The Primary Cilium of Connective Tissue Cells: Imaging by...

来自 : 发布时间：2024-05-16

The Anatomical RecordVolume 291, Issue 9 p. 1062-1073 Special Article Free Access The Primary Cilium of Connective Tissue Cells: Imaging by Multiphoton Microscopy Eve Donnelly, Corresponding Author eld26@cornell.edu Department of Biomedical Sciences, Cornell University, Ithaca, New YorkDepartment of Biomedical Sciences, College of Veterinary Medicine, Cornell University, T5-002 Veterinary Research Tower, Ithaca, New York 14853Search for more papers by this authorRebecca Williams, School of Applied and Engineering Physics, Cornell University, Ithaca, New YorkSearch for more papers by this authorCornelia Farnum, Department of Biomedical Sciences, Cornell University, Ithaca, New YorkSearch for more papers by this author Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URLShare a linkShare onEmailFacebookTwitterLinked InRedditWechat Abstract Although the role of the primary cilium as a sensory organelle in epithelial cells has been elucidated significantly over the past decade, the function of primary cilia in connective tissue cells has been studied less extensively. Primary cilia have been implicated as mechanotransducers in connective tissues, but the mechanisms by which the cells sense loads and convert them to biochemical signals for tissue formation and adaptation are poorly understood. Before hypotheses regarding the function of the primary cilium in connective tissue cells can be tested, methods for quantitation of incidence as well as three-dimensional visualization of primary cilia with respect to the extracellular matrix (ECM) are needed. The objective of this study was to develop a rapid method for visualizing primary cilia in their native ECM in a wide range of connective tissues. Whole-mount immunohistochemical and multiphoton microscopy techniques were developed to simultaneously image primary cilia, cell nuclei, and collagen and their relationships to each other in situ. Axonemes of primary cilia projecting into the ECM were successfully visualized in thick sections of growth plate cartilage, tendon, ligament, meniscus, intervertebral disc, and perichondrium. These methodologies will allow analysis of the incidence and three-dimensional orientation of primary cilia and enable investigation of the role of primary cilia in normal and pathological growth and adaptation in a variety of musculoskeletal tissues. Anat Rec, 291:1062–1073, 2008. © 2008 Wiley-Liss, Inc. Cilia are among the oldest of known cellular organelles, first described by van Leeuenhoek in 1675 (Satir, 1995). Motile cilia function as force-generating organelles producing a complex waveform that results in movement of fluid across an epithelium. In addition, it is now well established that on most cells of the body there also exists a single nonmotile primary cilium that acts as an antenna sensing the extracellular environment, be it the lumen of a tubular organ or the extracellular matrix of connective tissues. The primary cilium ultimately transduces stimuli that result in gene expression controlling fundamental biological responses of the cell. (Recent reviews include Avidor-Reiss et al., 2004; Quarmby and Parker, 2005; Caspary et al., 2007; Satir and Christensen, 2007.) In the past 10 years, a significant understanding has developed of the role of the primary cilium as a sensory organelle in epithelial cells throughout the body. The greatest breakthroughs have come from discoveries of abnormalities of the primary cilium linked to specific diseases, such as polycystic kidney disease in young children. (Recent reviews include Zhang et al., 2004; Davenport and Yoder, 2005; Badano et al., 2006; Bergmann et al., 2006; Bisgrove and Yost, 2006; Zariwala et al., 2007.) For parallel advancements to be made in understanding the functional role of the primary cilium in cells of connective tissues, methodology must be developed that allows easy detection, in vitro and in vivo, of primary ciliary presence, incidence, and orientation. In epithelia, the primary cilium of each cell projects into the lumen of the organ or to the surface of a monolayer culture. Experimental analysis of primary cilia in cells of connective tissues remains a significant challenge because (1) monolayer cultures are an artificial environment for most connective tissue cells, such as chondrocytes or tenocytes, that exist in vivo in a decidedly three-dimensional (3D) environment; and (2) in 3D culture systems or in vivo the primary cilium of a connective tissue cell projects not into the lumen of on organ where it is exposed for reaction with localizing antibodies, but rather into the surrounding extracellular matrix (ECM). Primary cilia have been described in multiple connective tissue cells. Examples include chondrocytes in elastic, articular, mandibular, sternal, and growth plate cartilage (Scherft and Daems, 1967; Cox and Peacock, 1977; Wilsman, 1978; Wilsman and Fletcher, 1978; Poole et al., 1985, 1997, 2001; Ascenzi et al., 2007); fibroblasts in ligaments, meniscus, perichondrium, and tendons (Hellio Le Graverand et al., 2001; Bray et al., 2005; and K. Kadler, personal communication); cells associated with a mineralizing matrix such as ameloblasts, osteoblasts, osteocytes, and odontoblasts (Garant et al., 1968; Tonna and Lampen, 1972; Federman and Nichols, 1974; Couve, 1986; Warshawsky, 1968); as well as in adipocytes (Geerts et al., 1990) and cells of the periodontal ligament (Beertsen et al., 1975). The basic structure of a primary cilium consists of an axoneme surrounded by plasma membrane and composed of nine outer doublet microtubules, but lacking a pair of central microtubules as found in motile cilia (9+0 vs. 9+2). The axoneme is contiguous with the basal body, a cylindrical structure of nine microtubular triplets that organizes the microtubules of the ciliary axoneme. The basal body, in turn, is associated with a centriole in the cytoplasm (Wilsman and Farnum, 1983; Badano et al., 2005; Yang et al., 2005). The axoneme has a variety of specialized proteins for anterograde and retrograde transport from the interior of the cell to the axonemal tip, as well as for signal transduction (Pazour and Rosenbaum, 2002; Blacque et al., 2005; Corbit et al., 2005; Germino, 2005; Gherman et al., 2006; Davis et al., 2006; Follit et al., 2006; Inglis et al., 2006; Michaud and Yoder, 2006; Scholey and Anderson, 2006; Singla and Reiter, 2006; Yoder, 2006; Haycraft et al., 2007; Song et al., 2007; Ruiz-Perez et al., 2007). The axonemal ultrastructure of primary cilia in articular chondrocytes was first described by Wilsman and Fletcher (1978), and is consistent with the morphology of nonmotile primary cilia from a wide range of cells, including epithelial cells, muscle cells, and neurons (recent reviews by Eley et al., 2005; Satir and Christensen, 2007). It also has been determined, by serial sectioning for transmission electron microscopy (TEM), that each articular chondrocyte has one cilium (Wilsman et al., 1980). In the past two decades, Poole and colleagues have been the principal champions of this organelle in articular chondrocytes, and have developed a model of the primary cilium, stressing its relationship with the Golgi apparatus as well as its projection into the pericellular matrix surrounding the chondrocyte (Poole et al., 1985, 1997, 2001; Jensen et al., 2004; McGlashan et al., 2006, 2007). The diameter of the axoneme of the primary cilium is typically ∼0.2 μm, putting it at the limit of resolution of brightfield microscopy. Even though the projecting axoneme may be as long as 10 μm, the primary cilium rarely is detected by routine light microscopy because it is unlikely that one section will contain the entire axonemal length and also because its profile is easily obscured by the ECM (Wilsman et al., 1980; Poole et al., 1985). Over the past decade primary cilia have been studied using imaging techniques that allow visualization by fluorescence microscopy, including confocal microscopy, in many cellular types in culture, including chondrocytic cultures, and in thin isolated slices of prefixed articular cartilage (Poole et al., 1997, 2001; Jensen et al., 2004; McGlashan et al., 2006; Song et al., 2007). This is possible because, of several hundred proteins found in cilia, a subset has been shown to be uniquely associated with primary cilia (Pazour and Witman, 2003; Marshall, 2004; Gherman et al., 2006). The ability to visualize primary cilia by fluorescence microscopy has meant that analysis of their presence both in normal cells and in cells in diseased organs is possible without the need for time-consuming serial sectioning for TEM. Our goal has been to develop a rapid in situ method for visualizing in situ the primary cilium in cells of a wide range of connective tissues, and to analyze its 3D spatial orientation. Our approach was to extend the work of Poole et al. (2001) and Jensen et al. (2004), who have studied the primary cilium of sternal and articular cartilage, to other connective tissues that differ from each other in such basic features as overall cellularity, cellular shape, spatial relationship of cells to each other, and features of the ECM such as composition and organization. A pre-embedment immunocytochemical approach, modified from that of Poole et al. (1997, 2001; Jensen et al., 2004) was developed and optimized for growth plate cartilage, a hyaline cartilage characterized by ovoid cells isolated from each other and an ECM whose composition is dominated by collagen II and aggrecan. This standard approach was then tested on several other connective tissues and modified specifically for tendon, a dense connective tissue characterized by irregularly shaped cells in a syncytium and an ECM of highly aligned fibers of collagen I (McNeilly et al., 1996; Canty et al., 2004). Multiphoton microscopy (MPM) allowed optical sectioning through the tissue to a depth of ∼200 μm and visualization of nonstained collagen fibrils (Williams et al., 2005). The long-term objective is to use this methodology to analyze the presence, incidence, and 3D orientation of the primary cilium of connective tissue cells, with the goal of using experimental manipulations to test the hypothesis that the primary cilium acts as a mechanosensory organelle involved in creating the specific ECM architecture associated with connective tissue organization. This methodology was adapted from Ward et al. (2003) and Chi et al. (2004) for visualization of primary cilia using immunocytochemistry of acetylated-α-tubulin. Wistar rats and C57BL/6J mice 3 to 5 weeks of age were kept under routine housing conditions, and all procedures were approved by the Institutional Animal Care Committee. Death was by intraperitoneal injection of an overdose of pentobarbital. For visualization of growth plate cartilage, the hindlimb was disarticulated at the coxofemoral joint, surrounding muscle was rapidly removed, and the distal limb was isolated by disarticulation at the femorotibial joint. The proximal half of the tibia was isolated and immersed in 100% methanol at 4°C, and all subsequent steps to isolate growth plate slabs were performed in 4°C methanol. Isolated slabs of growth plate cartilage were prepared by making a mid-sagittal cut of the tibia into two halves. Each half was additionally cut into two to three final slabs approximately 0.5–0.75 mm in thickness, each including the full height of the growth plate from the epiphyseal to the metaphyseal bone. Some slabs included the entire secondary center of ossification and the articular cartilage; for others, essentially all bone was trimmed leaving a piece consisting entirely of growth plate cartilage. Multiple variables, including antibody concentration, incubation time, and incubation temperature were tested to optimize the protocol for growth plate cartilage. The final protocol comprised fixation in methanol 4°C for 3 hr; wash 4× in 0.01 M phosphate buffered saline (PBS) at 25°C; incubation with primary antibody (monoclonal anti-acetylated-α-tubulin, Sigma), 1:100 at 2 hr at 25°C, followed by overnight at 4°C; wash 4× in PBS at 25°C; incubation with fluorescently-conjugated secondary antibody 1:100 (goat anti-mouse IgG [H+L], Molecular Probes) for 2 hr at 25°C, followed by overnight at 4°C; and wash 4× in PBS at 25°C. Slabs were stored at 4°C in PBS before imaging. It was established that a variety of secondary antibodies were equally effective at a 1:100 dilution, including Alexa Fluor 568, fluorescein (FITC), and rhodamine (TRITC; Molecular Probes). Also, primary antibody from Abcam (ab24610) was equally effective to that obtained from Sigma (T6793). Controls consisted of incubation with bovine serum replacing the primary antibody in the standard protocol. Several different connective tissues were collected and incubated following the standard procedure developed for growth plate cartilage. These included cranial cruciate ligament, meniscus, patellar tendon, intervertebral disc, and fibrocartilage of the perichondrium. Appropriate modifications of the dissection procedures were made with the consistent goal of isolating the tissue intact and separated from surrounding structures. As an example, for the intervertebral disc, essentially all surrounding bone and muscle were removed. The meniscus, patellar tendon and cranial cruciate ligament were each isolated from all other surrounding structures of the femorotibial joint. Perichondrium was imaged on slabs originally isolated for growth plate cartilage. Because of the consistent difficulty of visualizing primary cilia in connective tissues other than growth plate cartilage, the methodology was modified to address the challenges of localization in dense connective tissues, using tendon to develop the protocol. For the immunolabeling experiments, multiple variables were examined to achieve simultaneous imaging of cilia, nuclei, and collagen (Table 1). In the optimal protocol, 3-mm-long segments of common digital extensor tendons were collected bilaterally from 5-week-old rats. The tendon segments were then separated into individual fascicles after removal of the epitenon under a dissecting microscope. Fasicles were fixed for 2 hr in methanol at 4°C and permeabilized for 1 hr in 0.1% Triton X-100 at room temperature (RT). Tissues were incubated in mouse monoclonal anti-acetylated α-tubulin primary antibody (Abcam) for 2 hr at RT and then overnight at 4°C. Tissues were then incubated in goat anti-mouse Alexa Fluor 568 secondary antibody (Molecular Probes) overnight at 4°C. Finally, samples were stained with Hoechst 33258 (0.02 μg/mL) (Molecular Probes) for 30 minutes at 37°C to label cell nuclei. Negative controls underwent all tissue processing but were incubated in PBS during the primary antibody step. Table 1. Experimental variables for whole-mount immunohistochemistry to visualize primary cilia in tendon Multiphoton microscopy was used to visualize primary cilia and collagenous extracellular matrix. In this technique a fast-pulsed laser focused and scanned within the tissue excites both multiphoton fluorescence and second harmonic generation (SHG) (Zipfel et al., 2003). In two-photon excited fluorescence (TPEF) processes, two photons of equal energy interact with fluorescent molecules and produce an excitation equivalent to absorption of a single photon with twice the energy. Typically, infrared photons from the excitation laser excite fluorescent molecules in stains or secondary antibodies to produce visible fluorescence that is red-shifted from half the incident wavelength following molecular relaxation. In SHG processes, incident photons interact with noncentrosymmetric arrays of scattering centers and produce a beam of coherent light at twice the energy of the incident light (Williams et al., 2005). In our experiments, infrared photons from the excitation laser scatter from the ordered arrays of collagen molecules in the tissue and produce a coherent SHG signal at half the incident wavelength. In contrast to TPEF, the SHG signal is intrinsic to the tissue and does not require fluorescent dyes or stains. Both signals can be collected simultaneously to visualize the endogenous signal from the collagen and the TPEF signal from the fluorescently labeled organelles. Whole tissues were mounted in either PBS or Vectashield (Vector Labs) for imaging. Two similar MPM imaging systems were used, one for tendon, and one for all other tissues. The details of both microscopes have been described previously (Zipfel et al., 2003; Williams et al., 2005). Briefly, both systems relied on Ti:Sapphire lasers (Tsunami, Spectra Physics) for generating 100-fs pulses. Tendons were imaged with 800-nm excitation using a Bio-Rad 1024 scanner interfaced with an inverted Olympus IX-70 microscope and a 40×/1.15 NA water immersion objective (Olympus UApo/340). Three detector channels were equipped with filters corresponding to UV, blue, and visible emission (440 long pass dichroic (LPD) reflecting to a 390/70 emission filter (EF), 500 LPD to a BGG22 glass EF, and 670 LPD to a 580/150 EF; Chroma Technology), enabling the separation of endogenous SHG signal at 400 nm produced by the collagen, blue fluorescence from Hoechst-labeled cell nuclei, and yellow–green fluorescence from Alexa 568-labeled primary cilia. All other tissues were imaged with a Bio-Rad MRC-600 scanner interfaced with an upright (custom modified) Olympus AX-70 with 780-nm excitation and a 20×/0.75 NA air objective (Zeiss Flaur). Emission filters were chosen for a blue/green separation (BGG22 and 580/150 filters with a 500 LPD; Chroma Technology), allowing SHG at 395 nm to be separated from the various fluorescent antibody labels. Three-dimensional visualization of the tissue was achieved by raster scanning the laser within the tissue at 1- to 2-μm increments to collect image stacks. Samples of each connective tissue were prepared for light microscopy by immersion fixation for 2 hr in 2% glutaraldehyde/2% paraformaldehyde in 0.1 M cacodylate buffer, pH 7.3, with 0.7% ruthenium hexamine trichloride (RHT) (Hunziker et al., 1982). After rapid dehydration though graded ethanols and propylene oxide, samples were embedded in Epon-araldite and polymerized at 60°C for 3 days. No decalcification procedures were used. One-micron-thick sections were collected and stained with basic fuchsin/methylene blue/azure II. An additional collection of the entire femorotibial joint was made from three mice, fixed in 2% paraformaldehyde and embedded in paraffin. One-micron-thick sections were stained with Sirius Red F3BA (Electron Microscopy Sciences) and imaged using polarization microscopy (Junqueira et al., 1979). TEM images of primary cilia on growth plate chondrocytes were from a previous study of the distal radial growth plate of 4-week-old minipigs (Farnum and Wilsman, 1988). The primary fixative was the same as that used for light microscopy, followed by secondary fixation for two hours in 1% osmium tetroxide in 0.1 M cacodylate buffer, also containing 0.7% RHT. Sections approximately 60 nm thick were collected on formvar-coated grids, stained with uranyl-acetate-lead citrate, and viewed on a Philips 410 electron microscope at 60 kV. Multiphoton image stacks were processed in ImageJ (NIH) or Photoshop (Adobe Systems). Each channel was pseudocolored (blue for the collagen, red for the nuclei, and green for the primary cilia) and adjusted for optimal brightness and contrast. For the lower magnification images in which collagen thickness varied throughout the field of view and produced correspondingly dark and bright areas in the SHG signal, the shadows and highlights were adjusted to give a more uniform SHG brightness, and the images were despeckled. Figure 1a demonstrates the relationship of the proximal tibial growth plate of a 4-week-old rat to the epiphyseal and metaphyseal bone, and the surface articular cartilage. Staining is with alizarin red S (bone) and alcian blue (cartilage; Young et al., 2000). Figure 1b shows a 1-μm-thick histological section of the proximal tibial growth plate from a 4-week-old rat, stained with basic fuchsin/methylene blue/azure II. The proliferative and hypertrophic cell zones can be identified by the characteristic morphology of the chondrocytes. Although not visible with this staining, between the columns of cells collagen type II fibers are aligned essentially in parallel to the long axis of the bone. a: In this whole bone from a 4-week-old rat, cartilage of the growth plate (arrows) and of the articular surface stains with alcian blue, while bone of the secondary center (sc) and of the metaphysis (m) stains with Alizarin Red S. b: On a 1-μm-thick histological section the zonal architecture of the growth plate is seen, with columns of chondrocytes comprising a proliferative cell zone (p) and a hypertrophic cell zone (h). c: The challenge of visualizing the primary cilium by TEM is demonstrated here where the arrow points to a cross-sectional profile of a ciliary axoneme in a proliferative zone chondrocyte. A well-developed Golgi also is seen. d,e: The relationship of the basal body (arrow, bb) of the cilium to the associated centriole, at two different angles. f: The centriole is not visible, but the basal body (including the basal foot) is seen in longitudinal section with a short segment of axoneme projecting into the ECM. g: An almost complete profile of the ciliary axoneme paralleling the long axis of a proliferative zone chondrocytes is seen. Note also the extent of the juxtanuclear Golgi. All TEM images are of 60-nm-thick sections stained with uranyl acetate/lead citrate. Scale bar = 0.2 μm in e. Profiles of primary cilia in growth plate chondrocytes were identified on 60-nm-thick sections by TEM analysis. Figure 1c–g demonstrates the challenges associated with visualization of primary cilia by TEM in growth plate cartilage. The profiles of individual cilia varied depending upon the orientation of the cilium relative to the plane of section, as has been described for primary cilia in articular chondrocytes (Wilsman and Fletcher, 1978; Wilsman et al., 1980). It was unusual to be able to see profiles both of the cilium and of its associated centriole in one section. The low magnification view in Figure 1c demonstrates how cryptic the ciliary axoneme can be using TEM. In this micrograph of a proliferative zone chondrocyte, the complexity of the Golgi is illustrated and only the very tip of the ciliary axoneme can be identified in cross-section in the adjacent ECM. In Figure 1d, which is a section through the Golgi of a hypertrophic chondrocyte, grazing profiles of both the centriole and of the basal body are seen. In Figure 1e a cross-sectional profile of the centriole is seen orthogonal to the longitudinal section of the basal body of a primary cilium in a hypertrophic chondrocyte, and only a small section of the axoneme is seen projecting into the ECM. In Figure 1f the profile of the basal body is seen, but the centriole as well as essentially the entire ciliary axoneme are out of the plane of section. In Figure 1g the axoneme parallels the long axis of the proliferative zone chondrocyte, but the basal body and the associated centriole are not in the plane of section. The proximity of the cilium to the chondrocyte\'s juxtanuclear Golgi is seen in Figure 1c,d,g. Figure 2a demonstrates the morphology of a slice of tibial growth plate by light microscopy before incubation with the primary antibody. Epiphyseal and metaphyseal bone were left intact. Figure 2b is a similar slice after incubation in both the primary (anti-acetylated-α-tubulin) and secondary (fluorescein-labeled goat-anti-mouse) antibodies seen under epifluorescence microscopy. At this magnification the growth plate appears dark while the epiphyseal and metaphyseal bone are green, due to nonspecific absorbance of the secondary antibody onto the mineralized surface. Figure 2c is a section viewed by MPM near to the cut surface of an isolated slice from a murine proximal tibial growth plate demonstrating localization of chondrocytic cilia. The alignment of cells in columns, the generally ellipsoid shape of individual cells, and the separation of cells from each other were all features of growth plate cellular organization that allowed one to identify primary cilia as small positive-labeled dots. In an image taken at the surface of the cut slab, such as Figure 2c, there often was some degree of nonspecific staining in the matrix between columns of cells (diffuse streaks of green in the matrix). This was greater in murine growth plates than in rat growth plates. However, by its morphology this nonspecific staining could be distinguished from positive staining in cilia. Also, nonspecific surface staining disappeared after imaging deeper into the tissue in the z-direction. a,b: Slabs of growth plate cartilage with intact bone on either side are shown before (a) and after (b) pre-embedment immunocytochemistry for acetylated-α-tubulin. In a, the growth plate lies between the epiphyseal bone (e) and metaphyseal bone (m). In b, the growth plate appears as a dark, unstained region; significant nonspecific surface staining of the mineralized bone is evident. c: In the murine growth plate, positive identification of primary cilia (arrows) appeared as punctate dots; cells were barely visible. At the surface there was nonspecific staining in the longitudinal septae (*), but this did not penetrate more than a few microns into the section. d: In one focal volume ciliary profiles were usually seen in less than half of the cells, as demonstrated in this field of late proliferative and early hypertrophic chondrocytes. e,f: Examples could be found of cilia projecting from the same side of neighboring cells (e) as well as from opposite sides (f). Although most axonemal profiles were straight, examples of curved axonemes were seen (e). g: Positive Golgi staining often was seen; its morphology and intracellular location clearly distinguished it from ciliary staining, shown in this field of proliferative zone cells. Figure 2d–g is from the proximal tibial growth plates of 4-week-old rats. All panels are oriented with the epiphyseal end of the growth plate at the top of the image. Figure 2d–g exemplifies positive identification of primary cilia in the rat proximal tibial growth plate. Individual cellular profiles are seen, and there is no strong second harmonic signal from collagen in the ECM. In a given field of early hypertrophic zone cells imaged in a 2-μm-thick focal plane, ciliary profiles could be seen in less than half of the chondrocytes (Fig. 2d). By imaging deeper at 2-μm steps, cilia come into view in additional cells. In Figure 2e, cilia in adjacent hypertrophic chondrocytes project from the same side of their respective cells and have the same general orientation; the cilium of one chondrocyte clearly is curved. In Figure 2f cilia in two proliferative chondrocytes project from opposite sides of their respective cells, but both are similarly aligned relative to the longitudinal axis of the bone. Figure 2g demonstrates one chondrocyte with a cilium projecting orthogonal to the longitudinal axis of the bone. The positive cytoplasmic staining in three chondrocytes in this series is acetylated-α-tubulin in the Golgi, as seen in a grazing profile. The morphology of Golgi staining allows one to discriminate it from that of the ciliary axoneme, both because the Golgi staining is entirely intracellular and because of the characteristic rod-like shape of the ciliary axoneme. As examples, the chondrocyte at the top of the picture shows Golgi staining but no ciliary staining; the fourth cell in the column shows both Golgi staining and ciliary staining. Negative controls lacked both ciliary and Golgi-related staining (data not shown). The meniscus is a concave semilunar disc attached to the proximal tibial plateau by a series of ligaments. It is considerably thicker laterally than medially. The collagen component of the ECM of the meniscus consists of both types I and II, differing in their regional localization (Hellio Le Graverand et al., 2001; Kambic and McDevitt, 2005). The complexity of the cellular and ECM organization of the knee meniscus is seen after staining with Sirius Red and viewed under plane polarized light (Fig. 3a) (Junqueira et al., 1979). Collagen fibers within the meniscus change alignment abruptly, but a consistent circumferential alignment generally follows the curved contour of the adjacent articular surfaces as seen in the MPM image in Figure 3b showing the relationship of collagen orientation in the meniscus to the articular surface of the distal femur. The orientation of collagen fibers also varies with depth of the tissue (Xie et al., 2006). This changing orientation of the collagen fibers of the meniscal ECM made orientation during MPM imaging more challenging than for the growth plate. In addition, meniscal fibroblasts do not have a single characteristic shape, nor are they clearly spatially separated from each other, which also presents a challenge for visualization by MPM a: The complexity of meniscal collagen organization can be seen at the histological level after staining with Sirius Red and viewing under plane polarized light (4-week-old mouse). The meniscus (m) lies between articular cartilage on the surface of the epiphyseal bone of the distal femur (df) and of the proximal tibia (pt). The two semilunar discs of the meniscus are thicker laterally (left side of image) than medially (right side of image). b: Using MPM, the circumferential orientation of collagen fibers (SHG pseudocolored blue) of the meniscus (m) in relationship to the articular cartilage (chondrocytes pseudocolored green) of the femoral condyle can be seen. The joint space (j) occupies the region between the distal femur and the meniscus. c: At low magnification, cilia (arrows) on meniscal fibrochondrocytes (green) are more easily visualized medially (top of image), where the SHG collagen signal is less intense, in contrast to the lateral aspect of the meniscus (bottom of image), where the SHG signal of the circumferential collagen fibers (*) is relatively intense. d: At higher magnification, axonemal profiles (bright green rods) in multiple orientations can be seen on meniscal cells. Figure 3c,d shows positive staining for cilia using FITC secondary antibody in the murine meniscus as viewed by MPM. In these figures the SHG signal, which allows visualization of collagen orientation, is pseudocolored blue, and ciliary staining is pseudocolored green. In Figure 3c cilia can best be seen as fluorescent dots on the inner aspect of the meniscus where the collagen is less dense. This staining was absent in controls. Figure 3d demonstrates positive ciliary staining, with the cilia projecting in different directions, in the outer section of the meniscus. The homogeneous green staining is nonspecific surface staining of the isolated slice. These images demonstrate that visualization of primary cilia was considerably more difficult for the meniscus than for the growth plate, and also there were regional differences based primarily on the organization of the collagen of the ECM. Figure 4a demonstrates the perichondrium surrounding the cartilaginous growth plate of the proximal tibia of a 4-week-old mouse. The perichondrium blends into the periosteum of the epiphyseal and metaphyseal bone. At a higher magnification (Fig. 4b), the complexity of the perichondrium can be seen. Most laterally, type I collagen fibers are highly aligned parallel to the long axis of the bone. Irregularly shaped perichondrial fibroblasts stain dark blue and are found between the collagen fibers. Figure 4c is an MPM image of the perichondrium adjacent to the growth plate; the perichondrium blends into the periosteum adjacent to the epiphyseal bone. The collagen SHG demonstrates the complex orientation of collagen fibers of the perichondrium and periosteum. Figure 4d is localization of primary cilia in the perichondrium. Similar to the meniscus, the density of collagen fibers, their close association with cells, and the irregular shape of the cells themselves made visualization difficult. Three examples of ciliary axonemes, seen as fluorescent spots with positive staining, are seen in the lower part of the image. Nevertheless, it would be very difficult to quantitatively assess ciliary incidence in this tissue without further methodological optimization a: The perichondrium (p) surrounds the growth plate and blends into the periosteum of the bone (4-week-old mouse). b: The alignment of collagen fibrils can be seen in the 1-μm-thick section stained with basic fuchsin/azure II/methylene blue. Perichondrial fibroblasts can be identified by their dark-staining nuclei. c,d: In MPM images, the collagen SHG signal demonstrates the complexity of collagen fiber alignment. gp, growth plate; e, epiphyseal bone. The SHG signal is pseudo-colored blue in 4d. Although positive staining for cilia is present, it is difficult to make any assessment of incidence given the density of the collagen fibrils. The circular anulus fibrosus of the intervertebral disc is composed of dense type I collagen fibers. It surrounds a central nucleus pulposus (Fig. 5a). The primary collagen of the latter is type II. The structural organization of the anulus can be seen on a 1-μm-thick section of a rat intervertebral disc in Figure 5b, contrasting the highly aligned collagen of the anulus with the amorphous appearance of the nucleus. Figure 5c is an MPM image focusing specifically on rings of the anulus and demonstrating the complexity of ECM organization with layered changes in collagen organization and the relationship of disc fibroblasts to the ECM. There was significant nonspecific staining in the nucleus (Fig. 5d). This, coupled with its hypocellularity and its fragility under the MPM laser, made it impossible to assess the nucleus for the presence of primary cilia. It was possible to identify ciliary staining for acetylated-α-tubulin in cells of the anulus (Fig. 5e–g), but it was essentially impossible to follow this in the z-direction because of a rapid loss of signal. Whether this was due to the characteristics ofthe organization of the ECM or lack of penetration of antibody is unclear a: The intervertebral disc consists of a dense circumferential anulus fibrosus (a) surrounding a central soft nucleus pulposus, as shown in the wet specimen of a murine intervertebral disc. b: On a 1-μm-thick section stained with basic fuchsin/methylene blue/azure II, the amorphous nucleus (n) contrasts sharply with the highly aligned collagen fibers and cells of the anulus (a). This section is adjacent to bone of the vertebral body (b). c: Under MPM the SHG of collagen demonstrates the complexity of the collagen fiber alignment of the anulus and its relationship to disc fibroblasts, which appear as dark circular areas between bright lamellae. d: A high level of nonspecific staining precluded assessment of the presence of cilia in the nucleus (n). e–g: However, positive staining for cilia could be demonstrated in cells of the anulus, shown as punctate dots (arrows). Results for imaging primary cilia by this methodology were similar for the cranial cruciate ligament (Fig. 6a) and patellar tendon (Fig. 6b), as had been the case for meniscus, disc and perichondrium. In both the cranial cruciate ligament and the patellar tendon, positive staining for primary cilia was visible surrounded by an ECM of type I collagen fibers, but it was sparse and difficult to associate with cells. It was not possible to identify cellular boundaries, and there was a significant amount of nonspecific staining. a,b: Positive ciliary staining was visible as green dots in both the cranial cruciate ligament (a) and the patellar tendon (b). However, there was a significant amount of nonspecific staining, and it was essentially impossible to identify cellular boundaries. c: Rows of adjacent tenocytes between collagen bundles can be seen in a 1-μm-thick section of the common digital extensor tendon of the antebrachium stained with basic fuchsin/methylene blue/azure II. Tenocyte nuclei stain blue, with dark nucleoli, and the juxtanuclear Golgi (arrows) are identified by the lack of any uptake of the stain. This extensor tendon was used as the model system for the development of a methodology suitable for reliable ciliary staining in dense collagenous connective tissues. The three primary barriers to localization in all connective tissues analyzed, compared with growth plate cartilage, were the dense collagen I fiber alignment, the lack of an easily identifiable shape to the cells, and the lack of separation between cells. Also, given the density of the ECM, it was less clear that the antibodies were able to penetrate successfully, and there appeared to be a higher level of diffuse nonspecific staining compared with growth plate cartilage. It became clear that modifications of the basic methodology would be required for ciliary localization in dense connective tissues. For this, the common digital extensor tendon of the antebrachium was chosen as the model system. Figure 6c shows a 1-μm-thick section of this extensor tendon. Elongated tenocytes can be seen lined up adjacent to each other between the highly aligned type I collagen fibers. In this image, tenocyte nuclei stain blue, and the Golgi of individual cells can be seen adjacent to the nucleus as a nonstained area. Using specific modifications of the protocol described under Materials and Methods, primary cilia, tenocyte nuclei, and collagenous ECM were simultaneously visible in samples double-labeled with the α-tubulin primary antibody and Hoechst, enabling a clearer view of cellular position and ciliary orientation. Oblong nuclei were arranged between parallel bundles of collagen fibers (Fig. 7a, four panels). Profiles of axonemes of primary cilia were observed both paralleling the long axis of the cell and projecting into the surrounding ECM at a range of angles (Fig. 7b, four panels). Their in-plane lengths ranged from ∼1 to 5 μm. Cilia were frequently observed in close proximity to a nucleus and could be followed continuously in z-series throughout several frames (Fig. 7c, five panels). In the negative controls incubated without the α-tubulin primary antibody, cell nuclei and collagen were clearly visible, but no signal from the Alexa 568 secondary antibody was detected (data not shown). a,b: Simultaneous collection of three signals, endogenous SHG from collagen, blue fluorescence from Hoechst nuclear stain, and yellow–green fluorescence from Alexa 568 secondary antibody to α-tubulin primary antibody, allowed visualization of extracellular matrix, tenocyte nuclei, and primary cilia, respectively. c: A z-series of a tendon fascicle showing several nuclei and primary cilia in serial optical sections at 1-μm intervals. Collagen, nuclei, and primary cilia are pseudocolored blue, red, and green, respectively. The objective of this study was to develop a rapid method for visualizing the primary cilium of a wide range of connective tissue cells. A versatile experimental approach combining whole-mount immunohistochemistry and MPM allowed visualization of primary cilia and collagen through thick sections of a variety of connective tissues. We successfully imaged axonemes of primary cilia projecting into their native ECM in growth plate cartilage, tendon, ligament, meniscus, intervertebral disc, and perichondrium. To our knowledge, these are the first observations of primary cilia in intervertebral disc and the first published reports of primary cilia in tendon. In addition, this is the first time that localization of primary cilia, other than by TEM, has been described for ligaments or meniscus. In recent studies primary cilia were localized on paraffin-embedded sections in synchondroses and associated perichondrium, as well as on paraffin-embedded and frozen sections in growth plate cartilage (Koyama et al., 2007; Haycraft et al., 2007; Song et al., 2007). Even in dense connective tissues such as meniscus and tendon, antibody penetration into the tissue was sufficient to visualize cilia at depths of ∼200 μm, or approximately 40 cell layers. Tissue organization and ECM density are key factors that affect ease and quality of imaging, and our approach required optimization for each tissue. Overnight incubation times were typically required to ensure antibody penetration throughout the tissue and to minimize nonspecific staining. The use of thin tissues whenever possible, for example, the common digital extensor tendon rather than the common calcanean tendon, improved the extent of primary antibody penetration. All secondary antibodies examined were similarly effective. In the multi-labeling experiments in tendon, careful selection of secondary antibody conjugation, stains, emission filters, and incident wavelength was essential to achieving three well separated signals showing collagen, nuclei, and primary cilia. Multiphoton microscopy enabled excellent rapid imaging of cilia, nuclei, and ECM. The ability to focus the laser into the tissue in the z-direction allowed subsurface imaging ∼200 μm from the cut surface and epithelial layers, where nonspecific staining sometimes occurred. It also allowed visualization of the entire cilium throughout several frames, as opposed to the grazing sections usually obtained in single TEM sections. As has been reported for articular cartilage, the probability of finding a ciliary profile in a given cellular profile in a TEM section is only 1:100; therefore, assessment of incidence or orientation can only be done by time-consuming and technically challenging serial section analysis (Wilsman et al., 1980). The probability of seeing a ciliary profile is even less for hypertrophic cell zone chondrocytes, because the cells enlarge as much as ten-fold while the size of the cilium remains constant. Although the resolution of MPM images (∼0.4 μm in-plane and ∼2 μm in z; Zipfel et al., 2003) does not approach that of TEM, sample preparation and organization of 3D data sets using MPM is far less laborious, and \"serial sectioning” can be done optically. A key advantage of MPM relative to confocal microscopy, which has been used in immunohistochemical studies of primary cilia in chondrocytes of articular and sternal cartilage (Poole et al., 1997, 2001; Jensen et al., 2004; McGlashan et al., 2006), is that the endogenous SHG signal can be used to visualize collagen orientation in the tissue. This capability is particularly important in tissues with dense collagenous extracellular matrices, including tendon and meniscus, where spatial relationships between primary cilia, cells, and ECM may be relevant to mechanosensing and tissue formation. In addition, MPM photobleaches only within each imaging plane during image acquisition, rather than throughout the stack, as in confocal microscopy. Minimization of photobleaching is critical for identification of structures in 3D. Connective tissues such as tendon, bone, cartilage, intervertebral disc, and meniscus respond to mechanical forces with changes in metabolism, ultrastructure, and mechanical properties. As a corollary, mechanical inputs beyond those considered to be within the normal physiological range are a major cause of degeneration of connective tissues through both up- and down-regulation of multiple genes, including those responsible for synthesizing ECM components (Banes et al., 2001; Elder et al., 2001; Chen, 2003; Lammi, 2004; Niehoff et al., 2004; Fong et al., 2005). Experimental studies have focused on these responses at multiple levels of biological organization. Recent examples include demonstration of the ability of fibroblasts of the meniscus to respond to biomechanical loading or oscillatory fluid flow by changes in specific gene expression (Deschner et al., 2006; Upton et al., 2006; Eifler et al., 2006) and demonstration that disc cells respond to excessive mechanical stimulation by altering regulation of multiple ECM-producing genes, ultimately leading to degenerative changes in the ECM (Omlor et al., 2006; Roberts et al., 2006). Primary cilia have been implicated in such mechanotransduction processes in connective tissues (Poole et al., 1985; Whitfield, 2003; Xiao et al., 2006), but the mechanisms by which connective tissue cells sense mechanical loads and convert them to biochemical signals that lead to tissue formation and adaptation are still incompletely understood. A recent study of primary cilia in osteoblasts and osteocytes demonstrated that primary cilia are essential for key cellular bone formation and resorption responses to dynamic fluid flow (Malone et al., 2007). Furthermore, it has been suggested that the primary cilium may play a role in establishing cellular orientation and directing secretion of ECM molecules from the Golgi in response to biomechanical stimuli, because the primary cilium is located in close proximity to the Golgi apparatus (Poole et al., 1997). Therefore, the orientation of the primary cilium may relate to the establishment of anisotropic ECM organization in connective tissues. Multiple issues must be addressed before the hypotheses regarding the potential mechanosensory function of the primary cilium in normal and pathological growth, development, and adaptation of connective tissues can be tested. Although refinements of our methodology are required to meet the challenges of each individual connective tissue being studied, the MPM analyses presented here provide a flexible means of quantifying the incidence and spatial orientation of primary cilia in a range of connective tissue cells. Once these features have been characterized in normal tissues, experimental manipulations can then be used to test the hypothesis that the primary cilium acts as a mechanosensory organelle involved in creating the specific ECM architectures associated with connective tissue organization. The authors thank Wenhua Liu and Pat Fisher for help with the micrographs; Linda Jones and Barbara Linnehan for help with preparing the bibliography; and Jill Urban and Karl Kadler for helpful discussions.2007. Analysis of the orientation of primary cilia in growth plate cartilage: a mathematical method based on multiphoton microscopical images.J Struct Biol 158: 293– 206. Avidor-Reiss T,Maer AM,Koundakjian E,Polyanovsky A,Keil T,Subramaniam S,Zuker CS.2004. Decoding cilia function: defining specialized genes required for compartmentalized cilia biogenesis. 117: 527– 539. The ciliopathies: an emerging class of human genetic disorders.Annu Rev Genom Hum Genet 125– 148.2001. Mechanical forces and signaling in connective tissue cells: cellular mechanisms of detection, transduction, and responses to mechanical deformation.Curr Opin Orthop 389– 396.1975. Frequency of occurrence and position of cilia in fibroblasts of the periodontal ligament of the mouse incisor.Cell Tissue Res 163: 415– 431. Bergmann C,Frank V,Kupper F,Kamitz D,Hanten J,Berges P,Mager S,Moser M,Kirfel J,Buttner R,Senderek J,Zerres K.2006. Diagnosis, pathogenesis, and treatment prospects in cystic kidney disease.Mol Diagn Ther 163– 174. Blacque OE,Peren E,Boroevich KA,Inglis PN,Li C,Warner A,Khattra J,Holt RA,Ou G,Mah AK,McKay SJ,Huang P,Swoboda P,Jones SJM,Marra MA,Baillie DL,Moerman DG,Shaham S,Leroux MR.2005. Functional genomics of the cilium, a sensory organelle.Curr Biol 935– 941.2004. Coalignment of plasma membrane channels and protrusions (fibripositors) specifies the parallelism of tendon.J Cell Biol 165: 553– 563.2004. Communication between paired chondrocytes in the superficial zone of articular cartilage.J Anat 205: 363– 370. Wiley Online Library2005. An incredible decade for the primary cilium: a look at a once-forgotten organelle.Am J Physiol Renal Physiol 280: F1159– F1169.2006. The emerging complexity of the vertebrate cilium: new functional roles for an ancient organelle.Dev Cell 9– 19. Regulation of RANKL by biomechanical loading in fibrochondrocytes of meniscus.J Biomech 1796– 1803.2006. Oscillatory fluid flow regulates glycosaminoglycan production via in intracellular calcium pathway in meniscal cells.J Orthop Res 375– 384. Wiley Online Library2001. Chondrocyte differentiation is modulated by frequency and duration of cyclic compressive loading.Ann Biomed Eng 476– 482.1988. Lectin-binding histochemistry of intracellular and extracellular glycoconjugates of the reserve cell zone of growth plate cartilage.J Orthop Res 166– 179. Wiley Online Library2006. The intraflagellar transport protein IFT20 is associated with the Golgi complex and is required for cilia assembly.Mol Biol Cell 3781– 3792. Fong KD,Trindade MC,Wang Z,Nacamuli RP,Pham H,Fang TD,Song HM,Smith RL,Longaker MT,Chang J.2005. Microarray analysis of mechanical shear effects on flexor tendon cells.Plast Reconstr Surg 116: 1393– 1404. Ultrastructure and function of hepatic fat-storing and pit cells.J Electron Microsc Tech 247– 256. Wiley Online Library2006. The ciliary proteome database: an integrated community resource for the genetic and functional dissection of cilia.Nat Genet 961– 962.2001. The cells of the rabbit meniscus: their arrangement, interrelationship, morphological variations and cytoarchitecture.J Anat 198: 525– 535. Wiley Online Library1982. Improved cartilage fixation by ruthenium hexamine trichloride (RHT): a prerequisite for morphometry in growth cartilage.J Ultrastruct Res 1– 12.2004. Ultrastructural, tomographic and confocal imaging of the chondrocyte primary cilium in situ.Cell Biol Int 101– 110. Wiley Online Library1979. Picrosirius staining plus polarization microscopy, a specific method for collagen detection in tissue sections.Histochem J 447– 455. Koyama E,Young B,Nagayama M,Shibukawa Y,Enomoto-Iwamoto M,Iwamoto M,Maeda Y,Lanske B,Song B,Serra R,Pacifici M.2007. Conditional Kif3a ablation causes abnormal hedgehog signaling topography, growth plate dysfunction, and excessive bone and cartilage formation during mouse skeletogenesis.Development 134: 2159– 2169.2007. Primary cilia mediate mechanosensing in bone cells by a calcium-independent mechanism.Proc Natl Acad Sci U S A 104: 13325– 13330.2004. Human cilia proteome contains homolog of zebrafish polycystic kidney disease gene qilin.Curr Biol R913– R914.2006. Localization of extracellular matrix receptors on the chondrocyte primary cilium.J Histochem Cytochem 1005– 1014.2007. Articular cartilage and growth plate defects are associated with chondrocyte cytoskeletal abnormalities in Tg737orpk mice lacking the primary cilia protein polaris.Matrix Biol 234– 246.1996. Tendon cells in vivo form a three dimensional network of cell processes linked by gap junctions.J Anat 189: 593– 600.2004. Adaptation of mechanical, morphological, and biochemical properties of the rat growth plate to dose-dependent voluntary exercise. 899– 908.2006. Changes in gene expression and protein distribution at different stages of mechanically induced disc degeneration-An in vivo study on the New Zealand white rabbit.J Orthop Res 385– 392. Wiley Online Library1985. Analysis of the morphology and function of primary cilia in connective tissues: a cellular cybernetic probe? Cell Motil 175– 193. Wiley Online Library1997. Confocal analysis of primary cilia structure and colocalization with the Golgi apparatus in chondrocytes and aortic smooth muscle cells.Cell Biol Int 483– 494. Wiley Online Library2001. The differential distribution of acetylated and detyrosinated alpha-tubulin in the microtubular cytoskeleton and primary cilia of hyaline cartilage chondrocytes.J Anat 199: 393– 405. Wiley Online Library Ruiz-Perez VL,Blair HJ,Rodriguez-Andres ME,Blanco MJ,Wilson A,Liu Y-N,Miles C,Peters H,Goodship J.2007. Evc is a positive mediator of Ihh-regulated bone growth that localizes at the base of chondrocyte cilia.Development 134: 2903– 2912.2007. Development of the post-natal growth plate requires intraflagellar transport proteins.Dev Biol 305: 202– 216. Electron microscopy of aging skeletal cells. I. Centrioles and solitary cilia.J Gerontol 316– 324.2006. Biaxial strain effects on cells from the inner and outer regions of the meniscus.Connect Tissue Res 207– 214. Ward CJ,Yuan D,Masyuk TV,Wang X,Punyashthiti R,Whelan S,Bacallao R,Torra R,LaRusso NF,Torres VE,Harris PC.2003. Cellular and subcellular localization of the ARPKD protein; fibrocystin is expressed on primary cilia.Hum Mol Genet 2703– 2710. Xiao Z,Zhan S,Mahlios J,Zhou G,Magenheimer BS,Guo D,Dallas SL,Maser R,Calvet JP,Bonewald L,Quarles LD.2006. Cilia-like structures and polycystin-1 in osteoblasts/osteocytes and associated abnormalities in skeletogenesis and Runx2 expression.J Biol Chem 30884– 30895.2006. Use of polarization-sensitive optical coherence tomography to determine the directional polarization sensitivity of articular cartilage and meniscus.J Biomed Opt 064001. Yang J,Gao J,Adamian M,Wen X-H,Pawlyk B,Zhang L,Sanderson MJ,Zuo J,Makino CL,Li T.2005. The ciliary rootlet maintains long-term stability of sensory cilia.Mol Cell Biol 4129– 4137.2006. More than just the postal service: novel roles for IFT proteins in signal transduction.Dev Cell 541– 542.2000. Large-scale double-staining of rat fetal skeletons using Alizarin Red S and alcian blue.Teratology 273– 276. Wiley Online Library Please check your email for instructions on resetting your password. If you do not receive an email within 10 minutes, your email address may not be registered, and you may need to create a new Wiley Online Library account. Can\'t sign in? Forgot your username? Enter your email address below and we will send you your username If the address matches an existing account you will receive an email with instructions to retrieve your username