The Effect of Genipin Matrix Augmentation on the Retention of Glycosaminoglycans in the Intervertebral Disc—A Pilot Study

Abstract

The degradation of intervertebral disc proteoglycans, including the loss or shortening of their hydrophilic glycosaminoglycan chains, causes a loss of disc hydration, leading to an increase in solid matrix stresses. This illustrates one aspect of the complex multifactorial relationship between tissue degradation and the resulting mechanical dysfunction. Genipin matrix augmentation has previously been evaluated with regard to its ability to improve mechanical properties of the disc, increasing joint stability and permeability. The study aim was to evaluate the ability of genipin augmentation to increase retention of glycosaminoglycans in disc specimens subjected to free swelling. Three different models were utilized: whole bovine caudal discs, partial annulus specimens from bovine, and human thoracic discs. Total glycosaminoglycan release to a surrounding bath was quantified using a modified dimethyl-methylene blue assay. Genipin solution injections reduced glycosaminoglycan loss by 44.0% in intact bovine discs compared to buffer-only controls (p = 0.027), by 75.8% in partial bovine annulus specimens (p = 0.0004), and by 51.9% in human annulus specimens (p = 0.017). The combination of increased permeability and glycosaminoglycans retention may produce beneficial effects on nutritional flow, diurnal irrigation, and reduction of matrix solid phase stress. Combining these effects with the ability to improve joint stability and augment tissue mechanical properties suggests this nano-scale device may be capable of arresting ongoing degeneration.

1. Introduction

Disc degeneration is a term that has historically referred to numerous, often age-related, irreversible changes to intervertebral discs. Most of these degenerative traits have a strong genetic component [1] and are thought to stem from other core factors such as nutritional deficiency [2], mechanical insufficiency [3], and biochemical degradation of disc proteoglycans [2,4,5,6], including the loss of hydrophilic glycosaminoglycans (GAGs), particularly chondroitin sulphate, from the extracellular matrix. Studies have shown that this compositional change and the accompanying disc dehydration are correlated with discography in chronic low back pain patients [7], albeit with a high prevalence in asymptomatic individuals [8]. Conversely, lower disc hydration, reduced cellularity, and loss of disc height in older individuals is not correlated to increased incidence of low back pain [8,9,10]. Proteoglycan degradation begins early in life with readily observable changes by the end of the second decade of life [4,11]. Lower disc hydration translates to reduced fluid phase load support [12], causing tissue loading to be increasingly supported by the elastic solid phase of the matrix [13].

Even a small reduction in fluid phase load support can be presumed to exacerbate the vicious cycle of accumulating tissue damage due to increased annular shear strains [14] and “microtrauma” [3,15] in an avascular and, therefore, repair-deficient tissue. The accumulation of tissue microtrauma weakens the disc, leading to additional matrix overload and mechanical dysfunction. Loss of fluid phase load support could eventually lead to discontinuity in endplate loading [16] and subsequent endplate failures, increased disc bulge under load [17], and an increase in annular tears and fissures [18,19]. The presence of tears and fissures in a failed disc annulus would facilitate neoinnervation [20,21,22] coupled with the release of degradative and proinflammatory cytokines [23,24,25]. Adding to this degradative cascade, a drop in swelling pressure impairs proteoglycan synthesis [26,27]. Conversely, retention of even degraded proteoglycans can help maintain the swelling pressure within a disc [28]. The annulus provides the only route for aggrecan degradation products to escape the disc, but it is not an easy route. Therefore, despite the early onset of proteoglycan degradation, the decline in proteoglycan swelling pressure takes decades to occur. This suggests that providing resistance to the loss of degraded proteoglycans from the disc could be a potential avenue to resist disc degradation.

Because of the primary role that microdamage accumulation and increasing mechanical dysfunction plays in an avascular tissue with inadequate tissue repair, mechanical effects of matrix augmentation have been the focus in prior studies. The selection of genipin and optimization of buffer components to address the mechanical deficiencies of the degraded intervertebral disc annulus has been evaluated in many studies. These studies have demonstrated genipin’s ability to self-polymerize [29,30,31,32], forming an intra-discal mesh of load-supporting genipin oligomers of varying lengths [33,34] that covalently bond to the collagen matrix with multiple effects related to improving mechanical properties of the annulus [33,35], reducing joint instability [36] and disc bulge under load [33], and arresting enzymatic and mechanical loading-induced degradation of the annulus fibrosus [37,38]. Relative to disc hydration, genipin matrix augmentation has been shown to double hydraulic permeability of the disc [39]. A possible adverse effect of increased cross-linking, at least relative to endogenous, glycation-induced cross-links, is the potential reduction in tissue hydration, even when proteoglycan content is unaffected [40]. The aim of the present study was to evaluate the ability of genipin reagent injections to increase retention of GAGs in disc annulus subjected to the artificial GAG leaching conditions of free swelling.

2. Materials and Methods

Previous studies attempting to quantify GAG content in genipin-treated discs using Farndale’s dimethyl-methylene blue (DMMB) binding assay [41] met with limited success due to the colorimetric interference of the blue pigmentation generated by genipin upon its polymerization and reaction with amines [42]. In the present study, methods were developed to discern the effect of genipin matrix augmentation (in situ formation of short chains of genipin molecules) on GAG retention in the disc matrix by quantifying total GAG release to the surrounding bath. Note that DMMB only binds to sulfated GAGs, and so the signal obtained does not reflect hyaluronic content. Furthermore, DMMB does not differentiate between free GAGs and those bound to core proteins (i.e., proteoglycans), so for the purposes of this paper, references to GAGs released to the bath will reflect the total GAG content of the sample (i.e., free GAGs and those attached to the core proteoglycan proteins).

Urban and Maroudas [43] demonstrated that subjecting disc tissue to an aqueous solution without mechanical stress on the matrix results in free swelling- accelerated loss of proteoglycans from the matrix. While this free swelling does not closely mirror the diurnal inflow-outflow fluid convection that human discs experience, or the typically very slow loss of GAGs and other molecules from the matrix, its use as an accelerated method for GAG depletion has been previously used [44]. These methods were therefore modified and combined to provide a means for quantifying the effect of genipin-mediated matrix augmentation on the retention of proteoglycans/GAGs in the disc matrix when subjected to hyper-physiologic water content and fluid exchange. Three different ex vivo intervertebral disc models were used in this study, comprised of either bovine caudal disc tissues or human thoracic disc tissues (

Table 1. Ex vivo experimental model design.

Experimental Model	Genipin Treated (n)	Control (n)	Description
Bovine Motion Segments	8	13	Model representative of clinical delivery of buffered genipin reagent. Use of bovine joints to minimize inter-specimen variations.
Bovine Annulus Segments	9	9	Model ensures uniform distribution of genipin reagent injected into annulus matrix.
Human Thoracic Annulus Segments	9	9	Uniform distribution of genipin reagent injected into human disc annulus segments.

). In the first experiment, bovine tail motion segments (caudal intervertebral joints) were left intact during the treatment process. The second and third experimental models involved partial annulus fibrosus specimens cut from bovine caudal and human thoracic discs, respectively. Human disc degeneration is known to be associated with GAG loss, and the two bovine groups were selected for two purposes: (1) to evaluate the difference in retention of GAGs after treatment with a genipin cross-linking solution in an ex vivo model with less inherent variability [44] and (2) to demonstrate the differences and potential limitations to an annulus segment model compared to a whole disc injection model. Conversely, the human annulus group was included to examine the effects of genipin matrix augmentation on human tissue, which contains more variability with regards to composition, age, GAG content, and degeneration, and to confirm, with a more relevant tissue model, the effects seen in the bovine models.

In the bovine whole disc experiments, external tissues were removed, and bone–disc–bone motion segments were isolated. After establishing the protocol using five control specimens, the specimens were randomly divided into treatment and control groups. The 40 mM genipin concentration and EPPS-phosphate pH 9 buffer used in these experiments were determined from prior studies. Biomechanical experiments [45] demonstrated that lower than 40 mM concentrations of genipin resulted in less joint stabilization capabilities, and biochemistry experiments [32,33,34] demonstrated preferential reaction rate and diffusion through the annulus tissue using EPPS buffer with reaction rates enhanced by phosphate ions and elevated pH. As this was a pilot investigation to determine if this genipin reagent was capable of increasing retention of GAGs in the annulus matrix, the dose–response relationship between genipin reagents and GAG retention has not been investigated.

The disc annulus was injected from two posterolateral (dorsal-lateral) sites with a total of 1.5 mL of either 40 mM genipin in 50 mM EPPS (4-(2-hydroxyethyl)piperazine-1-propanesulfonic acid), 50 mM sodium phosphate, pH 9.0 buffer (n = 8), or with just the buffer alone (n = 13). The method and volume of the injection was based on a previously developed human disc injection protocol and the approximate relative volume of the bovine caudal discs compared to a medium-sized human lumbar disc (1/2). The human lumbar disc injection protocol is primarily directed at coverage of the annulus region of the disc for load support and joint constraint purposes. Therefore, there was no full penetration of the genipin reagent into the nucleus pulposus region of the whole bovine caudal discs. The specimens were then incubated for 4 h at 37 °C, followed by 18 h of overnight incubation at 4 °C to allow the genipin ample time to polymerize and bind to the tissue matrix. The discs were then excised, weighed, and soaked unloaded in a 20 mL phosphate buffered saline (PBS) bath at 4 °C for 18 h to allow GAGs to leach out of the tissue [43].

In the second set of experiments, nine bovine discs were excised, and the nucleus was removed prior to treatment. The annulus material was cut into smaller specimens with a wet weight between 400 and 1000 mg. The specimens were randomly divided into two groups, with one group receiving an injection of 40 mM genipin in the buffer described above (n = 9), while the other group was injected with buffer only (n = 9). The injection volumes were proportional to specimen size and limited to the amount required to saturate the specimen. The disc segments were incubated at 37 °C for 4 h and then placed in a 5 mL bath of PBS at 4 °C for a further 18 h. This latter step allowed free swelling-induced release of GAGs from the matrix [43].

In total, 18 200–400 mg human annulus fibrosus specimens from three thoracic discs removed from three human thoracic cadaveric spines were used in the third set of experiments. The human specimens were randomly divided into a 40 mM genipin in buffer (n = 9) and buffer-only groups (n = 9), similar to the bovine disc specimens. After the injections, the specimens were incubated at 37 °C for 4 h and then placed in a 5 mL bath of PBS at 4 °C for a further 18 h.

Following the 18 h free swelling incubation, the bath fluids from all groups were analyzed for GAG content colorimetrically using a modified version of the DMMB assay method described previously [46]. Previous methods had used an enzymatic digestion step to remove proteins that could potentially interfere with the DMMB signal, but preliminary experiments demonstrated that proteolysis in the system had little effect on the relative magnitudes of signal from different samples, but reduced the overall sensitivity of the assay by 20–30%. Bath fluids were therefore not treated with protease prior to GAG quantification.

Briefly, the bath fluid was centrifuged for 10 min at 4000 rpm to sediment the remaining fragments of tissue. The clarified supernatant was then incubated for 18 h at 56 °C to denature assay-interfering proteins [41]. The supernatant was diluted by adding 100 µL of the solution to 1 mL of 1,9-dimethylmethylene blue solution. The mixture was shaken vigorously for 30 min and then centrifuged for 20 min at 13,000 rpm in a 34 mm radius microfuge to pellet GAG-DMMB complexes. The supernatant was removed, 1 mL of the decomplexation solution was added (50 mM sodium acetate, pH 6.8, 10% (v/v) propan-1-ol, 4 M guanidinium hydrochloride), and the specimens were shaken for another 30 min. Absorbances were read at 656 nm using PBS as a blank and converted to concentrations using a standard curve consisting of 0–2 µg/mL of chondroitin sulfate. Results were further normalized to specimen wet mass and analyzed using a non-parametric Mann–Whitney U test.

3. Results

The results of all three studies demonstrated that injection of a buffered genipin reagent increased the retention of proteoglycans and glycosaminoglycans in disc matrices subjected to free swelling-induced accelerated release of accumulated proteoglycan degradation products (Figure 1).

Intact bovine caudal discs injected with buffer-only solution were found to release an average of 1.27 µg of GAGs per mg of tissue, compared to 0.72 µg/mg tissue for the genipin-treated specimens (mean difference: −0.56; 95% CI: −1.00, −0.10; p = 0.027), corresponding to a 44.0% reduction in the loss of GAGs from the matrix. The buffer-treated partial bovine annulus specimens released an average of 1.01 µg of GAGs per mg of tissue, compared to 0.24 µg/mg tissue for the genipin-treated specimens (mean difference: −0.77; 95% CI: −0.94, −0.60; p = 0.0004), corresponding to a 75.8% decrease in GAG loss from the matrix. The corresponding human thoracic annulus specimens released an average of 3.99 µg GAGs/mg tissue, which was reduced to 1.92 µg GAGs/mg tissue following genipin injection (mean difference: −2.07; 95% CI: −3.21, −0.93; p = 0.017), corresponding to a 51.9% decrease in GAG loss during free swelling.

The mean GAG loss per mg of buffer-only treated whole bovine discs was 26% greater than the mean loss per mg of buffer-only treated partial (cut) bovine annulus specimens, but this difference was not significant (mean difference: 0.26; 95% CI: −0.07, 0.59; p = 0.117). The mean GAG loss per mg of buffer-only treated partial (cut) human annulus specimens was 295% greater than the mean loss per mg of buffer-only partial bovine annulus specimens. This difference was statistically significant (mean difference: 2.98; 95% CI: 1.59, 4.37; p = 0.001).

4. Discussion

The breakdown and loss of proteoglycans from lumbar intervertebral discs causes a loss of disc hydration, leading to an increase in solid matrix stresses. In the circumstances where cell viability and proteoglycan synthesis are low, retention of proteoglycans in the matrix may be vital to disc homeostasis. All three experimental models in this pilot study demonstrated a reduction in GAGs lost from the disc matrix due to injection of a buffered self-polymerizing genipin reagent compared to buffer-only injections. The magnitudes of the present ex vivo study results agree with the approximately 50% (non-significant) greater retention of GAGs in the nucleus pulposus due to genipin cross-linking observed in the in vivo study by Yerramalli et al. [47]. Based on a previous quantification of GAGs per mg of dry tissue in rat tail and lumbar discs [42], approximately 9% of matrix GAGs were lost in the sham-treated specimens in this static, free swelling study.

It is important to note the non-physiological and static conditions of the free swelling GAG leaching experimental models used in these pilot experiments and previous studies. A cut section of annulus or even an unloaded whole disc is not a proxy for an intact, dynamically loaded disc. Instead, they were used in these experiments to ensure accelerated GAG loss from disc tissues in order to obtain a first indication of whether genipin cross-linking could provide a measurable resistance to GAG leaching. Therefore, there remains a corresponding uncertainty regarding the degree of effectiveness of genipin cross-linker resistance to GAG loss under physiological conditions. Free swelling accelerates the washout of proteoglycans and GAGs by eliminating some of the tissue stress, preventing water imbibition in the case of an intact but unloaded disc, and potentially eliminating a greater amount of this resistance to water influx and corresponding GAG leaching in the case of a cut specimen. In this way, it was expected that the cut bovine and human tissue models would elevate the potential GAG loss from the annulus tissues. However, any untreated portion of the nucleus in whole disc specimens would be expected to more freely release GAGs into the bath under free swelling in the control specimens. This untreated nucleus region would also experience less resistance to the release of GAGs resulting from the genipin treatment. The current models were unable to delineate these competing effects. Both cut and whole disc models are intended to accelerate GAG release from the matrix. The observed trend of 26% greater GAG loss per mg of tissue from control whole disc specimens compared to cut annulus bovine specimens was not significant. These different bovine tissue models were not expected or intended to show comparable control tissue results due to the competing mechanisms for GAG loss discussed here.

Furthermore, the convection-related expulsion of disc sub-components during physiological compressive and cyclic loading is not replicated in the free swelling model. Consequently, the static models used in this pilot study may overestimate retention efficacy by failing to account for the convection-related expulsion of GAGs during disc loading cycles.

The substantial increase in proteoglycan retention following genipin-mediated matrix augmentation demonstrated in this study (44–76%) could in part be due to the direct bonding of GAGs or the proteoglycan core protein to the matrix. This is supported by a study [48] in which genipin was found to be not only effective in bonding to collagen molecules but also in covalent adhesion of chondroitin sulfate and hyaluronan. They also found a decrease in the pore size of fabricated collagen type II sponges with genipin cross-linking. The results of this present study suggest that while the matrix is more permeable to water after treatment with genipin [39], large molecules may be trapped by the polymer-modified matrix. The covalent bonding of genipin oligomers to amine groups in the collagen matrix could also affect the electronegativity of the matrix, thereby affecting the attraction of GAGs to the matrix. It is unknown from the present experiments whether direct covalent bonding of GAG proteins to the disc annulus matrix via genipin oligomers is the dominant effect or whether physical containment and/or electrostatic resistance resulting from the added network of genipin oligomers in the collagen matrix is the primary factor impeding their release from the matrix. Essentially, matrix-bound genipin oligomers are capable of both trapping and covalently binding to GAGs and proteoglycans. A schematic representation of GAG trapping and binding mechanisms is shown in Figure 2.

While direct comparison of genipin cross-linker effectiveness in resisting GAG loss would not be appropriate between different ex vivo experimental models, genipin matrix augmentation appeared to be more effective in reducing GAG loss in the partial annulus specimens compared to whole bovine discs. This difference could have resulted from a higher volume of reagent per mass of disc tissue for the partial annulus specimens or from less uniform distribution of genipin delivered via injections into the whole discs compared to the partial annulus specimens. In particular, a reduced level of genipin diffusion into and throughout the nucleus pulposus region of the intact discs would be expected from the bilateral annular delivery method used in these experiments. It has also been observed in numerous ex vivo, animal in vivo, and clinical studies by the authors that the posterolateral intra-annulus injection procedure, while effective in producing genipin oligomer coverage of the majority of the inner and outer disc annulus for mechanical purposes, was not able to ensure uniform and complete coverage of the annulus tissues. This injection method is not intended to be directed to nucleus pulposus tissues due to the primary desired effect of supplementing annulus load support and motion constraint.

The biocompatibility of the genipin formulation used in this study was evaluated by a full battery of biocompatibility tests, which included acute/sub-acute/sub-chronic/chronic large animal testing, demonstrating satisfactory biocompatibility (unpublished reports). These studies were followed by initial clinical studies that demonstrated spinal joint stabilization, safety, and reductions of patient-reported pain and disability from 2 weeks through 2 years post-treatment [49]. The present data, as well as previous literature [47], may suggest that research beyond this pilot study should be conducted to determine whether an additional injection into the nucleus region of intact discs may be advisable, where maximum retention of proteoglycans is a primary goal of the treatment.

GAG loss from sham-treated human annulus specimens was 3-fold greater than GAG loss from comparable buffer-only bovine specimens. It is suggested that this difference primarily reflects the influence of age and degeneration on the human tissues, including accumulated proteolytic degradation of aggregating proteoglycans, resulting in greater amounts of unattached GAGs in the tissue matrix. Other factors that may have contributed to this difference include the species and anatomical (caudal versus thoracic) differences between the specimens. The fact that the variability of the results from the human tissues was greater than that from bovine tissues supports the suggestion that specimen age and level of degeneration are important factors in the susceptibility of disc tissue to GAG loss.

As noted above, free swelling-induced leaching of proteoglycans is an accelerated, non-physiological process. As only one timepoint and similar free swelling conditions were used in these experiments, no conclusions can be drawn from this data regarding the linearity of GAG loss as a function of free swelling time and conditions. Likewise, the ratio of GAG loss between genipin and sham-treated discs specimens under in vivo physiological conditions (including diurnal and loading-induced inflow and outflow) over different periods of time cannot be concluded from these experiments. Although genipin self-polymerization, covalent bonding to amines, and mechanical effects have been shown not to be inhibited by body temperatures, age, and disc degeneration level, the ex vivo tissue models used in this study cannot fully predict the influences of factors, such as tissue composition variations, tissue pH, or oxygen tension, on GAG retention capabilities. Composition and temperature variations have produced nonlinear effects in less complex synthetic materials such as hydrogels [50]. However, the GAG retention effects of genipin polymer additions to the disc matrix were unmistakable, and it seems reasonable that the free swelling model may indicate potential long-term effects by way of an accelerated model. Future less quantitative and more complex ex vivo and in vivo studies can investigate whether repeated intervals of inflow and outflow-simulating mechanical loading and diurnal conditions of lumbar discs, longer exposure to free swelling, or other proteoglycan-leaching conditions could reduce or magnify the genipin-induced GAG retention effects seen in this static ex vivo study. It remains unclear, and a subject for future studies, whether genipin oligomers preferentially bind large or degraded smaller proteoglycans or unbound GAGs to the matrix. It would also be valuable in future studies to investigate the effects of the combination of genipin treatment and free swelling leaching on collagen type II, Aggrecan, and cell viability using histology.

It has been previously demonstrated that genipin matrix augmentation was able to double the overall hydraulic permeability of the disc [39]. This is somewhat surprising in view of these present results, considering that greater proteoglycan content has been shown to be inversely related to tissue permeability [44]. However, other factors, such as water content and pore size (gap or pore size in the tissue matrix has been suggested to be influenced by a cross-linking-induced consolidation of microfibers [39,51]), can also have a strong effect on tissue permeability. It is reasonable to expect that the combination of increased retention of proteoglycans and increased permeability resulting from genipin matrix augmentation would produce beneficial effects on nutritional flow to the disc, diurnal irrigation, maintenance of disc height, and reduction of matrix stress. Indications from chronic [52] animal studies, in addition to early clinical studies [49], suggest that there may be counteracting genipin polymer effects, restricting an increase in disc volume or height. It is reasonable to assume that the matrix-binding effects of attached genipin polymers would restrict the expansion of disc volume in opposition to an increase in swelling pressure due to increased retention of water-imbibing GAGs (over long periods of time) and increased hydraulic permeability (immediately following treatment). Neither of these in vivo studies demonstrated a change in disc height from baseline at a relatively short-term (3 months) post-implantation imaging assessment. Likewise, there is currently no data to assess whether the ability of genipin polymers to increase retention of GAGs helps to preserve disc height over the long term.

Combining the ability to resist loss of GAGs from the disc matrix with the demonstrated ability of the genipin intra-annulus mesh to improve joint stability [36] and tissue strength [35] suggests that a genipin-based device intervention may be sufficient to arrest ongoing degeneration. It follows that these beneficial effects may provide maximum clinical impact if used in the treatment of early- to mid-stage disease. Initial clinical testing of 20 enrolled chronic or recurrent low back pain participants demonstrated the ability to improve patient reported outcomes (pain and disability relief) from a 2-week follow-up time point through 2 years post-treatment, with confirming objective kinematic data [49]. Additional clinical research is needed to better assess long-term effects of this treatment modality.

While not evaluated in the present study, genipin augmentation of the disc matrix could potentially have adverse effects on proteoglycan synthesis in the disc. Vickers et al. [53] demonstrated that highly cross-linked type II collagen scaffolds resisted cellular condensation (reduction of intercellular space), resulting in lower chondrogenesis. Likewise, it has been shown that a drop in nucleus pressure can lead to a decrease in proteoglycan synthesis [26,27]. Chuang et al. [54] found that genipin matrix augmentation reduced intradiscal pressure by altering the mechanics of compressive load support. This effect could potentially reduce proteoglycan synthesis in the nucleus region.

5. Conclusions

A self-polymerizing genipin reagent was investigated to determine its ability to increase the retention of GAGs within annulus fibrosus tissue from both bovine whole discs and partial annulus specimens and from partial human annulus specimens. Specimens were treated with a buffered genipin solution and evaluated for GAG loss after free swelling. A reduction in the loss of GAGs from the tissue specimens was seen in all genipin-treated test groups, and notably, a 51% decrease was seen in the human annulus specimens compared to buffer-only controls (p = 0.017). Under conditions associated with lumbar disc degeneration, where cell viability and proteoglycan synthesis are low, retention of proteoglycans in the matrix may be vital to disc homeostasis. Proteoglycan and GAG loss from intervertebral discs can cause an increase in solid matrix stress and ultimately add to the degenerative cascade of disc degeneration. The results of this study and previous work indicate that genipin’s ability to increase matrix retention of GAGs, as well as to provide mechanical support to the degraded disc, may be beneficial in the treatment of disc degeneration. Previous studies focused on delivery of genipin to the annulus region of discs in order to augment load support and joint stabilization. These current data suggest that if maximum retention of disc proteoglycans is considered a primary goal of an intervention, an injection of genipin reagent into the nucleus region of the disc, in addition to injections targeting the annulus, may be beneficial.