1.

Introduction

Low back pain (LBP) is the primary cause of activity limitation in patients below the age of 45. Upward of 80% of the U.S. population will be effected by LBP at some point in their lifetime.¹ Although LBP is a complicated condition with many possible etiologies, intervertebral disc degeneration (IDD) is believed to play a significant role (40% to 50% of all cases) in the development of this condition.² The intervertebral disc (IVD) consists of a fibrocartilaginous ring called the annulus fibrosus (AF), which surrounds the proteoglycan-rich nucleus pulposus (NP). Both structures are heavily imbibed with water, providing hydrostatic pressure to serve as a “shock absorber” of axial loads. IDD is thought to be a result of a combination of aging, abnormal mechanical stress, genetic predisposition, and environmental factors such as tobacco smoking. The common outcome of IDD is imbalanced disc matrix homeostasis leading to the net loss of extracellular matrix structure and water content.^3–5 Despite its prevalence, IDD has proven to be a difficult disease process to treat as the disc is a highly avascular structure with little inherent ability for repair or regeneration.⁶ Current areas of research focusing on treatment include physical therapy, spinal injections, molecular and gene therapy, stem cell therapy, and surgeries.^7–10 A unifying consensus of these treatment modalities is that early identification of disc tissue damage will aid in the treatment and recovery process.

For diagnosis of IDD via spine morphology imaging, several noninvasive imaging modalities are currently available, including computed tomography and magnetic resonance imaging (MRI).^11–13 As disc degeneration primarily involves changes in soft tissue structure, MRI has been a well-studied method for the evaluation of IDD.^14–17 Standard MR sequences such as T1 and T2 have significant limitations in detecting early IDD changes,^14–17 and commonly used MR contrast agents such as gadolinium (Gd) have also been shown to have limited clinical utility due to poor target specificity and low penetration efficiency within the disc.^18,19 The goal of this article was to develop and evaluate a novel fluorescent probe that specifically targets AF, the tissue that forms the outer layer of the disc. Fluorescence imaging is a relatively low-cost imaging method with high sensitivity and resolution. In addition, we incorporated a functional terminal carboxyl group in the probe design to allow for coupling with various signaling molecules for multimodal imaging applications. Fusing fluorescence imaging with other imaging modalities will combine the advantages of each imaging modality to improve diagnostic accuracy and better characterize the disease processes.²⁰ This will also allow translational imaging of IDD from the molecular level to clinical scales. Through imaging AF with a targeted probe, we will have the opportunity to visualize the molecular, cellular, and morphological changes that are associated with the early stage of IDD.

A recent study by Gibbs-Strauss et al. described the AF tissue-specific uptake of the fluorescent styryl pyridinium (FM) dyes in IVD.²¹ However the utility of these FM molecules is limited in IDD imaging due to the penetration limitation of the fluorescent signal, especially in the visible region, where FM dyes emit. Moreover, the lack of functional group on these FM molecules prevents them from being imaged by other imaging techniques. In this study, we developed a functional fluorescent IVD-targeted probe, HYK52, with a terminal carboxyl group that has the potential to couple with a wide range of signaling moieties, such as chelated radioisotopes (for positron emission tomography) and ${\mathrm{Gd}}^{3+}$ (for MRI). Multimodal imaging approach will address the penetration issue of fluorescence imaging in the visible region by introducing an additional imaging modality that has no penetration limitation. It also improves diagnostic accuracy and better characterizes disease processes through combining the advantages of each imaging modality. Therefore, such a tunable probe will have great potential in studying disc morphology by providing disc-targeted image contrast through noninvasive means.

2.

Methods and Materials

Unless otherwise stated, all chemical reagents were purchased from commercial sources and used without further purifications. Gadobenate dimeglumine (Gd-BOPTA, Multihance, Braco Diagnostics Inc., Princeton, New Jersey) was obtained from the vendor. Silica gel (240 to 400 mesh, Sorbtech) was used for column chromatography. NMR spectra were obtained from Bruker 400 MHz, and deuterated solvents were purchased from Cambridge Laboratory (Andover, Massachusetts). Mass spectrometry was performed using electrospray ionization mass spectrometry (ESI/MS; Waters 2998 photodiode array detector).

3.

Results and Discussion

Figure 1 shows the synthetic pathway of HYK52. The pyridinium salt, compound 1, was synthesized by a known procedure.^22,23 An aldol condensation reaction with $N-N^{\prime }$-dibutylaminobenzaldehyde in the presence of piperidine yielded HYK52 as a dark red powder with $>95\%$ yield. The structures of synthesized compounds were characterized by NMR and ESI/MS. The maximum absorption and emission of HYK52 occur at 490 and 613 nm (${\lambda }_{\mathrm{ext}}=495\mathrm{nm}$ in methanol), respectively, with extinction coefficient ($\epsilon$) of $\mathrm{12,863}{\mathrm{M}}^{-1}{\mathrm{cm}}^{-1}$ (Fig. 2). In water, the absorption and emission peaks were blue-shifted by roughly 10 nm, and the extinction coefficient decreased 2.4-fold ($\epsilon =5360{\mathrm{M}}^{-1}{\mathrm{cm}}^{-1}$).

Fig. 1

Synthesis of functional AF-targeted probe, HYK52.

Fig. 2

Absorption (a) and emission (b) spectra of HYK52 in methanol (blue) and water (red) (${\lambda }_{\mathrm{ext}}=495\mathrm{nm}$).

To test whether HYK52 has cytotoxic effects on disc cells, AF and NP cells were exposed to synthesized HYK52 and cell viability was measured using the XTT assay. Disc cells isolated from New Zealand white rabbits were incubated with two concentrations of HYK52 (50 and $100\mu \mathrm{g}/\mathrm{mL}$ or 108 and 217 μM) for 24 h (Fig. 3). Gadolinium (Gd-BOPTA) was included in the cytotoxicity assay as it is the modern standard for intravenous contrast in MRI. The concentrations of 50 and $100\mu \mathrm{g}/\mathrm{mL}$ (75 and 150 μM) were chosen because the peak plasma concentration of Gd-BOPTA after a standard intravenous dose was calculated at $100\mu \mathrm{g}/\mathrm{mL}$ from standard weight-based dosing protocols and average circulating blood volume. Thus, peak ($100\mu \mathrm{g}/\mathrm{mL}$) and half-strength ($50\mu \mathrm{g}/\mathrm{mL}$) doses were used in the cytotoxicity assay. Cells treated with 1 mM NaOH solution to induce complete cell death were included as a positive control. Compared to both untreated cells and clinically approved Gd-BOPTA, HYK52 showed no significant cell cytotoxicity at the experimental concentrations. Similar results were observed for cells exposed for 48 and 72 h (data not shown). As expected, cells treated with NaOH showed a significantly higher rate of cytotoxicity in both AF and NP cells at all time points.

Fig. 3

HYK52 has minimal cytotoxicity on disc cells. Annulus fibrosus (AF) (a) and nucleus pulposus (NP) (b) cells were incubated with 0, 108, and 217 μM HYK52 for 24 h and cell viability was measured using the XTT assay. Clinically approved gadobenate dimeglumine (Gd-BOPTA) was used for comparison. Error bars represent s.d. from triplicates.

Disc cells were incubated with the fluorescent HYK52 to evaluate cellular uptake. Time- and concentration-dependent cellular localizations were studied in both AF and NP cells (Figs. 4 and 5). Greater fluorescence was seen in AF cells compared to NP cells after 24 h of incubation, suggesting more efficient HYK52 uptake by AF cells (Fig. 4). After 48 h, NP disc cells also showed significant fluorescence signal. Cells treated with the higher concentration of HYK52 (122 μM, $56\mu \mathrm{g}/\mathrm{mL}$) showed more intense fluorescence signal than those treated with the lower concentration (61 μM, $28\mu \mathrm{g}/\mathrm{mL}$) (Fig. 5). The small lipophilic HYK52 dye was localized primarily to the cytosol of both AF and NP cells. It is possible that HYK52 was internalized through passive diffusion; however, elucidation of the exact mechanism is beyond the scope of this study.

Fig. 4

Time-dependent HYK52 cellular uptake. NP [(a), (b), (c), and (d)] and AF [(e), (f), (g), and (h)] cells were incubated with 122 μM HYK52 for 24 and 48 h. Bright field [(a), (b), (e), and (f)] and fluorescence [(c), (d), (g), and (h)] images were taken after fixation. Scale bar, 20 μm (magnification $\times 400$).

Fig. 5

Concentration-dependent HYK52 cellular uptake. NP [(a), (b), (c), and (d)] and AF [(e), (f), (g), and (h)] cells were incubated with 61 and 122 μM HYK52 for 48 h. Bright field [(a), (b), (e), and (f)] and fluorescence [(c), (d), (g), and (h)] images were taken after fixation. Scale bar, 20 μm (magnification $\times 400$).

Rabbit IVDs soaked in HYK52 showed strong fluorescence signals in the AF region of the IVD (Fig. 6). Three different incubation time periods, including 24, 48, and 72 h, were used for this study and the imaging results were similar. The data from 24-h incubation time were selected (Fig. 6). HYK52 binding in the AF highlighted the lamellar striations of the fibrocartilage structure, and stronger fluorescence was observed toward the disc’s periphery than at the inner annulus. The relatively weak fluorescence pattern observed in the NP region accurately reflects its less organized structure.

Fig. 6

Rabbit disc tissue microscopy images. The dissected rabbit discs were soaked in media containing HYK52 (122 μM) and incubated for 24 h. After washing and fixation, frozen tissue was cryosectioned in 10 μm thickness for imaging. Bright field (a) and fluorescence (b) images at $2\times$ magnification.

4.

Summary

In summary, HYK52 was synthesized with high yield and purity via a two-step chemical reaction. Cell viability test indicated negligible cytotoxicity of HYK52 in tested concentrations compared to the clinically approved MRI contrast agent, Gd-BOPTA. Disc cells readily uptake HYK52, with AF cells appearing to be more efficient than NP cells in this process. This is evident by the higher fluorescence signal in AF cells compared to that in NP cells after 24 h of incubation. Fluorescence imaging of rabbit disc specimen demonstrated the high localization of HYK52, which highlighted the lamellar striations of the fibrocartilage structure. These results indicate that HYK52 is biocompatible and has great potential in IVD-targeted imaging. Although HYK52 is not suitable for in vivo imaging due to the unfavorable emission wavelength in the visible region, introduction of additional imaging moieties that do not have penetration limitation can be readily achieved through bioconjugation with the carboxyl group on HYK52. Synthesis and in vivo imaging of HYK52-based multimodal imaging probes is currently ongoing.