INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

VOCAL FOLD HYDRATION MUST be regulated to maintain the ease of phonation in an environment of disparate and rapidly changing humidities. For example, both superficial and intrinsic hydration of the vocal fold are critical for vocal fold oscillation to occur in an ex vivo canine model (8, 14). Challenges that predictably decrease vocal fold hydration, such as dry air and hyperosmolar fluids, also increase the subglottic pressure necessary to achieve and maintain vocal fold oscillation [phonation threshold pressure (8, 14)]. Dehydration challenges are thought to inhibit vocal fold oscillation by increasing the viscosity and decreasing the thickness of the vocal fold cover (8, 11). Such volumetric and rheological properties of the vocal fold cover govern the biomechanics of vocal fold oscillation [e.g., body cover theory (13) and mucoviscoelastic-aerodynamic theory (18, 32, 33)]. Whether the vocal fold mucosa generates transepithelial water fluxes toward and/or away from the lumen has not been explored. Similarly, in the vocal fold, any cellular mechanisms that regulate ion and fluid transport and, therefore, volumetric and rheological homeostasis have not been identified.

The vocal fold cover consists of mucosa that contains stratified squamous epithelium, basal lamina, and superficial layer of lamina propria (13). Wet, stratified squamous epithelia of similar structure to that of the vocal fold (e.g., those of the buccal, esophageal, and vaginal mucosae) generate a potential difference (PD) and short-circuit current (I_sc) indicative of transepithelial ion fluxes (7, 17, 22, 23) and consistent with transepithelial water fluxes. In the stratified, squamous epithelia of the aerodigestive tract, the Na⁺-K⁺-ATPase contributes to the generation of this PD (23, 24); however, any contribution of Na⁺ transport to vectorial water flux has not been demonstrated. In tracheal epithelia (consisting of pseudostratified, ciliated columnar cells and goblet cells), Phillips et al. (26) have shown that vectorial water transport is coupled to sodium transport and the activity of basolateral Na⁺-K⁺-ATPase (39). Thus, in the trachea, where pseudostratified epithelial cells are known to be electrically and structurally polarized, the active Na⁺-K⁺-ATPase pump maintains the electrochemical gradients responsible for the PD across the epithelia and supplies most of the energy driving the I_sc. We questioned whether stratified vocal fold epithelium might be electrically polarized and, if so, whether the Na⁺-K⁺-ATPase may regulate cation transport and thus vocal fold water flux.

We hypothesized that the Na⁺-K⁺-ATPase is present in vocal fold epithelia and contributes to its lumen-negative PD, I_sc, and transepithelial water flux. We identified the -subunit protein of the Na⁺-K⁺-ATPase in the vocal fold epithelium and submucosal glands of the canine larynx. We subsequently demonstrated the contribution of the Na⁺-K⁺-ATPase to the transepithelial PD, I_sc, and water flux by luminal application of acetylstrophanthidin (a specific Na⁺-K⁺-ATPase inhibitor) to native canine and ovine vocal fold epithelium. The results support the presence of an active transport system that regulates transepithelial ion and water flux in the vocal fold. This system may provide the foundation for adaptive cellular mechanisms capable of responding to the osmotically challenging environment of the vocal folds.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The identification and localization of -subunit protein of the Na+-K⁺-ATPase. In accordance with approved protocols at Northwestern University and University of Illinois at Chicago, the larynx and kidneys were excised from 11 dogs (canis familiaris, age 3-4 yr, 25-35 kg) that were undergoing other nonsurvival procedures. Surgical anesthesia was induced with pentobarbital sodium (25 mg/kg iv). Vocal fold and kidney tissues were frozen or fixed immediately after excision. Animals were killed by a lethal dose of pentobarbital sodium (100 mg/kg iv) followed by thoracotomy. Masson's trichrome and hematoxylin and eosin preparations were used to stain the vocal fold tissues. Western blot and immunohistochemistry were used to detect the presence of the -subunit protein of the Na⁺-K⁺-ATPase. For Western blots (n = 3), the frozen tissue was rapidly homogenized in the presence of 1% Triton X-100 in 20 mM Tris buffer and 150 mM NaCl, as well as the protease inhibitors benzamidine, leupeptin (5 µg/ml each), and 1 mM phenylmethylsulfonyl fluoride. For histology and immunohistochemistry (n = 11), tissues were fixed in 4% paraformaldehyde and embedded in paraffin for subsequent sectioning and staining.

Contribution of the Na+-K⁺-ATPase to electrophysiology of native vocal fold mucosa. To perform the electrophysiological experiments, canine (n = 6) and ovine (n = 14) larynges were obtained. Canine larynges (from mongrel dogs, 45-60 kg, age 3-4 yr) were excised under surgical anesthesia, which was induced with pentobarbital sodium (25 mg/kg iv). Dogs were killed by a lethal dose of pentobarbital sodium (100 mg/kg iv) followed by thoracotomy. Ovine larynges that were larger than the canine larynges were obtained from a local abattoir. As soon as possible after excision, larynges were placed in 4°C Hanks' balanced salt solution (HBSS) and transported directly to the laboratory.

The role of Na+-K⁺-ATPase in vocal fold transepithelial water flux. The method used for simultaneous measurement of vectoral water fluxes and electrophysiological parameters has been validated elsewhere in studies of airway mucosae (25, 26). The technique employed 8-aminonaphthalene-1,3,6-trisulfonic acid (ANTS; Molecular Probes, Eugene, OR), a fluorescent probe that has a threefold greater quantum yield when in ²H₂O-HBSS than in H₂O-HBSS. When 1 mM ANTS ²H₂O and H₂O solutions bathed the basal and luminal sides of a membrane, respectively, temporal changes of fluorescence emitted by ANTS were used to detect unidirectional movements of ²H₂O and H₂O across the membrane. The method yields independent estimates of the simultaneous, unidirectional water fluxes.

w BL

w LB

w BL

w LB

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The identification and localization of -subunit protein of the Na+-K⁺-ATPase. A coronal section of a canine vocal fold is shown in Fig. 1A. The canine vocal fold mucosa (Fig. 1A) overlies the thyroarytenoid muscle complex. It can be seen that the mucosa consists of epithelium and a highly vascularized lamina propria with numerous lymphatic vessels. Glands and their associated ducts are most numerous at the inferior and superior margins of the vocal fold. These ducts as well as lymphatics are shown at higher magnification in Fig. 1, B and C. In all 11 dogs, the epithelium of the vibratory margin of the vocal fold consisted of a nonkeratinized, stratified, squamous epithelium of five to eight cell layers, as typified in Fig. 2, A and B. There was a brief transitional segment where the morphology of the cells and tissue changed from stratified, squamous epithelium to pseudostratified, ciliated, columnar epithelium, with goblet cells that lined the trachea and laryngeal ventricles. This transitional segment of the epithelium was characterized by a gradual increase in cell height (Fig. 2, C and D). There were numerous submucosal glands found beneath the ciliated epithelium of the trachea and laryngeal ventricles (not shown) and less frequently under transitional epithelium lining the inferior vocal fold (Fig. 2C). Numerous glands were seen under transitional epithelium that lined the inferior, posterior, cartilaginous glottis (Fig. 2, E and F). One glandular duct was observed to open onto the stratified, squamous epithelium of the vocal fold (Fig. 1, A and B). Numerous ducts were associated with the transitional region of the epithelium (Fig. 1C). The glands were of the mixed seromucous type with numerous serous demilunes (Fig. 2, E and F).

View larger version (97K): [in this window] [in a new window]   Fig. 1.   A: hematoxylin and eosin stain of the midmembranous canine vocal fold (magnification ×8), in cross section along the coronal plane, showing the mucosa (cover) overlying the body (muscular portion). B: inset shown in A magnified ×80. Numerous lymphatics (L) and the duct (D) of an exocrine gland underlying the stratified squamous epithelium are identified. C: insert shown in A magnified ×70. Shown are lymphatics and ducts underlying transitional epithelium.

View larger version (128K): [in this window] [in a new window]   Fig. 2.   A: stratified squamous epithelium and submucosa of the midmembranous canine vocal fold (magnification ×100). B: inset shown in A magnified ×400. C: transitional epithelium of the inferior glottis (magnification ×100). D: inset shown in C magnified ×400. E: seromucous glands from the posterior, cartilaginous vocal fold (magnification ×100). F: inset shown in E magnified ×400.

View larger version (147K): [in this window] [in a new window]   Fig. 3.   A: saggital section (magnification ×100) of canine kidney cortex showing the immunoreactivity of the thick ascending limb (TAL) of the outer medulla distal (D) tubules but not the renal corpuscle (RC). Some of the proximal tubules (P) were immunopositive, whereas others were not. The range of intensity of staining is shown in the magnified view of the panel inset (×360; B) and in a saggital section of renal cortex from another dog (×360; C). D: no immunopositive stain was observed in procedural controls (magnification ×390).

View larger version (92K): [in this window] [in a new window]   Fig. 4.   Immunoreactivity of the vocal fold epithelium was observed in the basal cells and immediately adjacent 1-2 layers of suprabasal cells (magnification ×500; A) but not in a negative procedural control (magnification ×560; B).

View larger version (131K): [in this window] [in a new window]   Fig. 5.   Immunoreactivity of vocal fold submucosal glands (magnification ×360). The immunopositive staining associated with basolateral plasma membranes is shown by the black arrows. Yellow arrowheads denote serous components.

Contribution of the Na+-K⁺-ATPase to the electrophysiology of native vocal fold mucosa. The electrophysiological viability of the tissue preparations was maintained for >3 h. The 14 ovine tissues had a mean baseline PD of 9.3 ± 1.3 mV (lumen negative) and generated an I_sc of 31 ± 4 µA/cm². This resulted in a mean baseline tissue resistance of 342 ± 48  · cm² for ovine membranes. Similarly, the six canine tissues had a mean PD of 8.1 ± 2.8 mV (lumen negative) and generated I_sc of 41 ± 10 µA/cm². This resulted in a mean baseline tissue resistance of 254 ± 81  · cm² for the canine specimens. Figure 6 shows an example of the current generated by a selected PD of a tissue from a canine vocal fold. The baseline current-voltage relation, representative of R_m, was linear. The offset of the I_sc at zero PD was the baseline I_sc for this membrane.

View larger version (15K): [in this window] [in a new window]   Fig. 6.   The current-voltage relation for a canine vocal fold epithelium [obtained by clamping the potential difference (PD) of the tissue between 50 and +50 mV]. The current at a selected PD was recorded. This current was normalized to the surface area of the tissue. The data were fit by the line y = 0.08x  4.42 to yield a surface area normalized resistance of 80  · cm2.

View larger version (17K): [in this window] [in a new window]   Fig. 7.   Reduction of the lumen-negative PD of canine (A) and ovine (B) vocal fold epithelium induced by luminal application of acetylstrophanthidin. A: baseline electrophysiological properties for the canine membrane were PD = 5 mV, short-circuit current = 59 µA/cm2, and membrane resistance = 85  · cm2. The time constant for the canine membrane was 500 s. B: for the ovine membrane, baseline PD = 8.5 mV, short-circuit current = 22 µA/cm2, and membrane resistance = 390 cm. The time constant for this membrane was 1,400 s.

The role of Na+-K⁺-ATPase in vocal fold transepithelial water flux. The reduction in bidirectional water flux (P = 0.014) by luminal acetylstrophanthidin (n = 7) is shown in Fig. 8. The baseline J_w BL of 5.1 ± 0.3 µl · min¹ · cm² was reduced by 15% to 4.3 ± 0.3 µl · min¹ · cm² within 45 min. The baseline J_w LB of 5.2 ± 0.2 µl · min¹ · cm² was reduced by 25% to 3.9 ± 0.3 µl · min¹ · cm² within 45 min. Application of acetylstrophanthidin reduced J_w LB to a greater extent than J_w BL, as shown by the significant treatment by flux direction interaction (P = 0.035). Two post hoc comparisons by Student's paired t-test confirmed that acetylstrophanthidin reduced J_w LB (P = 0.005) to a greater extent than J_w BL (P = 0.027) and showed both responses to be significant. The baseline open-circuit PD of 8.7 ± 1.2 mV and its associated I_sc of 44 ± 11 µA were reduced to 1.3 mV ± 0.5 mV and 27 ± 12 µA, respectively, replicating results from the electrophysiological experiments.

View larger version (21K): [in this window] [in a new window]   Fig. 8.   Significant reductions of unidirectional water fluxes (J) across ovine vocal fold epithelia (n = 7) were caused by luminal application of acetylstrophanthidin. B to L, basal to luminal chamber; L to B, luminal to basal chamber.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In the stratified, squamous epithelium of the vocal fold, the 100-kDa -subunit protein of the Na⁺-K⁺-ATPase was observed predominantly in association with the plasma membranes of the basal and adjacent one to two layers of suprabasal epithelial cells and perhaps, to a lesser extent, the luminal surface cells (Fig. 4). In seromucous glands, it also was found predominantly in association with the basolateral membranes of mucous cells, as well as the plasma membranes and cytosol of serous cells (Fig. 5). These glands were located beneath the tracheal epithelium and the transitional epithelium of the inferior and superiolateral vocal fold margins but not beneath the stratified squamous epithelium of the membranous vocal fold (Figs. 1 and 2). The vocal fold mucosa was polarized with the lumen negative, indicating an asymmetric cellular distribution or activity of the Na⁺-K⁺-ATPase. As the basal and possibly the luminal epithelial cells stained for the -subunit protein of the Na⁺-K⁺- ATPase, these cell layers are likely responsible for much of the observed PD. This PD was reduced by the addition of acetylstrophanthidin, a specific inhibitor of the Na⁺-K⁺-ATPase added to the luminal rather than the basal surface. The pronounced luminal response demonstrates the ready luminal accessibility of the epithelial Na⁺-K⁺-ATPase to acetylstrophanthidin and the presence, on the basal side, of a barrier to its diffusion. We demonstrated that bidirectional water fluxes across vocal fold mucosa were reduced by acetylstrophanthidin. Together these data indicate that vocal fold hydration can be controlled by the Na⁺-K⁺-ATPase in the cells of squamous epithelium through the regulation of transepithelial ion and water fluxes.

The vocal folds are subjected to osmotic challenges. Superficial hydration of the vocal fold has been attributed to secretions of the tracheobronchial airways and laryngeal glands extrinsic to the vibratory margin of the vocal fold (10, 19, 29, 31). Secretions from the lower airways, however, amount only to 0.5 mg · kg body wt¹ · day¹ (34, 41). The secretions are transported up the trachea and converge to pass almost exclusively through the posterior commissure of the larynx (3) during quiet respiration. Here we observed that ducts from mucosal glands open onto the transitional epithelium with some opening onto squamous epithelium of the vibratory vocal fold surface (Fig. 1, A and B). We showed that the electrically polarized, stratified epithelium of the vocal fold possesses an intrinsic Na⁺-K⁺-ATPase-dependent mechanism that regulates ion and water transport. In addition, the presence of the Na⁺-K⁺-ATPase provides the basis for the active volume regulation of vocal fold epithelium, important in this tissues' role as a tunable baffle for mechanical stress of phonation. Shear and longitudinal stress of phonation provide an additional challenge to ion transport, as well as the barrier properties of the epithelium, as has been shown in other tissues and cells (22, 37, 38). Both mechanical and osmotic challenges are implicated in numerous pathological states of vocal fold mucosa, most of which occur in the epithelium, basal lamina, or superficial layer of the lamina propria (e.g., region shown in Fig. 2B).

Albeit the ovine vocal fold and tracheal epithelial tissue both demonstrate a marked temporal response to acetylstrophanthidin of similar time course [time constant = 23.8 vs. 28.5 min (26)], these responses were obtained by application of acetylstrophanthidin to the luminal and basal surfaces, respectively. As the vocal fold generates a lumen-negative 9-mV PD, there must be an obligatory preponderance of the Na⁺-K⁺-ATPase or its activity in the basal direction of the cells in these tissues. The delayed and reduced action of acetylstrophanthidin when applied to the basal surface may be due to the magnitude of the unstirred layer on the basal side caused by some 2 mm of connective tissue vs. only <0.2 mm of superficial cells on the luminal side (Fig. 2, A and B).

We have considered the possibility that some of the PD measured was due to the epithelium on the superior and inferior margins of the vocal fold cover, where there is a transition between vocal fold epithelium and ciliated columnar epithelium covering the adjacent tracheal mucosa. The surface areas of the canine and ovine stratified, squamous vocal fold epithelia used for these electrophysiological studies were >2 cm², thus minimizing any active transitional or columnar epithelia within the chamber. This was confirmed by the observations that luminal rather than basal acetylstrophanthidin more effectively inhibited the PD in the epithelia tested.

The Na⁺-K⁺-ATPase contributes to lumen-negative polarization of other wet epithelia, including those of the bronchial (16), tracheal (26), esophageal (23, 28), and buccal (24) mucosa. Similarly, vaginal mucosa is electrically polarized (7, 17), consistent with active ion transport. Similar to the vocal fold, esophageal, buccal, and vaginal epithelia are subjected to substantial mechanical stress in addition to other environmental insults. All of these epithelia are also characterized by a layer of basal cells that have numerous desmosomes and are anchored to the basal lamina via hemidesmosomes. Such epithelia also have numerous layers of suprabasal cells. The frog skin, however, provides an example in which the Na⁺-K⁺-ATPase can be preferentially localized to cell layers of a stratified, squamous epithelium, specifically the stratum spinosum and stratum germinativum [i.e., stratum basale (21)]. In comparison, the wet, stratified, squamous epithelia of the mammalian aerodigestive tract have no stratum spinosum per se, and all but the most superficial desquamating cells are viable. To our knowledge, the intense staining of the Na⁺-K⁺-ATPase with selected cell layers of a wet, stratified, squamous epithelium has not been demonstrated previously in humans or other mammals.

The stratified, squamous epithelium of the vocal fold sustains an electrophysiological PD and I_sc. There are, however, obvious structural differences from surrounding pseudostratified, ciliated, columnar epithelium. The mean baseline PD (9.3 ± 1.3 and 8.1 ± 2.8 mV for ovine and canine vocal folds, respectively) approximated that of ovine tracheal epithelium (lumen negative, 11.7 ± 1.1 mV) when it was obtained with the use of the same apparatus (26). These estimates are lower than those of the trachea obtained in vivo and in vitro by using a Ringer-filled exploring bridge, where lumen-negative PD is approximately equal to 30 and 37 mV for canine and ovine tracheae, respectively (1, 2, 6, 20). It is possible that glandular and other secretory cells, numerous in the trachea but absent beneath the stratified squamous vocal fold epithelium (Figs. 1 and 2), could contribute to a greater PD for tracheal mucosa than for the vocal fold. On the other hand, these differences could be due to the "state" of the tissues.

Our membranes generated mean I_sc (ovine I_sc = 31 ± 4 µA/cm², canine I_sc = 41 ± 10 µA/cm²) much like those shown when the same apparatus is used for the ovine trachea [mean = 37 ± 4 µA/cm² (26)]. In this preparation, vocal fold membranes yielded electrical resistance values (ovine R_m = 342  · cm² and canine R_m = 254  · cm²) similar to the reported values for the trachea [ovine R_m = 362 ± 33  · cm² (26)]. These values of resistance are notably lower than the buccal mucosa [canine R_m = 1,090 ± 100  · cm² (24)] and vaginal mucosa [rat R_m = 827-2,366 , depending on the stage of the oestrus cycle (7)]. The PD and I_sc of the tracheal epithelium are generated by a pseudostratified epithelial layer with tight junctions near the apical membrane, providing a high-resistance pericellular return path. The PD and I_sc of the vocal fold epithelium appear to be generated by multiple cells in series and a convoluted pericellular return path. It is not clear whether the PD is generated by the luminal squamous cells and/or the multiple basal and immediately adjacent cells in series together with a convoluted pericellular return path between these multiple layers of cells (Fig. 9). The maintenance of an electrochemical gradient and transmembrane resistance indicates the presence of a barrier to the diffusion of ions through the pericellular path. Some models of ion and water transport have incorporated cells in series (6, 30, 36). It could be that the luminal squamous cells are primarily responsible for the PD across the vocal fold, whereas the basal and immediately suprabasal cells are specialized for volume regulation. In the vocal fold epithelium, tight junctions or complete junctional complexes have not yet been identified in luminal cells, yet our physiological data and transmission electron microscopy (Fig. 9A) may suggest their presence. The mechanism(s) and structures underlying polarization of the vocal fold and its vectorial fluxes of water await further elucidation.

View larger version (102K): [in this window] [in a new window]   Fig. 9.   Transmission electron microscopy of ovine vocal fold epithelium showing a possible junctional complex and a convoluted pericellular pathway between luminal squamous cells (magnification ×122,000; A) and spinous cells of the ovine vocal fold epithelium (magnification ×74,000; B). Cells with spinous projections were 4 cells beneath the luminal cell within the same ovine specimen. Numerous desmosomes (D) are noted. The location of plasma membrane specializations in the regions expected for tight (T) and intermediate (I) junctions are indicated.

The bidirectional, transepithelial water fluxes of the ovine vocal folds (J_w BL of 5.1 ±0.3 µl · min¹ · cm² and J_w LB of 5.2 ± 0.2 µl · min¹ · cm²) approximated the magnitude found by using the same method in ovine tracheae [J_w BL of 4.5 ± 0.3 µl · min¹ · cm² and J_w LB of 6.1 ± 0.3 µl · min¹ · cm² (25, 26)]. As any pressure differential was <3 mmH₂O, its contribution to the measured water flux would be negligible. When tritiated water is used, Phipps and colleagues (27) have reported similar water flux values for the tracheal epithelium of sheep. The tracheal mucosa, however, favors a baseline net flux of absorption (25, 26), whereas this degree of asymmetry was not evident in the baseline water fluxes across the vocal fold mucosa. In this respect, we note that the method reported here is good for measuring unidirectional water fluxes. Subtraction of the fluxes propagates the experimental error, leading to uncertainty in the net flux, given the number of tissues studied. The present method, however, is improved over those using tritiated water, as the water fluxes in both directions are measured simultaneously in the same tissue.

The magnitude of acetylstrophanthidin-induced inhibition of water fluxes depended significantly on flux direction, with J_w LB reduced by 25% and J_w BL by 15%. This suggests that the water flux toward the mucosa involves active Na⁺ transport. Water flux toward the lumen may involve the transport of an indirectly regulated counterion such as Cl or HCO. In the trachea also, acetylstrophanthidin decreased the J_w LB (26). When aerosolized acetylstrophanthidin is administered to the trachea in vivo, there is an increase in airway hydration (1, 2) and mucociliary transport (40). Superficial application of acetylstrophanthidin to the vocal folds can be predicted to induce swelling of epithelial cells as well as reduce fluid absorption, thereby increasing the hydration of the vocal fold epithelium as well as the surface fluid. Present findings suggest that topical pharmacological approaches may be able to exploit active ion transport to fine-tune the volume, rheological properties, and superficial hydration of vocal fold mucosa.