SEARCH

SEARCH BY CITATION

Keywords:

  • primary cilium;
  • chondrocyte;
  • 3D orientation;
  • articular cartilage;
  • axoneme

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. LITERATURE CITED
  9. Supporting Information

Primary cilia have functions as sensory organelles integral to signal transduction and establishment of cell polarity. In articular cartilage the primary cilium has been hypothesized to function as an antenna to sense the biomechanical environment, regulate the secretion of extracellular matrix components, and maintain cellular positional information, leading to high tissue anisotropy. We used analysis of electron microscopy serial sections to demonstrate positional attributes of the primary cilium of adult equine articular chondrocytes in situ. Data for ∼500 axonemes, comparing superficial to radiate chondrocytes from both load-bearing and non-load-bearing regions, were graphed using spherical co-ordinates θ, φ. The data demonstrate the axoneme has a definable orientation in 3D space differing in superficial and radiate zone chondrocytes, cells that differ by 90° in the orientation of their major axes to the articular surface. Axonemal orientation is more definable in load-bearing than in non-load-bearing areas. The position of emergence of the axoneme from the cell also is variable. In load-bearing regions of the superficial zone, extension of the axoneme is from the cellular side facing the subchondral bone. In radiate zone cells, axonemes extend from either face of the chondrocyte, that is, both toward the articular surface or toward the subchondral bone. In non-load-bearing regions this consistency is lost. These observations relate to current hypotheses concerning establishment of tissue anisotropy in articular cartilage during development, involving both migration of cells from the joint periphery and a restricted zone of division within the tissue resulting in the columnar arrangement of radiate zone cells. Anat Rec, 2011. © 2011 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. LITERATURE CITED
  9. Supporting Information

Recent experimental investigations have established essential functions for vertebrate primary cilia as sensory organelles integral to signal transduction pathways and to the establishment of epithelial cell polarity. Potential sensory functions of the primary cilium have been investigated over the last decade, with a major focus on the primary cilium in kidney tubule epithelial cells where ciliary dysfunction can lead to early-onset polycystic kidney disease. In the kidney, fluid flow within the tubule lumen leads to passive bending of the ciliary axoneme, which in turn activates multiple signal transduction pathways (Pazour and Witman,2003; Lehman et al.,2008; Berbari et al.,2009; Cardenas-Rodriguez and Badano,2009; Gerdes et al.,2009; Nigg and Raff,2009; Zhou,2009; Deltas and Papagregoriou,2010; Goetz and Anderson,2010; Hoyer-Fender,2010; Menezes and Germino,2010; Moser et al.,2010; Rydholm et al.,2010; Satir et al.,2010).

The extension of the ciliary axoneme of kidney epithelial cells into a matrix-free fluid space contrasts with what is known about the primary cilium in cartilaginous tissues. In articular cartilage the ciliary axoneme of some chondrocytes is observed by electron microscopy (EM) to be invaginated deep within a membrane-bound in-pocketing of the plasma membrane. In other cells, the axoneme can be seen with its full length extending into the extracellular matrix (ECM) with attachments of the axonemal membrane to the ECM (Wilsman,1978; Wilsman et al.,1980; Poole et al.,1985; Jensen et al.,2004; McGlashan et al.,2006; Knight et al.,2009). Interestingly, within a population of chondrocytes, the ciliary axoneme can be seen at different angles relative to the major axis of the chondrocyte. How the angle of extension is established and whether, for a given chondrocyte, it remains constant once established, has not been investigated. Additionally, it is not known whether the chondrocytic ciliary axoneme responds to fluid flow within the cartilaginous ECM with movement of the axoneme relative to the chondrocyte as seen in the passive bending of cilia in epithelial cells (Schwartz et al.,1997; Jensen et al.,2004; Rydholm et al.,2010). However, both straight and curved axonemes have been observed in chondrocytes (Jensen et al.,2004; Donnelly et al.,2008; Farnum et al.,2009). Therefore, while it is established that the ciliary axoneme of epithelial cells has directional flexibility in three-dimensional (3D) space, the degree of comparable flexibility for primary cilia with direct attachments to an ECM is unknown.

Several well-established observations form the basis for a dominant hypothesis concerning the function of the primary cilium in connective tissues, including articular cartilage. First, the high anisotropy of cellular and ECM organization in these tissues must be maintained for normal function (Eggli et al.,1988; Hunziker et al.,2002). Second, articular cartilage exhibits exquisite mechanosensitivity and there is experimental evidence that the primary cilium can function as a mechanotransducer in vivo and in vitro (Banes et al.,2001; Chen,2003; Delmas,2004; Poirier and Iglesias,2007; Anderson et al.,2008; Knight et al.,2009; Papachristou et al.,2009; McGlashan et al.,2010; Dalabiorgou et al.,2010; Deltas and Papagregoriou,2010; Kwon et al.,2010; Temiyasathit and Jacobs,2010; Xiao and Quarles,2010). Third, the location of the primary cilium in close association with the Golgi suggests a potential relationship between the cilium and the direction of cellular secretion of ECM components and/or the maintenance of cellular shape via cytoskeletal microtubules. The microtubular organizing center from which the cilium emerges is required to maintain cellular shape, which in turn regulates tissue organization (Poole et al.,1985,1997; Jensen et al.,2004; Haycraft et al.,2005,2007; McGlashan et al.,2007; Anderson et al.,2008; Haycraft and Serra,2008).

Based on these observations, the primary cilium of chondrocytes has been hypothesized to function as a cellular antenna able to sense the surrounding biomechanical environment and perhaps to regulate the directed secretion of ECM components and maintain cellular positional orientation, thus establishing tissue anisotropy (Marshall and Nonaka,2006; Jensen et al.,2004; Anderson et al.,2008; Haycraft and Serra,2008; Serra,2008; Knight et al.,2009). This hypothesis predicts that signal directionality from the surrounding biomechanical environment is integrated through the primary cilium, for processing within an individual cell or among cells (Whitfield,2008). The ability to define spatial characteristics of the ciliary axoneme in situ, including its position of emergence from the chondrocytic plasma membrane and its angles of extension in 3D space, is a prerequisite to test the validity of this hypothesis.

The purpose of this study is to use analysis of serial sections at the EM level to characterize the orientation of the axoneme of the primary cilium of chondrocytes in articular cartilage using the articular surface as a reference plane, comparing superficial and radiate zones, as well as load-bearing and non-load-bearing regions. The premise is that, if the primary cilium is a sensory receptor involved in establishment of tissue anisotropy and positioning of chondrocytes with defined alignment, then the ciliary axonemes within a population of chondrocytes would have a definable, nonrandom orientation in 3D space. The hypothesis is that axonemal orientation would differ in superficial and radiate zone chondrocytes, whose cellular alignment differs with respect to the articular surface. Because cellular alignment is more highly anisotropic in load- bearing compared to non-load-bearing regions, axonemal orientation in 3D space is expected to reflect this relative difference in tissue anisotropy.

Our results indicate that axonemal 3D orientation has random components, which may be significant for integrating sensory information throughout the population of chondrocytes, as well as nonrandom components. The latter, when considered together with the position of extension of the axoneme from the chondrocyte, may carry meaning for positioning of individual chondrocytes and for maintaining chondrocytic polarity.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. LITERATURE CITED
  9. Supporting Information

Animals, Tissue Collection, Fixation, and Microscopy

Full thickness slabs of articular cartilage were collected from two regions of the lateral femoral condyles of the left femur of four normal 4 to 7-year-old horses. Region A (Fig. 1A), referred to as the loaded region, was located centrally in a position on the middle condylar cartilage that would be constantly under load during standing, and would articulate with the meniscus during the swing phase of ambulation. Region B (Fig. 1A), referred to as the nonloaded region, was a cranial aspect of the condyle that would never experience load, even during full extension of the femorotibial joint. Horses used for this study were being euthanized for a reason not associated with this project (normal anatomical dissection in the veterinary medical anatomical curriculum); use of the tissue was approved by the Institutional Animal Care and Use Committee.

thumbnail image

Figure 1. Tissue collection for equine articular cartilage. (A) Articular cartilage samples were collected from the lateral femoral condyle of mature horses in two regions. The area indicated in A as cranial condylar cartilage represents a region that is non-load-bearing, even with the femorotibial joint fully extended. The area indicated in A as middle condylar cartilage is load-bearing through most phases of the gait cycle. (B) Individual slabs of articular cartilage were removed and sliced into blocks, with the lateral edge notched to indicate the lateral surface. (C) For orientation, 1 μm thick sections were cut and the superficial and radiate zones identified by the characteristic alignment of chondrocytes with their long axes either parallel (superficial, asterisk) or perpendicular (radiate, double asterisks) to the articular surface. (D) 70nm serial sections were cut, ∼40 sections per ribbon, and placed on one-hole formvar covered grids. For superficial zone chondrocytes the articular surface remained intact. For radiate zone chondrocytes, the orientation of the articular surface was maintained by trimming the block with cuts made parallel to the articular surface.

Download figure to PowerPoint

During all steps of tissue collection and subsequent analysis, the original orientation of the sample in 3D space was maintained, utilizing a combination of biological landmarks and created fiducial marks. Slabs were trimmed into 1 × 1 × 3 mm blocks (Fig. 1B). The characteristic morphology of the articular surface and the subchondral bone were used to identify the position of the sample relative to the femur on each block (Fig. 1C). During trimming, the lateral face of each block was notched, identifying the medial-lateral axis. Since collection was always from the left femur, the cranial-caudal axis was unambiguous, having established the medial-lateral axis and using morphological landmarks for the articular surface and subchondral bone.

Blocks were fixed in 2% purified glutaraldehyde/2.5% paraformaldehyde (Ted Pella, Redding CA) in 0.1 M phosphate buffer (pH 7.4, 752 mOsm) at room temperature for one hour, rinsed, and then postfixed in phosphate buffered 1% osmium tetroxide-1.5% potassium ferrocyanide for an additional hour, and embedded in Epon-araldite (Farnum and Wilsman,1983). Blocks were embedded with the orientation of the section plane perpendicular to the articular surface.

A random numbers table was used to select blocks to be cut from each region of each horse. Using a Reichert Ultracut E microtome (Leica Microsystems, Bannockburn, IL), 70 nm thick serial sections were cut from two areas of each block: an area exclusively of superficial zone chondrocytes, with the articular surface included as part of the section, and an area exclusively of radiate zone chondrocytes. For the latter, a cut parallel to the articular surface was made before completion of trimming to maintain orientation to the articular surface. From each area a ribbon of ∼40 serial sections was picked up on formvar-coated 1 × 2 mm, one-hole copper grids, again being sure that ribbons of sections maintained their original orientation in the animal [i.e., sections always were floated without flipping the section from its original orientation (Fig. 1D)]. In addition, 1 μm-thick sections were cut and stained with basic fuchsin/methylene blue/azure II for visualization of cellular morphology at the light microscopical level (Fig. 1C) (Humphrey and Pittman,1974).

Serial sections were stained with uranyl acetate and lead citrate and examined in a Zeiss EM 10A electron microscope equipped with goniometer stage. Incidence of primary cilia in adult articular chondrocytes has been shown to be one per cell (Wilsman and Fletcher,1978; Wilsman et al.,1980). For this study chondrocytic profiles were identified and numbered and then followed in preceding or succeeding sections until the cilium for each chondrocyte was located. Since the probability of visualizing an axonemal profile in a given chondrocytic profile on 70 nm sections is only ∼0.5% (one out of 200), multiple sets of serial sections were required to follow any given chondrocyte in its entirety and locate its cilium. Additional blocks were sectioned until a minimum of ∼100 cilia were sampled from each zone of each region. Each axonemal profile was photographed together with the articular surface.

Representation of Orientation of the Ciliary Axoneme in 3D Space

Orientation of the ciliary axoneme in 3D space can be expressed in terms of spherical coordinates θ, φ (Ascenzi et al.,2007; Donnelly et al.,2010). These angles for the ciliary axoneme (θcilium, φcilium) are the primary outputs of interest. The graphic (Fig. 2A) illustrates measurement of θcilium (orange) and φcilium (blue) for a cilium (gray cylinder) with respect to the 3D spherical measuring frame. The angle φcilium (conceptually comparable to latitude) is defined with respect to the articular surface and the subchondral bone. Angle φcilium varies from 0° to 180° (blue lines and arrows) with 0° pointing directly toward the articular surface (north pole) and 180° pointing directly toward the subchondral bone (south pole). On sections cut perpendicular to the articular surface, φcilium is measured directly with very small measuring errors by knowing and transferring the position of the articular surface to a given micrograph. The differences in intensity in the graphic (dark and light blue) are used to indicate the distinction of the hemisphere toward the articular surface (0°–90°) versus the hemisphere toward the subchondral bone (90°–180°), which is useful for the data plots in the Results section.

thumbnail image

Figure 2. Spherical co-ordinates for presentation of data of axonemal 3D orientation. This diagram (A) (adapted from Figure 1, Marshall and Kitner,2008) demonstrates the orientation and values for θ, φ, the angles used for graphing the orientation of the ciliary axoneme of articular chondrocytes in 3D space. The angle θcilium, which is the orientation angle of the ciliary axoneme measured parallel to the articular surface, varies from 0° to 360° in a plane defined by cranial/caudal and medial/lateral relative to the bone. The dark orange line represents values for axonemes projecting cranially; the light orange represents axonemes projecting caudally. This color-coding will be used in subsequent plots. The angle φcilium, which is the orientation angle of the ciliary axoneme measured orthogonal to the articular surface, has a range of values from 0° (pointing directly to the articular surface) to 180° (pointing directly to the subchondral bone). The dark blue line represents values of axonemes projecting toward the articular surface; the light blue line represents values for axonemes projecting away from the articular surface. Again, this color-coding will be used in subsequent plots. The gray cylinder represents a ciliary axoneme, and the green asterisk is the position of the axoneme on the sphere, where φcilium has a value of ∼135° and θcilium has a value of ∼42°. (B,C) are equatorial and polar Schmidt-net equal area projections for graphing the data. Data from this study can be plotted using either an equatorial view (B), which is analogous to a 2D projection of the data viewing the articular cartilage as in the micrographs of Fig. 1, or a polar view (C), which is analogous to a 2D projection of the data as if one were positioned on the articular surface and looking through the cartilage to the subchondral bone. These are Schmidt-net equal area projections of data from sections cut perpendicular to the articular surface, with 10° between lines and axes identical to the graphic in A. Data point dots are color-coded so that the cranial/ caudal projection of the axoneme on the equatorial view can be discriminated (dark orange/light orange, respectively), and so projection of the axoneme toward the articular surface versus toward subchondral bone can be discriminated on the polar view (dark blue/light blue, respectively). This also is consistent with the coordinates in A. Although both views plot the resultant of φcilium and θcilium for each axoneme, φcilium is most strikingly represented on the equatorial view and θcilium is most strikingly represented on the polar view. The cilium identified by the green asterisk from A has φcilium ∼135° and θcilium ∼42°. Because of the convention of viewing histological sections of articular cartilage cut orthogonal to the articular surface, the equatorial plot is the most intuitively meaningful. However, the polar plot is a particularly useful way to view the distribution of θcilium in a population of chondrocytes.

Download figure to PowerPoint

The angle θcilium is measured in the plane of the articular surface and varies between 0° and 360°. Thus θcilium describes the orientation of the ciliary axoneme in the plane defined by the cranial-caudal and lateral-medial directions (conceptually comparable to longitude). Specifically, 0°-90° is cranio-lateral; 90°–180° is caudo-lateral, 180°–270° is caudo-medial, and 270°–360° is cranio-medial. Locating a cilium to a given quadrant is done by direct observation, knowing that the right-hand edge of the section is lateral and by the observation of adjacent serial sections, knowing that deeper sections are more caudal. Again, the dark orange shading depicts the hemisphere on the cranial aspect, and the light orange is the hemisphere on the caudal aspect, and this color notation will be used in data plots in the Results section. On sections cut perpendicular to the articular surface, the angle θcilium within a quadrant is determined by assuming that cilia are right circular cylinders. Sectioning a cylinder at any angle except 0° results in an ellipse with the length of the short axis equal to the diameter of the cylinder and the length of long axis equal to 1/cos (or secant) θcilium (Rutledge and Pond,1956).

The green asterisk in Fig. 2A symbolizes where an axoneme encounters the measuring frame, specifically the position of a ciliary axoneme with φcilium = ∼135° and θcilium = ∼42°. Putting this in a biological framework, consider that the asterisk represents the cilium of a superficial zone chondrocyte in the load-bearing region with its long axis parallel to the articular surface. The value 135° of the angle φcilium indicates that this axoneme is projecting from the side of the chondrocyte facing the subchondral bone, pointing away from the articular surface at an angle of ∼135° and therefore at an angle of ∼42° relative to the long axis of the chondrocyte. The value of the angle θcilium indicates that the ciliary axoneme is essentially centered in the cranio-lateral quadrant.

Measuring θcilium, φcilium on Electron Micrographs of Serial Sections

The angles θcilium and φcilium were measured on sections cut perpendicular to the articular surface, the width of the section being medial-lateral and deeper sections being cranial to preceding sections (Fig. 1B). Since the articular surface was always known, for each ciliary profile φcilium was measured directly as the angle between the articular surface and the long axis of the axonemal profile as it projected from the chondrocyte.

Measurements to derive an estimate of θcilium required three steps. First, the angle of the axonemal profile with respect to the section plane was estimated by the stereological principle that this angle is the secant of the axial ratio (long axis/short axis)(DeHoff and Rhines,1961; Underwood,1970). Whether the profile was pointing medially or laterally was determined by keeping track of this orientation during sectioning. Third, whether the profile was pointing cranially or caudally was determined by following the axonemal profile in preceding and succeeding serial sections.

As a methodological control, θcilium and φcilium were measured on sections cut parallel to the articular surface. On these sections θcilium is measured directly, whereas φcilium is computed as the secant of the axial ratio, with serial sections used to determine if the axoneme is pointing toward the articular surface or toward the subchondral bone.

Data Analysis and Graphing of θcilium, φcilium

Consistent with visualization of the axonemal orientation in 3D, data were plotted as Schmidt-net equal area projections from two different orientations, equatorial or polar, shown in Fig. 2B,C, with lines of constant φcilium or θcilium every 10° (Allmendinger,2010; Wallbrecher,2010). The equatorial view (Fig. 2B) is a 2D projection of the data as if viewing the bone from cranial to caudal, with the articular cartilage as the north pole, and the subchondral bone as the south pole, utilizing the same coordinates as in Fig. 2A. In the equatorial view the visualization of values of φcilium is particularly striking. In this example for the green asterisk φcilium = ∼135°; the dark orange color of the data point indicates the axoneme is pointing cranially.

The polar view (Fig. 2C) is as if one is viewing from the articular surface to the subchondral bone in a 2D projection. In this view the visualization of values of θcilium is particularly striking because θcilium can be viewed directly. This example plots the coordinates for the green asterisk from Fig. 2A with θcilium = ∼42° (in the cranio-lateral quadrant); the light blue coloration of the data point indicates that the axoneme is pointing toward the subchondral bone.

Numerical values for θcilium and φcilium for a given axoneme can be read directly from these plots. However, it is important to stress that in this study we are interested less in the spherical coordinates of a specific axoneme than in visualizing the 3D orientation of axonemes in a population of chondrocytes. The graphs are designed to give an immediate visual impression of axonemal orientation in 3D space, projected onto the bone as it is positioned in the living animal.

For each region (loaded vs. nonloaded) for each type of chondrocyte, (superficial vs. radiate), using Fisher descriptive statistics a Resultant (R) and Spherical Variance of the Resultant (S) were calculated (Fisher et al.,1987; Moon and Spencer,1988; Lehmann and Romano,2005). Resultants are a numerical expression of the extent to which a set of vectors (cilia) are in alignment. In general, Resultants vary from 0 ≤ RN with 0 = no alignment and N = perfect alignment. In this study maximum R = 100 = sample size.

To calculate resultants, the angular coordinates φcilium and θcilium were converted to Cartesian (l, m, n) coordinates:

l = sin (φ) cos (θ)

m = sin (φ) sin (θ)

n = cos (θ)

Resultant: R = v (√ li )2 + (∑ mi )2 + (∑ ni )2

Spherical Variance: S = (NR)/N

Rose diagrams were used as a visual graphic representation of the quantitative aspects of axonemal orientation (Wallbrecher,2010).

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. LITERATURE CITED
  9. Supporting Information

Contrasting Morphology of Chondrocytic Cilia as Seen in the Superficial and Radiate Zones of Equine Articular Cartilage

Visualization of chondrocytic axial alignment in equine articular cartilage is critical for documenting its tissue zonal anisotropy. Superficial zone cells, when viewed in sections cut perpendicular to the articular surface (positioned at the top of all micrographs in Fig. 3), are aligned with their major axes paralleling the articular surface (Fig. 3A,B). Figure 3C shows the characteristic relationship of the ciliary basal body, ciliary axoneme and the associated centriole to the Golgi of the chondrocyte. Note that, in this example, the ciliary axoneme is embedded deep within an in-pocketing of the plasma membrane, and is projecting from the cell on the side of the chondrocyte away from the articular surface and at an angle essentially parallel to the articular surface.

thumbnail image

Figure 3. Morphology of chondrocytes and the associated primary cilium in equine articular cartilage. (A,B) The orientation of the long axis of superficial zone chondrocytes parallel to the articular surface is seen at both the light (A) and electron microscopical (B) levels. Staining of the 1 μm thick sections is with basic fuchsin/methylene blue/toluidine blue (Humphrey and Pittman,1974). (C) At higher magnification, the close association of the centriole (c), basal body (b) and ciliary axoneme (a) with the Golgi stacks of the chondrocyte is seen. Note that the axoneme of this cilium is deeply invaginated into the chondrocytic plasma membrane, and is oriented essentially parallel to the major axis of the cell, and hence parallel to the to the articular surface. (D,E) Micrographs at approximately the same magnification as seen for superficial cells demonstrate the shape and size differences of radiate zone cells compared to superficial zone cells. In D, nuclei and lipid droplets stain blue, and the Golgi is seen as a relatively clear juxtanuclear area. In E a grazing cut through two adjacent radiate cells shows their relationship to each other and to a relatively large pericellular matrix of each cell. Collagen fibrils aligned perpendicular to the articular cartilage are seen in the matrix distant from the cells. (F) A grazing cut through the axoneme (a), basal body (b) and associated centriole (c) of the radiate zone cell show the close association of the cilium with juxtanuclear Golgi stacks. The axoneme emerges on the side of the chondrocyte facing the articular surface and is pointing directly toward the articular surface.

Download figure to PowerPoint

Radiate zone chondrocytes are aligned with the major axis perpendicular to the articular surface (Fig. 3D,E). These cells often can be seen in small columns of 2-4 cells, with the alignment of the column also perpendicular to the articular surface. The electron micrograph in Fig. 3E shows at higher magnification the relationship of adjacent cells and their shared matrix. In Fig. 3F an axonemal profile can be seen projecting from one chondrocyte toward the adjacent chondrocyte. This axoneme both emerges from the side of the chondrocyte that is oriented toward the articular surface, and also projects directly toward the articular surface.

Measurement of Ciliary 3D Orientation on Serial Sections by EM

Details of ciliary identification and the challenges of following axonemal profiles in serial sections are presented in Fig. 4. When first encountered on a given set of serial sections, ciliary profiles were cryptic, as seen in superficial zone chondrocytes from sections cut perpendicular to (Fig. 4A) or parallel to (Fig. 4B) the articular surface (arrowheads). After initial identification, each axonemal profile was followed in adjacent serial sections until its relationship with the associated centriole was viewed optimally. In Fig. 4C the centriole is barely apparent; centriolar identification is optimal in Fig. 4D. At this point a line (L2, yellow line) was drawn through the axonemal profile (Fig. 4E,F) and was extended to intersect a line paralleling the articular surface (L1, yellow line). On sections cut perpendicular to the articular surface (Fig. 4E,F), φcilium was measured directly as the angle between L1 and L2; θcilium was calculated, as previously described. Figure 4E,F demonstrate cilia pointing both toward and away from the articular surface. For a cilium of a superficial chondrocyte to project toward the articular surface, it must emerge from the articular side of the superficial cell, given the spatial constraint that the long axis of the cell parallels the articular surface.

thumbnail image

Figure 4. Identification and measurement of axonemal profiles in articular chondrocytes. (A)–(D) The elusive nature of axonemal profiles when first encountered on a 70 nm thick section is demonstrated in A and B of a superficial zone chondrocyte from sections cut perpendicular and parallel to the articular surface, respectively. In both these chondrocytes an axonemal profile emerging from the region of the Golgi is visible but cryptic (arrowheads). Once identified, axonemal profiles were followed in adjacent serial sections until the section that most optimally demonstrated the relationship of the cilium to the associated centriole was reached. As an example, although the axoneme in C is fully visible, the associated centriole has not yet come into full view. In D both the centriole and the axoneme are clearly visible. (E,F) These micrographs give examples of drawing lines (L1, L2, in yellow) for measurement of angles of orientation of the ciliary axoneme in 3D space. In E the articular surface is identified in a load-bearing region as the junction between the cartilage matrix and electron dense material at the articular surface, which is preserved with ferrocyanide fixation and presumed to be hyaluronic acid and/ or lubricin (Farnum and Wilsman,1983). L1 marks the articular surface; L2 represents the angle of projection as the axoneme emerges from the cell. In F similar lines are drawn, although the clearest indicator of the direction of the articular surface is at the edge of the electron dense material. Note the more irregular cellular shape and matrix organization in the non-weight-bearing region (F) compared to the load-bearing region (E).

Download figure to PowerPoint

Figure 4E,F demonstrate two additional differences between the general morphology of weight-bearing versus non-weight-bearing regions. In non-weight-bearing regions (Fig. 4F) collagen fibrils were of larger average diameter and organized into more definable bundles than in weight-bearing regions (Fig. 4A,E). Cellular shape was more irregular in non-weight-bearing regions (cf. especially Fig. 4A,F).

In Weight-Bearing Regions, Axonemal Orientation and Extension from the Chondrocyte Differ in Superficial and Radiate Zone Cells

The comparisons in Fig. 5 demonstrate two striking differences in axonemal directional positioning between superficial and radiate zone chondrocytes in the load-bearing region. First, the position of emergence of the axoneme from the chondrocyte differs. In the superficial zone (5A), essentially no axonemes emerge from the face of the chondrocyte positioned toward the articular surface; in radiate zone chondrocytes (5B) there is an almost even split between axonemes emerging so they are directed toward the articular surface, and axonemes emerging so they are directed toward the subchondral bone. Second, in superficial zone cells, axonemes extending away from the articular surface have all possible values of φcilium and θcilium, essentially taking on a random orientation in 3D space. This is demonstrated in a striking way in Fig. 5A, where essentially all data points are in the “southern hemisphere.” The generally equal numbers of dark and light circles indicates that axonemes are essentially pointing equally in the cranial and caudal directions. This can also be seen in Fig. 5B, where the relative paucity of dark data points indicates that very few axonemes point toward the articular surface.

thumbnail image

Figure 5. Axonemal orientation in 3D space: superficial zone chondrocytes and radiate zone chondrocytes in a load-bearing region. Axes for these plots follow the convention of Fig. 2, as described in the Methods section, with 10° between all coordinate lines. Two observations can be made comparing the 3D orientation of ciliary axonemes in superficial zone chondrocytes in the loaded-region (A,B) to radiate zone chondrocytes in the loaded region (C,D). First, φcilium of superficial zone chondrocytes in the loaded region essentially is restricted to values pointing away from the articular surface. In contrast, in radiate zone chondrocytes from the same region, almost half the axonemes are directed toward the articular surface and half toward the subchondral bone. Second, as shown in the polar plots (B,D), despite the restricted values of φcilium, θcilium takes on all possible values for superficial zone chondrocytes of the load-bearing region for axonemes emerging from the subchondral face of the cell. θcilium has a smaller range of values in radiate zone cells, since the position of emergence of the axoneme is restricted to the narrow sides of the chondrocyte facing either the articular surface or the subchondral bone (Figs. 3D,E).

Download figure to PowerPoint

In contrast, in radiate zone chondrocytes φcilium had a restricted range, essentially being present only between 0° to 60° and 130° to 180°. Viewed in the equatorial plot (Fig. 5C), this means that φcilium never had values ±30° from the equator, and that an almost equal number of axonemes point toward the articular surface or toward the subchondral bone. The polar plot (5B) emphasizes that axonemes emerging from the subchondral bone side of superficial zone cells take on all possible values for θcilium, consistent with the extensive surface of the long axis of the cell paralleling the articular surface (Fig. 3A,B). In contrast, the polar plot for radiate zone cells (5D) shows that θcilium for these cells was more restricted. This is consistent with the restricted emergence of the axoneme from the narrow faces of radiate cells that parallel the articular surface (Fig. 3D,E). Thus, in both the superficial and radiate zones, axonemal directional positioning was highly defined, but strikingly different.

In Both Zones of Articular Cartilage, Axonemal Orientation has Less Variability in Weight-Bearing than in Non-Weight-Bearing Regions

Ciliary axonemal projection was less defined for both superficial and radiate zone cells in the nonloaded region, as seen in equatorial plots in Fig. 6A,B, when compared directly to Fig. 5A,C for superficial and radiate zone cells, respectively, in the loaded region. In the superficial zone in the non-load-bearing region (Fig. 6A), it is visibly apparent that more axonemes project to face toward the articular surface than in the load-bearing regions (Fig. 5A). Similar to the loaded region, the angle θcilium appears to take on all values from 0° to 360° for axonemes pointing either toward the articular surface or toward the subchondral bone for the superficial zone cells in the nonloaded region (Fig. 6A). The data for θcilium are less easily defined for radiate zone cells in the non-load-bearing region (Fig. 6B), with some impression of a preferential projection medial/lateral. For radiate zone cells, the angle φcilium took on a fuller range of values in the nonloaded regions (Fig. 6B) than in the loaded region (Fig. 5C).

thumbnail image

Figure 6. Axonemal orientation in 3D space: Comparison of superficial and radiate zone chondrocytes in the non-load-bearing region. (A,B) present a comparison of equatorial views of the data for superficial (A) and radiate (B) zone chondrocytes in the nonloaded region, with axes for these plots following the convention of Fig. 2, as described in the Methods section, with 10° between all coordinate lines. In both superficial and radiate zone chondrocytes the defined orientations of ciliary axonemes seen in the load-bearing region is lost in the non-load-bearing region. The number of data points is insufficient to demonstrate if there is some clustering of values in chondrocytes of either zone of the nonloaded regions; however, the highly defined 3D orientation of axonemes seen in the loaded regions (Fig. 5A,C) clearly has been lost.

Download figure to PowerPoint

Resultant data are presented in a table for the four groups (Fig. 7). The high resultant values for axonemal orientation for both superficial and radiate zone chondrocytes in the loaded regions indicate more alignment of ciliary axonemes in both of the chondrocytic populations than in the nonloaded region. Rose diagrams, which are a graphic way to demonstrate the degree of alignment, give a visual image that characterizes the different way axonemal alignment is achieved in the superficial and radiate zone cells of the load-bearing region. In superficial zone cells (Fig. 7A), the resultant value is high because essentially all axonemes point away from the articular surface, even though θcilium is random for ciliary axonemes in these cells. In radiate zone cells (7B), the resultant value is high because axonemes point either toward or away from the articular surface, with a more restricted value of θcilium.

thumbnail image

Figure 7. Resultants and rose diagrams indicating degree of alignment of ciliary axonemes. Resultants and spherical variance for axonemal 3D orientation were calculated as described in the Materials and Methods section. The higher resultant values for both superficial and radiate zone chondrocytes indicates a higher degree of alignment of ciliary axonemes in the loaded region than in the nonloaded region. This is shown visually in the rose diagrams. Alignment of ciliary axonemes in superficial zone chondrocytes in the loaded region (A) is associated with consistency of orientation away from the articular surface, despite θcilium taking on all possible values (essentially random). Alignment of ciliary axonemes in radiate zone chondrocytes (B) is associated with ciliary axonemes pointing essentially equally toward the articular surface or toward the subchondral zone, with a more restricted range of θcilium than seen for superficial zone cells. The lower resultant values in the nonloaded region indicate that ciliary axonemes are less aligned than in the load-bearing region, as shown visually in the rose diagrams in C,D.

Download figure to PowerPoint

It is clear both from the lower resultant values and visually from the rose diagrams that, comparatively, ciliary axonemes have less alignment with each other in nonloaded regions (Fig. 7C,D) than in loaded regions (Fig. 7A,B), with the least alignment being for ciliary axonemes in the superficial zone of the nonloaded region. It is important to put the data in context with the plots. The data demonstrate that axonemal positional projection in 3D space is less defined in non-weight-bearing regions (Fig. 7C,D) compared to weight-bearing regions (Fig. 7A,B).

Values of θ and φ from Sections Parallel Versus Orthogonal to the Articular Surface

In sections cut orthogonal to the articular surface, the profile of a superficial cell shows a long major axis parallel to the articular cartilage, and a shorter minor axis perpendicular to the articular surface (Fig. 4A). Figure 8A is the characteristic profile of a superficial zone chondrocyte from a section cut parallel to the articular surface, demonstrating that these superficial zone chondrocytes are, in fact, essentially pancake-shaped and approach being circular when viewed from the articular surface. Ciliary axonemes can be identified in either view, from sections cut orthogonal to (Fig. 8B) or parallel to (Fig. 8C) the articular surface. As these micrographs demonstrate, since the cilium is located in a juxtanuclear position, the nuclear profile often appears irregular, as would be seen in a grazing profile at the far edge of the nucleus.

thumbnail image

Figure 8. Measurement of θcilium and φcilium both directly and indirectly in the superficial zone of the load-bearing region, as a methodology control. (A)–(C) The characteristic shape of a superficial zone cell from a section cut parallel to the articular surface is illustrated in A, demonstrating that superficial cells are discoid-shaped with an almost round profile when viewed from the articular surface. The standard convention is to view the cellular organization of articular cartilage in sections cut perpendicular to the articular cartilage (B), which was our standard protocol. In these sections, φcilium is measured directly and θcilium is calculated as a function of the secant of the axial ratio of the axonemal profile. As a check on the accuracy of methodology utilizing the secant calculation, a set of sections was cut parallel to the articular surface (C), where θcilium is now measured directly and φcilium is calculated. The data are presented in two equatorial view graphs (D,E) where it can be seen immediately that the data are essentially identical. Axes for these equatorial plots follow the convention of Fig. 2, as described in the Methods, with 10° between all coordinate lines.

Download figure to PowerPoint

In sections cut orthogonal to the articular surface, φcilium is measured directly, and θcilium is calculated. In sections cut parallel to the articular surface, θcilium is measured directly and φcilium is calculated. Our standard protocol was to section orthogonal to the articular surface, thereby measuring φcilium directly. As a control of the methodology we cut one set of serial sections parallel to the articular surface, allowing us to measure θcilium directly and calculate φcilium. The two graphs of these data (equatorial views, Fig. 8D,E) demonstrate that the data essentially are identical. The conclusion is that, by using the standard approach of analyzing sections cut orthogonal to the articular surface, no technical error is introduced by measuring θcilium indirectly using the secant function.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. LITERATURE CITED
  9. Supporting Information

Articular cartilage displays high tissue anisotropy, which is established in development, is essential for normal function, and requires a sustained biomechanical environment to be maintained. We used analysis of EM serial sections to demonstrate two important positional attributes of the primary cilium of articular chondrocytes. First, the axoneme has a definable orientation in 3D space, which can be separated into random and non–random components. The orientation of the ciliary axoneme changes in superficial zone and radiate zone chondrocytes, cells that differ by 90° in the orientation of their major axes relative to the articular surface. Orientation of the ciliary axoneme is more tightly definable in load-bearing areas compared to non-load-bearing areas, mirroring comparative differences in matrix and cellular organization of these two regions. Secondly, the position of extension of the axoneme from the cell also is variable. In load-bearing regions of the superficial zone, the position of extension of the axoneme into the ECM essentially is always from the side of the cell facing the subchondral bone; in radiate zone cells axonemes extend from both faces of the chondrocyte, that is, both toward the articular surface or toward the subchondral bone. In non-load-bearing regions the striking consistency observed in weight-bearing regions is lost.

Directional Positioning of the Axoneme: Emergence Versus 3D Orientation

Our results describe 3D directional positioning of the axoneme in adult articular cartilage. When and how these positional attributes of the ciliary axoneme of an individual chondrocyte are established, and the degree to which they are transitory or permanent, are unknown. When considering directional positioning of the ciliary axoneme, it is essential to consider two components. First is the position on the chondrocyte from which the axoneme emerges. Several studies of primary cilia in articular cartilage have suggested that the position of emergence of the axoneme reflects chondrocytic polarity within the tissue (Song et al.,2007; McGlashan et al.,2007; de Andrea et al.,2010). Second is the orientation of the axoneme in 3D space after emergence from the cell. Our current study is the first to address this critical aspect of axonemal directional positioning in articular cartilage. These two components need to be separated conceptually when considering how they might be established initially and their stability over time.

For primary cilia in articular cartilage, the hypothesis is that positioning of the cilium at a juxtanuclear location initially is associated with an in-pocketing of the cilium into an invagination of the plasma membrane. In-pocketing is a consistent feature of the primary cilium in articular chondrocytes on EM images (Wilsman,1978; Poole et al.,1985,1997,2001; Jensen et al.,2004), but not resolved using either confocal or multiphoton microscopy (MPM). This feature has been called a ciliary pocket, a membrane domain found at high frequency in nonepithelial cells, enriched in clatrin-coated pits, and associated with actin filaments. The ciliary pocket is hypothesized to have a role in membrane trafficking in the vicinity of the cilium, potentially involved with basal body and axonemal positioning, signal transduction, and endocytosis (Molla-Herman et al.,2010; Moser et al.,2010; Rattner et al.,2010). However, how in-pocketing might be involved with establishment of orientation of the ciliary axoneme in 3D space has not been investigated.

An interesting comparison is of potential differences between primary cilia of epithelial cells and chondrocytic primary cilia for these two components of directional positioning of the axoneme. The position of emergence of the axoneme from epithelial cells is considered to be fixed during the establishment of cellular polarity; the axoneme is generally defined as being positioned centrally on the luminal side of the cell (Pazour and Bloodgood,2008; Gerdes et al.,2009; Zhou,2009). In epithelial cells the orientation of the axoneme is 3D space is considered to be variable and temporally transitory. In kidney tubular epithelial cells, as an example, the 3D orientation of the ciliary axoneme is subject to change by deflection under fluid flow (Pazour and Witman,2003; Nauli and Zhou,2004; Marshall and Kitner,2008; Resnick and Hopfer,2008; Berbari et al.,2009; Gerdes et al.,2009; Zhou,2009; Xiao and Quarles,2010).

For ciliary axonemes in chondrocytes, the position of emergence may be associated with either planes of cell division or directions of cellular migration (Menezes and Germino,2010; Moseley and Nurse,2010; Quyn et al.,2010; Satir,2010; Schneider et al.,2010; Seeley and Nachury,2010). However, given that the axoneme of chondrocytes has been shown to have direct attachments to the ECM (Jensen et al.,2004; McGlashan et al.,2006; Knight et al.,2009), its 3D orientation in space should be considered to be relatively nonflexible, although potentially variable and related to changes induced by compression during loading that would carry the axoneme with the matrix and thus change its 3D directionality. Our results do not address the question of variability of chondrocytic 3D orientation over time for an individual chondrocyte. However, our results do emphasize that the 3D orientation of the ciliary axoneme is a variable within a population of chondrocytes. These observations should be incorporated into any hypotheses concerning function of the primary cilium in chondrocytes or in other connective tissue cells where the axoneme extends into, and is attached to, the ECM.

Position of Emergence of the Axoneme from Articular Chondrocytes: Comparison with Previous Studies of Growth Plate Chondrocytes

The alignment of centrally positioned ciliary axonemes on adjacent chondrocytic profiles along what has been called a virtual axis of the tissue has been described in murine and human growth plates viewed by confocal microscopy (Haycraft and Serra,2008; de Andrea et al.,2010). This alignment is visible in published micrographs from our previous studies of rat growth plate cartilage on methanol-fixed sections viewed by (MPM) (Ascenzi et al.,2007; Donnelly et al.,2008; Farnum et al.,2009) (Supporting Information S1, S2). In all zones—proliferative, prehypertrophic and hypertrophic—the ciliary axoneme characteristically emerges from a central location adjacent to the nucleus, either from the epiphyseal (proximal) or the metaphyseal (distal) side of the chondrocyte. Consequently, pairs of adjacent chondrocytes show axonemes either directly projecting toward each other or directly away (video, S1).

Our EM serial sectioning methodology did not provide an appropriate scale for viewing this relationship, if present, in articular chondrocytes. However, the EM methodology did allow consistent affirmation that the primary cilium has an intimate relationship with the Golgi, which in turn is found in a juxtanuclear position (Wilsman et al.,1980; Poole et al.,1985,1997; Jensen et al.,2004). Such consistency of positioning of the axoneme for emergence from the cell implies a consistency of positioning of the ciliary basal body to the plasma membrane before ciliogenesis, as is required for ciliary regulation of planar cell polarity in epithelial tissues (Fischer et al.,2006; Jones et al.,2008; Santos and Reiter,2008; Goetz and Anderson,2010; Hoyer-Fender,2010; Menezes and Germino,2010) and for asymmetrical localization and posterior tilting of motile nodal cilia (Nonaka et al.,2005; Borovina et al.,2010).

The consistency of the position from which the ciliary axoneme extends from the chondrocyte suggests that the position, once established, does not change and may be related to the initial division plane of chondrocytes within columns (Milenkovic et al.,2009; Moseley and Nurse,2010; Quyn et al.,2010; Satir,2010). Observations of the division planes of dividing cells and the positioning of the Golgi relative to the nucleus, previously reported in the literature and documented in our own studies of growth plate chondrocytes (Supporting Information S2), provide support for this hypothesis (Dodds,1930; Morales,2007). First, proliferative chondrocytes of the growth plate divide in a plane orthogonal to the long axis of the bone, and, immediately after division, the nuclei of mother and daughter cells are situated at opposite sides of the division plane (Dodds,1930; Gould et al., 1974). Second, proliferative chondrocytes soon after division appear to interdigitate as mirror images. Third, the juxtanuclear Golgi of growth plate chondrocytes alternates its position relative to the nucleus in adjacent cells of a column. Finally, the chondrocytic primary cilium characteristically is associated with the Golgi (Wilsman et al.,1980; Poole et al.,1985,1997). These observations are consistent with the hypothesis that the location of the basal body with respect to the cell reflects its juxtanuclear location within the Golgi and its initial postproliferative docking position.

Post mitotic cells in the growth plate are thought to move relative to each other to establish the cellular column (Dodds,1930; Gould et al.,1974; Morales,2007; de Andrea et al.,2010). This two-step process (division followed by movement), ending with chondrocytes aligned in the direction of elongation, both minimizes the potential disruptive effect of mitosis on column alignment and results in appropriate cellular positioning for enlargement during hypertrophy, which is the primary engine of elongation (Baena-Lopez et al.,2005; Wilsman et al.,1996; Li and Dudley,2009). This hypothetical time-based scenario of dynamic events of division and realignment into columns has been refined to include cilia-related positioning of growth plate chondrocytes in columns and dependence on regulation via β1 integrins and ciliary mechanotransduction (Hynes,2002; Aszodi et al.,2003; Grashoff et al.,2003: Millward-Sadler and Salter,2004; Praetorius et al.,2004; Song et al.,2007; Haycraft and Serra,2008; de Andrea et al.,2010). This hypothesis predicts that following docking of the basal body and during axonemal elongation, the cilium will maintain the original juxtanuclear position for emergence of the axoneme (Milenkovic et al.,2009). It is our assumption that the initial positioning of the cilium in articular chondrocytes is analogous to that of growth plate chondrocytes, occurring after cell division and remaining stable over time in its postproliferative juxtanuclear position embedded in the Golgi.

Comparison of Axonemal Positional Characteristics in Articular Chondrocytes Versus Growth Plate Chondrocytes; Does this Provide Clues About the Development of Tissue Anisotropy in Articular Cartilage?

The alignment of the long axis of articular chondrocytes changes from paralleling the articular surface in superficial zone cells, to being orthogonal to the articular surface in radiate zone cells. Indeed, the columnar arrangement of radiate cells presents a mini-version of the columnar alignment of chondrocytes in growth plates. Interestingly, in load-bearing regions, the pattern of ciliary projection in the population of radiate zone chondrocytes mirrors what is seen in columns of growth plate chondrocytes, that is, approximately equal numbers of axonemes projecting toward the articular surface and toward the subchondral bone. The establishment of tissue anisotropy in articular cartilage is poorly understood. However, one hypothesis for maintenance of columnar organization within growth plates links the significance of orientation of mitotic spindles of cells during division and the relative positioning of cells to each other post division to ciliary function, as has been shown for migrating cells and for planar polarity systems that operate in motile populations undergoing dynamic morphogenetic changes (Albrecht-Buelher and Bushnell,1979; Katsumoto et al.,1994; Ueda et al.,1997; Fischer et al.,2006; Pan and Snell,2007; Zallen,2007; Jonassen et al.,2008; Pouthas et al.,2008; Saburi et al.,2008; Santos and Reiter,2008; Schneider et al.,2010; Zhou,2009; Goetz and Anderson,2010; Hoyer-Fender,2010; Menezes and Germino,2010; Mirzadeh et al.,2010; Satir et al.,2010).

For actively dividing chondrocytes, if axonemal position is established soon after cellular division, there may be spatial constraints to axonemal positioning involved with the division plane and internal cellular organization, specifically of the Golgi and its spatial proximity to the nucleus (Follit et al.,2006; Feldman et al.,2007; Gonçalves et al.,2010; Moseley and Nurse,2010). This hypothesis would predict that, for proliferative cells of the growth plate, positional consistency within the Golgi and adjacent to the nucleus would be carried through each division; for postmitotic cells (prehypertrophic and hypertrophic cells considered to be in Go), positional consistency of the axoneme would be maintained following their last division. Given the uniform positioning of the axoneme with respect to the chondrocyte (despite significant changes in both shape and volume of the chondrocyte during differentiation), the assumption is that, once positioned, the cilium does not migrate within the plasma membrane (Pazour and Bloodgood,2008; Hu et al.,2010).

The source of stem cells for articular chondrocytes and the chronology of cellular events leading to the establishment of the characteristic zonation of mature articular cartilage have been debated over many years (Mankin,1962,1963; Hayes et al.,2001; Dowthwaite et al.,2003; Hunziker et al.,2007; Simkin,2008). A recent addition to this debate has been the suggestion that mesenchymal stem cells from the synovium migrate over the articular surface and position themselves in the presumptive superficial cell zone, where they remain as nonproliferative superficial zone cells. A group that migrates deeper into the transitional zone of the tissue maintains proliferative potential and forms the columns of radiate zone cells (Simkin,2008). Ciliary axonemal projection in these deeper cells would be expected to follow the axonemal projection pattern seen in growth plate chondrocytes, which mirrors the columnar arrangement seen in radiate zone cells. Indeed, our data show that axonemes of radiate zone articular chondrocytes project both toward and away from the articular surface, analogous to axonemal emergence direction seen in columns of growth plate chondrocytes.

This hypothesis suggests that chondrocytes in immature articular cartilage are migratory, that migration into the cartilage occurs from a division site at the synovial margins, that a subset of cells remains at the articular surface while a group moves into the transitional zone guided by tensile stress patterns in the ECM, and that the latter cells continue proliferation, resulting in columns of radiate chondrocytes (Fig. 9) (Simkin,2008). Several recent investigations have documented the ability of cells isolated from both skeletally immature and skeletally mature articular cartilage to migrate through the ECM (Chang et al.,2003; Davies et al.,2008). Our data would suggest that the group of chondrocytes that remains as nonproliferative superficial zone cells at the articular surface are the ones with a highly uniform direction of axonemal emergence, that is, always projecting away from the articular surface (Fig. 9).

thumbnail image

Figure 9. Position of emergence of the ciliary axoneme from chondrocytes in the context of development of tissue anisotropy in articular cartilage. Simkin (2008) has proposed a time-based scenario for the development of tissue anisotropy in articular cartilage based on the migration of mesenchymal stem cells from the margins of the joint. A subset of these cells remains on the articular surface as nonproliferative superficial zone cells; a subset migrates into the putative transitional zone and divides to form columns of radiate zone cells. In this figure the basal body (mother centriole) is a red dot; the daughter centriole is a green dot; the ciliary axoneme is a red rod. Consistent with the findings of this study, the position of emergence of the axoneme from the cell is depicted as variable in the nonweight bearing region of the cartilage (arrowheads), and as consistently pointing away from the articular surface for cells remaining in the superficial zone. Radiate zone chondrocytes postdivision align into columns. The axoneme of individual radiate chondrocytes emerges to extend either toward or away from the articular surface, as also seen in columns of growth plate chondrocytes (Supporting Information S1, S2 A, B, C). Axonemal emergence from mesenchymal stem cells at the edge of the joint is depicted as highly variable in direction, but this was not analyzed in this study. This figure is adapted from diagrams in Fig. 2, Morales,2007; Fig. 2.2, Pedersen and Rosenbaum,2008; Fig. 1, Simkin,2008; Fig. 6, de Andrea, et al.,2010.

Download figure to PowerPoint

Serial Section EM as an Approach for Studying Ciliary Orientation

Analysis of EM serial sections, in which a ciliary axonemal profile appears on only 1/200 chondrocytic profiles, is tedious, time intensive and a technically demanding procedure at every step. Nevertheless, it is the optimal way to visualize ciliary morphology and multiple aspects of axonemal and basal body architecture. It is the only method available for visualizing multiple features of the ciliary complex such as alar sheets and striated rootlets, as well as subtle asymmetrical features such as the position of the basal foot and the characteristic loss of outer microtubular doublets that, in articular chondrocytes, has been shown to begin at approximately the midpoint of the axonemal shaft (Wilsman,1978; Wilsman et al.,1980). No ambiguity is involved in identifying primary cilia at the EM level, including specifics of the relationship of the ciliary complex to the Golgi and other cellular organelles and the morphology involved with in-pocketing of the ciliary axoneme within the plasma membrane and relationships to the pericentriolar material (Wilsman,1978; Poole et al.,1985,2001; Jensen et al.,2004; Molla-Herman et al.,2010; Moser et al.,2010). Thus, serial section EM analysis is an optimal way for examining multiple aspects of ciliary structure, even the most subtle, but it is an impractical approach for analyzing changes in axonemal 3D orientation between experimental and control groups in a reasonable period of time.

Analysis of ciliary 3D orientation by confocal or MPM with ciliary axonemes identified by immunofluorescence-based IHC has many positive features compared to EM analysis. The two most significant are rapid preparation of the tissue, including staining for ciliary identification, and rapid data collection through optical sectioning. However, limitations of IHC include that ciliary identification can be ambiguous, both due to some level of nonspecific staining and because beam penetration into the tissue becomes limiting. In addition, it must be assumed that positive reactivity occurs along the entire length of the ciliary axoneme and that in-pocketing of the axoneme within the plasma membrane does not interfere with access of the antibody. Although double staining of the axoneme and the basal body have been achieved (Pazour et al.,2002; Anderson and Stearns,2009; Farnum et al.,2009; Hu et al.,2010; Kim et al.,2010), resolution with confocal or MPM is not sufficient to analyze asymmetries of the basal body or the angular relationship of the basal body to the associated centriole. Because cellular morphology with fixation appropriate for IHC is suboptimal, cellular shapes must be idealized if analysis for cell polarity is included (Ascenzi et al.,2007).

It should be recognized that, in this study and all previously published studies of ciliary emergence and 3D orientation, only a very small sample of tissue is analyzed, and observations are limited to one species at one age. Our EM analysis involved a comprehensive experimental design with the goal of making qualitative and quantitative analyses of differences among four different populations of chondrocytes, sampled over four animals. For collection of data specifically about ciliary orientation in 3D space, EM data is nonambiguous (Wheatley,2008; Farnum et al.,2009).

The Cilium as the Chondrocyte's Antenna

The hypothesis that primary cilia function as mechanosensors as part of a cellular global positioning system to detect changes in the surrounding environment and to initiate responses would predict some constancy of positioning of the cilium relative to the cell (Poole et al.,1985; Satir and Christensen,2008), especially given the physical attachment of the ciliary axoneme to the ECM and the possibility that the cilium via its attachments to the ECM will be physically influenced in multiple directions (tension, compression, shear, rotation) as its biomechanical environment changes during ambulation (Jensen et al.,2004). The randomized distribution of θcilium for superficial zone cells but with φcilium essentially absent over half of its potential range (since the axoneme essentially never projects in the entire hemisphere on the articular cartilage side of the major axis of the chondrocyte) would be equivalent to having antennae positioned in all possible directions for the population of chondrocytes, but always pointing away from the articular surface. This is in contrast to axonemes of radiate zone cells, which extend both toward and away from the articular surface on individual cells but never have angles of elevation (φcilium) in the central 40° of the potential range. Again, with the analogy to antennae, almost half of the possible directional angles of the axoneme are not used.

Why axonemes within a population of chondrocytes display this characteristic and highly definable orientation, with both random and nonrandom components, currently is unknown. An intriguing follow-up question for future investigation is to ask the question, is the positional orientation of the ciliary axoneme significant only for signal transduction within an individual cell or is input received by individual chondrocytes involved with intercellular signaling integrated via the ECM throughout the population of chondrocytes, thus influencing the ability of the articular cartilage to sense the magnitude and duration of extrinsic forces and to integrate appropriate responses (Huang and Ingber,1999; Ingber,2003; Boudreau and Weaver,2006; Singla and Reiter,2006; Anderson et al.,2008)? This opens up possibilities of considering that the primary cilium could be involved with both receiving and transmitting signals (Wheatley,2010) and thus with integrating a higher order memory of the mechanical loading environment of articular cartilage as a whole by imparting a long-term cellular and tissue memory, similar to that hypothesized for bone cells and neuronal networks (Huang and Ingber,1999; Turner et al.,2002).

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. LITERATURE CITED
  9. Supporting Information

The authors thank Dr. Rebecca Williams and the DrBio Facility at Cornell University for use of the multiphoton microscope and creation of the graphic in Fig. 9. Dr. John Starkey was very helpful with advice for statistics of spherical data. They also extend their thanks to Wenhua Liu for the picture of growth plate chondrocytes during multihour imaging (S2L), to Debra Reed-Aksamit for her expertise in electron microscopy, and to Karen Lyons and Maria-Grazia Ascenzi for their helpful discussions. Special thanks go to Jen Patterson for help with Figures and submission of the manuscript.

LITERATURE CITED

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. LITERATURE CITED
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. LITERATURE CITED
  9. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
AR_21330_sm_suppinfofig2.tif14197KSupporting Information Figure.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.