New aspects of the esophageal histology of the domestic goat (Capra hircus) and European roe deer (Capreolus capreolus)

Abstract The present study examines the esophageal wall of animals from two distinct families of the Ruminantia: domestic goats and European roe deer. Five fragments were collected from the entire length of the esophageal wall in five goats and four roe deer and subjected to microscopic and morphometric analyses. All layers of the esophageal wall except the tela submucosa were found to be thicker in the goats. In both species, the esophagus was lined by parakeratinized stratified squamous epithelium, and the tela submucosa was deprived of glands along its entire length. However, the structure of the lamina muscularis mucosae was better developed in goats: it was found to be discontinuous in the proximal part, and then became fused in the cervical part, that is around the most proximal quarter of its length. In contrast, in roe deer, the lamina muscularis mucosae began as sparse, thin muscle bundles at the pharyngeal‐esophageal junction, which thickened and clustered further down the esophagus, but did not fuse. Our findings regarding the microscopic structure of the ruminant esophagus are not fully consistent with the widely‐accepted view and suggest that the histological structure of the esophagus demonstrates interspecies variation within this large suborder. More precisely, species‐specific differences can be seen regarding the presence of esophageal glands and parakeratinized epithelium, and in the organization of the lamina muscularis mucosae.

On the contrary, Ruminantia is a very large suborder of mammals that includes six families living in varied environmental conditions in almost all continents. Although domestic ruminants belong to the Bovidae family, they exist within two subfamilies: Bovinae (cattle) and Caprinae (goats, sheep). It is widely known that some anatomical differences exist between large and small domestic ruminants (König & Liebich, 2007;Kumar et al., 2018) and among various families of domestic and wild ruminants (Mbassa, 1988;Vidyadaran et al., 1999). As such, it is reasonable to assume that such differences may also concern the histological structure of the esophagus. Therefore, the aim of the present study was to compare the structures of the esophageal wall of two species, viz. the domestic goat (Capra hircus) and European roe deer (Capreolus capreolus), representing the Bovidae and Cervidae respectively, based on microscopic and morphometric analysis.
The goats were kept in confinement (i.e., a barn housing system).
They were fed maize silage and crushed oats, which were administered once a day, and good quality meadow hay was provided each morning.
Water and salt-lick were constantly available to the animals ad libitum.

Tissue collection and preparation
The tissue specimens were collected within 1 h of animal death. All esophagi were removed together with pharynx. Incision of the distal end of esophagus was made just above the rumen inlet.
Five fragments of esophagus approximately 3 cm in length were collected from each animal. The first (cranial) included the area directly adjacent to the pharynx; the next samples were taken from the most caudal (caudal) and middle part of esophagus (middle). The last two fragments (mid-cranial and mid-caudal) were collected from the middle area of the proximal and distal parts of the esophagus that remained after collection of the first three fragments. Following this, each esophageal fragment was cut into two equal parts: the first was used for longitudinal sectioning and the second for transverse sectioning.
All esophageal fragments were fixed in 10% neutral buffered formalin, processed by the common paraffin technique and cut into 3 µm specimens. In total, 24 slides were prepared for each esophageal fragment from each animal: 12 longitudinal sections and 12 transverse sections. In each case the tissue specimens collected for analysis were separated from each other by a distance of 100 µm that is after obtaining each section 100 µm of paraffin block was cut and discarded. All specimens were stained with hematoxylin and eosin, Ayoub-Shklar and Weigert-Van Gieson methods. The Weigert-Van Gieson method visualizes collagen and elastic fibers and differentiates connective tissue and muscle tissue (Bagiński, 1965). The Ayoub-Shklar method differentiates keratin from other tissue components (Ramulu et al., 2013).

Microscopic and morphometric analysis
All tissue specimens were subjected to histological examination by light microscopy. In addition, the following individual morphometric measurements were made: the thickness of the epithelium, lamina propria and lamina muscularis mucosae, as well as the tela submucosa and tunica muscularis (thickness of each muscle layer was measured separately). The tunica adventitia was excluded from the analysis, because it was removed during tissue collection. Only tissue areas without any pathological lesions were included in the morphometric analysis.
For each animal, nine randomly-selected specimens were analyzed from each esophageal fragment: three from the longitudinal sections and six from the transverse sections. In each of these specimens, measurements were performed in five randomly-selected, equidistant microscopic fields, localized along the entire length of each longitudinal section, or along the perimeter of each transverse section. Each microscopic field was displaced from the previous one by 300 µm for the longitudinal sections and 500 µm for the transverse sections.
All specimens were viewed at a magnification of 100x. Morphometric analysis was performed under a Nikon Eclipse 80i microscope equipped with a Nikon DS-Ri1 camera (Nikon, Tokyo, Japan) using NIS-Elements BR microscope imaging software (Nikon). The micrographs were taken in all microscopic fields designed according to the algorithm described above. They represented a part of microscopic field at magnification of 100x with dimensions of 942.5 × 1178.4 µm.
These images were used for morphometric analysis. Measurements perpendicular to the surfaces of particular layers were taken using specialized distance tools ('Horizontal' and 'Parallel lines'). As the thickness of the esophageal wall differed considerably not only between, but also within individual micrographs, especially in the transverse sections, three measurements were taken for each layer per micrograph. These values were used to calculate the mean thickness of each layer. This value was included in further analysis. The three measurements were taken at the thinnest and thickest points of the layer, and at a point of medium thickness. Where no considerable variation in esophageal wall thickness was observed, the measurements were taken from the opposite sides of the micrographs, with third one in the middle.
Thus, the following parameters were calculated: the thickness of each layer and the total thickness of the esophageal wall for each analyzed esophageal fragment and for the whole organ.
The results are presented as median and range (min-max). As only a small number of individuals were tested, no statistical analysis was performed of the obtained data. All presented data refer to all the analysed animals within a single species. The measurements taken from individual animals are presented in Supplementary files.

RESULTS
The esophagi of both species were organized into distinct layers: tunica mucosa, tela submucosa and tunica muscularis ( Figure 1). None of them contained obvious pathological lesions, except the presence of single Sarcocystis spp. scattered throughout the tunica muscularis in all roe deer.
All esophagi possessed longitudinal folds created by the tunica mucosa and tela submucosa. These folds were clearly visible on trans-verse sections, giving a star-like appearance to the esophageal lumen ( Figure 1a). Usually they were high, with deep grooves between them, pushing the epithelium into close proximity to the lamina muscularis mucosae.
The median thickness of entire esophageal wall was 1840.9 µm (704.1-3999.8µm) in goats and 1778.4 µm (870.7-5111.6µm) in roe deer. All layers except the tela submucosa were thicker in goats. A graphical comparison of the thicknesses of particular layers of the whole esophagus with regard to species is given in Chart 1.
Tunica mucosa. In both species, the tunica mucosa consisted of typical layers: the epithelium, lamina propria and lamina muscularis mucosae. Also, in both species, the epithelium was the stratified squamous type. Remnants of nuclei were observed in the superficial layers of the epithelium; in addition, Ayoub-Shklar staining confirmed the presence of keratin, indicating parakeratinization of the epithelium ( Figure 2).
In both species, the epithelium was of unequal thickness (Figures 1b,2a,3a,c,4d,5c,d,6b and 7c,d F I G U R E 1 General histology of esophagus (a) and its wall with all layers marked (b) (transverse sections, goat). Tunica mucosa and tela submucosa create longitudinal folds giving the esophageal lumen a star-like appearance (a). Hemotoxylin and eosin staining. Scale bars = 1000 µm (a) and 500 µm (b) Abbreviations: E, epithelium; IML, inner muscular layer; LMM, lamina muscularis mucosae; LP, lamina propria; OML, outer muscular layer; TS, tela submucosa. however, in one animal, they were shorter, less numerous and more widely spaced (data not shown). In transverse section, the lamina muscularis mucosae was visible as a relatively thin layer consisting of small All values are given as median (min-max). muscle bundles lying individually or in small groups (Figures 4c,d).

C H A RT 1 Comparison of median thickness of particular layers of esophagus between species
Some of them were separated by large areas of connective tissue, whereas others formed a relatively compact layer.
In the mid-cranial fragments of the esophageal wall visible in longitudinal section, the lamina muscularis mucosae became continuous; however, in three animals, some muscle bundles were still separated by a short distance. In transverse section, the muscle bundles were larger and located closer than in the previous fragment of the esophageal wall. Moreover, this layer was more evident than in the cranial fragment (data not shown).
In the remaining fragments of esophagus, the lamina muscularis mucosae was arranged in a continuous manner in longitudinal section (Figures 5a,b). In transverse section, it formed a clearly-visible continuous layer consisting of relatively closely-packed muscle bundles (Figures 5c,d).
The median thickness of this layer throughout the whole esophagus was 69.7 µm (15.9-549.5 µm) (Chart 1). Its thickness progressively increased from the cranial to the caudal fragment (Table 1, Chart 2).
In roe deer, the lamina muscularis mucosae remained noncontinuous along its entire length. In the transverse section of the cranial fragment of the esophageal wall, the lamina muscularis mucosae of all animals could be seen to comprise small muscle bundles, which were sparsely and irregularly distributed within the connective tissue ( Figure 6b). Although most were separate, some were arranged  (Figures 7a,b).
In the transverse sections, the lamina muscularis mucosae of the mid-cranial fragment was easier to recognize than that of the cranial fragment. The muscle bundles were larger than in the cranial fragment F I G U R E 3 Localization of collagen and elastic fibers within lamina propria and tela submucosa of goat (a and b) and roe deer (c and d) esophagus in transverse sections. Elastic fibers (black) are the most numerous in lamina propria (a-c). The tela submucosa consists mainly of collagen fibers (red). In tela submucosa elastic fibers are sparse, slightly more numerous at the border of the tela submucosa and tunica muscularis (d). The muscles of lamina muscularis mucosae and tunica muscularis, as well as epithelium, are stained yellow. Weigert-Van Gieson staining. Scale bars = 100 µm (a and c) and 50 µm (b and d) and usually grouped together (Figures 7c,d). Although they were still distinctly separated by connective tissue, they were closer together, with a consistent distance between them.
In the longitudinal sections of the middle fragments of the esophagi, the muscle bundles were arranged in a similar manner as in previous fragments; however, the bundles differed in length: in one animal, they were visibly longer than in other animals and were of comparable size to the muscle bundles observed in the mid-caudal and caudal fragments, sometimes extending beyond the visual field at 100x magnification. In transverse section, the lamina muscularis mucosae was well developed and formed a continuous layer consisting of large, closelypacked muscle bundles (data not shown).

F I G U R E 7 Representative image of mid-cranial fragment of roe deer esophagus in longitudinal (a and b) and transverse (c and d) sections.
Lamina muscularis mucosae is non-continuous, but smooth muscle bundles are thicker and, in transverse section, located closer and usually arranged in groups. Smooth muscle bundles are marked with brackets (a and c) or arrows (b and d). Hemotoxylin and eosin staining. Scale bars = 500 µm (a and c) and 100 µm (b and d) In the longitudinal sections of the mid-caudal and caudal fragments of the esophageal wall, the muscle bundles were longer than in previous fragments; however, they were of comparable length (Figures 8a,b).
The gaps between them were either similar to those observed previously or smaller. In transverse section, the lamina muscularis mucosae in the mid-caudal and caudal fragments had a similar appearance to the previous fragment (Figures 8c,d).
The median thickness of this layer was 51.6 µm (10.7-277.6 µm) along the entire esophagus (Chart 1), and the value progressively increased from cranial to caudal fragment (   ,d). However, in one roe deer, large numbers of adipocytes were observed in the cranial fragment ( Figure 6). Adipose tissue was more abundant in goats than in roe deer. No esophageal glands were observed in either species (Figures 1-8).
In both species, the tela submucosa demonstrated great variations in thickness resulting from the presence of high longitudinal folds ( Figure 1). The median thickness of the tela submucosa along the entire esophagus was 282.9 µm (48.6-1243.8 µm) in goats and 370.2 µm (51-1808.3 µm) in roe deer (Chart 1). In goats, the tela submucosa was thinner in three first fragments than the last two and was thickest in the caudal fragment. In roe deer, it was thinnest in the middle fragment and thickest in the cranial fragment (Table 1, Charts 2,3).
Tunica muscularis. In both species, the tunica muscularis was composed exclusively of skeletal muscle fibers (Figures 1 and 3-8 In goats, the tunica muscularis and its inner layer were thickest in the caudal fragment, whereas the outer layer was thickest in the cranial fragment. In contrast, the tunica muscularis and both its layers were thinnest in the middle fragment (Table 1, Chart 2). In roe deer, the whole tunica muscularis and its two layers were thickest in the caudal fragment, while the muscular layer was thinnest in the middle fragment of the esophagus (inner muscular layer) or the mid-cranial fragment (outer muscular layer and whole tunica muscularis) (

DISCUSSION
This study compared the histological structure of esophagus between two ruminant species: the domestic goat and the European roe deer.
While some previous studies have examined the morphology of the esophagus in goats (Ebraheem et al., 2018;Islam et al., 2005;Islam et al., 2008;Kumar et al., 2009), no such studies appear to have been performed in roe deer. Our present approach is more detailed than the four studies on goats given above: whereas Ebraheem et al. (2018) include only one small sample of the anterior portion, and Kumar et al. Our present findings reveal considerable differences in thickness of esophageal wall layers along the entire esophagus, especially in the tela submucosa. The structure of this layer allows the relatively inelastic mucosa to be pushed into longitudinal folds which obliterate the esophageal lumen, except during deglutition (König & Liebich, 2007).
The thickness was frequently found to differ even as much as tenfold in the case of high folds with deep grooves. The differences in thickness of tunica mucosa also reflects, to a large extent, its histological structure. The lamina propria protrudes at intervals into the epithelium to form dermal papillae, and these can protrude into as much as 50% of the epithelium in humans (Long & Orlando, 2002). This can explain the enormous range of thicknesses observed for esophageal layers in our present analysis.
However, in carcasses, the thickness of the esophageal wall can also be attributed to contraction of the esophageal musculature associated with rigor mortis. Generally, this phenomenon starts developing within 1-2 h after death (Shivpoojan, 2018); however, it has been observed at 12-15 h post-mortem in beef carcasses (Bodwell et al., 1965). In wild-tailed deer, rigor mortis usually develops in the jaw or of the neck 2 h after death, with a 75% chance of none or partial rigor mortis 1 h after death (Oates et al., 1984). As the present samples of roe deer esophagi were collected within 1 h after death, the impact of rigor mortis on the thickness of esophageal wall appears to be minimal. In addition, as the goats were euthanized by pentobarbital overdose, the skeletal muscles were unlikely to contract (Ingalls et al., 1996;Lapointe & Côté, 1999;Thesleff, 1956).
In contrast, the roe deer were shot during hunting, which would result in cadaveric spasm: a condition in which a group of muscles that were used profusely just before death becomes stiff and rigid immediately after death (Fierro, 2013;Madea, 2013). However, in these cases, this would have affected the skeletal muscles of musculoskeletal system, with only minimal impact to the muscles of the internal organs.
It is possible that the varying thickness of particular layers of esophageal wall in roe deer could have resulted from increased muscle tone created by stress. Hunting, as a stress factor, is known to increase cortisol concentration, and high serum cortisol levels can increase smooth muscle tone (Xiao et al., 2003) and influence the activity of skeletal muscles (Tosato et al., 2015). However, deer shot in the field has been found to have low or average serum cortisol concentrations (Smith & Dobson, 1990); furthermore, the epithelium and lamina muscularis mucosae were not strongly folded, and the degree of waviness was similar to that observed in goats, suggesting similar muscle tension.
Our study indicates that all the layers of the esophageal wall are thicker in goats; with the exception of the tela submucosa, which was thicker in roe deer. This can reflect facts that goats eat firmer food (thicker epithelium), and they are able to swallow food boluses of larger volume (thicker tunica muscularis) (Ebraheem et al., 2018). Furthermore, in both species, the tunica muscularis was relatively thick in the cranial fragment (especially the outer muscular layer in goats), but thinned in the next one or two fragments before expanding gradually in the final two fragments. Increased thickness of tunica muscularis in the cranial esophageal fragment can reflect the presence of muscles oesophagei longitudinales that bind the esophagus to adjacent tissues, and lie externally to the esophageal muscle layer (Schaller & Constantinescu, 2007). These muscles attach to the external surface of esophagus, thus they could increase the thickness of the outer muscular layer. Indeed, we found that the outer muscular layer was thicker in the cranial fragment than the mid-cranial fragment, especially in goats.
However, we did not observe any distinct additional muscle layer presented in this esophageal fragment or areas of outer layer with disturbed arrangements of muscle fibers indicating the presence of muscles oesophagei longitudinales. To clarify this issue, further study is needed.
However, our study only includes a small group of animals, and further studies on larger number of individuals are needed to confirm our morphometrical results.
Possibly our most surprising finding is that no esophageal glands were present in the tela submucosa of both examined species. All textbooks of veterinary histology indicate that ruminants possess esophageal glands, with only slight differences existing in their precise localization: either in the cervical region of the esophagus (Banks, 1993;Islam et al., 2005;Kuryszko & Zarzycki, 2000) or at the pharyngeal-esophageal junction (Bacha & Bacha, 2006;König & Liebich, 2007). However, our findings indicate a complete absence of esophageal glands in any of the histological sections taken from either species, including in the cranial fragment, which was always collected from the area directly adjoining the pharynx. It is surprising that no esophageal glands were identified at the pharyngeal-esophageal junction area, considering the large number of sections examined from the anterior esophagus. Although similar findings were noted in two previous studies on goats (Ebraheem et al., 2018;Kumar et al., 2009), these should be regarded as tentative, as the methodology only included one tissue specimen collected from the cervical part of esophagus and did not specify the exact area from which it was taken. Even so, Saxena and Klimbacher (2019) also report a lack of esophageal glands in samples from sheep and cattle. Taken together, these findings, and our present ones, raise the question of whether ruminant species have esophageal glands at all, and if so, whether all species have them.
A number of histology textbooks (Bacha & Bacha, 2006;Kuryszko & Zarzycki, 2000;Samuelson, 2007) indicate that the ruminant esophageal epithelium is cornified. This belief results from the widelyaccepted opinion that the degree of keratinization of the epithelium depends on the coarseness of the diet (Banks, 1993;König & Liebich, 2007), and being herbivores that ingest hard and dry food, it is reasonable to assume that ruminants possess a keratinized epithelium (Kuryszko & Zarzycki, 2000). However, our present observations did not indicate the presence of fully keratinised anucleated scale-like cells (Deo & Deshmukh, 2018) anywhere in the esophagi originating from either species. Rather, the esophageal epithelium of both species appeared to be parakeratinized which is a form of epithelial keratinization (Rao et al., 2014). In humans, such parakeratinized stratified squamous epithelia, and more commonly orthokeratinized epithelia, are associated with masticatory mucosa such as the hard palate, gingiva, dorsal surface of the tongue and the masticatory surfaces of the dental arches in the edentulous mouth (Groeger & Meyle, 2019;Nanci, 2018).
The presence of this form of keratinization on the skin is considered a disease state, such as psoriasis, where it typically indicates an increased rate of epidermal turnover (Nanci, 2018). Therefore, parakeratinization is believed to be a unique histological feature of a healthy oral cavity (Nanci, 2018): a highly-keratinized masticatory mucosa serves to dissipate shearing forces to a greater degree than a more flexible non-keratinized epithelium, which is also found on the esophagus (Squier & Kremer, 2001). Therefore, it is possible that in goat and roe deer esophagi, parakeratinization of the epithelium provide protection against rough food while maintaining greater flexibility than a highly cornified epithelium.
This is the first record of such parakeratinization in either the masticatory or esophageal mucosa of ruminants. However, such pattern of epithelial maturation has been observed in the rumen, either as a pathological condition (rumen acidosis) or linked to feed type (concentrated diet) (Bull et al., 1965;Steele et al., 2011). Our results suggest that parakeratinization could be a more common phenomenon in the ruminant digestive tract than previously thought; this is supported by previous descriptions of goat esophageal epithelium (Kumar et al., 2009) and images of goat esophagus (Ebraheem et al., 2018) that suggest the presence of epithelial parakeratosis.
Information regarding the exact structure of lamina muscularis mucosae in ruminants is sparse; however, the consensus is that it consists of longitudinally-arranged smooth muscle cells, forming isolated, scattered muscle bundles towards the anterior part of the esophagus, and that these bundles fuse to form a continuous layer posteriorly (Bacha & Bacha, 2006;Banks, 1993;Kuryszko & Zarzycki, 2000). However, no detailed studies have been conducted on ruminant species.
Although our present findings partially fill this gap, not all of our observations are consistent with available data, and clear interspecies differences were found in the structure of the lamina muscularis mucosae. Our results indicate that the lamina muscularis mucosae is better developed and thicker in goats than in roe deer. In the roe deer, it starts from individual, thin muscle bundles which gradually become thicker, longer and move closer together, achieving its final form in the middle fragment of the esophagus, that is, the cervical region. In contrast, in the goats, the lamina muscularis mucosae consisted of thick, long, well-developed muscle bundles throughout the entire length of the esophagus; however, while the bundles were separated by narrow gaps in the cranial fragment, this layer became continuous in more distal fragments. The area of transition was localized between cranial and mid-cranial fragments, in cervical region of esophagus.
The changes in the structure of lamina muscularis mucosae refer not only to the size of the individual muscle bundles and distance between them but also to their number. The number of individual muscle bundles scattered along perimeter of esophagus gradually increased together with increasing of their size. Such arrangement of lamina muscularis mucosae may provide a wide enough room for the wide expansion of the esophageal lumen during swallowing of foods in its cranial fragment and maintaining tension against the pressure from the internal cavity in the remaining part of esophagus (Nagai et al., 2003).
Our findings confirmed the structure of goat lamina muscularis mucosae described by veterinary textbooks (Bacha & Bacha, 2006;Banks, 1993;Kuryszko & Zarzycki, 2000); however, our observations indicate that the area where the separated muscle bundles fuse to form a continuous layer appears to be significantly closer to the pharynx, that is, at around the most proximal 1/4 of the esophagus. In addition, our findings suggest that this arrangement of muscle bundles is not characteristic for all ruminant species, and that at least in European roe deer, this layer is non-continuous throughout the length of esophagus.
The present study describes a comparative analysis of the histological structure of the esophagus in two species belonging to the Ruminantia suborder that is, domestic goats and European roe deer.
Certain interspecies variations were observed, especially regarding the arrangement of lamina muscularis mucosae. Moreover, some discrepancies were identified between our findings and the widelyaccepted description of the ruminant esophagus presented in histology textbooks. The main differences refer to the presence of esophageal glands, keratinisation of the epithelium and the organization of the lamina muscularis mucosae. Our findings suggest that interspecies variations exist in this large suborder of mammals, and that its members do not share a single common scheme regarding the histological structure of the esophagus. Therefore, to gain a fuller understanding of the histology of the ruminant esophagus, further studies are needed based on more detailed microscopic examinations of the esophagi of other species of domestic and wild ruminants.