The effects of acute hydrogen peroxide exposure on respiratory cilia motility and viability

Animals

All animal procedures were conducted in accordance with the James Cook University Animal Ethics Committee (Ethics# A2783). C57/BL6 mice of mixed sex and age destined for euthanasia during routine colony maintenance were donated to this study by the Australian Institute of Tropical Health & Medicine small animal colony. Animals were housed in an air-conditioned room in racked mouse cages connected to a Smart Flow ventilation system (Tecniplast, Buguggiate, Italy). Animals had access to food (standard rodent chow) and water ad libitum. Enrichment of animal cages was achieved by including a range of bedding materials (sawdust and shredded paper), and cardboard tubes for mice to hide/sleep in. Mice were delivered weekly and were euthanised the same week using carbon dioxide asphyxiation. Euthanasia was confirmed by loss of corneal and toe pinch reflexes. All animals delivered for this study were euthanised and used for data collection.

Ciliated epithelia isolation and preparation

Sections of whole-mount ciliated trachea epithelia were harvested, treated, and imaged in L-15 media lacking phenol red (21083027; ThermoFisher Scientific, Waltham, MA, USA) supplemented with 10% FBS (16000036; ThermoFisher Scientific, Waltham, MA, USA), 100 units/ml of penicillin G sodium, and 100 μg/ml of streptomycin sulfate (15140122; ThermoFisher Scientific, Waltham, MA, USA). Mice were euthanised via CO₂ asphyxiation and trachea were immediately isolated and placed into L-15 media on ice before being prepared for video microscopy as previously described (Francis & Lo, 2013). In brief, trachea segments were cut longitudinally through the middle of the trachealis muscle then mounted with ~100 µl of L-15 media in an imaging chamber constructed using two 24 × 50 mm #1 coverslips (EPBRCS24501GP; Bio-Strategy, Hobsonville, New Zealand) sandwiching a 0.254 mm thick silicone sheet gasket (CASS-.010X36-64909; AAA Acme Rubber Co, Tempe, AZ, USA). 1–2 drops of 0.50 μm microspheres (17152-10; Polysciences, Inc., Warrington, PA, USA) were added to ~4 ml of L-15 media used to mount trachea sections for tracking of cilia generated flow.

H₂O₂ treatment

H₂O₂ doses and treatment times were based on recent CAM literature, with most recommending nebulization of 3% H₂O₂ for 10 to 15 min (Cervantes Trejo et al., 2021) (Table S1). L-15 media containing different H₂O₂ concentrations (0.1%, 0.2%, 0.5%, 1%) was freshly made before each experiment. Four trachea samples were collected from each animal and randomly assigned to either a sham treatment group, or one of the four H₂O₂ treatment concentrations. Sham treatment was L-15 media without added H₂O₂. Tissue samples were incubated for 10 min in their assigned treatment, then washed three times in fresh L-15 media before storage at 37 °C before imaging. Division of each trachea into four samples allowed tissue from the same animal to be imaged at four different timepoints (0, 30, 60, 120 min) following H₂O₂ or sham treatment.

Imaging cilia motility

Samples were imaged using a Zeiss Axiovert 200 microscope with a 63x/1.4 Oil objective (420782-9900; Zeiss, Jena, Germany), Immersol 518F Immersion Oil (444960-0000; Zeiss, Jena, Germany), and DIC microscopy as previously described (Scopulovic et al., 2022). In brief, samples were imaged at 37 °C and recordings of cilia motility were collected using a Sony Exmore CMOS sensor (EM101500A; ProSciTech, Kirwan, QLD, Australia). One second movies (AVI; uncompressed) were collected at ~300 fps for quantification of CBF, while 20 s movies (mp4; HEVC) were collected at 30 fps for quantification of motile cilia percent and cilia generated flow.

Quantifying cilia motility and cilia generated flow

The impact of H₂O₂ treatment on cilia motility was first assessed by calculating the percent of motile ciliated cells (%MC) visible in each 20 s movie field of view, this was done by dividing the number of ciliated cells with motile cilia by the total number of ciliated cells visible (both motile and non-motile).

To measure cilia beat frequency (CBF), kymographs of motile cilia were generated from each 1 s (300 fps) movie in ImageJ (FIJI 2.3.0/1.53f) (Schindelin et al., 2012) as previously described (Scopulovic et al., 2022). CBF was then quantified from the kymographs using a custom MATLAB script (version 9.9.0, R2020b; MathWorks Inc, Natick, Massachusetts, USA).

Cilia generated flow was quantified from each 20 s movie (30 fps) using the Manual Tracking plugin for ImageJ (https://imagej.nih.gov/ij/plugins/track/track.html) to track the velocity of the 0.50 μm microspheres added to the bathing media (Movie S2). Cilia generated flow was assessed by tracking microsphere movement within the bathing media via two parameters, microsphere velocity and microsphere directionality. Microsphere directionality was calculated from the velocity data using Microsoft excel by dividing net microsphere displacement over time by total distance travelled (i.e., was microsphere motion random or more directed). Microspheres moving in a straight-line display directionality ≈1; microspheres moving randomly display directionality ≈0. Thus, microsphere velocity reflected microsphere speed, while microsphere directionality reflected the linearity of microsphere movement.

Microspheres moving within empty imaging chambers were also tracked to determine Brownian motion values (Fig. S1; Movie S3). Values for Brownian motion were essential for proper interpretation of microsphere movement caused by cilia motility (or lack thereof).

Quantification of epithelial damage by fluorescent microscopy

H₂O₂ induced cytotoxicity on respiratory epithelia samples was assessed using live/dead staining, immunohistochemistry, and fluorescent microscopy. Tracheal samples were stained using a Live/Dead Fixable Violet Dead Cell Stain Kit (L34963; ThermoFisher, Waltham, MA, USA) as per manufacturer instructions 120 min after treatment (Sham or H₂O₂). In brief, tracheal samples where washed three times in PBS, then incubated for 30 min at room temperature in a 1:1,000 dilution of the blue-fluorescent reactive dye in PBS. Samples were then washed once in PBS before being fixed in 4% PFA for 15 min. After fixation samples were washed three times in PBS, permeabilized for 10 min in PBST (0.2% Triton X-100 in PBS), blocked for 1 h in PBSGS (PBS + 5% Goat Serum), then incubated for 2 h at room temperature with a mouse anti-acetylated tubulin antibody (T7451; Sigma, Kawasaki, Japan) diluted 1:500 in antibody dilution buffer (PBS + 5% Goat Serum, + 0.1% Triton X-100). Secondary fluorescent labelling was subsequently performed by incubating samples for 1 h with a goat anti-mouse FITC-conjugated antibody (115-095-003; Jackson ImmunoResearch Laboratories, West Grove, PA, USA) diluted in PBST containing a 1:1,000 dilation of TRITC-conjugated phalloidin (P1951; Sigma, Kawasaki, Japan). Trachea sections were mounted lumen side up on glass slides (7105-PPA; Livingstone, London, UK) with a drop of mounting media (F6182, Fluoroshield; Sigma, Kawasaki, Japan) under a 24 × 32 mm #1 coverslip (EPBRCS24321GP; Bio-Strategy PTY Ltd., Hobsonville, New Zealand) and 0.127 mm thick gasket cut from silicone sheet (CASS-.005X24-65908; AAA Acme Rubber Co, Tempe, AZ, USA) to prevent whole-mount samples from being crushed. Coverslip edges were sealed using clear nail polish (quick dry top coat; Revlon, Manhattan, NY, USA) and stored at 4 °C until imaged.

Fluorescently stained samples were imaged on a Zeiss LSM 710 confocal microscope using a 40x oil objective (Zeiss EC Plan-Neofluar 40x/1.30 Oil DIC M27) and ZEN black software (2.31 SP1). The Live/Dead fluorescent stain was imaged using 405 nm excitation (Laser Diode 405–30) and 494–552 nm emission spectra. FITC fluorescence was imaged using 488 nm excitation (Argon laser) and 494–552 nm emission spectra. TRITC fluorescence was imaged using 561 nm excitation (DPSS 561-10 laser) and 566–669 nm emission spectra. Samples were imaged using the line sequential scanning mode and z-stacking. Blue (Live/Dead violet) and orange (TRITC) fluorescence was imaged simultaneously followed by green (FITC) fluorescence. Image z-stack resolution was 1024/1024/~30 pixels (x/y/z), equating to pixel sizes of 0.21 µm/pixel (x/y), and 0.4–0.5 µm/pixel (z).

Respiratory epithelia cell damage was quantified from the collected fluorescent images using ImageJ (FIJI 2.3.0/1.53f) (Schindelin et al., 2012). The three colour channels in each z-stack image were collapsed using max intensity projection then combined to generate a single three colour image. The ImageJ ‘Multi-point’ tool was then used to count the total number of epithelial cells (TRITC-conjugated phalloidin), the total number of ciliated epithelial cells (FITC-labelled acetylated tubulin), the total number of dead epithelial cells (Live/Dead violet-stained cells without FITC-labelled cilia), and the total number dead ciliated epithelial cells (Live/Dead violet-stained cells with FITC-labelled cilia).

Data analysis and statistics

Multiple measurements (≥3) of each parameter (%MC, CBF, flow velocity, flow directionality, live/dead counts) were made in each tissue sample for each treatment and each timepoint; resultant averages were then compared using two-way ANOVA and post-hoc Šídák’s multiple comparisons test (Prism 9; GraphPad Software, San Diego, CA, USA). P > 0.05 was considered significant.

Cilia motility: control values following sham treatment

Sham treated tissue displayed constant well maintained cilia motility and cilia generated flow at all time points imaged (Figs. 1 and 2; Movie S1). The proportion of ciliated epithelial cells with motile cilia remained ≥97% in control samples (Figs. 1A and 2A), while CBF remained constant at ~20 Hz at all time points imaged (0–120 min) (Figs. 1B and 2B). Cilia generated flow remained constant at ~33 µm/sec (Figs. 1E and 2C), while flow directionality remained high (~0.9) (Figs. 1F and 2D) in all control samples at all time points indicating linear fluid flow across the ciliated epithelia was well maintained following sham treatment.

Acute cilia motility response to H₂O₂ treatment

The immediate effect of H₂O₂ treatment on respiratory cilia was a significant impairment in cilia motility and cilia generated flow (Fig. 1; Movie S4). All H₂O₂ treatment concentrations produced a significant decrease in %MC (Fig. 1A) from the control value of 98.0 ± 3.7% to 34.1 ± 34.1% (P < 0.0001) in 0.1% H₂O₂ treated tissue and a further significant decrease to 5.4 ± 7.3% (P < 0.0001) in 0.2% H₂O₂ treated tissue. Higher H₂O₂ treatments (0.5% & 1%) both caused an immediate cessation (P < 0.0001) in all cilia motility (Fig. 1A; Movie S4). All H₂O₂ treatments also resulted in an immediate significant impairment in CBF, from 22.3 ± 3.9 Hz in control samples to 4.9 ± 1.2 Hz (P < 0.0001) in 0.1% H₂O₂ treated tissue and 4.9 ± 0.4 Hz (P < 0.0001) in 0.2% H₂O₂ treated tissue (Fig. 1B). CBF values were only calculated for cilia that displayed movement, except for the highest H₂O₂ concentrations (0.5% & 1%) which caused a complete cessation in all cilia motility, for which CBF was set as 0.0 ± 0 Hz (Fig. 1B).

Cilia generated flow was assessed by quantifying microsphere movement across the surface of the ciliated epithelium. Figure 1 shows representative microsphere traces following sham and 1% H₂O₂ treatment respectively (Figs. 1C and 1D; Movie S2). All H₂O₂ treatments resulted in a significant decrease (P < 0.0001) in cilia generated flow velocity from a control value of 32.6 ± 8.4 µm/sec to 7.3 ± 4.1 µm/sec in 0.1% H₂O₂ treated tissue, 5.7 ± 1.1 µm/sec in 0.2% H₂O₂ treated tissue, 6.5 ± 0.9 µm/sec in 0.5% H₂O₂ treated tissue, and 5.8 ± 1.8 µm/sec in 1% H₂O₂ treated tissue (Fig. 1E). Flow velocity in all H₂O₂ treated tissues was not significantly different from each other (P > 0.05) and were also not significantly different (P > 0.05) from the values obtained for microspheres moving in the absence of ciliated tissue via Brownian motion (Red dotted line in Fig. 1E). Flow directionality was also significantly reduced (P < 0.0001) in all H₂O₂ treatment groups (Fig. 1F) from a control value of 0.89 ± 0.04 to 0.40 ± 0.21 in 0.1% H₂O₂ treated tissue, 0.20 ± 0.06 in 0.2% H₂O₂ treated tissue, 0.21 ± 0.07 in 0.5% H₂O₂ treated tissue, and 0.22 ± 0.05 in 1% H₂O₂ treated tissue. Flow directionality in all H₂O₂ treated tissues was not significantly different from each other (P > 0.05) or values obtained for microspheres moving in the absence of ciliated tissue via Brownian motion (Red dotted line in Fig. 1F). However, flow directionality following 0.1% H₂O₂ treatment showed a trend for higher directionality suggesting linear flow (albeit very slow) was maintained in a subset of these samples.

Recovery of cilia motility following H₂O₂ treatment

Recovery of cilia motility and cilia generated flow following H₂O₂ treatment was found to be treatment specific (Fig. 2; Movies S5–S7). While trending lower than sham treated animals, 30 min after the lowest H₂O₂ dose (0.1%) %MC had recovered to 80.1 ± 16.5% vs 97.2 ± 4.6% in controls and was not significantly different to the control values (P > 0.05) at the later timepoints (Fig. 2A). Conversely, while there was a graded recovery in %MC following 0.2% and 0.5% H₂O₂ treatment, %MC remained significantly lower compared to both control (P < 0.0001) and 0.1% H₂O₂ (P < 0.01) treatment groups (Fig. 2A). There was no recovery of %MC observed in airway tissues treated with the highest H₂O₂ dose (1%) at any timepoint (Fig. 2A). A total of 1% H₂O₂ treatment caused the complete cessation of all cilia motion which did not return after 120 min (Fig. 2A; Movie S7).

CBF displayed a similar recovery profile following H₂O₂ treatment. Namely, while trending lower than sham treated animals, 30 min after the lowest H₂O₂ dose (0.1%) CBF had recovered to 16.9 ± 5.9 Hz vs 19.7 ± 4.9 Hz in controls and was not significantly different from control values (P > 0.05) at 30- and 60-min post-treatment (Fig. 2B). However, after 120 min the 0.1% H₂O₂ treatment group displayed a significant drop in CBF to 14.2 ± 4.8 Hz vs 23.5 ± 4.5 Hz in the control group (P < 0.0001). There was also a partial recovery in CBF following 0.2% and 0.5% H₂O₂ treatment after 30 min (Fig. 2B). CBF remained significantly lower following 0.2% and 0.5% H₂O₂ treatment vs control values (P < 0.0001) at all timepoints, but while these CBF values trended lower they were not significantly different from each other or the 0.1% H₂O₂ (P < 0.01) treatment group (Fig. 2B). No recovery in CBF was observed in airway tissues treated with the highest H₂O₂ dose (1%) at any timepoint (Fig. 2B).

Whereas %MC and CBF displayed graded recoveries depending on H₂O₂ treatment dose, cilia generated flow velocity, as assessed by microsphere tracking, showed a significant impairment in all H₂O₂ treatment groups at all timepoints (Fig. 2C). While samples 30 min after the lowest H₂O₂ dose (0.1%) did display a small improvement in cilia generated flow velocity to 17.8 ± 4.7 µm/sec vs 32.8 ± 4.7 µm/sec in controls, this change was not significantly different from the other treatment groups (P > 0.05), or the values obtained for microspheres moving in the absence of ciliated tissue via Brownian motion (red dotted line in Fig. 2C).

While cilia generated flow velocity remained depressed, linear flow as assessed by microsphere directionality was seen to return in a graded manner across the surface of the ciliated epithelium depending on H₂O₂ treatment concentration (Fig. 2D). Thirty minutes after the lowest H₂O₂ treatment (0.1%) linear flow had recovered to 0.82 ± 0.03 which was not significantly different from the control value of 0.92 ± 0.02 (P > 0.05) at all subsequent timepoints (Fig. 2D). Linear flow (Fig. 2D) also recovered in the 0.2% and 0.5% H₂O₂ treatment groups but remained significantly lower compared to control values (P < 0.0001) following 0.2% H₂O₂ treatment, and significantly lower compared to both control (P < 0.0001) and 0.1% H₂O₂ treatment (P < 0.01) following 0.5% H₂O₂ treatment (Fig. 2D). There was no recovery of linear flow observed in airway tissues treated with the highest H₂O₂ dose (1%) at any timepoint, which was not significantly different from the values obtained for microspheres moving in the absence of ciliated tissue via Brownian motion (P > 0.05) (red dotted line in Fig. 2D).

A small number of tissues were examined 0–120 min following 2% (n = 2) and 3% (n = 2) H₂O₂ treatment and were found to give identical results as seen following 1% H₂O₂ treatment, i.e., a complete cessation of all cilia motion and cilia generated flow which did not return after 120 min (data not shown).

Viability of tracheal epithelia cells following H₂O₂ treatment

Live/dead staining was used to assess the viability of tracheal epithelia cells 120 min following sham or H₂O₂ treatments (Fig. 3). Tracheal epithelia of control samples displayed a regular cobblestone arrangement which became more disorganized in samples treated with higher H₂O₂ concentrations (Fig. 3A). Cell counting revealed that all samples displayed the same proportion of ciliated epithelia cells vs non ciliated epithelia cells, with ~60% of mouse tracheal epithelia being non-ciliated vs ~40% ciliated (Fig. 3B). A dose-response trend for elevated cell death following treatment with higher H₂O₂ concentrations was observed (Fig. 3C). While both non-ciliated and ciliated tracheal epithelia cells showed a graded increase in cell death following increased H₂O₂ concentration, ciliated epithelia cells appeared more sensitive to H₂O₂ induced toxicity than non-ciliated cells (Fig. 3C). As highlighted by 1% H₂O₂ treatment which resulted in 35.3 ± 7.0% of the ciliated epithelia cells staining dead vs 2.0 ± 0.9% in control samples (P < 0.0001), whereas only 9.3 ± 2.3% of non-ciliated epithelia cells stained dead following 1% H₂O₂ treatment vs 0.7 ± 0.4% in control samples (P > 0.05) (Fig. 3C).

Control cilia motility data

The control trachea tissue samples displayed identical cilia activity across all the time points examined (0–120 min), as characterised by ~100% of cilia imaged displaying motility, CBF being maintained at ~20 Hz, cilia generated flow being maintained at ~35 µm/sec, and minimal to no respiratory epithelial cell death after 2 h in culture. This control cilia activity is consistent with previously published data from our laboratory (Scopulovic et al., 2022). It should be noted that a large heterogeneity exists within the published literature when reporting respiratory cilia motility in control samples. CBF is the most commonly assessed parameter (often the only assessed parameter) in respiratory cilia studies, probably due to the relatively simple nature of its measurement. Respiratory cilia studies report a broad range of control CBFs, ranging between 4–25 Hz (Jing et al., 2017; Yasuda et al., 2020; Zahid et al., 2020). While animal model may influence reported CBF, it’s more likely this heterogeneity is caused by the different experimental protocols used, including different culture media, imaging modalities, and sample/temperature setups. However, while CBF is the most quantified parameter, CBF may be fairly meaningless by itself. The role of cilia in the lungs is to generate mucus flow across the surface of the ciliated epithelium to remove mucus and trapped contaminants, to accomplish this cilia need to beat in a coordinated manner using an optimal beat pattern (Bustamante-Marin & Ostrowski, 2017; Raidt et al., 2014), and previous studies have shown situations where CBF is maintained (or even elevated), but cilia generated flow is impaired due to defects in cilia beat pattern (Chilvers, Rutman & O’Callaghan, 2003; Raidt et al., 2014). Thus, the added impact offered by this study which not only assesses CBF, but also cilia generated flow.

Immediate effect of single H₂O₂ exposures on respiratory cilia motility

The current study clearly demonstrates that H₂O₂ has an immediate negative effect on respiratory cilia function, with ≥1% H₂O₂ causing an immediate cessation of all cilia motion and the complete inhibition of cilia generated flow. Lower H₂O₂ concentrations also dramatically impacted cilia activity, resulting in significant reductions in CBF even at the lowest H₂O₂ concentration assessed (0.1%). Furthermore, while some cilia activity was maintained at the lowest H₂O₂ dose, cilia generated flow as assessed by tracking microsphere movement was significantly impaired, as highlighted by a reduction in linear flow and flow velocity which was slowed to the same levels as observed during simple Brownian motion (Fig. S1).

Only a handful of previous studies have examined the effect of H₂O₂ on respiratory cilia activity (Burman & Martin, 1986; Feldman et al., 1994; Honda et al., 2014; Kakuta, Sasaki & Takishima, 1991; Kobayashi et al., 1992; Nakajima et al., 1999). One major difference between this study and past studies is that they all utilized H₂O₂ concentrations at least 10x lower than the lowest dose used in the current study, and ~300x lower than the 3% concentration recommended by CAM practitioners (Table S1). Another major difference between this study and past studies is the wide variety of animal tissues used to model the respiratory cilia, which includes airway tissue isolated from rats (Burman & Martin, 1986), guinea pigs (Kakuta, Sasaki & Takishima, 1991), sheep (Kobayashi et al., 1992), humans (Feldman et al., 1994; Honda et al., 2014), and bovines (Nakajima et al., 1999). H₂O₂ treatment times also varied greatly between studies, ranging from 10–15 min (Burman & Martin, 1986; Kakuta, Sasaki & Takishima, 1991) up to 24 h of continuous H₂O₂ treatment (Honda et al., 2014; Nakajima et al., 1999). However, while direct comparisons between this study and past studies may be difficult due to these major differences in experimental designs, all studies agree on one important point, namely, that respiratory epithelia exposure to H₂O₂ causes significant impairment to cilia motility and CBF. The power of this study is that it directly tests H₂O₂ concentrations and treatment times matching CAM recommendations, thus provides a more accurate appraisal of what may be occurring to the respiratory cilia of people following CAM advice.

Only one previous study assessed the effect of H₂O₂ on respiratory cilia motile function beyond simple CBF measurements. Honda et al. (2014) found that 24-h treatment of human bronchial cultures with 500 µM H₂O₂ caused a significant reduction in cilia generated flow as assessed by tracking migration rates of fluorescent microspheres. While the experimental design of the Honda et al. (2014) study was quite different from the current study, it does support this study finding H₂O₂ having a negative impact on cilia generated flow. Peculiarly, Honda et al. (2014) didn’t measure CBF, so comparisons between CBF and cilia generated flow in their study is not possible.

Recovery of cilia motility following H₂O₂ exposure

The current study monitored recovery of cilia motility up to 2 h following H₂O₂ treatment and found that respiratory cilia regained nearly normal motile function 30 min after treatment with the lowest H₂O₂ dose (0.1%), as assessed by recovery of %MC and CBF. Conversely, while some recovery in cilia motile function was observed after 0.2–0.5% H₂O₂ treatment (in a dose-response manner), cilia motility never recovered to pre-exposure levels, and remined impaired until the end of observation (2 h). This H₂O₂ toxicity was best highlighted following treatment with H₂O₂ concentrations ≥1% which resulted in no recovery of cilia motility at any time point assessed, with cilia remaining completely immotile until the end of the observation period. Only one previous study has attempted to assess the recovery of respiratory cilia motility following an initial H₂O₂ exposure. Kakuta, Sasaki & Takishima (1991) treated isolated guinea pig trachea rings for 15 min with 2 mM H₂O₂ and found H₂O₂ caused an initial ~60% reduction in CBF compared to controls, which then recovered to control levels 15 min after the H₂O₂ was removed. However, this recovery was dependant on the presence of a surfactant, and full recovery of CBF was not seen in samples without surfactant (Kakuta, Sasaki & Takishima, 1991). As Kakuta, Sasaki & Takishima (1991) only monitored their samples for 30 min after initial H₂O₂ exposure it’s possible that full recovery of cilia motility may have occurred in non-surfactant samples at later timepoints. This data along with the current study suggests that respiratory cilia motility can recover if the H₂O₂ exposure concentration is low, but higher H₂O₂ concentrations comparable with CAM recommendations for H₂O₂ nebulizer treatment may cause irreversible impairment to respiratory cilia motility.

Recovery of cilia generated flow following H₂O₂ exposure

The current study also assessed the recovery of cilia generated flow up to 2 h following H₂O₂ treatment, which provided one of the more interesting findings of the study. Namely, the study found that while there was recovery of linear flow across the surface of ciliated respiratory epithelia following treatment with the lowest H₂O₂ dose (0.1%), as assessed by recovery of microsphere flow directionality; cilia generated flow velocity remained significantly impaired following all H₂O₂ treatment doses, even the lowest. This observation highlights an important problem that lies within the published respiratory cilia literature. As mentioned previously, the vast majority of respiratory cilia studies only assess CBF (Burman & Martin, 1986; Feldman et al., 1994; Kakuta, Sasaki & Takishima, 1991; Kobayashi et al., 1992; Nakajima et al., 1999), probably due to the relatively simple nature of its measurement, and CBF may be fairly meaningless by itself as an assessment for mucociliary clearance.

This study is the first time that recovery of cilia generated flow following H₂O₂ treatment was assessed, and clearly shows a disconnect between CBF and cilia generated flow velocity, as highlighted in the 0.1% H₂O₂ treatment group. One explanation for this finding is that cilia coordination and/or cilia beat pattern may be impaired by H₂O₂ treatment, which significantly reduces their ability to generate flow even if CBF is maintained at normal levels. This hypothesis is supported by previous studies which have shown situations where CBF is maintained (or even elevated), but cilia generated flow is impaired due to defects in cilia beat pattern (Chilvers, Rutman & O’Callaghan, 2003; Raidt et al., 2014). However, more studies are required to determine if perturbations of cilia coordination and/or cilia beat pattern are caused by H₂O₂ treatment.

Respiratory epithelia H₂O₂ cytotoxicity

The cytotoxicity of H₂O₂ on respiratory epithelia was assessed 2 h after H₂O₂ exposure using fluorescent live/dead staining. Non-H₂O₂ exposed control tissues showed a well-organized cobblestone epithelia morphology with a ∼3:2 non-ciliated to ciliated cell distribution ratio. H₂O₂ treatment caused epithelia morphology to become noticeably disorganized, severity of this disorganization increased with H₂O₂ concentration, but no change in non-ciliated to ciliated epithelia cell distribution ratio was observed at any H₂O₂ dose suggesting no overall loss of ciliated cells or cilia. A clear dose response effect was seen with cell survival, with increased H₂O₂ concentration causing significant increases in epithelia cell death. Most importantly, ciliated epithelial cells appeared considerably more sensitive to the cytotoxic effects of H₂O₂ than non-ciliated epithelial cells, with 1% H₂O₂ treatment causing death in ~35% of ciliated epithelia cells but only in ~9% of their non-ciliated counterparts.

Three previous studies have reported that H₂O₂ is cytotoxic to tracheal epithelia (Burman & Martin, 1986; Kobayashi et al., 1992; Nakajima et al., 1999). These studies included a ⁵¹Cr Cytotoxicity Assay following a 4-h treatment of rat trachea rings with of 3 mM H₂O₂ (Burman & Martin, 1986); lactate dehydrogenase measurement following 60–90 min treatment of sheep trachea cultures with 10⁻¹⁰ to 10⁻⁴ M H₂O₂ (Kobayashi et al., 1992); and a gel electrophoresis DNA fragmentation assay following 24 h treatment of bovine trachea cultures with 100–1,000 µm H₂O₂ (Nakajima et al., 1999). The problem with these past studies is that they are all utilized non-specific cytotoxicity assays, and can’t directly determine which cells were dying, and while Nakajima et al. (1999) suggests from their TEM images that ciliated cells are more sensitive to the cytotoxic effects of H₂O₂, no quantification for this was offered (Nakajima et al., 1999). Thus, this study clearly demonstrates for the first time that ciliated respiratory epithelial cells are significantly more sensitive to H₂O₂ cytotoxicity than their non-ciliated epithelial counterparts.