Data & Analytics

11 pages

Impaired Intervertebral Disc Development and Premature Disc Degeneration in Mice With Notochord-Specific Deletion of CCN2

of 11
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Impaired Intervertebral Disc Development and Premature Disc Degeneration in Mice With Notochord-Specific Deletion of CCN2
   ARTHRITIS & RHEUMATISMVol. 65, No. 10, October 2013, pp 2634–2644DOI 10.1002/art.38075© 2013, American College of Rheumatology Impaired Intervertebral Disc Development andPremature Disc Degeneration in Mice WithNotochord-Specific Deletion of CCN2 Jake Bedore, Wei Sha, Matthew R. McCann, Shangxi Liu, Andrew Leask, and Cheryle A. Se´guin Objective.  Currently, our ability to treat inter- vertebral disc (IVD) degeneration is hampered by anincomplete understanding of disc development andaging. The specific function of matricellular proteins,including CCN2, during these processes remains anenigma. The aim of this study was to determine thetissue-specific localization of CCN proteins and to char-acterize their role in IVD tissues during embryonicdevelopment and age-related degeneration by using amouse model of notochord-specific CCN2 deletion.  Methods.  Expression of CCN proteins was as-sessed in IVD tissues from wild-type mice beginningon embryonic day 15.5 to 17 months of age. Given theenrichment of CCN2 in notochord-derived tissues, wegenerated notochord-specific CCN2–null mice to assessthe impact on the IVD structure and extracellularmatrix composition. Using a combination of histologicevaluation and magnetic resonance imaging (MRI), IVDhealth was assessed.  Results.  Loss of the CCN2 gene in notochord-derived cells disrupted the formation of IVDs in embry-onic and newborn mice, resulting in decreased levels of aggrecan and type II collagen and concomitantly in-creased levels of type I collagen within the nucleuspulposus. CCN2-knockout mice also had altered expres-sion of CCN1 (Cyr61) and CCN3 (Nov). Mirroring itsrole during early development, notochord-specificCCN2 deletion accelerated age-associated degenerationof IVDs. Conclusion.  Using a notochord-specific gene tar-geting strategy, this study demonstrates that CCN2expression by nucleus pulposus cells is essential to theregulation of IVD development and age-associated tis-sue maintenance. The ability of CCN2 to regulate thecomposition of the intervertebral disc suggests that itmay represent an intriguing clinical target for thetreatment of disc degeneration. Globally, the prevalence of low back pain isincreasing at an alarming rate, with the most recentsystematic review reporting a lifetime prevalence of 39%, which is predicted to increase substantially overthe coming decades as the population ages (1). Inter- vertebral disc (IVD) degeneration, an underlying causeof low back pain, begins with changes in the cellularmicroenvironment that are initiated long before theappearance of associated symptoms, such as decreasedmobility and acute or chronic pain (2). The lack of effective treatment for this widespread clinical problemis related to our limited understanding of the mecha-nisms that regulate the processes of IVD development,maintenance, and degeneration. In particular, there isan incomplete understanding of the relative importanceof individual growth factors, secreted molecules, andmatrix components that constitute the unique micro-environment of the IVD. Supported by the Canadian Institutes of Health Research(CIHR) (grants MOP-77603 to Dr. Leask and MOP-115718 toDr. Se´guin). Mr. Bedore and Mr. McCann’s work was supported by theCIHR Joint Motion Program/Training Program in MusculoskeletalHealth Research and Leadership. Dr. Sha is recipient of a Postdoc-toral Fellowship from the Canadian Scleroderma Research Group.Mr. McCann is also recipient of a CIHR Doctoral Research award.Dr. Se´guin is recipient of a Scholar award from The Arthritis Society/ Canadian Arthritis Network.Jake Bedore, HBSc, Wei Sha, PhD (current address: Chao- yang Hospital, Chaoyang, Beijing, China), Matthew R. McCann, BSc,Shangxi Liu, PhD, Andrew Leask, PhD, Cheryle A. Se´guin, PhD:Schulich School of Medicine and Dentistry, University of WesternOntario, London, Ontario, Canada.Dr. Leask owns stock or stock options in FibroGen. Address correspondence to Cheryle A. Se´guin, PhD, Depart-ment of Physiology and Pharmacology, Schulich School of Medicineand Dentistry, University of Western Ontario, London, Ontario N6A 5C1, Canada. E-mail: for publication November 20, 2012; accepted inrevised form June 25, 2013.2634  IVDs are specialized connective tissue structuresthat anchor adjacent vertebral bodies along the spine,conferring flexibility and providing mechanical stabilityduring axial compression. Anatomically, the IVD con-sists of 3 structurally distinct, yet interdependent tissues:the annulus fibrosus (AF), the central nucleus pulposus(NP), and the cartilage end plates that anchor the discto the adjacent vertebral bodies.The mechanical properties of the IVD resultfrom the specific organization of the extracellular matrix (ECM), which is maintained by the distinct cell popula-tions found within the component tissues. The AFconsists of concentric lamellae formed by parallel bun-dles of type I collagen fibers, which encapsulate the NPand provide tensile strength (3). The NP consists largelyof a proteoglycan and water gel supported by an irreg-ular network of type II collagen and elastin fibers. Themajor proteoglycan of the disc is aggrecan, which, due toits highly anionic glycosaminoglycan content, providesosmotic properties, enabling the NP to maintain heightand turgor against compressive loads (2,4). In contrast tothe cartilage end plates and AF, in which mesenchyme-derived cell types remain relatively invariant, the NPundergoes a substantial change in its cellular and ECMcomposition throughout life (5). Notochord cells consti-tute the primordium for the NP during development;however, in humans, their numbers decrease rapidlyafter birth and the NP becomes gradually populated bysmaller chondrocyte-like cells (6). Recent studies havedemonstrated that in mice, all cells of the mature NP arederived from the embryonic notochord (7,8); however,the change in cell phenotype associated with notochordcell maturation has yet to be established.Named after the first 3 members to be discov-ered, cysteine-rich protein 61 (Cyr61), connective tissuegrowth factor (CTGF), and nephroblastoma overex-pressed protein (Nov), the CCN family consists of 6secreted matricellular proteins (CCN1–CCN6) thatserve as multifunctional signaling mediators that regu-late the interactions of cells, growth factors, and theECM (9). Data generated using mouse knockout modelsrevealed that CCN proteins play key roles in angio-genesis, embryonic cartilage and bone formation, andmediate inflammation and fibrosis in adults (10,11). Of the CCN family members, CCN1 appears to possess anexpression pattern and activity similar to that of CCN2(12), whereas CCN3 appears to act antagonistically toCCN2 (13–15).CCN2, formerly known as CTGF, was identifiedas a protein secreted by hypertrophic chondrocytesthat promotes chondrocyte proliferation and differenti-ation (16,17). In mice, loss of the CCN2 gene resultsin skeletal dysmorphisms and perinatal lethality linked with improper rib cage formation due to impairedendochondral ossification (18). In addition to its role incartilage and bone development, several studies havesuggested that CCN2 may also play a role in both thenotochord and IVD. In mice, microarray-based geneexpression profiling demonstrated that expression of CCN2 is enriched in the embryonic node and notochord(19). Moreover, in zebrafish, knockdown of CCN2 ex-pression leads to notochord malformation and earlyembryonic lethality (20). Within the IVD, CCN2 wasfirst identified as an anabolic factor secreted by noto-chord cells, which induces NP cell proliferation andaggrecan production in vitro (21–23). Recent studieslocalized CCN2 expression to NP cells in degeneratedhuman IVDs and proposed that CCN2 promotes angio-genesis and accelerates IVD fibrosis and degeneration(24,25). In contrast, the association between CCN2 andIVD degeneration has led to speculation that the expres-sion of CCN2 is associated with initiation of a reparativeresponse that influences ECM remodeling and cellproliferation (26). Despite these correlational observa-tions, the specific role played in vivo by CCN2 in thedevelopment, aging, and degeneration of the IVD re-mains unclear.The objective of the present study was to examineselected CCN proteins that had previously been local-ized to IVD tissues for their expression during distinctstages of development and aging. To precisely delineatethe role of CCN2 in the IVD, we used a conditionalknockout strategy in mice harboring an allele that drivesthe expression of Cre recombinase under the control of a notochord-specific promoter (  Noto Cre ) (8) and CCN2alleles flanked with loxP sites (27). We found that loss of CCN2 in notochord-derived cells in vivo resulted inimpaired development of IVDs and marked accelerationof age-associated IVD degeneration. Our results providenew and valuable insights into the role played by CCN2in the IVD and suggest that modulation of CCN levelscould regulate disc degeneration. MATERIALS AND METHODS  Animals.  Genetically modified mice harboring anotochord-specific deletion of   Ccn2  were generated using theCre/loxP method. Mice carrying the  Ccn2  gene flanked by loxPsites ( Ccn2  Fl/Fl C57BL6 mice) (27) were mated with  Noto Cre mice (8), to generate mice bearing  Noto Cre and a floxed  Ccn2 allele in their germline. These mice were backcrossed tohomozygous floxed mice (  Noto Cre/    /  Ccn2  Fl/     Ccn2  Fl/Fl ) togenerate mice with inactivation of both alleles in notochord- CCN2 REGULATES IVD DEVELOPMENT AND MAINTAINS TISSUE INTEGRITY 2635  derived cells (genotype  Noto Cre/    /  Ccn2  Fl/Fl ). Homozygous dis-ruption of the  Noto  locus is perinatal lethal (28); viableoffspring have genotypes of either  Noto Cre/    /  Ccn2  Fl/Fl (herein-after referred to as CCN2-knockout) or  Noto   /    /  Ccn2  Fl/Fl (hereinafter referred to as wild-type). Mice were housed instandard cages and maintained on a 12-hour light/dark cycle, with rodent chow and water available ad libitum. Genotyping was performed as previously described (8,27).Mice were euthanized at the following ages: embryonicday 15.5, postnatal day 1, postnatal day 28, 12 months, and 17months. Time points were selected to reflect distinct stages of IVD development, including notochord segmentation (embry-onic day 15.5) and intervertebral disc formation (postnatal day1). To assess IVD aging, tissues were examined in fully formedIVDs prior to skeletal maturity (postnatal day 28), and in miceof advanced age prior to the onset of significant degenerativechanges (12 and 17 months), which were previously reported in wild-type mice (29,30). Experimental results at each time pointexamined were derived from the analysis of 5–6 IVDs fromeach mouse, using 4 wild-type and 4 CCN2-knockout mice, with the exception of the month 17 time point where data arereported for 4 wild-type and 3 CCN2 - knockout mice. All aspects of this study were conducted in accordance with the policies and guidelines set forth by the CanadianCouncil on Animal Care and were approved by the AnimalUse Subcommittee of the University of Western Ontario. Magnetic resonance imaging (MRI).  At 12 or 17 monthsof age, wild-type and CCN2-knockout mice were subjected toMRI. Three-dimensional spinal IVD imaging was conducted with a 9.4T small-animal MRI scanner equipped with a 30-mmmillipede coil (Agilent) and using a TrueFISP 3-dimensionalpulse sequence with the following parameters: repetition time5.4 msec, echo time 2.7 msec, field of view 51.2    22.4    16mm 3 , matrix 640    280    32 voxels, spatial resolution 80   80    500   m, and scan time 60 minutes. The anesthetizedmouse was placed supine on a tray and taped to minimizemotion artifacts. The body temperature of the mouse wasmaintained at 37°C using a MR-compatible physiologic mon-itoring and gating system (SA Instruments) during the scan.For quantification of signal intensity (an indicator of tissue hydration), a region of interest was standardized basedon wild-type animals, using a box size encompassing the entirelumbar IVD. Relative signal intensities within this region werenormalized to that of adjacent skeletal muscle to account forgray-level shifting and differences in receptive properties (seeSupplementary Figure 1, available on the  Arthritis & Rheuma-tism  web site at As an independent measure of IVD degener-ation, pixel standard deviation was assessed within the IVDin order to quantify the progressive loss of a distinct AF–NPboundary. As an inhomogeneity descriptor, the standard devi-ation is a relative measurement that is independent of gray-level shifting (31). Signal intensity and standard deviation values were measured using ImageJ software, and Student’s2-tailed  t -test was used to compare values between CCN2-knockout mice and wild-type control mice. Histologic and immunofluorescence analyses.  Formalin-fixed tissue samples were decalcified for 5 days with gentlerocking, using a Shandon TBD-2 decalcifier (Thermo Scien-tific) at a fluid-to-tissue ratio of 10:1. Following standardhistologic processing, samples were embedded in paraffin,and 5   m–thick serial sections were cut in the coronal plane.Sections were deparaffinized in xylene and rehydrated bysuccessive immersion in descending concentrations of alcohol.Serial sections were processed with either hematoxylin andeosin or 0.1% Safranin O–0.02% fast green, and images wereacquired using a Leica DM1000 microscope with Leica Ap-plication Suite software. For evaluation of IVD degenera-tion, sections of the lumbar spine from wild-type and CCN2-knockout mice were stained with Safranin O–fast green, andchanges were scored according to the Thompson grading scale(32) by 2 independent observers who were blinded with regardto the experimental group.For immunohistochemistry, samples were processedas described above and antigen retrieval was performed with10 m  M   sodium citrate (pH 6.0) for 12 minutes at 95°C. Slides were blocked by incubation for 1 hour with the appropriatespecies-specific serum albumin (5%) in phosphate bufferedsaline plus 0.1% Triton X-100 (Sigma). Primary antibodies were added and incubated overnight at 4°C in a humidifiedchamber, followed by incubation with secondary antibodies for1 hour at room temperature. Samples were mounted usingVectashield mounting medium with DAPI (Vector). The fol-lowing antibodies and dilutions were used: anti-COL1A2 at a1:200 dilution (catalog no. sc-28654; Santa Cruz Biotechnology), Figure 1.  Localization of CCN proteins in the intervertebral discs of  wild-type mice during skeletal development. Representative imagesdemonstrate the immunolocalization of CCN1, CCN2, and CCN3 in wild-type mice of the following ages: embryonic day 15.5 (E15.5)(  A–C ), newborn (NB) ( D–F ), postnatal day 28 (P28) ( G–I ), 12 months(  J–L ), and 17 months ( M–O ). The expression of CCN1 and CCN2, butnot CCN3, is enriched in the notochord and notochord-derivednucleus pulposus ( arrows ). Images are oriented with rostral at the topand are representative of 4 mice per time point. Bar  50   m.2636 BEDORE ET AL   anti-COL2A1 at a 1:200 dilution (catalog no. 70R-CR008;Fitzgerald), antiaggrecan at a 1:50 dilution (catalog no. sc-25674; Santa Cruz Biotechnology), anti-CCN1/Cyr61 1:100(catalog no. sc-13100; Santa Cruz Biotechnology), anti-CCN2/ CTGF 1:200 (clone L-20; Santa Cruz Biotechnology), anti-CCN3/Nov 1:1,000 (catalog no. AF-1976; R&D Systems),DyLight 488–conjugated anti-goat, anti-rabbit, and anti-mouseIgG at a 1:500 dilution (Jackson ImmunoResearch), andDyLight 594–conjugated anti-goat and anti-rabbit IgG at a1:500 dilution (Jackson ImmunoResearch). Species-specificisotype IgG controls (Invitrogen) were used to assess antibodyspecificity (see Supplementary Figure 2, available on the  Arthritis & Rheumatism  web site at http://onlinelibrary. Images were cap-tured with a Zeiss Axio Imager.M1 fluorescence microscopeand processed with Northern Eclipse software. RESULTSTissue-specific expression patterns of CCN pro-teins in murine IVDs.  To determine the distribution of CCN proteins within both notochord- and mesenchyme-derived tissues of the IVD, immunolocalization studies were conducted on spinal segments from wild-type miceranging in age from embryonic day 15.5 to 17 months of age (Figure 1). Following notochord segmentation andthe initiation of disc formation on embryonic day 15.5,CCN1 expression was not detected (Figure 1A), CCN2expression was specific for the notochord-derived NP(Figure 1B), and CCN3 expression was detected in theNP, early AF, and mesenchymal cells that give rise to the vertebral bone (Figure 1C). In newborn mice, CCN1 wasdetected in the NP and hypertrophic chondrocytes of the vertebral bone (Figure 1D). CCN2 expression was mostrobust in the NP, but detectable in mesenchymal cells of the AF and hypertrophic chondrocytes of the adjacent vertebrae (Figure 1E). In contrast, CCN3 expressionappeared more robust in the AF and hypertrophicchondrocytes than in the NP (Figure 1F). From post-natal day 28 to 17 months of age, the expression of CCN1 and CCN2, but not CCN3, within the IVD was Figure 2.  Histologic appearance and localization of extracellular matrix (ECM) proteins in intervertebral discs (IVDs) obtained on embryonic day15.5 from wild-type mice and CCN2-knockout mice.  A,  Representative images of putative IVD tissues following notochord segmentationdemonstrate the histologic appearance and localization of CCN proteins. Immunolocalization demonstrates reduced CCN1 expression and enhancedCCN3 expression in CCN2-knockout mice as compared to wild-type littermate controls.  Arrows  indicate nucleus pulposus (NP) tissue.  B, Representative images of the ECM composition show decreased levels of type II collagen and aggrecan proteins and increased levels of type Icollagen within the NP of CCN2-knockout mice as compared to wild-type littermate controls. Images are oriented with rostral at the top and arerepresentative of 4 mice per genotype. Bars  50   m. H&E  hematoxylin and eosin.CCN2 REGULATES IVD DEVELOPMENT AND MAINTAINS TISSUE INTEGRITY 2637  enriched in the NP (Figure 1), suggesting that theseproteins may play a role in NP function. Disruption of embryonic IVD formation innotochord-specific CCN2-knockout mice.  To investigatethe role of CCN2 in NP function, notochord-specificCCN2–knockout mice were generated. Notochord seg-mentation and IVD patterning did not appear to bedisrupted in CCN2-knockout mice as compared to their wild-type littermate controls (Figure 2). On embryonicday 15.5, loss of CCN2 resulted in increased expressionof CCN3 within the putative NP (Figure 2A). Comparedto wild-type littermate controls, CCN2-knockout micedemonstrated decreased aggrecan and type II collagenprotein expression within the putative NP, concomitant with increased type I collagen protein expression (Figure2B). By postnatal day 1, when IVD development iscompleted, perturbation of the ECM composition of theIVD in CCN2-knockout mice was more pronounced(Figure 3). Notably, deletion of CCN2 in notochord-derived cells resulted in decreased CCN1 expression andincreased CCN3 expression within the NP (Figure 3A).Moreover, compared to wild-type littermate controls,CCN2-knockout mice displayed decreased levels of ag-grecan in the NP, as well as decreased levels of type IIcollagen and increased levels of type I collagen through-out the IVD (Figure 3B).Based on these observations, we anticipated thatthe alterations in the IVD matrix composition detectedduring development in notochord-specific CCN2–knockout mice would disrupt disc function in the matureIVD. However, on postnatal day 28, neither the tissuearchitecture nor the ECM content was appreciably al-tered in CCN2-knockout mice as compared to theirlittermate controls (Figures 4A and B). Intriguingly,although the CCN2 gene was deleted in notochord-derived cells of the NP, CCN2 protein was detected Figure 3.  Histologic appearance and localization of ECM proteins in IVDs obtained from 1-day-old wild-type mice and CCN2-knockout mice.  A, Representative images of IVD tissues on postnatal day 1 demonstrate the histologic appearance and localization of CCN proteins. Similar to thefindings at embryonic time points, CCN1 expression is reduced and CCN3 expression is enhanced in CCN2-knockout mice as compared to wild-typelittermate controls.  Arrows  indicate NP tissue.  B,  Representative images examining the ECM composition show decreased levels of type II collagenthroughout the IVD, decreased levels of aggrecan within the NP, and increased levels of type I collagen within the NP of CCN2-knockout mice ascompared to wild-type littermate controls. Images are oriented with rostral at the top and are representative of 4 mice per genotype. Bars  50   m.See Figure 2 for definitions.2638 BEDORE ET AL 
Related Documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks