Quantifying Baseline Fixed Charge Density in Healthy Human Cartilage Endplate: A Two-point Electrical Conductivity Method

Abstract

Study Design

Regional measurements of fixed charge densities (FCDs) of healthy human cartilage endplate (CEP) using a two-point electrical conductivity approach.

Objective

To determine the FCDs at four different regions (central, lateral, anterior, and posterior) of human CEP, and correlate the FCDs with tissue biochemical composition.

Summary of Background Data

The CEP, a thin layer of hyaline cartilage on the cranial and caudal surfaces of the intervertebral disc, plays an irreplaceable role in maintaining the unique physiological mechano-electrochemical environment inside the disc. FCD, arising from the carboxyl and sulfate groups of the glycosaminoglycans (GAG) in the extracellular matrix of the disc, is a key regulator of the disc ionic and osmotic environment through physicochemical and electrokinetic effects. While FCD in the annulus fibrosus (AF) and nucleus pulposus (NP) have been reported, quantitative baseline FCD in healthy human CEP has not been reported.

Methods

CEP specimens were regionally isolated from human lumbar spines. FCD and ion diffusivity were concurrently investigated using a two-point electrical conductivity method. Biochemical assays were used to quantify regional glycosaminoglycan (GAG) and water content.

Results

FCD in healthy human CEP was region-dependent, with FCD lowest in the lateral region (p=0.044). Cross-region FCD was 30–60% smaller than FCD in NP, but similar to the AF and articular cartilage (AC). CEP FCD (average: 0.12±0.03 mEq/g wet tissue) was correlated with GAG content (average: 31.24±5.06 μg/mg wet tissue) (p=0.005). Additionally, the cross-region ion diffusivity in healthy CEP (2.97±1.00×10⁻⁶ cm²/s) was much smaller compared to the AF and NP.

Conclusions

Healthy human CEP acts as a biomechanical interface, distributing loads between the bony vertebral body and soft disc tissues and as a gateway impeding rapid solute diffusion through the disc.

INTRODUCTION

Comprising the largest avascular tissue in the human body, the intervertebral disc (IVD) includes three components: cartilage endplate (CEP), annulus fibrosus (AF), and nucleus pulposus (NP). The CEP, a thin layer of hyaline cartilage on the cranial and caudal surfaces of the disc, is the main pathway for solute exchange (e.g. ions and nutrient solutes) through the IVD ^1–4. Pathological changes in CEPs (e.g. calcification or degeneration) significantly alter the extracellular mechano-electrochemical environment inside the disc by changing mechanical and transport functions of the CEP ^5,6, which may initiate and accelerate disc degeneration by triggering a cascade of biological sequelae at cellular or molecular levels ^7–10. Healthy CEPs play a critical role in maintaining the physiological extracellular mechano-electrochemical environment inside the disc.

All three IVD components are negatively charged, due to the carboxyl and sulfate groups on the glycosaminoglycan (GAG) in the extracellular matrix ¹¹. Fixed charge density (FCD), representing the negative charges attached to the matrix in each tissue unit (per weight or volume), is a key factor regulating the unique physiological extracellular mechano-electrochemical environment of healthy IVD. Electrostatic interactions between FCD and mobile free ions results in physicochemical and electrokinetic effects, including high cation and low anion concentrations, Donnan osmotic pressure and swelling, and streaming potential/current ^12–14. Studies have shown marked effects of changes to the ionic/osmotic environment on isolated or in situ disc cell metabolism ^15,16. Additionally, FCD significantly impacts the mechanical (e.g. swelling pressure and compressive modulus) and transport properties (e.g. hydraulic permeability) of IVD tissues ^17,18. GAG loss, which reduces FCD, is correlated to human IVD degeneration ¹⁹. Therefore, FCD may indicate the degree of disc degeneration. Although FCDs in AF and NP have been measured ^20,21, to our knowledge, no quantitative assessment of FCD in healthy human CEP has been reported.

Previous FCD measurements have used the traditional tracer cation method (radio-labeled ions), imaging techniques (e.g. MRI), and mechanical indentation testing ^11,22,23. Recently, an electrical conductivity approach has proved to be a simple and reliable FCD measurement in bovine NP and AF ²¹. Human CEP is a thin hyaline cartilage layer with an average thickness of ~0.7 mm ²⁴, which could limit imaging techniques. Consequently, the electrical conductivity approach is more appropriate for FCD measurements in thin CEP specimens in vitro. Previous studies have revealed regional differences in morphology, biochemical composition, biphasic mechanical properties, and glucose/lactate diffusion rates between the central and peripheral regions of the CEP ^4,24–26. Therefore, we hypothesized that FCD in healthy human CEP was region-dependent and correlated with biochemical composition (i.e., GAG content). The objective of this study was to measure FCD at four regions (central, lateral, anterior, and posterior) in healthy human CEP using a two-point electrical conductivity method ²¹. At the same time, ion diffusivity was determined. Regional biochemical composition was characterized, including GAG content and porosity, to determine potential correlations with FCD and ion diffusivity in the CEP ^21,27, respectively. By determining baseline measurements of FCD and ion diffusivity in healthy human CEPs, this study may establish a role of the CEP in maintaining a healthy mechano-electrochemical environment inside human IVD, and a potential mechanism of disc degeneration.

MATERIALS AND METHODS

Specimen preparation

Twelve human lumbar spines (aged 33–65 years) obtained from an Organ Procurement Organization (LifePoint Inc., Charleston, SC) under institutional approval were screened based on the Thompson grading system ²⁸. Spines with degenerated discs (Grade III–V) were not used for baseline measurements. After dissection, it was confirmed that the CEPs from the spines with healthy discs (Grade I & II) had no signs of the artifacts (e.g. fissures and calcification). To limit tissue variation, only L4–L5 motion segments were used. Considering these criteria, four L4–L5 motion segments were harvested from four lumbar spines within 24 hours post-mortem (33 year old male, 42 year old male, 54 year old female, and 58 year old female).

Motion segments were immediately opened through the median plane of the IVD with a scalpel. Disc-shaped CEP specimens with a 5 mm diameter were prepared from superior and inferior surfaces at four disc regions (central, lateral, anterior, and posterior) (detailed in Figure 1). Due to disc symmetry, specimens harvested from left and right lateral regions were included in the same group. Prepared CEP specimens had an average height of 0.685±0.094 mm, similar to previous studies ^4,24. Four groups of CEP specimens harvested from four L4–L5 discs were tested in this study: central (n=8), lateral (n=9), anterior (n=8), and posterior (n=6). No posterior specimen was harvested from one of the discs due to its small size. Seven specimens harvested from lateral regions were used for testing protocol development, not data collection.

Figure 1.

Schematic view of specimen preparation. Motion segments were opened through the median plane of the IVD with a scalpel. A cylindrical plug (8 mm diameter) of NP or AF/endplate/bone was extracted from superior and inferior surfaces at four regions of the disc (center, lateral, anterior, and posterior) with a serrated-edge cutter. Each plug was then mounted onto the freezing stage of a sledge microtome (Leica SM2400, Leica Biosystem, Chicago, IL) to remove the overlying NP/AF (relatively more transparent than CEP tissue) and vertebral bone. The final disc-shaped CEP specimens were punched with a 5 mm corneal trephine for electrical conductivity measurements. Specimen preparation was performed in a humidified hood to prevent tissue dehydration.

Electrical conductivity measurements

FCD for each specimen was determined using a previously established two-point electrical conductivity method ²¹. Following preparation, specimens were first immersed in 0.15 M KCl solution (isotonic) for a 12 hour equilibration period at 4°C before conductivity measurements. During incubation, specimens were axially confined between two hydrophilic polyethylene porous platens (50–90 μm, Small Parts, Inc., Miami Lakes, FL), and radially confined within an acrylic chamber with an inner diameter of 5 mm to minimize swelling and proteoglycan (PG) leaching (Figure 2). Electrical conductivity was measured with a previously developed conductivity apparatus (Figure 2) ^29,30. Using a four-wire method, the resistance across the CEP specimen was measured at a low and constant current density of 0.015 mA/cm² supplied by a Keithley Sourcemeter (Model 2400, Keithley Instruments, Inc., Cleveland, OH). CEP electrical conductivity (χ) was calculated by:

$$\chi =\frac{\mathrm{h}}{\text{RA}}$$ (1)

where R, h, and A are the resistance, height, and cross sectional area of the specimen, respectively. Specimen height was measured with a current sensing micrometer. After the initial electrical conductivity measurement, each specimen was re-equilibrated in 0.03 M KCl solution (hypotonic) for 12 hours and electrical conductivity measurements were retested. All conductivity measurements were performed at room temperature (22°C).

Figure 2.

Experimental flow of the two-point conductivity approach for FCD and ion diffusivity measurements in healthy human CEPs. Electrical conductivity was measured with a previously developed conductivity apparatus ^29,30, consisting of two stainless steel current electrodes coaxial to two Teflon-coated Ag/AgCl voltage electrodes placed on the top and bottom of a cylindrical nonconductive Plexiglass chamber (5mm diameter). Measurements were achieved by sequentially equilibrating tissue specimens in two concentrations (i.e., isotonic and hypotonic) of potassium chloride (KCl) solution and measuring the corresponding electrical conductivity of the tissue.

FCD and ion diffusivity calculations

Specimen FCD was determined by measuring the electrical conductivities of the specimen in two KCl solutions (isotonic condition: 0.15 M KCl; hypotonic condition: 0.03 M KCl). According to the two point electrical conductivity method ²¹, tissue FCD (c^F) was calculated by:

$$c^F=2\sqrt{\frac{{\chi }_{\mathit{Iso}}^2[{(c_{\mathit{Iso}}^{\ast })}^2-{(c_{\mathit{Hypo}}^{\ast })}^2]}{{\chi }_{\mathit{Iso}}^2-{\chi }_{\mathit{Hypo}}^2}-{(c_{\mathit{Iso}}^{\ast })}^2}$$ (2)

where χ_i and $c_i^{\ast }$ (i = Iso or Hypo) are the electrical conductivity in the CEP and the concentration of the KCl solution under isotonic or hypotonic conditions. This method also determined the K⁺ or Cl⁻ ion diffusivity in CEP tissue. The diffusivity of cations (D⁺) was assumed to be the same as anions (D⁻) due to similar Stokes’ radii (K⁺: r_s=0.137 nm; Cl⁻: r_s=0.142 nm) ²⁷. Thus, ion diffusivity (D=D⁺=D−) was determined by ²¹:

$$D=\frac{RT}{2F_c^2{\varphi }^w}\sqrt{\frac{{\chi }_{\mathit{Iso}}^2-{\chi }_{\mathit{Hypo}}^2}{{(c_{\mathit{Iso}}^{\ast })}^2-{(c_{\mathit{Hypo}}^{\ast })}^2}}$$ (3)

where F_c is the Faraday constant, ϕ^w is porosity, and R and T are the gas constant and absolute temperature.

Porosity and GAG measurements

The porosity (water volume fraction) of the CEP specimen was determined using a buoyancy method ³¹. The specimen weight in air (W_wet) and in phosphate buffered saline (PBS) solution (W_PBS) were measured using the density-determination kit of a Sartorius analytical balance (Sartorius YDK01, Germany). Then, specimens were lyophilized and the dry weight (W_dry) recorded. Porosity was calculated by ³¹:

$${\varphi }^{\mathrm{w}}=\frac{{\mathrm{W}}_{\text{wet}}-{\mathrm{W}}_{\text{dry}}}{{\mathrm{W}}_{\text{wet}}-{\mathrm{W}}_{\text{PBS}}}\frac{{\rho }_{\mathit{PBS}}}{{\rho }_w}$$ (4)

where ρ_PBS and ρ_w are the mass densities of the PBS solution and water, respectively. Lyophilized tissues were then assayed for GAG content by Blyscan Glycosaminoglycan Assay kit (Biocolor Ltd., Newtonabbey, Northern Ireland) based on 1,9-dimethylmethylene blue dye binding ³², with standards provided by the manufacturer.

Statistical analysis

All measurements were reported using the means and standard deviations (SD). Significant differences by disc region were determined by one-way ANOVA for isotonic and hypotonic electrical conductivities, FCD, ion diffusivity, GAG content, and porosity. Correlations between FCD and GAG content, and between ion diffusivity and porosity were determined ³³ by linear mixed effects model with a random effect for spine. Due to the limited sample size, sex and age effects were not considered. The statistical analysis was conducted in R (R Core Team, 2015) and statistical significance was determined at p-values < 0.05.

RESULTS

Electrical conductivity

The cross-region electrical conductivities of CEP were 3.04±1.37 mS/cm and 1.65±0.75 mS/cm under isotonic and hypotonic conditions, respectively (Table 1). Significant regional variation was detected at both isotonic and hypotonic conditions in four regions (p<0.0001; Figure 3). The electrical conductivities in the central region were significantly higher (p<0.0001) than all other regions (Table 1).

Figure 3.

Region-dependent electrical conductivities of healthy human CEPs under isotonic and hypotonic conditions. Electrical conductivity in the central region was significantly higher than in other regions under both isotonic (p<0.0001) and hypotonic conditions (p<0.0001).

FCD and Ion diffusivity

FCD was reported as milliequivalent moles (mEq) per gram wet or dry tissue. The cross-region FCD of CEP was 0.124±0.032 mEq/g wet tissue or 0.363±0.118 mEq/g dry tissue (Table 1). Significant regional variation, with the lowest value in lateral region (Lateral: 0.101±0.018 mEq/g wet tissue), was found for FCD (p=0.044, Figure 4A). By contrast, no significant regional variation was found for FCD per gram dry tissue (p=0.254, Table 1), although the lowest mean value was found in lateral region (Lateral: 0.314±0.128 mEq/g dry tissue). Cross-region ion (K⁺/Cl⁻) diffusivity in CEP was 2.97±1.00×10⁻⁶ cm²/s (Table 1). Significant regional variation, with the highest value in the central region (Central: 4.67±0.50×10⁻⁶ cm²/s), was found for ion diffusivity (p<0.0001; Figure 5A).

Figure 4.

(A) Region-dependent FCDs from electrical conductivity measurements in healthy human CEPs. FCD in the lateral region was significantly lower (p=0.04) than in other regions. (B) The linear correlation between FCD and GAG content in healthy human CEPs was significant (r=0.472, p=0.005).

Figure 5.

(A) Region-dependent relative ion diffusivities of healthy human CEPs. Ion diffusivity in the central region was significantly higher (p<0.0001) than in other regions. (B) The linear correlation between ion diffusivities and porosities in healthy human CEPs was significant (r=0.555, p=0.001).

GAG content and Porosity

Cross-region GAG content in CEP was 31.24±5.06 μg/mg wet tissue or 93.64±18.43 μg/mg dry tissue, with no significant regional variations detected (p=0.121 for wet tissue, p=0.063 for dry tissue; Table 1). Cross-region CEP porosity was 0.726±0.069. Significant regional variation in CEP porosity was detected, with the central region (Central: 0.798±0.055) having higher values than other regions (p=0.0003; Table 1).

Correlation between FCD/Ion diffusivities and tissue biochemical composition

The correlation between FCD and GAG content was statistically significant (r=0.472, p=0.005; Figure 4B), while there was no statistical significant correlation between FCD and porosity (r=0.135; p=0.470; not shown). The correlation between ion diffusivity and porosity was also statistically significant (r=0.555, p=0.001; Figure 5B).

DISCUSSION

This study determined baseline FCD in healthy human CEP and correlated results with relevant biochemical composition (GAG content). The amount of FCD in human CEP was significant and similar to human AF and articular cartilage (AC) (Table 2), however 30–60% smaller than human NP ^14,20,34. Significant FCD creates a unique extracellular physicochemical environment inside the CEP, elevating cation and reducing anion concentrations, and raising osmotic pressure surrounding CEP cells. Changes to the ionic/osmotic environment have shown significant effects on the behavior of AF, NP, and AC cells, such as energy metabolism and cell volume regulation ^7,35. While CEP cells have rarely been studied, it can be reasonably expected that maintaining the unique ionic/osmotic environment is critical for CEP cell homeostasis. Additionally, as in the NP and AF, osmotic pressure due to high FCD significantly contributes to the CEP compressive modulus. Based on the Donnan equation, electroneutrality condition, and van’t Hoff’s equation for osmotic pressure ¹², the average measured FCD (0.124±0.032 mEq/g wet tissue) can generate 0.081±0.039 MPa osmotic pressure in healthy human CEP. It has been shown that decreases in FCD due to PG depletion reduces osmotic/swelling pressure, decreasing equilibrium and dynamic compressive moduli of AF tissues ¹⁸. Together with the larger tensile modulus and lower hydraulic permeability found in the CEP compared to other disc components (i.e., NP and AF) ^26,36, this study demonstrated that healthy human CEP acts as a unique biomechanical interface distributing loads between the bony vertebral body and soft disc tissues. Under pathological conditions (e.g. calcification or degeneration), the FCD of disc tissues decreases significantly ¹⁹. Therefore, the FCD of healthy CEP measured in this study, as well as in other disc components (i.e., NP and AF) ²⁰, provides baseline information for the evaluation of the mechano-electrochemical conditions (e.g. ionic concentrations, osmotic/swelling pressure, streaming potential and current) inside human IVD under physiological and pathological conditions ^13,14,37.

FCD in healthy human CEP was significantly correlated with its GAG content, not porosity, which is similar to other IVD and cartilage tissues ^21,23. This observation validated that the negatively charged nature of the CEP results from charged groups on the glycosaminoglycan (GAG) molecules on the proteoglycans (PGs) in the extracellular matrix. Moreover, there are two primary types of disaccharides in IVD tissues contributing to FCD, including chondroitin sulfate (CS), which has two negative charges per disaccharide, and keratin sulfate (KS), which carries one charged per disaccharide ^21,38,39. Previous studies have shown that the ratio of KS to CS varies regionally through the human IVD ^38,40, however the GAG content essay in this study wasn’t able to determine the ratio of KS to CS. It may explain the regional effect discrepancies between FCD and GAG content in the human CEP: significant regional FCD variations were revealed using the electrical conductivity method, with the lowest value in the lateral region, while no significant regional variation was found for GAG content measured using the biochemical assay. This difference further demonstrates that the electrical conductivity approach is a sensitive method to detect the physicochemical composition of the CEP ^21,27,29,34.

The two-point conductivity method was also able to estimate the diffusivity of Cl⁻ (or K⁺) ions in human CEPs, which were much smaller than in human NP, AF, and articular cartilage tissues (Table 2). Previous studies have found a lower glucose/lactate diffusivity and hydraulic permeability in the CEP compared to other disc and cartilaginous tissues ^25,26. The lower diffusivity values for CEP tissues is likely due to the lower porosity, as diffusivity has been shown to be dependent upon tissue hydration. Ion diffusivity was significantly correlated with porosity in the CEP. Region-dependent ion diffusivity in healthy CEP was highest in the central region, corresponding to the highest porosity in the central CEP, supporting the hypothesis that water content is a dominant determinant of solute diffusion properties in IVD tissues ^29,41. Due to the avascular nature of the human IVD, the healthy CEP acts as a gateway by blocking the rapid transport of ions, nutrient solutes, and angiogenic molecules, such as vascular endothelial growth factor (VEGF), through the disc and maintaining a stable physiological physicochemical environment inside the IVD ^1,6,42–44. As such, a delicate extracellular environment may be vulnerable to any CEP pathological changes, such as calcification ^5,6.

A few limitations should be noted for this study. A screening criteria (e.g. disc level: L4–L5; degeneration level: Grade I–II) was used to establish baseline characteristics of the healthy human CEP. As a result, 31 human CEP samples were harvested from 4 fresh, healthy and mature lumbar spines. In future work, more CEP specimens from a larger number of human lumbar spines will be collected to systematically characterize the effect of other factors (e.g. disc level, age, sex, diffusion direction, calcification, and degeneration) ¹⁹. To uncouple the effects of FCD and ion diffusivity on electrical conductivity (Equations 2 & 3), the diffusivities of cation and anion were assumed to be the same. Since the Stokes’ radii (r_s) of K⁺ and Cl⁻ are approximately equal (K⁺: r_s=0.137 nm; Cl⁻: r_s=0.142 nm), the value of their diffusivities can also be assumed to be the same (D=D⁺=D⁻). By contrast, the diffusivity of Na⁺ can be smaller than Cl⁻ due to its larger Stokes’ radii (r_s) (Na⁺: r_s=0.197 nm) ^21,27. Therefore, KCl instead of NaCl solution was used during the conductivity measurements to satisfy the assumption of equal diffusivities of cation and anion for calculating the FCD and ion diffusivity values ²¹.

In summary, this study measured baseline FCDs and ion (Cl⁻/K⁺) diffusivities in healthy human CEP, and further studied the effect of disc region on CEP properties. FCD and ion diffusivity in healthy human CEP were region-dependent due to its unique tissue biochemical composition (i.e., GAG content and porosity). FCD in healthy human CEP was similar to human AF and articular cartilage, while smaller than human NP. Significant FCD creates a unique extracellular ionic/osmotic environment critical for disc cell homeostasis. Osmotic pressure due to high FCD also contributes significantly to CEP mechanical function. Additionally, the ion diffusivity in human CEP were much smaller than in human NP an AF. Together, these results suggest that the healthy human CEP acts as a unique biomechanical interface distributing loads between the bony vertebral body and soft disc tissues, and acts as a gateway blocking the rapid diffusion of solutes through the IVD. These results further develop our understanding of the role of the human CEP in IVD biomechanics and physiology, and provide new insights into the mechanisms of disc degeneration and regeneration.