- Email: [email protected]
Esophagocardiac convergence onto thoracic spinal neurons: comparison of cervical and thoracic esophagus
Brain Research 1008 (2004) 193 – 197 www.elsevier.com/locate/brainres
Research report
Esophagocardiac convergence onto thoracic spinal neurons: comparison of cervical and thoracic esophagus Chao Qin *, Margaret J. Chandler, Robert D. Foreman Department of Physiology, University of Oklahoma Health Sciences Center, P.O. Box 26901, Oklahoma City, OK 73190, USA Accepted 28 December 2003 Available online 8 April 2004
Abstract The aim of this study was to characterize thoracic spinal neurons receiving convergent inputs from the esophagus, heart and somatic receptive fields. Extracellular potentials of single T3 – T4 spinal neurons were recorded in pentobarbital anesthetized male rats. Thoracic and cervical esophageal distensions (TED, CED) were produced by water inflation of a latex balloon. A catheter was placed in the pericardial sac to administer bradykinin or a mixture of algogenic chemicals. 96/311 (31%) neurons responded to both TED and intrapericardial chemicals (IC) and 48/177 (27%) neurons responded to both CED and IC. Long-lasting excitatory responses were more frequently encountered ( P < 0.05) in esophagocardiac spinal neurons responding to TED (T-ECSNs, 62/91) than in neurons responding to CED (C-ECSNs, 23/47). Ninety-one percent of T-ECSNs and 98% of C-ECSNs had somatic fields on chest, axilla and upper back areas. Esophagocardiac convergence on thoracic spinal neurons provided a spinal mechanism that might mediate viscerovisceral nociception and reflexes. D 2004 Elsevier B.V. All rights reserved. Theme: Sensory system Topic: Pain modulation: anatomy and physiology Keywords: Esophageal pain; Angina pectoris; Visceral nociception; Esophagocardiac reflex; Spinal cord; Rat
1. Introduction Esophageal chest pain is often misdiagnosed as cardiac pain because the sensory characteristics and referred sites are very similar. These similarities are often explained by the convergence-projection theory [8,19]. This theory is supported by evidence in cats that afferent inputs from the heart and esophagus converge on the same spinal neurons that also receive nociceptive inputs from somatic fields [9]. In rats, esophagocardiac convergence to the upper thoracic spinal cord has not been systemically examined, although a few viscerosomatic spinal neurons that receive convergent inputs from esophagus and heart have been reported [7]. Furthermore, no comparisons have been made between the effects of thoracic and cervical esophageal distensions (TED, CED) on spinal neurons. The purpose of this study was to quantitatively characterize and compare the * Corresponding author. Tel.: +1-405-271-2226-x210; fax: +1-405271-3181. E-mail address: [email protected] (C. Qin). 0006-8993/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2003.12.056
responses of upper thoracic esophagocardiac spinal neurons (ECSNs) to cardiac stimulation and to TED and/or CED. A preliminary report of this work has been presented [3].
2. Materials and methods Experiments were performed in 68 male Sprague – Dawley rats (Charles River) weighing between 350 and 460 g. After initial anesthesia with sodium pentobarbital (60 mg/kg, i.p), catheters were inserted into the right carotid artery to monitor blood pressure and into the left jugular vein to infuse sodium pentobarbital (10 – 15 mg/kg/h) to maintain a constant level of anesthesia throughout the experiment. Animals were paralyzed with pancuronium bromide (0.4 mg/kg, i.p.) and paralysis was maintained at an infusion rate of 0.2 mg/kg/ h (i.v.) during the experiment. A tracheotomy was performed for artificial ventilation using a constant-volume pump (55 – 60 strokes/min, 3.0– 5.0 ml stroke volume). A thermostatically controlled heating pad and overhead infrared lamps was used to keep rectal temperature between 37 and 38 jC.
194
C. Qin et al. / Brain Research 1008 (2004) 193–197
Thoracic laminectomies were performed to expose T3 –T4 spinal segments for recording activity of spinal neurons. After rats were mounted in a stereotaxic headholder, the dura mater was carefully removed and the spinal cord was covered with warm agar (3– 4% in saline) to improve stability for neuronal recording. Carbon-filament glass microelectrodes were used to record extracellular action potentials of single spinal neurons. All recordings were made 0.5– 2 mm lateral from the midline and at depths between 0 and 1.4 mm from dorsal surface of spinal cord. Neuronal activity was recorded online with the Spike 2 data acquisition system (CED, Cambridge). A neuronal response (imp/s) to visceral stimulation was calculated by subtracting the mean of 10 s of control activity from the mean of 10 s of the greatest response. A neuron was classified as responding to various stimuli if the maximal change in activity was at least 20% compared to control activity. Data are presented as mean F S.E. Statistical comparisons were made using Student’s paired or unpaired ttest and Chi-square test. Differences were considered statistically significant at P < 0.05. Esophageal distension and intrapericardial injections of algogenic chemicals were used to determine responses of cells to noxious visceral stimuli. Esophageal distension (ED) was produced by warm water inflation of a small latex balloon (1.0 cm in length) at the end of PE-240 tubing. Balloons were inserted perorally 5 – 6 cm or 9 – 10 cm into esophagus from the front incisors to the end of the tubing for cervical or thoracic esophageal distension (CED or TED), respectively [13,17]. Distending volumes of 0.4 and 0.5 ml (20 s) were used as stimuli for identifying responses to noxious ED [6,17]. Intrapericardial chemical injections were made to produce noxious cardiac inputs. A silicone tube (0.020 ID, 0.037 OD, 14 – 16 cm in length) with 8– 10 small holes in distal 2 cm was carefully inserted into the pericardial sac over the left ventricle [7,16]. To activate cardiac receptors, solutions of bradykinin (10 5 M, 0.2 ml) or a mixture of algogenic chemicals (0.2 –0.3 ml, adenosine 10 3 M, bradykinin, histamine, serotonin, prostaglandin E2 10 5 M each) were injected into the pericardial sac via a 1-ml syringe that was connected to the catheter. After 1 min, a syringe (1 ml) was used to remove the intrapericardial chemicals (IC) and two to three saline flushes (0.2 ml each) were used to rinse the pericardial sac. Cutaneous receptive fields of neurons were tested for responses to innocuous brushing with a camel-hair brush and to noxious pinch of skin and muscles with a blunt forceps. Neurons were categorized as follows: wide dynamic range (WDR) neurons were responsive to brushing the hair and had a greater response to noxious pinching of the somatic field; high-threshold (HT) neurons only responded to noxious pinching; low-threshold (LT) neurons responded primarily to hair movement. If a cutaneous receptive field was not found, forelimb or shoulder joint movement was examined. To mark the locations of spinal neurons, electrolytic lesions (50 AA DC, 20 s) were made at recording sites after
neurons with visceral inputs had been studied. At the end of the experiment, animals were euthanasized with an overdose of pentobarbital. The thoracic spinal cord was removed and placed in 10% buffered formalin solution. After at least 3 days, frozen sections (55 –60 Am) of the upper thoracic cord were made and lesion sites in the spinal cord were viewed under a microscope. Laminae of gray matter were identified using the cytoarchitectonic scheme reported by Molander et al. [15].
3. Results 3.1. Responses to TED Of 311 spinal neurons examined for responses to both TED and IC, 96 (31%) neurons had convergent inputs from the heart and thoracic esophagus. Eighty-seven of 96 (91%) of ECSNs with inputs from thoracic esophagus (T-ECSNs) were excited by both TED and IC; the remainder exhibited multiple patterns of neuronal responses to visceral stimuli (Fig. 1A). Characteristics of spontaneous activity and excitatory responses of T-ECSNs to TED are shown in Table 1. The excitatory responses were subdivided into two groups based on the time of recovery to control activity after TED (0.4 ml) was terminated [6,17]; responses with recovery time V 5 s were classified as short-lasting excitatory (SL-E) and responses with after-discharges >5 s were classified as long-lasting excitatory (LL-E) (Table 1) Excitatory responses of T-ECSNs to cardiac stimulation also were subdivided into SL-E and LL-E groups with recovery time V 50 and >50 s, respectively [16]. Mean duration of LL-E responses to IC was significantly longer than for SL-E neurons (Table 2, P < 0.01). Of the 89/96 (93%) T-ECSNs that responded to mechanical stimulation of somatic structures, 36 neurons were classified as WDR and 47 neurons were HT. Excitatory receptive fields were distributed on the chest, forelimb, upper back and shoulder area; 6 neurons responded only to movement of forelimbs. No neuron was classified as LT. 3.2. Responses to CED CED (0.4 ml) changed activity of 48/177 (27%) neurons with cardiac inputs. Of these ECSNs with input from cervical esophagus (C-ECSNs), 90% (43/48) were excited by both CED and IC (Fig. 1A). Spontaneous activity and excitatory responses of C-ECSNs to CED are shown in Table 1. Similar to T-ECSNs, neuronal responses to CED were classified as SL-E or LL-E. Mean duration of LL-E responses to CED was significantly longer than for SL-E neurons (Table 1, P < 0.01). Also, average duration of LL-E responses to IC was significantly longer than for SL-E neurons (Table 2, P < 0.01). Forty-seven of 48 (98%) C-ECSNs also responded to mechanical stimulation of somatic structures: 15 WDR, 30 HT and 2 neurons with joint input were found.
C. Qin et al. / Brain Research 1008 (2004) 193–197
195
Fig. 1. Response patterns of ECSNs to stimulation of both heart and cervical and/or thoracic esophagus. (A, AV) Proportions of response patterns of ECSNs to thoracic or cervical esophageal distension (TED or CED) and intrapericardial chemicals (IC). R, response; NR, no response; E, excitatory response; I, inhibitory response. Black area shows viscerovisceral convergent neurons; white area shows non-convergent visceral neurons. First response is to TED or CED and second response is to IC. (B) A neuron with excitatory responses to CED, TED and IC. (C) A neuron with inhibitory responses to CED, TED and IC. (D) A neuron excited by CED and TED but inhibited by IC.
Thus, somatic field properties of C-ECSNs were not different from T-ECSNs. 3.3. Comparison of T- and C-ECSNs Esophagocardiac convergent neurons were found in the dorsal horn and intermediate zone of the spinal cord gray
matter. No difference was found in distribution of T-ECSNs and C-ECSNs. ECSNs with LL-E responses were more frequently encountered with TED than CED (63/91 vs. 23/ 47, P < 0.05). The mean duration of excitatory responses to TED for 91 T-ECSNs was significantly longer than responses to CED for 47 C-ECSNs (32.1 F 1.6 vs. 23.8 F 0.7 s, P < 0.01). This difference resulted from the
Table 1 Comparison of excitatory responses of ECSNs to TED or CED (0.4 ml) Stimuli
Groups
n
Spontaneous activity (imp/s)
Latency (s)
Responses (imp/s)
Duration (s)
TED
Total SL-E LL-E Total SL-E LL-E
91 29 62 47 24 23
10.8 F 0.9 9.4 F 1.4 11.4 F 1.1 8.7 F 1.1 8.0 F 1.6 9.4 F 1.5
1.9 F 0.2 1.4 F 0.3 2.2 F 0.2 2.2 F 0.3 1.9 F 0.4 2.5 F 0.5
14.4 F 0.9 12.5 F 1.7 15.3 F 1.1 15.3 F 1.4 14.1 F 2.1 16.5 F 2.0
32.1 F 1.6 21.2 F 0.4 37.1 F 2.1* 23.8 F 0.7# 20.6 F 0.4 27.1 F 0.8*,#
CED
* P < 0.01 compared to corresponding durations of SL-E responses in ECSNs to TED or CED. # P < 0.01 compared to corresponding durations of LL-E responses in T-ECSNs.
196
C. Qin et al. / Brain Research 1008 (2004) 193–197
Table 2 Comparison of excitatory responses of C-ECSNs and T-ECSNs to IC Neurons
n
Spontaneous activity (imp/s)
Latency (s)
Responses (imp/s)
Duration (s)
T-ECSN SL-E LL-E C-ECSN SL-E LL-E
89 40 49 43 27 16
10.5 F 0.8 10.7 F 1.4 10.3 F 1.1 9.5 F 1.3 9.5 F 1.9 9.5 F 1.8
10.3 F 1.0 8.8 F 0.8 11.5 F 1.6 7.9 F 0.8 7.6 F 0.8 8.4 F 1.8
22.8 F 1.7 23.4 F 2.6 22.3 F 2.2 20.5 F 2.3 18.3 F 3.0 24.0 F 3.5
158.8 F 13.2 72.8 F 4.2 229.0 F 18.3* 128.4 F 18.0 67.6 F 3.9 231.1 F 35.1*
* P < 0.01 compared to corresponding duration of SL-E responses to IC.
variation in mean duration of LL-E responses of T-ECSNs and C-ECSNs (Table 1). However, no difference was found between mean duration of excitatory responses of T-ECSNs and C-ECSNs to a noxious cardiac stimulus (Table 2). For 122 ECSNs examined for responses to both TED and CED, 33 (27%) neurons were T-C-ECSNs, i.e. neurons with convergent inputs from thoracic esophagus, cervical esophagus and heart (Fig. 1B – D). Thirty of 33 (91%) T-C-ECSNs had excitatory responses to TED, CED and IC; 1 neuron was inhibited by TED, CED and IC; 2 neurons were excited by both TED and CED but inhibited by IC. For 32 T-CECSNs, LL-E responses to TED (n = 25) were more often observed than LL-E responses to CED (n = 17, P < 0.05). Mean duration of excitatory responses of T-C-ECSNs to TED was significantly longer than to CED (33.4 F 2.9 vs. 24.3 F 0.7 s, n = 32, P < 0.01). This difference was due to longer LL-E responses to TED compared to LL-E responses to CED (37.4 F 3.5 vs. 27.0 F 1.1 s, P < 0.01).
4. Discussion In the present study, 96/311 (31%) T3 –T4 spinal neurons responded to both TED and IC and 48/177 (27%) neurons responded to CED and IC. In a much smaller sample size reported in a previous study [7], 5/12 (42%) upper thoracic spinal neurons with mid-chest esophageal input also responded to pericardial administration of algogenic chemicals. The higher percentage of ECSNs found previously in rats might result from the different search stimuli used to examine neuronal responses to visceral stimuli. EuchnerWamser et al. [7] used esophageal distension as the search stimulus to select neurons for further examination of neuronal responses to chemical irritation of cardiac afferent fibers. In the present study, all spinal neurons were examined for responses to both esophageal and cardiac stimuli, even if a neuron did not respond to stimulation of one of the visceral organs. A higher percentage of ECSNs has been reported in cats [9]; 76% of thoracic spinal neurons (T2 – T7) with esophageal input also responded to bradykinin injections into the left atrium. Possible reasons for the greater number of ECSNs in cats might be species differences, the different routes of administration for cardiac chemical stimulation, or the extensive surgery used for open thorax preparation in cats.
Both TED and CED changed the activity of 27% of upper thoracic spinal neurons that also received noxious cardiac input. This percentage of convergent inputs from thoracic and cervical esophagus is consistent with observations in our previous study [17], in which 26% of upper thoracic spinal neurons received inputs from both thoracic and cervical esophagus. This finding is presumed to be a spinal neural basis for intersegmental integration of esophageal sensory information from different regions of the esophagus. Furthermore, LL-E responses of thoracic ECSNs to TED were more frequently encountered than LL-E responses to CED. This finding also agrees with a comparison of thoracic spinal neuronal responses to TED and CED [17] and implies that there is a differentiation of spinal sensory processing for afferent inputs originating from different regions of the esophagus. In support of this concept, Loomis et al. [12] report that distension of the lower esophagus is a more effective stimulus for pseudoaffective responses than distension of the upper esophagus. The majority of ECSNs in the present study responded to noxious stimulation of somatic receptive fields located on the chest, axilla, forearms, shoulder and upper back areas. This generally agrees with previous studies, in which thoracic spinal neurons with esophageal or cardiac inputs have been examined separately for neuronal responses to somatic stimuli [6,16,17]. Viscerosomatic convergence onto thoracic ECSNs could explain why a patient experiencing angina pectoris or esophagitis might falsely perceive the pain as coming from the same somatic areas [5,8,18 – 20]. Also, convergence of inputs from cervical and thoracic esophagus onto thoracic ECSNs might be a neural basis for referred non-cardiac chest pain. Studies have shown important interactions between the esophagus and heart with respect to visceral sensation and motility, i.e. esophagocardiac or cardioesophageal bidirectional hyperesthesia and reflexes in humans [1,2,14] and in animals [4,11]. For example, stimulation of the esophagus in patients with acid infusion, distension or hot/cold liquid results in physiological changes in coronary blood flow and cardiac rhythm [10,21]. Also, patients with diffuse esophageal spasm have exaggerated esophagocardiac reflexes with esophageal distension, indicating viscerovisceral hyperalgesia [1]. Coronary ischemia triggers a gastroesophageal reflex that lowers the esophageal pain threshold [14]. Finally, in a
C. Qin et al. / Brain Research 1008 (2004) 193–197
study in rats, noxious stimulation of the esophagus and heart produces contractions of the same paraspinal muscles, indicating motor reflexes [11]. Therefore, it is reasonable to presume that viscerovisceral convergence of esophageal and cardiac inputs onto single thoracic spinal neurons might contribute to afferent –afferent or afferent –efferent ‘‘crosstalk’’ between esophagus and heart. In summary, the present study in rats quantitatively characterized activity of upper thoracic spinal neurons with convergent inputs from heart, esophagus and somatic receptive fields. The findings might explain the similarities and differences in esophagocardiac referred pain originating in cervical and thoracic esophagus. Viscerovisceral afferent convergence of afferent inputs from the heart and esophagus also provided a spinal mechanism that might mediate bidirectional ‘‘cross-talk’’ between two visceral organs.
Acknowledgements The authors thank Diana Holston for excellent technical assistance. This work was supported by NIH grant NS-35471.
References [1] M. Bortolotti, P. Sarti, F. Brunelli, M. Mazza, L. Barbara, Abnormal esophagocardiac inhibitory reflex in patients with diffuse esophageal spasm, Digestion 56 (1995) 488 – 492. [2] M. Bortolotti, N. Pandolfo, M. Miglioli, Abnormal esophagocardiac reflex in patients with non-cardiac chest pain, Dis. Esophagus 14 (2001) 57 – 59. [3] M.J. Chandler, C. Qin, R.D. Foreman, Esophagocardiac convergence onto thoracic spinal neurons in rats: comparison of cervical and thoracic esophagus, Program No. 451.9. 2002 Abstract Viewer/Itinerary Planner. Washington, DC: Society for Neuroscience, 2002. Online. [4] S.B. Cheng, T. Hayakawa, S. Kuchiiwa, S. Maeda, H. Ito, M. Seki, S. Nakagawa, Evidence for the collateral innervation of the esophagus and the heart from neurons in the compact formation of the nucleus ambiguus of the rat, Brain Res. 832 (1999) 171 – 174. [5] H.A. Davis, D.B. Jone, J. Rhodes, Esophageal spasm as the cause of chest pain, JAMA 248 (1982) 2274 – 2278.
197
[6] I. Euchner-Wamser, J.N. Sengupta, G.F. Gebhart, S.T. Meller, Characterization of responses of T2 – T4 spinal cord neurons to esophageal distension in the rat, J. Neurophysiol. 69 (1993) 868 – 883. [7] I. Euchner-Wamser, S.T. Meller, G.F. Gebhart, A model of cardiac nociception in chronically instrumented rats: behavioral and electrophysiological effects of pericardial administration of algogenic substance, Pain 58 (1994) 117 – 128. [8] R.D. Foreman, Mechanisms of cardiac pain, Annu. Rev. Physiol. 61 (1999) 143 – 167. [9] D.W. Garrison, M.J. Chandler, R.D. Foreman, Viscerosomatic convergence onto feline spinal neurons from esophagus, heart and somatic field: effects of inflammation, Pain 49 (1992) 373 – 382. [10] P.A. Gayheart, P.A. Gwirtz, J.S. Bravenec, N. Longlet, C.E. Jones, An alpha-adrenergic coronary constriction during esophageal distention in the dog, J. Cardiovasc. Pharmacol. 17 (1991) 747 – 753. [11] C.J. Jou, J.P. Farber, C. Qin, R.D. Foreman, Convergent pathways for cardiac- and esophageal-somatic motor reflexes in rats, Auton. Neurosci. 99 (2002) 70 – 77. [12] C.W. Loomis, D. Yao, D. Bieger, Characterization of an esophagocardiovascular reflex in the rat, Am. J. Physiol., Regul. Integr. Comp. Physiol. 272 (1997) R1783 – R1791. [13] W.Y. Lu, D. Bieger, Vagovagal reflex motility patterns of the rat esophagus, Am. J. Physiol., Regul. Integr. Comp. Physiol. 274 (1998) R1425 – R1435. [14] L.J. Makk, M. Leesar, A. Joseph, C.P. Prince, R.A. Wright, Cardioesophageal reflexes: an invasive human study, Dig. Dis. Sci. 45 (2000) 2451 – 2454. [15] C. Molander, Q. Xu, C. Rivero-Melian, G. Grant, Cytoarchitectonic organization of the spinal cord in the rat: II. The cervical and upper thoracic cord, J. Comp. Neurol. 289 (1989) 375 – 385. [16] C. Qin, M.J. Chandler, K.E. Miller, R.D. Foreman, Chemical activation of cardiac receptors affects activity of superficial and deeper T3 – T4 spinal neurons in rats, Brain Res. 959 (2003) 77 – 85. [17] C. Qin, M.J. Chandler, C.J. Jou, K.E. Miller, R.D. Foreman, Comparison of responses of T3 – T4 spinal neurons to cervical and thoracic esophageal distention in rats, Auton. Neurosci. 109 (2003) 10 – 20. [18] J.E. Richter, L.A. Bradley, D.O. Castell, Esophageal chest pain: current controversies in pathogenesis, diagnosis and therapy, Ann. Intern. Med. 110 (1989) 66 – 78. [19] T.C. Ruch, Pathophysiology of pain, in: T.C. Ruch, H.D. Patton (Eds.), Physiology and Biophysics, Saunders, Philadelphia, 1961, pp. 345 – 363. [20] C. Sylven, Angina pectoris: clinical characteristics, neurophysiological and molecular mechanisms, Pain 36 (1989) 145 – 167. [21] R.A. Wright, S.A. Miller, B.F. Corsello, Acid-induced esophagobronchial-cardiac reflexes in humans, Gastroenterology 99 (1990) 71 – 73.
Research report
Esophagocardiac convergence onto thoracic spinal neurons: comparison of cervical and thoracic esophagus Chao Qin *, Margaret J. Chandler, Robert D. Foreman Department of Physiology, University of Oklahoma Health Sciences Center, P.O. Box 26901, Oklahoma City, OK 73190, USA Accepted 28 December 2003 Available online 8 April 2004
Abstract The aim of this study was to characterize thoracic spinal neurons receiving convergent inputs from the esophagus, heart and somatic receptive fields. Extracellular potentials of single T3 – T4 spinal neurons were recorded in pentobarbital anesthetized male rats. Thoracic and cervical esophageal distensions (TED, CED) were produced by water inflation of a latex balloon. A catheter was placed in the pericardial sac to administer bradykinin or a mixture of algogenic chemicals. 96/311 (31%) neurons responded to both TED and intrapericardial chemicals (IC) and 48/177 (27%) neurons responded to both CED and IC. Long-lasting excitatory responses were more frequently encountered ( P < 0.05) in esophagocardiac spinal neurons responding to TED (T-ECSNs, 62/91) than in neurons responding to CED (C-ECSNs, 23/47). Ninety-one percent of T-ECSNs and 98% of C-ECSNs had somatic fields on chest, axilla and upper back areas. Esophagocardiac convergence on thoracic spinal neurons provided a spinal mechanism that might mediate viscerovisceral nociception and reflexes. D 2004 Elsevier B.V. All rights reserved. Theme: Sensory system Topic: Pain modulation: anatomy and physiology Keywords: Esophageal pain; Angina pectoris; Visceral nociception; Esophagocardiac reflex; Spinal cord; Rat
1. Introduction Esophageal chest pain is often misdiagnosed as cardiac pain because the sensory characteristics and referred sites are very similar. These similarities are often explained by the convergence-projection theory [8,19]. This theory is supported by evidence in cats that afferent inputs from the heart and esophagus converge on the same spinal neurons that also receive nociceptive inputs from somatic fields [9]. In rats, esophagocardiac convergence to the upper thoracic spinal cord has not been systemically examined, although a few viscerosomatic spinal neurons that receive convergent inputs from esophagus and heart have been reported [7]. Furthermore, no comparisons have been made between the effects of thoracic and cervical esophageal distensions (TED, CED) on spinal neurons. The purpose of this study was to quantitatively characterize and compare the * Corresponding author. Tel.: +1-405-271-2226-x210; fax: +1-405271-3181. E-mail address: [email protected] (C. Qin). 0006-8993/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2003.12.056
responses of upper thoracic esophagocardiac spinal neurons (ECSNs) to cardiac stimulation and to TED and/or CED. A preliminary report of this work has been presented [3].
2. Materials and methods Experiments were performed in 68 male Sprague – Dawley rats (Charles River) weighing between 350 and 460 g. After initial anesthesia with sodium pentobarbital (60 mg/kg, i.p), catheters were inserted into the right carotid artery to monitor blood pressure and into the left jugular vein to infuse sodium pentobarbital (10 – 15 mg/kg/h) to maintain a constant level of anesthesia throughout the experiment. Animals were paralyzed with pancuronium bromide (0.4 mg/kg, i.p.) and paralysis was maintained at an infusion rate of 0.2 mg/kg/ h (i.v.) during the experiment. A tracheotomy was performed for artificial ventilation using a constant-volume pump (55 – 60 strokes/min, 3.0– 5.0 ml stroke volume). A thermostatically controlled heating pad and overhead infrared lamps was used to keep rectal temperature between 37 and 38 jC.
194
C. Qin et al. / Brain Research 1008 (2004) 193–197
Thoracic laminectomies were performed to expose T3 –T4 spinal segments for recording activity of spinal neurons. After rats were mounted in a stereotaxic headholder, the dura mater was carefully removed and the spinal cord was covered with warm agar (3– 4% in saline) to improve stability for neuronal recording. Carbon-filament glass microelectrodes were used to record extracellular action potentials of single spinal neurons. All recordings were made 0.5– 2 mm lateral from the midline and at depths between 0 and 1.4 mm from dorsal surface of spinal cord. Neuronal activity was recorded online with the Spike 2 data acquisition system (CED, Cambridge). A neuronal response (imp/s) to visceral stimulation was calculated by subtracting the mean of 10 s of control activity from the mean of 10 s of the greatest response. A neuron was classified as responding to various stimuli if the maximal change in activity was at least 20% compared to control activity. Data are presented as mean F S.E. Statistical comparisons were made using Student’s paired or unpaired ttest and Chi-square test. Differences were considered statistically significant at P < 0.05. Esophageal distension and intrapericardial injections of algogenic chemicals were used to determine responses of cells to noxious visceral stimuli. Esophageal distension (ED) was produced by warm water inflation of a small latex balloon (1.0 cm in length) at the end of PE-240 tubing. Balloons were inserted perorally 5 – 6 cm or 9 – 10 cm into esophagus from the front incisors to the end of the tubing for cervical or thoracic esophageal distension (CED or TED), respectively [13,17]. Distending volumes of 0.4 and 0.5 ml (20 s) were used as stimuli for identifying responses to noxious ED [6,17]. Intrapericardial chemical injections were made to produce noxious cardiac inputs. A silicone tube (0.020 ID, 0.037 OD, 14 – 16 cm in length) with 8– 10 small holes in distal 2 cm was carefully inserted into the pericardial sac over the left ventricle [7,16]. To activate cardiac receptors, solutions of bradykinin (10 5 M, 0.2 ml) or a mixture of algogenic chemicals (0.2 –0.3 ml, adenosine 10 3 M, bradykinin, histamine, serotonin, prostaglandin E2 10 5 M each) were injected into the pericardial sac via a 1-ml syringe that was connected to the catheter. After 1 min, a syringe (1 ml) was used to remove the intrapericardial chemicals (IC) and two to three saline flushes (0.2 ml each) were used to rinse the pericardial sac. Cutaneous receptive fields of neurons were tested for responses to innocuous brushing with a camel-hair brush and to noxious pinch of skin and muscles with a blunt forceps. Neurons were categorized as follows: wide dynamic range (WDR) neurons were responsive to brushing the hair and had a greater response to noxious pinching of the somatic field; high-threshold (HT) neurons only responded to noxious pinching; low-threshold (LT) neurons responded primarily to hair movement. If a cutaneous receptive field was not found, forelimb or shoulder joint movement was examined. To mark the locations of spinal neurons, electrolytic lesions (50 AA DC, 20 s) were made at recording sites after
neurons with visceral inputs had been studied. At the end of the experiment, animals were euthanasized with an overdose of pentobarbital. The thoracic spinal cord was removed and placed in 10% buffered formalin solution. After at least 3 days, frozen sections (55 –60 Am) of the upper thoracic cord were made and lesion sites in the spinal cord were viewed under a microscope. Laminae of gray matter were identified using the cytoarchitectonic scheme reported by Molander et al. [15].
3. Results 3.1. Responses to TED Of 311 spinal neurons examined for responses to both TED and IC, 96 (31%) neurons had convergent inputs from the heart and thoracic esophagus. Eighty-seven of 96 (91%) of ECSNs with inputs from thoracic esophagus (T-ECSNs) were excited by both TED and IC; the remainder exhibited multiple patterns of neuronal responses to visceral stimuli (Fig. 1A). Characteristics of spontaneous activity and excitatory responses of T-ECSNs to TED are shown in Table 1. The excitatory responses were subdivided into two groups based on the time of recovery to control activity after TED (0.4 ml) was terminated [6,17]; responses with recovery time V 5 s were classified as short-lasting excitatory (SL-E) and responses with after-discharges >5 s were classified as long-lasting excitatory (LL-E) (Table 1) Excitatory responses of T-ECSNs to cardiac stimulation also were subdivided into SL-E and LL-E groups with recovery time V 50 and >50 s, respectively [16]. Mean duration of LL-E responses to IC was significantly longer than for SL-E neurons (Table 2, P < 0.01). Of the 89/96 (93%) T-ECSNs that responded to mechanical stimulation of somatic structures, 36 neurons were classified as WDR and 47 neurons were HT. Excitatory receptive fields were distributed on the chest, forelimb, upper back and shoulder area; 6 neurons responded only to movement of forelimbs. No neuron was classified as LT. 3.2. Responses to CED CED (0.4 ml) changed activity of 48/177 (27%) neurons with cardiac inputs. Of these ECSNs with input from cervical esophagus (C-ECSNs), 90% (43/48) were excited by both CED and IC (Fig. 1A). Spontaneous activity and excitatory responses of C-ECSNs to CED are shown in Table 1. Similar to T-ECSNs, neuronal responses to CED were classified as SL-E or LL-E. Mean duration of LL-E responses to CED was significantly longer than for SL-E neurons (Table 1, P < 0.01). Also, average duration of LL-E responses to IC was significantly longer than for SL-E neurons (Table 2, P < 0.01). Forty-seven of 48 (98%) C-ECSNs also responded to mechanical stimulation of somatic structures: 15 WDR, 30 HT and 2 neurons with joint input were found.
C. Qin et al. / Brain Research 1008 (2004) 193–197
195
Fig. 1. Response patterns of ECSNs to stimulation of both heart and cervical and/or thoracic esophagus. (A, AV) Proportions of response patterns of ECSNs to thoracic or cervical esophageal distension (TED or CED) and intrapericardial chemicals (IC). R, response; NR, no response; E, excitatory response; I, inhibitory response. Black area shows viscerovisceral convergent neurons; white area shows non-convergent visceral neurons. First response is to TED or CED and second response is to IC. (B) A neuron with excitatory responses to CED, TED and IC. (C) A neuron with inhibitory responses to CED, TED and IC. (D) A neuron excited by CED and TED but inhibited by IC.
Thus, somatic field properties of C-ECSNs were not different from T-ECSNs. 3.3. Comparison of T- and C-ECSNs Esophagocardiac convergent neurons were found in the dorsal horn and intermediate zone of the spinal cord gray
matter. No difference was found in distribution of T-ECSNs and C-ECSNs. ECSNs with LL-E responses were more frequently encountered with TED than CED (63/91 vs. 23/ 47, P < 0.05). The mean duration of excitatory responses to TED for 91 T-ECSNs was significantly longer than responses to CED for 47 C-ECSNs (32.1 F 1.6 vs. 23.8 F 0.7 s, P < 0.01). This difference resulted from the
Table 1 Comparison of excitatory responses of ECSNs to TED or CED (0.4 ml) Stimuli
Groups
n
Spontaneous activity (imp/s)
Latency (s)
Responses (imp/s)
Duration (s)
TED
Total SL-E LL-E Total SL-E LL-E
91 29 62 47 24 23
10.8 F 0.9 9.4 F 1.4 11.4 F 1.1 8.7 F 1.1 8.0 F 1.6 9.4 F 1.5
1.9 F 0.2 1.4 F 0.3 2.2 F 0.2 2.2 F 0.3 1.9 F 0.4 2.5 F 0.5
14.4 F 0.9 12.5 F 1.7 15.3 F 1.1 15.3 F 1.4 14.1 F 2.1 16.5 F 2.0
32.1 F 1.6 21.2 F 0.4 37.1 F 2.1* 23.8 F 0.7# 20.6 F 0.4 27.1 F 0.8*,#
CED
* P < 0.01 compared to corresponding durations of SL-E responses in ECSNs to TED or CED. # P < 0.01 compared to corresponding durations of LL-E responses in T-ECSNs.
196
C. Qin et al. / Brain Research 1008 (2004) 193–197
Table 2 Comparison of excitatory responses of C-ECSNs and T-ECSNs to IC Neurons
n
Spontaneous activity (imp/s)
Latency (s)
Responses (imp/s)
Duration (s)
T-ECSN SL-E LL-E C-ECSN SL-E LL-E
89 40 49 43 27 16
10.5 F 0.8 10.7 F 1.4 10.3 F 1.1 9.5 F 1.3 9.5 F 1.9 9.5 F 1.8
10.3 F 1.0 8.8 F 0.8 11.5 F 1.6 7.9 F 0.8 7.6 F 0.8 8.4 F 1.8
22.8 F 1.7 23.4 F 2.6 22.3 F 2.2 20.5 F 2.3 18.3 F 3.0 24.0 F 3.5
158.8 F 13.2 72.8 F 4.2 229.0 F 18.3* 128.4 F 18.0 67.6 F 3.9 231.1 F 35.1*
* P < 0.01 compared to corresponding duration of SL-E responses to IC.
variation in mean duration of LL-E responses of T-ECSNs and C-ECSNs (Table 1). However, no difference was found between mean duration of excitatory responses of T-ECSNs and C-ECSNs to a noxious cardiac stimulus (Table 2). For 122 ECSNs examined for responses to both TED and CED, 33 (27%) neurons were T-C-ECSNs, i.e. neurons with convergent inputs from thoracic esophagus, cervical esophagus and heart (Fig. 1B – D). Thirty of 33 (91%) T-C-ECSNs had excitatory responses to TED, CED and IC; 1 neuron was inhibited by TED, CED and IC; 2 neurons were excited by both TED and CED but inhibited by IC. For 32 T-CECSNs, LL-E responses to TED (n = 25) were more often observed than LL-E responses to CED (n = 17, P < 0.05). Mean duration of excitatory responses of T-C-ECSNs to TED was significantly longer than to CED (33.4 F 2.9 vs. 24.3 F 0.7 s, n = 32, P < 0.01). This difference was due to longer LL-E responses to TED compared to LL-E responses to CED (37.4 F 3.5 vs. 27.0 F 1.1 s, P < 0.01).
4. Discussion In the present study, 96/311 (31%) T3 –T4 spinal neurons responded to both TED and IC and 48/177 (27%) neurons responded to CED and IC. In a much smaller sample size reported in a previous study [7], 5/12 (42%) upper thoracic spinal neurons with mid-chest esophageal input also responded to pericardial administration of algogenic chemicals. The higher percentage of ECSNs found previously in rats might result from the different search stimuli used to examine neuronal responses to visceral stimuli. EuchnerWamser et al. [7] used esophageal distension as the search stimulus to select neurons for further examination of neuronal responses to chemical irritation of cardiac afferent fibers. In the present study, all spinal neurons were examined for responses to both esophageal and cardiac stimuli, even if a neuron did not respond to stimulation of one of the visceral organs. A higher percentage of ECSNs has been reported in cats [9]; 76% of thoracic spinal neurons (T2 – T7) with esophageal input also responded to bradykinin injections into the left atrium. Possible reasons for the greater number of ECSNs in cats might be species differences, the different routes of administration for cardiac chemical stimulation, or the extensive surgery used for open thorax preparation in cats.
Both TED and CED changed the activity of 27% of upper thoracic spinal neurons that also received noxious cardiac input. This percentage of convergent inputs from thoracic and cervical esophagus is consistent with observations in our previous study [17], in which 26% of upper thoracic spinal neurons received inputs from both thoracic and cervical esophagus. This finding is presumed to be a spinal neural basis for intersegmental integration of esophageal sensory information from different regions of the esophagus. Furthermore, LL-E responses of thoracic ECSNs to TED were more frequently encountered than LL-E responses to CED. This finding also agrees with a comparison of thoracic spinal neuronal responses to TED and CED [17] and implies that there is a differentiation of spinal sensory processing for afferent inputs originating from different regions of the esophagus. In support of this concept, Loomis et al. [12] report that distension of the lower esophagus is a more effective stimulus for pseudoaffective responses than distension of the upper esophagus. The majority of ECSNs in the present study responded to noxious stimulation of somatic receptive fields located on the chest, axilla, forearms, shoulder and upper back areas. This generally agrees with previous studies, in which thoracic spinal neurons with esophageal or cardiac inputs have been examined separately for neuronal responses to somatic stimuli [6,16,17]. Viscerosomatic convergence onto thoracic ECSNs could explain why a patient experiencing angina pectoris or esophagitis might falsely perceive the pain as coming from the same somatic areas [5,8,18 – 20]. Also, convergence of inputs from cervical and thoracic esophagus onto thoracic ECSNs might be a neural basis for referred non-cardiac chest pain. Studies have shown important interactions between the esophagus and heart with respect to visceral sensation and motility, i.e. esophagocardiac or cardioesophageal bidirectional hyperesthesia and reflexes in humans [1,2,14] and in animals [4,11]. For example, stimulation of the esophagus in patients with acid infusion, distension or hot/cold liquid results in physiological changes in coronary blood flow and cardiac rhythm [10,21]. Also, patients with diffuse esophageal spasm have exaggerated esophagocardiac reflexes with esophageal distension, indicating viscerovisceral hyperalgesia [1]. Coronary ischemia triggers a gastroesophageal reflex that lowers the esophageal pain threshold [14]. Finally, in a
C. Qin et al. / Brain Research 1008 (2004) 193–197
study in rats, noxious stimulation of the esophagus and heart produces contractions of the same paraspinal muscles, indicating motor reflexes [11]. Therefore, it is reasonable to presume that viscerovisceral convergence of esophageal and cardiac inputs onto single thoracic spinal neurons might contribute to afferent –afferent or afferent –efferent ‘‘crosstalk’’ between esophagus and heart. In summary, the present study in rats quantitatively characterized activity of upper thoracic spinal neurons with convergent inputs from heart, esophagus and somatic receptive fields. The findings might explain the similarities and differences in esophagocardiac referred pain originating in cervical and thoracic esophagus. Viscerovisceral afferent convergence of afferent inputs from the heart and esophagus also provided a spinal mechanism that might mediate bidirectional ‘‘cross-talk’’ between two visceral organs.
Acknowledgements The authors thank Diana Holston for excellent technical assistance. This work was supported by NIH grant NS-35471.
References [1] M. Bortolotti, P. Sarti, F. Brunelli, M. Mazza, L. Barbara, Abnormal esophagocardiac inhibitory reflex in patients with diffuse esophageal spasm, Digestion 56 (1995) 488 – 492. [2] M. Bortolotti, N. Pandolfo, M. Miglioli, Abnormal esophagocardiac reflex in patients with non-cardiac chest pain, Dis. Esophagus 14 (2001) 57 – 59. [3] M.J. Chandler, C. Qin, R.D. Foreman, Esophagocardiac convergence onto thoracic spinal neurons in rats: comparison of cervical and thoracic esophagus, Program No. 451.9. 2002 Abstract Viewer/Itinerary Planner. Washington, DC: Society for Neuroscience, 2002. Online. [4] S.B. Cheng, T. Hayakawa, S. Kuchiiwa, S. Maeda, H. Ito, M. Seki, S. Nakagawa, Evidence for the collateral innervation of the esophagus and the heart from neurons in the compact formation of the nucleus ambiguus of the rat, Brain Res. 832 (1999) 171 – 174. [5] H.A. Davis, D.B. Jone, J. Rhodes, Esophageal spasm as the cause of chest pain, JAMA 248 (1982) 2274 – 2278.
197
[6] I. Euchner-Wamser, J.N. Sengupta, G.F. Gebhart, S.T. Meller, Characterization of responses of T2 – T4 spinal cord neurons to esophageal distension in the rat, J. Neurophysiol. 69 (1993) 868 – 883. [7] I. Euchner-Wamser, S.T. Meller, G.F. Gebhart, A model of cardiac nociception in chronically instrumented rats: behavioral and electrophysiological effects of pericardial administration of algogenic substance, Pain 58 (1994) 117 – 128. [8] R.D. Foreman, Mechanisms of cardiac pain, Annu. Rev. Physiol. 61 (1999) 143 – 167. [9] D.W. Garrison, M.J. Chandler, R.D. Foreman, Viscerosomatic convergence onto feline spinal neurons from esophagus, heart and somatic field: effects of inflammation, Pain 49 (1992) 373 – 382. [10] P.A. Gayheart, P.A. Gwirtz, J.S. Bravenec, N. Longlet, C.E. Jones, An alpha-adrenergic coronary constriction during esophageal distention in the dog, J. Cardiovasc. Pharmacol. 17 (1991) 747 – 753. [11] C.J. Jou, J.P. Farber, C. Qin, R.D. Foreman, Convergent pathways for cardiac- and esophageal-somatic motor reflexes in rats, Auton. Neurosci. 99 (2002) 70 – 77. [12] C.W. Loomis, D. Yao, D. Bieger, Characterization of an esophagocardiovascular reflex in the rat, Am. J. Physiol., Regul. Integr. Comp. Physiol. 272 (1997) R1783 – R1791. [13] W.Y. Lu, D. Bieger, Vagovagal reflex motility patterns of the rat esophagus, Am. J. Physiol., Regul. Integr. Comp. Physiol. 274 (1998) R1425 – R1435. [14] L.J. Makk, M. Leesar, A. Joseph, C.P. Prince, R.A. Wright, Cardioesophageal reflexes: an invasive human study, Dig. Dis. Sci. 45 (2000) 2451 – 2454. [15] C. Molander, Q. Xu, C. Rivero-Melian, G. Grant, Cytoarchitectonic organization of the spinal cord in the rat: II. The cervical and upper thoracic cord, J. Comp. Neurol. 289 (1989) 375 – 385. [16] C. Qin, M.J. Chandler, K.E. Miller, R.D. Foreman, Chemical activation of cardiac receptors affects activity of superficial and deeper T3 – T4 spinal neurons in rats, Brain Res. 959 (2003) 77 – 85. [17] C. Qin, M.J. Chandler, C.J. Jou, K.E. Miller, R.D. Foreman, Comparison of responses of T3 – T4 spinal neurons to cervical and thoracic esophageal distention in rats, Auton. Neurosci. 109 (2003) 10 – 20. [18] J.E. Richter, L.A. Bradley, D.O. Castell, Esophageal chest pain: current controversies in pathogenesis, diagnosis and therapy, Ann. Intern. Med. 110 (1989) 66 – 78. [19] T.C. Ruch, Pathophysiology of pain, in: T.C. Ruch, H.D. Patton (Eds.), Physiology and Biophysics, Saunders, Philadelphia, 1961, pp. 345 – 363. [20] C. Sylven, Angina pectoris: clinical characteristics, neurophysiological and molecular mechanisms, Pain 36 (1989) 145 – 167. [21] R.A. Wright, S.A. Miller, B.F. Corsello, Acid-induced esophagobronchial-cardiac reflexes in humans, Gastroenterology 99 (1990) 71 – 73.