Identification and properties of parietal pleural afferents in rabbits
1 Laboratoire de Physiopathologie Respiratoire (Unité Propre de Recherche de l'Enseignement Supérieur, Equipe d'Accueil 2201), Institut Jean Roche, Faculté de Médecine, Université de la Méditerranée, Marseille, France
Abstract
Although pain and dyspnoea are common symptoms in pleural diseases, there are few studies on the sensory innervation of the pleura. Using rabbits, after removal of all muscles in the intercostal space to be studied, we investigated the afferents of the internal intercostal nerve by applying to the internal thoracic wall pieces of gauze soaked in warmed (37°C), buffered saline (mechanical stimulation) or solutions containing lactic acid, inflammatory mediators or capsaicin (chemical stimulation). The afferent conduction velocity ranged from 0.5 to 14 m s–1. Most units (97%) were activated by mechanical stimulation of the pleura (local positive pressure range = 4.5–8.5 cmH2O) and we found a linear relationship between the discharge rate of afferents and the force applied to the thoracic wall. The majority of mechanosensitive units (70%) also responded to one or several chemical agents. Thus, the afferents were activated by lactic acid (49%) and/or a mixture of inflammatory mediators (50%). Local application of capsaicin elicited an initial increased or decreased background afferent activity in 57% of the afferents, a delayed decrease in firing rate being noted in some units initially activated by capsaicin. Capsaicin blocked the afferent response to a further application of inflammatory mediators but did not affect the mechanosensitive units. Thus, sensory endings connected with thin myelinated and unmyelinated fibres in the internal intercostal nerve detect the mechanical and chemical events of pleural diseases.
, 百拇医药
Introduction
Although pleuritic pain and dyspnoea are cardinal symptoms in pleural diseases and accompany both inflammation (pleural effusions) and mechanical stress (pneumothorax and pleural aspiration), there are few data in the literature on the identification and properties of the pleural afferents.
Pneumothorax is a disease of the pleura, which suppresses the normal negative pressure and often necessitates a therapeutic action involving a continuous pleural aspiration, the two situations activating mechanosensitive nerve endings. Pleural effusion is another disease, often associated with pleural fluid acidosis, due to the synthesis of lactic acid, and local release of inflammatory mediators (Sahn, 1983). The pleural mesothelial cells play a key role by secreting such different metabolites as lactic acid and inflammatory substances (Potts et al. 1978; Lo et al. 1987; Hott et al. 1994; Kroegel & Antony, 1997), which could activate chemosensitive nerve endings. Thus, sensory endings in the pleural layers should be submitted in pathological circumstances to both mechanical and chemical stimulations.
, 百拇医药
Anatomical studies reveal that the parietal costal pleura and the peripheral part of the diaphragmatic pleura are innervated by the internal intercostal nerves whereas the central portion of the diaphragmatic pleura is innervated by the phrenic nerves (Chrétien, 1983; Peng & Wang, 2004). The neurohistological study of internal intercostal nerves by Duron & Marlot (1980) showed in cats that the afferent fibres, which degenerate after the ablation of the spinal ganglia, constitute 18–24% of the total nerve population, depending on the intercostal space. These authors also found that the number of non-myelinated fibres was higher in the internal intercostal nerves than in external intercostal ones. They attributed these differences to the existence of a cutaneous contingent in the internal intercostal nerve but they also hypothesized that a certain number of these fibres may go to the pleura. Neurohistological and neurohistochemical studies have also demonstrated the existence of an autonomic innervation of both the visceral and parietal pleura of various mammals, including man (Amenta et al. 1982; Artico et al. 1998). The catecholaminergic innervation prevails in the rat parietal pleura (Artico et al. 1998), but cholinergic nerves are also present (Amenta et al. 1982). In the parietal pleura, the sensory nerve endings terminate in corpuscular endings (Golgi-Mazzoni type) and mostly in free nerve endings (Dubreuil & Beaudrimont, 1950; cited by Chrétien, 1983).
, 百拇医药
We found only one electrophysiological study on the receptive properties of afferent fibres from the pleura in rabbits (Wedekind, 1997). This study was limited to an in vitro protocol on isolated portions of mediastinal and pericardial pleura and the tendinous centre of the diaphragm, all these portions being innervated by the phrenic nerve. The author tested the response of phrenic afferents to locally applied punctuate mechanical stimulation (von Frey hairs) and chemicals represented by capsaicin and an inflammatory mixture combining the effects of bradykinin, serotonin, histamine and prostaglandin E2 (PGE2) (Kessler et al. 1992). The response to lactic acid, commonly present in pleural effusion, was not tested. Wedekind identified group III (thin myelinated) and group IV (non-myelinated) fibres which showed slowly adapting response to mechanical stimulation. Some of these afferents were activated by capsaicin (21% of the units tested), but the inflammatory mixture constituted the most efficient stimulus for 82% of the afferent fibres challenged. Wedekind concluded that ‘the rabbit pleura appears as a tissue mainly innervated by multimodal mechano- and chemosensitive afferent units’ (Wedekind, 1997). However, this study did not entirely discard the possibility that the test agents had also activated nerve endings in the diaphragm muscle itself, which are also very sensitive to the chemicals tested by Wedekind (Kaufman et al. 1982; Graham et al. 1986; Jammes et al. 1986; Hussein et al. 1990; Jammes & Balzamo, 1992; Hoheisel et al. 2004).
, 百拇医药
The aim of the present study was to identify, in internal intercostal nerves, afferent activities responding to local mechanical and/or chemical stimulation of the parietal pleura. We destroyed the muscle fibres in these intercostal spaces, leaving intact the pleural layer, in order to avoid the activation of muscle afferents by the test agents.
Methods
Animal care and general preparation
The animal experiments were performed in accordance with national guidelines and the experiments conformed with the requirements of the ethics committee of Jean Roche Institut.
, 百拇医药
Sixteen adult rabbits (body weight = 2.3–3.1 kg) were used. They were anaesthetized by an injection of ethyl-carbamate, 1 g kg–1 (Urethane, Coopération Pharmaceutique Franaise, France) in the marginal ear vein. The external jugular vein was cannulated to continue anaesthesia by subsequent injections and to perfuse the animals with saline containing adrenalin (0.3 μg kg–1 h–1), allowing maintenance of the systolic blood pressure in the range 100–130 mmHg after thoracotomy. A tracheotomy was performed and the animals were ventilated at constant volume (10 ml kg–1) and frequency (30 min–1) with a Harvard volumetric pump. They inhaled a gas mixture containing 30% oxygen. End-tidal O2 and CO2 fractions were, respectively, measured with rapid pyrolytic (Gauthier, France) and infrared gas analysers (Godart, Netherlands). A heating pad allowed maintenance of the rectal temperature in the range 37–38°C. The animals were neuromuscularly blocked at hourly intervals by I.V. injections of pancuronium bromide (Pavulon, Organon Technika, France, 0.4 mg kg–1). The left carotid artery was catheterized for measurements of blood pressure and heart rate, with an electromanometer (Gould Statham P23Db, Hato Rey, PR, USA), and for blood gas analyses (Radiometer ABL 330, Copenhagen).
, 百拇医药
Throughout and after the operative procedure, the adequacy of the level of anaesthesia was judged from the changes in blood pressure, heart rate and pupil size, the changes in these variables governing the injection of supplementary doses of ethyl-carbamate. At the end of the experiments, the rabbits were killed by an I.V. injection of a hyperosmolar potassium chloride solution.
Recordings of internal intercostal afferent activities
, 百拇医药
Figure 1 schematizes the experimental set-up used to record the afferent activities and analyse their discharge.
Distal stimulation of the nerve allowed to evoke either a compound nerve potential (A, B, and C waves correspond to the different fibres with different axonal conduction velocities) or single potentials in discriminated units having a spontaneous background discharge.
A large median sternotomy was performed from the xyphoid to the manubrium sterni to bilaterally expose the entire thoracic pleural area.
, 百拇医药
First, on one side, the thoracic muscles covering the 5th to 10th intercostal spaces, such as the pectoralis major and minor, and serratus anterior, were dissected and removed. We carefully removed the external and internal intercostal muscles in one selected space, leaving the parietal pleura intact, by using an operating microscope (x 40, OPM 11 Zeiss, Germany). A pair of small steel hooks were implanted into the upper rib making up the intercostal space, in the vicinity of the sternum. The hooks were connected to a Grass S8800 stimulator delivering rectangular pulses through an isolation unit. As described by De Troyer & Legrand (1995), the internal intercostal nerve was dissected and freed from surrounding tissues over at least 3 cm, and its proximal portion was cut at the mid-axillary level. To record activity from the intercostal afferents, the distal nerve ending was divided into several filament bundles on a 6 mm x 6 mm bakelite nerve support covered by warm paraffin oil, by using the operating microscope (x 40, OPM 11 Zeiss, Germany). In each nerve, one to three bundles were placed sequentially on a monopolar tungsten electrode. The nerve activity was referred to a nearby ground electrode, amplified (x 50 000–100 000) and filtered (30 Hz to 10 kHz) by a differential amplifier.
, http://www.100md.com
Prior to testing the response of the selected afferent units to different test agents, the compound evoked action potential was recorded in the whole nerve bundle to identify the different afferent populations with respect to their conduction velocities. Thus, the raw nerve signal was displayed on a storage oscilloscope (DSO 400, Gould, France) to average the action potentials evoked by the stimulation of the distal nerve with single shocks (0.1 or 1 ms-long rectangular pulses, supramaximal) delivered by the Grass S8800 stimulator throughout an isolation unit. We measured the conduction time and calculated the conduction velocity of the different nerve populations taking into consideration the interelectrode distance, which varied from 2.0 to 2.8 cm.
, 百拇医药
To identify selective afferent activities from multiunit recordings, the raw nerve signal was displayed on a two-channel window discriminator built in our laboratory, which allowed the simultaneous analysis of two afferent populations with positive and negative polarities. The output of these discriminators provided noise-free tracings (discriminated units) which were counted by two frequency meters at 1-s intervals (fimpulses, expressed in s–1) and then displayed on the chart recorder. The positive and negative discriminated units were counted and recorded on separate tracings. Due to the small size of the action potentials of the thin afferent fibres in each filament bundle, the window discriminators allowed us to select two to three units with the same polarity in each afferent population, i.e. four to six units with positive and negative polarities per filament bundle. In eight nerve bundles (4 animals), it was possible to record and count single action potentials which had the highest conduction velocity (10.8–14 m s–1) and thus the largest diameters (Fig. 2).
, http://www.100md.com
A, examples of single afferent activities activated by the application of a gauze soaked in warmed (37°C), buffered (pH 7.40) isotonic saline against the thoracic pleura, in front of the dissected intercostal spaces. A strain gauge allowed measurement of the force produced during the gauze application. Upper to lower traces are: discharge rate (fimpulses), discriminated unit, raw nerve activity, and force. B, the relationship between force and discharge rate. The lowest force levels (less than 5 g) correspond to simple touchs.
, 百拇医药
The discriminated units were displayed on a storage oscilloscope (DSO 400 Gould, France) to average the nerve action potentials evoked by the stimulation of the distal nerve with single shocks (0.1 or 1 ms-long rectangular pulses, supramaximal) and to measure the conduction velocity of each selected afferent unit. Figures 3A and 5A show examples of activities evoked in the discriminated units by peripheral nerve stimulation. In Figs 3, 4 and 5, fimpulses corresponded to the global discharge rate of all discriminated units with the same polarity, but in Fig. 2 single units only were counted.
, 百拇医药
Examples of afferent activities elicited by mechanical (B) or chemical stimulation with the 0.04 M lactic acid solution (C). A, conduction velocities of the three discriminated positive units (discriminated units have the same amplitude at the output of the window discriminator). Touching the thoracic pleura with a gauze soaked in warmed, buffered saline allowed us to identify the mechanosensitive unit (1) having the largest amplitude and the fastest conduction velocity (5.4 m s–1). Stimulation with lactic acid elicited activities in small units (2 and 3) having the lowest conduction velocities (0.8 and 0.7 m s–1). fimpulses corresponds to the global counting of discriminated units with positive polarity.
, 百拇医药
A, the measurement of conduction velocities in the three discriminated negative units. B, mechanical stimulation (touch) of the thoracic pleura activated unit 1 (large potential, high conduction velocity). C, the inflammatory mixture activated new units (2 and 3) whereas the discharge of the mechanosensitive unit had adapted during the first 10 s of gauze application. Then, the afferent nerve discharge displayed a rhythmic pattern in phase with the frequency of the ventilatory pump. Capsaicin applications elicited the same pattern of response (mechano- then chemostimulation) and also suppressed the excitatory effects of the inflammatory mixture. The mechanosensitivity was attenuated but persisted during capsaicin application. fimpulses corresponds to the global counting of discriminated units with negative polarity.
, 百拇医药
Protocol
The following sessions were performed for each selected filament bundle. (1) Measurements of the conduction velocities of all units in a selected bundle. (2) Measurement of the conduction velocity of discriminated units. (3) Response of discriminated units to mechanical stimulation (touch stimulus): a small piece of gauze (10 mm x 20 mm) soaked in buffered isotonic saline (NaCl 0.9%; pH 7.0) was placed for a 10-s period close to the internal thoracic wall at the level of the studied intercostal space. This also constituted a sham chemical test. In four rabbits, a strain gauge (Narco Bio-Systems, Detroit, MI, USA), linear from 0 to 30 g, was used to measure the force produced during the gauze application against the thoracic pleura. A steel hook was fixed in the upper rib making up the space, in the vicinity of the sternum, and connected to the strain gauge with a nylon cable. The force was measured perpendicularly to the thoracic wall. (4) Response to acid: a piece of gauze soaked in lactic acid solution (0.04 M; pH 4.7) was positioned for a 10-s period at the same thoracic level. In three rabbits, we also compared the response of nerve afferents to 0.02 M and 0.04 M lactic acid solutions. (5) At the end of experiment, response to the application of a piece of gauze soaked in an inflammatory mixture consisting of bradykinine–5-HT–histamine–PGE2, at 10–5M concentration (pH 6.7) (Kessler et al. 1992; Wedekind, 1997). The contact of the pleura with the inflammatory mixture was maintained for a 10-s or most often a 30-s period. In five filament bundles, we also tested the afferent response to a lower concentration (10–8M) of inflammatory mediators. (6) Response to a buffered capsaicin solution (10–5M; pH 7.0) applied for several minutes. (7) In three rabbits (5 filaments bundles), the effects of the inflammatory mixture were tested 2 min after the beginning of capsaicin application.
, 百拇医药
All the chemical solutions were stored at 37°C in a thermostatic water bath before soaking the pieces of gauze. Fifteen minutes lapsed between applying two successive test agents.
On the same hemithorax only one internal intercostal nerve (7th or 8th space) was studied due to the long lasting effects of capsaicin, which also diffused on the pleura covering the adjacent spaces.
In six rabbits, a contralateral internal intercostal nerve was also studied and the battery of test agents repeated on each filament bundle.
, 百拇医药
Statistical analysis
The baseline value of the afferent fibre activity was averaged for a 1-min period prior to applying each test agent and expressed as the mean ±S.E.M. Then, the significant changes in the afferent activity induced by each test agent were determined with respect to the corresponding averaged baseline value using analysis of variance for repeated measures followed by the Student-Newman-Keuls post hoc test to indicate the direction and magnitude of the variations between the different conditions. Data processing was carried out on absolute values of impulse frequency (fimpulses) using the software program SigmaStat (SPSS Inc., Chicago, IL, USA).
, 百拇医药
Results
In the 16 rabbits, 31 filament bundles were studied.
Conduction velocities of afferent fibres
Recordings of the compound action potentials evoked in the whole nerve trunk by distal stimulation showed different conduction velocities corresponding to large myelinated (45–88 m s–1) fibres, small myelinated fibres (mean value: 11.2 ± 1.5 m s–1) (pulse shock duration = 0.1 ms), and also unmyelinated fibres (1.2 ± 0.1 m s–1) (pulse shock duration = 1.0–1.5 ms). The large myelinated fibres probably corresponded to motor fibres and afferents from the muscle spindles and Golgi tendon organs in intercostal muscles. However, after we had destroyed the muscle layers, the maximal conduction velocity of the discriminated units, having a spontaneous background afferent discharge, was 14 m s–1 and most units had very low conduction velocities (less than 2.8 m s–1). It must be pointed out that the test agents used in our protocol of parietal pleural stimulation were unable to elicit afferent activities having rapid conduction velocities. These observations concerned the destruction of the intercostal muscles, suppressing the muscle afferents in the selected bundles.
, 百拇医药
A total of 154 discriminated units were examined.
The mean background afferent discharge of small size potentials (2–3 in each filament bundle) was 3.8 ± 0.9 s–1 (i.e. 1.2–1.9 s–1 for a single unit). Their conduction velocity was low (0.5–2.8 m s–1).
The spontaneous discharge of units having the largest potentials (maximally two units in most of bundles and one in eight of them) was less than 0.5 s–1 and their conduction velocity ranged from 5 to 14 m s–1.
, 百拇医药
Response to test agents
The response to LA was tested in 142 discriminated units, to a pure mechanical stimulus (warmed buffered saline solution) in 131 units, to the inflammatory mixture in 110 units, and to capsaicin in 114 units.
Response to pure mechanical pleural stimulation (Figs 2, 3B and 5B and C). One hundred and twenty-seven (i.e. 97%) of the 131 tested units were immediately activated when the gauze piece soaked in buffered isotonic saline were applied with differing force. Figure 2A shows that the baseline activity of single units (two examples) increased in parallel with the rise of force applied to the thoracic wall and that there was a very slow adaptation of the discharge rate throughout the 10-s period of contact. Figure 2B indicates that there is a linear relationship between the discharge rate of units and the force applied to the thoracic wall, the lowest force levels corresponding to simple touch stimulation. This was also noted in multiunit recordings (Figs 3B and 5B). When there was a prolonged contact of the gauze piece with the pleural space, i.e. in the cases of the inflammatory mixture (Figs 3B and 5C) and capsaicin (Fig. 5C), the mechanosensitive units adapted and their activation totally disappeared in less than 20 s. In some cases (Fig. 5C), the afferent nerve discharge displayed a rhythmic pattern in phase with the frequency of the ventilatory pump. This may represent a periodic mechanical stimulus of pleural afferents due to the gauze motion in phase with lung inflation. Among these mechanosensitive units, only 38/131 (29%) solely responded to touch and thus were purely mechanosensitive. However, 92/131 units (70%) were also activated by several chemical agents (lactic acid, the inflammatory mixture, and capsaicin) and were regarded as multimodal receptors. As shown in Figs 2, 3B and 5B and C, mechanosensitive units had often the largest action potential amplitude and the fastest conduction velocity (11.0 ± 1.3 m s–1). The maximal global discharge rate of multiunits in response to gauze application considerably varied among the tests (14 to 78 s–1; mean = 30.0 ± 8.0 s–1). However, the strength of the mechanical stimulation was not standardized. In four rabbits, we measured the force produced by local gauze application against the thoracic wall. The values of force elicited by touching ranged from 9 to 17 g, corresponding to a pressure of 4.5–8.5 g cm–2 (area of the gauze = 2 cm2), i.e. 4.5–8.5 cmH2O.
, http://www.100md.com
Response to lactic acid (Fig. 3C). Sixty-nine (49%) of the 142 tested units were activated after applying the gauze soaked in LA solution. The gauze was maintained close to the pleura for a 10-s period. The latency for a 20% increase in discharge rate was 27 ± 3 s, i.e. the response to LA began after the gauze was removed and persisted for a further 2–4-min period. The maximal discharge rate measured in response to 0.04 M LA was 15.8 ± 1.4 s–1 and thus lower than the maximal response to mechanical stimulation. Among the 69 units responding to LA, 15 units (21.7%) did not respond to the mechanical stimulation, and thus were considered purely chemosensitive. In six filament bundles, the response to 0.04 M and 0.02 M LA solutions was tested. In these units, the maximal discharge rate to 0.02 M and 0.04 M LA was 8.2 ± 2.0 s–1 and 16.9 ± 1.8 s–1, respectively.
, 百拇医药
Response to the inflammatory mixture (Figs 4 and 5C). Fifty-nine (50%) of 118 tested units were activated after applying the inflammatory mixture at a concentration of 10–5M. We noted that the inflammatory mixture must be maintained in contact with the pleura for 30 s to elicit any measurable activation of the afferents. The latency of the response, as well as the magnitude of the maximal activation, greatly varied among the bundles: a 20% increase in discharge rate occurred within 12–80 s and the maximal discharge rate was 28.1 ± 6.0 s–1. The recovery of the background discharge rate needed several minutes after the removal of the gauze with the inflammatory mixture. A few (8/59, i.e. 13.6%) of the afferent units activated by the inflammatory mixture did not respond to the mechanical stimulation. In five filament bundles (16 units), we also tested a solution of inflammatory mediators at a lower concentration (10–8M). This test agent was never effective in activating these units.
, 百拇医药
fimpulses corresponds to the global counting of discriminated units with positive polarity.
Effects of capsaicin on the spontaneous afferent discharge (Fig. 5C). The baseline activity of 65 (57%) of the 114 tested units was modified by capsaicin. Among them, 41/65 (63%) were activated within 36 ± 2.9 s (peak discharge rate = 58.2 ± 2.9 s–1) whereas capsaicin elicited an initial decrease in firing rate in 24/65 units (37%). In 10 of the 41 units (24.4%) activated by capsaicin, we noted a further decrease in the firing rate, which fell below the initial background discharge.
, 百拇医药
Twenty units were studied during three successive sessions: (1) activation by the inflammatory mixture; (2) activation by capsaicin after the background discharge rate had been recovered or was reduced; (3) further application of the inflammatory mixture during continuous application of capsaicin. Figure 5C shows that during capsaicin treatment, although the background discharge rate was low, a further application of the inflammatory mixture did not elicit any afferent response.
, 百拇医药
In all cases, the application of the gauze (whether soaked in the lactic acid solution, the inflammatory mixture, or capsaicin) also transiently activated mechanosensitive afferents (Figs 3C and 5C).
Discussion
Our data showed that, after the removal of the intercostal muscles and their sensory endings, the background discharge rate of the afferent activities in the internal intercostal nerve was enhanced by mechanical and/or chemical stimulation of the parietal pleura covering the intercostal space. Because the conduction velocity of these afferents ranged between 0.5 and 14 m s–1, we deduced that they belonged to the group III (thin myelinated) and IV (unmyelinated) fibres. Most of these units (97%) were immediately activated by the local application of a gauze soaked in warmed, buffered saline. However, only 29% of this nerve population was purely mechanosensitive. The other units, also responding to several chemical agents, were considered multimodal receptors. The mechanosensitive units were slowly adapting, because their discharge rate remained rather stable throughout a 10-s period of stimulation. The internal intercostal nerve also contained afferent units activated by lactic acid and inflammatory mediators. The response to lactic acid seemed to be roughly proportional to its concentration, whereas a lower concentration (10–8 compared to 10–5M) of inflammatory mediators was inefficient in activating the pleural afferents. Local application of capsaicin increased or decreased the background activity in 57% of the afferents. A delayed decrease in firing rate was also noted in some units initially activated by capsaicin. Capsaicin application blocked the response of the afferents to a further application of inflammatory mediators whereas the activation of mechanosensitive units then persisted.
, 百拇医药
Most of our data confirm the observations by Wedekind (1997) in the phrenic nerve, which innervates the diaphragmatic pleura. Indeed, we also identified mechano- and chemosensitive units and mostly multimodal receptors in the parietal pleura covering the chest wall. However, Wedekind did not examine the response to lactic acid, a chemical often present in pleural effusion, and he was not able to discard the possibility that some afferents may arise from the diaphragmatic muscle itself. Indeed, the inflammatory mediators, lactate and capsaicin tested by Wedekind are also powerful stimuli for the group III–IV afferents in the diaphragm (Graham et al. 1986; Jammes et al. 1986; Hussein et al. 1990; Jammes & Balzamo, 1992).
, 百拇医药
Mechanosensitive afferents were activated by stretching the thoracic pleura at the level of intercostal spaces in which the muscles had been removed. This surgical preparation avoided the recording of muscle afferents. Using the strain gauge to measure the displacement of the thorax during the gauze application revealed first that pleural afferents were activated by a wide range of mechanical stimulations varying from a simple touch (1 g, i.e. around 0.5 cmH2O) to a relatively high stretch of the thoracic wall (30 g corresponded to 15 cmH2O). In most of these cases, the absolute pressure developed (4.5–8.5 cmH2O) was in the range of the pleural pressure measured in rabbits breathing spontaneously with intact chest wall, and anaesthetized with the same agent (Badier et al. 1989). However, we were only able to mimic positive pressure changes in the pleural cavity. This situation solely occurs in rare circumstances, such as cough, forced expiration, pneumothorax and surgical pleuroscopy. Thus, there is a major difference between our experimental protocol and the natural condition where pleural nerve endings are normally exposed to a negative (subatmospheric) pressure. We cannot conclude that these mechanosensitive pleural afferents may be activated during spontaneous breathing movements before performing further experiments reproducing a negative pressure.
, 百拇医药
The normal pleural fluid contains a low level of lactate dehydrogenase (half that found in the serum) (Miserocchi & Agostoni, 1971; Sahn, 1983). This probably explains why lactic acidosis frequently occurs in pleural effusions. In humans, the concentration of lactic acid in pleural effusions varied from 8 to 54 mmol l–1, i.e. 0.008–0.054 M (Potts et al. 1978; Brook, 1980; Riley, 1985). Thus, in our study the 0.02 M and 0.04 M lactic acid solutions were in the range of lactic acid levels measured in pleural diseases.
, 百拇医药
The release of inflammatory mediators is documented in the experimental kaolin-induced pleurisy in rats (Hori et al. 1988). In this experimental model, the histamine concentration in pleural fluid went up to 3 x 10–6M, that of kinins to 3 x 10–9M, and of PGE2 to 6 x 10–10M. Indeed, the pleural mesothelial cells release PGE2 (Hott et al. 1994; Harada et al. 1996) and other inflammatory mediators including the interleukins (Kroegel & Antony, 1997). It was already suspected by Costa et al. (2002) that kinin generation could activate the sensory nerves innervating the pleura. In our study, only the 10–5M and not the 10–8M concentration of inflammatory mediators was effective in activating pleural afferents. Such a high concentration of inflammatory substances has been used already by Wedekind (1997) to activate pleural afferents and by Kessler et al. (1992) to stimulate cutaneous afferents. This concentration was higher than that measured in kaolin-induced pleurisy by Hori et al. (1988) (10–6–10–8M) but their experimental model of pleurisy needed 5 h to develop whereas in our study the direct application of the mixture of inflammatory substances against the thoracic pleura was limited to a 3-min period. Thus, we suppose that our protocol should necessitate a higher concentration to rapidly activate nerve afferents.
, http://www.100md.com
The effects of capsaicin on unmyelinated afferents in the vagus nerve (Coleridge et al. 1965; Kaufman et al. 1982; Solway & Leff, 1991; Marantz et al. 2001; Kollarik et al. 2003; Lee et al. 2003) and also the limb muscle (Dousset et al. 2004; Hoheisel et al. 2004) and the diaphragmatic afferents (Hussein et al. 1990) are well documented. In all the aforementioned studies capsaicin was injected into the general circulation at lower concentrations (10–7–10–8M) than in our study (10–5M). Indeed, we choose to reproduce the protocol by Wedekind (1997) who directly applied the capsaicin solution to the isolated diaphragm preparation. Capsaicin action on afferents is regarded as involving an initial excitation, which leads to a transmitter release, followed by desensitization after a prolonged exposure, the latter mostly resulting from the diffusion of the drug and its direct effect on the nerve fibres (Solway & Leff, 1991; Holzer, 1998; Blackshaw et al. 2000). Desensitization of the C-fibre pleural afferents seems to occur with the high capsaicin concentration (10–5M) used in the present study because we reported a marked depression or a suppression of the nerve response to a further chemical stimulation under capsaicin application. Based on these observations, we may suppose that a large proportion (57%) of internal intercostal afferents, which were initially either activated (41/65) or inhibited (24/65) by capsaicin and then became unresponsive to inflammatory mediators, were unmyelinated fibres. This is consistent with our observation that the majority of the discriminated units in each filament bundle (3–5/6 units) had a rather low (0.5–2.8 m s–1) conduction velocity. The treatment of rats with capsaicin resulted in an inhibition of the leucocyte migration in a model of pleurisy (Costa et al. 2002), suggesting that the activation of the sensory nerves in the pleura may contribute to a neurogenic inflammatory mechanism.
, 百拇医药
Like Wedekind (1997), we denominated the pleural afferents as ‘groups III and IV’ fibres in the limb and respiratory muscles. However, this nerve classification is specific to the somatic nervous system and our study, like the study by Wedekind, cannot discard the possibility that sensory neurones in the vegetative system may also innervate the parietal pleura, their afferent fibres going with the somatic afferents in intercostal nerves. In such a case the afferents here recorded should be A and C fibres and not III and IV ones. This hypothesis is supported by some data indicating the existence of a rich sympathetic innervation of the parietal pleura (Amenta et al. 1982; Artico et al. 1998).
, 百拇医药
These electrophysiological data have provided physiological evidence in support of the existence of pleural afferents activated by mechanical and/or chemical stimuli which are encountered in pleural diseases. It is now crucial to study the ventilatory, bronchomotor and cardiovascular responses to the activation of these pleural afferents and to explore their projections on the cortical areas involved in the sensation of thoracic pain.
References
, 百拇医药
Amenta F, Cavallotti C, Fabrizi F, Fiorenti C & Laglia G (1982). Cholinergic nerve in the parietal pleura. Acta Histochem 71, 77–81.
Artico M, Iannetti G, Trangilli Leali FM, Malinovsky L & Cavallotti C (1998). Nerve fibers-mast cells correlation in the rat parietal pleura. Respir Physiol 113, 181–188.
Badier M, Jammes Y, Romero-Colomer P & Lemerre C (1989). Tonic activity in inspiratory muscles and phrenic motoneurons by stimulation of vagal afferents. J Appl Physiol 66, 1613–1619.
, http://www.100md.com
Blackshaw LA, Page AJ & Partosoedarso ER (2000). Acute effects of capsaicin on gastrointestinal vagal afferents. Neuroscience 96, 407–416.
Brook I (1980). Measurements of lactic acid in pleural fluid. Respiration 40, 344–348.
Chrétien J (1983). Embryology and anatomy of the pleura. In Diseases of the Pleura, ed. Chrétien J & Hirsch A, pp. 1–9. Masson, Paris, New York.
Coleridge HM, Coleridge JCG & Luck JC (1965). Pulmonary afferent fibres of small diameter stimulated by capsaicin and by hyperinflation of the lungs. J Physiol 179, 248–262.
, 百拇医药
Costa SK, Moreno RA, Esquisatto LC, Juliano L, Brain SD, De Nucci G & Antunes E (2002). Role of kinins and sensory neurons in the rat pleural leukocyte migration induced by Phoneutria nigriventer spider venom. Neurosci Lett 318, 158–162.
De Troyer A & Legrand A (1995). Inhomogeneous activation of the parasternal intercostals during breathing. J Appl Physiol 79, 55–62.
Dousset E, Marqueste T, Decherchi P, Jammes Y & Grélot L (2004). Effects of neonatal capsaicin deafferentation on neuromuscular adjustements of performance, and afferent activities from adult tibialis anterior muscle during exercise. J Neurosci Res 76, 734–741.
, 百拇医药
Dubreuil G & Beaudrimont A (1950). Manuel Théorique et Pratique d'Histologie. Vigot, Paris.
Duron B & Marlot D (1980). The non-myelinated fibers of the phrenic and intercostal nerves in the cat. Z Mikrosk Anat Forsch 94, 257–268.
Graham R, Jammes Y, Delpierre S, Grimaud C & Roussos C (1986). The effects of ischemia, lactic acid and hypertonic sodium chloride on phrenic afferent discharge during spontaneous diaphragmatic contraction. Neurosci Lett 67, 257–262.
, 百拇医药
Harada Y, Hatanaka K, Kawamura M, Saito M, Ogino M, Majima M, Ohno T, Yamamoto K, Takenati Y, Yamamoto S & Katori M (1996). Role of prostaglandin H synthase-2 in protaglandin E2 formation in rat carrageenin-induced pleurisy. Prostaglandins 51, 19–33.
Hoheisel U, Reinohl J, Unger T & Mense S (2004). Acidic pH and capsaicin activate mechanosensitive group IV receptors in the rat. Pain 110, 149–157.
Holzer P (1998). Neural injury, repair, an adaptation in the gastrointestinal tract. II. The elusive action of capsaicin on the vagus nerve. Am J Physiol 275, G8–G13.
, http://www.100md.com
Hori Y, Jyoyama H, Yamada K, Takagi M, Hirose K & Katori M (1988). Time course analyses of kinins and other mediators in plasma exudation of rat kaolin-induced pleurisy. Eur J Pharmacol 152, 235–245.
Hott JW, Godbey SW & Antony VB (1994). Mesothelial cell modulation of pleural repair: thrombin stimulated mesothelial cells release prostaglandin E2. Prostaglandins Leukot Essent Fatty Acids 51, 329–335.
Hussein SNA, Magder S, Chatillon A & Roussos C (1990). Chemical activation of thin-fiber phrenic afferents: respiratory responses. J Appl Physiol 69, 1002–1011.
, 百拇医药
Jammes Y & Balzamo E (1992). Changes in afferent and efferent phrenic activity during electrically-induced diaphragm fatigue. J Appl Physiol 73, 894–902.
Jammes Y, Buchler B, Delpierre S, Rasidakis A, Grimaud C & Roussos C (1986). Phrenic afferents and their role in inspiratory control. J Appl Physiol 60, 854–860.
Kaufman MP, Iwamoto GA, Longhurst JC & Mitchell JH (1982). Effects of capsaicin and bradykinin on afferent fibers with ending in skeletal muscle. Circ Res 50, 133–139.
, 百拇医药
Kessler W, Kirschhoff C, Reeh PW & Handwerker HO (1992). Excitation of cutaneous afferent nerve endings in vitro by a combination of inflammatory mediators and conditioning effect of substance P. Exp Brain Res 91, 467–476.
Kollarik M, Dinh QT, Fischer A & Undem BJ (2003). Capsaicin-sensitive and -insensitive vagal bronchopulmonary fibres in the mouse. J Physiol 551, 869–879.
Kroegel C & Antony VB (1997). Immunobiology of pleural inflammation: potential implication in pathogenesis, diagnosis and therapy. Eur Respir J 10, 2411–2418.
, 百拇医药
Lee LY, Shuei Lin Y, Gu Q, Chung E & Ho CY (2003). Functional morphology and physiological properties of bronchopulmonary C-fiber afferents. Anat Rec 270, 17–24.
Lo TN, Saul WF & Lau SS (1987). Carrageenan-stimulated release of arachidonic acid and of lactate dehydrogenase from rat pleural cells. Biochem Pharmacol 36, 2405–2413.
Marantz MJ, Vincent SG & Fisher JT (2001). Role of vagal C-fiber afferents in the bronchomotor response to lactic acid in the newborn dog. J Appl Physiol 90, 2311–2318.
, http://www.100md.com
Miserocchi G & Agostoni E (1971). Contents of the pleural space. J Appl Physiol 30, 208–213.
Peng M & Wang N (2004). Anatomy of the pleura. In Pleural Disease, ed. Bouros D, pp. 23–43. Marcel Dekker, New York.
Potts DE, Willcow MA, Good JT, Taryle DA & Sahn SA (1978). The acidosis of low-glucose pleural effusions. Am Rev Respir Dis 117, 665–671.
Riley TV (1985). Lactic acid levels in pleural fluid from patients with bacterial pleuritis. J Clin Microbiol 21, 280–281.
, http://www.100md.com
Sahn SE (1983). Cells, proteins, and the acid-base status of pleural fluid in health and disease. In Diseases of the Pleura, ed. Chrétien J & Hirsch A, pp. 120–130. Masson, Paris, New York.
Solway J & Leff AE (1991). Sensory neuropeptides and airway function. J Appl Physiol 71, 2077–2087.
Wedekind C (1997). Receptive properties of primary afferent fibres from rabbit pleura in vitro. Somatosens Mot Res 14, 229–236., 百拇医药(Yves Jammes, Delphine Tro)
Abstract
Although pain and dyspnoea are common symptoms in pleural diseases, there are few studies on the sensory innervation of the pleura. Using rabbits, after removal of all muscles in the intercostal space to be studied, we investigated the afferents of the internal intercostal nerve by applying to the internal thoracic wall pieces of gauze soaked in warmed (37°C), buffered saline (mechanical stimulation) or solutions containing lactic acid, inflammatory mediators or capsaicin (chemical stimulation). The afferent conduction velocity ranged from 0.5 to 14 m s–1. Most units (97%) were activated by mechanical stimulation of the pleura (local positive pressure range = 4.5–8.5 cmH2O) and we found a linear relationship between the discharge rate of afferents and the force applied to the thoracic wall. The majority of mechanosensitive units (70%) also responded to one or several chemical agents. Thus, the afferents were activated by lactic acid (49%) and/or a mixture of inflammatory mediators (50%). Local application of capsaicin elicited an initial increased or decreased background afferent activity in 57% of the afferents, a delayed decrease in firing rate being noted in some units initially activated by capsaicin. Capsaicin blocked the afferent response to a further application of inflammatory mediators but did not affect the mechanosensitive units. Thus, sensory endings connected with thin myelinated and unmyelinated fibres in the internal intercostal nerve detect the mechanical and chemical events of pleural diseases.
, 百拇医药
Introduction
Although pleuritic pain and dyspnoea are cardinal symptoms in pleural diseases and accompany both inflammation (pleural effusions) and mechanical stress (pneumothorax and pleural aspiration), there are few data in the literature on the identification and properties of the pleural afferents.
Pneumothorax is a disease of the pleura, which suppresses the normal negative pressure and often necessitates a therapeutic action involving a continuous pleural aspiration, the two situations activating mechanosensitive nerve endings. Pleural effusion is another disease, often associated with pleural fluid acidosis, due to the synthesis of lactic acid, and local release of inflammatory mediators (Sahn, 1983). The pleural mesothelial cells play a key role by secreting such different metabolites as lactic acid and inflammatory substances (Potts et al. 1978; Lo et al. 1987; Hott et al. 1994; Kroegel & Antony, 1997), which could activate chemosensitive nerve endings. Thus, sensory endings in the pleural layers should be submitted in pathological circumstances to both mechanical and chemical stimulations.
, 百拇医药
Anatomical studies reveal that the parietal costal pleura and the peripheral part of the diaphragmatic pleura are innervated by the internal intercostal nerves whereas the central portion of the diaphragmatic pleura is innervated by the phrenic nerves (Chrétien, 1983; Peng & Wang, 2004). The neurohistological study of internal intercostal nerves by Duron & Marlot (1980) showed in cats that the afferent fibres, which degenerate after the ablation of the spinal ganglia, constitute 18–24% of the total nerve population, depending on the intercostal space. These authors also found that the number of non-myelinated fibres was higher in the internal intercostal nerves than in external intercostal ones. They attributed these differences to the existence of a cutaneous contingent in the internal intercostal nerve but they also hypothesized that a certain number of these fibres may go to the pleura. Neurohistological and neurohistochemical studies have also demonstrated the existence of an autonomic innervation of both the visceral and parietal pleura of various mammals, including man (Amenta et al. 1982; Artico et al. 1998). The catecholaminergic innervation prevails in the rat parietal pleura (Artico et al. 1998), but cholinergic nerves are also present (Amenta et al. 1982). In the parietal pleura, the sensory nerve endings terminate in corpuscular endings (Golgi-Mazzoni type) and mostly in free nerve endings (Dubreuil & Beaudrimont, 1950; cited by Chrétien, 1983).
, 百拇医药
We found only one electrophysiological study on the receptive properties of afferent fibres from the pleura in rabbits (Wedekind, 1997). This study was limited to an in vitro protocol on isolated portions of mediastinal and pericardial pleura and the tendinous centre of the diaphragm, all these portions being innervated by the phrenic nerve. The author tested the response of phrenic afferents to locally applied punctuate mechanical stimulation (von Frey hairs) and chemicals represented by capsaicin and an inflammatory mixture combining the effects of bradykinin, serotonin, histamine and prostaglandin E2 (PGE2) (Kessler et al. 1992). The response to lactic acid, commonly present in pleural effusion, was not tested. Wedekind identified group III (thin myelinated) and group IV (non-myelinated) fibres which showed slowly adapting response to mechanical stimulation. Some of these afferents were activated by capsaicin (21% of the units tested), but the inflammatory mixture constituted the most efficient stimulus for 82% of the afferent fibres challenged. Wedekind concluded that ‘the rabbit pleura appears as a tissue mainly innervated by multimodal mechano- and chemosensitive afferent units’ (Wedekind, 1997). However, this study did not entirely discard the possibility that the test agents had also activated nerve endings in the diaphragm muscle itself, which are also very sensitive to the chemicals tested by Wedekind (Kaufman et al. 1982; Graham et al. 1986; Jammes et al. 1986; Hussein et al. 1990; Jammes & Balzamo, 1992; Hoheisel et al. 2004).
, 百拇医药
The aim of the present study was to identify, in internal intercostal nerves, afferent activities responding to local mechanical and/or chemical stimulation of the parietal pleura. We destroyed the muscle fibres in these intercostal spaces, leaving intact the pleural layer, in order to avoid the activation of muscle afferents by the test agents.
Methods
Animal care and general preparation
The animal experiments were performed in accordance with national guidelines and the experiments conformed with the requirements of the ethics committee of Jean Roche Institut.
, 百拇医药
Sixteen adult rabbits (body weight = 2.3–3.1 kg) were used. They were anaesthetized by an injection of ethyl-carbamate, 1 g kg–1 (Urethane, Coopération Pharmaceutique Franaise, France) in the marginal ear vein. The external jugular vein was cannulated to continue anaesthesia by subsequent injections and to perfuse the animals with saline containing adrenalin (0.3 μg kg–1 h–1), allowing maintenance of the systolic blood pressure in the range 100–130 mmHg after thoracotomy. A tracheotomy was performed and the animals were ventilated at constant volume (10 ml kg–1) and frequency (30 min–1) with a Harvard volumetric pump. They inhaled a gas mixture containing 30% oxygen. End-tidal O2 and CO2 fractions were, respectively, measured with rapid pyrolytic (Gauthier, France) and infrared gas analysers (Godart, Netherlands). A heating pad allowed maintenance of the rectal temperature in the range 37–38°C. The animals were neuromuscularly blocked at hourly intervals by I.V. injections of pancuronium bromide (Pavulon, Organon Technika, France, 0.4 mg kg–1). The left carotid artery was catheterized for measurements of blood pressure and heart rate, with an electromanometer (Gould Statham P23Db, Hato Rey, PR, USA), and for blood gas analyses (Radiometer ABL 330, Copenhagen).
, 百拇医药
Throughout and after the operative procedure, the adequacy of the level of anaesthesia was judged from the changes in blood pressure, heart rate and pupil size, the changes in these variables governing the injection of supplementary doses of ethyl-carbamate. At the end of the experiments, the rabbits were killed by an I.V. injection of a hyperosmolar potassium chloride solution.
Recordings of internal intercostal afferent activities
, 百拇医药
Figure 1 schematizes the experimental set-up used to record the afferent activities and analyse their discharge.
Distal stimulation of the nerve allowed to evoke either a compound nerve potential (A, B, and C waves correspond to the different fibres with different axonal conduction velocities) or single potentials in discriminated units having a spontaneous background discharge.
A large median sternotomy was performed from the xyphoid to the manubrium sterni to bilaterally expose the entire thoracic pleural area.
, 百拇医药
First, on one side, the thoracic muscles covering the 5th to 10th intercostal spaces, such as the pectoralis major and minor, and serratus anterior, were dissected and removed. We carefully removed the external and internal intercostal muscles in one selected space, leaving the parietal pleura intact, by using an operating microscope (x 40, OPM 11 Zeiss, Germany). A pair of small steel hooks were implanted into the upper rib making up the intercostal space, in the vicinity of the sternum. The hooks were connected to a Grass S8800 stimulator delivering rectangular pulses through an isolation unit. As described by De Troyer & Legrand (1995), the internal intercostal nerve was dissected and freed from surrounding tissues over at least 3 cm, and its proximal portion was cut at the mid-axillary level. To record activity from the intercostal afferents, the distal nerve ending was divided into several filament bundles on a 6 mm x 6 mm bakelite nerve support covered by warm paraffin oil, by using the operating microscope (x 40, OPM 11 Zeiss, Germany). In each nerve, one to three bundles were placed sequentially on a monopolar tungsten electrode. The nerve activity was referred to a nearby ground electrode, amplified (x 50 000–100 000) and filtered (30 Hz to 10 kHz) by a differential amplifier.
, http://www.100md.com
Prior to testing the response of the selected afferent units to different test agents, the compound evoked action potential was recorded in the whole nerve bundle to identify the different afferent populations with respect to their conduction velocities. Thus, the raw nerve signal was displayed on a storage oscilloscope (DSO 400, Gould, France) to average the action potentials evoked by the stimulation of the distal nerve with single shocks (0.1 or 1 ms-long rectangular pulses, supramaximal) delivered by the Grass S8800 stimulator throughout an isolation unit. We measured the conduction time and calculated the conduction velocity of the different nerve populations taking into consideration the interelectrode distance, which varied from 2.0 to 2.8 cm.
, 百拇医药
To identify selective afferent activities from multiunit recordings, the raw nerve signal was displayed on a two-channel window discriminator built in our laboratory, which allowed the simultaneous analysis of two afferent populations with positive and negative polarities. The output of these discriminators provided noise-free tracings (discriminated units) which were counted by two frequency meters at 1-s intervals (fimpulses, expressed in s–1) and then displayed on the chart recorder. The positive and negative discriminated units were counted and recorded on separate tracings. Due to the small size of the action potentials of the thin afferent fibres in each filament bundle, the window discriminators allowed us to select two to three units with the same polarity in each afferent population, i.e. four to six units with positive and negative polarities per filament bundle. In eight nerve bundles (4 animals), it was possible to record and count single action potentials which had the highest conduction velocity (10.8–14 m s–1) and thus the largest diameters (Fig. 2).
, http://www.100md.com
A, examples of single afferent activities activated by the application of a gauze soaked in warmed (37°C), buffered (pH 7.40) isotonic saline against the thoracic pleura, in front of the dissected intercostal spaces. A strain gauge allowed measurement of the force produced during the gauze application. Upper to lower traces are: discharge rate (fimpulses), discriminated unit, raw nerve activity, and force. B, the relationship between force and discharge rate. The lowest force levels (less than 5 g) correspond to simple touchs.
, 百拇医药
The discriminated units were displayed on a storage oscilloscope (DSO 400 Gould, France) to average the nerve action potentials evoked by the stimulation of the distal nerve with single shocks (0.1 or 1 ms-long rectangular pulses, supramaximal) and to measure the conduction velocity of each selected afferent unit. Figures 3A and 5A show examples of activities evoked in the discriminated units by peripheral nerve stimulation. In Figs 3, 4 and 5, fimpulses corresponded to the global discharge rate of all discriminated units with the same polarity, but in Fig. 2 single units only were counted.
, 百拇医药
Examples of afferent activities elicited by mechanical (B) or chemical stimulation with the 0.04 M lactic acid solution (C). A, conduction velocities of the three discriminated positive units (discriminated units have the same amplitude at the output of the window discriminator). Touching the thoracic pleura with a gauze soaked in warmed, buffered saline allowed us to identify the mechanosensitive unit (1) having the largest amplitude and the fastest conduction velocity (5.4 m s–1). Stimulation with lactic acid elicited activities in small units (2 and 3) having the lowest conduction velocities (0.8 and 0.7 m s–1). fimpulses corresponds to the global counting of discriminated units with positive polarity.
, 百拇医药
A, the measurement of conduction velocities in the three discriminated negative units. B, mechanical stimulation (touch) of the thoracic pleura activated unit 1 (large potential, high conduction velocity). C, the inflammatory mixture activated new units (2 and 3) whereas the discharge of the mechanosensitive unit had adapted during the first 10 s of gauze application. Then, the afferent nerve discharge displayed a rhythmic pattern in phase with the frequency of the ventilatory pump. Capsaicin applications elicited the same pattern of response (mechano- then chemostimulation) and also suppressed the excitatory effects of the inflammatory mixture. The mechanosensitivity was attenuated but persisted during capsaicin application. fimpulses corresponds to the global counting of discriminated units with negative polarity.
, 百拇医药
Protocol
The following sessions were performed for each selected filament bundle. (1) Measurements of the conduction velocities of all units in a selected bundle. (2) Measurement of the conduction velocity of discriminated units. (3) Response of discriminated units to mechanical stimulation (touch stimulus): a small piece of gauze (10 mm x 20 mm) soaked in buffered isotonic saline (NaCl 0.9%; pH 7.0) was placed for a 10-s period close to the internal thoracic wall at the level of the studied intercostal space. This also constituted a sham chemical test. In four rabbits, a strain gauge (Narco Bio-Systems, Detroit, MI, USA), linear from 0 to 30 g, was used to measure the force produced during the gauze application against the thoracic pleura. A steel hook was fixed in the upper rib making up the space, in the vicinity of the sternum, and connected to the strain gauge with a nylon cable. The force was measured perpendicularly to the thoracic wall. (4) Response to acid: a piece of gauze soaked in lactic acid solution (0.04 M; pH 4.7) was positioned for a 10-s period at the same thoracic level. In three rabbits, we also compared the response of nerve afferents to 0.02 M and 0.04 M lactic acid solutions. (5) At the end of experiment, response to the application of a piece of gauze soaked in an inflammatory mixture consisting of bradykinine–5-HT–histamine–PGE2, at 10–5M concentration (pH 6.7) (Kessler et al. 1992; Wedekind, 1997). The contact of the pleura with the inflammatory mixture was maintained for a 10-s or most often a 30-s period. In five filament bundles, we also tested the afferent response to a lower concentration (10–8M) of inflammatory mediators. (6) Response to a buffered capsaicin solution (10–5M; pH 7.0) applied for several minutes. (7) In three rabbits (5 filaments bundles), the effects of the inflammatory mixture were tested 2 min after the beginning of capsaicin application.
, 百拇医药
All the chemical solutions were stored at 37°C in a thermostatic water bath before soaking the pieces of gauze. Fifteen minutes lapsed between applying two successive test agents.
On the same hemithorax only one internal intercostal nerve (7th or 8th space) was studied due to the long lasting effects of capsaicin, which also diffused on the pleura covering the adjacent spaces.
In six rabbits, a contralateral internal intercostal nerve was also studied and the battery of test agents repeated on each filament bundle.
, 百拇医药
Statistical analysis
The baseline value of the afferent fibre activity was averaged for a 1-min period prior to applying each test agent and expressed as the mean ±S.E.M. Then, the significant changes in the afferent activity induced by each test agent were determined with respect to the corresponding averaged baseline value using analysis of variance for repeated measures followed by the Student-Newman-Keuls post hoc test to indicate the direction and magnitude of the variations between the different conditions. Data processing was carried out on absolute values of impulse frequency (fimpulses) using the software program SigmaStat (SPSS Inc., Chicago, IL, USA).
, 百拇医药
Results
In the 16 rabbits, 31 filament bundles were studied.
Conduction velocities of afferent fibres
Recordings of the compound action potentials evoked in the whole nerve trunk by distal stimulation showed different conduction velocities corresponding to large myelinated (45–88 m s–1) fibres, small myelinated fibres (mean value: 11.2 ± 1.5 m s–1) (pulse shock duration = 0.1 ms), and also unmyelinated fibres (1.2 ± 0.1 m s–1) (pulse shock duration = 1.0–1.5 ms). The large myelinated fibres probably corresponded to motor fibres and afferents from the muscle spindles and Golgi tendon organs in intercostal muscles. However, after we had destroyed the muscle layers, the maximal conduction velocity of the discriminated units, having a spontaneous background afferent discharge, was 14 m s–1 and most units had very low conduction velocities (less than 2.8 m s–1). It must be pointed out that the test agents used in our protocol of parietal pleural stimulation were unable to elicit afferent activities having rapid conduction velocities. These observations concerned the destruction of the intercostal muscles, suppressing the muscle afferents in the selected bundles.
, 百拇医药
A total of 154 discriminated units were examined.
The mean background afferent discharge of small size potentials (2–3 in each filament bundle) was 3.8 ± 0.9 s–1 (i.e. 1.2–1.9 s–1 for a single unit). Their conduction velocity was low (0.5–2.8 m s–1).
The spontaneous discharge of units having the largest potentials (maximally two units in most of bundles and one in eight of them) was less than 0.5 s–1 and their conduction velocity ranged from 5 to 14 m s–1.
, 百拇医药
Response to test agents
The response to LA was tested in 142 discriminated units, to a pure mechanical stimulus (warmed buffered saline solution) in 131 units, to the inflammatory mixture in 110 units, and to capsaicin in 114 units.
Response to pure mechanical pleural stimulation (Figs 2, 3B and 5B and C). One hundred and twenty-seven (i.e. 97%) of the 131 tested units were immediately activated when the gauze piece soaked in buffered isotonic saline were applied with differing force. Figure 2A shows that the baseline activity of single units (two examples) increased in parallel with the rise of force applied to the thoracic wall and that there was a very slow adaptation of the discharge rate throughout the 10-s period of contact. Figure 2B indicates that there is a linear relationship between the discharge rate of units and the force applied to the thoracic wall, the lowest force levels corresponding to simple touch stimulation. This was also noted in multiunit recordings (Figs 3B and 5B). When there was a prolonged contact of the gauze piece with the pleural space, i.e. in the cases of the inflammatory mixture (Figs 3B and 5C) and capsaicin (Fig. 5C), the mechanosensitive units adapted and their activation totally disappeared in less than 20 s. In some cases (Fig. 5C), the afferent nerve discharge displayed a rhythmic pattern in phase with the frequency of the ventilatory pump. This may represent a periodic mechanical stimulus of pleural afferents due to the gauze motion in phase with lung inflation. Among these mechanosensitive units, only 38/131 (29%) solely responded to touch and thus were purely mechanosensitive. However, 92/131 units (70%) were also activated by several chemical agents (lactic acid, the inflammatory mixture, and capsaicin) and were regarded as multimodal receptors. As shown in Figs 2, 3B and 5B and C, mechanosensitive units had often the largest action potential amplitude and the fastest conduction velocity (11.0 ± 1.3 m s–1). The maximal global discharge rate of multiunits in response to gauze application considerably varied among the tests (14 to 78 s–1; mean = 30.0 ± 8.0 s–1). However, the strength of the mechanical stimulation was not standardized. In four rabbits, we measured the force produced by local gauze application against the thoracic wall. The values of force elicited by touching ranged from 9 to 17 g, corresponding to a pressure of 4.5–8.5 g cm–2 (area of the gauze = 2 cm2), i.e. 4.5–8.5 cmH2O.
, http://www.100md.com
Response to lactic acid (Fig. 3C). Sixty-nine (49%) of the 142 tested units were activated after applying the gauze soaked in LA solution. The gauze was maintained close to the pleura for a 10-s period. The latency for a 20% increase in discharge rate was 27 ± 3 s, i.e. the response to LA began after the gauze was removed and persisted for a further 2–4-min period. The maximal discharge rate measured in response to 0.04 M LA was 15.8 ± 1.4 s–1 and thus lower than the maximal response to mechanical stimulation. Among the 69 units responding to LA, 15 units (21.7%) did not respond to the mechanical stimulation, and thus were considered purely chemosensitive. In six filament bundles, the response to 0.04 M and 0.02 M LA solutions was tested. In these units, the maximal discharge rate to 0.02 M and 0.04 M LA was 8.2 ± 2.0 s–1 and 16.9 ± 1.8 s–1, respectively.
, 百拇医药
Response to the inflammatory mixture (Figs 4 and 5C). Fifty-nine (50%) of 118 tested units were activated after applying the inflammatory mixture at a concentration of 10–5M. We noted that the inflammatory mixture must be maintained in contact with the pleura for 30 s to elicit any measurable activation of the afferents. The latency of the response, as well as the magnitude of the maximal activation, greatly varied among the bundles: a 20% increase in discharge rate occurred within 12–80 s and the maximal discharge rate was 28.1 ± 6.0 s–1. The recovery of the background discharge rate needed several minutes after the removal of the gauze with the inflammatory mixture. A few (8/59, i.e. 13.6%) of the afferent units activated by the inflammatory mixture did not respond to the mechanical stimulation. In five filament bundles (16 units), we also tested a solution of inflammatory mediators at a lower concentration (10–8M). This test agent was never effective in activating these units.
, 百拇医药
fimpulses corresponds to the global counting of discriminated units with positive polarity.
Effects of capsaicin on the spontaneous afferent discharge (Fig. 5C). The baseline activity of 65 (57%) of the 114 tested units was modified by capsaicin. Among them, 41/65 (63%) were activated within 36 ± 2.9 s (peak discharge rate = 58.2 ± 2.9 s–1) whereas capsaicin elicited an initial decrease in firing rate in 24/65 units (37%). In 10 of the 41 units (24.4%) activated by capsaicin, we noted a further decrease in the firing rate, which fell below the initial background discharge.
, 百拇医药
Twenty units were studied during three successive sessions: (1) activation by the inflammatory mixture; (2) activation by capsaicin after the background discharge rate had been recovered or was reduced; (3) further application of the inflammatory mixture during continuous application of capsaicin. Figure 5C shows that during capsaicin treatment, although the background discharge rate was low, a further application of the inflammatory mixture did not elicit any afferent response.
, 百拇医药
In all cases, the application of the gauze (whether soaked in the lactic acid solution, the inflammatory mixture, or capsaicin) also transiently activated mechanosensitive afferents (Figs 3C and 5C).
Discussion
Our data showed that, after the removal of the intercostal muscles and their sensory endings, the background discharge rate of the afferent activities in the internal intercostal nerve was enhanced by mechanical and/or chemical stimulation of the parietal pleura covering the intercostal space. Because the conduction velocity of these afferents ranged between 0.5 and 14 m s–1, we deduced that they belonged to the group III (thin myelinated) and IV (unmyelinated) fibres. Most of these units (97%) were immediately activated by the local application of a gauze soaked in warmed, buffered saline. However, only 29% of this nerve population was purely mechanosensitive. The other units, also responding to several chemical agents, were considered multimodal receptors. The mechanosensitive units were slowly adapting, because their discharge rate remained rather stable throughout a 10-s period of stimulation. The internal intercostal nerve also contained afferent units activated by lactic acid and inflammatory mediators. The response to lactic acid seemed to be roughly proportional to its concentration, whereas a lower concentration (10–8 compared to 10–5M) of inflammatory mediators was inefficient in activating the pleural afferents. Local application of capsaicin increased or decreased the background activity in 57% of the afferents. A delayed decrease in firing rate was also noted in some units initially activated by capsaicin. Capsaicin application blocked the response of the afferents to a further application of inflammatory mediators whereas the activation of mechanosensitive units then persisted.
, 百拇医药
Most of our data confirm the observations by Wedekind (1997) in the phrenic nerve, which innervates the diaphragmatic pleura. Indeed, we also identified mechano- and chemosensitive units and mostly multimodal receptors in the parietal pleura covering the chest wall. However, Wedekind did not examine the response to lactic acid, a chemical often present in pleural effusion, and he was not able to discard the possibility that some afferents may arise from the diaphragmatic muscle itself. Indeed, the inflammatory mediators, lactate and capsaicin tested by Wedekind are also powerful stimuli for the group III–IV afferents in the diaphragm (Graham et al. 1986; Jammes et al. 1986; Hussein et al. 1990; Jammes & Balzamo, 1992).
, 百拇医药
Mechanosensitive afferents were activated by stretching the thoracic pleura at the level of intercostal spaces in which the muscles had been removed. This surgical preparation avoided the recording of muscle afferents. Using the strain gauge to measure the displacement of the thorax during the gauze application revealed first that pleural afferents were activated by a wide range of mechanical stimulations varying from a simple touch (1 g, i.e. around 0.5 cmH2O) to a relatively high stretch of the thoracic wall (30 g corresponded to 15 cmH2O). In most of these cases, the absolute pressure developed (4.5–8.5 cmH2O) was in the range of the pleural pressure measured in rabbits breathing spontaneously with intact chest wall, and anaesthetized with the same agent (Badier et al. 1989). However, we were only able to mimic positive pressure changes in the pleural cavity. This situation solely occurs in rare circumstances, such as cough, forced expiration, pneumothorax and surgical pleuroscopy. Thus, there is a major difference between our experimental protocol and the natural condition where pleural nerve endings are normally exposed to a negative (subatmospheric) pressure. We cannot conclude that these mechanosensitive pleural afferents may be activated during spontaneous breathing movements before performing further experiments reproducing a negative pressure.
, 百拇医药
The normal pleural fluid contains a low level of lactate dehydrogenase (half that found in the serum) (Miserocchi & Agostoni, 1971; Sahn, 1983). This probably explains why lactic acidosis frequently occurs in pleural effusions. In humans, the concentration of lactic acid in pleural effusions varied from 8 to 54 mmol l–1, i.e. 0.008–0.054 M (Potts et al. 1978; Brook, 1980; Riley, 1985). Thus, in our study the 0.02 M and 0.04 M lactic acid solutions were in the range of lactic acid levels measured in pleural diseases.
, 百拇医药
The release of inflammatory mediators is documented in the experimental kaolin-induced pleurisy in rats (Hori et al. 1988). In this experimental model, the histamine concentration in pleural fluid went up to 3 x 10–6M, that of kinins to 3 x 10–9M, and of PGE2 to 6 x 10–10M. Indeed, the pleural mesothelial cells release PGE2 (Hott et al. 1994; Harada et al. 1996) and other inflammatory mediators including the interleukins (Kroegel & Antony, 1997). It was already suspected by Costa et al. (2002) that kinin generation could activate the sensory nerves innervating the pleura. In our study, only the 10–5M and not the 10–8M concentration of inflammatory mediators was effective in activating pleural afferents. Such a high concentration of inflammatory substances has been used already by Wedekind (1997) to activate pleural afferents and by Kessler et al. (1992) to stimulate cutaneous afferents. This concentration was higher than that measured in kaolin-induced pleurisy by Hori et al. (1988) (10–6–10–8M) but their experimental model of pleurisy needed 5 h to develop whereas in our study the direct application of the mixture of inflammatory substances against the thoracic pleura was limited to a 3-min period. Thus, we suppose that our protocol should necessitate a higher concentration to rapidly activate nerve afferents.
, http://www.100md.com
The effects of capsaicin on unmyelinated afferents in the vagus nerve (Coleridge et al. 1965; Kaufman et al. 1982; Solway & Leff, 1991; Marantz et al. 2001; Kollarik et al. 2003; Lee et al. 2003) and also the limb muscle (Dousset et al. 2004; Hoheisel et al. 2004) and the diaphragmatic afferents (Hussein et al. 1990) are well documented. In all the aforementioned studies capsaicin was injected into the general circulation at lower concentrations (10–7–10–8M) than in our study (10–5M). Indeed, we choose to reproduce the protocol by Wedekind (1997) who directly applied the capsaicin solution to the isolated diaphragm preparation. Capsaicin action on afferents is regarded as involving an initial excitation, which leads to a transmitter release, followed by desensitization after a prolonged exposure, the latter mostly resulting from the diffusion of the drug and its direct effect on the nerve fibres (Solway & Leff, 1991; Holzer, 1998; Blackshaw et al. 2000). Desensitization of the C-fibre pleural afferents seems to occur with the high capsaicin concentration (10–5M) used in the present study because we reported a marked depression or a suppression of the nerve response to a further chemical stimulation under capsaicin application. Based on these observations, we may suppose that a large proportion (57%) of internal intercostal afferents, which were initially either activated (41/65) or inhibited (24/65) by capsaicin and then became unresponsive to inflammatory mediators, were unmyelinated fibres. This is consistent with our observation that the majority of the discriminated units in each filament bundle (3–5/6 units) had a rather low (0.5–2.8 m s–1) conduction velocity. The treatment of rats with capsaicin resulted in an inhibition of the leucocyte migration in a model of pleurisy (Costa et al. 2002), suggesting that the activation of the sensory nerves in the pleura may contribute to a neurogenic inflammatory mechanism.
, 百拇医药
Like Wedekind (1997), we denominated the pleural afferents as ‘groups III and IV’ fibres in the limb and respiratory muscles. However, this nerve classification is specific to the somatic nervous system and our study, like the study by Wedekind, cannot discard the possibility that sensory neurones in the vegetative system may also innervate the parietal pleura, their afferent fibres going with the somatic afferents in intercostal nerves. In such a case the afferents here recorded should be A and C fibres and not III and IV ones. This hypothesis is supported by some data indicating the existence of a rich sympathetic innervation of the parietal pleura (Amenta et al. 1982; Artico et al. 1998).
, 百拇医药
These electrophysiological data have provided physiological evidence in support of the existence of pleural afferents activated by mechanical and/or chemical stimuli which are encountered in pleural diseases. It is now crucial to study the ventilatory, bronchomotor and cardiovascular responses to the activation of these pleural afferents and to explore their projections on the cortical areas involved in the sensation of thoracic pain.
References
, 百拇医药
Amenta F, Cavallotti C, Fabrizi F, Fiorenti C & Laglia G (1982). Cholinergic nerve in the parietal pleura. Acta Histochem 71, 77–81.
Artico M, Iannetti G, Trangilli Leali FM, Malinovsky L & Cavallotti C (1998). Nerve fibers-mast cells correlation in the rat parietal pleura. Respir Physiol 113, 181–188.
Badier M, Jammes Y, Romero-Colomer P & Lemerre C (1989). Tonic activity in inspiratory muscles and phrenic motoneurons by stimulation of vagal afferents. J Appl Physiol 66, 1613–1619.
, http://www.100md.com
Blackshaw LA, Page AJ & Partosoedarso ER (2000). Acute effects of capsaicin on gastrointestinal vagal afferents. Neuroscience 96, 407–416.
Brook I (1980). Measurements of lactic acid in pleural fluid. Respiration 40, 344–348.
Chrétien J (1983). Embryology and anatomy of the pleura. In Diseases of the Pleura, ed. Chrétien J & Hirsch A, pp. 1–9. Masson, Paris, New York.
Coleridge HM, Coleridge JCG & Luck JC (1965). Pulmonary afferent fibres of small diameter stimulated by capsaicin and by hyperinflation of the lungs. J Physiol 179, 248–262.
, 百拇医药
Costa SK, Moreno RA, Esquisatto LC, Juliano L, Brain SD, De Nucci G & Antunes E (2002). Role of kinins and sensory neurons in the rat pleural leukocyte migration induced by Phoneutria nigriventer spider venom. Neurosci Lett 318, 158–162.
De Troyer A & Legrand A (1995). Inhomogeneous activation of the parasternal intercostals during breathing. J Appl Physiol 79, 55–62.
Dousset E, Marqueste T, Decherchi P, Jammes Y & Grélot L (2004). Effects of neonatal capsaicin deafferentation on neuromuscular adjustements of performance, and afferent activities from adult tibialis anterior muscle during exercise. J Neurosci Res 76, 734–741.
, 百拇医药
Dubreuil G & Beaudrimont A (1950). Manuel Théorique et Pratique d'Histologie. Vigot, Paris.
Duron B & Marlot D (1980). The non-myelinated fibers of the phrenic and intercostal nerves in the cat. Z Mikrosk Anat Forsch 94, 257–268.
Graham R, Jammes Y, Delpierre S, Grimaud C & Roussos C (1986). The effects of ischemia, lactic acid and hypertonic sodium chloride on phrenic afferent discharge during spontaneous diaphragmatic contraction. Neurosci Lett 67, 257–262.
, 百拇医药
Harada Y, Hatanaka K, Kawamura M, Saito M, Ogino M, Majima M, Ohno T, Yamamoto K, Takenati Y, Yamamoto S & Katori M (1996). Role of prostaglandin H synthase-2 in protaglandin E2 formation in rat carrageenin-induced pleurisy. Prostaglandins 51, 19–33.
Hoheisel U, Reinohl J, Unger T & Mense S (2004). Acidic pH and capsaicin activate mechanosensitive group IV receptors in the rat. Pain 110, 149–157.
Holzer P (1998). Neural injury, repair, an adaptation in the gastrointestinal tract. II. The elusive action of capsaicin on the vagus nerve. Am J Physiol 275, G8–G13.
, http://www.100md.com
Hori Y, Jyoyama H, Yamada K, Takagi M, Hirose K & Katori M (1988). Time course analyses of kinins and other mediators in plasma exudation of rat kaolin-induced pleurisy. Eur J Pharmacol 152, 235–245.
Hott JW, Godbey SW & Antony VB (1994). Mesothelial cell modulation of pleural repair: thrombin stimulated mesothelial cells release prostaglandin E2. Prostaglandins Leukot Essent Fatty Acids 51, 329–335.
Hussein SNA, Magder S, Chatillon A & Roussos C (1990). Chemical activation of thin-fiber phrenic afferents: respiratory responses. J Appl Physiol 69, 1002–1011.
, 百拇医药
Jammes Y & Balzamo E (1992). Changes in afferent and efferent phrenic activity during electrically-induced diaphragm fatigue. J Appl Physiol 73, 894–902.
Jammes Y, Buchler B, Delpierre S, Rasidakis A, Grimaud C & Roussos C (1986). Phrenic afferents and their role in inspiratory control. J Appl Physiol 60, 854–860.
Kaufman MP, Iwamoto GA, Longhurst JC & Mitchell JH (1982). Effects of capsaicin and bradykinin on afferent fibers with ending in skeletal muscle. Circ Res 50, 133–139.
, 百拇医药
Kessler W, Kirschhoff C, Reeh PW & Handwerker HO (1992). Excitation of cutaneous afferent nerve endings in vitro by a combination of inflammatory mediators and conditioning effect of substance P. Exp Brain Res 91, 467–476.
Kollarik M, Dinh QT, Fischer A & Undem BJ (2003). Capsaicin-sensitive and -insensitive vagal bronchopulmonary fibres in the mouse. J Physiol 551, 869–879.
Kroegel C & Antony VB (1997). Immunobiology of pleural inflammation: potential implication in pathogenesis, diagnosis and therapy. Eur Respir J 10, 2411–2418.
, 百拇医药
Lee LY, Shuei Lin Y, Gu Q, Chung E & Ho CY (2003). Functional morphology and physiological properties of bronchopulmonary C-fiber afferents. Anat Rec 270, 17–24.
Lo TN, Saul WF & Lau SS (1987). Carrageenan-stimulated release of arachidonic acid and of lactate dehydrogenase from rat pleural cells. Biochem Pharmacol 36, 2405–2413.
Marantz MJ, Vincent SG & Fisher JT (2001). Role of vagal C-fiber afferents in the bronchomotor response to lactic acid in the newborn dog. J Appl Physiol 90, 2311–2318.
, http://www.100md.com
Miserocchi G & Agostoni E (1971). Contents of the pleural space. J Appl Physiol 30, 208–213.
Peng M & Wang N (2004). Anatomy of the pleura. In Pleural Disease, ed. Bouros D, pp. 23–43. Marcel Dekker, New York.
Potts DE, Willcow MA, Good JT, Taryle DA & Sahn SA (1978). The acidosis of low-glucose pleural effusions. Am Rev Respir Dis 117, 665–671.
Riley TV (1985). Lactic acid levels in pleural fluid from patients with bacterial pleuritis. J Clin Microbiol 21, 280–281.
, http://www.100md.com
Sahn SE (1983). Cells, proteins, and the acid-base status of pleural fluid in health and disease. In Diseases of the Pleura, ed. Chrétien J & Hirsch A, pp. 120–130. Masson, Paris, New York.
Solway J & Leff AE (1991). Sensory neuropeptides and airway function. J Appl Physiol 71, 2077–2087.
Wedekind C (1997). Receptive properties of primary afferent fibres from rabbit pleura in vitro. Somatosens Mot Res 14, 229–236., 百拇医药(Yves Jammes, Delphine Tro)