Nanoparticles for Imaging Diagnostics
Although radiography is still the first and most widely used imaging method for assessment of a patient with a suspected or known diagnosis of OA, precise measurement of articular structures is not possible by X-ray. Computed tomography (CT), which is another radiogrphic technique, can also only indirectly assess cartilage degeneration by 3D observation of joint space narrowing, which is a feature of late-stage OA [
82,
83]. While it is possible to use a contrast agent in CT to directly visualize articular cartilage in 3D in vivo [
84], this approach is invasive as it requires an injection of the contrast agent directly into the joint space.
More sensitive non-invasive imagining techniques, for example, ultrasound and magnetic resonance imaging (MRI), allow visualization of changes in the cartilage volume or thickness. In clinical medicine, the most commonly used non-invasive medical imaging technique to visualize the structural changes associated with functional changes in tissues is MRI [
85‐
87]. Moreover, by developing new contrast agents, MRI can be applied to permit the more accurate visualization of structural tissue changes. For example, due to high biocompatibility and low toxicity, iron-based magnetic nanoparticles (MNPs) have been developed for clinical oncology imaging as novel biomarker-specific agents [
88]. Furthermore, MNPs could also be used as multifunctional agents because it is possible to combine diagnostic and therapeutic properties into them [
89]. For example, superparamagnetic iron oxide nanoparticles (SPIONs) exhibit a high degree of saturation magnetization, which is lost in the absence of magnetic field. These NPs are considered relatively less toxic than optical agents. For active targeting, SPIONs can be embedded into polymer cores and conjugated with different peptides, antibodies, or small molecules. In this way, modified SPIONs could serve as agents for the detection of OA biomarkers in synovial fluid [
90].
Iron oxide NPs have also been proposed for use in apoptosis detection in transplanted stem cells in arthritic joints. A significant problem for the long-term success of the matrix-associated stem cell implants (MASI) or chondrocyte implants (MACI) is the loss of cells after implantation due to cell migration, necrosis, and apoptosis [
91,
92]. Clinically, the success of implantation is usually assessed only a few weeks following the cell implantation by invasive arthroscopy and biopsy [
93‐
95]. Therefore, early detection and visualization of cell implants is an essential aspect in the development of similar strategies for OA treatments. In vitro studies have shown that iron oxide particles directly accumulates in the cytoplasm of viable cells, while ferumoxides are dispersed in large amounts in smaller cellular fragments following apoptotis. In addition, dispersed iron oxides showed a stronger T2 signal in apoptotic cell fragments compared to iron oxide clusters in viable cells [
96].
Non-iron-based NPs, for instance, gadolinium (Gd)-based contrast agent and caspase-3-sensitive nano aggregation MRI probe (C-SNAM), may also be successfully used for MACI and MASI implantation assessment. C-SNAM is a small molecular probe that can be easily delivered to MASI in cartilage defects by injection and passive diffusion. Induction-spectral plasma (ICP-MS) analysis showed significantly higher levels of Gd concentrations in apoptotic adipose-derived stem cells (ASCs) than in viable ASCs. Moreover bioluminescence imaging studies confirmed apoptosis of mitomycin C-exposed cells. In vivo studies demonstrated that strong bioluminescent signals were detected in all ASCs implants immediately after implantation. In addition, C-SNAM-exposed apoptotic cells also showed significantly stronger T1 signals on MRI than viable cells [
97]. However, it should be noted that many visualization methods are still in the experimental phase and are not yet standardized sufficiently for use in classification during daily clinical practice [
97]. However, it should be noted that many visualization methods are still in the experimental phase and are not yet standardized sufficiently for use in classification during daily clinical practice [
98].
A novel and innovative alternative to MRI for detection of cartilage degeneration is photoacoustic imaging (PAI). PAI is a hybrid bioimaging technology that combines the benefits of ultrasound with deep penetration into tissues and optical imaging with high spatial resolution [
99,
100]. Although the optical image has limited use for detection of OA, PAI can visualize neovascularity in arthritic joints, as well as provide morphological information on the degeneration of finger joints [
101,
102]. However, contrast agents used currently for PAI lack the sensitivity and specificity for detecting cartilage lesions in the early stages of the disease. NP-based contrast agents such as cationic poly-L-lysine-enveloped water-soluble anionic melanin nanoparticles (PLL–MNPs), which have desirable properties of biocompatibility and low cytotoxicity, could be used to improve the sensitivity of PAI [
103‐
105]. Research shows that PLL–MNPs can enhance PAI with significantly different accumulation in OA compared to healthy joints [
106]. Interestingly, comparative analysis of results from PAI, radiography, and MRI show that PAI based on PLL–MNPs could be more useful for the detection of cartilage degeneration than standard joint examination methods. Moreover, histological results are consistent with PAI and confirmed the feasibility of PAI using PLL–MNPs to detect cartilage degeneration in early-stage disease [
106].
One of the obstacles in evaluating pharmacological intervention involves the definition, identification, and quantification of early OA, as well as following up the efficacy of the applied therapies. Therefore, early detection and accurate visualization of cartilage degeneration and other synovial joint alterations are crucial for the appropriate treatment of OA.
Nanomaterial-Based Biosensors and Biomarkers for Early Diagnosis
The current clinical diagnosis of OA is based traditionally on clinical symptoms (e.g., pain and loss of function) and radiographic criteria (e.g., joint space width), which often occur late in the disease course. An attractive and practical alternative could be the measurement of biochemical markers, which can reflect dynamic events such ECM synthesis and degradation. Biomarkers may be measured in body fluids such as synovial fluid, blood (serum or plasma), and urine [
107]. The discovery of a definitive biomarker and its utilization in clinical practice could help to diagnose the disease much earlier and, importantly, to distinguish between phenotypes [
108], the fast and slow progressive forms of the disease, and the erosive and non-erosive forms of OA [
109,
110]. For these reasons, the main proteins of the ECM of cartilage, the metabolic products, and inflammatory mediators are widely examined. Many of these biomarkers are associated with the metabolism of collagen type II or aggrecan in cartilage or of collagen type I in subchondral bone [
111‐
113].
Other biomarkers that could reflect an initial change of ECM structure are related to a range of non-collagenous matrix proteins, including glycoproteins and proteoglycans, as well as matrix-degrading metalloproteinases. These molecules are constituents of both cartilage and synovium, and they have a role in other metabolic pathways in the joint.
Technological methods that have been developed for OA biomarker detection include enzyme-linked immunosorbent assay (ELISA), real-time polymerase chain reaction (RT-PCR), quartz crystal microbalance, mass spectrometry, and electrochemical methods. Some of these methods offer high sensitivity or selectivity; however, they have essential disadvantages not only because they are time-consuming and costly, but also they have reduced precision. For this reason, methods with high sensitivity and selectivity, as well as small sample volume requirements, are under investigation for their potential to detect OA in its early stages. These include nanoparticle-based devices or label-free and real-time biosensors for specific detection of OA biomarkers such as glycosaminoglycans (GAGs) released from degrading cartilage, HA, cytokines, free radicals (NO), and proteinases.
Loss of aggregating proteoglycans and constituent GAGs, which are essential contributors to the structure and biomechanical properties of articular cartilage, usually occurs before significant morphological changes [
114].
HA is a linear polysaccharide that is commonly found in synovial fluid, as well as in the skin and other tissues and organs. Its molecular mass ranges from 10
5 to 10
7 Da. This corresponds to 250–25,000 units of disaccharides [
115]. High-molecular-weight HA (> 1000 kDa) displays immunosuppressive properties [
116]. On the other hand, low-molecular-weight HA (typically < 500 kDa) has pro-inflammatory effects and can stimulate the production and secretion of inflammatory cytokines [
117]. Therefore, the size distribution and abundance of HA is considered a promising biological indicator of pathophysiology and could be assessed as a target for disease-specific diagnostics [
98,
118,
119].
Currently, to determine the molecular mass of HA or similar molecules, ELISA, size exclusion chromatography (SEC), and multi-angle laser light scattering (MALLS) methods can be used. However, they all have significant limitations, such as limited ability to distinguish molecular weights of molecules (ELISA), limited practical constraints on the number of fractions and samples that can be tested (SEC) or limited accuracy, and are relatively insensitive for low molecular weight fragments (MALLS). For these reasons, a label-free solid-state nanopore sensor was developed using a nanometer-scale aperture formed in a thin membrane as the only fluid connection between two reservoirs of an electrolyte solution [
120]. After analyzing the translocation properties of the molecules, the molecular weight of HA or other molecules could be determined on a per molecule basis, and finally, overall size distribution is obtained from only a few hundred events. For example, SS-nanopore-based detection method has been tested in the studies of an equine model of OA [
120]. Results confirmed that analysis of translocation properties using this method is sufficient for determining the size distribution and physiological concentration of HA in biological fluids and could be a good alternative for assessment of another OA significant molecular biomarkers.
Proteolytic enzymes, such as MMP-3, MMP-13, or ADAMTS, could be other essential biomarkers for the early diagnosis of OA. However, for determination of the activity of MMPs and ADAMTS, anti-neoepitope antibodies that recognize relatively large substrates are usually used. In some cases, new anti-neoepitope antibodies lack specificity for a particular enzyme because other proteases can also produce similar neo-peptides [
121,
122]. Therefore, the use of non-antibody-based methods could be more appropriate for the detection of ADAMTS or other MMPs than current techniques. For instance, a new fluorescent probe (probe ADAMTS-4-D-Au) based on AuNPs to determine the activity of ADAMTS-4 have been developed. The method employs fluorescein isothiocyanate (FITC) linked to the N-terminal of the ADAMTS-4 specific peptide DVQEFRGVTAVIR (Asp-Val-Gln-Glu-Phe-Arg-Gly-Val-Thr-Ala-Val-Ile-Arg) with the FITC-peptide conjugated to AuNPs with a diameter of 7 nm through cysteine by a gold-thiol bond [
123]. This probe is stable under physiological conditions, and fluorescence intensity is proportional to the concentration of active ADAMTS-4. When this ADAMTS-4-D-Au probe was used to determine ADAMTS-4 activity in human synovial fluid, the strongest fluorescence signals were detected in patients with acute joint injury and patients with late-stage OA [
123].
Moreover, the high activity of ADAMTS-4 was consistent with the results obtained from the arthroscopy analysis, while the MRI results were different. For instance, for a patient from the group with acute joint damage, arthroscopy revealed second-degree cartilage damage; however, T1-weighted and T2-weighted MRI did not specify any difference related to cartilage damage, while fluorescence intensity in this patient was comparatively high. This probe can be used to identify ADAMTS as a potential biomarker associated with cartilage damage at an early stage of the disease [
123].
Alternative non-antibody-based method for non-invasive, real-time evaluation of OA has been developed for the monitoring of the nitric oxide (NO) release in OA chondrocytes, which overexpress the gene encoding inducible NO synthase (NOS2) and its product, NO [
124,
125]. NO has been considered as a biomarker for OA [
126]. The nanosensors were synthesized by encapsulating the NO-sensing molecules (4-amino-5-methylamino-2′,7′-difluorofluorescein diaminofluorescein-FM (DAF-FM)) within the biodegradable poly-(lactic-co-glycolic acid) NPs. In vitro studies have demonstrated that there is a positive correlation between the increase in the fluorescence intensity and the change in NO concentration in the chondrocytes. The efficacy of this approach was tested in the rat model of OA due to anterior cruciate ligament transection (ACLT). After ACLT surgery, the level of NO in the joint fluid increased with cartilage degeneration and was positively correlated with increased NO nanosensor fluorescence [
127].
Another promising technology for OA diagnostics with improved sensitivity and analysis time is based on the chemical properties of gold nanoparticles (AuNPs), which are being extensively studied for the development of new multimodal contrast elements or biosensors. Gold nanoparticle biosensing involves the interaction between a target biomarker molecule and a AuNP crosslinker or a AuNP-containing antibody [
128].
For biosensor applications, AuNPs are attractive for their chemical stability and convenient spectral window in the visible range. The AuNP-based biosensors are designed to be selective for the detection of various biomolecules, including small molecules, peptides, and nucleic acids.
Particle plasmon resonance (PPR) or localized surface plasmon resonance (LSPR) methods are used widely for the determination of chemical and biochemical species, because of the sensitivity of their electron-rich surfaces to the surrounding environment [
129‐
131]. For example, a fiber-optic particle plasmon resonance (FOPPR)-sensing platform, based on AuNP-modified optical fiber for the detection of the OA related pro-inflammatory cytokine interleukin (IL)-1β in synovial fluid samples, has been developed [
132]. In this FOPPR sensing platform, the molecular binding of IL-1β on the AuNP-conjugated anti-IL-1β transduces a local increase in the refractive index of the medium surrounding the AuNP, enhancing the plasmon absorbance of the AuNP. Linear regression analysis showed a good correlation coefficient for both ELISA and modified FOPPR detection methods. However, using a FOPPR sensor, the analysis time for detecting IL-1β in synovial fluid was significantly shorter than with ELISA, thereby reducing the chance of potential experimental errors. Similar results were obtained using FOPPR sensor for detection of TNF-α and MMP-3, achieving an excellent refractive index resolution (5.18 × 10–7 RIU) with limits of detection as low as 0.48 pM and 1.56 pM for TNF-α and MMP-3, respectively. This suggests that the label-free and real-time detection capabilities of the FOPPR sensor for protein analysis could be an excellent alternative to immunoassay [
133].