We present a method to investigate spatial chondrocyte organization in the anulus fibrosus of the intervertebral disc using an optical sectioning method.
Intervertebral disc (IVD) degeneration is a leading cause of low back pain and it entails a high degree of impairment for the affected individuals. To decode disc degeneration and to be able to develop regenerative approaches a thorough understanding of the cellular biology of the IVD is essential. One aspect of this biology that still remains unanswered is the question of how cells are spatially arranged in a physiological state and during degeneration. The biological properties of the IVD and its availability make this tissue difficult to analyze. The present study investigates spatial chondrocyte organization in the anulus fibrosus from early embryonic development to end-stage degeneration. An optical sectioning method (Apotome) is applied to perform high resolution staining analyses using bovine embryonic tissue as an animal model and human disc tissue obtained from patients undergoing spine surgery. From a very high chondrocyte density in the early embryonic bovine disc, the number of cells decreases during gestation, growth, and maturation. In human discs, an increase in cellular density accompanied the progression of tissue degeneration. As had already been demonstrated in articular cartilage, cluster formation represents a characteristic feature of advanced disc degeneration.
The intervertebral disc (IVD) is a cartilage-based structure that biochemically and with respect to cellular architecture, at first sight, resembles in many ways the articular cartilage1. Indeed, both IVD degeneration and osteoarthritis (OA) of articular cartilage are characterized by joint space narrowing due to cartilage wear, subchondral cyst and osteophyte formation, and subchondral sclerosis2,3. Despite these seeming similarities architecture and functional role of both tissues differ. While the matrix of articular cartilage is mainly formed of an arcade-forming collagen type II network, the IVD consists of three different types of tissue: the collagen type II-rich nucleus pulposus in the center takes up axial loads and transmits them to an encompassing ring of densely packed circular collagen type I fibers which is called anulus fibrosus. Their function is to absorb the translated axial pressures received by the proteoglycan- and water-rich nucleus with their tensile longitudinal fiber strength. At the top and bottom of each nucleus and anulus a hyaline cartilaginous endplate forms the junction to the adjacent vertebrae4 (Figure 1).
In articular cartilage, four distinct spatial chondrocyte patterns can be found: pairs, strings, double strings, small respectively big clusters5,6,7 (Figure 2). Changes in this pattern are associated with OA onset and progression8,9. Spatial chondrocyte organization is also indicative for a direct functional property of cartilage, namely its stiffness, underlining the functional relevance of this image-based grading approach10,11. These patterns can additionally be identified with already existing clinically available technology12. Due to the similarities between the IVD and articular cartilage, it can be hypothesized that characteristic chondrocyte patterns are also present in the IVD. Cluster formation is a phenomenon also observed in the degenerated IVD13,14.
When trying to analyze spatial cellular organization in the IVD, it is necessary to overcome several technical difficulties that are not present when investigating articular cartilage:
First, processing of the tissue itself is much more challenging than with the homogeneous hyaline cartilage which is largely composed of collagen type II. The IVD's main fiber component is collagen type I, which makes it much more difficult to generate thin histologic sections. While in the hyaline articular cartilage even thick sections can easily be analyzed due to the "glass-like" nature of the tissue, the collagen type I network of the IVD is optically highly impenetrable. For this reason, a strong background noise is a common problem in the histology of the IVD. A fast and cheap way to penetrate this optically dense tissue is the use of an optical sectioning device e.g., by means of an Apotome. In such an Apotome, a grid is inserted in the illumination pathway of a conventional fluorescence microscope. In front of the grid a plane-parallel glass plate is placed. This tilts back and forth thus projecting the grid in the image in three different positions. For each z-position, three raw images with the projected grid are created and superimposed. By means of special software, the out of focus light can be calculated out. The underlying principle is that, if the grid is visible, that information is in focus, if not it is considered to be out of focus. With this technique, well focused and high-resolution images can be acquired in a reasonable amount of time.
Secondly, the tissue is hard to come by from human donors. When doing total knee replacement, the entire surface of the joint can be obtained for further analysis during surgery. Although osteoarthritis of a diarthrodial joint is also a disease of the whole joint, there are nevertheless strong focal differences in the quality of the cartilage with usually some areas of the joint still being intact, for example due to reduced loading in that area. This situation is different in the IVD, where surgery is usually only performed when the disc is globally destroyed. When obtaining tissue from human donors from the operation room, the tissue is also highly fragmented and it is necessary to correctly allocate the tissue to one of the three cartilage types of the IVD before doing further analyses. To allow more detailed analyses of also larger tissue sections and to look into the embryonic development of the IVD the choice of an animal model organism is, therefore, necessary.
When choosing such a model organism it is important to have a system which is comparable with the human disc with respect to its anatomy and dimensions, its mechanical loading, the present cell population as well as its tissue composition. For the purpose of the presented technique here we suggest the use of bovine lumbar disc tissue: A critical property of the human disc resulting in its low regenerative potential is the loss of notochordal cells during maturation in the nucleus. However, in numerous model organisms notochordal cells can be detected their entire life long. Most of the few animals which lose their notochordal cells such as sheep, goats or chondrodystrophig dogs have an IVD that is much smaller than human discs. Only lumbar bovine discs present with a comparable sagittal disc diameter to those of human IVDs15.
A key factor leading to early disc degeneration is excessive mechanical loading. The intradiscal pressures of a standing cow in the lumbar spine are around 0.8 MPa with the spine aligned horizontally. Surprisingly these pressures are comparable to the lumbar intradiscal pressures reported for the erect human spine (0.5 MPa)15,16. Also the amount of water and proteoglycans in bovine discs is comparable to that of the IVD from young humans17. Therefore, although the actual movement pattern of the motion segments might differ in quadrupedal animals from the bipedal human, with respect to total loading and disc characteristics, the cow is much closer to human biology than other established animal models for the IVD such as sheep and dogs.
In this protocol we present a technique how to analyze changes in the IVD from the point of view of spatial chondrocyte organization from early embryonic development to end stage degeneration.
For the analysis of embryonic development and maturation, bovine discs were used. To evaluate degeneration of the IVD, human samples were analyzed.
Human IVD tissue was obtained from patients undergoing surgery for lumbar disc degeneration, disc prolapse, or spinal trauma in the Department of Orthopaedic Surgery, University Hospital of Tübingen and the BG Trauma Centre Tübingen. Full ethical committee approval was obtained before the commencement of the study (project number 244/2013BO2). Written informed consent was received from all patients before participation. The methods were carried out in accordance with the approved guidelines.
Bovine tissue was obtained from the Bavarian State Office for Health and Food Safety/Oberschleißheim and from a rendering plant in Warthausen (Germany). Local and veterinary authorities' approval was received for tissue from dead animals.
1. Sample harvest
2. Sample preparation
3. Grading of sample age, integrity, and degeneration
4. Tissue fixation
5. Histologic sectioning
6. Fluorescence staining
7. Microscopic imaging and processing
8. Cellular pattern identification and density assessment
Using mosaic images, the architecture of the IVD with its dense collagen fiber network in the anulus and the softer nucleus can clearly be recognized (Figure 4). A continuous decrease in cellular density can be observed during embryonic development (Figure 5). While in the early stages of IVD development a cell density of 11,435 cells/mm² in the bovine anulus fibrosus and 17,426 cells/mm² in the bovine nucleus pulposus can be found, these numbers decrease rapidly to 1,011 cells/mm² (bovine anulus fibrosus) and 488 cells/mm² (bovine nucleus pulposus) until birth. In the adult cattle 71 cells/mm² (anulus fibrosus) resp. 106 cells/mm² (nucleus pulposus) are seen (Figure 6 A-B). Using bi-channel imaging with the Apotome allows to visualize the 3D architecture of the spatial patterns (Figure 7).
Figure 1. Macroscopic anatomy of the intervertebral disc. Schematic drawing of the intervertebral disc showing the nucleus pulposus (red), directly around it the intermediate zone (pink), and then in circular layers around it the anulus fibrosus. Note the ply-angle direction of the collagen type I fibers in the anulus fibrosus. The axial load to the nucleus can thus be translated to axial tensile forces of the collagen fibers. Please click here to view a larger version of this figure.
Figure 2. Schematic illustration of different spatial organizational patterns of chondrocytes. Depending on the tissue chondrocytes are found as single cells, pairs or strings in healthy cartilage. With beginning degeneration these patterns change to form double strings, small clusters and then big clusters in end-stage degeneration. This figure has been modified from Danalache, M. et al.21. Please click here to view a larger version of this figure.
Figure 3. Different conditions of the intervertebral disc (IVD). (A–D') Sagittal T2-weighted magnetic resonance imaging of the human lumbar spine (A–D), with the magnified motion segment L4/L5 (A'–D'). The ventral side of the patient faces left, the dorsal side faces right with the spinal canal with its white signal for the cerebrospinal fluid. (A,A') Intact IVD with the anulus displayed with a hypointense (black) signal due to the high collagen type I content and the nucleus displayed much brighter (hyperintense) due to the high content of water-binding proteoglycans (Pfirrmann grade19 I). (B,B') Beginning IVD degeneration with a loss of the water signal from the nucleus pulposus and where the distinction between the anulus and nucleus is lost (Pfirrmann grade19 IV). (C,C') Acute nuclear prolapse with still a prominent water signal in the area of the nucleus indicative for an otherwise intact IVD and the disc tissue protruding dorsally into the spinal canal. (D,D') Advanced disc degeneration with a largely destroyed IVD with a complete loss of the water signal within the disc, ventral and dorsal spondylophyte formation and subchondral sclerosis of the vertebrae corresponding to a Pfirrmann grade19 of V. Please click here to view a larger version of this figure.
Figure 4. Mosaic fluorescence imaging of the intervertebral disc (IVD). The architecture of the IVD with its dense collagen fiber network in the anulus fibrosus and the softer nucleus can clearly be recognized. DAPI-nuclear staining (white) in the axial (A1) and sagittal (A2) plane shows the cell distribution and arrangement within the IVD. Magnified representative areas from these mosaic images are displayed in B1–B4 and C1–C4 illustrating the spatial chondrocyte organization – in this case single cells (green box), pairs (blue box) and strings (yellow box). A: scale bar 1,000 µm, B1–B4: scale bar 200 µm, C1–C4: scale bar: 50 µm. This figure has been modified from Bonnaire, F. C. et al.22. Please click here to view a larger version of this figure.
Figure 5. Different developmental and maturation stages of bovine anulus fibrosus and human degenerating anulus fibrosus. DAPI nuclear staining. The mosaic images show a whole early embryonic disc in an axial section (A1) and representative images of the bovine anulus during embryonic development, maturation, and beginning degeneration (A2–A8). (B1–B3) Anulus from human IVDs was obtained intraoperatively. A continuous decrease in cellular density can be observed during embryonic disc development. A higher spatial organization cellular pattern seems to be present especially around birth. In the adult human disc during degeneration, the cellular density increases again and increasing cluster formation can be observed. Scale bar 100 µm. Wog: weeks of gestation. This figure has been modified from Bonnaire, F. C. et al.22. Please click here to view a larger version of this figure.
Figure 6. Reduction in cellular density throughout development and maturation of the bovine intervertebral disc. The mean (standard deviation) cell count per mm² is illustrated by bar diagrams for the bovine anulus fibrosus (A) and the bovine nucleus pulposus (B). A clear reduction of cellular density can be observed especially during the embryonic period which continues to a lesser extent at least until full maturation of the disc (n=72). Wog: weeks of gestation. This figure has been modified from Bonnaire, F. C. et al.22. Please click here to view a larger version of this figure.
Figure 7. Apotome imaging of the intervertebral disc (IVD). Bi-channel image showing the cytoplasm (red, Actin staining) and the nucleus (blue, DAPI nuclear staining). (A) In the intact IVD, in addition to the predominant spatial pattern of single chondrocytes, pairs are also found. (B) In degenerated anulus cell clusters can be found. Scale bar 20 µm. Please click here to view a larger version of this figure.
Video 1. Apotome imaging of a pair in the intervertebral disc (IVD) as a 3D Model. Bi-channel images showing the cytoplasm (red, Actin staining) and the nucleus (blue, DAPI nuclear staining) of a pair. Please click here to download this Video.
Video 2. Apotome imaging of a cluster in the intervertebral disc (IVD) as a 3D Model. Bi-channel images showing the cytoplasm (red, Actin staining) and the nucleus (blue, DAPI nuclear staining) of a cluster. Please click here to download this Video.
Table 1: Bovine embryonic development, maturation, and growth after birth with its different milestones. Wog – weeks of gestation. Gestation period ca. 283 days, natural life-expectancy 20-25 years20,23,24. Please click here to download this Table.
Using fluorescence microscopy augmented by mosaic imaging and optical sectioning, we evaluated the spatial arrangement of chondrocytes in the anulus of the lumbar IVD throughout development, maturation, and degeneration. While degenerative tissue could be harvested from patients receiving spine surgery for disc degeneration, analysis of the embryonic period and maturation phase required the use of a model organism (bovine). High cellular densities were noted in the anulus during early embryonic development. In the further course of development, postnatal growth and maturation a pronounced decrease in cellular density could be observed. In human tissue with advanced disc degeneration, we could then quantify an increase in cellular density in the anulus fibrosus.
The rapid increase in tissue volume combined with actively dividing and biosynthetically active notochordal cells are probable reasons for the changes in cellular density observed in the embryo25. The mechanisms by which cellular density again increases with degeneration remain, however, still unclear. Degenerative processes of the anulus lead to a series of pathological changes including increased tissue innervation and inflammation, upregulation of matrix-degrading enzymes and growth factor production, and also changes in cellularity26,27,28.
When considering cell presence as a function of cellular spatial organization based on patterns known from articular cartilage7,9 we found no recognizable spatial organization in early embryonic discs where the cells seem to be densely packed and which we considered to be present as single cells22. This finding is consistent with the results from embryonic articular cartilage7. In the healthy mature anulus, single cells represent the predominant spatial pattern22. Pairs and string-formations can, however, also be observed22. The more degenerate the tissue in human discs, the higher the proportion of cells that can be found in clusters22. The physiopathological model suggested by Rolauffs in articular cartilage7,9,29 shows an intriguing similarity to our findings. Previous studies on IVD degeneration had also already outlined cluster formation as a hallmark of disc degeneration14,30,31,32,33. Since these clusters can mostly be found in highly degenerated tissues, cluster formation might indicate a failed attempt of the tissue to repair the degenerative damage34. Establishing a strong correlation between the locally predominant chondrocyte pattern and tissue elasticity, a high functional relevance of these patterns could be demonstrated in articular cartilage10,11. It can be speculated that such a functional relevance also applies to spatial chondrocyte organization in the IVD.
Fluorescence histologic analyses are an easy to come by and attractive means to analyze morphologic changes to tissues. When trying to histologically analyze the IVD, there are distinct technical difficulties that need to be overcome: First, the limited availability of human tissue makes it important to choose an adequate animal model organism with which those aspects of the disease can be studied where human samples cannot be obtained. For the research question addressed in this study, we chose a bovine animal model.
Secondly, processing of the collagen type I-rich IVD is much more challenging than for most other human tissues. The dense collagen type I-fiber network strongly scatters the fluorescing light creating a high background noise signal. This problem is best addressed by using a technique that allows the subtraction or elimination of such background signal. A well-known method to do so is confocal microscopy. While the image quality of confocal acquired LASER-images is usually excellent, the disadvantages of this technique are that it is relatively time-consuming and as such it does not allow the analysis of larger tissue areas by means of mosaic imaging. Secondly confocal microscopes are relatively expensive and not available everywhere. A much faster and cheaper way to filter out background noise is to perform optical sectioning e.g., by means of an Apotome.
To be able to correctly interpret the findings, it is essential to also correctly allocate the tissue that is to be analyzed to its origin in the IVD. While this task is relatively simple when having an entire bovine disc to select from, it can be very challenging when receiving human tissue from the operation theatre. The characteristic feature of the anulus fibrosus is its very dense collagen type I network in an angle-ply fiber orientation. The nucleus, in contrast, is an amorphous gelatinous structure where no higher collagen architecture is visible to the naked eye. The intermediate zone lies between those two extrema and also possesses a clear collagen fiber architecture, but it is much softer and less dense than the anulus fibrosus. The cartilaginous endplate consists of hyaline cartilage, does not present with a collagen architecture recognizable by the naked eye but it is rather "glass-like" as the term "hyaline" suggests. Unlike the nucleus pulposus it is, however, also very stiff, cannot be deformed, and is situated on the subchondral bone which can often still be recognized by inspection of the tissue.
Once the tissue origin has been correctly identified, the tissue still needs to be correctly oriented for sectioning to obtain standardized images which also allow a proper qualitative and quantitative read-out. In orientation to other imaging techniques such as magnetic-resonance imaging we suggest two standard analysis-planes: one median-sagittal section and one in the axial plane. These two planes will also provide a good impression of the architecture of the collagen type I-fiber network of the anulus. When interpreting the findings obtained from the F-actin staining it needs to be kept in mind that freezing of the tissue samples may cause cytoskeletal structure alterations35 which should, however, not affect the spatial chondrocyte organization.
To fully comprehend the cellular biology of the IVD is one of the still open challenges in our understanding of disc degeneration and regeneration36. Questions that sometimes seem trivial, such as the physiological arrangement of cells in healthy tissue or the question of cellular reorganization during degeneration, remain, to date, unanswered. The obtained knowledge may allow us to use this image-based degeneration marker to evaluate tissue quality, investigate the reversibility of cluster formation and thus possibly define a therapeutic window in which the degenerating processes can still be successfully targeted.
The authors have nothing to disclose.
We thank our co-authors from the original publications for their help and support. We thank Charlotte Emma Bamberger for helping to acquire the apotome images.
Amphotericin B | Merck KGaA, Germany | A2942 | |
Adhesion Microscope Slides SuperFrost Plus | R. Langenbrinck, Germany | 03-0060 | |
ApoTome | Carl Zeiss MicroImaging GmbH, Germany | 462000115 | |
AxioVision Rel. 4.8 with Modul MosaiX | Carl Zeiss MicroImaging GmbH, Germany | ||
CellMask Actin Tracking Stain | Thermo Fischer Scientific, US | A57249 | |
Cryostat | Leica Biosystems, US | CM3050S | |
DAPI | Thermo Fischer Scientific, US | D1306 | |
Dulbecco's modified Eagle's medium (DMEM) | Gibco, Life Technologies, Germany | 41966052 | |
Ethylenediaminetetraacetic acid | Sigma-Aldrich, US | 60004 | |
Fluorescence Miscoscope – Axio Observer Z1 with Axio Cam MR3 and Colibri | Carl Zeiss MicroImaging GmbH, Germany | 3834000604 | |
Formaldehyde | Merck KGaA, Germany | 104002 | |
Image J 1.53a, with Cell counter plugin | National Insittute of Health (NIH), US | ||
Invitrogen Alexa Fluor 568 Phalloidin | Thermo Fischer Scientific, US | A12380 | |
Microscopic Cover Glasses | R. Langenbrinck, Germany | 01-1818/1 | |
PAP Pen Liquid Blocker | Science Sevices GmbH, Germany | N71310 | |
Penicillin-Streptomycin | Sigma-Aldrich, US | P4333 | |
Phosphate buffered saline | Sigma-Aldrich,US | P5119 | |
Scalpel | pf medical AG, Germany | 2023-01 | |
Tissue-tek O.C.T. Compound | Sakura Finetek, Netherlands | SA6255012 |