Supportive fibrous tissues of the nasal epithelium with special reference to the site-dependent difference

Abstract

The nasal mucosa and submucosa likely contain both vascular beds against cold and dry air and resident immunoreactive cells against various antigens. Therefore, a specific fibrous structure seems to be necessary. Using histological specimens from 20 elderly cadavers, we examined the nasal mucosal and submucosal architecture. The ciliated columnar epithelium of the nasal mucosa was characterized by 1) a thick basal lamina, 2) few elastin-positive fibers beneath the epithelium, that was quite different from the nearby mucocutaneous junction area with a thick layer (0.3–0.8 mm) of elastic and oxytalan fibers corresponding to the skin dermis, 3) CD34-positive cells distributing diffusely in the submucosal tissue, and 4) few smooth muscle actin (SMA)-positive fibers beneath the epithelium. Some of submucosal fibrous structure appeared to express both elastin and CD34. CD34-positive arterioles were abundant beneath the ciliated epithelium, but they appeared negative for SMA antibody that cross-reacts with endothelium. Notably, the ciliated columnar epithelium was thin in the lateral wall of the nasal cavity, while the inferior concha carried the thick pseudostratified columnar epithelium. Strangely, the inferior or palatal wall of the nasal cavity was covered by the thick stratified epithelium. We found SMA-positive mucosal venous plexus in the lateral wall of nasal cavity, but the submucosa was filled with glands in the inferior concha. Vascular beds might be replaced by glands in the nasal submucosa. The site-dependent difference in the mucosal morphology as well as the absence of vascular beds might be a result of secondary change with aging.

Introduction

To resist against cold and dry air, the nasal submucosal tissue contains well-developed venous plexus with arteriovenous shunts [1-3]. Simultaneously, the nasal epithelium is always suffering from various antigens including dusts and viruses and, thus, allergic reaction and inflammation frequently occur in usual daily life. Therefore, the nasal subepithelial fibrous tissue should maintain and mechanically support both the vascular bed and the immunoreactive cell. The resident lymphocytes and macrophages are densely distributed in mucosal specimens from donated cadavers of anatomical gift [4]. Therefore, when they died at the 80–100 years old of age, numerous immunoreactive cells were still alive in the nasal mucosa and submucosa.

Our group recently demonstrated that the nasal squamous epithelium is lined and supported by elastin-positive, thick fibrous structures that was composed of both elastic and oxytalan fibers [4, 5]. In histological sections with specific stains for elastic fibers, oxytalan fibers are identified as wavy fibers much thinner and shorter than elastic fibers [6, 7]. In the human body, elastic fibers usually accompany smooth muscles, as typically seen in the thick arterial wall [8] as well as the pelvic floor [9, 10]. However, commercially-obtained antibodies of the smooth muscle actin (SMA) co-react with endothelia of the artery and vein [11, 12].

CD34, a transmembrane phosphoglycoprotein, is a major marker of stromal (mesenchymal) and vascular progenitor cells in both adults and fetuses [13-15]. Thus, previous immunohistochemical studies focused CD34-positive cells at hematopoietic areas along the embryonic abdominal aorta [16-18]. Limited exceptions are found in Katori et al. [19] and Abe et al. [20] who reported abundant CD34-positive cells in developing connective tissues at various sites and organs in human fetuses at and around 15 weeks of gestational age. In adults, details of CD34-positive cell distribution were described in the brain [21], retina [22] and lung [23]. These research groups focused on vascular progenitors, and appeared to overlook mesenchymal or stromal progenitors. Recently, in adult human lymph nodes, Jin et al. [24] found the adult nodal capsule and some stromal cells expressing CD34. Therefore, in the capsule, CD34-positive cells exist with elastic fibers and smooth muscles.

Consequently, according to immunohistochemistry of elastin, SMA and CD34, we aimed to characterize the nasal subepithelial tissue with special reference to the site-dependent difference.

Materials and Methods

This study was performed in accordance with the provisions of the Declaration of Helsinki 1995 (as revised in 2013). We dissected 20 cadavers (8 males and 12 females) those had been donated to Tokyo Dental College Anatomical Gift for research and education of human anatomy, and Tokyo Dental College committee approved their use for research (No. 922-2). The cause of death was heart failure or ischemic brain disease and ages at the death ranged from 70–94 years old (mean, 84.2). None of the cadavers carried tumors in and around the upper respiratory tract according to macroscopic observations after bisection of the head. We found no evidence suggesting that the elderly had needed nasal tube feeding when they were alive (e.g., a deformity or injury of the nasal vestibulum). The cadavers had been fixed by arterial injection of 10% neutral formalin solution in aqua after 12–24 hours after they death and stored in 50% ethanol for more than three months.

A 50–60 mm long, belt-like specimen, that contained parts of the nasal bone and alar cartilages, was obtained from the lateral wall (the ala) of the nasal cavity: it included the alar external skin as well as the vestibular skin, mucocutaneous junction (MCJ) and respiratory ciliated epithelium. We did not choose the septal epithelium because the MCJ area is short and small according to our preliminary observations [25]. To correspond the epithelial type to its subepithelial architecture, we paid a special attention to discriminate the MCJ area from the nasal mucosa. In addition, an entire layer of the palate at and near the transverse palatine suture as well as an antero-inferior part of the inferior concha was also obtained from a half of the 20 cadavers. The former specimens contained both nasal and oral epithelia.

The removed specimens were fixed in 10% w/w neutral formalin solution for 7 days and decalcified by incubating them at 4°C in 0.5 mol/L EDTA solution (pH 7.5; Decalcifying solution B; legitimate manufacturer) for 14 days. After a routine procedure for paraffin-embedded histology, we prepared 10–15 serial sections from a belt-like tissue. Two sections were stained with hematoxylin and eosin (H&E) stain and elastica-Masson stain, while the other was used for immunohistochemistry (see below). The elastica-Masson stain is a variation of Masson-Goldner staining [26, 27] and it colors elastic fibers clear black in paraffin section from formalin-treated donated cadavers. The residual sections had been used for another study of immunoreactive cells [4].

The primary antibodies, together with their dilutions and antigen retrieval procedures, are listed in Table 1. Briefly, we used mouse monoclonal antibodies for α elastin, CD34, α SMA and S100 protein. After incubation with primary antibodies, the sections were incubated for 30 minutes with horseradish peroxidase-conjugated secondary antibodies (Histofine Simple Stain Max-PO; Nichirei) diluted 1:1,000. Immunoreactive proteins were detected by incubation with diaminobenzidine for 3–5 minutes (Histofine Simple Stain DAB; Nichire). Each sample was counterstained with hematoxylin, and a negative control without the primary antibody was used for all specimens. Treated sections were counterstained with hematoxylin, dehydrated in ethanol, and cleared in xylene. All histological images were captured using a Nikon Eclipse 80i microscope (Nikon).

Table 1.

Primary monoclonal antibodies, their dilution, and specific treatment

Legand	Ig types	Sources	Final dilution	Antigen retrival
α elastin	Mouse	Abcam ab9519 (Cambridge, UK)	1:20	Trypsin
CD34	Mouse	Nichirei413361 (Tokyo, Japan)	1:200
α SMA	Mouse	Dako M0851 (Glostrup, Denmark)	1:800	Trypsin
S100 protein	Mouse	Dako Z0311 (Glostrup, Denmark)	1:100

SMA, smooth muscle actin.

Results

Lateral wall of the nasal cavity

Two specimens are shown first for connection between the topographical anatomy and immunohistochemical observations (Figs. 1–4). For each of the specimens, the topographical anatomy of the vestibular skin, the MCJ and the respiratory ciliated mucosa is shown first at a low magnification (Figs. 1, 3) and, next, we present the higher magnification view (Figs. 2, 4). In the alar external skin of the nose (Fig. 1B) as well as the vestibular skin (Figs. 1B, 3C), the dermis was filled with elastin-positive thick fibers and the subcutaneous tissue also contained abundant positive thin fibers. Elastin-positive fibers displayed a distribution similar to CD34-positive cells (Fig. 3A vs. Fig. 3B).

Fig. 1.

Difference in architecture among the external skin, vestibular skin, mucocutaneous junction (MCJ), and ciliated mucosa in the lateral wall of the nasal cavity obtained from an 80-year-old male (M). H&E staining (A), immunohistochemistry of elastin (B), CD34 (C), and smooth muscle actin (SMA) (D). The right-hand side of each panel corresponds to the superficial or vestibular side of the nasal cavity. Star in each panel indicates a corresponding site between sections. Asterisks in panel (A) indicate an artifact space during histological procedure. In the external skin of the nose (B), the dermis is filled with elastin-positive fibers and the subcutaneous tissue also contains abundant positive fibers. CD34-positive cells tended to be rich beneath the mucosa (C). Immunohistochemistry of SMA reveals the mucosal venous plexus (D). Squares in panels (B–D) are shown in Fig. 2. All panels were prepared at the same magnification. (A) Scale bar=1 mm.

Fig. 2.

Subepithelial fibrous structure in specimens from an 80-year-old male (M). The specimen same as in Fig. 1. The left-hand side column shows the mucocutaneous junction area, while the right-hand side column displays the ciliated mucosa. Elastica Masson staining (EM) (A, F) and immunohistochemistry of elastin (B, G), CD34 (C, H), smooth muscle actin (SMA) (D, I), and S100 protein (E, J). In the mucocutaneous junction, a thick layer beneath the epithelia (semicircle in A, B) contains thin fibers (possibly oxytalan fibers), while a deep layer (star in A, B) contains relatively thick fibers (possibly elastic fibers). Arrows in panels (A) and (F) indicate the basal lamina of epithelial cells. Submucosal fibrous structures (arrowheads) appear to express both elastin and CD34 (G, H). Arterial endothelia strongly express CD34 (arrows in C). Abundant arterioles are seen beneath the ciliated epithelium (arrows in H). SMA-positive fibers are few in the mucosa (I) and absent in the mucocutaneous junction (D). Endothelia of arteries and veins show reactivity of the SMA, but submucosal arterioles appear negative (I vs. H). Thin nerves are rich in glands in the submucosa (J). All panels were prepared at the same magnification. (A) Scale bar=0.1 mm.

Fig. 3.

Elastin- and CD34-positive fibrous structures in the lateral wall of the nasal cavity of an 83-year-old male (M). Immunohistochemistry of elastin (A, C), CD34 (B), and smooth muscle actin (SMA) (D). The lower side of each panel corresponds to the superficial or inferior side of the nasal cavity. Star in (A, B, D) indicates a corresponding site between sections. Elastin-positive fibers display a distribution similar to CD34-positive cells (B). Panel (C) exhibits a dense distribution of elastin-positive fibers. SMA-positive vessels are abundant in the mucocutaneous junction (MCJ) and mucosa (D). A square in panel (D) is shown in Fig. 4. Asterisks in panel (A) indicate an artifact space during histological procedure. Panels (A), (B), and (D) were prepared at the same magnification. (A, C) Scale bars=1 mm.

Fig. 4.

Submucosal fibrous structure in specimens from an 83-year-old male (M). The specimen same as in Fig. 3, but most panels display a site outside of Fig. 3. The left-hand side column shows a superficial mucosa near the mucocutaneous junction area, while the right-hand side column displays the deep ciliated mucosa. Immunohistochemistry of elastin (A, E), CD34 (B, F), smooth muscle actin (SMA) (C, G), and S100 protein (D, H). Elastin-positive fibers are few beneath the deep mucosa (E), relative to the superficial mucosa (A). The basal lamina is thick beneath the deep mucosa (arrows in E), but it is unclear in the superficial mucosa (A). Elastin-positive fibrous structures appear negative for CD34 (A vs. B). CD34-positive arterioles are seen beneath the epithelium (arrows in B, F), but they appeared negative for SMA (C, G). CD34-positive submucosal vessels are densely distributed in the deep mucosa (F), relative to the superficial mucosa (B). Few myoepithelial cells are seen (arrowheads in G). Thin nerves are rich in glands in the submucosa (H). All panels were prepared at the same magnification. (A) Scale bar=0.1 mm.

The MCJ was identified as a long area along the alar cartilages in which hair follicles were absent and glands were few (Figs. 1, 3). The MCJ likely contain abundant thick arteries and veins (Fig. 3C, D) that expressed SMA reactivity (Fig. 3D). In the MCJ (Fig. 2A, B), a thick layer beneath the squamous epithelia contained elastin-positive thin fibers (possibly oxytalan fibers), while a deep layer (star in Fig. 2A, B) contained relatively thick fibers (possibly elastic fibers). The elastic and oxytalan fiber-rich layer, possibly corresponding to the dermis of the vestibular skin (Figs. 1B, 3C), ranged from 0.3–0.8 mm in thickness. The basal lamina was thick beneath the squamous epithelium (Fig. 2A, F) and elastin-positive fibers appeared not to insert into the lamina (Fig. 2B). The basal margin of the epithelium was wavy (Fig. 2A): this morphology resembled the skin epidermis that makes a wavy margin corresponding to the dermal papillae. SMA-positive fibers were absent beneath the squamous epithelium (Fig. 2D). The elastin-positive fibrous layer was thick beneath the MCJ and it contained few CD34-positive cells (Fig. 2B vs. Fig. 2C).

The ciliated (pseudostratified) columnar epithelium of the mucosa was characterized by few elastin-positive fibers beneath the epithelium (Figs. 2G, 4E). The elastin-positive fiber density appeared to decrease from the superficial side to the deep side (Fig. 4A vs. Fig. 4E). The basal lamina was thick beneath the ciliated epithelium (>20 micron; Figs. 4E, 5A, B). CD34-positive cells tended to diffusely distribute in the submucosal tissue, but they were absent in glands (Figs. 2H, 5C). Some of submucosal fibrous structure appeared to express both elastin and CD34 (Fig. 2G, H), but the other fibrous structures appeared negative for CD34 (Fig. 4A vs. Fig. 4B). CD34 immunostaining revealed abundant arterioles arranged in line along the thick basal lamina (Figs. 2H, 4B, F), but they appeared negative for SMA (Fig. 4C, G). SMA immunohistochemistry clearly demonstrated the mucosal venous plexus (Fig. 1D) and the other thin veins (Figs. 2I, 4G). Few SMA-positive fibers were seen beneath the mucosa (Figs. 2I, 4G, 5D): they tended to accumulate between acini of glands as well as along the deep side of glands (Fig. 5D). Endothelia of arteries and veins show SMA reactivity, but submucosal arterioles appear negative for SMA (Fig. 2I vs. Fig. 2H). SMA-positive myoepithelial cells were rather rare in the submucosal glands in the lateral wall (Fig. 4G). Most of the submucosal glands were mucinous and some of them were serous. Thin nerves were rich in these glands (Figs. 2J, 4H).

Fig. 5.

A difference in fibrous structures between the lateral and inferior walls of the nasal cavity from a 92-year-old female (F). The left-hand side column shows a ciliated mucosa in the lateral wall of the nasal cavity, while the right-hand side column displays a mucosa in the inferior wall (a level above the transverse palatine suture). Elastica Masson staining (A, E) and immunohistochemistry of elastin (B, F), CD34 (C, G), and smooth muscle actin (SMA) (D, H). In this specimen, a thick basal lamina is evident in the lateral nasal wall (arrows in A–C). In contrast, a thick, stratified epithelium without cilia above the palate has a thin basal lamina lined by elastic fibers (arrows in E, F) and accompanies a thick layer of elastin-positive fibers (semicircle in F). CD34-positive cells diffusely distribute in the lateral wall (C), while they tended to be dense beneath the stratified epithelium (semicircle in G). In the lateral wall, SMA-positive fibers accumulate beneath the epithelium (stars in D) as well as glands (semicircle in D). Note abundant myoepithelial cells (arrowheads) in the submucosa above the palate, relative to the lateral wall (H vs. D). All panels were prepared at the same magnification. (A) Scale bar=0.1 mm.

Inferior wall of the nasal cavity

The epithelium in the inferior nasal wall above the palate lost both cilia and a thick basal lamina. Thus, it appeared to be a stratified squamous epithelium. However, since the composite cells resembled those of the columnar epithelium, the authors regarded it as a variation of the pseudostratified columnar epithelium (Fig. 5F–H). The basal margin of the epithelium was straight: this was different from a wavy basal margin at the MCJ in the lateral wall of the nasal cavity. The specific columnar epithelium had a thin basal lamina lined by elastic fiber bundles (Fig. 5E, F). Resembling the MCJ, elastin-positive fibers provided a thick layer (0.1 mm in thickness) extending along the epithelium (Fig. 5F) and, in contrast to no or few subepithelial veins in the MCJ (Fig. 2D), it contained abundant SMA-positive veins (Fig. 5H). CD34-positive cells also tended to be dense beneath the epithelium (Fig. 5G). Thus, CD34-positive cells existed with elastin-positive fibers. In contrast to mucinous glands in the lateral wall of the nasal cavity, almost all myoepithelial cells expressed SMA reactivity (Fig. 5H vs. Fig. 5D).

Inferior concha of the nasal cavity

Soft tissues around the bony and cartilaginous inferior concha were filled with glands (most of them, mucinous), not with a venous plexus (Fig. 6A). The ciliated pseudostratified columnar epithelium of the concha was much thicker than the lateral wall mucosa because of the thick stratification (strictly, pseudo-stratification; Fig. 6B–E). The thick basal lamina (10–20 micron in thickness) was lined by a bundle of thin fibrous structures that expressed reactivities of elastin (Fig. 6C) and CD34 (Fig. 6D). CD34-positive arterioles were also arranged along the basal lamina. Few SMA-positive fibers were seen beneath the mucosa but accumulated between acini of glands (Fig. 6E). In glands, some myoepithelial cells were positive for SMA (Fig. 6E).

Fig. 6.

Submucosal tissue of the inferior concha from an 80-year-old male (M). The specimen same as in Figs. 1, 2. Panel (A) displays a topographical anatomy of an inferomedial part of the inferior concha. A square by dotted line indicates the site shown in panels (B–E). H&E staining (A), elastica Masson staining (EM) (B), immunohistochemistry of elastin (C), CD34 (D), and smooth muscle actin (SMA) (E). A thick basal lamina (arrows in B, C) is lined by elastin-positive thin fibers (C) and CD34-positive cells (arrows in D). The fibrous lining resembles that in the lateral wall of the nasal cavity (Fig. 5B, C). Few myoepithelial cells (arrowheads in E) are seen in the submucosal gland. Panels (B–E) were prepared at the same magnification. (A) Scale bar=1 mm, (B) scale bar=0.1 mm.

Overall, much or less elastin-positive fibers existed with CD34-positive cells beneath the thick stratified epithelium (the MCJ and the inferior or palatal wall of the nasal cavity; Fig. 7A). In contrast, the ciliated columnar epithelium had a thick basal lamina that was lined by a fibrous layer containing both elastin-positive fibers and CD34-positive cells (the lateral wall of the nasal cavity and the inferior concha; Fig. 7B). In the latter sites, abundant CD34-positive cells were also seen in the deep layer.

Fig. 7.

Co-existence of elastin-positive and CD34-positive fibrous structures: the site- and epithelium-dependent difference. Beneath the stratified epithelium (the mucocutaneous junction and the inferior or palatal wall of the nasal cavity), elastin-positive thick fibers existed with CD34-positive cells (A). In contrast, beneath and along a thick basal lamina of the ciliated columnar epithelium (the inferior concha and the lateral wall of the nasal cavity), elastin-positive fibers were thin and existed with CD34-positive cells. The latter cells were also seen in the submucosa with no or few elastin-positive fibers (B).

Discussion

The epithelial barrier has been studied in the cellular and molecular level [28, 29]. In this context, Chason et al. [30] reported age-associated changes in the respiratory epithelial response to influenza infection. However, there may be relatively little information on the barrier in the level of lower magnification histology.

According to the present observations, the nasal mucosal epithelium was characterized by 1) a thick basal lamina, 2) few elastin-positive fibers beneath the epithelium, that was quite different from the nearby MCJ with a thick layer of elastic and oxytalan fibers corresponding to the skin dermis, 3) CD34-positive cells distributing diffusely in the submucosal tissue, 4) few SMA-positive fibers beneath the epithelium, and 5) CD34-positive SMA-negative arterioles arranged along the basal lamina. However, beyond our understanding, the mucosal epithelium was quite different between sites around the nasal cavity: the ciliated columnar epithelium in the lateral wall of the nasal cavity; thick pseudostratified columnar epithelium in the inferior concha and; the thick pseudostratified columnar-like epithelium without cilia in the inferior or palatal wall of the nasal cavity. The last, thick columnar-like epithelium suggested a mild pathology due to the unusual mechanical and/or chemical stimulation to the epithelium. A candidate cause was refluxed saliva with disturbance of feeding and swallowing in the elderly. A nasal tube feeding also likely produces such a pathology of nasal epithelia.

We found SMA-positive mucosal venous plexus in the lateral wall of the nasal cavity, but the submucosa was filled with glands in the inferior concha. Vascular beds might also be replaced by glands in the other sites of the nasal submucosa. The absence of vascular beds or the dominancy in glands might be a result of secondary change with aging. However, there seems to be no information on topographical anatomy of the so-called nasal swell body [1, 3]. The vascular bed may not be localized in the lateral wall but in the septal mucosa that was not examined in the present study.

A presence and role of CD34-positive structures in adults was still obscure. Lee et al. [31] reported that connective tissue progenitor cells are negative for CD34 in adults. Multiple research groups found CD34-positive stromal cells in limited cases of adult skin fibroma [32-34]. In contrast, using flow cytometry, Young et al. [13] found CD34-positive cells in both fetal and adult skeletal muscle. According to Katori et al. [19], CD34-positive fibrous tissues cover pharyngeal and laryngeal striated muscles but not extraocular and tongue muscles in human fetuses. Notably, almost all visceral interstitial tissues were negative for CD34 in fetuses. Therefore, CD34-positive fibrous tissues show a site-dependent distribution. Invernici et al. [35] suggested cross-reactions of CD34 antibody with E-selectin, inter-cellular adhesion molecule 1 and vascular cell adhesion protein 1. Jin et al. [24] also found, in fatty degeneration of elderly nodes, CD34- elastin-double positive stromal cells. In mouse lymph nodes, elastic fibers are likely to support mechanically sinus macrophage residency [36, 37]. Overall, we speculated that subepithelial immunoreactive cells favor or use elastin- and CD34-positive structures for guidance of migration in early stage of inflammation. Nevertheless, these fibers are most likely to be destroyed with neutrophil elastase in the later stage of inflammation.

Although the present immunohistochemistry had a limitation to reveal a presence of elastin- CD34-double positive structures, a co-existence of these cells and fibers likely showed epithelium-dependent differences. A thick stratified epithelium (typically, a stratified squamous epithelium) was associated with elastin-positive thick fibers as well as abundant CD34-positive cells (Fig. 7). In contrast, a ciliated columnar epithelium accompanied elastin-positive thin fibers, but the latter seemed to contain no oxytalan fibers. There might be no or few previous reports about the epithelium-dependent difference in subepithelial fibrous architecture. A further study using cytokeratin immunohistochemistry seemed necessary to define the epithelium.

Katori et al. [38] extensively compare immunoreactivity of SMA, S100 and glial fibrillary acidic protein between major and minor salivary glands and they found site-dependent differences in immunoreactivity of myoepithelial cells. Likewise, the present study strongly suggested heterogeneity of myoepithelial cells in nasal mucosal glands: the myoepithelial cell of mucous glands in the lateral wall of the nasal cavity usually negative for SMA, while they always expressed SMA reactivity in mucous glands in the inferior wall of the nasal cavity. However, we had no evidence suggesting a correlation between a component of nasal fibrous structures and the difference in myoepithelial cells.