Fourier-transform infrared spectroscopic imaging of articular cartilage and biomaterials: A review

Abstract

Fourier transform infrared spectroscopy (FTIR) has the potential to mark up the chemical changes of the materials, as almost all the materials contain their signatures in infrared region. Spectroscopy combined with spatial resolution enables the possibility of characterizing samples up to microscopic level. The emerging development of instrumentation to provide spatial information for infrared (IR) spectroscopy, termed as IR microscopy, provides an opening for newer applications in terms of image analysis, novel data processing tools, etc. Characterization of biomaterials using IR spectroscopy has a trace back to 1950s. The advent of FTIR with imaging capability made characterization possible in cartilage tissue and other biological systems. Extensive analysis of chemical constituents of cartilage and tendon, collagen orientation and polarization property of cartilage using FTIR imaging (FTIRI) has been actively explored during the last two decades. Also, studies using specialized instrumentations like synchrotron FTIR imaging have been attempted to understand the characteristics of biological samples like cartilage. This review covers most of those investigations on cartilage with FTIRI to characterize the same in terms of component characteristics and quantification, collagen orientation, zonal boundary determination, influence of mechanical compression on tissue nature and its correlation to other techniques in last 20 years.

1. INTRODUCTION

Articular cartilage is a thin layer of connective tissue that covers the load-bearing ends of bones in joints (Fig. 1). It is highly specialized in its molecular and histological structures and provides a smooth surface for easy motion of the joints and prevents the joints from shock impact. The molecular composition of cartilage consists primarily of water, collagen and proteoglycans (PG) [1–3]. The morphological structure across the entire depth of articular cartilage is unique [4–9]. It is composed of three sub-tissue zones based on the local orientation of collagen fibrils. These three histological zones are (1) the superficial zone (SZ) where collagen is oriented parallel to the articular surface, (2) the transitional zone (TZ) where the collagen is oriented rather randomly, and (3) the radial zone (RZ) where the collagen is oriented perpendicularly to the articular surface (Fig. 2).

Fig. 1.

A schematic of the articular capsule of the knee joint. (The image was from the Wikipedia Commons, http://en.wikipedia.org/wiki/File:Gray350.png.)

Fig. 2.

A schematic of articular cartilage, showing the depth-dependent changes of the collagen fibers (the short lines) and the chondrocytes (the circles and ovals) in different histological zones (not drawn to scale).

In cartilage degeneration studies, it is essential to understand the kinetics of swelling and deformational behavior as well as change in morphology and molecular compositions. The depth-dependent structure of the collagen matrix in articular cartilage is important in resisting the swelling pressure created by the negatively charged PG, thereby preserving the tissue’s integrity. Disruption of the fibrillar network has been linked to the early stages of osteoarthritis (OA) and may represent a functional failure of the tissue [10–13]. Studies of articular cartilage and its degradation are motivated by the critical role of cartilage in the development of osteoarthritis [14, 15]. An accurate diagnosis of early OA, however, remains elusive in clinical practice because many changes in tissue’s fine structure and delicate chemical/molecular composition precede significantly prior to the development of OA as a clinical disease. Therefore, every sensitive technique for detecting the early changes in cartilage leading to OA would be valuable for monitoring disease progression and evaluating the efficacy of treatment. The depth-dependent anisotropy of the collagen fibrillar structure in connective tissues becomes evident when examined by several imaging techniques including microscopic MRI (μMRI) [16, 17], polarized light microscopy (PLM) [18–20] and Fourier Transform Infrared Imaging (FTIRI) [21–24].

Magnetic resonance imaging (MRI) has the potential to directly evaluate the morphology and cartilage wear in injuries, due to its excellent soft tissue contrast [25]. Since the values of T2 relaxation in MRI of tendon and cartilage is known to be sensitive to collagen structure and fibril orientation relative to the static magnetic field B₀ [26–29], T2 has been used as a molecular-level parameter to assess the structural integrity of the tissue [10, 30]. The use of T2 to quantitatively monitor the collagen distribution in cartilage requires the rotation of the tissue block in B₀ and the analysis of the anisotropy of T2 at different orientations. Using MRI, attempts have been made to detect structural damage in the joint and to evaluate the patterns of compartmental involvement in knees with pain, between men and women, and in different age and body mass index (BMI) categories [31]. Correlation of macromolecular structure of cartilage studied using quantitative MRI measurements including T1, T2 and T1ρ relaxometry, diffusion weighted imaging and magnetization transfer provides quantitative parametric MRI of articular cartilage [32]. Also, to detect the early onset of OA in articular cartilage (AC), T1ρ technique has been used in MRI [33]. In addition to multicomponent analysis [34], the depth-dependent profile of glycosaminoglycans (GAGs) in articular cartilage has been investigated by μMRI and correlated to histochemistry [35].

Optical transmission principle is the basis of PLM technique. It provides information based on the birefringence of the sample. The optical retardance/birefringence is proportional to the thickness of the sample. Collagen orientation in cartilage is determined from the birefringence information. PLM is considered as the golden standard for determining the collagen orientation in cartilage [36]. PLM was among the first techniques used to establish the curved architecture of collagen fibres in adult cartilage and remains one of the most commonly used techniques for the investigation of cartilage microstructure [37].

With the potential to provide quantitative and spatially-resolved information about the chemical composition of degrading tissue, FTIRI has been used extensively in cartilage studies in recent years. This technique has been applied to examine the feasibility of measuring changes in the composition and distribution of collagen and proteoglycan macromolecules in human osteoarthritic cartilage [38]. Also, FTIRI is extensively used to investigate the amide anisotropies at different surfaces of a three-dimensional cartilage block. With the change in the polarization state of the incident infrared light, the resulting anisotropic behavior of the tissue structure to its full depth has been investigated to throw additional light on collagen orientation, especially in the superficial zone [39]. The induced osteoarthritic cartilage changes in a rabbit model via ligament transection and medial meniscectomy and monitored disease progression has also been studied using infrared microscopic imaging [40]. Attempts have been made to correlate different techniques to understand cartilage tissue characteristics [41–43]. To understand and to correlate the key parameters of MRI, PLM and FTIRI to evaluate the healthiness of cartilage tissue, studies have been initiated. In septal cartilage (a substitute for articular cartilage), the existence of anisotropy in collagen orientation is detected by the above said techniques. The findings may have important clinical implications where collagen misalignment in cartilage grafts may hinder the long-term success of the transplant [44]. The depth dependent anisotropic information in articular cartilage using μMRI, PLM and FTIRI indicates that neither of the imaging technique has the resolution to identify individual collagen fibrils, but their anisotropy maps reveal the angular-dependent variations in the tissue. Hence, it is feasible to correlate the infrared absorbance with PLM using the same tissue section and to correlate PLM with μMRI [45]. This may help in the better understanding of the earliest degradations in articular cartilage before it becomes a clinical disease.

The rest of this review is divided into four major sections. In Section 2, the molecular constituents of articular cartilage is briefly introduced, which provides a background for better understanding of some of the experimental work described in this review. In Section 3, the general applications of infrared microscopic imaging are discussed, from the basic physics to the use of Fourier-transform infrared imaging in material research. In Section 4, the studies of articular cartilage by FTIRI are discussed in details, with ample examples from many published works. This review paper is concluded in Section 5, which anticipates that the applications of some novel procedures and approaches in FTIRI could further extend the usefulness of this relatively new imaging technique.

2. Constituents of articular cartilage

Articular cartilage plays an important role in proper joint functioning. It has a unique load bearing capacity (up to 10 times of body weight) and very low frictional co-efficient [46, 47]. It is continually synthesized and repaired to replace damage from mechanical wear, and is able to last for up to seven or eight decades [48]. The major constituents of this matrix are fibril forming collagen type II (for tensile strength), large aggregating proteoglycans (for its load-bearing properties) and water. Collagen fibers and aggregating proteoglycans form a network together with non-collagenous proteins responsible for the functional properties of the tissue [49]. As mentioned earlier, articular cartilage is structurally divided into superficial, transitional and radial zones along with calcified zone [50]. The collagen orientation is parallel to the surface in superficial zone and changing to perpendicular direction in radial zone. Hence, in the transitional zone, the fiber orientation is almost random. The collagen as well as water content is higher in superficial zone, whereas in radial zone the collagen content is lesser as compared to superficial zone. The proteoglycan content shows a reverse trend over the tissue thickness [51]. The bottom layer is mineralized and separated from the non-mineralized articular cartilage by a distinct thin layer of enhanced calcification-known as the tidemark [52]. As a result of the slow, but progressive advance of the ossification front (aging), the cartilage gets thinned. Also, due to mechanical wear, the superficial layer may damage gradually. The physical and mechanical integrity of cartilage decreases with age resulting in pain and development of OA. Due to the importance of cartilage for our mobility, there is considerable research into the biology and mechanical characteristics of articular cartilage [53].

3. Infrared microscopic imaging

3.1. Theory of spectroscopy and FTIR

In the electromagnetic spectrum, infrared (IR) region span from 0.7–1000 μm (14000 to 10 cm⁻¹). The region 4000–400 cm⁻¹ mostly used for chemical analysis corresponds to changes in molecular vibrational energy states. Above absolute zero temperature (0K), vibrational transition between energy states happen in all molecules. This transition is governed by degree of freedom of molecules involved. For a molecule with N number of atoms, there will be 3N degrees of freedom corresponding to the Cartesian coordinates of atoms involved. For a non-linear molecule, there are 3 rotational and 3 translational degrees resulting into 3N-6 vibrational degrees. For a linear molecule, the rotational degrees are reduced into 2 and hence the available fundamental vibrations are 3N-5. Transitions between the vibrational energy levels of atoms/molecules are recorded by infrared spectroscopy. Molecules absorb IR light when their vibrational frequency is equal to the frequency of incident IR radiation. This fundamental frequency of transition is of the order 0.1 eV and is dependent on the bond strength and bond angle. IR spectroscopy is sensitive to the changes in the dipole moment of the molecules, and has been used to obtain structural information of molecules.

For a vibration to be IR active, the vibration must result in a change of dipole moment during the vibration. Hence, homonuclear molecules like Hydrogen, Nitrogen and Oxygen are infrared inactive. IR vibrations are broadly classified into two categories named stretching and bending in case of a polyatomic molecule. Stretching involves change in the inter-nuclear distance between two nuclei, whereas bending involves change in the angle between two bonds. The bending vibrations can occur in four ways: scissoring, rocking, wagging, and twisting. Also, coupling of vibrations can occur if the vibrations involve bonds to a single atom. The result of coupling is a change in the characteristics of the vibrations involved. The stretching of a diatomic molecule can be understood by an analogy of a spring with two metallic balls. By applying Hook’s law, the two atoms and the connecting bond are treated as a simple harmonic oscillator composed of two masses (m₁ and m₂): the frequency of the vibration of the spring is related to the mass and the force constant of the spring, k, by the following formula:

$$v=\frac{1}{2\pi }\sqrt{\frac{k}{m}}$$

where k is the force constant, m is a reduced mass, v is the frequency of the vibration.

The disadvantage of grating based spectrometer is its monochromatic scanning. As every wavelength is scanned one after the other, it takes a lot of time in acquiring a spectrum. In addition, during scanning if any error occurs, baseline correction becomes an issue to be addressed. To overcome this problem, an instantaneous scanning method to cover the complete wavelength region has been developed, which is termed as Fourier Transform (FT) Spectroscopy. In FT based technique, the background noise is quite reduced and the optical throughput (Signal-to-noise ratio) has been increased more than 10 times as compared to dispersive technique. FT is applicable not only in infrared, but also in magnetic resonance spectroscopy. In principle, the time domain data can be converted into the frequency domain data using FT in a very short period, which is an advent for most of the present day biological as well as dynamic spectroscopic studies. In FT based spectrometers, Michelson Interferometer is the key component to analyze composite frequency spectrum. The principle of Michelson interferometer is shown in Fig. 3. Using a beam splitter, a varying path difference of the light is produced. The recombination of the path difference introduced beams provides an interferogram. This is later converted by the FT process into an IR spectrum of the sample under observation. Special techniques like Attenuated Total Reflectance (ATR), fiber optic probe and diffuse reflectance are also being used with FTIR to characterize materials. In addition, the physics principles like polarization, dichroism, transition moment determination, ellipsometry and 2D correlation spectroscopy are also extensively used to extract the characteristics of the material of interest to a greater extent.

Fig. 3.

The configuration of the Michelson interferometer, where the re-combination of the two light beams causes an interference pattern.

3.2. Infrared imaging and instrumentation

The advantage of FTIR extends the utility of infrared spectroscopy from material development to tissue characterization. But, the spatial information was absent till the imaging concept has been introduced. Today, IR microscopy is employed generally for chemical identification, which is used to complement traditional light microscopy. Contrary to the images integrated over entire wavelength region, IR microscopic images are spectral images that contain spectral response of the sample over individual wavelength or narrow wave-band. It reveals the finer details of morphology of a given sample. IR imaging was developed primarily for military applications and secondarily for astronomical as well as remote sensing applications. Later the same was extended to laboratory applications. The spatial (X, Y) as well as spectral (Z) content enriched data cubes make this technique versatile for inter-disciplinary applications. Though the concept of coupling of an infrared (IR) spectrometer to a microscope that provides widespread use in a variety of research and industrial applications was suggested by Barer [54] in 1945, significant improvement in addressing spatial resolution for IR imaging started in 1980s. The description of figure of merit of infrared microspectrometer is available in 1951 [55]. Analysis of sample characteristics over a limited area using infrared microscopy (micro-spectroscopy) can be achieved in two ways:

1. Mapping and 2. Imaging. Sequential spectral measurement of adjacent locations of a sample is called mapping. This requires a single detector element to record data at a point and to move to the other till covering the entire sampling area. In imaging, an array of detector elements called pixels in a CCD that covers the desired sample area to record spectral data at one shot. The development of IR multichannel detector array is the key reason for that [56]. Also, a hybrid technique combining both mapping and imaging exists using a linear array instead of 2D array detector. In this case, spectra are recorded simultaneously over a line of points on sample. Sequentially obtained such line maps are stitched to form complete image of the sample [57]. The approach of infrared imaging and the limitations of instrumentation have been explored by various researchers [58–61] including imaging with the Focal-plane array (FPA) detectors [62].

A detailed case study has been carried out to demonstrate the value of imaging and mapping in specific situations. Two dynamic systems were studied using both the techniques and the differences in experiment and instrument design philosophy, instrumentation, data collection and processing for the two FT-IR microspectroscopic techniques are outlined clearly. It is found that for a static sample, mapping offers a convenient, fast, and cheaper route for analysis. However, for small morphological features that may change in minutes require the potentiality of imaging [59]. The versatility of IR imaging in remote sensing, astronomy, chemistry and medicine to highlight the interdisciplinary nature of this field was explored. The capability of IR imaging was detailed from planetary to cellular system to introduce the basic concepts of IR spectral imaging [60]. The enhanced capability of recently developed FTIR spectroscopic imaging to examine the spatial distribution of vibrational spectroscopic signatures of materials spanning from physical to biomedical disciplines on a microscopic scale was investigated extensively [61]. A new approach to infrared chemical imaging that combines the multiplexing power of interferometry performed in a step-scan mode with the multichannel characteristics of an indium antimonide focal plane array detector has been demonstrated [62] to explore the better utility of the same in comparison with Raman imaging. Also, the advantage of source as well as image modulations in IR imaging was discussed.

Attempt has been made to combine microscope with synchrotron source [63, 64] as well as IR diode laser [65] to achieve higher spatial resolution. Though efforts to develop near filed approach seem promising [66, 67], for laboratory environment it remains slow and uncommon. To increase throughput for FTIR imaging studies, attenuated total reflection (ATR) technique has been explored [68, 69]. Increase in spatial resolution of an IR imaging system is at a cost of its temporal resolution. Hence, attempt has been made to introduce the concept and instrumental details of a time-resolved infrared spectroscopic imaging modality that permits the examination of repetitive dynamic processes whose half-lives are of the order of milliseconds [70]. As the data collection using FTIR imager depends on the consistency of clock signals that drives both interferometer and detector FPA, asynchronous rapid-scan approach to FTIR imaging has been explored to have faster data acquisition with no compromise in data quality [71]. Explorations are still going on to understand the limits of infrared imaging using Michelson interferometer.

3.3. Infrared investigations of heterogeneous materials (Biomaterials)

Infrared spectroscopy is being used as an analytical tool to analyze the structural composition of materials, including those from biological origin. It is also used for quantitative analysis of structural networks. The spectroscopic techniques have been widely used for biological studies since 1960s. These studies are ranging from polymers, oriented purple membrane, polypeptides and proteins, fibrous protein structures, proteins in solution, human serum albumin, synthetic and biological apatites [72–81]. The versatility of infrared spectroscopy for biological/medical/clinical applications has been detailed extensively [82–87]. The presence of water limits the studies of structural elements in biological systems using infrared. The water absorption peaks around 3400 cm⁻¹ and 1600 cm⁻¹ restrict the analysis of vibration frequencies like C=O, N-H and C=N. Hence, to understand the limitation of water absorption peaks, infrared spectra of water peaks were analyzed [88]. Also, the influence of relative humidity in IR absorption measurements has been investigated. Despite the limitations, studies about atherosclerotic plaques in arteries have been done [89]. FTIR spectral features investigation in the diagnosis of esophageal cancer revealed specific structural changes and thus spectral criteria were established for the detection of malignancy in esophagus tissues [90]. Regardless of the biochemical explanation, proliferative status of mammalian cells determined from the RNA to lipid ratio of its IR spectra provide a valuable non-invasive tool to measure pathologies in vivo, such as cancer [91].

In cell research, using suspensions of viable cells in aqueous suspension reduces IR measurement artifacts and facilitates determining the concentration of the major biochemical components via a linear least squares fit of the component spectra to the spectrum of the cells [92]. The extensive exploration of infrared spectroscopy in biological application took a turn when infrared imaging became possible for microscopic studies. Studies have shown that FTIR imaging enables determination of the distribution of molecules of interest for tissue analysis without pre-analytical modification of the sample such as staining. Molecular structure information is also available from the same analysis. Thus, FTIR imaging has been proposed to analyze the metabolic and structural status of a cerebral tumor tissue [93]. Also, several cancer markers can be identified from FTIR tissue images, enabling accurate discrimination between healthy and tumor areas. Estrogen deficiency had the effect of retarding the fracture-healing process relative to the estrogen-sufficient animals. FTIR imaging is used to characterize the changes in fracture callus mineral content, carbonate content, mineral crystallinity and collagen maturity in femurs of 3-month-old ovariectomized rats treated with estrogen sufficiency or estrogen deficiency [94]. The structural changes of brain tissues were related to qualitative and quantitative changes in lipids and were correlated to the degree of myelination, an important factor in several neurodegenerative disorders. To identify the molecular origin that gave rise to the specific changes between healthy and diseased states, FTIR microspectroscopy is used [95]. Also, quantitative analysis on human breast cells for malignancy using cluster analysis and artificial neural networks (ANN) with FTIR microscopic imaging has been performed [96].

Similar research has been conducted on bio-polymers, collagen, etc., using IR spectroscopy/imaging. The transition moment of amide I and II has been investigated for α-helix and interpretation of various proteins has been revised. Also, dichroic ratio of amide bands to estimate the degree of orientation of fiber axes has been discussed [97]. In addition, infrared absorption spectra of poly-L-lysine, poly-L-glutamic acid, β-lactoglobulin, myoglobin, and αs-casein in the region of absorption of the amide I band have been observed in H₂O solution, D₂O solution as well as in the solid state. The results indicate that characteristic frequencies exhibited by specific conformations of the investigated synthetic polypeptides are not transferable to corresponding conformations of globular proteins [98]. In another study, infrared spectra of poly(d,l-alanine), poly(l-glutamic acid), poly(l-lysine), silk fibroin and tropomyosin have been registered for various conformations of the polypeptide chain. Assuming additivity of the main- and side-chain absorption, spectral parameters of amide I and II absorption bands corresponding to α-, β-, and random coil conformations have been derived. The amide I band parameters for H₂O and D₂O have been compared [99]. As a precursor to analyze collagen in cartilage, collagen model peptides were extensively investigated using infrared. Polypeptide Models of Collagen-Solution properties of (Pro-Gly-Phe)_n , amide I band of IR spectrum and structure of collagen and related polypeptides and complementary experimental and simulation approaches to conformation and unfolding of collagen model peptides are to name a few of such investigations using infrared techniques [100–102]. In addition, infrared spectroscopic ellipsometry, reflection absorption IR spectroscopy, IR transmission spectroscopy, and best-fit calculations are applied in a cooperative study to determine the anisotropic optical properties of a thin polyimide layer in the spectral range 4000–500 cm⁻¹ [103]. The list of infrared applications in biological/medical/clinical investigations thus becomes very long to summarize.

4. FTIR imaging (FTIRI) of articular cartilage

As discussed earlier, molecular signatures of biological samples can be obtained using FTIR imaging with fine spatial resolutions. This molecular mapping/imaging capability of FTIRI in a tissue has great importance in cartilage study. In a calcified Turkey tendon, the spatial variation of mineral phase has been studied qualitatively and quantitatively using infrared microscopy. Mineral-to-matrix ratio was used to understand the tissue [104]. Attempt was made to investigate synovial fluids from joints affected by rheumatoid arthritis, osteoarthritis, spondyloarthropathies, and meniscal injuries using FTIR spectroscopy. Significant differences in the composition of the fluid as a result of the disease processes were observed and quantitatively analyzed using multivariate technique. The results demonstrate the potential of infrared spectroscopy for the differential diagnosis of arthritis [105]. Investigations were carried out on bovine cartilage using infrared microscopy to understand the contribution of cartilage components to its infrared spectrum. The variation in absorption intensities of components in SZ, TZ and RZ of bovine cartilage was analyzed. Also, to identify spectral markers, synthetic samples of cartilage components were also studied [21].

Adding new dimension to cartilage analysis, the temperature dependency of absorption coefficient of water in cartilage and cornea was investigated. The water temperature was altered through IR free electron laser and the absorption spectra were analyzed and it was found that the absorption coefficient of cartilage decreases at temperatures higher than 50°C because of the thermal alterations of water–water and water–biopolymer interactions. The degree of absorption coefficient decrease for the 2.9 μm band was found to be much more than that for the 6.1 μm band [106]. As sampling plays a crucial role in infrared measurements, efforts have been made to understand the error estimation of measurements. Studies show that variation in cartilage cryosection thickness permits quantitative FT-IRI analysis only. Hence, an inexpensive reference sample method for quantitative analysis was explored. In this technique, normalization of the measured cartilage absorbance with respect to external reference has been proposed [107]. Since collagenase treatment of cartilage serves as an in vitro model of the pathological collagen degradation, collagenase-treated cartilage has been analyzed to elucidate the molecular origin of the spectral changes at the articular surface of human OA cartilage. It was found from the results that the spectral changes observed in the collagenase-treated cartilage and in OA cartilage arise from changes in collagen structure [108].

The abnormal organization of connective tissue and/or collagen network formation results in many muscular diseases; attempts have been made using FTIR spectroscopy to analyze the collagen network by differentiating the type of collagen [109]. The inability to identify early cartilage changes during the development of the disease, and the lack of techniques to evaluate the tissue response to therapeutic and tissue engineering interventions are the reasons for the complications in managing osteoarthritis. Out of various analytical tools available, FT-IRIS permits evaluation of early-stage matrix changes in the primary components of cartilage, collagen and proteoglycan. Studies reveal that the ratio of peak areas at 1338 cm⁻¹/amide II found to correspond to the histological Mankin grade serves as a marker to evaluate cartilage degeneration [110]. Surprisingly, a Chinese literature reveals that the ratio of organic (collagen) to inorganic substance (phosphate, carbonate etc.) of cartilage in youth is 5–7 times larger than that in elders, but there was a big difference in the content of lipids between youth and elders. Also, an important finding revealed is that less organic substance and more inorganic substance with increasing age is a symbol of the degeneration of cartilage [111]. The applications of FT-IR microscopy and imaging for analyses of bone and cartilage in healthy and diseased tissues and the application of these techniques for the characterization of tissue-engineered bone and cartilage discussed in detail in a review [112] paves path for cartilage characterization using IR microscopy.

To add a dimension to the tissue analysis, NIR spectroscopy was explored as a means to quantify tissue alterations in low-grade cartilage defects [113], especially low-grade cartilage defects in a preliminary clinical study [114]. In the light of earlier studies, a new attempt has been made to understand cartilage in angular space using a polarizer between the sample and detector (actually analyzer). The intensity responses of all the major species were analyzed for the analyzer angles from 0 to 180°. The cartilage anisotropy has been mapped for all the major chemical species-amide I, II, III and sugar, throughout the cartilage depth. In contrast to most previous studies that used cartilage explants (tissue separated from the underlying bone), here the specimens included the entire thickness of the articular cartilage and the interface between the cartilage and the underlying bone. Based on the analysis, a 2D anisotropy map was created for the measured chemical components and reported for the first time (Fig. 4) [24]. Also, a correlation of amide I anisotropy was established with PLM retardation data as shown in Fig 5. In continuation to the earlier anisotropic studies of cartilage, investigations were carried out to further explore the visualization of the anisotropy in different orientations of cartilage. Fig. 6 shows the different orientations in which tissues can be analyzed for their structural changes.

Fig. 4.

Anisotropy of amide and sugar components in articular cartilage in the angular space between 0°–180°. (Reprinted from Xia, Y, Ramakrishnan, N. and Bidthanapally, A. 2007, Osteoarthritis Cartilage, 15(7), 780–788 with permission from Elsevier).

Fig. 5.

Correlation between the depth-dependent FTIRI and PLM results in articular cartilage. (Reprinted from Xia, Y, Ramakrishnan, N. and Bidthanapally, A. 2007, Osteoarthritis Cartilage, 15(7), 780–788 with permission from Elsevier).

Fig. 6.

Various orientations of the cartilage tissue section from a block of articular cartilage. (Reprinted from Ramakrishnan, N., Xia, Y., and Bidthanapally, A. 2008, J. Orthop. Surg Res. 3(1), 48 with permission from BioMed Central).

An out of the box thinking of comparing a single structured tendon to a multi structured cartilage emerged out in understanding anisotropic changes with respect to the tissue sample orientation using FTIRI. The analysis of parallel sections revealed that for tendon, the anisotropy of amide I and amide II components in parallel sections is comparable to that of regular sections. For articular cartilage, parallel sections in the superficial zone have the expected infrared anisotropy that is consistent with that of regular sections, whereas the parallel sections in the radial zone, however, have a nearly isotropic amide II absorption and a distinct amide I anisotropy as shown in Fig. 7 [39]. Extending this idea, further studies have been done extensively to compare regular/perpendicular sections with parallel sections for the construction of anisotropic information, which will provide the structural similarity/change in 3D structure of the tissue. Using a curve fit equation the anisotropies of amide I and II were compared over the entire depth of cartilage for the first time that provides new information about the bond arrangement. Fig. 8 shows the vital information of this study that is required for future studies [115]. As spectroscopic analysis requires special skill, methods that are currently in use such as peak area analysis and other chemometric techniques to extract the information from the articular cartilage spectra were reviewed [116] and their pros and cons have been discussed for better understanding, to utilize this knowledge to differentiate intact and repaired articular cartilage [117].

Fig. 7.

Anisotropy of the parallel tissue sections in FTIRI. (Reprinted from Ramakrishnan, N., Xia, Y., and Bidthanapally, A. 2008, J. Orthop. Surg Res. 3(1), 48 with permission from BioMed Central).

Fig. 8.

Reconstruction of the tissue anisotropy between the parallel and perpendicular sections. (Reprint from Xia, Y., Mittelstaedt, D., Ramakrishnan, N., Szarko, M. and Bidthanapally, A. 2011, Microsc. Res. Tech., 74(2), 122–132, with permission from John Wiley and Sons).

4.1. Cartilage component characterization and quantification

The concept of composition changes in cartilage due to factors like aging, disease, etc., is always a concern. Hence, investigations in every possible direction to quantify the chemical components to bench mark the healthy and diseased cartilage is continuing. Proteoglycans were extracted from the articular cartilage of fetal, calf and adult bovine metacarpal-phalangeal joints with 4 M-guanidinium chloride to understand the Age-related changes in the chemical composition [118]. Soluble protein chondroitin keratin sulfate as proteoglycan (PG) aggregate was studied for its IR spectra to directly detect PG present in cartilage tissue and to discover the changes in its macromolecules in different states in the body [119]. Similarly, infrared spectra of collagen were acquired and analyzed for different temperatures and the spectral overlapping has been addressed through de-convolution technique [120]. With the advent of FTIR imaging, concentration measurements of biological fluids, cells and tissues were attempted [121]. To quantify the chemical components of cartilage, mixtures of model compounds with type II collagen and aggrecan with varying concentrations were studied using FTIRI. This spectral information was used to analyze the IR spectra of cartilage in different zones [21].

With an assumption that cartilage is predominantly chondroitin sulfate (CS) and type II collagen, quantitative analysis of CS has been attempted in native as well as engineered cartilage using FTIR imaging. The data analyses were carried out using two chemometric techniques, the Euclidean distance algorithm and a least-squares approach. The results have been discussed in detail for the two methods and the correlation found that the FTIR imaging technique overestimated the collagen content because it lacks specificity for different proteins. Hence, the lack of specificity of this technique for different types of proteins may be a limitation [23]. To understand the evaluation of native, repaired and engineered cartilage at the molecular level, the determination of PG concentration has been attempted for the first time in engineered cartilage using FTIR microscopic imaging. This result was correlated with the gold standard technique of histological determination of PG [122]. Samples from human cartilage were used to acquire spectral images from the superficial, intermediate, and deep layers for each sample. Results were analyzed using Euclidean distance mapping and quantitative partial least squares analysis (PLS) using reference spectra for type-II collagen and chondroitin 6-sulphate (CS6). FTIRI results were correlated to the histology-based Mankin scoring system. PLS analysis found relatively low-concentrations of collagen and proteoglycan in osteoarthritic cartilage [38].

Extending the line of chemometrics, principal component regression (PCR) method was explored to quantitatively determine collagen and proteoglycan concentrations in bovine nasal cartilage (BNC). Using model collagen and chondroitin 6-sulphate mixture at different ratio, the spectral library has been created. Using this library and PCR technique, relative concentrations of collagen and proteoglycan in BNC were estimated. These PCR-determined concentrations agreed with the molecular concentrations determined biochemically using an enzyme digestion assay. Use of the imaging approach revealed that proteoglycan loss in the specimens occurs first at the surface of the tissue block when compared with the middle portion of the tissue block [123]. Using polarization technique, the anisotropy of sugar peak (~1050 cm⁻¹) was investigated over the entire depth of cartilage. The absorption profile of the sugar band shows an anisotropic flipping at the deeper part in the radial zone, just above the tidemark. This anisotropy flipping of sugar might be caused by the orientational change in the collagen-attaching PG from orthogonal to parallel when the fibrils are entering the calcified zone.

Also, from the quantitative PG mapping throughout the cartilage depth, it was found that PG concentration in this region is reduced. The component anisotropy of collagen and sugar along with the concentration of PG in cartilage is shown in Fig. 9. The sugar band is seen to have an anisotropic flipping not near the surface but at the deepest part of RZ approaching the tidemark. Since the reduction of the chondroitin sulfate PG concentration and the increased amount of calcified material with the fibrils entering the calcified zone, this infrared anisotropy of sugar was attributed to the orientational changes in PG. This observation will be helpful to further understand the related structural and functional properties of connective tissues by the quantitative concentration calculation and anisotropies of principal components [124]. Yet another investigation used multivariate regression technique on cartilage to quantify its PG content using FTIR imaging. As the earlier studies focused on univariate techniques, an attempt has been made using partial least squares regression (PLSR) and principal component regression (PCR) methods for the analysis of the PG content of AC. And these results were compared with earlier one. This shows that multivariate technique provides better results for determining PG concentration [125].

Fig. 9.

Anisotropic flipping of a) collagen and b) sugar components in FTIRI. The concentration profile of PG by PCR algorithm is also shown in F ig. 9b. (Reprint from Yin, J. H., Xia, Y. and Ramakrishnan, N. 2011, Vib. Spectrosc., 57(2), 338–341, with permission from Elsevier).

4.2. Polarization studies and Quantitative analysis of cartilage using FTIRI

To study the orientation of polymer samples using FTIR microspectroscopic studies, polarization technique had been adopted [126]. Occurrence of partial polarization of light when reflected by a mirror is well known for physicists. While measuring isotropic samples, this partial polarization does not have any influence. But for heterogeneous materials like cartilage, this partial polarization does have an influence. The sample orientation introduces measuremental errors. Efforts were made to understand the polarization effects in biological samples. From the collagen tape and human trabecular, which have higher degree of orientation, it was found that the relative intensities of certain absorption bands significantly varied while the sample was rotated and measured through FTIR microscopic imaging [127]. This study provided profound information regarding orientation of collagen in cartilage for quantitative analysis. Potential characterization by FTIR imaging for the anisotropic materials using polarized radiation was attempted on poly (vinylidene fluoride) films that have been uniaxially elongated below and above the threshold temperature of the II (alpha) to I (beta) phase transition [128].

In tendon, polarized IR microscopy was used to assess the mineral-matrix components. The parallel and perpendicular orientations of tendon samples were analyzed using FTIR microscopy and it was found that the collagen has dichroic nature that provides orientation information of the fibre [22]. This dichroic nature of collagen in cartilage was further exploited using polarized IR microscopic imaging. The established results of cartilage specimen with an analyzer [24] prompted to experiment cartilage using i) a polarizer (before the sample), and ii) using a polarizer and an analyzer. It was concluded from the studies that presence of either a polarizer or an analyzer with cartilage provides same spectral information throughout the depth of cartilage. In case of studies with both polarizer and analyzer, the results are surprising. It was found that under cross polarization state (PװA⊥/P⊥Aװ), instead of observing null/negligible intensity of chemical components responses, intensity profile closer to perpendicular polarizer (P⊥)/Analyzer (A⊥) state were obtained. This suggests that a substantial rotation of infrared radiation by the cartilage specimen causes this. Also, under the cross-polarization condition, the spectra in all three zones exhibit clear peak splitting as shown in Fig. 10, which may reveal additional information regarding the anisotropy and chemical structure of the macromolecular components involved. The splitting of amide bands in cartilage suggests the possibility of resolving overlapped peaks of collagen triple helix and/or proteoglycan, which may help in better characterization of major biochemical components [129].

Fig. 10.

Infrared spectra of cartilage tissue at three different depths (one in each histological zone) for different polarization: (a) at 31.75 μm (superficial zone), (b) at 81.25 μm (transitional zone) and (c) at 300 μm (radial zone). While the P0 spectra are nearly identical to the P0A0 spectra, the P0A90 spectra show clear peak splitting under cross polarization, especially for the amide I peak. (Reprinted from Ramakrishnan, N., Xia, Y. and Bidthanapally, A. 2007, Phys. Med. Biol., 52(15), 4601–14, with permission from IOP publishing).

4.3. Collagen orientation and zonal boundary determination

Analysis of Infrared dichroism of biomaterials has started way back in 1950s [75, 130]. To determine the fibre/membrane orientation in receptor-mediated processes, drug actions and membrane probes, the important techniques and methodologies namely FTIR linear dichroism, UV-Vis linear dichroism, Time-resolved fluorescence anisotropy, NMR and Surface Plasmon Resonance have been reviewed [131]. Also, the errors involved in calculating the orientation functions for polymers from their dichroic measurements were analyzed [132]. Investigations were carried out to simulate transition moment of stretching vibrations of functional groups, calculating dichroic absorption data and order parameters and correlating the results with experimental data [133]. In view of the above, attempts were made in understanding the collagen orientation in cartilage. Using polarization studies, the amide I/amide II ratio was used as a marker to define collagen orientation [134].

Extending the previous study, studies have been carried out extensively to understand the cartilage/tendon isotropy/anisotropy in 3D point of view. From the results, it was observed that for articular cartilage, parallel sections in the superficial zone have the expected infrared anisotropy, which is consistent with that of regular/perpendicular sections. The parallel sections in the radial zone, however, have nearly isotropic amide II absorption and a distinct amide I anisotropy [39]. This new information provides certain understanding about the 3D structure of cartilage which may help in examining OA. The reconstruction of anisotropies of cartilage components [115] provides a completely different idea about the collagen orientation. The flipping of amide dichroic profile in the beginning of transitional zone and the anisotropy of sugar in upper part of radial zone may provide insight to zonal determination of cartilage. In addition, it was found that the clustering of IR spectra using fuzzy algorithm reveals histological zones in healthy cartilage tissue [135]. To determine the cartilage zones, PLM supported FTIR amide I/amide II ratio derived from parallel and perpendicular polarization data of tendon were used [134]. But due to the dependency of PLM results to derive cartilage amide ratio, the method was not effective. Hence, to overcome this issue, a novel technique was designed that used the dichroic behavior of amide I and amide II components of cartilage. As both the amide components have perpendicular polarization and the collagen itself has perpendicular orientation between superficial and radial zones, the dichroic cross-over of both the amides twice within the cartilage can aid to define the three zones of cartilage. This very idea was tested in both transmission and reflection modes as shown in Fig. 11 on various samples with the help of a curve fit [136]. The agreement of FTIR results with PLM for the same specimen established this as a standard technique.

Fig. 11.

Dichroic ratio profile of articular cartilage in (a) transmission mode and (b) reflection mode. The dichroic ratio values in both the figures for amide I (open circles) and amide II (solid rectangles) reaches the unity in the transitional zone. Hence, the curve fit with arbitrary values becomes indifferent for the transmission/reflection modes of experiment. (Reprinted from Ramakrishnan, N., Xia, Y. and Bidthanapally, A. 2007, Phys. Med. Biol., 52(15), 4601–14, with permission from IOP publishing).

4.4. Effect of mechanical loading on cartilage

In mature joints, the place where the cartilage fluid pressure as well as contact pressure is high, the tissue is very thick. The onset and spread of OA depends on the following mechanical factors: (a) reduction of fluid pressure, which activates the subchondral growth, (b) the subchondral growth due to local shear stress, (c) cartilage surface wear due to mechanical abnormalities, and (d) other mechanical factors [137]. Hence, collagen network and proteoglycan were analyzed in articular cartilage after cyclic loading. The loading caused condensation of non-calcified portion of the tissue by 12.8%. Birefringence of collagen in superficial zone got increased due to loading [138]. Though PLM was used for static compression studies of cartilage [139], efforts were made to understand the IR spectral effects due to loading. Cartilage specimens were observed for 80 cumulative hours of in vivo cyclical joint loading using PLM and FTIRI. FTIRI results show that loading increased PG content by 46% leaving the collagen content unchanged [140]. Molecular changes in terms of amide anisotropy were analyzed by comparing the control to cartilage specimens compressed for different strain values from 0 to 50%. The results show that the zonal boundaries were altered by compression, which is shown in Fig. 12.

Fig. 12.

The depth profiles of the amide II anisotropy at two infrared polarizations, 0° (solid line) and 90° dashed line), from the tissue at the 0% (a) and 50% (b) strains . (Reprinted from Xia, Y., Alhadlaq, H., Ramakrishnan, N., Bidthanapally, A., Badar, F. and Lu, M. 2008, J. Struct. Biol., 164(1), 88–95, with permission from Elseiver).

In addition, after compression, the two amide components with bond direction perpendicular to the external compression retain anisotropy (amide II in the superficial zone and amide I in the radial zone), whereas the measured anisotropy from the two amide components with bond direction parallel to the external compression changes their anisotropy significantly (amide I in the superficial zone and amide II in the radial zone). The trend change of anisotropy is depicted clearly in Fig. 13. Correlation of the compressed cartilage results obtained through PLM (minimum retardation) to FTIRI (amide II polarization crossover) provided promising results, as shown in Fig. 14 that can lead towards further investigation of compressed cartilage towards the onset of OA [141]. The studies were further extended in near infrared (NIR) region by correlating osmatic re-swelling of cartilage with NIR spectra of the same using Partial Least Squared Regression. The results show a strong relationship (R² = 95.89%, p < 0.0001) between the spectral data and re-swelling and hence the conclusion was drawn that the re-swelling of cartilage - both biochemical (osmotic) and mechanical (hydrostatic pressure) in origin, could be a better marker for characterizing the tissue, especially in regions surrounding focal cartilage defects in joints [142].

Fig. 13.

FTIRI anisotropy plots at cartilage’s “superficial zone” (RD 0.058), or 37.5 μm from the articular surface when the tissue is not compressed) (a and b) and “radial zone” (RD 0.6), about 360 μm from the articular surface when the tissue is not compressed) (c and d). The solid lines are the fitted lines from a theoretical equation that models the anisotropic changes in the amide absorption. (Reprinted from Xia, Y., Alhadlaq, H., Ramakrishnan, N., Bidthanapally, A., Badar, F. and Lu, M. 2008, J. Struct. Biol., 164(1), 88–95, with permission from Elseiver).

Fig. 14.

The linear regression correlation between the depth of the minimum retardation in PLM and the depth of the amide II cross-over in FTIRI, as the function of strains. The central shaded band marks the 95% confidence limits. Two outlines mark the 95% prediction limits. The Pearson correlation coefficient between the two variables is 0.9812. (Reprinted from Xia, Y., Alhadlaq, H., Ramakrishnan, N., Bidthanapally, A., Badar, F. and Lu, M. 2008, J. Struct. Biol., 164(1), 88–95, with permission from Elseiver).

4.5. Additional Techniques

The versatility of IR imaging can further be enhanced by various spectroscopic techniques. An example of it is application of second derivative spectroscopy in resolving the overlapping spectral analysis of articular cartilage specimen. Histological bovine cartilage sections were studied after enzymatic removal of PG. The results using second derivative analysis show that the alteration in PG content of peaks at 1064 cm⁻¹ and 1376 cm⁻¹ were notable [143]. 2D correlation spectroscopy, another popular technique generally used to increase spectral resolution, has also been explored to estimate the concentration gradient of cartilage constituents [144]. To avoid the error introduced due to sample thickness non-uniformity, attenuated total reflectance (ATR) technique has been examined for cartilage characterization. In freeze-dried bovine articular cartilage, ATR studies revealed the presence of hyaluronan and the phospholipids in superficial zone [145]. Even chemical and structural distributions of cellular components in chondrocytes have been analyzed using FTIRI-ATR technique [146]. Fiber optic probe (FOP) technique, which follows multiple internal reflection of infrared light within a probe, is another widely-used technique in IR spectroscopy. The probe provides advantage in examining those parts of a sample which generally can’t be prepared for regular absorption studies. Such an attempt to investigate the degeneration of human articular cartilage has been made [147]. Extending the studies, it was aimed to develop FOP based chemeometric method using partial least squares to estimate the degree of cartilage degradation [148]. To predict the histological Mankin score as an indicator of tissue quality, FOP spectra of tibial plateau articular cartilage were used to derive required spectral parameters [149].

4.6. Correlation with other techniques

The objective of characterization of cartilage using various techniques is to unify the different parameters to determine the onset of OA. In a rabbit model, OA was induced and the disease progression has been monitored by FTIR imaging, MRI and FOP [40]. Also, in a rabbit osteochondral defect model, short-term articular cartilage repair was monitored using MRI and FTIRI [150]. Also, FTIRI results (amide anisotropy) were correlated with those of PLM (retardation) to understand the tissue structure as well as zonal boundary determination [24, 136]. In addition, the correlation of PG-bound water fraction determined through MRI and FTIRI found satisfactory [151]. More cross technique correlations are required to arrive at a complete picture in understanding the various dimensions of cartilage.

5. CONCLUSION

The understanding of cartilage tissue has been extensively attempted using various techniques to determine the onset of osteoarthritis. It is also amazing to know the potential of FTIR imaging/microscopy/micro-spectroscopy in revealing the cartilage characteristics qualitatively and quantitatively. The polarization studies and the ability of ATR and FOP techniques with the aid of chemometrics have brought the cartilage characterization to a significant point. Still, extending the 2D correlation spectroscopy and ellipsometry to the cartilage characterization and correlating FTIR imaging to MRI, PLM and histochemical techniques can make the tissue as an open book to define clinically important biomarkers for the determination of the onset of OA. As a final note, the FTIRI cartilage research discussed in this review can be used to understand the potentials of FTIRI in other important structural materials.