The soft palate (velum palatinum) is the musculo-membranous part of the partition separating the proximal portions of the respiratory and digestive tracts. It continues caudally from the hard palate at a transverse level, overlapping the epiglottis to a varied degree. The harmonized functioning of the two organs is extraordinarily important, being involved in both swallowing and breathing. Actually, dysfunction of the soft palate is implicated in the pathogenesis of well-recognized respiratory syndromes, such as obstructive sleep apnea syndrome in human patients (Hamans et al., 2000; Berger et al., 2002) and intermittent dorsal displacement of the soft palate in exercising horses (Holcombe et al., 1999).
In dogs, the soft palate is commonly very long, particularly in the brachycephalic breeds, whose skull bone shortening is not paralleled by a decreased development of soft tissues. The limit between hard and soft palate passes just caudal to last upper molar teeth in mesaticephalic heads and can be over 1 cm caudal in the brachycephalic ones (Evans, 1993).
The correlation between craniofacial morphology and disordered breathing is well known in the dog. Relatively longer soft palate is one of the main factors contributing to pharyngeal narrowing during normal respiratory activity of the brachycephalic breeds (Stockard, 1941; Harvey, 1989), which are frequent carriers of the brachycephalic airway obstructive syndrome (BAOS), affecting the most part of them during their postnatal life (Koch et al., 2003).
The gross anatomy and overall structure of the soft palate has been described in the average dog's head (Evans, 1993). Nevertheless, no descriptive microanatomical study of the dog soft palate has been published, even if it would be useful in view of the manifold and important repercussions of this organ physiology. This study deals with the morphological characteristics of the soft palate in adult mesaticephalic dog breeds and will be followed by a companion paper evaluating the soft palate in brachycephalic breeds with Grade 1 BAOS.
To this aim, the caudal part of the soft palate was considered in both studies, in consideration of the conventional surgical procedure of soft palate resection (Hobson, 1995) that is in use to prevent or resolve the main clinical problems linked to the BAOS in the brachycephalic breeds.
MATERIALS AND METHODS
Specimens from eight homogeneously sized mesaticephalic adult dogs (age range: 5–12 years, mostly German Shepherd mongrels) were obtained following the owner's authorization, within 2 hr after euthanasia at the Department of Clinical Veterinary Sciences of the University of Milan (Italy). Dogs were euthanized for a variety of causes (predominantly for tumors), but had a normal oral cavity, and no clinical history of upper airway disease. The distal part of the soft palate was dissected along a curved line crossing transversally from the caudal border of the fossa tonsillaris to the same point on the opposite side and inspected to confirm the absence of any gross lesion. The soft palate specimens were distended, pinned out on a thin cardboard (Fig. 1a), and fixed by immersion in 10× neutral formalin, for at least 7 days.
After fixation, each specimen of soft palate was subdivided transversely into two equal parts (A and B in Fig. 1a), dehydrated in a graded series of ethanol, and paraffin embedded. Starting from the more oral side, and proceeding aborally with serial sections, each block was further subdivided into three parts and the first sections from each third were collected to obtain six different comparable levels per each initial sample (Fig. 1b).
For microscopic evaluation of qualitative histological features, 4- to 6-μm thick sections from each level were routinely stained with hematoxylin and eosin technique (H&E) and with Mallory trichromic stain to provide a better definition of the major tissue types constituting the soft palate.
A sequential staining technique using Alcian Blue (AB) at pH 2.5 (A3157-Alcian blue 8GX, Sigma–Aldrich, Milano, Italy) followed by Periodic Acid-Schiff (PAS) was applied to assess mucin types within glandular tissue and to investigate the connective tissues. With this technique, acidic glycoconjugates stain blue, and neutral PAS positive glycoconjugates stain magenta. Mixtures of the two stain blue/purple.
High iron diamine (HID)/AB 8GX, pH 2.5, reaction was applied to differentiate sulfated (brownish-black stained) from carboxylated-nonsulfated (blue stained) glycoconjugates within the acidic glycoconjugate category.
At the initial phase of the research, one additional sample of soft palate was serially sectioned in its entirety and stained with H&E, to obtain precise anatomical information, and to better map the sites of dissection.
Serial sections were observed and photographed under an Olympus BX51 photomicroscope equipped with a digital camera and DPsoft software (Olympus, Italy) for computer-assisted image acquisition.
The tissue layers composing the distal part of the soft palate in adult mesaticephalic dogs were characteristically organized into a major deep musculo-connective axis mixed with salivary glands and covered by the mucosal lining on either the nasopharyngeal or the oral sides (Fig. 2).
The palatine mucosa was lined by a pluristratified, noncornified, squamous epithelial lining, consisting of a fewer number of cell layers at the nasopharyngeal side (Fig. 3a) than at the oral one (Fig. 3b), with a symmetrical progressive thinning all along the margins. In the most aboral sections, the oral epithelium was so reduced in thickness that a difference with the nasopharyngeal side could not be evidenced. A continuous thin PAS-positive basement membrane was evidenced at the epithelium-connective junction. The basement membrane had a straight aspect at the nasopharyngeal side and a wavy course at the oral side, as the connective tissue was organized in irregular finger-like papillae, projecting, and interconnecting with the superficial epithelium to allow anchorage with the superficial epithelium organized in finger-like projections.
Two major connective tissue layers were evidenced composing the chorion. A thin band-like homogeneous stratum of fine, tightly packed, connective tissue fibrils could be seen at subepithelial level (Figs. 2, 3) with fibrils arranged in short irregularly interlacing bundles. The underlying thicker connective tissue layer had broad but more dispersed fibrils, irregularly organized in collagen bundles mostly parallel to the mucosal epithelium. The fibrils of the subepithelial connective tissue were intensely PAS-positive, whereas the deep chorion was characterized by weak PAS positivity mixed with interfibrillar AB-positive acidic mucopolysaccharides.
A central, longitudinal laminar structure composed of dense and compact, intensely eosinophilic connective fibers, representing the aponeurosis palatina, was demonstrated starting from the more oral sections (Fig. 4a). Characteristically, the aponeurosis was thicker in the lateral parts of the soft palate and contained prominent venous and arterial blood vessels. Toward the median portions of the soft palate, the aponeurosis shifted dorsally and incorporated the muscular and glandular components of this area. Thickness progressively decreased and the aponeurosis gradually disappeared in the last third of the tissue specimens, being completely absent in the last sixth of the soft palate.
Glandular and muscular tissues filled the axial part of the organ (Fig. 2). Palatine glands were distributed in the respiratory and oral side with distinctive differences in the quality of the secretory products, as demonstrated by acinar morphology (Fig. 4c,f) and histochemical (AB/PAS and HID/AB) stainings (Fig. 4d,e,g,h) The nasopharyngeal side of the soft palate contained a moderate number of palatine glands just underneath the nasopharyngeal epithelium, somewhere mixed with the upper musculature (Fig. 2). The majority were serous glands, and only few were mixed glands with separate mucous and serous acini (Fig. 4c), whereas mixed acini were rarely observed. Serous cells had a finely granular, intensely basophilic, PAS-positive cytoplasm, and small round nuclei. Mixed acidic and neutral glycoconjugates characterized the mucous acini (Fig. 4d). Among acidic glycoconjugates, mainly HID-positive sulfated mucins were detected (Fig. 4e). Tubular or tubuloalveolar glands of the oral side produced a thick layer in the soft palate, between the mucosa and the palatine musculature (Fig. 2), organized into several lobules. They were mostly composed of mucous acini having lightly eosinophilic, foamy, and pale cytoplasm, with basal nuclei, whereas intensely eosinophilic, thin and flattened serous cells surrounded mucous cells in mixed acini that represented a minority of the glands (Fig. 4f). By AB/PAS stain, acidic and neutral glycoconjugates were evidenced as the secretory products, with a prevalence of the acidic glycoconjugates inside the glandular ducts (Fig. 4g). After HID/AB reaction, only HID-positive sulfated components were observed, whereas few acini contained carboxylated sialomucins (Fig. 4h). Both the nasopharyngeal and the oral glands decreased in number and size toward the aboral sections. Occasional lymphocytes and plasma cells were present in the glandular stroma. Excretory ducts from serous glands were empty or showed PAS-positive secretion. In glandular lobules with mixed acini, ducts contained mixed PAS- and AB-positive secretory material (Fig. 4d).
In general, the epithelial lining of the glandular ducts was cubic to columnar in the larger ducts, and merged into the superficial epithelium of palatal mucosa. Mild but morphologically visible duct dilation was occasionally present, and a focal mucous spilling was found associated with areas of lumen dilation in two mesaticephalic dogs.
The muscular component was represented by the two longitudinally coupled palatini muscles (Fig. 4a,b), which composed the midsagittal plane and by dorso-lateral muscles less well organized and morphologically distinct. Dorsally to the palatine aponeurosis few oblique and thick fibers composed the so-called levator veli palatini muscle (Fig. 4a). Additional smaller muscle fiber bundles localized in the caudal palate laterally to the palatine muscles most likely represented the endings of the palatopharyngeus muscle (Fig. 4b).
The palatinus muscle was composed of homogeneously sized, small and rounded and fibers, whereas the levator veli palatini and palatopharyngeus demonstrated loosely arranged larger fibers with a certain degree of variability in size and shape (Fig 4a,b). Extremely rare and focal areas of muscular degeneration were noticed in the soft palate of mesaticephalic dogs.
In the more aboral sections, glandular and especially muscular tissues decreased together with the palatine aponeurosis and could not be observed in the last sixth of the soft palate.
Microscopical analysis of the caudal soft palate in a group of eight homogeneously sized mongrel mesaticephalic dogs identified a tidy and constant stratigraphy, characterized by a major deep musculo-connective axis mixed with a thick layer of palatine salivary glands enveloped by a flat dorsal nasopharyngeal mucosa and a more folded, ventral oral one.
The epithelial-connective tissue boundary was found to be differently organized in the nasopharyngeal versus the oral side of the soft palate, with well-developed connective tissue papillae at the oral side, which are absent at the nasal one. The connective tissue papillae are regarded as adaptive structures which enlarge the epithelial-connective tissue interface to achieve a broader anchorage for the epithelium and to provide a larger exchange surface for nutritional purposes (Paulsen and Thale, 1998). Thus, the differently arranged epithelial-connective tissue boundary might sustain a different functional need at the two mucosal sides.
Two main layers were observed constituting the lamina propria of canine soft palate, characterized by differences in collagen fiber size and arrangement. Also, similar findings have been observed in the normal human soft palate (Paulsen and Thale, 1998). The combination and organization of elastic, reticular, and collagen fibers mixed with glycosaminoglycans are responsible for the mechanical properties of tensile strength and simultaneous flexibility that characterize the healthy soft palate (Paulsen and Thale, 1998).
As far as the thick layer of palatine salivary glands is concerned, they ensure constant humidity and lubrication of the palatine mucosa, to prevent friction damage during food intake (Kuehn and Moon, 2005). At the oral side, AB/PAS stain showed acidic and neutral glycoconjugates as glandular secretory products, and after HID/AB reaction, mostly HID-positive sulfated components were observed, which are known to make the mucous material thicker. This feature, together with the more conspicuous number of glandular lobules at the oral side of the soft palate, is probably linked to a higher functional demand, where diet factors are mostly involved (Kurahashi, 2002). A prevalence of serous acini was in fact present at the nasopharyngeal side, providing for a more liquid, less protective saliva.
The palatine musculature controls the shape and position of the soft palate. The paired muscles arise from nearby hard tissue structures and fan out to form the palatine aponeurosis in the center of the soft palate, except the palatines which disperse into the caudal free border of the soft palate. The lack of skeletal attachment at one end is likely to have implications for the capacity of these muscles to develop static contractions and high tension (Stål and Lindman, 2000).
The anatomy of the palatine musculature of mesaticephalic dogs recalls that described in human (Stål and Lindman, 2000) and equine (Hawkes et al., 2010) healthy soft palate, including the finding of smaller fiber diameters in the palatinus compared with the other soft palate muscles. In the human soft palate, Stål and Lindman (2000) suggested this may be due to specific functional requirements of the uvula (equivalent of the canine palatinus muscle) that could be related to the specific anatomy of such muscles, where skeletal insertions are absent at one end. Also, a certain degree of size variability, which was observed in the dog soft palate muscles other than the palatines, was considered normal either in the human (Stål and Lindman, 2000) or in the equine (Hawkes et al., 2010) healthy soft palate.
In conclusion, the data presented in this paper add information on the histology of the caudal part of the soft palate in mesaticephalic dogs, filling a gap of the scientific literature on this argument. Moreover, it will also be a baseline in the forthcoming analysis of the soft palate thickening in brachycephalic dogs and consequent reduction of soft palate function linked to BAOS.
Mr. Paolo Stortini (VSA, Università degli Studi di Milano) is gratefully acknowledged for his experienced and skillful technical support. Excellent technical assistance in image editing made by Mr. Marco M. Colombo (digital creative and webmaster, Università degli Studi di Milano) is specially acknowledged.