Chemonucleolysis combined with dynamic loading for inducing degeneration in bovine caudal intervertebral discs

Abstract

Chemonucleolysis has become an established method of producing whole organ culture models of intervertebral disc (IVD) degeneration. However, the field needs more side-by-side comparisons of the degenerative effects of the major enzymes used in chemonucleolysis towards gaining a greater understanding of how these organ culture models mimic the wide spectrum of characteristics observed in human degeneration. In the current work we induced chemonucleolysis in bovine coccygeal IVDs with 100 µL of papain (65 U/mL), chondroitinase ABC (chABC, 5 U/mL), or collagenase II (col'ase, 0.5 U/mL). Each enzyme was applied in a concentration projected to produce moderate levels of degeneration. After 7 days of culture with daily dynamic physiological loading (0.02-0.2 MPa, 0.2 Hz, 2 h), the cellular, biochemical and histological properties of the IVDs were evaluated in comparison to a PBS-injected control. Papain and collagenase, but not chABC, produced macroscopic voids in the tissues. Compared to day 0 intact IVDs, papain induced the greatest magnitude glycosaminoglycan (GAG) loss compared to chABC and col'ase. Papain also induced the greatest height loss (3%), compared to 0.7%, 1.2% and 0.4% for chABC, col'ase, and PBS, respectively. Cell viability in the region adjacent to papain and PBS-injection remained at nearly 100% over the 7-day culture period, whereas it was reduced to 60%-70% by chABC and col'ase. Generally, enzyme treatment tended to downregulate gene expression for major ECM markers, type I collagen (COL1), type II collagen (COL2), and aggrecan (ACAN) in the tissue adjacent to injection. However, chABC treatment induced an increase in COL2 gene expression, which was significant compared to the papain treated group. In general, papain and col'ase treatment tended to recapitulate aspects of advanced IVD degeneration, whereas chABC treatment captured aspects of early-stage degeneration. Chemonucleolysis of whole bovine IVDs is a useful tool providing researchers with a robust spectrum of degenerative changes and can be utilized for examination of therapeutic interventions.

Keywords: chemonucleolysis (CN); degeneration; extracellular matrix; intervertebral disc; organ culture.

FIGURE 1

Workflow for developing the enzyme-induced model of IVD degeneration. Bovine coccygeal IVDs were isolated. An intact IVD from each tail was set aside as the day 0 control, and the remaining samples subjected to dynamic axial loading. The IVDs were subsequently injected with 100 µL of PBS (control), papain (65 U/mL), chABC (5 U/mL) or col’ase (0.5 U/mL) and cultured for 7 days with daily physiological loading applied. IVD height was measured daily prior to each loading cycle. At the end of the study period, the specimens were sliced along the sagittal plane and evaluated for gene expression, GAG content, histological staining, and cell viability analyses. A total of n = 4 replicates were used per enzyme/control group, isolated from two to four donors.

FIGURE 2

(A) Gross morphology of the IVD study groups. At day 7 post-enzyme treatment, macroscopic voids were observed in the papain and col’ase-injected specimens, indicated by the yellow boxes, but not in chABC-injected specimens. (B) Height loss during 7 days of IVD culture with daily loading in the bioreactor, calculated relative to the initial IVD height after dissection. Papain produced the highest magnitude height loss in the IVD specimens compared to the PBS control. Data are shown as the mean and standard deviation of n = 4 replicates, where ** indicates p = 0.0065.

FIGURE 3

Biochemical and cellular changes in the IVDs at day 7 after enzyme treatment. GAG content/wet tissue mass of (A) iAF and (B) oAF tissue samples at day 7 normalized to day 0 of the respective donor. Papain produced the largest drop in GAG content of the iAF and oAF compared to the PBS control, the drop being statistically significant for the iAF (p = 0.0143.) (C) Cell viability (%) at day 7 normalized to day 0 of the respective donor for the iAF and (D) oAF of the IVDs quantified from LDH staining images. No significant changes in cell viability were measured compared to the PBS control in the iAF or oAF, although cell viability in the iAF for the papain-treated samples trended highest (close to 100%) compared to the chABC and col’ase (60%–70%). Results are shown as the mean and standard deviation of four replicates, except for (A) col’ase, where one replicate could only be recovered due to digestion.

FIGURE 4

Safranin-O and fast green staining across whole sagittal cross-sectional areas of the IVD sample groups highlighting the morphology of GAG (red) and collagens (green) in the matrix for (A) Day 0 intact control, and at day 7 for (B) PBS control, (C) papain, (D) chABC, (E) col’ase, specimen 1, and (F) col’ase, specimen 2. While voids and significant GAG loss could be identified at day 7 in papain and col’ase treated specimens, chABC produced the mildest changes, with no void produced and some GAG staining remaining. Two representative specimens are shown for the col’ase group to demonstrate a region of concentrated GAG staining observed in the oAF of one of the samples (red arrow). Yellow boxes highlight macroscopic voids produced by enzyme digestion. Scale bars = 2 mm. (G,H,I) Histological scoring values based on criteria detailed in Table 1 for Safranin O/Fast green staining of whole sagittal cross-sections. Overall, papain and col’ase produced higher degeneration scores than chABC, PBS and day 0 intact controls. Grading was assessed by two blinded observers on three sections per group, where *p < 0.05, **p < 0.01, and ***p < 0.001.

FIGURE 5

High magnification images of Safranin-O and fast green staining highlighting the morphology of GAG (red) and collagens (green) in the nucleus pulposus (NP), inner annulus fibrosus (iAF) and outer annulus fibrosus (oAF) of the intact (day 0), PBS-treated (day 7), and enzyme-treated (day 7) groups. Overall, papain and col’ase digestion resulted in complete loss of GAG staining in the NP region. GAG staining could still be identified in the NP and iAF of chABC treated specimens. Two representative specimens are shown for the col’ase group to demonstrate a region of concentrated GAG staining observed in the oAF of one of the samples (yellow arrow). Scale bars = 200 µm.

FIGURE 6

Bovine mRNA expression at day 7 for the major IVD ECM components, COL1, COL2 and ACAN, relative to day 0 and RPLP0 for the PBS-injected and enzyme-treated sample groups. In the iAF, papain and col’ase resulted in mild downregulation in the major disc NP matrix markers, ACAN and COL2. ChABC treatment resulted in a statistically significant upregulation of a major ECM marker, COL2, within the iAF compared to papain (*p = 0.0286), suggesting chABC could be inducing the mildest degenerative response. Results are shown as the mean and standard deviation of four replicates.

FIGURE 7

Bovine mRNA expression at day 7 for catabolic enzymes, ADAMTS5 and MMP3, relative to day 0 and RPLP0 for the PBS-injected and enzyme-treated sample groups. The highest magnitude shift was observed for MMP3 in the oAF, although no statistically significant shifts were detected (p > 0.05). Results are shown as the mean and standard deviation of four replicates.

FIGURE 8

Bovine mRNA expression at day 7 for oAF markers, TAGLN, ELN, MCAM, and MKX, relative to day 0 and RPLP0 for the PBS-injected and enzyme-treated sample groups. Notably, in the iAF, col’ase treatment tended to downregulate expression of all oAF markers compared to the PBS control, although the shifts were not statistically significant (p > 0.05). In the oAF, col’ase treatment tended to upregulate oAF markers compared to PBS (p > 0.05). Results are shown as the mean and standard deviation of four replicates with *p = 0.0360.

FIGURE 9

Bovine mRNA expression at day 7 for proinflammatory markers relative to day 0 and RPLP0 for the PBS-injected and enzyme-treated sample groups. Papain treatment produced the highest upregulation in the expression of a proinflammatory marker, 1L-1β, although the none of the shifts were statistically significant relative to each other or the PBS control (p > 0.05). Results are shown as the mean and standard deviation of four replicates.