The role of biomechanical factors in models of intervertebral disc degeneration across multiple length scales

Abstract

Low back pain is the leading cause of disability, producing a substantial socio-economic burden on healthcare systems worldwide. Intervertebral disc (IVD) degeneration is a primary cause of lower back pain, and while regenerative therapies aimed at full functional recovery of the disc have been developed in recent years, no commercially available, approved devices or therapies for the regeneration of the IVD currently exist. In the development of these new approaches, numerous models for mechanical stimulation and preclinical assessment, including in vitro cell studies using microfluidics, ex vivo organ studies coupled with bioreactors and mechanical testing rigs, and in vivo testing in a variety of large and small animals, have emerged. These approaches have provided different capabilities, certainly improving the preclinical evaluation of these regenerative therapies, but challenges within the research environment, and compromises relating to non-representative mechanical stimulation and unrealistic test conditions, remain to be resolved. In this review, insights into the ideal characteristics of a disc model for the testing of IVD regenerative approaches are first assessed. Key learnings from in vivo, ex vivo, and in vitro IVD models under mechanical loading stimulation to date are presented alongside the merits and limitations of each model based on the physiological resemblance to the human IVD environment (biological and mechanical) as well as the possible feedback and output measurements for each approach. When moving from simplified in vitro models to ex vivo and in vivo approaches, the complexity increases resulting in less controllable models but providing a better representation of the physiological environment. Although cost, time, and ethical constraints are dependent on each approach, they escalate with the model complexity. These constraints are discussed and weighted as part of the characteristics of each model.

FIG. 1.

Forces and torques acting on the intervertebral disc.

FIG. 2.

Load components and force vectors during walking up the stairs obtained from a telemeterized vertebral body replacement. Results from walking up five steps are shown. Videoclips of the original measurements with the synchronously shown loading data file wp1_140307_1_123 from database OrthoLoad. G. E. Bergmann, see http://www.OrthoLoad.com for “Charité Universitaetsmedizin Berlin, ‘OrthoLoad’ (2008)” (accessed April 1, 2021). Copyright 2008 Authors, licensed under a Creative Commons Attribution (CC BY) Unported License.

FIG. 3.

Requirements for IVD models which should be considered during experimental design.

FIG. 4.

Model comparison based on physiological resemblance and data acquisition capabilities.

FIG. 5.

The effect of tensile strain on the actin cytoskeleton of intervertebral disc cells cultured on type I collagen. Visualization using Alexa-488™-phalloidin (scale bar = 8 μm). F-actin organization in nucleus pulposus (NP) and outer annulus fibrosus (OAF) cells before and after 10% strain (60 min). Punctate F-actin labeling in NP and OAF cells (white arrows) were replaced with extensive F-actin stress fibers upon tensile strain. Reprinted with permission from Li et al., Eur. Cells Mater. 21, 508–522 (2011). Copyright 2011 Authors, licensed under a Creative Commons Attribution (CC BY) Unported License.

FIG. 6.

Structure and composition of fresh, death and culture IVDs (control and after injury). (a) GAGs quantification; (b) disc height in all groups; (c)–(f) Safranin‐O staining in all disc. Reprinted with permission from Abraham et al., J. Orthop. Res. 34(8), 1431–1438 (2016). Copyright 2016 John Wiley and Sons.

FIG. 7.

MRI T2^* mapping of transverse sections of bovine discs at day 9 (before loading) and day 16 (after loading). The blue scale represents higher content of water while the orange-yellow indicate lower water content. PDD represents papain disc digestion. The untreated control has been injected with Phosphate Buffered Saline (PBS); PDD + PBS injection; PDD + thermo-hydrogel (TR‐HG, material control); PDD + TR‐HG + autologous bovine NP cells (bNPCs); and PDD + TR‐HG + hMSCs Reproduced with permission from Malonzo et al., J. Tissue Eng. Regener. Med. 9(12), E167–E176 (2013). Copyright 2013, John Wiley & Sons, Ltd.

FIG. 8.

Load capability and comparison of bioreactor studies.

FIG. 9.

Midsagittal sections of rabbit intervertebral discs (IVDs). UPAL, ultra-purified alginate. AF; annulus fibrosus, and NP; nucleus pulposus. (a) Hematoxylin and Eosin (H&E); (b) safranin O. (c) Histological grading in rabbit discs. Midsagittal sections of sheep IVDs stained (d) H&E; (e) safranin O; and (f) Histological grading in sheep discs. Reproduced with permission from Tsujimoto et al., EBioMedicine 37, 521–534 (2018). Copyright 2018 Authors, licensed under a Creative Commons Attribution (CC BY) Unported License.

FIG. 10.

Cost, time and ethics implication of in vitro and in vivo IVD models. The cost of animals was estimated using catalogues from companies selling cell lines, research models (e.g., Charles River 2020 Catalogue¹⁴⁷) and animal husbandry markets and prices reports in the UK. Prices for animal upkeeping were acquired from internet sources (e.g., costs of Raising Goats in 2019¹⁴⁹) and anecdotal information from colleagues working with small and large models in the USA and Europe. The estimated expenses present a rough estimate and might vary in different geographies and be impacted by bulk purchases and infrastructure in place.

FIG. 11.

Examples of studies using the different approaches to investigate the degeneration and treatment of the IVD. (a) and (b) NP cells cultured at different pH at day 0 and day 7. Green cells represent live cells, while red cells represent dead cells. Non-degenerate environment was represented by pH 7.4 while a severely degenerated environment was simulated with pH 6.2. Reproduced with permission from Gilbert et al., Sci. Rep. 6, 28038 (2016). Copyright 2016 Authors, licensed under a Creative Commons Attribution (CC BY) Unported License. (c) and (d) Live/dead (green/red) staining of IVD cells seeded on chitosan hydrogel for 24 h and 3 D [cytoskeleton filaments (green), nuclei (blue)]. Reproduced with permission from Yang et al., RSC Adv. 8, 68 (2018). Copyright 2018 Authors, licensed under a Creative Commons Attribution (CC BY) Unported License. (e)–(g) Histological evaluation of ex vivo cultures, stained with aldehyde fuchsin and alcian blue to identify sGAG and deep purple to indicate GAG accumulation. Images shown represent a healthy disc, a disc with nucleus injury and primed microencapsulated bone marrow stromal cells (BMSCs) (all after 28 days of culture). Reproduced with permission from Naqvi et al., Eur. Cells Mater. 37, 134–152 (2019). Copyright 2019 Authors, licensed under a Creative Commons Attribution (CC BY) Unported License. (h)–(i) Safranin O/fast green-stained IVD sections after 15 D organ culture including annulotomised discs and annulotomised discs repaired with polyurethane and collagen scaffold seeded with TGF-β1-pre-treated AF cells (PU-Col-AFCs) and compressed for 1 h at 0.02–0.2 MPa, 0.2 Hz daily. Reproduced with permission from Du et al., Eur. Cells Mater. 39, 1–17 (2020). Copyright 2020 Authors, licensed under a Creative Commons Attribution (CC BY) Unported License. (j)–(l) Genetic deletion of miR-141 suppresses spontaneous and surgically induced IDD. Safranin O staining of intervertebral disc from 6-month-old mice and 22 month-old mice with and without gene deletion (WT: wild-type). Reproduced with permission from Ji et al., Nat. Commun. 9, 5051 (2018). Copyright 2018 Authors, licensed under a Creative Commons Attribution (CC BY) Unported License. (m)–(o) Mid-sagittal histological sections of the disc intact, with nucleus injury and with treatment composed of dodecyl-amide of hyaluronic acid (DDAHA) hydrogel and bone marrow derived mononuclear cells (BMC) after 12 weeks. Reproduced with permission from Reitmaier et al., Eur. Spine J. 23, 19–26 (2014). Copyright 2014 Authors, licensed under a Creative Commons Attribution (CC BY) Unported License.