Thioredoxin liquefies and decreases the viscoelasticity of cystic fibrosis sputum

Raymond C. Rancourt, Shusheng Tai, Malcolm King, Sonya L. Heltshe, Churee Penvari, Frank J. Accurso, Carl W. White


The persistent and viscous nature of airway secretions in cystic fibrosis (CF) disease leads to airway obstruction, opportunistic infection, and deterioration of lung function. Thioredoxin (Trx) is a protein disulfide reductase that catalyzes numerous thiol-dependent cellular reductive processes. To determine whether Trx can alter the rheological properties of mucus, sputum obtained from CF patients was treated with TRX and its reducing system (0.1 μM thioredoxin reductase + 2 mM NADPH), and liquid phase-gel phase ratio (percent liquid phase) was assessed by compaction assay. Exposure to low Trx concentrations (1 μM) caused significant increases in the percentage of liquid phase of sputum. Maximal increases in percent liquid phase occurred with 30 μM Trx. Additional measurements revealed that sputum liquefaction by the Trx reducing system is dependent on NADPH concentration. The relative potency of the Trx reducing system also was compared with other disulfide-reducing agents. In contrast with Trx, glutathione and N-acetylcysteine were ineffective in liquefying sputum when used at concentrations <1 mM. Sputum viscoelasticity, measured by magnetic microrheometry, also was diminished significantly following 20-min treatment with 3, 10, or 30 μM Trx. Similarly, this reduction in viscoelasticty also was dependent on NADPH concentration. Further investigation has indicated that Trx treatment increases the solubility of high-molecular-weight glycoproteins and causes redistribution of extracellular DNA into the liquid phase of sputum. Recognizing that mucins are the major gel-forming glycoproteins in mucus, we suggest that Trx alters sputum rheology by enzymatic reduction of glycoprotein polymers present in sputum.

  • sputum viscoelasticity
  • mucin
  • mucus
  • glutathione
  • N-acetylcysteine
  • deoxyribonucleic acid

cystic fibrosis (CF) is a common lethal genetic disease that results from a mutation in the gene encoding a chloride channel protein, the CF transmembrane conductance regulator (CFTR). As a result of this defect, epithelia within the body are impermeable to chloride ion transport (4). Although several organs are affected, including the pancreas, intestine, and male genital tract, complications within the lung account for 95% of the morbidity and mortality (35). In the lung, impaired chloride transport into the airway lumen leads to excessive absorption of Na+ and fluid, reducing the volume of airway surface liquid (14). Desiccation of airway surface liquid leads to the concentration of mucin macromolecules, which are the gel-forming constituents of mucus (21). The viscoelastic properties of normal mucus are dependent on the concentration, molecular weight, and entanglements between mucin polymers (34). Further interaction of mucins with DNA (18, 19, 23) and f-actin polymers (27, 33) released from dying inflammatory cells is responsible for the dense and viscous nature of CF sputum. The inability to clear such mucus by cough or mucociliary clearance worsens lung function and facilitates colonization of the lung with opportunistic pathogens. Chronic lung injury results from a persistent cycle of bacterial infection and inflammatory response (16).

Despite some promising advances, correction of CF by gene therapy is not yet attainable. Currently, antibiotic regimens, coupled with drugs that facilitate the clearance of purulent airway secretions, remain the mainstay treatments for progressive airway disease. Inhalation of purified recombinant human DNase (Pulmozyme, Genentech), which digests extracellular DNA present in the CF airway, is widely used as a respiratory decongestant. Such treatment is clinically effective for diminishing sputum viscosity and has some capacity to improve the forced expiratory volume (11). Other investigative therapies aimed at breaking down mucin or actin polymers, including N-acetylcysteine (NAC), nacystelyn (an N-acetyl-l-cysteine derivative), and gelsolin, can also reduce sputum viscosity experimentally but have yet to attain approval for the application of CF treatment in the United States.

In this investigation, we examined the effects of the thioredoxin (Trx) reducing system [Trx, Trx reductase (TR), and NADPH] on the physical and rheologic properties of CF sputum in vitro. Trx is a 13-kDa protein containing two redox-active cysteines at its active site (Trp-Cys-Gly-Pro-Cys), which are highly conserved across species. In their oxidized form, these cysteines form a disulfide bridge that protrudes from the three-dimensional structure of the protein (12). Reduction of this active center by the NADPH-dependent TR enzyme allows Trx to function as an electron carrier with dithiol/disulfide exchange capability (22). Protein disulfides are a preferred substrate for Trx-mediated reducing action, and these are readily reduced by Trx with a potency exceeding dithiothreitol (DTT) (12). Intracellular concentrations of Trx range from 1 to 20 μM with distribution in cytoplasmic, membrane, and mitochondrial compartments (12). Thiol-redox control by Trx can regulate a number of biological functions involved in cell signaling and growth. These include the provision of reducing equivalents for ribonucleotide reductase, the enzyme that synthesizes deoxyribonucleoside triphosphates for DNA synthesis. Trx also facilitates the hormone binding activity of the glucocortocoid receptor and the DNA binding activities of the arylhydrocarbon receptor as well as basal transcription factor III C complex, activator protein-1, hypoxia-inducible factor-1, NF-κB, and other transcription factors (reviewed in Ref. 20). In addition to being an antioxidant itself (6), Trx is an essential cofactor for Trx-dependent peroxidases (5) and is capable of increasing manganese superoxide dismutase expression (7).

Cells also have the capacity to release Trx into extracellular space. The mechanism of release is unknown but appears to deviate from the “classical” endoplasmic reticulum/Golgi pathway of protein secretion due to the lack of an NH2-terminal secretory sequence on the Trx molecule (24). In terms of function, exogenous Trx in vitro can increase IL-2 receptor expression (29) and displays chemotactic or mitogenic properties toward hematopoietic cell types (3). It is uncertain whether extracellular Trx has any additional physiological role.

Because respiratory mucins contain several cysteine domains that are believed to play an essential role in polymerization (1, 2), we hypothesized that Trx could serve as an effective extracellular mucolytic by reduction of mucin disulfides. The present study demonstrates that in vitro treatment with catalytic amounts of Trx and its reducing system can liquefy and decrease the viscoelasticity of purulent CF sputum. The increased solubility of high-molecular-weight glycoproteins present in Trx-treated sputum suggests that mucin macromolecules may be the substrates reduced by Trx during the mucolytic process.


Reagents and materials. Lyophilized recombinant Escherichia coli Trx was obtained from Promega (Madison, WI). E. coli TR was from American Diagnostica (Greenwich, CT). β-NADP, reduced-form NADPH, reduced GSH, glutathione reductase, DTT, disopropylflurophosphate, aprotinin, N-ethyl maleimide, Schiff reagent, salmon testes DNA, and Hoechst dye were all obtained from Sigma Chemical (St. Louis, MO). NAC was from Fisher Scientific (Pittsburgh, PA). All other chemicals were of the highest possible grade.

Sputum collection. Sputum was obtained from adult and pediatric patients with CF at The Children's Hospital of Denver (Denver, CO). Patients were diagnosed with CF if they had sweat chloride values in excess of 60 mM in two separate pilocarpine iontophoresis sweat tests and exhibited two allelic CF-producing mutations in subsequent genetic analysis. All samples were donated by either spontaneous expectoration or hypertonic saline induction. Sputum samples containing visibly detectable saliva were discarded. After expectoration, samples were stored at -80°C until their time of use. Sputum collection protocol, data collection, and consent/assent forms were approved by the Institutional Review Board of the University of Colorado Health Science Center and affiliated hospitals. The sputum samples analyzed in all experiments were not pooled but obtained from one patient at a single time of donation. Data regarding the sputum samples and the donors providing them are provided in Table 1.

View this table:
Table 1.

Summary of sputum donor information

Compaction assay. CF sputum stored at -80°C was thawed at room temperature and aliquotted into 1.5-ml Eppendorf centrifuge tubes at volumes of 275 μl via a positive displacement pipette (Rainin, Emeryville, CA). Sputum samples were subjected to either diluent (H2O), Trx + TR + NADPH, GSH + glutathione reductase + NADPH, DTT, or NAC treatment by the addition of 25 μl of H2O containing the appropriate molar concentration of each agent. After brief vortexing (∼1 s), sample tubes were loaded onto a microtube rotisserie (Barnstead, Dubuque, IA) and incubated at 37°C for 20 min. Samples were then processed for compaction assay according to methodology originated by Daugherty et al. (8). To perform the assay, we loaded the contents of each sample into 100-μl glass microcapillary tubes (Fisher Scientific) that had been previously welded to 200-μl pipette tips to achieve a tight fit. Three modified capillary tubes were used to draw up >90% of each sputum sample. Capillary tubes were then removed from their pipette tip, sealed with clay, and centrifuged for 10 min in a hematocrit centrifuge (IEC, Needham Heights, MA), followed by measurement of the length in millimeters of the gel (solid) and aqueous (liquid) phases in each tube. We calculated the percent liquid fraction of each capillary tube by dividing aqueous phase length by total length (gel + aqueous) × 100. The three measurements of the liquid fraction (%) derived from each sample were then averaged to generate a single value for each treatment condition. Compaction assay measurements were performed on five separate sputum samples for both Trx and NADPH dose-response measurements (Fig. 1), three independent samples for Fig. 2, five samples for GSH, and four samples for NAC (Fig. 3).

Fig. 1.

A: liquefaction of cystic fibrosis (CF) sputum by exposure to the thioredoxin (Trx) reducing system is dose dependent. Equal volumes of sputum samples were treated with 25 μl of H2O containing 0, 1, 10, or 30 μM Trx; 0.1 μM Trx reductase (TR); and 2 mM NADPH (final concentration). After a 20-min incubation, the liquid fraction of each sample was determined by compaction assay. B: liquefaction of CF sputum by Trx is dependent on NADPH. Sputum was treated with 30 μM Trx; 0.1 μM TR; and 0, 0.2, 0.6, 1.0, or 2.0 mM NADPH for 20 min. Determination of the liquid fraction following treatment was determined by compaction assay. Values are means from 5 independent experiments. *P < 0.05 vs. H2O-exposed samples.

Fig. 2.

Assessment of compaction assay reproducibility. Freshly isolated CF sputum from 3 different donors (A, B, C) was separated into 275-μl aliquots and frozen. After being thawed, aliquots were incubated without treatment or with H2O, 10 μM Trx (+ 0.1 μM TR and 2 mM NADPH), or dithiothreitol (DTT, 1 or 5 mM) for 20 min, and the percent liquid was measured by compaction assay. Results obtained from 3 independent experiments performed on each donor sample were used to evaluate assay reproducibility.

Fig. 3.

Trx reducing system (Trx + 0.1 μM TR + 2 mM NADPH) is more potent at sputum liquefaction than the glutathione-reducing system (GSH + 0.1 μM Gr + 2 mM NADPH, A) or N-acetylcysteine (NAC, B). Sputum samples were aliquoted and treated with Trx or GSH for 20 min, and percent liquid was determined by compaction assay. No treatment, 2.7 ± 2.3%; H2O, 4.5 ± 2.7%; 10 μM Trx, 34.8 ± 6.6%; 30 μM Trx, 54.6 ± 10.4%; 60 μM Trx, 67.0 ± 8.0; 30 μM GSH, 7.8 ± 5.7; 100 μM GSH, 15.9 ± 9.3%; and 1 mM GSH, 27.6 ± 3.9%. The analysis of Trx and NAC efficacy also was determined after a 20-min incubation but on a different set of sputum samples. Values are means from 5 (GSH) or 4 (NAC) experiments.

Magnetic microrheometry. Viscoelastic change in response to treatment was measured by means of a magnetic microrheometer as developed by King (15). An 80- to 120-μm steel sphere was placed in a 10-mg sputum sample. An electromagnet was used to oscillate this sphere, whose image was projected onto a pair of photocells via a microscope. The mucus retarded the motion of the sphere, and we revealed this effect by plotting the motion of the sphere against the driving force of the magnet on an oscilloscope, from which G* was measured. G* was the mechanical impedance or vector sum of viscosity and elasticity. For Trx and NADPH dose-response experiments, log G* at 10 rad/s was measured before any treatment (baseline) and then after 20-min incubation with no treatment, diluent (H2O), or Trx with reducing system. All treatments were administered to the sample in a volume of H2O equal to 10% of total sample volume. One measurement was performed per aliquot for each of the four sputum samples tested.

Glycoprotein extraction from sputum. Extraction of soluble glycoproteins from sputum was performed according to methodology outlined by Davies and Carlstedt (9). CF sputum (275 μl) was treated for 20 min at 37°C with 25 μl of H2O alone or H2O containing the Trx reducing system such that final concentrations were Trx (10 or 30 μM) + NADPH (2 mM) and TR (0.1 μM). After treatment, 100 μl of H2O containing 1 mM disopropylflurophosphate and 10 μg/ml aprotinin were added to each sample, followed by 15-min centrifugation at 22,000 g at 4°C. The resulting supernatant (aqueous phase) of each sample was transferred to a new microcentrifuge tube and stored at -20°C. The remaining solid gel portion of each sample was carefully unseated from the tube bottom in the presence of 250 μl of guanidinium extraction buffer (6 M guanidinium chloride, 5 mM EDTA, 10 mM sodium phosphate buffer, pH 6.5, 1 mM N-ethyl maleimide, 100 μM disopropylflurophosphate, and 1 μg/ml aprotinin) using a pipette tip and rotated for 14 h at 4°C. After centrifugation, the resulting supernatant from this gel phase extraction was then transferred to a clean tube and frozen at -20°C until time of electrophoresis.

Analysis of glycoprotein content. The glycoprotein content of aqueous and gel phase samples was evaluated by staining with periodic acid-Schiff (PAS) reagent according to methodology outlined by Thornton et al. (32). Aqueous and gel samples were thawed, and 80-μl aliquots of each were loaded onto a 1.0% agarose gel (150 × 125 mm) housed within a Biomax horizontal electrophoresis apparatus (Kodak, Rochester, NY). Electrophoresis reagents were as follows: electrophoresis buffer: 40 mM Tris-acetate, 1 mM EDTA, pH 8.0, and 0.1% SDS; sample loading buffer: 60% electrophoresis buffer, 40% glycerol (vol/vol), and 0.005% (wt/vol) bromphenol blue. Gel contents were transferred to polyvinylidene (PVDF) membrane by vacuum blotter (Boeckel Scientific, Feasterville, PA) using 0.6 M NaCl and 60 mM sodium citrate as a transfer solution. After transfer, membranes were washed in three changes of water and transferred to 200 ml of a 1% periodic acid (vol/vol)-3% acetic acid (vol/vol) solution for 30 min at room temperature. The membrane was then rinsed twice with 0.1% sodium metabisulfite in 1 mM HCl and placed in Schiff reagent for 6 min.

Measurement of total DNA content. Aliquots from five different CF sputum samples (275 μl each) were incubated with no treatment, 25 μl of H2O, or 25 μl of H2O containing the Trx reducing system such that final concentrations were: Trx (30 μM) + TR (0.1 μM) and NADPH (2 mM). After a 20-min incubation at 37°C, 100 μl of H2O were added to each sample, followed by centrifugation (22,000 g) for 10 min. Resulting gel and liquid phases were separated and incubated with an equal volume of digestion solution consisting of 100 mM Tris·Cl, 5 mM EDTA, 200 mM NaCl, 0.5% Tween 20, and 1 mg/ml proteinase K for 4 h at 50°C. DNA was purified from liquid and gel phases by phenol/chloroform extraction and resuspended in 100 μl of TE buffer (45 mM Tris base; 1 mM EDTA), pH 8.0. DNA concentrations were determined by Hoechst assay (17) using an F-2000 fluorometer (Hitachi, Schaumburg, IL) with an excitation wavelength of 575 nm and an emission wavelength of 555 nm. Salmon testis DNA, dissolved in TE buffer, was used to establish the standard curve.

Muc5AC Western blot analysis. CF sputum was treated for 20 min at 37°C with 25 μl of H2O alone or H2O containing the Trx reducing DTT system. Samples were separated into gel and liquid fractions with gel fractions further subjected to a 16-h guanidinium extraction. The supernatant of each sample was then cleared by centrifugation and frozen at -20°C overnight. Samples were loaded into an agarose gel, electrophoresed, and transferred to PVDF membrane according to the methodology described above. The membranes were blocked for 1 h in Tris-buffered saline with 0.5% Tween 20 (TBST) containing 5% nonfat dry milk followed by incubation with the primary antibody, mouse anti-MUC5AC human mucin [MAB2011, 1:1,000 (Chemicon, Temecula, CA) or Ab-1, 1:1,000 (Labview, Freemont, CA)] for 20 h. Nonspecific interactions were removed by three 10-min washes in TBST followed by 1-h incubation with goat anti-mouse secondary antibody (1:5,000; Southern Biotechnogy, Birmingham, AL). The blots were washed again in TBST for 1 h, and the conjugates were visualized by chemiluminescence (Amersham, Arlington Heights, IL).

Statistics. Data in the figures are presented as means ± SD, except Fig. 4, A and B, which displays SE. A linear mixed-effects modeling approach was used to analyze the effect of treatments on the liquefaction, viscoelasticity, and DNA solubility of sputum samples. We applied a Dunnett correction when comparing several treatments against a single control. All analyses were performed using SAS version 8.2 (SAS Institute, Cary, NC). Significance was defined as P < 0.05.

Fig. 4.

Effect of Trx and NADPH dose on viscoelasticity (log G*) of CF sputum in vitro. A: sputum samples were incubated with H2O or 3, 10, or 30 μM Trx + 0.1 μM TR and 2 mM NADPH for 20 min, and log G* was determined by magnetic microrheometry. B: change in log G* after incubation with H2O or 10 μM Trx, 0.1 μM TR, and 0.2, 0.6, or 2 mM NADPH. Data are presented as the difference in log G* measurements recorded before and after exposure (displayed on the y-axis). Each column represents means ± SE for 4 samples. Statistical analysis by linear mixed-effect modeling indicate that significant declines in log G* occur with increasing concentration of either Trx (P < 0.0001) or NADPH (P < 0.005).


Effect of the Trx reducing system on release of liquid from CF sputum. Due to abnormal ion transport caused by defects in the CFTR gene, airway secretions in CF patients often are desiccated. As a consequence, purulent CF sputum comprises largely a rigid and nonflowing biopolymer matrix, often referred to as gel phase, and lesser amounts of soluble, liquid phase. To assess the effect of Trx on the ratios of these two phases in sputum, we performed compaction assay measurements. The mean (± SD) percentage of CF sputum present in the liquid phase was 3.5 ± 2.9% before Trx exposure (Fig. 1A). Aliquots treated with diluent (H2O) equal to 10% of the sputum demonstrated a small, nonsignificant increase (6.2 ± 6.6%) in the proportion of sputum present in the liquid phase. In contrast, the liquid phase of CF sputum was significantly increased after treatment with the Trx reducing system (Trx + 0.1 μM TR and 2 mM NADPH). Treatment of sputum with Trx (1 μM) increased the liquid fraction of sputum to 37.8 ± 15.4%. Maximal increases in liquid fraction occurred in samples incubated with a higher Trx (30 μM) concentration (74.5 ± 15.6%). To examine the effect of NADPH, we treated additional samples from different donors with 30 μM Trx, 0.1 μM TR, and either 0, 0.2, 0.6, 1.0, or 2 mM NADPH (Fig. 1B). Samples treated with Trx and TR without NADPH had a low percentage of sputum present in the liquid phase (2.9 ± 1.3%). Aliquots treated with NADPH demonstrated a significant dose-dependent linear increase (P < 0.001) in the liquid fraction, with maximal increase occurring at 2 mM (70.6 ± 18.1%). Treatment of sputum with NADPH in the absence of Trx did not cause any increase in the proportion of sputum present in the liquid fraction (not shown).

Reproducibility of compaction assay measurements. To assess the reproducibility of compaction assay measurements, we separated into aliquots and froze sputum samples from three different CF donors. On three consecutive days, aliquots were thawed and treated with water, Trx and its reducing system, DTT, or no treatment. As shown in Fig. 2, aliquots from donor A that had been treated with no additions or diluent (H2O) had a liquid-phase fraction of <10% of their total volume. Treatment of sputum aliquots from donor A with DTT (1 or 5 mM) or Trx (30 μM) with reducing system increased the liquid fraction of these samples to >90% of the total sample volume. These percent liquid fraction values of sputum samples from donor A did not fluctuate to any appreciable degree with identical treatment upon the second and third determinations in subsequent independent experiments. The changes occurring in donor B and C samples in response to Trx or DTT exposure were less extensive than those occurring in sputum samples from donor A. For the sample from donor B, the 3-day range of variation in percentage of sputum present in the liquid state after drug treatment was: 1 mM DTT, 24–31%; 5 mM DTT, 42–57%; and 30 μM Trx, 46–51%. For sputum from donor C, the range of percent liquid values was: 1 mM DTT, 57–61%; 5 mM DTT, 77–79%; and 30 μM Trx, 62–81%. These results show that the compaction assay has sufficient intrasample reproducibility to validate its use as a method for measuring drug-induced liquefaction in a heterogeneous group of sputum samples.

Trx is a more potent sputum liquefaction agent than GSH or NAC. The effectiveness of Trx in liquefying sputum was compared with other monothiol- and dithiol-reducing agents. Initial compaction assay experiments compared the potencies of the Trx and GSH reducing systems in liquefaction of sputum in the presence of equimolar concentrations of NADPH. Compared with control (no additions), a progressive and significant increase in percent liquid fraction was observed in sputum treated with 10, 30, or 60 μM Trx (Fig. 3A). Although GSH also showed indications of a linear increase (4.0 ± 0.4% liquid fraction per 10-fold increase in concentration), Trx is significantly more efficient with its dosing by showing an increase of 16.4 ± 0.7% liquid fraction per 10-fold increase in concentration. In separate studies, the use of NAC across a range of concentrations (Fig. 3B) also was observed to be less effective than Trx in causing liquefaction of sputum.

Effect of Trx reducing system on sputum viscoelasticity. Magnetic microrheometry was performed to determine the effect of Trx and its reducing system on sputum viscosity. Measurements were performed on sputum samples before and after incubation with the Trx reducing system (Table 2) to determine the change in log G* (viscoelasticity). Trx caused a concentration-dependent decrease in sputum viscoelasticity (Fig. 4A). On the basis of results obtained from use of a linear mixed-effect model that assumed a compound symmetric covariance matrix, an intercept and slope were generated to model the relationship of ΔG* to log dose. The model fit was: Δlog G* = -0.073 - 0.15*log(TRX μM) [slope 95% confidence interval (CI) = (-0.11, -0.19); P < 0.0001]. The effect of varying concentrations of NADPH also was examined (Fig. 4B). NADPH caused a concentration-dependent decrease in sputum viscoelasticity. The model fit was: Δlog G* = -0.12 - 0.25*log(NADPH mM) [slope 95% CI = (-0.13, -0.38); P < 0.005]. These analyses indicate concentration-dependent effects of both Trx and NADPH on CF sputum viscoelasticity.

View this table:
Table 2.

Viscoelasticity data of untreated, H2O-, or thioredoxin-treated CF sputum

Effect of Trx on the solubility of sputum glycoproteins. Disulfide bonds on mucin glycoprotein polymers are potential targets for reduction by Trx. To examine the effect of Trx on glycoproteins present in sputum, we incubated samples for 20 min with diluent or Trx with its reducing system. After treatment, the resulting soluble and insoluble phases of each sample were separated and analyzed for glycoprotein content by PAS reagent staining. As shown in Fig. 5, a discrete population of high-molecular-weight glycoproteins was detected in both the soluble and gel fractions derived from sputum treated with diluent. In contrast, greater amounts of PAS-reactive glycoproteins were evident in both phases derived from sputum treated with 10 or 30 μM Trx. During the processing of these samples it was observed that the gel-phase matrix from diluent-treated samples retained a high degree of insolubility, despite overnight guanidinium treatment. This insolubility is the likely reason for the quantitative difference in amounts of glycoprotein observed in gel-phase lanes from diluent- and Trx-treated sputum. In addition to being more abundant, a substantial proportion of the glycoforms in Trx-exposed samples also exhibited greater electrophoretic mobility than those moieties present in diluent-treated samples. These findings suggest that Trx increases the solubility and reduces the size of glycoprotein polymers in sputum.

Fig. 5.

Glycoprotein mass profile of CF sputum after diluent or Trx exposure. Aliquots of sputum were incubated with H2O or Trx (10 or 30 μM) and separated into aqueous (Aq) and gel fractions by centrifugation. Each insoluble gel fraction was further treated for 14 h with guanidinium (G). Fractions were then loaded and electrophoresed in a 1% agarose (wt/vol) gel, transferred to polyvinylidene difluoride membrane, and stained with periodic acid-Schiff reagent as described in methods. Molecular weight standards are shown in lane at far right. Results are representative of 3 independent experiments.

Effect of Trx on the solubility of DNA in sputum. The presence of high amounts of extracellular DNA in CF airway secretions contributes to the excessive viscoelasticity of CF sputum. To evaluate what effect Trx has on the solubility of DNA in sputum, we exposed samples to either no additions, diluent (H2O), or Trx (30 μM) + reducing system and then separated them into gel and liquid phases. Measurement of DNA content by Hoechst assay (Fig. 6) revealed that most of the DNA present in the untreated samples was retained in the gel phase (gel, 0.94 ± 0.26 mg; liquid, 0.05 ± 0.03 mg). Diluent-treated samples demonstrated a modest increase in mean DNA content in their liquid phase (gel, 0.80 ± 0.24 mg; liquid, 0.21 ± 0.23 mg). With Trx treatment, a further shift in DNA from gel to liquid phase (gel, 0.55 ± 0.31 mg; liquid, 0.57 ± 0.37 mg) was observed.

Fig. 6.

Trx exposure increases the solubility of DNA present in CF sputum. Sputum (275 μl) was incubated with no additions, H2O, or 30 μM Trx for 20 min. Samples were separated into gel (insoluble) and liquid [soluble (Sol)] fractions, and the total amount of DNA present in each phase was determined by Hoechst assay. Shown are mean DNA content ± SD from each fraction (n = 5 experiments). *P < 0.05 vs. no treatment soluble phase.


Mucus obstruction of the airways can cause significant morbidity and mortality in patients with CF. In this report, we demonstrate that the viscoelastic properties facilitating the persistence of these secretions within airways are markedly diminished by Trx. This conclusion is supported by two lines of experimental evidence. First, compaction assay results indicate that large amounts of liquid are released from the gel matrix of CF sputum during incubation with Trx. Occurring simultaneously with this release were decreases in the volume of solid matter, indicating that the gel-forming constituents of sputum were being solubilized. This liquefaction of CF sputum could often be observed grossly in CF sputum samples during the incubation period and, therefore, is not an artifact of centrifugal disruption. The liberation of liquid by Trx might have important therapeutic implications, since restoration of water volume at airway surfaces can restore the mucociliary transportability of CF epithelium (14). Second, magnetic microrheometry measurements provide direct evidence that sputum viscoelasticty declines as a result of reduction of sputum components by Trx.

CF sputum is a non-Newtonian fluid exhibiting both liquid and solid characteristics. Polymers when present in solutions at low concentration are able to rotate freely. When polymers become concentrated or cross-linked to such a degree that their rotation is hindered, a solution has reached a transition phase called the percolation threshold (10). At the percolation threshold the solution begins to acquire characteristics of a solid, and the elastic moduli continue to increase as more cross-polymer interactions are added, until each filament in the sample is incorporated into the matrix. Biochemical analyses have revealed that mucins MUC5AC and MUC5B, secreted by cells lining the respiratory tract, are the major gel-forming polymer components of airway mucus (13, 30, 31). Cysteine domains present on these mucins contribute to polymer formation and possibly interaction with neighboring mucin chains, by disulfide bond formation (1, 2). Because disulfide bonds on proteins are the preferred substrates for Trx enzymatic activity, it was hypothesized that mucin polymers were targets for reduction during the liquefaction of sputum by Trx. Initial attempts to support this by immunochemical analysis using a MUC5AC-specific antibody were unsuccessful (data not shown). The failure of this approach may relate to the cryptic nature of the mucin epitope and/or structural alterations resulting from reduction by Trx. A second approach using PAS staining did reveal changes in the solubility of high-molecular-weight glycoforms in Trx-treated sputum. Detection of greater concentrations of glycoproteins in the liquid phase of Trx-exposed sputum was not surprising, since these samples often exhibited a more intense yellow color and had greater opacity than liquid phase derived from diluent-treated samples. The enhanced electrophoretic mobility of PAS-detectable glycoproteins in Trx-exposed sputum also suggests that these macromolecules may decrease in size during enzymatic reduction. Findings from this electrophoretic analysis are in agreement with compaction assay measurements by demonstrating that glycoprotein release into liquid phase coincides with the decrease in mass of the gel matrix during exposure to Trx. Although PAS staining has been useful for the detection of mucin molecules in previous studies, we draw no definitive conclusions regarding the site of action for Trx in sputum, since this assay lacks the capacity to distinguish specific mucin gene products.

Neutrophil lysis within the airways of diseased CF lungs results in the deposition of extracellular DNA into airway secretions (19). By noncovalent interactions, this DNA becomes entangled within mucin glycoproteins, increasing mucus gel viscoelasticity (26). In this study we found that DNA present in sputum becomes increasingly soluble following Trx treatment. A logical explanation is that Trx activity causes structural changes within the gel matrix that are sufficient to relieve entanglement interactions between DNA and the affected macromolecules. It is uncertain what the relative contribution of this increased DNA solubility has toward viscoelastic changes observed during exposure of CF sputum to Trx. Nonetheless, from a clinical standpoint, removal of DNA from the insoluble gel phase of sputum could render it more susceptible to DNase activity during such treatment in CF.

In studies comparing the sputum-liquefying abilities of other thiol-reducing agents, Trx demonstrated greater efficacy than the GSH-reducing system. Because Trx has two redox-active cysteine residues (dithiol), whereas GSH contains only one (monothiol), Trx may be more efficient in reduction of disulfide bonds on the gel-forming constituents of CF sputum. Such a difference is unlikely to occur merely on the basis of stoichiometry, because even a 10-fold increase in [GSH] to 1 mM does not approach the sputum liquefaction capacity of Trx (Fig. 3, top). With regard to nonrecycling mucolytic drugs, DTT was more effective on an equimolar basis than NAC (or Mucomyst) solutions (Figs. 2 and 3; and data not shown). Efficacy of these compounds may again be dependent on number of redox-active cysteine residues, DTT having two, NAC only one. On the basis of these compaction assay measurements, enzymatic disulfide bond reduction using proteins or compounds with dual redox-active cysteines could be a potent mucolytic strategy. Further testing with other standardized rheological methods is necessary, however, to fully evaluate the structure/activity relationship of these compounds. In addition, potential effects of such potent mucolytic agents could, at least in theory, cause excessive liquefaction of CF sputum. Such effects could have deleterious effects on clearance of airway secretions in patients with poor pulmonary function (25).

To date there have been no published reports regarding administration of Trx with its reducing system into pulmonary airways. Any discussion on the safety and efficacy of Trx as a lung therapy is therefore speculative. However, since Trx is a native protein lacking glycosylation and posttranslational modification and normally appears at low levels within extracellular space, its chronic administration could be tolerated well by the immune system. Delivery of the Trx system into airways at concentrations tested in these present studies should be attainable, since all three components are highly soluble in aqueous solutions, and recombinant proteins have previously been administered by nebulization at similar concentration in other drug studies (28). As a three-component system, Trx therapy would face significant challenges during clinical testing. Current work is now focused toward simplification of the Trx reducing system to facilitate practical application.

In summary, Trx increases the liquid fraction and diminishes the viscoelasticity of CF sputum. Increases in glycoprotein solubility occur during treatment of sputum with Trx, and this may be the mechanism for these rheological changes. The development of mucus-reducing systems that stimulate release of liquid and reduce the viscosity of airway secretions has therapeutic potential for CF.


National Jewish Medical and Research Center and C. W. White hold a pending patent on various aspects of the use of Trx for liquefaction of viscous sputum.



This study was supported by a Max and Yetta Karasik Foundation grant. R. C. Rancourt was supported by National Heart, Lung, and Blood Institute T32 training grant in pediatric pulmonary disease HL-07670.


  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


View Abstract