Materials and Methods

Animals—The dogs included in the study were a subset of those reported in a previous publication²⁸ in which tracheobronchial brushing cytologic findings in healthy dogs and dogs with chronic cough were compared. Dogs from the previous study were excluded if there was insufficient RNA available for measurement, RNA quality was inadequate, or a diagnosis of chronic bronchitis had not been made in dogs with chronic cough.

Twelve dogs (7 healthy dogs and 5 dogs with chronic bronchitis) were included in the study. Healthy dogs were obtained from the Laboratory Animal Resources of North Carolina State University, and the group comprised 3 Beagles, 3 mixed-breed dogs, and 1 Labrador Retriever. The exact age of each dog was uncertain, but none of the dogs was < 2 years old, and the 3 oldest dogs were known to be at least 5 years old. All of the healthy dogs were female (3 sexually intact and 4 of unknown reproductive status), and their weight ranged from 10 to 23 kg (mean ± SD, 16 ± 6 kg; median, 15 kg).

Dogs were considered healthy on the basis of various criteria including a medical history of no coughing or nasal discharge for at least the preceding 6 months during daily observation, results of physical examination and thoracic radiography, negative results of a heartworm antigen test, and negative results for pulmonary parasites via fecal examinations (flotation, sedimentation, and Baermann examinations). Arterial blood gas measurement was performed in 5 healthy dogs; PaO₂ in all dogs was ≥ 80 mm Hg. No airway inflammation was identified grossly during bronchoscopy or via cytologic examination of specimens of tracheobronchial brushings or washings.

The 5 dogs with chronic bronchitis were patients at the North Carolina State University Veterinary Teaching Hospital that underwent bronchoscopy for diagnostic purposes. This group comprised a Cocker Spaniel, Samoyed, Standard Poodle, Miniature Poodle, and Boxer. Dogs were 3 to 12 years old (mean ± SD, 9 ± 3 years; median, 10 years). Three of the dogs were spayed females, and 2 were castrated males. Body weight ranged from 7 to 30 kg (mean, 19 ± 8 kg; median, 21 kg).

The diagnosis of chronic bronchitis was made on the basis of a history of coughing (duration ≥ 2 months), evidence of airway inflammation detected via thoracic radiography, gross bronchoscopic findings, cytologic examination of bronchoalveolar lavage fluid or tracheobronchial brushing specimens, and an absence of evidence for a specific underlying cause for cough. In addition to the aforementioned tests, results of serologic testing for heartworm antigen were negative for all dogs and fecal flotation yielded negative results for parasites in 1 dog.

All procedures were approved by the Institutional Animal Care and Use Committee. Signed informed consent was obtained from owners of the dogs with cough.

Tracheobronchial brushing procedure—The anesthetic protocol used for all healthy dogs and most of the dogs with cough included premedication with glycopyrrolate^a (0.01 mg/kg, IM) and hydromorphone^b (0.05 mg/kg, IM), followed by administration of propofol^c to achieve anesthesia. Anesthetic protocols for clientowned dogs were adjusted to meet the needs of each patient. Bronchoscopy was performed by use of a 5.0-mm (outer diameter) flexible pediatric bronchoscope.^d Bronchial secretions were subjectively classified for quantity as either normal, mildly increased (strands), moderately increased (globs), or severely increased (occlusion of airways). The character of mucous secretions was subjectively described as clear; cloudy; opaque; or white, yellow, or green.

All brushing specimens were collected through the biopsy channel of the bronchoscope. Brushings were obtained first, followed immediately by bronchial washings or BAL. A 3.0-mm sheathed cytology brush^e was passed through the biopsy channel of the bronchoscope. The brush was extended past the end of the bronchoscope and out of the sheath, rubbed gently back and forth across the mucosal surface approximately 5 times, pulled back into the sheath, and removed from the bronchoscope. The brush was again extended out of the sheath, briskly agitated in 2 to 10 mL of sterile cell culture medium, and rinsed in sterile saline (0.9% NaCl) solution. The procedure was repeated 8 to 10 times/dog by use of the same brush, and the material from each brushing was pooled in the same vial of medium. In each dog, half of the brushing samples were collected from the distal portion of the trachea and half were collected in the mainstem and lobar bronchi. Grossly excoriated areas or sites of visible secretions were avoided during specimen collection.

A portion of each brushing specimen was used for determination of total epithelial cell count, differential cell count, and viability of epithelial cells, as previously described.²⁸ The remainder of each specimen was used for isolation of RNA.

RNA isolation from brushing specimens—Total RNA was extracted from each of the brushing samples by use of a commercially available kit according to the manufacturer's instructions.^f For all samples analyzed by use of the real-time PCR assay, total RNA was quantified via UV spectrophotometry^g; the RNA integrity was assessed via microcapillary electrophoresis.^h All samples used for mRNA quantification had an RNA integrity number ≥ 7.

Preparation of cDNA—For mRNA quantification by use of the real-time PCR assay, 100 ng of total RNA was used to make cDNA with random hexamers. The RNA was mixed with 2 μL of random hexanucleotide mixtureⁱ (0.17 μg/μL) in a total volume of 15 μL. This mixture was heated to 75°C for 5 minutes to remove secondary RNA structures and cooled on ice before addition of 5X first-strand buffer^j (5 μL), 0.1M dithiothreitol (1.25 μL), 10mM dNTPs (1.25 μL), RNase inhibitor^k (50 units), and Moloney murine leukemia virus transcriptase^l (200 units) in a 25-μL final reaction volume. The reaction was carried out at 42°C for 1 hour. Negative control samples were prepared as described without the addition of Moloney murine leukemia virus reverse transcriptase. The cDNA was stored at –20°C until used in real-time PCR reactions. The preparation of cDNA from canine gastric mucosa was performed as previously described.²⁶

Primer design—The MUC5AC-like cDNA was amplified from canine gastric mucosa by use of degenerate oligonucleotides corresponding to 2 octapeptide motifs within the C-terminal regions of MUC5AC mucins (conserved among species).²⁶ Gastric mucosa was obtained from an apparently healthy dog immediately following euthanasia for reasons other than respiratory tract disease. The resulting product was approximately 550 bp in length and was cloned into a plasmid vector (according to the manufacturer's instructions) and sequenced. Excluding the primer regions, this amplicon aligned to 4 discontinuous regions of canine chromosome 18 contig (positions 20199639 to 20199837, 20200138 to 20200316, 20200661 to 20200749, and 20200855 to 20200886 of GenBank accession No. NW876266) and differed by only 1 nucleotide. This sequence was submitted to GenBank (accession No. DQ989238). This amplicon sequence was compared with the existing sequences in GenBank by use of BLAST.²⁹ In brief, the sequence excluding the primer regions (500 bp) was compared by use of the discontiguous mega-BLAST program with the default settings.

To avoid amplification of genomic DNA during the real-time PCR procedure, forward (5′–ATTCTGAG-CAAGGTGTTTGG–3′) and reverse (5′–AGGTGAAT-GGGCACGTGTG–3′) primers were designed to span predicted intron-exon boundaries on the basis of BLAST comparison of the partial mucin mRNA sequence with the canine genome. The resulting amplification product was 199 bp.

Standards for quantification—The mucin 199-bp reverse transcription–PCR product was cloned into a plasmid vector,^m and both strands were sequenced by use of primers M13F (5′–GTAAAACGACGGCCAG–3′) and M13R (5′–CAGGAAACAGCTATGAC–3′) to confirm identity. Plasmids containing the mucin PCR fragment were quantified via UV spectrophotometry^g and were serially diluted 10-fold with concentrations ranging from 100,000 to 1 copies/μL; 5 μL of each dilution was used as a template in triplicate, according to standards for quantification.

Real-time PCR procedure—The real-time PCR process was carried out in triplicate in a thermal cyclerⁿ; 1.25 μL of cDNA was used as a template in 50-μL reactions with 25 μL of 2X SYBR green master mix° and 0.5μM each of forward and reverse primers under conditions as follows: initial 95°C for 6 minutes, followed by 45 cycles at 95°C for 15 seconds, then 64°C for 45 seconds. Melting curve analysis was initiated at 55°C, and data were captured at increasing increments of 0.5°C for 80 time points. The identity of all amplification products was confirmed with ethidium bromide staining after electrophoresis in a 2% agarose gel with appropriate size standards for all reactions, and at least 1 reaction from each dog was also confirmed as the expected target via direct sequencing.^p Standard procedures for prevention of amplicon contamination were used, including the use of aerosol-resistant pipette tips and gloves, physical and temporal separation of nucleic acid extraction, reaction setup, and post-PCR processing of samples. Positive and negative control samples consisting of cloned gene targets and water (no DNA) were used with every reaction. The predicted amino acid sequence of the partial MUC5AC-like mRNA transcript was aligned with mucin amino acid sequences from humans, pigs, horses, cows, mice, and rats.

Statistical analysis—The mean starting quantity of MUC5AC-like mRNA from healthy dogs and dogs with chronic bronchitis was compared. Mean starting quantities were also adjusted for relative epithelial cell counts, presuming a lack of mucin production by WBCs within the brushing specimens,³⁰ and compared between the 2 groups. Goblet cell numbers in tracheobronchial brushings (as a percentage of all nucleated cells and as a percentage of epithelial cells) were compared between groups. The Mann-Whitney rank sum test was used for all comparisons, and a value of P < 0.05 was considered significant. Commercially available software^q was used for analyses. A descriptive comparison between mean starting quantity of MUC5AC-like mRNA and gross assessment of airway mucus via bronchoscopy was also made.

Results

The 500-bp partial canine mucin mRNA transcript shared 99.8% identity with a predicted canine mucin mRNA (XM_540775). Fifteen of 19 total potential matches were either actual or predicted MUC5AC mRNA sequences from other species, including horses, humans, pigs, mice, rats, and cows. Of the remaining potential matches, 2 were predicted canine MUC5B mRNA sequences and 2 were unnamed clones from humans and mice, respectively.

The PCR assay amplified a 199-bp amplicon from all airway-derived cDNA samples and did not produce any amplicons when canine genomic DNA was used as template. The assay efficiently amplified and detected the cloned target over a wide dynamic range (500,000 to 5 copies/reaction) with a high correlation coefficient (r² = 0.99). Amplicons were not detected in any of the negative control samples. The mean melting temperature for the specific amplicons was 89.53 ± 0.022°C. Sequencing of these amplicons determined that they all shared 100% sequence identity with the expected target. The predicted amino acid sequence of this partial mucin mRNA transcript shared the highest degree of similarity with human, porcine, and rat MUC5AC and lower identities with MUC5B (Table 1). The degree of similarity of the predicted canine partial mucin mRNA transcript was within the range of homology found among MUC5AC proteins from other species within this region. There was less similarity to the human and murine MUC5B sequences (Table 2), although there are fewer MUC5B sequences in GenBank. On the basis of these data, the canine real-time reverse transcription PCR products will be hereafter referred to as MUC5AClike mRNA.

The mean starting quantity of MUC5AC-like mRNA from dogs with chronic bronchitis was significantly (P = 0.003) greater than that from healthy dogs (Figure 1). There was a 3.0-fold difference in median values. The difference persisted when starting quantities were adjusted for epithelial counts (P < 0.001; median difference, 3.3-fold).

View Full Size

Mean starting quantity of MUC5AC-like mRNA from 7 healthy dogs and 5 dogs with chronic bronchitis. Median value for each group is indicated by a bar. Mean starting quantity of MUC5AC-like mRNA was significantly (P = 0.003) different between groups.

Citation: American Journal of Veterinary Research 68, 4; 10.2460/ajvr.68.4.435

• Download Figure
• Download figure as PowerPoint slide

There was no difference in numbers of goblet cells between specimens from healthy dogs and dogs with chronic bronchitis, regardless of whether values were expressed as a percentage of WBCs (median value for healthy dogs, 3%; median value for dogs with chronic bronchitis, 2%) or as a percentage of epithelial cells (median value for healthy dogs, 3%; median value for dogs with chronic bronchitis, 2%). Subjective assessments of mucus made during bronchoscopy appeared to correlate with mean starting quantities of MUC5AC-like mRNA in individual dogs. When ranked from lowest to highest starting quantity, subjective assessments of mucus volume (compared with volumes expected in healthy dogs) and character among the 5 dogs with chronic bronchitis were the following: normal volume, clear; normal volume, clear; normal volume, cloudy; mildly increased volume, cloudy; and severely increased volume, green (pus).

Discussion

In the present study in dogs, the real-time PCR assay was able to specifically detect and quantify canine MUC5AC-like mRNA transcripts in samples obtained via tracheobronchial brushing. Results indicated that the assay is sensitive and specific (amplification from genomic DNA samples was not detected) and able to accurately quantify gene expression over a wide dynamic range. Dogs with chronic bronchitis had increased expression of MUC5AC-like mRNA in tracheobronchial epithelium. The degree of increased expression was of a similar magnitude as that previously reported in other species.^24-26,31,32 In humans with asthma, MUC5AC mRNA in bronchial biopsy specimens was increased 60%, compared with findings in humans without asthma.^24,31 In horses with reactive airway disease, expression was increased approximately 2-fold in first-generation airways.²⁵ In rats with induced bronchitis, MUC5AC mRNA expression was increased 3-fold.³² In a study²⁶ of ovalbumin-sensitized guinea pigs, MUC5AC mRNA expression was found to increase by 150% (mean value) following challenge with ovalbumin.

To our knowledge, the mucin sequences investigated in the present study represent the first canine MUC5AC-like mRNA sequences to be characterized. Full-length canine MUC5AC or MUC5B mRNA transcripts have not been characterized. On the basis of a computer-predicted mRNA sequence,^r our assay might amplify a putative MUC5B precursor (XM_540775). There were no computer-predicted mRNA sequences in GenBank for canine MUC5AC at the time of preparation of this report. However, at the time of the assay design (February 2006), the same regions that are now predicted to represent MUC5B precursors were predicted to encode mRNA for MUC5AC (XM_540776). These computer predictions of mRNA molecules were based on genomic DNA sequences and not actual mRNA products; hence, they should be used as guidelines. The computer predictions are revised frequently, and their names (eg, XM_540776; Canis familiaris similar to MUC5B precursor) are interim names that are likely to change as more evidence becomes available. However, it is possible that mRNA transcripts detected with the real-time PCR assay in the present study represent a combination of both subtypes. In humans, MUC5AC mRNA is detected primarily in goblet cells and MUC5B mRNA is detected primarily in submucosal glands.^20,33 However, in inflamed airways, MUC5B is also present in goblet cells.²⁰ The sampling technique of tracheobronchial brushing likely preferentially collects superficial epithelial cells. Relative expression of each gene in dogs has not been reported, to our knowledge, and there may or may not be clinical relevance to which specific gene expression is upregulated.

Ideally, the finding of increased gene expression would be correlated with a measurable increase in airway mucins. The measurement of airway mucins in sputum or lavage specimens from patients is problematic for numerous reasons, including extensive glycosylation that may mask antibody binding sites or block antibody attachment in western blot or ELISA techniques, difficulty in solubilizing specimens, and destruction of the large protein backbone by the action of proteases and glycosidases in those specimens.^18,34 Furthermore, representative sample collection from the airways is difficult because of the large surface area and complex anatomy of the bronchial tree. Dogs do not produce collectable sputum, and the mucin is present in extremely low concentrations in BAL fluid. It is difficult to control for the dilution factor introduced during BAL, and factors such as the concentrations of other proteins within a complex solution such as BAL fluid can affect concentrations measured by use of an ELISA.¹⁹

In the present study, we were unable to associate the increased gene expression with an increase in goblet cell numbers. In a related study²⁸ involving slightly larger populations of dogs, the number of goblet cells in dogs with chronic cough was not greater than that in healthy dogs. Although the stimulus for increased mucus production might include goblet cell hyperplasia, an increase in goblet cell numbers is not necessary for an increase in mucin gene expression.^23,35 However, the subjective assessment of mucus during bronchoscopy and the mean starting quantity of MUC5AC-like mRNA did appear to be related. Although the study population in the present study was small and larger numbers of dogs with chronic bronchitis and healthy dogs should be studied, there was no overlap in results between the 2 groups and differences were statistically significant.

Overall, the results of our study suggest that the ability to measure mucin gene expression from airway specimens collected from client-owned dogs during routine bronchoscopy may be useful in future studies of the pathogenesis and treatment of canine chronic bronchitis. That ability may also expand the usefulness of airway inflammation in dogs as a method for studying bronchitis in humans.