INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Viscosity and elasticity are the fundamental rheologic properties of respiratory mucus (1), and are important determinants of transportability of mucus in the mucociliary system (2, 3). In chronic upper and lower respiratory diseases, such as chronic sinusitis (CS) and chronic bronchitis (CB), both the viscosity and the elasticity of respiratory mucus are much higher than the optimal values for mucociliary transport (4, 5). Such abnormal rheologic properties could be a cause of decelerated mucociliary clearance in upper and lower respiratory diseases.

Respiratory mucus is a mixture of constituents derived from the respiratory mucosa and serum. The respiratory mucosa produces epithelial and glandular mucous glycoproteins, as well as immunoglobulins and proteins produced locally. The major serum contributions to respiratory mucus are immunoglobulins, albumin and plasma-type glycoproteins (6). Under pathologic conditions, the relative amounts of such constituents depends on the degree of hyperplasia or hypertrophy of secretory cells and on the transudation or exudation of serum components. It would therefore be useful to study the contributions of such biochemical constituents to the high viscoelasticity of nasal mucus for the purpose of improving the abnormal rheologic properties of mucus in CS. Although many biochemical factors affecting the viscoelastic properties of respiratory mucus have been reported (7, 8), the factors that predominantly contribute to the high viscoelasticity of respiratory mucus have never been fully elucidated.

The purpose of the present study was to investigate the relative contribution of biochemical constituents derived from respiratory mucosa and serum to the viscoelasticity of nasal mucus in patients with CS.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects and Collection of Nasal Mucus

We defined CS in our patients as an inflammation of the nasal and sinus mucosae with a persistent mucoid or mucopurulent nasal discharge for longer than 3 mo that resisted repeated antimicrobial therapy and antral irrigation (9). Twenty-four patients (10 males and 14 females, aged 9 to 73 yr) were entered into our study. No patients received any antibiotics, steroids, or mucokinetic agents during the period of study. The study protocols were approved by the Review Board for Human Studies of the Mie University School of Medicine, and informed written consent was obtained from the subjects or their surrogates if required by the institutional review board.

Nasal mucus was collected by aspiration from the nasal cavity. Care was taken to avoid stimulating the mucosal surface during the collection. The collected mucus samples were frozen at 80° C and were preserved for rheologic and biochemical measurements.

Measurement of Dynamic Viscoelasticity

The viscoelasticity of the nasal mucus was determined with an oscillating sphere magnetic rheometer (5). Approximately 4 µl of the test sample, together with a small iron sphere with a radius of 100 µm, was placed into the sample cell, which was placed in the space between the upper and lower magnetic poles of the rheometer. The iron sphere was oscillated in a vertical direction by applying sinusoidally varying magnetic field gradients. The motion of the iron sphere was followed through a light microscope and transduced to an electrical signal with an optoelectric tracker. The dynamic viscosity (') and elastic modulus (G') of the mucus were calculated from the obtained values of the amplitude of displacement of the iron sphere and the phase lag of the sphere with respect to the oscillatory driving force. ' is a measure of the viscous behavior of the sample, and G' is indicative of its elastic behavior. Both ' and G' were determined at 25° C at frequencies of 1.0 Hz and 10 Hz.

Biochemical Analysis

Fucose and N-acetyl neuraminic acid. Both fucose and N-acetylneuraminic acid (NANA) were separated from other sugars before assay. For this, nasal samples were mixed with 9 volumes of 10 mM Na phosphate buffer (pH 7.4) containing 0.15 M NaCl, and the mixture was homogenized in an ice-bath for 1 min with a Polytron homogenizer (Typ PT10/35; Kinematica AG, Lucerne, Switzerland).

For assay of fucose, the whole homogenates were hydrolyzed with 0.5 N H₂SO₄ for 3 h at 100° C in a sealed tube at 100- or 200-fold final dilution, and were filtered through a filter with a pore size of 0.22 µm (PVDF, GV type; Japan Millpore Ltd., Tokyo, Japan). Twenty microliters of the filtrate was subjected to high-pressure liquid chromatography (HPLC) on an Aminex HPX-87 column (7.8 × 300 mm; Bio-Rad Laboratories, Hercules, CA) with 0.01 N H₂SO₄ as the mobile phase at 60° C. The fucose isolated was subjected to postcolumn carbohydrate labeling with 2-cyanoacetamide (10), and its content was estimated from the absorbance at 276 nm with L-fucose (Wako Chemical Co., Tokyo, Japan) as a standard.

For assay of NANA, the homogenates were hydrolyzed with 0.05 N H₂SO₄ for 1 h at 80° C in a tube at 20- or 40-fold final dilution, and were subjected to HPLC on HPX-87 column with 0.01 N H₂SO₄ as the mobile phase at 40° C in the same manner as described earlier. The content of NANA isolated was estimated from the absorbance at 200 nm with NANA from goat submandibular gland (Fluka AG, Buchs, Switzerland) as a standard.

Albumin and IgG. Samples were diluted 2 × 10⁵-fold for albumin, and 2 × 10⁴-fold for IgG with phosphate-buffered saline containing 1% Tween. Albumin and IgG were measured with a sandwich-type enzyme-linked immunosorbent assay (11). Rabbit and goat antihuman serum albumin (IgG fraction; Cappel, Costa Mesa, CA), and rabbit peroxidase-labeled antigoat IgG serum (Seikagaku Kogyo, Tokyo, Japan) were used in the assay for albumin. Rabbit antihuman IgG serum (Behringwerke AG, Marburg, Germany) and goat alkaline phosphatase-labeled antihuman IgG (Tago Immunologicals, Camarillo, CA) were used in the assay for IgG. Standard serum was obtained from Behringwerke AG.

Secretory-IgA. Secretory IgA (S-IgA) was measured through the electroimmunodiffusion method (12), with purified human S-IgA from colostrum (Behringwerke AG) as a standard. The 10% homogenate of nasal mucus was centrifuged at 10,000 × g for 10 min at 4° C, and samples of 3 µl of the supernatant were used for the measurement.

Lysozyme. Lysozyme activity was measured according to the method of Smolelis and Hartsell (13), with Micrococcus lysodeikticus (Sigma Chemical Co., St. Louis, MO) as substrate. Egg-white lysozyme (EWL) was used as a standard. The amount of lysozyme was expressed as microgram equivalents of EWL.

Statistical Analysis

Because the variations in ' and G' were logarithmically distributed, the following statistics were performed after ' and G' values were subjected to common logarithmic transformation. Cook's statistics (14) were applied to the values of viscosity and elasticity for eliminating abnormally low or high values that would affect the statistical analysis. Because this method eliminated one sample from the viscosity analysis, 24 samples tested for elasticity and 23 samples tested for viscosity were used in the subsequent statistical analyses.

Linear regression analysis was done for the viscoelastic values versus each biochemical parameter, using Spearman's rank correlation coefficient (one-sided analysis).

A multiple stepwise regression (backward elimination procedure) was done to analyze whether viscosity and elasticity were influenced by the biochemical parameters examined in the study. In this, the contribution of the biochemical parameters to the viscoelasticity of mucus were determined by the square of the multiple correlation coefficient, which is known as the "coefficient of multiple determination adjusted by the numbers of samples" of a fitted regression model, and which is denoted by r²adj. The analysis was performed with StatView version 4.5 software (Abacus Concepts, Inc., Berkeley, CA). Values of p < 0.05 were considered significant. Data are expressed as mean ± SD.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Distributions of Rheologic and Biochemical Parameters

Table 1 shows the mean ± SD of ', G', and the concentrations of the biochemical parameters examined in the study. The ' and G' values at an oscillatory frequency of 1 Hz were 1.6 ± 1.5 Pa/s and 31.8 ± 31.0 Pa, respectively. At a frequency of 10 Hz, ' was 0.38 ± 0.34 Pa/s, and G' was 49.5 ± 45.2 Pa.

Relationship between Rheologic and Biochemical Parameters

Table 2 presents the results of the linear regression analyses of viscoelasticity measured at frequencies of 1 and 10 Hz versus the biochemical parameters. The viscosity significantly correlated with fucose, NANA, and IgG, respectively. Similar correlations were observed between elasticity and these same three parameters. The relationships between viscoelasticity at 1 Hz and fucose and IgG concentrations are shown in Figures 1 and 2.

View larger version (11K): [in this window] [in a new window]   Figure 1.   Linear regression of viscoelasticity against fucose concentration in nasal mucus from CS patients. There were positive correlations between ' and G' and fucose concentrations. ' and G' were determined at an oscillatory frequency of 1 Hz.

View larger version (13K): [in this window] [in a new window]   Figure 2.   Linear regression of viscoelasticity against IgG concentration in nasal mucus from CS patients. ' and G' were positively correlated with IgG concentration. ' and G' were determined at an oscillatory frequency of 1 Hz.

Biochemical Parameters Influencing to Rheologic Properties

Table 3 shows the results of multiple stepwise regression analysis of viscosity and elasticity determined at a frequency of 1 Hz with respect to fucose, NANA, IgG, S-IgA, albumin, and lysozyme. The results indicate that fucose had the highest influence on both viscosity and elasticity. IgG also affected viscosity and elasticity. The coefficient of multiple determination, r²adj, determined for fucose together with IgG, was 0.732 and 0.733 when the response variables were viscosity and elasticity, respectively. Table 4 shows the results of multiple regression analysis of biochemical factors influencing viscoelasticity determined at 10 Hz. Fucose was the only factor influencing viscosity, and r²adj determined with fucose was 0.689. Elasticity was influenced by fucose, IgG, and lysozyme, and r²adj as determined with these parameters was 0.794. 

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In order to relate rheologic properties to chemical structure and physiologic function, a nondestructive test that can be repeated over a suitable range of shear rates is needed. Such data can be obtained readily from dynamic oscillatory tests (15). The oscillating sphere magnetic rheometer used in the present study provides one such dynamic oscillatory test, in which a mucus sample is subjected to cyclic deformation and the deformation response of the mucus is observed. The resulting rheologic parameters are expressed as ' and G'. Cilia beat in a cyclic pattern, with a fast effective stroke and slow recovery stroke at about 10 Hz in the respiratory tract (16). By contrast, the ciliary beat may be 1 to 3 Hz under viscoelastic mucus load (16). For this reason, measurement of the dynamic viscoelasticity of mucus at frequencies in the vicinity of the ciliary beat frequency with and without mucus is best for examining the rheologic properties of mucus in relation to mucociliary transport mechanisms resulting from the interplay of  mucus and cilia. We therefore measured the dynamic viscoelasticity of mucus at the oscillatory frequencies of 1 and 10 Hz. As shown in Table 1, the mean value of viscosity of nasal mucus from CS patients was 1.6 Pa/s, and that of elasticity was 31.8 Pa (25° C) at a frequency of 1 Hz. In the CS nasal mucus, we observed that ' values of mucoid mucus and mucopurulent mucus were 6.8 ± 4.1 poise (0.68 ± 0.41 Pa/s) and 32.1 ± 27.9 poise (3.21 ± 2.79 Pa/s) (1 Hz, 25° C), respectively. The value of G' in mucoid mucus was 105.6 ± 83.6 dyne/cm² (10.56 ± 8.36 Pa), and in mucopurulent mucus it was 638.5 ± 520.0 dyne/cm² (63.85 ± 52.00 Pa) (1 Hz, 25° C) (17). These results indicate that the nasal mucus used in the present study had viscoelastic values typical of CS nasal mucus. The viscoelastic values of nasal mucus from our CS patients were also similar to the values for mucoid or mucopurulent sputum from patients with CB (18).

In a study of middle ear effusions (mucus) (3), the maximum mucociliary transport rate (MTR) of the effusion on mucus-depleted frog palate was achieved when the value of ' was 0.2 Pa/s (2 poises), and that of G' was 2 Pa (20 dyne/cm²) at a frequency of 1 Hz (25° C). There was a significant positive correlation between MTR and ' or G' at values below the optimal viscoelasticity. A significant negative correlation was also observed between MTR and ' or G' at values above the optimal viscoelasticity. The present study indicates that the viscoelasticity of nasal mucus in CS is much higher than the optimal viscoelasticity for mucociliary transport.

In addition to our previous findings (3), it is well established that the viscosity and the elasticity of mucus are important determinants of mucociliary transport. King and colleagues reported that there were negative linear correlations between the mucus transport rate on frog palate and logarithmic values of viscosity and elasticity in canine tracheal mucus (19). Dulfano and Adler reported a positive correlation between the frog-palate transport rate and the recoverable strain, a measure of compliance that is effectively the inverse of the elastic modulus, of sputum of bronchitic patients (4). Giordano and coworkers showed a decrease in mucociliary transport with increasing elastic modulus of whole mucus samples collected with the tracheal pouch technique (20). In the model analysis of mucociliary pumping by Ross and Corrisin (21), mucociliary velocity decreased with an increase in the mucus relaxation time, expressed as the ratio of viscosity to elasticity. A negative correlation between in vitro mucociliary transport velocity and the viscosity/elasticity ratio was confirmed experimentally (19). Lorenzi and colleagues suggested that a combination of the vectorial sum of viscosity and elasticity and the viscosity/elasticity ratio is the most important determinant of mucociliary transport (22). In any event, both the viscosity and the elasticity of mucus are essential for determining mucociliary clearance. It was for this reason that we evaluated the contributions of biochemical constituents to the viscosity and the elasticity of nasal mucus in our study.

In the present study, a multiple stepwise regression was performed to predict the viscoelasticity of mucus as a function of the biochemical parameters. Such an analysis has never previously been done of upper and lower respiratory mucus. The results indicate that fucose is the most important determinant of both the viscosity and the elasticity of nasal mucus from CS patients at any oscillatory frequency examined in this study (Tables 3 and 4). As shown in Figure 1, there was a highly significant positive correlation between the fucose concentration and the viscoelasticity of this mucus. Fucose is present in purified respiratory mucous glycoproteins and is virtually absent in serum glycoproteins (6). Thus, fucose is a marker of mucous glycoproteins. Because respiratory mucous glycoproteins are produced by goblet cells and submucosal gland cells in the respiratory mucosa (23), locally produced mucous glycoproteins from these secretory cells may largely contribute to the high viscoelasticity of nasal mucus in patients with CS.

NANA is present in similar amounts in mucous and serum glycoproteins (6). NANA has an axial carboxylic group and, owing to its negative charge, was hypothesized to play a central role in establishing crosslinks between glycoprotein molecules (24). Therefore, the NANA content of mucus has often been considered an index of mucus viscosity. In our present study, NANA was positively correlated with the viscoelasticity of nasal mucus (Table 2). This finding agrees with the view that a significant linear correlation existed between NANA and viscosity in the mucoid and mucopurulent sputum of CB patients (7). However, the multiple regression analysis in our study suggested that NANA has no significant effect in determining the viscosity and elasticity of nasal mucus in CS. Although the specific physicochemical action of NANA is still the subject of controversy, the result of our multiple regression analysis is consistent with the failure of total enzymatic removal of NANA (sialic acid) residues to affect the viscoelasticity of bovine cervical mucus (25).

IgG is a major component of serum immunoglobulins. In respiratory mucus it is mainly derived from serum and partly derived from local production. Our multiple regression analysis showed that IgG is the next most important determinant after fucose of the viscosity at 1 Hz and elasticity at 1 and 10 Hz of CS patients' nasal mucus. The coefficient of multiple determination (r²adj) determined with fucose and IgG was 0.732 and 0.733 when the response variables were viscosity and elasticity, respectively, at an oscillatory frequency of 1 Hz (Table 3). This indicates that fucose together with IgG accounted for about 73% of the variance of viscoelasticity of nasal mucus in CS. Because a significant positive linear correlation between IgG and the viscoelasticity of sputum was observed in patients with CB (18), the importance of IgG in viscoelasticity may not be specific to the nasal mucus. Marriott and colleagues (8) reported that the addition of IgG increased the viscoelasticity of purified human sputum, but this increase in viscoelasticity was not as great as the increase when S-IgA was added to the sputum. Thus, no exact explanation for the large contribution of IgG to viscoelasticity can be offered by the present study. The role of IgG as a determinant of viscoelasticity must be clarified by further study.

S-IgA is produced locally and is suggested to bind to mucous glycoprotein molecules by hydrogen or disulfide bonds (26). Close correlations were found between the S-IgA level and both apparent viscosity and retarded recoverable strain (elasticity) in bronchial mucus (7). Furthermore, the viscosity of purified human sputum was remarkably increased by the addition of S-IgA (8). However, S-IgA was not an important determinant of the viscoelasticity of nasal mucus in the present study.

List and associates (27) reported that the incubation of human serum albumin with pig gastric mucin increased the viscosity of the solution. They suggested that hydrophobic bonding may play a role in stabilizing the mucin-albumin interaction. They also showed that the maximum enhancement of mucin viscosity occurred with an albumin/mucin ratio of 2:1 (27). However, the present study clearly showed that albumin concentration was not an important determinant of the viscoelasticity of nasal mucus.

Lysozyme is a product of serous gland cells. Jenssen and colleagues reported that the addition of lysozyme to sputum from patients with chronic obstructive lung diseases significantly increased the viscosity of this material (28). They suggested that lysozyme acts as a crosslinking agent in mucus by an electrostatic mechanism that takes place between the negatively charged mucin molecules and the positively charged lysozyme molecules. However, our present study indicates that lysozyme probably does not play an important role in determining the viscoelasticity of nasal mucus in CS.

Using a pooled and reconstituted mucus, Litt and Khan studied the effects of concentrations of nondialyzable solids, pH, and ionic strength on the viscoelasticity of the mucus, and found that it was largely affected by nondialyzable-solid concentrations but was less affected by pH and ionic strength (29). Moreover, the pH of CS nasal mucus was found to be distributed within a narrow range (30). Since the nondialyzable solids of mucus include the biochemical constituents examined in the present study, the viscoelasticity of CS nasal mucus could depend on nondialyzable solids, especially fucose and IgG, rather than on pH and ionic strength.

DNA is a high-molecular-weight polymer mainly released from host neutrophils and suspected to contribute to the viscoelasticity of purulent sputum (31). Nevertheless, some controversy exists in the literature about the specific role of DNA. In the sputum of CB patients, no correlation was found between DNA content and viscoelasticity of sputum (7). Picot and associates suggested that other biochemical constituents than DNA may participate in the increased viscosity of sputum in CB (32). The role of DNA in the viscoelasticity of CS nasal mucus was not examined in the present study, but must be clarified in further experiments.