Abstract
This paper describes some important differences in normal histology of the upper respiratory tract of laboratory animals. It also provides examples of lesions observed or reported in the upper respiratory tract of laboratory animals, predominantly rodents, exposed via inhalation. The anatomy and physiology of upper respiratory tract tissues play a major role in the response to an insult, given that different epithelial types vary in susceptibility to injury and toxicant exposure concentrations throughout the airway vary due to airflow dynamics. Although dogs and nonhuman primates are utilized for inhalation toxicology studies, less information is available regarding sites of upper respiratory injury and types of responses in these species. Awareness of interspecies differences in normal histology and zones of transition from squamous to respiratory to olfactory epithelium in different areas of the upper respiratory tract is critical to detection and description of lesions. Repeated inhalation of chemicals, drugs, or environmental contaminants induces a wide range of responses, depending on the physical properties of the toxicant and concentration and duration of exposure. Accurate and consistent fixation, trimming, and microtomy of tissue sections using anatomic landmarks are critical steps in providing the pathologist the tools needed to compare the morphology of upper respiratory tract tissues from exposed and control animals and detect and interpret subtle differences.
Introduction
This paper describes some important differences in normal histology of the upper respiratory tract of laboratory animals and describes some examples of lesions observed or reported in the upper respiratory tract of laboratory animals, predominantly rodents, exposed via inhalation for varying periods to toxic chemicals, drugs, or complex mixtures. The majority of this report discusses lesions in laboratory rodents because these are the species most frequently utilized in toxicology studies and the source of most published literature pertinent to this subject.
Rats and mice exposed via inhalation to toxic or irritating drugs, chemicals, or environmental contaminants have a relatively high incidence of lesions in the respiratory tract. The most frequent target tissues include the mucosa of the nasal cavity and larynx. As a result, many regulatory guidelines for inhalation exposure to environmental toxicants and pharmaceuticals are derived from studies in which key conclusions were based on lesions in the upper respiratory tract of rodents. Results of sensory irritation assays in mice (Alarie et al., 1980) have also been used in setting occupational exposure guidelines in the United States. Much less information is published on respiratory tract lesions induced in dogs and primates, yet regulatory agencies increasingly require toxicity studies in nonrodent species. Thus, there is a need for more information on the normal upper respiratory tract histology of dogs and primates, the location and morphology of lesions induced in this area by inhalation exposure, and interspecies differences in response to inhaled materials.
There are distinct differences in the normal anatomy and histology of the upper respiratory tract of different laboratory species and humans. Nasal turbinates are complex in rodents and dogs but much simpler in nonhuman primates and humans. Rodents are obligate nose breathers whereas other species are not, further complicating application of the results of rodent studies to other species, especially man. Laryngeal anatomy and histology also vary among species, including the location of the transition from relatively tough squamous epithelium to more sensitive respiratory epithelium. These differences make interpretation of the results challenging and become significant when one attempts to extrapolate histopathology data from rodents to other laboratory species or humans.
Materials and Methods
The information presented in this paper is from publications by the authors or others as cited, or from unpublished work in the authors’ laboratory. A number of excellent reviews have been published on the pathology and pathophysiology of the upper respiratory tract of laboratory animals (Bogdanffy and Keller, 1999; Miller, 1999; Feron et al., 2001; Harkema, 2006; Morris, 2006), and the reader is referred to these and other sources in the literature cited here for in-depth information on the subjects reviewed here.
Most published descriptions of morphologic effects of inhaled toxicants on nasal mucosa of rodents utilize the histologic methods published by John Young (1981). Transverse sections are cut from the decalcified nose at several sites using landmarks on the roof of the mouth or teeth. It is important to make these cuts consistently in the same locations to enable comparable sections through the areas of interest. Close examination of the trimmed nasal sections provides the opportunity for tissue trimmers to observe visible nasal tumors or other gross nasal cavity lesions. Mapping of nasal mucosal lesions provides a precise description of location of lesions and epithelial types affected. Publications providing key information on mapping and more precise localization of nasal lesions are available for rodents (Mery et al., 1994; Morgan, 1994; Hardisty et al., 1999) and nonhuman primates (Kepler et al., 1995).
Consistent access to toxicologically important areas of the mucosal epithelium is equally important for accurate microscopic examination of the larynx. One key is access to interpretable sections through the area of transition from squamous to respiratory epithelium at the base of the epiglottis. The submucosal glands and cartilages at the level of the base of the epiglottis, ventral pouch, and vocal processes of the arytenoids in rodent larynges can be visualized in unstained sections at microtomy by adjusting the light diffraction in the microscope (Renne and Gideon, 2006). Microscopic examination of unstained sections through these key areas at microtomy enables the histotechnologist to accurately determine the location of the section and provide the pathologist with consistent sections through key areas of the larynx. Publications providing information on the most frequent sites and typical morphology of laryngeal lesions induced in rodents from inhalation exposure are available in the literature (Lewis, 1991; Miller and Renne, 1996; Renne and Gideon, 2006).
Results and Discussion
As stated previously, there is a standard set of transverse sections of nasal cavity prepared and examined from rodents in inhalation studies. However, the most rostral of these is taken caudally to the root of the upper incisor teeth and thus the nasal vestibule is not routinely examined using this standard method. This may result in important exposure-induced lesions in the atrioturbinates and adjacent structures in the vestibule being missed. The nasal vestibule is lined with stratified squamous epithelium, thus is more resistant to injury than transitional or respiratory epithelium. Nonetheless, toxicologically important lesions have been described in the nasal vestibule as a result of inhalation of highly volatile, water-soluble chemicals such as glutaraldehyde (Gross et al., 1994), dimethylamine (Buckley et al., 1985), ammonia (Bolon et al., 1991), hydrogen fluoride (Rosenholtz et al., 1963; Morris and Smith, 1982), and hydrogen chloride (Jiang et al., 1983). Exposure to volatile, water-soluble chemicals such as these can induce extensive necrosis, inflammation, hyperplasia, and hyperkeratosis of the squamous epithelium lining the vestibule. Repeated inhalation of glutaraldehyde led to accumulation of sloughed keratin and inflammatory debris, resulting in occlusion of the nasal lumen (Gross et al., 1994; Van Birgelen et al., 2000). Sections through the nasal vestibule should be routinely prepared and examined in toxicology studies involving inhalation of highly volatile substances.
The nasal lumen of the most rostral transverse section, standard section 1 (Young, 1981), located immediately caudal to the upper incisor teeth, is lined by stratified squamous epithelium in the ventral meatus, transitional epithelium on the lateral aspect of the nasal and maxillary turbinates and lateral walls, and respiratory epithelium on the median septum, dorsal meatus, and medial aspect of the turbinates (Young, 1981; Monteiro-Riviere and Popp, 1984; Hebel and Stromberg, 1986; Uraih and Maronpot, 1990). Responses to injury of transitional and respiratory mucosa include vacuolization and degeneration, loss of cilia, inflammation, necrosis, ulceration, necrosis or hyperostosis of turbinates, and metaplasia, hyperplasia, hyperkeratosis, and neoplasia of the mucosal epithelium and adjacent glands.
One of the frequent sites of early morphologic evidence of injury from inhaled xenobiotics in this section is the transitional epithelium covering the distal third of the nasoturbinates and maxilloturbinates, the lateral walls, and the adjacent median septum. Degeneration of the bilayered cuboidal epithelium characterized by nuclear pyknosis and cytoplasmic vacuolation, accompanied by suppurative inflammation in the adjacent lamina propria may lead to necrosis and sloughing of surface epithelium followed by regeneration. Loss of cilia in adjacent affected respiratory epithelium provides additional early evidence of exposure-related injury. Repeated exposures lead to squamous metaplasia, hyperplasia, and hyperkeratosis of surface epithelium and/or mucous cell metaplasia of transitional epithelium and mucous cell (goblet cell) hyperplasia of respiratory epithelium at these sites. Factors leading to the relatively high incidence of lesions at these most sensitive sites in nasal section 1 in rodents include the high volume of air flow through the ventral aspect of the nasal cavity over the nasal and maxillary turbinates (Morgan et al., 1991; Morgan, 1994) and the increased susceptibility of transitional epithelium to oxidants such as ozone (Harkema, 2006). Transitional and respiratory epithelium both contain Cytochrome P-450 and other metabolic enzymes within abundant smooth endoplasmic reticulum (ER), suggesting a role for these epithelial types in metabolism of inhaled xenobiotics (Matulionis and Parks, 1973; Yamamoto and Masuda, 1982; Bogdanffy et al., 1987; Harkema et al., 1987a). In most reported subchronic (13-week) studies and in the authors’ experience, removal of the causative agent results in regression of hyperplastic and metaplastic changes in nasal respiratory epithelium following a recovery period of several weeks.
Gross anatomic features of the anterior nasal cavity of non-rodent laboratory species are quite different from rodents. The reader is referred to several excellent sources of detailed information on these interspecies differences (Negus, 1958; Schreider and Raabe, 1981; Harkema et al., 1987a, 2006; Harkema, 1991; Wako et al., 1999; Morris, 2006). Despite these differences, the arrangement of squamous, transitional, and respiratory epithelium of the anterior nasal cavity of nonhuman primates and dogs and is similar to that of rodents (Harkema et al., 1987a; Harkema, 1992; Wako et al., 1999) stratified squamous epithelium lines the vestibule and ventral meatus, transitional epithelium lines the lateral walls and lateral turbinates, respiratory epithelium lines the median septum and medial turbinate surfaces. However, these epithelia are composed of many more layers of cells in dogs and monkeys (Harkema, 1991; Wako et al., 1999) compared to rodents. Figures 1 and 2 illustrate the nasal turbinate histology of a rat at nasal level 1 and the nasal turbinate of a dog (Beagle) at the level of the canine teeth. At the level of upper canine teeth, the canine turbinates are lined with a mixture of squamous and respiratory epithelium of varying thickness. At the level of the premolar teeth, the turbinates of dogs are lined entirely by respiratory epithelium.
Perhaps in part because of this increased thickness, the mucosa of the anterior nasal cavity of dogs and primates is apparently more resistant than rodents to injury from inhaled toxicants. Numerous reports in the literature describe nasal lesions induced in rodents by inhaled chemicals with no corresponding nasal effects in dogs or monkeys exposed to similar concentrations (Vernot et al., 1985; Dudley et al., 1989; Hiles et al., 1994; Pino et al., 1999). Other factors such as interspecies differences in breathing pattern (oronasal versus obligatory nasal breathing), mucociliary apparatus, and distribution of mucosubstances are also important determinants of nasal toxicity in nonrodents compared to rodents. However, there are important examples in the literature of similar nasal effects induced in rodent and nonhuman primates by inhaled chemicals. Research on inhaled ozone (Harkema et al., 1987b, 1994, 1999, 2006; Cho et al., 1999) and formaldehyde (Swenberg et al., 1980; Chang et al., 1983; Kerns et al., 1983; Monticello et al., 1989; Morgan et al., 1991, 1997), 2 important environmental toxicants demonstrated similar nasal lesions induced in rodents and nonhuman primates. Research on these 2 chemicals has provided an important bridge to improve the usefulness of rodent toxicology data for assessing human risk from inhaled chemicals or environmental toxicants.
The mucosal surface of rodent nasal sections taken at the level of the incisive papilla in standard section 2 (Young, 1981) is lined by squamous epithelium in the ventral meatus and respiratory epithelium elsewhere except for the dorsal meatus, which is lined by olfactory epithelium. Effects on respiratory epithelium are similar to those described above for section 1, but often are less severe and extensive than in section 1. The junction of respiratory and olfactory epithelium in rats and mice is a frequent site of intracytoplasmic accumulations of proteinaceous material described in the literature as eosinophilic globules (Monticello et al., 1990). Also frequently present in adjacent glands, these structures are increased in association with toxic effects on the nasal mucosa but are also observed in untreated control rats and mice, especially in aged animals. The olfactory epithelium lining the dorsal meatus at this level in rodents is the most anterior location for this epithelial type, and is especially susceptible to injury from many inhaled toxicants. A comparable area of canine nasal cavity is the level of the first molar teeth. The epithelium lining the canine nasal cavity at this level is a mixture of olfactory and respiratory epithelium; the ethmoturbinates are lined with a mixture of olfactory and respiratory epithelium and the median septum is lined with olfactory epithelium dorsally and respiratory epithelium ventrally. Sections through the second premolar and first molar tooth of the marmoset (Wako et al., 1999) and the bonnet monkey (Harkema, 1991) also contain a mixture of respiratory epithelium ventrally and olfactory epithelium dorsally.
The epithelium lining the rodent nasal cavity section 3 (Young, 1981) at the level of the second palatal ridge is predominantly olfactory epithelium covering the ethmoturbinates, with respiratory epithelium lining the ventral portion of the median septum, maxillary sinuses, and nasopharyngeal duct. There is a transition between the 2 epithelial types in the ethmoturbinates, the location of the transition depending on the level of the cut. Mery et al. (1994) proposed a numbering system for rodent ethmoturbinates in which they are numbered sequentially from dorsal to ventral, and those with the scrolls of ethmoturbinates 2, 3, 4, and 6 are labeled as dorsal or ventral scrolls. Although the ethmoturbinates comprise about 50% of the surface area of the rodent nose, only 10–15% of inspired air passes over them (Morris, 2006). In the authors’ experience, the most frequent site of injury in nasal section 3 of rodents is the olfactory epithelium lining the dorsal medial meatus and adjacent ethmoturbinates 1 and 3.
Olfactory mucosa contains 3 types of cells: neuroepithelium, sustentacular cells, and basal cells. Sustentacular nuclei are aligned perpendicular to the surface. The cytoplasm of sustentacular cells surrounds and supports neuroepithelial cells, and contains abundant smooth ER with important metabolizing enzymes. Microvilli extend from the distal portion of sustentacular cells into the lumen. Neuorepithelial cells are basal, with axons extending into lamina propria and terminal microcilia extending into the nasal lumen. Bowman’s glands produce abundant enzymes important in metabolism of xenobiotics, and secrete the mucus that protects the surface epithelium in this area. Considerable research has been done on the metabolic capacity of the enzyme-rich olfactory epithelium (Bogdanffy and Keller, 1999). Cytochrome p450s are present in higher levels in olfactory than in respiratory epithelium, and metabolism by these enzymes is a key factor in the olfactory toxicity of methyl bromide, styrene, chloroform, and bromobenzene (Morris, 2006). Carboxylesterases and aldehyde dehydrogenases present in rodent olfactory epithelium hydrolyze esters to acidic metabolites highly toxic to olfactory mucosa (Bogdanffy et al., 1987). The best example of this is acetaldehyde.
The response of olfactory epithelium to repeated injury typically starts as degeneration, vacuolation, and/or necrosis, followed by regeneration, basal cell hyperplasia, and then respiratory and/or squamous metaplasia, or atrophy. Atrophy of olfactory axons is visible in the adjacent lamina propria, and changes in Bowman’s glands may include hyperplasia, atrophy, or the presence of intracytoplasmic eosinophilic material (eosinophilic globules; Monticello et al., 1990). Progression to neoplasia is a rare sequel to this process. The paper by Hardisty et al. (1999) provides a consensus description by a group of experienced nasal pathologists on the pathogenesis of olfactory lesions induced by several chemicals inhaled for durations ranging from 2 weeks to 2 years. Early lesions were degeneration and loss of both sustentacular and neuroepithelial cells with decreased thickness of olfactory epithelium, followed by regeneration through hyperplasia of the basal cell layer, atrophy of the overlying neural/sustentacular layer, and loss of axon bundles and Bowman’s glands in the lamina propria. Lesions were most frequent and severe in the dorsal medial meatus of section 2, but also involved more caudal sections in some studies. The medial and dorsal surfaces of ethmoturbinate 3 were especially prone to lesions, probably due to its location adjacent to the median septum and thus near the airstream. In studies lasting more than 13 weeks, acute degenerative and necrotizing olfactory epithelial lesions were less extensive but areas of respiratory metaplasia of olfactory epithelium and proliferation of Bowman’s glands were observed.
Lymphoid tissue is typically present in the lamina propria adjacent to the nasopharyngeal duct of rodents and other laboratory animal species. NALT in this location has similar characteristics and function to the bronchus-associated lymphoid tissue (BALT) in the lungs, and has been the subject of considerable research on its immunologic properties (Harkema, 2006). Recently NALT has been recommended for inclusion as a tissue to be examined routinely during histopathologic examination of tissues from inhalation toxicity studies.
The larynx, although not as complicated as the nasal cavity, is a bilaterally symmetrical organ lined by several epithelial types and containing various protuberances, pouches, folds, and cartilages, with considerable anatomic and histologic variation among laboratory animal species (Table 1). The distribution of the types of epithelium lining the laryngeal lumen is similar to the nose, in that the mucosal epithelium changes from stratified squamous rostrally to pseudostratified ciliated columnar (respiratory) epithelium caudally. These areas of transition from a relatively durable stratified squamous epithelium to a much more fragile respiratory epithelium are the most sensitive sites for cellular changes in rodents inhaling xenobiotics (Gopinath et al., 1987; Lewis, 1991). As in the nose, the predilection of these areas in the larynx for exposure-induced lesions in rodents is probably also related to airflow dynamics as well as regional epithelial sensitivity (Chevalier and Dontenwill, 1972; Lewis, 1981, 1991; Gopinath et al., 1987). The normal laryngeal histology and the sites most susceptible to injury from inhalation of toxicants in rats, mice, and hamsters have been described in the literature (Lewis, 1991; Renne and Miller, 1996).
The normal morphology of the rostral larynx of Syrian hamsters varies slightly from mice and rats, with potentially important differences in effects from exposure to inhaled toxicants (Lewis, 1981; Renne et al., 1992a). Exposure-induced laryngeal lesions are frequently reported in inhalation studies in rodents; thus, there is considerable information available on laryngeal lesions in these species (Lewis, 1991; Miller and Renne, 1996; Renne and Gideon, 2006).
The laryngeal epithelium responds similarly to the epithelium lining the nose. The response depends on the physical and chemical properties and concentration of the inhaled material, and duration/frequency of exposure. As in the nasal mucosa, the rostral areas of the larynx lined by stratified squamous epithelium are more exposed to trauma by direct contact with inhaled substances. Although the thickness of this epithelium and its inherent resistance to damage provide more protection than other epithelial types, the stratified squamous epithelium lining the rostral larynx of rodents is still more susceptible than in other areas of the oropharyngeal cavity because it lacks keratin or is poorly keratinized (Lewis, 1991; Renne and Miller, 1996). Thus, exposure to irritants can induce edema, inflammation, and if prolonged and severe enough, ulceration, necrosis, and epithelial sloughing. Death due to occlusion of the laryngeal lumen from edema and inflammation has been described (Miller and Renne, 1996).
In rodents, the epithelium lining the base of the epiglottis is the area of transition from stratified squamous to respiratory epithelium and is the area most frequently affected by inhaled materials. The normal epithelium at the base of the rat epiglottis is a mixture of ciliated and nonciliated columnar to round cells 2–3 cells thick (transitional epithelium), with no definite basal cell layer. Depending on the plane of section, a small area in the ventral midline at the rostral and caudal borders of the submucosal glands may be covered by squamous epithelium (Renne et al., 1992b), but these areas do not have the prominent basal cell layer typical of stratified squamous epithelium. The epithelium lining the base of the epiglottis in mice is more homogenous than in rats, consisting of a relatively pure population of tall columnar ciliated cells.
Loss of cilia and slight flattening are subtle morphologic changes in the surface epithelium lining the base of the epiglottis; their detection requires prompt fixation and careful comparison of exposed with control tissues at precisely the same level of section. Acute inflammation with edema may lead to necrosis and ulceration, most often at the base of the epiglottis in rodents. If no further insult occurs, the epithelium may regenerate by migration of epithelial cells from the sides of the ulcerated area. Repeated inhalation of materials sufficiently irritating to induce an inflammatory or degenerative response eventually stimulates squamous metaplasia and hyperplasia of the affected transitional epithelium. Depending on severity and duration of exposure, this metaplastic epithelium also becomes hyperkeratotic.
Severity of squamous metaplasia and hyperplasia depends on concentration of the inhaled toxicant and duration of exposure. In most reported studies and in the authors’ experience, removal of the inciting stimulus for a period of 6 to 13 weeks following subchronic (13 weeks) exposure results in regression of the lesions and return of the affected epithelium to normal morphology. Morphometry using image analysis tools to measure the total area of epithelium lining the base of the epiglottis has proven useful in quantitating hyperplasia and hyperkeratosis in rodents (Renne and Gideon, 2006).
Although usually seen first and most severe at the base of the epiglottis in rodents, squamous metaplasia and hyperplasia may occur throughout the laryngeal mucosa. Areas normally lined by stratified squamous epithelium such as the medial surface of the vocal processes of the arytenoids, may undergo hyperplasia and hyperkeratosis. Areas normally covered with a thin layer of squamous epithelium such as the mucosa adjacent to the ventral pouch may also become hyperplastic and hyperkeratotic, and may be accompanied by inflammatory cell infiltrates in the adjacent submucosa, depending on severity and duration of the stimulus (Renne and Gideon, 2006). If squamous metaplasia extends downward into the ducts of submucosal glands, the ducts may plug with hyperkeratotic debris, causing dilatation and inflammation of the affected glands. The granulomatous inflammatory response to this trapped material may lead to formation of polypoid lesions, which may extend into the lumen and impede airflow (Bucher et al., 1990).
Although squamous metaplasia and hyperplasia of laryngeal epithelium are reported frequently in subchronic and chronic rodent inhalation studies, progression to neoplasia at this site with repeated exposures is apparently very rare. Dontenwill et al. (1973) reported induction of squamous cell carcinomas in the larynges of Syrian hamsters by chronic exposure to cigarette smoke and dimethylbenzanthracene. Homburger (1975) reported induction of laryngeal carcinomas in Syrian hamsters exposed to cigarette smoke for up to 24 months. Feron et al. (1982, 1990) reported the induction of a small number of laryngeal carcinomas in rats and hamsters inhaling acetaldehyde or propylene oxide for 24 months. Conversely, there are numerous published reports of chronic inhalation exposures of rodents to irritant compounds such as tobacco smoke (Dalbey et al., 1980), ozone (Boorman et al., 1994), hexachlorocyclopentadiene (National Toxicology Program [NTP], 1994), and molybdenum trioxide (NTP, 1997) in which laryngeal squamous metaplasia was induced but progression to neoplasia was not present, even after lifetime exposure. Squamous metaplasia of the laryngeal epithelium is described as occurring in at least half of the healthy human population, yet carcinoma of the larynx is very uncommon (Stell et al., 1982). It appears that, although severe squamous metaplasia and hyperplasia of laryngeal epithelium often precede neoplastic lesions, the metaplastic change by itself is simply a response to repeated irritation in which a resistant type of epithelium replaces a susceptible one (Gopinath et al., 1987; Burger et al., 1989).
In dogs and monkeys, the stratified squamous epithelium lining the epiglottal cartilage is much thicker than in rodents and extends further caudally, resulting in a more formidable barrier to inhaled materials impacting the base of the epiglottis (Figures 3 and 4). Diverticula (lateral ventricles and saccules) extend laterally rather than ventrally and are lined by stratified squamous epithelium in dogs and respiratory epithelium in monkeys. In these species, the transition from stratified squamous to respiratory epithelium occurs in the epithelium lining the vocal processes of the arytenoid cartilages, just caudal to the junction of the lateral ventricles with the laryngeal lumen (Renne and Gideon, 2006). The vocal processes of the arytenoids extend closer to the ventral surface of the epiglottis in dogs and monkeys compared to rodents. In dogs and nonhuman primates the epithelium lining the caudal larynx is similar to rodents—slightly flattened ciliated columnar epithelium dorsally, sparsely ciliated tall columnar epithelium ventrally.
Based on these histologic differences and investigations of the responses of rodent larynges to inhaled toxicants, one would expect to find lesions further caudally in the larynges of dogs and monkeys inhaling toxicants. However, there are very few descriptions in the literature of inhalation exposure-induced laryngeal lesions in either of these species. Squamous metaplasia of laryngeal epithelium was reported in rhesus monkeys inhaling formaldehyde for 6 weeks (Monticello et al., 1989). Our experience includes observation of a very few lesions in the larynges of dogs and monkeys. These were foci of inflammation and/or ulceration of the squamous epithelium in or near the junction of the lateral ventricles with the laryngeal lumen.