduced Safranin-O staining in NP compared with sham IVD, very faint or no Safranin O staining in OAF/MAF, fast green staining only in OAF
	B. IVD structure/lesion characteristics
0	Normal IVD structure with well-defined annular lamellae, central NP
1	Lesion evident in MAF, normal NP morphology
2	Lesion evident in MAF/IAF, lesion but may not be apparent in OAF due to spontaneous repair, IAF lamellae may be inverted and have anomalous distortions in normal lamellar architecture
3	Bifurcation/propagation of lesion from MAF/IAF into NP margins, mild delamination, when more extensive may lead to concentric tears between lamellae in MAF/IAF
4	Propagation of lesion into NP, with disruption in normal NP structure, distortion of annular lamellae into atypical arrangements-severe delamination, separation of translamellar cross bridges
	C. Formation of clefts in vicinity to lesion (mainly radially oriented)
0	No clefts in AF
1	Small cleft area in AF
2	Moderate cleft area in AF
3	Large cleft area in AF
4	Vast cleft area in AF and also in NP

AF = annulus fibrosus; IAF = inner AF; IVD = intervertebral disc; MAF = mid AF; NP = nucleus pulposus; OAF = outer AF.

Cell viability was assessed using the lactate dehydrogenase (LDH) method, by staining with ethidium homodimer and LDH.²⁵ Five random images (300 × 300 μm²) were taken from different areas that were about 20 µm from the defect location in each section, 3 sections from each disc and four discs per group were analyzed. The number of alive and dead cells was counted using ImageJ. Cells stained blue (LDH) or blue/red (ethidium homodimer) were assigned to alive cells. Cells stained red only were labeled as dead cells. In each selected region (300 × 300 μm²), about 400 AF cells and 50 NP cells in total were counted.

Animal Study

A pilot animal study with 4 sheep was performed to evaluate the feasibility and efficacy of CCL5 treatment for AF rupture repair in an in vivo situation. Four female Swiss White Alpine sheep, with an age of 2 to 4 years, and a mean body weight of 67.5 ± 7.55 kg (range, 62-78 kg), were included in this study. All procedures were performed in an Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC) International approved facility and according to Swiss animal protection law and regulations under the approval of the local governmental animal use committee (Graubünden TVB 21E/2014). Sheep were assessed to be healthy and free of orthopedic disease, based on physical examination, including blood examination prior to enrolment in the study. Animals were acclimatized for at least 2 weeks to the housing conditions (12 hours light/dark). During this time, they were group housed and fed with hay, a mineral lick, and hand-fed grain to gain familiarity with the animal care givers.

Animals were sedated (0.04 mg/kg subcutaneously of detomidine; Domosedan, Provet AG, Lyssach, Switzerland) and induced for intubation (0.2 mg/kg intravenously of diazepam; valium, Roche Pharma AG, Reinach, Switzerland) and 6 mg/kg intravenously of ketamine (Ketasol-100, Roche Pharma AG, Reinach, Switzerland). Anesthesia was maintained using isoflurane (isofluran Baxter, Baxter AG, Opfikon) in oxygen and air.

For systemic analgesia animals received fentanyl intraoperatively (5 µg/kg/h intravenous). At the end of surgery, a nonsteroidal anti-inflammatory drug (4 mg/kg intravenously of carprofen; Rimadyl Rind, Zoetis Schweiz GmbH) and opioids (0.05 mg/kg of buprenorphine intramuscularly; Temgesic, Reckitt Benckiser AG) were injected.

For the surgery, the sheep were in dorsal recumbency and a standard anterior approach from the right side was performed. A longitudinal incision and blunt preparation between the sternocleidoideus and longus colli muscles/carotid sheath (lateral) and the esophagus/trachea (medial) exposed the anterior aspect of (C2-C6). A 1 mm Kirschner wire was placed in the median plane in disc C2/3, C3/4 and C4/5. After confirmation of the right location with the image intensifier ( Fig. 2A ), the AF defects were created, again in C2/3, C3/4, and C4/5 with a biopsy punch (diameter 2 mm, depth 7 mm; Kai Medical Europe GmbH, Solingen, Germany), by pushing the whole length of the punch into the disc. The defects were randomly filled with 30 µL fibrin (treated control group), or 30 µL fibrin with 500 ng CCL5 ( Fig. 2B ) and sealed with fibrin-genipin glue^26-28 (treated study group), or left empty (untreated control group). Each group contained 3 IVD samples from 3 different animals. The source and composition of fibrin gel and CCL5 were equal to the organ culture study. The concentration of the CCL5 was selected based on the results of the release study and considerations of side effect.

Figure 2.

Animal surgery. (A) Radiographic control for correct location and direction of the 2-mm defect. (B) Pipetting of the fibrin gel with or without CCL5 into the defect.

The postoperative analgesia protocol consisted of carprofen (4 mg/kg subcutaneous every 48 hours for a minimum of 3 days), buprenorphine (0.05 mg/kg intramuscularly every 8 hours for 24 hours), and fentanyl (2 µg/kg/h of fentanyl; Durogesic Matrix patches, Janssen-Cilag AG, Zug, Switzerland) for 72 hours. After operation, the sheep were housed in groups of 2 with free access to water and food. Ceftiofur was given perioperatively and for 5 days postoperatively every 24 hours. The animal welfare was assessed post operatively twice per day for the first 3 days, then daily until 7 days postoperatively, afterward weekly using a study specific score sheet. Weights were recorded 1 and 2 weeks after surgery, then monthly. Computed tomography scans were performed pre- and postoperatively, monthly, and at the end of the study. After 14 weeks, all the animals were sacrificed, and the cervical spine underwent magnetic resonance imaging (MRI) scan before IVD motion segments were fixed in 70% methanol for histological analysis. IVDs with endplates were embedded within paraffin after decalcification. Safranin-O/Fast Green staining was performed on 6 µm thick sagittal sections as described above. The disc degenerative degree was evaluated both by histological analysis²⁴ ( Table 2 ) and Pfirrmann classification based on MRI images.²⁹ Examiners of histological and MRI images were blinded during their assessment.

Statistical Analysis

The statistical analysis was carried out by SPSS version 22 software (IBM SPSS Statistics for Windows, Armonk, NY, IBM Corp. Released 2013). Chemotaxis and chemokinesis data were analyzed with Bonferroni-corrected Mann-Whitney U test. Gene expression differences between groups were analyzed using Wilcoxon matched-pairs signed rank test. The number of ethidium homodimer and LDH-stained cells at different areas, and the histopathological scoring were compared using the Tukey-Kramer multiple-comparisons posttest analysis of variance (ANOVA). Data were considered statistically significant when P < 0.05.

Results

Effect of CCL5 on AF Cells In Vitro

In the chemotaxis test of CCL5 on AF cells, the number of migrated cells in 100 ng/mL group was higher than that of 0 ng/mL (P < 0.05) and 25 ng/mL groups (P < 0.01) ( Fig. 3B , left bars). Whereas in the chemokinesis test, the number of the migrated cells did not increase under higher concentrations of CCL5 ( Fig. 3B , right bars). These results indicate a dose-dependent recruitment effect of CCL5 on AF cells, with a chemotactic effect under the highest dose 100 ng/mL. SDF1 showed a chemokinetic effect on AF cells, indicated by increased random cell migration with 75 ng/mL SDF1 in both top and bottom chambers ( Fig. 3A , right bars).

Figure 3.

Chemotaxis and chemokinesis effects of SDF1 (A) and CCL5 (B) on annulus fibrosus cells. Top—top chamber, bottom—bottom chamber. Chemotaxis effect means a dose-dependent cell migration behavior toward the chemokine, which is tested by no chemokine in the top chamber (Top = 0 ng/mL), and increased dose of chemokine in the bottom chamber (left bars). Chemokinesis effect means a random cell migration behavior when cells are exposed to the chemokine, which is tested by adding increased dose of chemokine in both the top and bottom chambers at the same concentration (Top = Bottom, right bars). Migrated cell number normalized to 0 ng/mL: Means ± standard error of the mean (SEM), n = 3, *P < 0.05, **P < 0.01.

The gene expression of catabolic related genes (ADAMTS4, TIMP1, MMP1, MMP13), TNF-α, COL1, and CCL5 receptor (CCR3, CCR5) were measured in AF cells cultured with or without 100 ng/mL CCL5 after 17 hours and normalized to the expression level on day 0 before cells were seeded into well plates. None of the genes showed any significant change after treatment with CCL5 ( Fig. 4 ).

Figure 4.

Gene expression of catabolic genes (ADAMTS4, TIMP1, MMP1, MMP13), TNF-α, COL1, and CCL5 receptor (CCR3, CCR5) were measured in AF cells cultured with or without 100 ng/mL CCL5 and normalized to the expression level on day 0 before cells were seeded into well plates. Means ± standard error of the mean (SEM), n = 3.

CCL5 In Vitro Release from Fibrin Gel

In vitro release study showed that the CCL5 was sustainably released from the fibrin gel into the culture media ( Fig. 5 ). Fibrin gel containing 50 ng reached a release plateau after 2 days. Fibrin gel containing 500 ng CCL5 showed higher release compared to the 50 ng groups after 4 hours of incubation (P < 0.05) and reached 100 ng at the endpoint. Fibrin gel containing 2500 ng CCL5 showed higher release compared to the other 2 groups (P < 0.05), and had not reached the release plateau after 14 days of culture. According to these results, all the 3 concentrations were tested in the following organ culture study to investigate the dose dependent effect of CCL5 for AF rupture repair in situ.

Figure 5.

CCL5 release study. Fibrin gel containing 50 ng CCL5 reached a release plateau after 2 days. Fibrin gel containing 500 ng CCL5 showed higher release compared to the 50 ng groups after 4 hours of incubation and reached 100 ng at the endpoint. Fibrin gel containing 2500 ng CCL5 showed higher release compared with the other 2 groups after 4 hours of incubation and did not reach the release plateau after 14 days of culture. Means ± standard error of the mean (SEM,) n = 3, *P < 0.05, **P < 0.01, ***P < 0.001 significant difference between 2500 ng and 500 ng, 2500 ng and 50 ng; ^#P < 0.05, ^##P < 0.01,^###P < 0.001 significant difference between 500 ng and 50 ng.

AF Defect Model of Organ-Cultured IVDs and Repair with Fibrin Gel Containing CCL5

In the organ culture study, all the discs were harvested after the 14-day culture with 1-hour physiological loading every day. The images of Safranin-O/Fast Green stained transverse disc sections are shown in Figure 6 . The implanted fibrin gel could not be identified within the AF defect. In some of the discs, NP tissue protruded into the AF defect ( Fig. 6A and C ). According to the histopathological scoring, there was no significant difference among all the groups ( Fig. 6A and B ). From the LDH-stained disc sections, a clear band with markedly lower number of cells was observed at the edge of the AF defect in all groups ( Fig. 6C ). The alive and dead cell numbers in the native disc tissue region close to the AF defect were counted ( Fig. 6D and E ). There was no significant difference on the number of alive cells and dead cells both in AF and NP among all the groups, which indicates no induction of cell migration within the native tissue toward the fibrin with CCL5. These results were not in line with the chemotaxis effect of CCL5 on AF cells in vitro. Considering the lack of whole body effect in the organ culture model, a pilot sheep study was performed to reveal the potential feasibility and efficacy of CCL5 treatment for AF rupture repair in an in vivo situation.

Figure 6.

Histopathological results ex vivo. (A) representative images of the Safranin-O/Fast Green–stained transverse disc sections, scale bar—upper images 1000 µm, lower images 500 µm. (B) Histological scoring values based on Safranin O/Fast Green staining. (C) Representative images of the lactate dehydrogenase (LDH)/ethidium homodimer stained transverse disc sections, scale bar—upper images 1000 µm, lower image 200 µm. A clear band without any cells was observed at the edge of the annulus fibrosus (AF) defect (inside the dashed lines). (D and E) The alive and dead cell number in the native disc tissue region close to the AF defect was counted, blue arrows indicate alive cells (stained blue, or blue/red), red arrows indicate dead cells (stained red only). Means ± SD, n = 4 (For interpretation of the references to colours in this figure legend, refer to the online version of this article).

Animal Study

No postoperative morbidity or mortality was recorded. All animals recovered rapidly after surgery and quickly resumed normal activities. None of the 4 animals showed any remarkable change in weight. One sheep developed a mild seroma, which resolved within a week. No spinal deformity was found in the sheep by the radiological assessment. Computed tomography scans at 14 weeks postoperatively showed mild to moderate new bone formation (osteophytes) in several locations in 3 out of 4 sheep at the cartilage endplates. Macroscopic examination revealed minimal fibrotic changes and no obvious prolapse. In the Safranin-O/Fast Green–stained sections some gel was still recognized. However, no difference among the groups was identified except for some gel left in the treated groups ( Fig. 7A ). The histopathological scoring also confirmed this result ( Fig. 7B ). T2-MRI images demonstrated that the discs in all the groups had a degenerative grade of 3 to 4 using Pfirrmann classification ( Fig. 7C and D ).

Figure 7.

(A) Representative images of Safranin-O/Fast Green–stained sagittal disc sections showed no difference among the untreated, fibrin, and fibrin + CCL5 groups, except for some gel left in the treated groups (black squares). Scale bar: 1000 µm. (B) The semiquantitative scoring values based on Safranin-O/Fast Green staining. (C and D) T2-weighted magnetic resonance images demonstrated that the discs in all groups had a degenerative grade of 3 to 4 (red arrows) using Pfirrmann classification. Means ± SD, n = 3 (For interpretation of the references to colours in this figure legend, refer to the online version of this article).

Discussion

Recently, CCL5 was reported to have a close relationship with disc degeneration as a chemotactic cytokine, which is highly expressed in degenerative organ–cultured bovine IVD, human degenerative IVD, and in systemic blood plasma of patients with moderate/severe lumbar disc degeneration.^17,18 Its ability to recruit mesenchymal stem cells into degenerative organ–cultured IVD makes it a highly promising chemokine as a new drug for enhancement of endogenous repair.¹⁷ However, whether CCL5 has a positive or negative impact on the native disc AF cells and whether it can stimulate regeneration of disc tissue after AF injury is not clear. In this study, we found that CCL5 demonstrated a chemotactic effect on primary AF cells in vitro. AF cells cultured with CCL5 in vitro did not show any change of gene expression of CCL5 receptors, which indicates that the chemotaxis movement of AF cells was not correlated with an induction of CCL5 receptor expression; post-transcriptional processes and activation of receptors present on the AF cells may have induced the observed cell migration. To investigate if CCL5 itself had any side/negative effect on the AF cells, the gene expression of catabolic and proinflammatory markers was measured. Results showed that CCL5 did not induce degenerative or inflammatory responses on the AF cells, which supported further test in organ culture and animal study. We further tested if CCL5 could stimulate the AF cells to migrate toward injury sites and if it facilitates regeneration of AF tissue after rupture, in an IVD organ culture model and a pilot sheep study.

Previous publications show that CCL5 had a strong chemotactic effect on several types of cells expressing CCR3/CCR5 receptor, such as T cells,³⁰ eosinophils,³¹ and mesenchymal stem cells.³² Recently, Liu et al.¹⁹ found that CCR5 was expressed at the mRNA level and on the cell surface of human NP and AF cells. They also suggested that the CCR5 receptors in AF cells were functional due to increased levels of ERK phosphorylation, and AF cells migration could be detected in the presence of CCL5. The results from the current study also showed that CCL5 exhibited a chemotaxis effect on bovine AF cells. However, all these findings were based on in vitro cell culture studies. In our organ culture model, there was no evidence that AF cells could migrate from the healthy site toward the defect area where the CCL5 was sustainably released, even at a very high concentration (2500 ng per disc). In the pilot sheep study, annulotomized IVDs treated with CCL5 also did not show obvious advantage in tissue regeneration compared with the IVDs without treatment or treated with fibrin gel alone.

Several potential reasons may be addressed here to explain the discrepancy that the chemotaxis effect of CCL5 on the AF cells could be detected in vitro but not in the organ culture or animal study. First, the results of the animal study may not be predicted by cell culture experiments,³³ especially for the dense tissue such as AF with a sophisticated natural structure.^34-37 In conventional 2-dimensional conditions, the extracellular matrix components, cell-to-cell and cell-to-matrix interaction that influence differentiation, proliferation and cellular functions in vivo are lost.^38,39 In the cell culture system, the AF cells had easy access to the CCL5 in the media, and the cell surface receptors such as CCR3/CCR5 were directly exposed to the CCL5, therefore, it was straightforward for AF cells to sense the concentration gradient of CCL5, which activates the chemotaxis effect. However, the microenvironment of cells in the AF tissue is markedly different, where the cells are located within a dense extracellular matrix network predominantly composed of collagen.^34-37 Organ and explant culture models have been verified to mimic the in vivo condition more closely.⁴⁰ In this study, the results from the IVD organ culture model were in line with those from the pilot sheep study, which further confirmed that organ culture models may be used as a prescreening tool before testing in animals. Second, the fibrin gel used in the current study for chemokine delivery may not be the optimal biomaterial for AF cells to migrate in. Fibrin glue is a topical biological adhesive, the effect of which imitates the final stages of coagulation. The gel consists of a solution of concentrated fibrinogen, which is activated by the addition of thrombin. This gel is noncytotoxic and has been verified as carrier for mesenchymal stem cells^41,42 and for migration studies of endothelial cells⁴³ and neutrophil cells⁴⁴ in vitro. However, the crosslinking of the fibrin gel only occurs within the gel itself. This leads to lack of integration of biomaterials with surrounding native tissue, which plays an essential role in migration of endogenous cells.⁴⁵ Further study using a tissue adhesive hydrogel which contains bonding effects with the components of native disc tissue may facilitate the migration of endogenous disc cells toward the chemokine containing biomaterials. In addition, the high density of fibrin gel may also cause an impediment for cell migration. Third, the high density of extracellular matrix within AF tissue may reduce endogenous cell migration. It has been shown that delivery of degradative enzyme in meniscus enhanced the cell density and integrative tissue formation within the wound.⁴⁶ This strategy may be applied for repair of other dense connective tissues, such as AF, in combination with chemokine delivery. However, the concentration and delivery system of enzyme needs to be optimized to avoid acute degradation of remaining healthy tissue.

Conclusions

The current study demonstrated the chemotactic homing effect of CCL5 on AF cells in cell culture study. The addition of CCL5 in IVD organ culture and the sheep AF defect model did not improve the healing response, and results suggested that CCL5 may not home AF cells to repair the defect area after disc puncture with the current delivery system. Further study should focus on alternative hydrogels that are more conducive to cell motility and may better complement the chemotactic benefits of CCL5 to better promote cell migration and healing of injured and degenerated IVDs. Using the organ culture and animal model established within the current study, various bioactive agents with chemotactic and/or regenerative effect may be screened and validated for repair of AF rupture.