Abstract
Introduction
Lower back pain affects millions of people worldwide, and has been linked to degenerative changes in the intervertebral disc (IVD) of the spine. In the “NPmimetic” project, a multidisciplinary team has come together to create a biomimetic nanopolymer based implant and develop a minimally invasive therapy to reconstruct and regenerate diseased nucleus pulposus (NP). The biodegradable nanofibers of the implant can also be designed to carry anti-inflammatory drugs, which can be released in situ promoting healing and preventing inflammation (http://npmimetic.com/).
An IVD consists primarily of a proteoglycan-water gel embedded in a randomly arranged collagen network in the NP, and highly ordered collagen lamella in the annulus fibrosus (AF). Disc cells in the AF are elongated parallel to the collagen fibers and produce predominantly collagen I in the outer AF in response to deformation. NP cells are responsive to hydrostatic pressure and synthesize mostly proteoglycan and collagen II. In adults, the cell density in an IVD is very low and cell phenotypes can change in response to altered matrix and stress distribution.1 For successful regeneration, tissue integrity together with the right mechanical environment is essential for normal cell function. To develop suitable biomimetic implants a thorough characterization profile, that can be used as an aspirational target is important. In this study, we use Fourier transform infrared (FTIR) microscopic imaging to obtain chemical maps of control and in-vivo CABC-degenerated goat IVDs.2,3
Materials and Methods
Goat IVDs were kindly provided from VU University Medical Centre (VUMC), Amsterdam. The IVDs were formalin fixed (10%, overnight), paraffin embedded, and 4 µm sections were mounted on custom- made reflective steel slides. FTIR microscopy in transflection mode was used to generate chemical distribution maps from unstained paraffin sections. One microscopic image covers only a very small area (350 × 350 µm) of an IVD sample (∼2.8 × 2.3 cm, transverse section). Bigger areas are measured by sequential sample movement and image acquisition covering a user-defined mosaic image area. In this example, an area of 80 × 64 images was measured resulting in a total of 10,720 IR spectra per IVD section. FTIR mosaic imaging generally generates many thousands of data points. A major challenge is handling and analyzing such large and chemically complex datasets to extract meaningful information. However, using iterative multivariate curve resolution (MCR) techniques on the reduced data matrix from principal component analysis analysis of second derivative spectra it is possible to deconvolute highly overlapping infrared peaks into single contributions of different molecular species.
Results
The chemical identity of the extracted component using an iterative MCR algorithm is determined by comparing the extracted spectral profiles with the spectral profiles of reference materials for proteoglycan and collagen. Spectral features matching typical proteoglycan and collagen spectral characteristics are observed. The distribution of the extracted MCR factors for collagen and proteoglycan are compared with consecutive IVD sections, which were stained using traditional histological methods; Masson trichrome staining for collagen and Alcian blue staining for proteoglycan. A good match can be observed between the distribution maps of collagen and proteoglycans derived from FTIR microscopic imaging measurements and the traditionally stained sections.
Conclusion
FTIR microscopic imaging in transflection mode has been successfully used to generate molecular images of the collagen and proteoglycan distribution of natural IVD material without the use of contrast-enhancing agents. The use of second derivative spectra together with the application of iterative MCR algorithms shows great potential to enhance the chemical specificity of FTIR mosaic imaging and opens up the possibility to distinguish between different proteclycans and proteins as well as their secondary structure, which usually are expressed only through minor peak shifts.
None declared
Adams MA, Roughley PJ. What is intervertebral disc degeneration, and what causes it? Spine 2006;31(18):2151–2161
Hoogendoorn RJW, Wuisman PIJM, Smit TH, Everts VE, Helder MN. Experimental intervertebral disc degeneration induced by chondroitinase ABC in the goat. Spine 2007;32(17):1816–1825
Hoogendoorn RJW, Helder MN, Kroeze RJ, Bank RA, Smit TH, Wuisman PI. Reproducible long-term disc degeneration in a large animal model. Spine 2008;33(9):949–954