Induction of abnormal respiratory sounds by capsaicin in rats previously infected with Bordetella pertussis

Abstract

Sprague Dawley rats, previously infected with Phase-I Bordetella pertussis, developed more severe abnormal respiratory sounds than normal animals, but not coughing, when exposed to aerosolized capsaicin, one of several cough-inducing agents tested. Stethoscope examination suggested that greater production of pulmonary mucus might be occurring after capsaicin challenge of the infected animals, compared to the uninfected controls. Rats of three other strains gave characteristically different responses from the Sprague Dawleys. The administration of capsaicin to B. pertussis-infected rats may provide useful insights into the pathophysiology of excess mucus secretion in human pertussis.

Bordetella pertussis, Capsaicin, Sprague Dawley rat, Cough

1 Introduction

Experimental infections of the lung with Bordetella pertussis, the causative agent of whooping cough in man, can be established in a variety of laboratory animals but only certain species of monkey [1] and rats, especially Sprague Dawleys [2–7], develop a spontaneous paroxysmal cough. In these rats, the coughing is most reliably detected by sound-activated tape recording particularly during the hours of darkness and at days 8–14 post-infection. The coughing is specific for Phase-I B. pertussis and is not elicited by transposon mutants lacking pertussis toxin; neither is it elicited by Phase-IV B. pertussis nor by B. parapertussis[6]. Cough in the rats is accompanied by leukocytosis, another of the cardinal signs of pertussis in man. Different strains of rat responded differently to the B. pertussis challenge. Sprague Dawley rats gave more coughs after B. pertussis infection than animals of the Brown Norway, Lewis or Hooded Lister strains [7]. Immunization with pertussis toxoid protected the Sprague Dawleys to some extent against the coughing induced by bacterial pulmonary challenge, but was not as effective as whole-cell or acellular pertussis vaccines (manuscript submitted), as used in recent human field trials.

The object of the present work was to explore the possibility that a cough-inducing agent, such as capsaicin administered by aerosol, might facilitate the observation and study of coughing in the B. pertussis-infected rats. If so, it could be more convenient and reproducible than relying on the capture of spontaneous coughing on sound-activated tape recorders.

2 Materials and methods

2.1 Experimental animals

Outbred female Sprague Dawley and Hooded Lister rats and inbred Brown Norway/Ssn and Lewis rats were purchased at 5 weeks of age from Harlan Olac Ltd. (Shaws Farm, Blackthorn, Bicester, UK). The animals were barrier-reared and each batch certified as being free from a range of viral, protozoal and bacterial pathogens including Bordetella bronchiseptica. They were allowed to acclimatise for one week before use.

2.2 Infection procedure

Rats were infected with 0.1 ml containing 10⁸ CFU of B. pertussis strain 18-323 encased in fine agarose beads and instilled into the bronchial tree via a tracheostomy under Hypnorm/Hypnovel anaesthesia [5]. Normal rats of each strain were kept as controls. In previous studies on the spontaneous paroxysmal coughing due to B. pertussis infection, sham-infected animals given sterile beads were used as controls [4,5]. In all respects, however, these rats behaved essentially as normal rats and for this reason it was not considered necessary to include them in the present work. These previous studies had also shown that the spontaneous coughing in B. pertussis-infected Sprague Dawley rats occurred from day 5 onwards with a peak around days 8–14 [4,5] .

2.3 Exposure to cough-inducing agents

Capsaicin (Sigma) 100 µM was dissolved in 10% ethanol and 10% Tween 20 and diluted in saline. Histamine 10 mM, serotonin 10 mM and citric acid 0.8 M were made up in saline, ammonia 0.33 M and sodium metabisulfite 1.0 M were prepared in water.

The exposure to aerosols of capsaicin and other cough-inducing agents was done on day 9 after infection with B. pertussis. Rats in pairs, usually one infected and one control, were placed in a lidded plexiglass box (12×16×30 cm) inside a safety cabinet. The box had a 1 cm diameter hole near the lid at one end for the nozzle of the nebulizer and a similar hole at the other to allow escape of the fumes into the safety cabinet. The aerosol was administered, generally for 2 min, with a DeVilbiss glass nebulizer (The DeVilbiss Company, Somerset, PA 15501, USA) connected to an electrical air pump.

The animals were observed for their responses, particularly any abnormal respiratory sounds, for up to 10 min after the nebulizer had been switched off. The exposure was stopped if distress was apparent and the animals were taken out of the box to recover. Each rat, after aerosol exposure, was gently restrained and examined for respiratory sounds through a small-animal stethoscope held against the front of the chest. In some experiments the stethoscope examination was also done before exposure to the aerosol.

3 Results

3.1 Exposure of Sprague Dawley rats to capsaicin

Table 1 presents the results with nine infected and seven normal Sprague Dawley rats that were exposed to an aerosol of 100 µM capsaicin for 2 min. Respiratory responses to the capsaicin were variable in intensity but were much stronger in B. pertussis-infected than in normal rats. However, coughing was not observed in any of the animals. Acute respiratory distress, requiring termination of exposure, developed in one of the infected rats, while most of the other infected animals showed wheezing, squeaking, chirping or loud, coarse croaking. These sounds were not heard from control animals. The stethoscope observations paralleled the visible signs and allowed detection of abnormal respiratory sounds in two of the infected rats where no abnormal sounds were apparent to the unaided ear. The moist rales heard in the stethoscope suggested that excess mucus production had occurred in the infected animals.

1

Responses of B. pertussis-infected and normal Sprague Dawley rats to exposure in pairs to aerosols of 100 µM capsaicin for 2 min

Rat pair no.	Type of rat	Response
		Visual observation	Stethoscope
1	infected	−	+
	normal	−	−
2	infected	++++	+++
	normal	−	−
3	infected	+	+
	normal	−	−
4	infected	+	++
	normal	−	−
5	infected	++	++
	normal	−	−
6	infected	+	++
	normal	−	−
7	infected	±	++
	normal	−	−
8	infected	+	++
	infected	−	±

No obvious respiratory abnormality; ±, +, wheezing, squeaking, chirping; ++, loud, coarse croaking; +++/++++, acute respiratory distress, exposure stopped.

Normal breathing sounds; ±, slight wheeziness; +, grating, squeaking noise; ++, stronger than previous, with rales; +++, very pronounced respiratory sounds.

3.2 Exposure to other agents

The other cough-inducing agents (10 mM histamine, 10 mM serotonin, 0.8 M citric acid, 0.33 M ammonia, and 1.0 M sodium metabisulfite) all produced various visible behavioural responses, such as rapid breathing, sneezing, rubbing of the nose, salivation, urination, defaecation, and agitation often giving way to passivity with the eyes shut. None of the agents provoked coughing and none was as effective as capsaicin in producing obvious differences in respiratory responses between normal and B. pertussis-infected animals.

3.3 Exposure of Hooded Lister rats to capsaicin

When Hooded Lister rats were exposed to capsaicin, at 100 µM for up to 2 min, the responses were quite different from those seen with the Sprague Dawleys. Table 2 shows that with seven of the pairs, the exposure to capsaicin had to be stopped prematurely because of the onset of convulsions after between 50 s and 1.75 min of aerosol. In most cases it was a normal rat that had the convulsions, suggesting that infection with B. pertussis had a protective effect against the capsaicin challenge. There was rapid and complete recovery of the convulsed animals after removal from the exposure chamber. Stethoscope examination of the infected animals before exposure to capsaicin revealed that most of them had some degree of wheeziness or were producing a grating, squeaking sound in their airways. These sounds were intensified considerably after the capsaicin exposure, whereas normal animals were mostly free from such abnormalities. Nevertheless, it was the normals that showed most convulsions.

2

Responses of B. pertussis-infected and normal Hooded Lister rats to exposure in pairs to aerosols of 100 µM capsaicin for 2 min

Rat pair no.	Type of rat	Response
		Visual observation	Stethoscope
			Pre-exposure	Post-exposure
1	infected	C at 1.5 min	±	±
	normal	A	−	−
2	infected	A, +	−	++
	normal	C at 1 min	±	±
3	infected	A, +	±	++
	normal	C at 1 min, +	nd	+
4	infected	±	+	++
	normal	C at 1.75 min	+	+
5	infected	±	+	++
	normal	±	−	±
6	infected	±	+	++
	normal	C at 1.75 min	−	−
7	infected	±	+	++
	normal	C at 50 s	−	−
8	infected	C at 1 min	±	+
	normal	A	−	−

+, wheezing, squeaking, chirping; A, agitated; C, convulsion, exposure stopped.

−, normal breathing sounds; ±, slight wheeziness; +, grating, squeaking noise; ++, stronger than previous, with rales.

3.4 Comparison of other rat strains

Four pairs each of Lewis and Brown Norway rats, one infected and one normal in each pair, were tested in a similar manner, with two pairs of Sprague Dawleys as positive and negative controls. The results in Table 3 show that the two new strains were relatively unresponsive to the capsaicin challenge, as judged by visual observation, although stethoscope examination revealed various degrees of wheeziness or grating, squeaking sounds. However the pattern of the sounds did not correlate with previous B. pertussis infection, although there was a tendency for the infected animals to be less noisy than the normals. The two pairs of Sprague Dawleys included as comparators followed the pattern of Table 1, with the capsaicin producing strong abnormal respiratory sounds in the infected animals but not in the controls.

3

Responses of B. pertussis-infected, and normal, Lewis and Brown Norway rats to exposure in pairs to aerosols of 100 µM capsaicin for 2 min

Rat strain	Pair no.	Type of rat	Response
			Visual observation	Stethoscope
				Pre-exposure	Post-exposure
Lewis	1	infected	±	±	+
		normal	±	±	++
	2	infected	±	−	+
		normal	−	±	−
	3	infected	−	±	+
		normal	−	±	+
	4	infected	−	−	−
		normal	−	−	±
Brown Norway	5	infected	+	−	++
		normal	+	−	+
	6	infected	+	−	+
		normal	+	−	++
	7	infected	−	−	+
		normal	−	−	++
	8	infected	−	−	+
		normal	−	−	++
Sprague Dawley	9	infected	−	−	+++
		normal	−	−	−
	10	infected	−	−	++
		normal	−	−	−

−, no obvious respiratory abnormality; ±, +, wheezing, squeaking, chirping.

−, normal breathing sounds; ±, slight wheeziness; +, grating, squeaking noise; ++, stronger than previous, with rales; +++, very pronounced respiratory sounds.

When the four strains of rats were assessed for typical spontaneous paroxysmal coughing in response to B. pertussis infection by our usual procedure with sound-activated tape recorders overnight, only the Sprague Dawleys coughed to an appreciable extent, as reported elsewhere [7]. Hooded Lister rats gave only a few paroxysmal coughs and Brown Norway and Lewis were even less affected.

4 Discussion

Despite remarkable advances in the molecular biology of B. pertussis, the pathophysiology of the paroxysmal cough in pertussis is still unknown. In the human infant, the severe coughing does not start until relatively late after infection and may persist long after the infectious agent has been cleared from the lungs [8]. The direct involvement of one or more of the toxins of B. pertussis in cough induction seems likely, so that the prolonged paroxysmal coughing phase of the disease can be viewed as a post-infection intoxication. This would be consonant with the clinical observations that antibiotics given after the onset of the paroxysmal phase have little effect on the duration and severity of the coughing [9,10]. Pertussis toxin has been suggested to play a major, but as yet undefined, role in the paroxysmal cough [11]. Immunization of infants with pertussis toxoid gave good protection against naturally acquired disease [12] but since the toxin has a variety of aggressive activities against host defences and is also an adhesin [13] it is not clear whether the protection so given is at the earlier, infectious stage of the disease. Also there is the problem that B. parapertussis, which does not produce pertussis toxin, may on occasion be the apparent causative agent of whooping cough that is indistinguishable in signs and severity from that due to B. pertussis[14,15] .

In the coughing rat model of pertussis, the most demanding and least controllable part of the procedure is the observation of the coughing itself. The infected rats do not cough continuously or when handled but, as in human pertussis, have intermittent episodes of paroxysms with long periods of normal respiration inbetween. Only irregularly, and by chance, have we heard an infected rat coughing. Instead reliance has had to be placed on sensitive microphones suspended above the cages and attached to sound-activated tape recorders which are switched on in the early evening when the lights go off and the rats become active. The tape recorder thus captures the sounds produced by the group of typically eight rats housed in a sound-insulated booth, but does not allow observations on individual animals. The present experiments with capsaicin and other cough-inducing agents were therefore done with the intention of seeing if coughing in the individual rat could be produced on demand. In the event, none of the agents we tested induced coughing that was identical to the spontaneous coughing that we have routinely tape-recorded, and occasionally heard directly. This type of coughing is paroxysmal and ranges in volume and pitch from a chirping sound to a hoarse distinctive cough.

The present preliminary survey suggests that the combination of the Sprague Dawley rat and capsaicin may be best for provoking abnormal respiratory sounds, if not coughing, in B. pertussis-infected rats. It may be significant that Sprague Dawley rats were also the strain giving the highest level of spontaneous paroxysmal coughing in response to B. pertussis infection [7]. The wheezing, squeaking and chirping sounds induced by capsaicin were quite distinct from the aforementioned coughing and seemed to be related to excess production of mucus in the previously infected animals. This might have some relevance to human pertussis where inability to expel viscid mucus seems to be a factor in the paroxysmal cough [8,10]. The mechanisms of the capsaicin-induced convulsions in Hooded Lister rats and the apparent protective effect of B. pertussis infection are unknown but merit further study.

According to Korpáš[16], the rat is not widely used for studies on cough induction for various reasons: problems of reproducibility, insusceptibility to mechanical stimulation, and the requirement for stronger chemical stimuli than in other species. In one study, only 27% of Wistar rats coughed in response to capsaicin [17]. Other investigators, however, have found rat coughing to be a sufficiently definite phenomenon to allow the analysis of the neural pathways involved, together with receptors, agonists and anti-tussive agents [18,19]. In these studies on rats, capsaicin-induced coughing was not audible and was detected only by changes in airflow with the animals contained in a whole-body plethysmograph (J. Kamei, personal communication). Thus, our failure to detect audible capsaicin-induced coughing in uninfected rats does not conflict with previous work. So far, the use of a plethysmograph has not been applied to the B. pertussis-infected rat, but the observations reported here suggest that such studies might throw useful light on the mechanism of the paroxysmal cough in human pertussis and perhaps its control.

Acknowledgements

This work was supported by research grant K/MRS/50/C2235 from the Scottish Office Home and Health Department.