In studies of active tactile sensing, the motor plant of the rodent mystacial pad (MP) has been used to study coordinated and precise whisker movements (Hill et al.,2008; Knutsen et al.,2008; Grant et al.,2009; Simony et al.,2010). The orderliness of simultaneous vibrissal movements during whisking spurred our search for a mechanism that enables highly coordinated movements of the vibrissae, which may be similar to those previously described for cardiac and lung skeletons that consist of fibrous collagen frameworks (Caulfield and Borg,1979; Cavalcante et al.,2005; Pope et al.,2008). In the mouse MP, connective tissue fibers in bands parallel to the skin surface connect, in a plate-like fashion, the lower parts of two neighboring follicles in the same row (Dörfl,1982). In contrast, in the hamster MP, there is a loosely organized layer of connective tissue fibers in the immediate subcapsular zone (Wineski,1985). Involvement of connective tissue in tying the capsule to the hair follicle has been described in rats (Vincent,1913). Despite the obvious importance of collagen for the mechanical properties of the MP, until now little has been revealed about the structural organization of collagen in the entire MP.
In whisking rodents, MPs are specifically organized and contain four basic tissue types. Starting from the outside, the first is the corium, which is covered by epithelial tissue that rests on a layer of connective tissue. Within the last, follicle-sinus complexes (FSCs) and elements of nervous and muscular tissue are present. According to Dörfl (1982), the proximal ends of FSCs are interconnected by a fibrous plate. Furthermore, the subcapsular zone of hamsters contains rostrocaudally oriented connective tissue fibers (Wineski,1985). As rodent whiskers have no direct contact with the muscles, the aim of this study was to clarify the role of the extracellular matrix (ECM) in transmission of momentum from MP muscles to MP whiskers.
Collagen, which is produced by fibroblasts, is thought to determine the mechanical properties of the MP. Densely packed collagen fibers provide both strength and resistance to tearing and stretching (Caulfield and Borg,1979). Intrinsic collagenous and non-collagenous fluorophors of the skin and muscles can be directly visualized by various fluorescence microscopy methods (Skala et al.,2005). Autofluorescence (AF) of collagen occurs due to crosslinks, including non-enzyme-dependent ones, between collagen fibers. AF of the MP muscles is caused by the presence of collagen in the epi-, peri-, and endomysium, and of other fluorophores, such as nicotinamid and flavin adenine dinucleotides, proteins, and pigments (Andersen and Wold,2003). After tissue staining for cytochrome oxidase (CCO), the AF of free collagen fibers in the MP was not changed, whereas that of muscles due to the presence of collagen and other muscle fluorophors was no longer detectable (Haidarlui, unpublished data). The histochemical reaction for demonstration of CCO activity is based upon oxidative polymerization of diaminobenzidine (DAB), and reflects mitochondrial CCO activity (Seligman et al.,1968; Reith and Schuler,1972). However, conversion of the DAB into a dark homogenously distributed, non-fluorescent product quenches the fluorescence that was induced within the neurons by different fluorescent markers (Lübke,1993). Similar quenching of co-localized fluorescence caused by diffuse DAB staining of cell nuclei was also reported (Yamada et al.,2005). In muscles, mitochondria are distributed within different pools, such as, subsarcolemmal, perivascular, intersarcomeric, and paranuclear (Rothstein et al.,2005), so local absorption of exciting or emitted light by diffuse DAB polymerization products may extinguish muscle AF.
Here, we present evidence that supports the presence of a collagenous skeleton in the rat MP, a skeleton that we assume determines the anatomical and functional integrity of the entire MP. The organization of the collagen within the rat MP was revealed by collagen AF after excitation with UV light in histological slices stained for CCO. Histological samples were obtained from young and adult male albino Wistar rats and processed as previously described (Haidarliu et al.,2010). Briefly, MPs were sectioned into 60 μm thick slices in coronal, tangential (parasagittal), and horizontal planes. The slices were fixed with paraformaldehyde, stained for CCO, and coverslipped with Krystalon (Harleco; Lawrence, KS). Structures containing CCO and collagen were visualized by light and fluorescence microscopy, respectively, using a Nikon Eclipse 50i fluorescent microscope. Excitation of AF was attained at 350–370 nm using a standard Nikon excitation filter CY3, which causes emission of blue AF at ∼440 nm (Odetti et al.,1992; Kollias et al.,1998; Na et al.,2001). These wavelengths are a reliable fluorescent biomarker for collagen (Georgakoudi et al.,2002). Confirmation that collagen was the source of AF was achieved by incubating portions of unstained and CCO-stained tissue slices with Clostridiopeptidase A, EC 184.108.40.206 (0.5 mg/mL in 0.05 M Tris buffer, pH 7.6, containing 0.001 M CaCl2) at room temperature, which produces a collagenolytic effect (Johanson and Pierce,1972).
In unstained tangential slices of the rat MP (Fig. 1A), muscle fibers possessed a blue AF (Fig. 1B). Both extrinsic and intrinsic muscle AF was extensive, and interfered with visualization of collagen fibers by AF (Fig. 1C,D). Staining for CCO activity permitted visualization of tissue morphology by bright-field microscopy (Fig. 1E), while extinguishing the AF of collagen and other minor fluorophors in extrinsic and intrinsic muscles (Fig. 1F–H).
The general architecture of the MP and its collagenous structures was visualized in coronal slices of the snout of young rats by staining for CCO (Fig. 2A) and viewing the distribution of blue AF (Fig. 2B). The borders of the slices were easily determined by the bright AF of the superficial layers of the corium. The fluorescence of the distal ends of the capsules continued into the corium. In fact, at high magnification of slices obtained by cutting the corium at an angle of ∼45° to the surface of the MP (Fig. 3A,B), collagen fibers of the distal end of the capsule continued into the superficial mesh.
In transversal slices of the FSCs of adult rats, a spongy three-dimensional network composed of thin collagen fibers filled the lower part of the cavernous sinus (CS) (Fig. 4A). In the distal area of the CS, thick collagenous trabeculae, reported as early as 1913 (Vincent,1913), were clearly seen in longitudinally cut FSCs (Fig. 4B). The space between the inner surface of the capsule and mesenchymal sheath was filled by a collagen network and trabeculae. The lengths of the capsules appeared to determine the distance between the corium and plate, because their shape was not affected by different manipulations (folding, stretching, and flattening) during histological treatment. This supports the use of the term “rigid bodies” for FSCs in a model of the biomechanical plant of the rat MP (Simony et al.,2010), although it may refer mainly to the capsules of the FSCs, and only partially to their hair shafts.
The proximal ends of the follicle capsules come into contact with a collagen layer that was called the “plate” by Dörfl (1982). In rats, we observed that this plate consists of a spongy network of thin interlacing collagen fibers that form a continuous soft layer that encases the proximal ends of vibrissal follicles. In this collagen layer, which is between proximal ends of FSCs, large areas were more organized, and could be considered as a quasi-hexagonal mesh (Fig. 5). The fibers of this collagen layer, due to splitting and joining, formed hexagons of 30–40 μm in diameter, similar to the mesh of fibers previously described in muscle and adipose tissues of the heart (Caulfield and Borg,1979; Swatland,1987; Pope et al.,2008), and in lung (Cavalcante et al.,2005). This layer of the collagenous skeleton may correspond to the hypodermis of the skin that consists of adipose tissue. In mice, Dörfl (1982) revealed that follicles with their muscle slings are located in adipose tissue. In the MP of the rat, we observed only a thin (∼0.5 mm thick) layer of adipose tissue. More superficially, at the level of blood sinuses and the extremities of the intrinsic muscles, adipose tissue was absent (Fig. 6). The quasi-hexagonal collagenous mesh was attached to the proximal ends of the FSCs. It contained adipocytes filled with droplets of uncompressible fat and surrounded by the woven collagen network that resembles closed cell foam. These characteristics might allow the layer to serve both as a supporting framework for the follicles, as well as a resilience system that provides a uniform distribution of the stresses and strains due to whisking.
Under the layer formed by the quasi-hexagonal deep mesh, we revealed a subcapsular fibrous mat that consisted of thick fibers of collagen, which was similar to the subcapsular layer described by Wineski (1985). The fibers of this layer touched the proximal ends of the vibrissal follicles, and consisted mainly of long thick wavy fibrous fascicles of collagen, with a wide range of sizes, that were oriented rostro-caudal (Figs. 2 and 7). This layer was represented mostly by a single row of collagen fibers, the extensions of which could also be observed in the deep mesh (Fig. 5). However, in the central area of the maxillary part of the fibrous mat, under the FSCs that are represented in the thalamus by the largest barreloids (Haidarliu and Ahissar,2001), two to three rows of thick collagen fibers were observed (Figs. 2 and 8). More caudal, after passing the level of straddlers, collagen fibers of both nasal and maxillary parts of the MP converged and anchored to the maxilla.
The mechanical role of the collagen fibers of the quasi-hexagonal deep mesh and the subcapsular fibrous mat (the two layers described above) might consist of returning the proximal ends of the FSCs to their initial (resting) position after being displaced during a whisker protraction caused by muscle contraction. Furthermore, these collagen fibers might be essential for maintaining a constant position of the proximal ends of whisker follicles relative to each other.
Incubation with Clostridium collagenase of slices that contain the subcapsular fibrous mat revealed a progressive reduction of the intensity of blue AF, which confirms that collagen is a major component of the MP skeleton. Exposure to the collagenase caused the AF of the thin fibers to disappear completely, and although the thicker fibers were still visible, their diameter diminished (Fig. 9). After 90 min of exposure, cloudy material accumulated around thick collagen fibers.
Our findings support the collagenous skeleton of the MP being involved in maintaining follicle and whisker alignment, as well as MP shape and size. The collagen of the MP skeleton is intimately associated with intrinsic and extrinsic muscles, and with FSCs. This collagenous skeleton appears to be the principal force-bearing structure of the rat MP. Although not a bony (“true”) skeleton, it can provide structural support and function for the MP. Within the core of the MP, the collagen skeleton may maintain spatial orientation of FSCs during whisking. When tractive forces produced by the extrinsic muscles are applied to the superficial and deep layers of the collagenous skeleton, they are probably transmitted through linkage “joints” to the proximal and distal ends of the FSC capsules. As the whiskers have no direct contact with the muscles, their active movements are completely controlled by the movements of the FSCs. Within the FSCs, delicate fibers, continuous with the fibers of the capsule, extend across the CS and adhere to the glassy membrane that separates the mesenchymal sheath of the hair follicle from external root sheath (Melaragno and Montagna,1953).
The skeleton of the MP involves one superficial and two deep (quasi-hexagonal mesh and subcapsular fibrous mat) collagen layers, which are connected to each other by the capsules of the FSCs (Table 1). According to the classification proposed by Ottani et al. (2001), the subcapsular fibrous mat may be composed of fibrils that belong to the T-type form, and the quasi-hexagonal mesh may contain fibrils of the C-type. The T-type fibrils are arranged in large and heterogenous bundles oriented parallel to tensile stress along their axes, and are not extendable. During protraction, they straighten, and during retraction, they restore their initial wavy shape. The C-type fibrils are small and homogenous. They resist to multidirectional stresses, forming sheets with a quasi-hexagonal mesh. Similar organization of a meshwork of branched, small diameter fibers composed of Type III collagen was described in adipose tissue of the pig (Swatland,1987). In aquatic animals, such as harbor seals, high mobility of vibrissa can be facilitated by unsaturated fatty acids with low melting points that are present in adipose tissue near the vibrissa (Dehnhardt et al.,1998). In the mouse MP, the plate-like layer appears like a “fenestrated ribbon” (Dörfl,1982). This can now be explained as a representation of the adipose tissue maintained by a collagenous quasi-hexagonal network, the fibers of which are attached to the proximal ends of FSCs.
Multidirectional, forming quasi-hexagons in the plane parallel to the skin
Rostro-caudal, parallel to each other in parasagittal plane
To the distal capsular ends and skin collagen layer, along the perimeter of the MP
To the proximal capsular ends, and to subcapsular fibrous mat by the aid of fiber extensions
Caudal, to maxilla; rostral, to rostro-lateral pole of the premaxilla and nasal cartilage
Our findings support the previously reported tight relationship between hair follicles and adipose tissue in rats (Hausman et al.,1981). During development of the rat hypodermis, adipocytes are observed only around the lower part of the hair follicles, where they form a layer in direct contact with cells of the external root sheath. Cultured dermal papilla of the adult rat vibrissal follicle is also capable of spontaneous adipogenic differentiation (Jahoda et al.,2003).
In the rat mystacial pad, adipocytes and ECM are constantly subjected to spatial deformations caused by whisking. Though the impact of mechanical forces on the differentiation and morphology of the ECM has not yet been studied in depth, certain mechanisms that prevent tissue disruption by mechanical forces have been described (Mariman and Wang,2010). In the rat MP, we observed some features that were characteristic for the adipose tissue of the deep mesh.
iLayer II of the collagenous skeleton was thin (∼0.5 mm) and contained adipocytes encased in a quasi-hexagonal mesh. This mesh was reinforced by collagen fibers, similar to hypodermal adipose tissue (Motta,1975; Comley and Fleck,2010). Some of these collagen fibers originated from the subcapsular fibrous mat (Fig. 5).
iiIn Layer II, the adipocytes were ∼40 μm in diameter. The diameters of adipocytes in other regions of the rat body were reported to be 50–80 μm (Askew and Hecker,1976; Jamdar et al.,1981,1986) whereas those of pigs and humans were reported to be 70 μm and 30 to 70 μm, respectively, (Geerlings et al.,2008).
iiiIn the deep mesh, ECM was continuous and connected neighboring FSCs (Fig. 5). Collagen fibers of the ECM were attached to the lower capsular part of the FSCs.
ivIn the microtome slices of the rat MP, the surface of contacts between the neighboring adipocytes was flat which means that the shape of the adipocytes should be polyhedronic. In the other regions of the rat body, adipocytes are reportedly mostly spherical (Avram et al.,2005).
In light of the above, we conclude that Layer II of the rat MP can be considered to be a layer of adipose tissue in which adipocytes are encased in a strong collagenous mesh. This mesh is a part of the collagenous skeleton, is attached to vibrissal capsules and can function as springs and dampers simultaneously. Non-compressible fat droplets within the collagenous polyhedra facilitate the resilience of this layer during whisking.
In the core of the MP, the collagenous skeleton consists of strong rigid capsules of the FSCs. These capsules are comprised of thick collagen fibers that form a strong wall (Sakita et al.,1994), and can accomplish their function when the sinuses are filled with blood. The sinuses are supplied with blood directly from the branches of the arterioles. High blood pressure within the capsule increases turgidity, which is necessary to stabilize mechanical properties of the vibrissa during sensory exploration (Rice,1993). The distal ends of the capsules continue into the collagen mesh of the corium, and their proximal ends are attached to the collagen fibers of the two deep collagen layers (see above) to form a connection that can function as a semi-ball joint. Such connections enable the whisker shaft to move in all directions within a plane that is parallel to the whisker pad surface.
Attachments of the distal and proximal ends of the FSCs to the relevant collagenous layers create conditions favorable for synchronous vibrissal movements by facilitating side-to-side coupling and by minimizing the possibility of transverse vibrissa deviations during whisking. The wavy, coarse collagen bundles of the subcapsular fibrous mat are crucial viscoelastic structures of the skeleton, which retract whiskers after their protraction. These collagen fibers were oriented in parallel, and along the direction of the largest tension generation, similar to those of the fibrous heart skeleton (see Icardo and Colvee,1998), and were attached, rostral, to the rostrolateral pole of the premaxilla and nasal cartilage, and caudal, to the maxilla.
One unexpected function of the MP skeleton is to facilitate movements, but in opposite directions, of the corium and deep collagen layers, the planes of which are parallel. This is possible because the stiff capsules of the FSCs cannot be compressed or extended during muscle contraction. In fact, the distance between the deep collagen layers and corium increases during protraction, and is maximal when the long axes of FSCs form a rectangle with the deep collagen layers and corium. The collagenous skeleton of the rat MP also limits the extension of the MP tissues. Moreover, the skeleton is a dynamically active entity. The MP skeleton retracts whiskers after protraction caused by muscle contraction, prevents exceeding torsional deformation (Knutsen et al.,2008), and controls movement of the corium in relation to the deep collagen layers of the MP.
On the basis of what is now known about the network of the MP collagenous skeleton, we propose a new schematic bio-mechanical model of the rat MP (Fig. 10). This new 3D model includes three connective tissue layers, whereas a reduced 2D model we previously proposed (Haidarliu et al.,2010) only included two such layers. The first layer (Superficial mesh) consists of a subepidermal layer of connective tissue fibers that are arranged at random and form an interlacing plexus (Fig. 10, Layer I). It is a 2D layer in which forces can be applied along both the rostrocaudal (RC) and dorsoventral (DV) axes. This layer can be pulled caudal by the M. nasolabialis and M. maxillolabialis (corium whisker retractor, CR) to cause whisker retraction, and rostral, by the Pars media superior and inferior of the M. nasolabialis profundus (corium whisker protractor, CP) to cause whisker protraction. The second layer (Deep mesh), which consists of a fenestrated plate composed of a collagenous mesh with encased adipocytes (Figs. 5 and 10, Layer II), is also a 2D layer in which forces can be applied along both the RC and DV axes. The third layer (Subcapsular fibrous mat) consists of parallel collagen fibers anchored rostral to the premaxilla and nasal cartilage, and caudal to the rostral part of the maxilla (Fig. 10, Layer III). It is a 2D layer in which forces can only be applied in the RC direction. Layer III can be pulled rostral by the mat whisker retractor (MR), which is comprised of the three parts of the M. nasolabialis profundus that cause whisker retraction (Haidarliu et al.,2010). This is possible because, in the whisker resting position, the fibers of this layer are undulated, and can be straighten by muscle contraction. Layers I and II act locally during deformation of the surrounding tissues (at anchor points between follicles) whereas Layer III effectively connects all the bases of FSCs in a row, similar to the suggestion of Hill et al. (2008). Even though their bases are connected, removing one FSC from a vibrissal row should not affect the integrity of the collagenous layers of the MP.
The skeleton of the MP includes the origins and insertion sites for the MP muscles. The intrinsic muscles of the MP are attached to the capsules of two neighboring FSCs in a row, and to the corium. The extrinsic muscles of the MP originate from various sites on the facial skull bones or from the nasal cartilage. Their insertion (anchor) sites are in the corium or the subcapsular fibrous mat. In the corium, most of these insertion sites are specifically organized to form spreads of tapered fibers reminiscent of rosettes. Such spreading of insertion sites facilitates synchronous movements with uniform trajectories for all of an individual's vibrissae during whisking. The skeleton receives external loads from the extrinsic muscles, and transfers the impelling momentum to the whiskers.
Incubation with Clostridium collagenase (Clostridiopeptidase A) of MP slices containing fibers of the subcapsular fibrous mat caused a decrease in AF of collagen fibers. A similar progressive degradation of collagen was observed in liver slices upon incubation with bacterial collagenase (Perez-Tamayo and Montfort,1980). During application of Clostridium collagenase to structurally intact collagen fibers of the rat tail, the diameter of the fibers decreases continuously, which indicates a degradation process (Wyatt et al.,2009). Clostridium collagenase does not digest some types of collagen, such as Type VI collagen, a microfibrillar collagen that assembles into hexagonal-like networks (Wiberg et al.,2002; Shi et al.,2010). This might explain why large collagenous fibers, though diminished in diameter, were still seen when we exposed the subcapsular fibrous mat to Clostridium collagenase.
In summary, in the rat MP, we revealed a fibrous collagenous skeleton by fluorescent microscopy of intrinsic MP fluorophores. This skeleton consists of superficial and deep collagen layers interconnected by collagenous capsules of the FSCs. It is an integrated framework that satisfies the functional requirements of whisking behavior. We suggest that the fibrous collagenous skeleton of the MP is able to support the tissues of the MP, to transform muscle-impelling momentum into whisker movements, to facilitate passive whisker retraction, and to control the extension of MP tissue.
Ehud Ahissar holds the Helen Diller Family Professorial Chair of Neurobiology. The authors thank Dr. Barbara Schick for reviewing the manuscript.