‘. ~31 LIBRARY Michigan State University This is to certify that the dissertation entitled "Anatomical and Electrophysiological Characteristics of Splenic and Renal Sympathetic Nerves in Cats" presented by ROBERT LOUIS MECKLER has been accepted towards fulfillment of the requirements for Ph.D. degmwin Physiology 4232§W4x14 (L ko)flsl4*¢&2 Major professor Date M(%:{ \qg l "c'lllfl- Alfi-u—no‘ ‘ ‘ "1 '1‘ I: I ~ - 0-1 1 )V1ESI.J RETURNING MATERIALS: Place in book drop to LIBRARJES remove this checkout from .JIIIKSIIIL. your record. FINES will be charged if book is returned after the date stamped below. ANATOMICAL AND ELECTROPHYSIOLOCICAL CHARACTERISTICS OF RENAL AND SPLBNIC SYMPATHETIC REEVES IN CATS By Robert Louis Heckler A DISSERTATION Submitted to Michigan State University in pertitiel fulfillment of the requirenents for the degree of DOCTOR OF PHILOSOPHY Department of Physiology 1987 ABSTRACT ANATOMICAL AND ELECTROPHYSIOLOGICAL CHARACTERISTICS OF RENAL AND SPLENIC SYMPATHETIC NERVES IN CATS BY Robert Louis Heckler The sympathetic nervous system can control different organs or target tissues individually. The purpose of the present studies was to compare and contrast 1) ganglionic distributions of renal and splenic sympathetic neuronal cell bodies, 2) dependence of spontaneous multiunit discharge of renal and splenic sympathetic nerves upon supraspinal sources of excitatory drive, and 3) responses of single renal and splenic axons to stimulation of afferent nerves. Retrograde axonal transport of horseradish peroxidase was employed to identify ganglionic distributions of cell bodies of postganglionic nerves supplying the spleen and kidney. Most labeled cell bodies of renal nerves were clustered in groups within ganglia of the solar plexus. The remainder of labeled renal neurons were located in upper lumbar or lower thoracic paravertebral sympathetic ganglia. In contrast, 90% of labeled cell bodies of splenic neurons were scattered throughout the left and right celiac poles of the solar plexus. In electro- physiological studies of multifiber sympathetic nerve activity, it was demonstrated that ongoing activity of splenic nerves is less dependent Robert Louis Meckler upon supraspinal sources of excitation than is activity of renal or cardiac nerves. Neither increased nor decreased arterial pressure, systemic hypercapnia and acidosis, nor thoracolumbar dorsal rhizotomy revealed specific inputs responsible for preferential maintenance of splenic nerve activity in spinal cats. Potential heterogeneity of ongoing activity and reflex responses of individual fibers in splenic and renal nerves to stimulation of visceral receptors was assessed. Half of the splenic papulation and all renal fibers had cardiac-related discharge patterns. Of those tested for respiratory-related activity, 302 of splenic and 692 of renal fibers exhibited this pattern. Activity of splenic fibers was influenced less than that of renal fibers by changes in arterial pressure. Chemical stimulation of splenic afferent nerves caused large excitatory responses in activity of all splenic fibers and smaller excitatory responses in discharge of 752 of the renal fibers. Intestinal afferent nerve stimulation caused excitation, inhibition, or no change in splenic fiber discharge, whereas renal unit activity was almost always excited by this stimulation. These results indicate that the sympathetic system is organized to provide different viscera with discrete control. This dissertation is dedicated with love to my mother, Anne Heckler, and to my family. iv ACKNOWLEDGMENTS I am deeply indebted to Dr. Lynne Weaver for her years of patient guidance. In addition I would like to thank the members of my guidance committee: Drs. Franco Calaresu, Gerard Gebber, William Spielman, and Jacob Krier. Thanks also are owed to Dr. Robert Pittman for his friendship and moral support without which I neither could have survived nor succeeded. TABLE OF CONTENTS LIST OF TMLESCOOCOCCO0......OOOOOOCOOOOOOOOOOOOOOOOOOOOOOOOOOO LIST OF FIGURESOOCOO000......O...OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO Chapter I. II. INTRODUCTION............................................ LITERATURE REVIEW Early hallmarks in the study of the sympathetic nervous system...................................... Homeostasis and unitary action of the sympathetic nervous system...................................... Ongoing activity of sympathetic nerves A. B. C. 1. 2. Rhythmicity of sympathetic nerve activity a. cardiac-related rhythms..................... b. respiratory-related rhythms................. c. functional significance of rhythmic nerve activity.............................. Sources of sympathetic activity................. Reflex changes in sympathetic activity 1. 2. Pressoreceptors a. carotid sinus and aortic arch............... b. reflex influences of baroreceptors on peripheral targets....................... c. cardiOpulmonary and thoracic vasculature pressoreceptors................. d. abdominal visceral pressoreceptors.......... Viscerosympathetic reflexes..................... Functional specificity of the sympathetic nervous system 1. 2. 3. Selective connections in sympathetic ganglia.... Electrophysiological evidence for funtional organization of the sympathetic system.......... Neurochemical evidence for specificity of target tissue innervation....................... vi Page viii ix 12 13 18 21 22 27 28 31 35 40 III. Comparison of the distributions of renal and splenic neurons in sympathetic ganglia. A. B. C. D. IntrOductionCOOOOOOOOOOOOOOOOOOOIOOOOOO0.0.0.... MethOdBOOCOCCOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO. Re8u1t800000.00.00.00.00...OOOOOOOOOOOOOOOOIOOOO D13C“881°noooo00000000000000.0000...oeoeoecoeeoo IV. Splenic, renal, and cardiac nerves have unequal dependence upon tonic supraspinal inputs. A. B. C. D. IntrOductionIOOOCCOOOOOOOOOOOOOOOOOOOIOOOOO...O. MethOdSOOOOOO...OOOIOOOOOOOOIOOOOOOOOOOI.0....I. ResultSCOCO0.000000000000000IOI00.00.00.000....0 DiscuaionOOOOOCOCOOOOOOOOOOOOOOOOOOOOOOOOOOIOOI. V. Characteristics of ongoing and reflex discharge of single splenic and renal sympathetic postganglionic fibers in cats. A. B. C. D. Int rOduc tion 0 O O O O O O O O O I O O O O O I O O O O O O O O O C O I O O O O O O O methads O O O O O O O O O O O O O O O O O O O O O I I O O O C O O O O O O O O O O O O O 0 Results 0 O O O O O O O O O O O O O O O O O O O O O O O O O O O O I O O I O O O O O I O O DiscussionCOCCO......0.00.00...OOOOOOIOOOOOIOIO. VI. PRINCIPAL RESULTS AND CONCLUSIONS....................... APPENDIX A..................................................... APPENDIX B..................................................... BIBLIOGRAPHY................................................... vii 42 44 47 59 64 67 75 91 100 103 111 128 136 140 143 147 LIST or TABLES. Table Page 1. Numbers of labelled cells in sympathetic ganglia.......... SO 2. Effect of spinal cord transection on integrated sympathetic discharge.................................. 78 3. Effect of spinal cord transection on discharge rates of sympathetic nerves recorded simultaneously.......... 80 4. Average mean arterial pressure and heart rate before and after spinal cord transection...................... 81 5. Values of pH, PaCO and mean arterial pressure during normocapnia and hypercapnia before and after spinal cord transection................................ 84 6. Summary of the effects of splenic and intestinal receptor stimulation on discharge of splenic and renal sympathetic fibers........................... 117 A. Historical sketch of investigations of neural control of the cardiovascular system...........,.......... 140 viii LIST OF FIGURES Figure Page 1. Schematic diagrams of variations in the gross anatomy of ganglia of the solar plexus............... 49 2. High contrast photomontage of HRP label within two splenic sympathetic neurons...................... 51 3. The percent distributions of splenic and renal neurons among sympathetic ganglia of the thoracic chain (TC), lumbar chain (LC) and solar plexus (SP)................................ 53 4. Photomicrograph of peroxidase-labeled splenic (A) and renal (B) sympathetic neurons in the left celiac pole of the solar plexus ganglia of two cats............................................. 55 5. Camera lucida drawings showing the arrangement of labeled splenic neurons in the solar plexus ganglia of 4 cats............................. 56 6. Camera lucida drawings showing the arrangement of labeled left renal neurons in the solar plexus ganglia of 4 cats............................. 58 7. Photographs of oscillograph tracings of sympathetic nerve activity before and after spinal cord transeCtionOOOIOOOOOIOOOOOOOOOOOOOIOIO0.0 76 8. Ongoing discharge of sympathetic nerves before and after high cervical spinal cord transection...... 77 9. Effects of spinal cord transection on simultaneously recorded discharge rates of pairs of sympathetic nerves............................................... 82 10. Effect of hypercapnia on sympathetic discharge before and after spinal cord transection............. 86 11. Effect of increased arterial blood pressure on sympathetic discharge before and after spinal cord transection..................................... 86 12. Effect of hemorrhage on sympathetic discharge before and after spinal cord transection............. 88 13. Ongoing discharge of sympathetic nerves following dorsal rhizotomy in 6 cats........................... 90 ix Figure Page 14. Neuronal recording arrangement and methods of chemical stimulation of visceral afferent nerves...................................... 106 15. Distributions of ongoing discharge rates of splenic and renal sympathetic fibers................. 112 16. Cardiac and respiratory rhythmicities of a renal sympathetic fiber............................ 113 17. Assessment of cardiac rhythmicity in activity of two splenic sympathetic fibers....................... 114 18. Effects of arterial and cardiopulmonary pressoreceptor stimulation and unloading on discharge rates of renal and splenic sympathetic fibers................. 116 19. Group comparisons of splenic and renal unit responses to Visceral stimulation-00......0......0.00.00.00.00. 118 20. Oscillographic records of responses of discharge of one splenic and one renal sympathetic fiber to chemical stimulation of visceral afferent nerves..... 121 21. Paired comparisons of responses of sympathetic fibers to chemical stimulation of splenic and intestinal nerves...0.00.00.00.00...0.0.0.0...OOOOOOIOOOOOOOOOOO 123 22. Group comparisons of responses to visceral stimulation one hour after spinal cord transection............... 125 23. Paired comparisons of responses to visceral stimulation following spinal cord transection.................... 127 INTRODUCTION In the early twentieth century, the sympathetic nervous system was considered to respond to internal or external stimuli in a unitary fashion to provide a homeostatic mechanism for the maintenance of the internal environment of vertebrates (Cannon, 1930). This view influenced subsequent investigations in which the responses the sympathetic nerves to one organ, or responses of the adrenal medulla, were considered representative of the sympathetic nervous system as a whole. Had it been realized that sympathetic nerves could respond nonuniformly to various inputs, the Sherringtonian approach to studying nervous system function ("application of the reflex concept” in the analyses of neural systems) could have given information regarding the functional structure of the sympathetic nervous system. However, this was not understood until relatively recently. The concept of sympathetic mass-activation, rendered by Cannon (1930), initially was challenged by Folkow, Johansson, and L8fving (1961) and Ldfving (1961). These investigators made simultaneous measurements of the reflex changes in blood flow to hindlimb muscle and skin, kidney, and intestine that were initiated by stimulation of arterial chemoreceptors or by unloading of carotid baroreceptors. In these studies, renal and cutaneous blood flows were relatively unaffected whereas vessels of skeletal muscle constricted in response 2 to these perturbations; responses of intestinal blood flow were intermediate compared to the other vascular responses. These results were interpreted as an indication of reflex specificity among the sympathetic nerves supplying each vascular bed. However, the differences in regional distribution of blood flow may have arisen from heterogeneity of the neuroeffectors' sensitivities to equivalent magnitudes of sympathetic excitation. For example, blood vessels of different regions may have varying receptor densities (Bevan, 1979) leading to differences in vascular responses despite similar discharge rates of the sympathetic innervations. Therefore, simultaneous recordings of activity from different sympathetic nerves are necessary to distinguish between target responsiveness and specificity of the neural organization. - More recently, electrOphysiological studies have demonstrated nonuniform reactions of different sympathetic nerves in a variety of circumstances, such as during chemoreceptor or baroreceptor stimulation, or during exercise (Grignolo, Koepke, and Obrist, 1982; Iriki, walther, Pleschka, and Simon, 1971; Janig, 1982; Kollai and Koizumi, 1977; Ninomiya and Irasawa, 1975). These studies have been interpreted, keeping in mind the different responses of various vascular beds mentioned above, as indications of the ability of the sympathetic nervous system to influence the vasculature of different organs in a differential manner, while ignoring the fact that not all sympathetic nerve fibers influence vascular function. Most studies designed to investigate nonuniformity of visceral sympathetic responses have entailed multifiber recording techniques. This kind of data is difficult to interpret due to the lack of information pertaining to the 3 heterogeneity or homogeneity of responses of individual neurons within a given fiber bundle. Different fibers within individual visceral nerves may influence different target tissues within the innervated organ. For example, renal nerves are known to influence blood flow, renin secretion, and tubular reabsorption of electrolytes (DiBona, 1982); splenic nerves innervate arterioles and trabeculae, as well as the splenic capsule (Cleland and Tait, 1927; Fillenz, 1970; Utterback, 1944). Therefore, populations of axons within a nerve trunk may exhibit different responses to a given stimulus based on the type of target tissues innervated by the particular neurons in question. The following investigations will be included in this dissertation: 1. The distributions of renal and splenic sympathetic neuronal cell bodies within pre- and paravertebral sympathetic ganglia were compared and contrasted to provide the tapographic organization of renal and splenic neural cell bodies. This may be useful for future investigations in which horseradish peroxidase may be injected into specific ganglionic sites to label preganglionic neurons which project to different organs. 2. The origins of spontaneous sympathetic neural discharge of renal and splenic nerves were ascertained because it is not known whether discharge of these sympathetic nerves is dependant upon supraspinal sources of excitatory drive. 3. The reflex responses of single renal and splenic nerve fibers were compared following stimulation of systemic pressoreceptors and chemical stimulation of splenic and intestinal afferent nerves. This was done to quantify reflex responses of renal and splenic sympathetic 4 nerves in absolute units of measurement (number of action potentials per second). In addition, these data may provide information about heterogeneity or homogeneity of populations of individual axons contained within renal and splenic nerves. The literature review is focussed primarily upon data which provides evidence for the discrete neural control of individual vascular beds, individual organs or target tissues. The original data which comprises the body of the dissertation has been divided into three chapters. Each chapter contains an introduction, methods, results and discussion section pertaining to the project described in that chapter. A fourth chapter contains brief conclusions and a general discussion of the importance of the three studies. LITERATURE REVIEW EARLY HALLMARKS IN THE STUDY OF THE SYMPATHETIC NERVOUS SYSTEM Since Galen's time (130-200), the peripheral components of the sympathetic system have been known, by Western civilization, to innervate the viscera (Ackerknecht, 1974; Sheehan, 1936). During the next millenium, little seems to have been done to extend Galen's anatomical observations. Most studies of the nervous system.during the 16th, 17th, and 18th centuries were purely descriptive anatomical analyses, the interpretations of which were largely metaphysical. In those times, the brain was thought to be the source of "vital humours” or "sympathies" which were distributed throughout the organism to bring about harmonious action in, and reaction to, the external environment (Ackerknecht, 1974; Langley, 1916; Sheehan, 1936). Despite the lack of empirical data regarding the functions of the sympathetic system, valuable anatomical data was generated which served as the foundation for further study. Willis, in 1664, proposed that the viscera are not controlled voluntarily since he thought that the sympathetics descended from the cerebellum, rather than from.the cerebrum where volition was thought to reside (Ackerknecht, 1974; Sheehan, 1936). Remak was probably the first to recognize that the sympathetic fibers that innervate the viscera originate in cell bodies located in the sympathetic ganglia rather than in the central nervous system, although 6 it was widely accepted by this time that the ganglia were able to communicate with the central nervous system via the rami communicantes (Ackerknecht, 1974; Sheehan, 1936). Willis was the first to recognize the influence of sympathetic nerves on blood vessels. He thought that sympathetic nerves wrapped around blood vessels and mechanically caused constriction of the arteries resulting in a pulsatile increase in blood flow through the vessels (Sheehan, 1936). However, it wasn't until 1840 that Henle discovered muscle tissue in the walls of blood vessels and the close association of sympathetic nerves appearing to terminate in the muscle (Heymans and Neil, 1958; Sheehan, 1936). That same year, Stilling coined the term ”vasomotor system". However, eleven years elapsed before Claude Bernard and Brown-Sequard demonstrated independently that the sympathetics exert ongoing vasoconstrictor influences on blood vessels (Olmsted and Olmsted, 1952). Bernard later severed the spinal cord, observed the ensuing decreased arterial pressure, and concluded that ongoing activity of the sympathetic system which maintains vasomotor tone, is dependant upon activity arising from the brain. Owsjannikow and Dittmar, in the 1870's, extended these investigations by trying to locate a vasotonic center in the brain by observing the changes in arterial pressure produced by serial transections of the neuraxis (Gebber, 1982; Heymans and Neil, 1958). No consistent changes in blood pressure resulted from transections anterior to the pans. However, as transections were made progressively more posterior through the pons and medulla oblongata, arterial pressure drapped more and more, until the neuraxis was severed at the junction of the medulla and spinal cord. At this point, arterial pressure was the lowest; further 7 posteriorly placed transections did not result in a further decrease in blood pressure. The interpretation of these data was that a vasotonic center was located in the brain between midpontine and caudal medullary levels of the neuraxis. A table summarizing important hallmarks in the study of the sympathetic nervous system control of cardiovascular function, up to the work of Cannon, is presented in Appendix A. HOMEOSTASIS AND UNITARX ACTION OF THE SYMPATHETIC NERVOUS SYSTEM Focus upon the sympathetic system changed to encompass the pathways by which these brain areas exerted their influences on peripheral neural components as well as the influences of sympathetic nerves on visceral and cardiovascular function. Claude Bernard emphasized the physiological importance of maintaining the uniformity of the "milieu interne" or internal environment (Cannon, 1930). Cannon proposed the use of the term "homeostasis" to describe the ”general idea of uniformity or stability in the organism". He regarded the sympathetic division of the nervous system as an instrument indispensable for preservation of constant composition of this internal environment. Cannon reasoned that the sympathetic nervous system is organized to produce "widespread changes in smooth muscle and glands throughout the organism” (Cannon, 1930). This diffuse action of the sympathetic system is possible because of its "arrangement for simultaneous and unified action” (ibid;). Particularly, the extensive divergence of preganglionic innervation of postganglionic neurons, as well as of secretory cells of the adrenal medulla, would seem to ensure concordant physiological responses required to preserve homeostasis during mammals' 8 reponses to changes in the external environment. Cannon's postulated "en masse" action of the sympathetic nervous system has been used in many undergraduate, graduate, and medical textbooks to describe the "fight or flight” reaction in which the sympathetic system is known to participate. Although views of the uniformity of sympathetic action were strictly interpreted for decades, Cannon was aware of the dearth of information concerning influences of individual sympathetic nerves on particular target organs, as indicated by this statement in the Linarre Lecture (Cannon, 1930): ”In the present state of our knowledge of the nature of nerve impulses in the autonomic system no decisive conclusion can be drawn regarding the possible function of the outlying neurones as 'transformers,' adapting the nerve impulses received from the brain and spinal cord to the peculiar visceral structures which are affected. More information is needed. May I venture the suggestion that here lies an attractive realm for research, which may yield highly interesting new facts.” ONGOING ACTIVITY OF SYMPATHETIC NERVES Under control conditions, unevoked, repetitive discharge can be recorded from sympathetic nerves in anesthetized and unanesthetized animals and humans. This activity has been called tonic (ELEL’ Alexander, 1946), background (g;g;, Beacham and Perl, 1964), basal (SEER: Gebber, 1982), and ongoing (g;g;, Skok and Ivanov, 1983), and is necessary for the maintenance of arterial pressure. The presence of ongoing nerve activity enables the system to be finely controlled because activity can be increased or decreased to meet momentary demands 9 of the organism. In addition, ongoing activity of nerves increases the efficiency of target organ responses by ”priming" the effectors, thereby enabling quick target responses "without the lag inherent in building up a response in quiescent tissue" (Polosa, Mannard, and Laskey, 1979). Rhythmicity of Sympathetic Nerve Activity Cardiac-related Rhythms In the first electrOphysiological recordings of mammalian sympathetic nerve activity, Adrian, Bronk, and Phillips (1932) observed waxing and waning of sympathetic nerve discharge occurring in approximate synchrony with the period of cardiac or respiratory cycles. The sources of these rhythmic fluctuations in ongoing discharge observed in recordings of sympathetic activity have been a matter of speculation and much investigative effort ever since (3:5;9 Gebber, 1982). Adrian, et a1. thought the cardiac rhythm was due to the controlling influences of the phasic stimulation of arterial baroreceptors produced by systolic arterial pressure. This view was supported by Cohen and Gootman (1970) using crosscorrelation analyses. However, Gebber and coworkers (Barman and Gebber, 1980, 1981a,b; Gebber and Barman, 1980, 1981, 1982; McCall and Gebber, 1975; Taylor and Gebber, 1975) demonstrated that a periodic rhythm in sympathetic discharge persisted even after surgical denervation of carotid and aortic baroreceptors. The frequency of the rhythm was approximately the same as that of the cardiac cycle (2-6 Hz, in cats) but the fluctuations of sympathetic activity were no longer locked in a 1:1 relationship with the cardiac cycle. In addition, when 10 heart rate was slowed by a variety of pacing techniques in baroreceptor intact animals, baroreceptor afferent nerves continued to discharge in synchrony with the cardiac cycle, whereas rhythmic discharge of the sympathetic nerves was no longer locked to the cardiac cycle, but continued to be in the 2-6 Hz range. On the basis of these observations, Gebber has prOposed that sympathetic rhythmicity can occur in the absence of peripheral inputs and that baroreceptor afferent activity entrains this rhythm so as to produce a 1:1, or locked, relationship between the cardiac cycle and sympathetic activity. Respiratory-related Rhythms Most studies indicate that sympathetic nerves have a higher probability of firing during inspiration (Adrian et a1. 1932; Barman and Gebber, 1976; Cohen and Gootman, 1970; Gootmanand Cohen, 1971) and it has been presumed that brain stem inspiratory neurons facilitate (and/or brain stem expiratory neurons inhibit) sympathetic activity. Although lung stretch receptors with afferent axons in the vagi probably play a role in the genesis of respiratory rhythmicity of sympathetic nerves, this rhythm can still be observed when these sympathoinhibitory effects are eliminated by vagal transection (Adrian, et al., 1932). Prior to 1976, the respiratory periodicity of sympathetic discharge was thought to be due to direct coupling between the brainstem respiratory oscillator and neural circuits that contribute to ongoing sympathetic excitation (Cohen and Gootman, 1970; Gootman and Cohen, 1971; Roizumi, Seller, Kaufman, and Brooks, 1971; Preiss, Kirchner, and Polosa, 1975). If respiratory rhythmicity of sympathetic activity is simply due to 11 direct influences of the respiratory oscillator, then anything which disrupts the rhythmicity of phrenic nerve activity also should disrupt the respiratory-related rhythmicity of sympathetic activity identically. For example, the latency between peak phrenic discharge and peak sympathetic discharge should remain constant during changes in the length of intervals between the bursts of phrenic activity. In addition, elimination of phrenic nerve activity (g;g;_hyperventilation) should abolish the respiratory rhythmicity of sympathetic nerve discharge. However, Barman and Gebber (1976) provided evidence that the brainstem respiratory oscillator can be uncoupled from a presumedly separate sympathetic oscillator. First, changes in the phrenic activity cycle caused changes in the phase relationship between phrenic and sympathetic activities, which should not occur if the rhythm originates in the respiratory oscillator. Second, although the respiratory rhythmicity of sympathetic discharge should be eliminated by hyperventilationrinduced inhibition of phrenic nerve activity, these rhythms persist after hyperventilation sufficient to cause cessation of phrenic nerve activity. Therefore, these authors assert that the ”periodic components of sympathetic and phrenic nerve activity are generated by independent oscillators that normally are entrained to each other”. Since that time, others have provided evidence that at least a proportion of sympathetic pre- or postganglionic neurons are more likely to be active during the expiratory phase of the respiratory cycle (Bachoo and Polosa, 1986; Bainton, Richter, Seller, Ballentyne, and Klein, 1985). Therefore, some brain stem expiratory neurons may serve 12 to facilitate excitatory inputs to sympathetic neurons, either directly or indirectly. The functional significance of inspiratory-related versus expiratory-related sympathetic activity is not known. Functional Significance of Rhythmic Nerve Activity Ongoing activity of sympathetic nerves has physiological advantages, as mentioned above. The rhythmicity of nerve activity also may have functional consequences. A few investigators have compared the vascular effects of stimuli delivered to sympathetic nerves at constant rates to the effects of stimuli delivered in trains (to mimic the bursts of activity observed in ongoing discharge of sympathetic nerves). The average contractile responses (i.e. force of contraction) of arteries were significantly greater when stimuli were delivered in bursts at irregular frequencies than when stimuli were applied continuously at regular intervals (Nilsson, Ljung, Sjoblom, and Wallin, 1983). In addition, the magnitude of blood flow response were shown to be dependent upon 1) the frequency of impulses within a burst, and 2) the intervals between successive bursts (Andersson, 1983). If stimuli were delivered in bursts, maximal and maintained blood flow responses were produced, but if stimuli were delivered at regular intervals, responses were submaximal and faded. Andersson proposes that during the periods of quiescence, nerve terminals can take up or synthesize enough transmitter for repeated release, whereas continuous stimulation leads to depletion of transmitter stores due to the inadequate time for repletion. 13 Sources of Sympathetic Activity Maintenance of normal arterial pressure depends on intact pathways from the brainstem to spinal sympathetic centers; this has been known since the experiments of Bernard, Dittmar, and Owsjannikow (Gebber, 1982). Benson and Billingsley (1916) consequently probed the floor of the fourth ventricle with stimulating electrodes and measured the evoked changes in arterial pressure in an attempt to locate medullary areas important for maintenance of arterial pressure. When the medial medulla was stimulated near the obex, arterial pressure decreased, and when more lateral and more rostral sites were stimulated, arterial pressure increased. Wang and Ranson (1939) extended these observations by electrical stimulation of deeper medullary regions in addition to the dorsal surface. These investigators produced decreases in blood pressure by stimulation of the medial reticular formation and increased arterial pressure was elicited by stimulation of more lateral reticular areas. Although vasoactive points were not confined to specific nuclear groups in the reticular formation, these results were interpreted as evidence for the existence of discrete pressor and depressor areas. These results were confirmed by Alexander (1946) who monitored cardiac sympathetic nerve activity as well as arterial pressure. In addition to the pressor and depressor areas defined by previous investigators, Alexander elicited increases in arterial pressure and nerve activity from stimulation of the dorsomedial reticular formation. Since that time, others have explored the brainstem.with stimulating electrodes while recording discharge from sympathetic preganglionic neurons. Gootman and Cohen (1971) stimulated the dorsomedial reticular formation l4 (nucleus reticularis parvocellularis) and found short latency, large amplitude excitatory responses in splanchnic nerve activity. The next year, Nathan (1972) was able to elicit responses of even shorter latency from a more caudal area in the nucleus reticularis ventralis and proposed that neurons in this region comprised at least part of the final descending bulbospinal sympathetic pathway. However, as electrical stimulation can affect axons of passage as well as somata, this interpretation is equivocal. In addition to studies employing electrical stimulation of brainstem sites, lesions ”discretely placed” in the medulla have been used to identify loci important in cardiovascular control. Schlaefke and Loeschke (1967) used focal cooling of the rostral ventrolateral medulla to produce large decreases in arterial pressure. Dampney and Moon (1980) placed lesions in small areas of the rostral ventrolateral medulla and confirmed these results. Electrical stimulation of this region caused large increases in arterial pressure. Kumada et a1. (1979) found that destruction of dorsal medullary neurons results in lowered arterial pressure. Dampney and Moon (1980) suggested that the ventrolateral medullary neurons project through the dorsomedial reticular formation as the pathway descends to spinal sympathetic preganglionic loci because lesions placed in the dorsomedial reticular formation attenuated the increase in blood pressure elicited by stimulation of the ventrolateral medulla. To eliminate the confounding effects of lesioning or stimulating axons of passage, Guertzenstein and Silver (1974) used local applications of chemicals which act selectively on cell bodies and elicited decreases (glycine) and increases (glutamate) in arterial pressure. These results have been interpreted 15 as evidence that the ventrolateral medulla is at least one significant source of sympathetic activity (Dampney, 1981). Barman and Gebber (1983) have challenged the premise that ongoing sympathetic activity is generated exclusively in the ventrolateral medulla. Although these authors agree that these neurons may represent the final common descending pathway to sympathetic preganglionic neurons, they propose that other, presumably medullary, areas are antecedent to the ventrolateral medullary areas and therefore the ventrolateral medulla is more a relay station than a source for descending excitatory influences on the sympathetic system. This suggestion is based on studies in which the time lag between medullary unit action potentials and the peak in sympathetic activity have been compared when activity was recorded from neurons in dorsomedial and ventrolateral medullary areas. Ventrolateral neurons fired, on average, approximately 50 me later than dorsomedial neurons in relation to the peak discharge of sympathetic postganglionic nerves (Barman and Gebber, 1983). In addition, different pathways descend in parallel from a number of levels of the neuraxis to innervate neurons in the spinal sympathetic nuclei (Peiss, 1965). Determination of direct projections from supraspinal neurons to sympathetic preganglionic neurons has proved to be difficult despite advances in anatomical technology. Utilizing the retrograde transport of horseradish peroxidase (HRP) from injection sites restricted to the intermediolateral cell columns of thoracic and lumbar segments of spinal cord, Amendt, et al. (1979) located cell bodies of bulbospinal neurons in the nucleus tractus solitarius, ventrolateral reticular formation, and ventral portions of the raphe nuclei. Neurons near the surface of the ventrolateral medulla were labeled with horseradish peroxidase 16 following injections into both the intermediolateral cell column and central autonomic nuclei of the spinal cord (Caverson, et al., 1983b). The descending projections from these areas have been confirmed using orthograde transport of radio-labeled amino acids (see Barman, 1984). Another experimental approach used to identify potential sources of ongoing sympathetic activity has been to stimulate sites within the preganglionic cell columns of the lateral horn and record antidromic action potentials from brain sites. Henry and Calaresu (1974) were able to antidromically activate cells in several brainstem areas: nucleus reticularis parvicellularis, gigantocellularis, paramedian reticular nucleus, lateral reticular nucleus, raphe nuclei, and the inferior olivary nucleus. Electrical stimulation of these areas caused increased arterial pressure and heart rate (Henry and Calaresu, 1974). Others have antidromically activated neurons near the surface of the ventrolateral medulla by stimulation of the intermediolateral cell column or central autonomic area (Barman and Gebber, 1985; Caverson, Ciriello, and Calaresu, 1983a,b). Many of these ventrolateral medullary neurons responded to stimulation of the carotid sinus and aortic depressor nerves (Caverson, et al., 1983a), indicating their possible role in cardiovascular control. The possibility that discrete medullary areas are dedicated to particular vascular beds is discussed in section E.2. In addition to descending excitatory drive of sympathetic nerve activity, evidence indicates that ongoing inhibitory influences also participate in the control of sympathetic discharge. Although many investigators have demonstrated the existence of descending inhibitory pathways from the medulla by electrical stimulation of the classical 17 depressor areas of the mediocaudal medulla (Eifiiw Coote and Macleod, 1974a; Gootman and Cohen, 1971; Henry and Calaresu, 1974), few have investigated ongoing influences of descending inhibition. Alexander (1946) made serial transections of the brain stem and observed that arterial pressure was decreased progressively until maximal depression was achieved as the lower medulla was transected. Then, when the neuraxis was severed at the first cervical spinal cord segment, arterial pressure and sympathetic activity actually increased by a small amount. Thus he showed that the lower medulla exerted a Eggig descending inhibitory influence on sympathetic outlow. These results have been confirmed, in principle, by others using more modern techniques. Dembowski et al. (Dembowski, Czachurski, Amendt, and Seller, 1980; Dembowski, Lackner, Czachurski, and Seller, 1981) showed that the spinal components of somatosympathetic reflexes were exaggerated following reversible cold block of descending pathways within the cervical spinal cord. Furthermore, these effects were not altered by baroreceptor denervation or by midcollicular decerebration, indicating that this descending inhibition was not of baroreceptor origin (Dembowski, et a1. 1980). These ongoing inhibitory influences were thought to be mediated by descending catecholaminergic neurons situated in ventrolateral areas of the medulla (Dembowski, et al. 1981). In addition, chemical inhibition of neurons in the caudal ventrolateral medulla can cause increased arterial pressure and heart rate (Blessing, west, and Chalmers, 1981) as well as increased activity of renal nerves (Pilowsky, West, and Chalmers, 1985). The ongoing sympathoinhibition 18 deduced from these studies was thought to originate in the medullary region of Al, noradrenaline-containing neurons (Blessing, Chalmers, and Howe, 1978). REFLEX CHANGES IN SYMPATHETIC ACTIVITY Pressoreceptors Carotid Sinus and Aortic Arch Slowly adapting mechanoreceptors are embedded in the adventitia of the walls of the aortic arch and at the bifurcation of the common carotid arteries. These receptors are sensitive to deformations of the vessel walls caused by increased transmural pressure which results in circumferential stretch of the vessels (Korner, 1979). Changes in arterial pulse pressure produce changes in vessel radius (deformation) due to the elasticity and geometry of the vessel walls, and, consequently, stimulation of the mechanoreceptors. The receptors have been described morphologically as diffuse arborizations or ”circumscribed glomerular-like structures” (Heymans and Neil, 1958) at the ends of the afferent nerve fibers, the cell bodies of which are located in the petrosal ganglia. Axons of the sinoaortic baroreceptors can be myelinated (ca. two thirds of the fibers) or unmyelinated. Stimulus thresholds of the myelinated axons are between 40 and 70 mmHg, and the neurons increase their rates of activity in direct, linear proportion to the intensity of stimulation at arterial pressures of 75 to 150 mmHg; maximum discharge rates are elicited by pressures of 175 to 19 200 mmHg (Abboud and Themes, 1983). Unmyelinated fibres exhibit higher thresholds, higher saturation pressures, and lower sensitivities than do the myelinated afferent fibers (Abboud and Thames, 1983). As not all baroreceptor afferent neurons exhibit ongoing discharge at normal arterial pressures (Kircheim, 1976), increased activity of these nerves can be caused by recruitment of inactive neurons by increasing arterial pressure past the stimulus thresholds, as well as by increasing the discharge rates of active fibers by suprathreshold increases in blood pressure. The sympathoinhibition produced by stimulation of arterial baroreceptors, first directly assessed by Adrian, et al. (1932), is known (Abboud and Thames, 1981; Heymans and Neil, 1958; Kircheim, 1976; Korner, 1979; Spyer, 1981). The location of the first synapse in the baroreceptor mediated sympathoinhibitory reflex arc is the least controversial link in the entire pathway. Whereas recent authors agree that the primary baroreceptor afferents in the cranial nerves terminate in the NTS, the precise pathway(s) by which pressoreceptor-mediated inhibition of sympathetic activity remain in dispute or are unknown (Abboud and Thames, 1981; Dampney, 1981; Rircheim, 1976; Rorner, 1979; Spyer, 1981). Although central projection sites of the sinoaortic activity (beyond the NTS) have been investigated by numerous investigators, results are equivocal simply due to technical problems. For example, it is very difficult, if not impossible, to lesion cell bodies of neurons without also destroying fibers of passage. Therefore, if lesions are placed in the dorsomedial NTS with the intention of correlating the paths of fiber degeneration with specific NTS projection sites, destruction of passing fibers may lead to an incorrect 20 assessment. The same caveats apply to tract-tracing techniques utilizing HRP and radiolabeled amino acids. This necessitates the use of electrOphysiological studies to corroborate anatomical investigations. However, care must also be taken to interpret electrOphysiological data because baroreceptor activity may have effects on parts of the nervous system which seem to have nothing to do with the inhibition of sympathetic nerve activity (Spyer, 1981). Therefore, it is not possible to define the precise pathway(s) by which baroreceptor stimulation results in inhibition of sympathetic activity. Despite these problems associated with the identification of central baroreceptor pathways affecting sympathetic outflow, many investigations have been carried out to describe baroreceptor projections within the central nervous system. Projections from medial NTS to various areas have been demonstrated, including a direct projection to sympathetic preganglionic nuclei in the lateral horns of the thoracolumbar spinal cord (Loewy and Burton, 1978). In addition, neurons in the medial NTS, with activity related to the period of the cardiac cycle, project to the rostral ventrolateral medulla and to dorsomedial reticular areas. However, Barman and Gebber (1978) have demonstrated sympathoinhibition of baroreceptor origin following large lesions placed in the dorsomedial medulla (encompassing the paramedian reticular nucleus as well as the caudal raphe nuclei). Granata et a1. (1985) have provided evidence that baroreceptor-mediated sympathoinhibition utilizes rostral ventrolateral medullary neurons (the Cl area) containing phenylethanolamine-N-methyl transferase (PNMT), the enzyme necessary for the synthesis of epinephrine. Neurons in the Cl area project to spinal cord locations of sympathethic preganglionic cell bodies (Ross et al., 1984) In addition, 21 electrolytic lesions in the C1 area abolish the depressor response to carotid sinus baroreceptor stimulation (Granata, et al., 1985). Similar results of studies on rabbits (Dampney, 1981a, 1981b) support the conclusion that this area mediates the barorecptor reflex. That these PNMT-containing neurons do not seem to comprise a specific cytoarchitecturally defined nucleus may well be the reason for the lack of precise anatomical data pertaining to the baroreflex. Investigators have attempted to define locations of the inhibitory synapses in the pathways. Spinal (Coote and Macleod, 1981; McCall, Gebber, and Barman, 1977; Taylor and Gebber, 1973), as well as supraspinal (Biscoe and Sampson, 1970; Kirchner, Sato, and weidenger, 1971), inhibitory sites have been documented. Reflex Influences of Baroreceptors on Peripheral Taggets Folkow, Johansson, and L8fving (1961) made the first systematic investigation of baroreceptor influences on different vascular beds. These investigators made simultaneous measurements of the reflex changes in blood flow to hindlimb muscle and skin, kidney, and intestine that were initiated by unloading of carotid arterial baroreceptors. In these studies, renal and cutaneous blood flows were relatively unaffected whereas skeletal muscle vessels constricted in response to these perturbations; intestinal blood flow decreased less than skeletal muscle blood flow. These results were confirmed by Kendrick et a1. (1972) and by Pelletier and Shepherd (1975). Brender and Webb-Peploe (1969) found that baroreceptor stimulation caused decreased blood flow in hindlimb resistance vessels, decreased venous pressure in the isovolumetric 22 spleen, and produced no consistent changes in hindlimb capacitance vessels. The greater effects of baroreceptor stimulation on renal than cutaneous vasoconstrictors have been confirmed in recordings of renal and cutaneous sympathetic discharge (Ninomiya, et al., 1973). Cardiac sympathetic activity can be more inhibited than renal by stimulation of arterial baroreceptors (Ninomiya, et al., 1971). In addition to the differential effects of baroreceptor influences on distant components of the cardiovascular system, Ninomiya and coworkers have provided evidence that different vascular beds within the abdomen are affected differently by these cardiovascular afferent nerves. Baroreceptor stimulation caused greater inhibition of splenic than renal sympathetic nerve discharge (Ninomiya, et al., 1971). These results have been challenged by Tobey and Weaver (1987) who elicited greater inhibition of renal than splenic sympathetic activity by phenylephrine-induced increases in arterial pressure. It now is clear that baroreceptor reflexes must be considered individually with each particular target organ in mind. In addition, recording activity of single sympathetic neurons during baroreceptor stimulation may resolve differences of opinion regarding the organization of baroreceptor influences on individual target organs or, possibly, even on different tissues within an individual organ. Cardiopulmonary and Thoracic Vasculature Pressoreceptors Terminals of sensory fibers that innervate the heart are unencapsulated nerve endings within the subendocardium (Malliani, 1982). The axons of these afferent neurons are contained in the vagi (cell 23 bodies in the nodose ganglia) as well as in cardiac sympathetic nerves (cell bodies in spinal dorsal root ganglia), and may be either myelinated or unmyelinated. The myelinated vagal afferent axons terminate at the junctions of the great veins and atrium and increase their rates of discharge in response to increased cardiac volume. Other myelinated vagal afferent axons innervate the atria and respond more to increased contractility than to volume loads per se; these fibers exhibit ongoing activity during atrial systole. In addition, myelinated vagal afferent fibers also terminate in the ventricular endocardium and discharge in synchrony with ventricular systole; these fibers respond to changes in cardiac distention or contractility. Approximately three fourths of cardiac vagal afferent fibers are unmyelinated (Donald and Shepherd, 1978). These C fibers are either mechanoreceptive or chemically sensitive, but are not thought to be polymodal. Mechanoreceptor C fibers innervate the ventricles to a greater extent than the atria and their discharge rates are preportional to end-diastolic pressure. Some C fibers terminate in the epicardium of the left ventricle and respond to changes in contractility or ventricular volume. The pulmonary vasculature also is innervated by vagal afferent nerves that respond to changes in arterial and venous pressures. The pulmonary arterial baroreceptors exhibit characteristics of ongoing and reflex discharge similar to those of sinoaortic baroreceptors. These pulmonary afferent nerves have small, myelinated axons (Ardelta) and are active at normal pulmonary arterial pressure. Pulmonary vagal afferent neurons with unmyelinated (C) fibers 24 innervate pulmonary capillaries and/or veins and exhibit low discharge rates that have no consistent relationship to the cardiac cycle (Abboud and Thames, 1983). Cardiopulmonary vagal afferent nerves exert ongoing influences on sympathetic nerve activity; interruption of these afferent nerves causes increased vascular resistance in kidney, intestine, and skeletal muscle (Donald and Shepherd, 1978). This tonic sympathoinhibition is believed to be mediated by unmyelinated vagal afferent fibers. Mechanical stimulation of vagal afferent neurons causes increased cardiac, decreased renal, and no change in splenic or lumbar sympathetic nerve activity (Karim, Kidd, Malpus, and Penna, 1972), as well as diuresis (Gauer and Henry, 1976). Cardiac afferent axons contained within sympathetic nerves (cardiac sympathetic afferent nerves; cardiac spinal afferent nerves) are similar to those within the vagi, but many have been shown to be associated with polymodal receptors responsive to chemical £22 mechanical stimuli. Spinal afferent innervation of the heart is distributed predominantly to the ventricles, although the atria do have some receptors with spinal afferent axons (Donald and Shepherd, 1978). Pulmonary spinal afferent nerves exhibit ongoing activity which occurs during the systolic phase of pulmonary arterial or venous pressure. The arterial afferent nerves have lower discharge rates than do the venous afferents which ”appear particularly suitable to sense pulmonary congestion” (Malliani, 1982). In addition, ongoing, pulse-synchronous activity has been recorded from spinal afferent neurons that innervate the aorta; these neurons respond to increased blood pressure (Malliani, 1982). 25 In contrast to the effects of vagal afferent stimulation, cardiopulmonary spinal afferent nerves generally are thought to mediate increased sympathetic efferent activity and arterial pressure (Donald and Shepherd, 1978). These effects can be observed in animals following acute transection of the spinal cord (Malliani, Lombardi, Pagani, Recordati, and Schwartz, 1975). Chemical stimulation of cardiac spinal afferent nerves can produce increased renal nerve activity and, consequently, antidiuresis and increased retention of sodium and potassium (Heckler, Macklem, and Weaver, 1981). However, Weaver, SE ‘gl;_have demonstrated that cardiac spinal afferent nerves have high stimulus thresholds, as very large intracardiac pressures or deformations of the ventricular walls are necessary to evoke reflex effects in recordings of renal nerve activity (Weaver, Hacklem, Reimann, Heckler, and Oehl, 1979). It has been suggested that these afferent nerves exert more control during conditions of myocardial ischemia (Weaver, Danes, Oehl, and Heckler, 1981) or during chemical stimulation (Reimann and Weaver, 1980) of the afferent nerves, than during increased intracardiac pressure. Injections of vasoactive drugs such as phenylephrine and nitroprusside evoke changes in activity of many cardiovascular afferent nerves. The summation of reflexes caused by such perturbations is important, and individual reflex effects of stimulation of isolated groups of receptors probably are not particularly meaningful under normal physiological circumstances. Therefore, it is as important to evaluate the effects of general increases in pressure as it is to assess the individual contributions of each group of pressoreceptors to cardiovascular control. The combined effects of stimulation of caretid, 26 aortic, and cardiopulmonary vagal afferent pressoreceptors by large increases in arterial pressure all contribute to the ensuing sympathoinhibition (Guo, Thames, and Abboud, 1982). The carotid sinus baroreceptor afferent nerves appear to be dominant in this response (Guo, et al., 1982; Hancia, Donald, and Shepherd, 1976). However, any one set of pressoreceptors does compensate for the absence of the other set of receptors in producing sympathoinhibition (Guo, et al., 1982). This apparent contradiction has been resolved partially by Thames and Ballon (1984) who have presented descriptive evidence that concomitant activation of carotid and aortic baroreceptors does not produce much more sympathoinhibition than does activation of either input separately because activation of each input alone can cause almost maximal inhibition. Although pressoreceptive cardiac spinal afferent nerves are less likely to exert ongoing control of sympathetic activity when cardiac vagal afferent innervation remains intact (Abboud and Themes, 1983), cardiac spinal afferent nerves can have significant excitatory influences despite simultaneous inhibitory influences of vagal afferent nerves. Chemical stimulation of cardiac vagal and spinal afferent nerves simultaneously can produce inhibition, excitation, or no change in renal nerve activity (Reimann and Weaver, 1980). In addition, when both groups of afferent nerves are stimulated simultaneously by application of algogenic substances to the heart, renal nerve activity often is inhibited, while splenic nerve activity is increased (weaver, Fry, and Heckler, 1984). These results indicate that sympathetic outflow to different viscera can respond selectively to summation of multiple afferent influences. 27 Abdominal Visceral Pressoreeptors Gammon and Bronk (1935) described pacinian corpuscles closely associated with arteries supplying the small intestine. Many of the afferent axons innervating these receptors were shown to discharge during the systolic phase of the cardiac cycle, suggesting that they might play a role in transmitting information related to the arterial pressure in this region. Since that time, others have reported the presence of ”baroreceptors” in a variety of viscera, including the kidney (Beacham and Kunze, 1969), liver (Kostreva, Castaner, and Kampine, 1980), and spleen (Herman, et al., 1982). Although cardiovascular and sympathetic neural responses have been elicited by stimulation of these visceral pressure receptors, the significance of such reactions during systemic injections of pressor agents (such as phenylephrine) remains obscure (Thames and Ballon, 1984). Recently, Martin and Longhurst (1986) have conducted experiments to compare abdomdnal visceral afferent nerve fibers to arterial baroreceptors and refute the existence of ”high pressure baroreceptors" in the abdominal viscera. Although these authors describe l9 afferent fibers with ongoing activity related to the cardiac cycle, similar to the afferents described by Gammon and Bronk (1935), Martin and Longhurst could not produce consistent responses in these neurons by alterations in hemodynamic variables such as increased or decreased arterial pressure. Martin and Longhurst (1986) assert that these afferent neurons most likely respond to deformation of nearby tissue, changes in tissue fluids, and/or chemicals released into the local environment. 28 Viscerosympathetic Reflexes The vasculature and parenchyma of abdominal organs are innervated by afferent nerves with axons in the splanchnic nerves and vagi. Stimulation of visceral afferent nerves innervating gall bladder, intestine, kidney, liver, pancreas, spleen, and stomach, with few exceptions, causes excitation of sympathetic nerve activity, increased heart rate and dP/dt, total peripheral resistance, and arterial pressure. The primary afferent pathway for the majority of these reflexes is the greater and lesser splanchnic nerves. Afferent axons contained in the splanchnic nerves of cats have somata in the dorsal root ganglia of spinal segments T2 through L2 (Janig and Morrison, 1986). These afferent nerves project to Lissauer's tract in the dorsolateral spinal cord and travel rostrally and caudally to adjacent spinal segments where they terminate in Rexed's laminae V-VII and X. Cells in these areas, in turn, project to the brain in spinoreticular and spinothalamic tracts. Some of the secondary afferents project rostrally in the dorsal columns and terminate in the gracile and cuneate nuclei. Termination of at least some renal primary afferent neurons occurs in the medulla oblongata (Simon and Schramm, 1984). In addition to their supraspinal projections, visceral afferent nerves are known to influence spinal neurons which affect sympathetic preganglionic neurons via solely spinal pathways. It is not known whether the ascending pathways are viscerotopically organized, but some authors doubt the specifity of visceral afferent projections on the basis of the convergence of visceral sensory neurons upon cells with somatic afferent inputs, and the phenomenon of referred pain. However, at least the 29 autonomic components of some reflexes initiated from the viscera can be engaged preferentially. The present discussion will be limited to splenic and intestinal spinal afferent nerves. The interested reader will find information regarding the sensory innervation of these and other abdominal and pelvic organs in the reviews contained in a recent volume of Progress in Brain Research (Cervero and Morrison, 1986). Injections of algogenic chemicals into splenic or superior mesenteric arteries of lightly anesthetized animals produces the ”pseudaffective response": hyperpnea, increased arterial pressure, and vocalization (Guzman, Braun, and Lim, 1962; Moore and Singleton, 1933a). Floyd and Morrison (1974) recorded activity of splanchnic afferent fibers during mechanical perturbations of the spleen and intestine and described punctate receptive fields usually associated with blood vessels. The intestinal and splenic afferent fibers in the greater splanchnic nerve were unmyelinated as well as myelinated and often exhibited ongoing discharge which sometimes had cardiac or respiratory rhythms. The primary afferent pathways are the greater splanchnic nerves and lumbar splanchnic nerves for the spleen and intestine, respectively (Moore and Singleton, 1933b). Supraspinal neural circuits apparently are not necessary for some reflex responses to mechanical or electrical stimualtion of these nerves since intestinal vasoconstriction and increased arterial pressure can be elicited after the spinal cord is severed (Downman and McSwiney, 1946). Beyond that, little seems to be known about the anatomy of the afferent pathways. Electrical stimulation of the mesenteric afferent nerves causes greater excitation of visceral (splanchnic preganglionic and mesenteric efferent) sympathetic activity than cutaneous or skeletal muscle 30 vasoconstrictor discharge (Koizumi and Suda, 1963). Johansson and Langston (1964) found that electrical stimulation of superior mesenteric afferent nerves caused vasoconstriction in the kidney and dilation of skeletal muscle vasculature. Later, others recorded activity of single preganglionic neurons during electrical stimulation of the mesenteric nerve and found that not all preganglionic neurons were involved in the reflex, even if the preganglionic neurons exhibited spontaneous activity (Fedina, Katunskii, Khayutin, and Mitsanyi, 1966; Franz, Evans, and Perl, 1966). Distension of the small intestine to excite - mechanoreceptors causes greater excitation of mesenteric than renal or splenic sympathetic activity (Ninomiya and Irisawa, 1975; Ninomiya, Irasawa, and Woolley, 1974). Few studies of the effects of splenic afferent nerve stimulation on sympathetic outflow have been done. Electrical stimulation of splenic afferent nerves causes increased arterial pressure, heart rate, dP/dt, and excitation of renal and cardiopulmonary sympathetic nerve discharge (Herman, Kostreva, and Kampine, 1982). In addition, these investigators could elicit similar responses by pinching the spleen to alter intrasplenic pressure, leading them to conclude that splenic afferent nerves could be the afferent limb of a low-pressure baroreceptor reflex. Calaresu, Tobey, Heidemann, and Weaver (1984) demonstrated that chemical stimulation of splenic afferent nerves can produce greater excitation of splenic than renal sympathetic efferent nerve activity. Similar results could be achieved whether arterial baroreceptor and cardiopulmonary vagal afferent nerves remained intact or were severed (Tobey and Weaver, 1987), indicating that the pressoreceptor stimulation, produced by the reflex increase in arterial pressure, was not necessary for unequal 31 splenic and renal nerve responses. Calaresu et a1. speculated that the greater splenic than renal sympathetic responses to splenic stimulation may reflect the functional organization of these reflexes to provide ”the largest output to the organ from.which the reflex originates”. These authors gave no indication of the meaning of this interpretation: do a greater number of splenic than renal sympathetic neurons participate in the excitatory reflex, or is the activity of individual splenic neurons increased more than that of individual renal neurons? FUNCTIONAL SPECIFICITY OF THE SYMPATHETIC NERVOUS SYSTEM Selective Connections in Sympathetic Gamglia Langley was the first to demonstrate that different target organs could be activated selectively by electrical stimulation of the preganglionic axons contained within white rami of different spinal segments. Thus, electrical stimulation of the white rami of spinal segments T1 and T2 caused pupilary constriction, and contraction of the nictitating membrane; stimulation of T2 and T3 white rami led to vasoconstriction in the head; stimulation of T2 through T4 produced cardioacceleration; stimulation of T5 and T6 caused piloerection in the face and neck; stimulation of T6 through T8 produced responses of the sweat glands in the ipsilateral forefoot (Langley, 1892). Although segmental distributions of these effects were overlapping, it was clear that preganglionic fibers originating from different spinal segments preferentially influenced particular target areas. These results were confirmed later by N19 and Purves (1977) and extended by analyses of 32 intracellular recordings of postganglionic neuronal discharge. In addition to the segmental organization of preganglionic influences on different neuroeffector targets, individual postganglionic neurons are innervated by preganglionic neurons from distinctive sets of spinal cord segments (Purves, 1978). A particular postganglionic neuron is innervated predominantly by preganglionic neurons originating from one spinal segment. Additional preganglionic inputs are provided to this ganglion cell by spinal segments immediately adjacent to the dominant segment. Thus, for example, a neuron can be dominated by input from preganglionic cells from T4, and fewer preganglionic inputs to this ganglion cell originate from segments T3 and T5. As one proceeds farther and farther from the dominant spinal segment, activation of a postganglionic cell becomes progressively less likely. The selectivity of segmental influences of preganglionic cells on individual ganglion cells may be organized functionally to provide the selective activation of end organs with similar function (such as vasoconstrictor or pilomotor). This possibility will be discussed further in the following section. Does the position of postganglionic neurons within a given ganglion determine the preganglionic inputs received by the ganglion cells? Purves offers evidence that position of cells in a ganglion is not an important factor in the determination of pre- to postganglionic connectivity. He found that injections of horseradish peroxidase into particular targets (2:3;2 eye or ear) produced an apparently random distribution of labeled cells within the superior cervical ganglion. Therefore, selective segmental inputs to this ganglion were shown to exist in the absence of topographical orientation of postganglionic 33 neurons within the ganglion. However, others have speculated that clustering of postganglionic cell bodies which innervate individual targets does exist and that the topographical arrangement of these cell bodies may be conducive to selective preganglionic inputs (Archakova, Bulygin, and Netukova, 1982; Kelts, Whitlock, Ledbury, and Reese, 1979; Kuntz, 1938). Celiac ganglion cells in close proximity to one another often have dendrites which are incorporated into common fiber bundles (Archakova, et al., 1982; Kuntz, 1938). Terminal branches of preganglionic axons arborize within these bundles to provide circumscribed groups of postganglionic cells with a common preganglionic innervation (Archakova, 22.2i" 1982; Kelts, 25.2l3’ 1979). Therefore, the position of postganglionic neurons within a given ganglion may be important in the determination of preganglionic influences on peripheral targets. The distribution of neurons among different ganglia may be another important factor which determines the selectivity of preganglionic innervation of the ganglion cells. Postganglionic neurons in different thoracic paravertebral ganglia are innervated selectively by preganglionic neurons originating from different segments of the spinal cord (Lichtman, Purves, and Yip, 1980). Thus, thoracic chain ganglia from the stellate ganglion to the T5 ganglion receive preganglionic innervation from different sets of spinal cord segments. For example, the stellate ganglion is innervated by preganglionic neurons originating from spinal cord segments more rostral than segments which provide inputs to the T5 ganglion. Interestingly, postganglionic neurons also are segmentally distributed in sympathetic ganglia with respect to the 34 target tissues. Thus, the cell bodies of hindlimb muscle vasocon- strictors are located in ganglia of segments T13 to L3; cutaneous vasoconstrictors to the hindpaw are in L1 to L3; cutaneous vasoconstrictors to the tail are in L3 and L4; postganglionic neurons which innervate the lower abdominal and pelvic viscera are distributed in ganglia L3 to L5 (Jfinig, 1986). In addition, preganglionic cell bodies of individual spinal cord segments may be organized anatomically with respect to the target tissue innervated by their postganglionic counterparts. For example, preganglionic neurons situated in the border region between the grey matter of the lateral horn and the lateral funiculus of spinal cord segment L4, project to the caudal sympathetic trunk and innervate neurons which supply hindlimb vasculature (Jdnig and McLachlan, 1986; McLachlan, Oldfield, and Sittiracha, 1984). In contrast, preganglionic neurons in more medial areas of the L4 grey matter innervate postganglionic neurons in the inferior mesenteric ganglion, many of which regulate visceral motility of the lower abdominal organs (Janig and McLachlan, 1986; HcLachlan,pg£_gl,, 1984). In summary, preganglionic neurons originating from different spinal cord segments make selective connections with postganglionic neurons based upon 1) the particular ganglion containing the ganglion cell, 2) the position of the postganglionic some within a sympathetic ganglion, and 3) the particular target organ innervated by the postganglionic neuron o 35 Electrophysiolqgical Evidence for Functional Organization of the Sympathetic Nervous System Whereas the sympathetic nervous system is known to be organized to provide discrete control of blood flow to different organs, Jdnig and coworkers have presented much evidence for the functional organization of sympathetic outflow to different target tissues, even within an individual organ (JHnig, 1985, 1986). Initially, investigations were confined to sympathetic control of hindlimb blood vessels (cutaneous and skeletal muscle vascular beds), sweat glands, and hair. By careful comparison of responses of effector organs and single sympathetic fiber activity to natural stimulation of afferent nerves, and by the ongoing discharge patterns of the neurons, Jfinig has formulated criteria for the functional identification of sympathetic neurons. For example, sudomotor neurons and vasoconstrictor fibers to skin and skeletal muscle exhibit ongoing discharge whereas pilomotor neurons and vasodilator fibers to skin and skeletal muscle do not discharge under resting conditions. Further, ongoing activity of muscle vasoconstrictors shows strong correlation to the arterial pressure pulse and is strongly inhibited by stimulation of carotid sinus baroreceptors. In contrast, ongoing activity of cutaneous vasoconstrictors is only weakly correlated with arterial pressure, and does not respond to increased carotid sinus pressure; activity of sudomotor neurons has no correlation with arterial pressure and is not affected at all by baroreceptor stimulation. The hindlimb pilomotor and vasodilator neurons can be activated only by very specific stimuli or in specific behavioral contexts. For example, stimulation of hypothalamic defence areas or emotional stimuli (such as 36 confrontation) elicits activity from previously quiescent muscle vasodilator neurons, whereas no reflex effects are observed in response to stimulation of baroreceptors, chemoreceptors, or any somatic stimuli (Janig, 1985). Cutaneous vasodilator neurons could be activated only by warming of the spinal cord (Jdnig, 1985). In addition to the functional classification of postganglionic neurons, Jfinig has been able to find preganglionic neurons with activity patterns to match those of the postganglionic fibers. These data suggest that pre- and postganglionic neurons comprise functionally dedicated pathways to different effector organs. JHnig has broadened the scope of his investigations by considering lumbar sympathetic outflow to the lower abdominal and pelvic viscera. His interpretations are the same for the sympathetic innervation of these areas as for hindlimb sympathetic outflow (Janig, 1986a,b). Neurons with pulse-rhythmic activity and that are weakly excited by visceral afferent stimulation are vasoconstrictors. In contrast, cells that respond strongly to visceral stimuli and have no pulse wave correlation are thought to mediate regulation of visceral motility and secretion. Apparently, there are functionally dedicated pathways to lower abdominal and pelvic visceral vasculature and parenchyma. Therefore, according to Jdnig, the sympathetic nervous system has a more complex organization than had been anticipated previously, to provide discrete control of different effector organs under the differing demands encountered by behaving mammals during challenges to homeostasis. The possibility that medullary, as well as spinal, neurons are dedicated to particular vascular beds, has been investigated in 37 relatively few laboratories. Barman, Gebber, and Calaresu (1984) provided evidence that medullary reticular and raphe neurons nonuniformly affect the discharge patterns of renal, cardiac, and external carotid sympathetic nerves. A given brain stem neuron could be temporally related more strongly to the activity pattern of one than another of two sympathetic nerves. Although many medullary units were related similarly to discharge patterns of more than one sympathetic nerve, 502 of reticular and 302 of raphe neurons were more tightly correlated with the discharge of a particular sympathetic nerve. In addition, Barman and Gebber (personal communication) demonstrated that many ventrolateral medullary neurons project to the intermediolateral areas of particular spinal cord segments, whereas some neurons project to multiple segments of the spinal cord, indicating some degree of specificity of descending pathways. Similar descending trajectories of bulbospinal sympathoinhibitory neurons of the raphe nuclei have been described (Morrison and Gebber, 1985). Utilizing injections of excitant amino acids (such as D,L-Homocysteic acid, DLH) into the ventrolateral medulla (nucleus paragigantocellularis lateralis, PGL) to excite cell bodies without affecting axons of passage, Lovick (1985) demonstrated patterned changes in heart rate and vascular conductance of hindlimb arteries. Differing patterns of responses were elicited from chemical stimulation of different sites within PGL. Lovick preposed that specific neurons in PGL are dedicated to control of particular functional components of sympathetic outflows; the response pattern apparently were dependent upon the relative numbers of functionally different neurons excited during particular injections of the excitant amino acids. During simultaneous recordings of activity of sympathetic 38 fibers supplying cutaneous and skeletal muscle vasculature, McAllen and Dampney (1986) chemically stimulated sites in the ventrolateral medulla and produced differentiated sympathetic responses. Stimulation of medial areas of the subretrofacial nucleus produced excitation of cutaneous sympathetic activity, whereas stimulation of lateral areas of the nucleus caused excitation of sympathetic outflow to skeletal muscle; chemical stimulation of neurons in intermediate sites activated both components of sympathetic outflow. These results indicate the topographical organization of these ”pre-sympathetic” medullary neurons. As well as the identification of functional specificity of reflex discharge in vivo, it has been possible to classify sympathetic neurons on the basis of their firing characteristics produced in vitro. By their responses to intracellular depolarizing current, sympathetic ganglion cells of the guinea pig and cat have been classified as 'tonic' or 'phasic' (Cassell, Clark, and McLachlan, 1986; Decktor and weems, 1981, 1983; Hartman and Krier, 1984; Kreulen and Szurszewski, 1979; Weems and Szurszewski, 1978). Tonic cells are characterized by maintained repetitive discharge throughout a period of injecting a pulse of suprathreshold depolarizing current. In contrast, phasic cells respond to this stimulation with a transient burst of action potentials at the onset of current injection. The distributions of these two cell types in lower lumbar sympathetic paravertebral and prevertebral ganglia of guinea pigs coincide with the differential distribution of functionally specific postganglionic neurons (Baron, Jdnig, and McLachlan, 1985; Cassell et al., 1986). In guinea pigs, virtually all neurons in the caudal lumbar sympathetic chain can be classified as phasic, whereas a majority of neurons in the inferior mesenteric 39 ganglion (distal lobe) are tonic (Cassell, Clark, and McLachlan, 1986). Interestingly, most neurons in the caudal sympathetic chain probably are vasoconstrictors that innervate the vasculature of hindlimb skin and skeletal muscle (McLachlan and Jdnig, 1983). In contrast, neurons in the distal lobe of the inferior mesenteric ganglion send axons in the hypogastric nerves (Baron, Jfinig, and McLachlan, 1985) and are involved with regulation of motility and secretion of the pelvic viscera as well as with pelvic vasculature (Baron, Janig, and McLachlan, 1985; Langley and Anderson, 1895a,b,c). Central preganglionic input may also differ between tonic and phasic postganglionic neurons (Jfinig, 1986). Postganglionic neurons in the lower sympathetic chain of guinea pigs discharge action potentials only when "dominant" preganglionic axons are stimulated (see Janig, 1986). In this respect, cells of these ganglia appear to be similar to neurons in ganglia situated on pelvic nerves (Szurszewski, 1981). In contrast, neurons in the inferior mesenteric ganglion require convergence of many presynaptic inputs before these postganglionic neurons will discharge (Crowcroft and Szurszewski, 1971; Szurszewski, 1981). Therefore, perhaps preganglionic and peripheral synaptic inputs are required for activity to occur in lower lumbar, prevertebral ganglion cells, whereas preganglionic input alone may be sufficient for discharge of neurons in the chain ganglia. If it is true that phasic neurons innervate the vasculature, whereas tonic neurons regulate visceral motility, then these data may indicate that vascular reflexes require intact connections between the spinal cord and sympathetic ganglia, whereas reflexes directed at visceral parenchyma can occur in the absence of these connections. Some data indicate that reflexes affecting the 40 vasculature apparently do 225 occur in the absence of connections of sympathetic ganglia with central neural structures (Calaresu, Kim, Nakamura, and Sato, 1978; Johansson and Langston, 1964). However, the occurrance of peripheral reflexes is common to both superior and inferior mesenteric prevertebral outflow (Szurszewski, 1981). After interruption of all afferent and efferent connections between the central nervous system and the celiac plexus, distention of one intestinal segment elicits inhibition of motility in other intestinal segments (Kuntz, 1940; Kuntz and Saccomanno, 1944; Scuba, 1954). In addition, the gastric inhibitory effect of acid in the duodenum persists following decentralization of the celiac plexus (Schapiro and Weedward, 1959). These results suggest that sensory fibers in the intestine do make synaptic contact with neurons in prevertebral ganglia. Synaptic input from intestinal afferent neurons to prevertebral postganglionic neurons has been demonstrated electrOphysiologically (Croweroft and Szurszewski, 1971; Kreulen and Szurszewski, 1979) as well as immunohistochemically (Costa and Furness, 1983; Dalsgaard, H8kfelt, Schultzberg, Lundberg, Terenius, Dockray, and Goldstein). Neurochemical Evidence for Specificity of Target Tissue Innervation The preportion of phasic neurons in the lower abdominal sympathetic system corresponds to the preportion of noradrenergic neurons showing immunoreactivity to neuropeptide Y (Macrae, Furness, and Costs, 1986; McLachlan, 1986). This peptide has been co-localized with noradrenaline in sympathetic neurons which specifically innervate submucosal blood 41 vessels of the intestine (Sundler, Moghimzadeh, Hakanson, Ekelund, and Emson, 1983). Immunohistochemical evidence indicates that axons of sensory neurons from the intestine terminate selectively around celiac neurons which regulate intestinal motility, but not around neurons which are thought to innervate mesenteric blood vessels (Lindh, et al., 1986; Macrae, et al., 1986; Sundler, et al. 1983). Recent neurochemical studies have shown that the majority of neurons in the posterior lobes of the celiac ganglion (adjacent to the splanchnic input) are neuropeptide Y-positive. In the rest of the ganglion, some neuropeptide Y-positive cells are present but more neurons are stained for somatostatin in these regions; other cells do not contain either of the two peptides (Lindh, Hkaelt, Elfvin, Terenius, Fahrenkrug, Elde, and Goldstein, 1986; Macrae, et al. 1986). COMPARISON OF THE DISTRIBUTIONS OF RENAL AND SPLENIC NEURONS IN SYMPATHETIC GANGLIA Journal of the Autonomic Nervous System 11:189-200, 1984 (Reproduced with the permission of Elsevier Scientific Publications) INTRODUCTION Stimulation of visceral afferent nerves has been shown to cause differential responses in splenic and renal sympathetic efferent nerve activity. Stimulation of splenic receptors with algogenic chemicals, such as capsaicin or bradykinin, produces reflex excitation of splenic nerve activity which is greater than the increase in renal nerve activity (Calaresu, et al., 1984). Stimulation of afferent fibers within cardiac sympathetic nerves also causes greater reflex increases in splenic than renal nerve activity; this pattern of differential responses can be elicited in spinal cats as well as in animals with intact neuraxes (Weaver, et al., 1983). Since supraspinal neural pathways are not always necessary for these reflexes to occur, the autonomic organization responsible for at least some of these differential responses must reside within spinal or ganglionic neural circuits. The distribution of postganglionic neurons among different sympathetic ganglia may determine the specificity with which afferent and preganglionic neurons innervate the postjunctional cells. For example, postganglionic neurons in different thoracic paravertebral 42 43 ganglia are innervated selectively by preganglionic neurons originating from different segments of the spinal cord (Lichtman, Purves, and Yip, 1980). Also, the arrangement of cells within any ganglion may be conducive to interaction among individual postganglionic neurons (Archakova, Bulygin, and Netukova, 1982; Decktor and Weems, 1981; Kelts, Whitlock, Ledbury, and Reese, 1979; Kuntz, 1938). Celiac ganglion cells in close proximity to one another often have dendrites which are incorporated into common fiber bundles (Kuntz, 1938). Terminal branches of preganglionic axons arborize within these bundles to provide circumscribed groups of postganglionic cells with a common preganglionic innervation. This arrangement could contribute to specificity of reflex responses; and, consequently, the relative densities of specific types of postganglionic neurons within ganglia is important. Although the ganglionic location of efferent renal neurons has been demonstrated anatomically (Kuo,Krauthamer, and Nadelhaft, 1982) and electrOphysiologically (Decktor and Weems, 1981), the origin of splenic efferent nerves has been studied only cursorily (Kuo and Krauthamer, 1981). Therefore, retrograde transport of horseradish peroxidase (HRP) was employed to compare the inter- and intra-ganglionic distributions of cell bodies of sympathetic efferent neurons supplying the kidney and spleen in cats. 44 METHODS Twenty-three cats (2.7 1:0.2 kg) used in this study were anesthetized with 30-35 mg/kg sodium pentobarbital (Nembutal, Abbott Laboratories, North Chicago, IL) administered intravenously. A left or right flank incision was made and splenic (6 cats), left renal (6 cats), or right renal (7 cats) nerves were dissected from surrounding tissue and cut very close to the spleen or kidney. The central ends of the nerves were moistened and placed on a small sheet of parafilm. Surrounding tissue was covered with gauze soaked in 0.9 2 NaCl to prevent fluid loss by evaporation and to avoid contamination of the tissue with peroxidase. Ten to fifteen milligrams of crystalline HRP (Sigma VI; Sigma Chemical, St. Louis, MO) was applied to the central nerve stumps for one hour. Afterwards, excess HRP was blotted, and the nerves were rinsed with warm buffered physiological saline solution. In four additional animals the nerves (2 splenic, 2 renal) were transected 1-2 cm central to the site of HRP application to test for nonspecific ganglionic labelling and endogenous peroxidase activity. The incisions were sewn closed; the cats were given fluids subcutaneously and allowed to recover from the anesthesia. Following survival times ranging from 24-48 h, the animals were administered 40 mg/kg sodium pentobarbital, 500 units heparin (ICN Pharmaceuticals, Inc., Cleveland, OH), and 5 mg/kg hexamethonium (Mann Research Laboratories, Inc., New York, NY) intravenously to prevent blood clotting and to produce maximal 45 vasodilation for optimal tissue fixation. Animals then were perfused transcardially with 1-2 1 warm, heparinized saline, followed first by 1.5 l of a fixative solution composed of 0.15 M phosphate buffer, 12 (w/v) paraformaldehyde, and 22 gluteraldehyde, and then 1.5 l of the fixative with 10: (w/v) sucrose added. The perfusion schedule was a modified version of that described by Mesulam (Rescue and Mesulam, 1978). Immediately after perfusion, left and/or right thoraco-lumbar sympathetic chain ganglia (stellate through L3) and upper abdominal prevertebral ganglia (left and right celiac, superior mesenteric) were removed and placed in 30% sucrose in 0.15 M phosphate buffer and refrigerated for 1-4 h. Tissue sections were cut at 50 microns on a freezing stage at ‘20°C (Cryo-Histomat MKZ, Hacker Instruments, Inc., Fairfield, NJ) and processed utilizing tetramethylbenzidine (Sigma Chemical Co., St. Louis, MD) as the chromogen following a modification of Mesulam's method (Mesulam, 1978; 1980). Details of the histological procedure can be found in Appendix 8. Tissue was mounted on gelatinized slides, allowed to dry overnight, counterstained with neutral red or toluidine blue, coverslipped, and systematically inspected for reaction product using light- and/or dark-field microscopy. Labelled neurons were counted, and the range of cell diameters was noted using a calibrated ocular reticle under light-field illumination at a magnification of 200x. Every labelled cell in every section was counted and all numbers were corrected for double counting using the method of 46 Abercrombie (1946): P - (AM)/(L+M); where P - corrected number of cells, A - uncorrected number of cells, M - thickness of the tissue sections in microns, and L - average length of the cells. 47 RESULTS Variation was noted in the extent to which the left celiac, superior mesenteric, and right celiac ganglia were fused into a single complex ganglion (solar plexus; Figure 1). Even when the plexus appeared as three separate ganglia (n - 4), neuron cell bodies were observed in the connective strands between the ganglionic masses. Therefore it was impossible to delineate precise borders among the three ganglia. Junctions of greater splanchnic nerves and the solar plexus will be referred to as the left and right celiac poles of the solar plexus. Regardless of the configuration of the solar plexus, the relative distributions of cell bodies of renal and splenic nerves were consistent from animal to animal. Both labelled and unlabelled cells in the sympathetic ganglia were round, oval, or fusiform. The cell shapes likely were dependent upon the plane of the sections. Long axes of all cells measured from 10 to 50 microns. A total of 10,140 splenic, left renal, and right renal neurons (corrected counts) were labelled with HRP (Table 1). Labelled cells usually ranged in size from 30 to 50 microns. However, in two experiments the labelled perikarya (of one right renal nerve and of one left renal nerve) were 10-20 microns in size. HRP reaction product was packed densely within the soma and often filled cell processes (Figure 2). Ganglionic cells were never labelled in the 4 cats in which the 48 Figure 1. Schematic diagrams of variations in the gross anatomy of the solar plexus ganglia. A. (n - 4) Left celiac (LC), superior mesenteric (SH), and right celiac (RC) ganglia appear as 3 distinct ganglionic. masses connected by nerve trunks. B. (n - 3) Left celiac and superior mesenteric ganglia are fused; distinct right celiac ganglion is connected by a nerve trunk. C. (n - 2) Distinct left celiac ganglion connected by a nerve trunk to fused superior mesenteric and right celiac ganglia. D. (n - 14) All three ganglia are fused ‘together into a single ganglionic mass. LGSN, left greater splanchnic nerve; L1 and L2, first and second ganglia of the left lumbar sympathetic chain; Position of the diaphragm is indicated by the interrupted line. ‘AIDIAPHRACM l (3IDIAPHRAGH FIGURE 1 . 49 B:DIAPHRAGM L2 I L1 L2 I I 13|DIAPHRAGH L2 : ’ L1 L2 50 TABLE 1. Numbers of labelled cells in sympathetic ganglia. E _I:l_._ __1;_2_ 1:3 Sol . plex. Totals SPLENIC (n-6) # cells (total) 90 42 7 556 6,072 6,767 # sections (total) 229 16 12 25 143 425 (mean) 18 14 2 185 1,216 ---- # cells/cat (range) 0-46 0-40 0-6 0-556 512-2,168 ---- LEFT RENAL (n-6) # cells (total) 335 513 480 142 1,459 2,929 # sections (total) 330 36 24 25 164 579 (mean) 57 85 80 28 243 --- # cells/cat (range) 0-157 1-293 1-268 5-70 0-784 --— RIGHT RENAL (n-7) # cells (total) 118 151 201 139 569 1,178 # sections (total) 278 28 26 21 356 709 (mean) 17 22 33 28 81 --- # cells/cat (range) 0-111 0-93 0-98 1-120 0-441 --- TCG: thoracic chain ganglia (stellate through T13), L1: first lumbar chain ganglion, L2: second lumbar chain ganglion, L3: third lumbar chain ganglion, Sol. plex.: solar plexus (left celiac, superior mesenteric, and right celiac ganglia); n: number of cats, # cells: number of labelled cells from all animals (numbers are corrected for double counting), # section: number of 50 micron sections from all animals. 51 .msouowa oe we nuance unwwu Home: one :« soaumuoaamo .uosooue mm: sows maomnov voaonma one maaoo noon mo mommoooue one ohumxwume one .mnouso: swoonuoeamm cascade osu canoes Henna man no ammucoaOuonn ummuonou swam .N shaman IN... 4. x “as“: . 4......” .1... a}... 21"{1 . . .’£u\. .. ”P“v. . . so .. o .‘.w . ‘4 m“ G \..v A‘\ bflfl!..1|ll5 18' . .la xIl' u . a As. .Jt. .. . t.. \) .Q.‘v .HQ. 1) ‘i‘fl'. . 3.4 . .\ . ‘ O .. 1...}. .g 52 specific nerve bundles treated with HRP were severed central to the site of HRP application. Gamglionic distribution of splenic neurons A small percentage of labelled splenic neurons were located within the sympathetic chain ganglia (Table 1, Figure 3). Fewer than 3 labelled splenic neurons per ganglion were located in the thoracic chain from the stellate ganglion through T11. Only 90 splenic nerve cell bodies were labelled in thoracic chain ganglia T12 and T13. Ganglia of lumbar segments contained 605 splenic neurons, but 992 of these cells were labelled in only one experiment. Most labelled splenic neurons (6072 cells) were distributed randomly throughout the left and right celiac poles of the solar plexus (Table 1, Figure 3, Figure 4A, Figure 5). The center of the plexus (superior mesenteric region) was virtually devoid of labelled splenic cells. To determine if the splenic nerve axons were located only in nerves emanating from the left celiac pole of the solar plexus, nerves close to the left celiac pole of the ganglionic plexus were severed in one cat. HRP was applied, as usual, to splenic nerves close to the spleen. Labelled splenic neurons of this cat were distributed uniformly throughout both celiac poles, illustrating that the spleen is well innervated by nerves orginating from the right side of the solar plexus. Gamglionic distribution of left renal neurons Renal neurons were distributed more evenly between the chain ganglia and solar plexus (Table 1, Figure 3) than were splenic neurons. Fewer 53 100‘ SPLENIC LEFT RENAL RIGHT RENAL (6072 (1459) (569) (1135) h of labelled neurons l (118) 0|- TC LC SP TC LC SP TC LC SP Figure 3. The percent distributions of splenic and renal neurons among sympathetic ganglia of the thoracic chain (TC), lumbar chain (LC), and solar plexus (SP). Numbers of labeled neurons are given in parentheses above each bar. 54 .aa m.o mum wuosuoo unmau Home: ecu ow macaumunfiamo .mumo osu mo meawomm mnxoae umaoo emu mo mace unease umma ecu :H accuse: ofiuosumoamm Amv Hosea one Aouuwa on one maoauouowaou Hmoauuo> .Acov anemonuoaoxos mo coaumuumfiofiaem mafiaoaaow one .noauoomsmuu muoo nouns :HE owe was .00 .om um .AuumusHv unsung was mfixmunom ecu some moaeamm mma mo>uom umwvumo one .Hmaou .uanoanm mo muw>auu< .nOHuoomomuu muou Homage Houwm one ouomon mua>fiuom o>uom uguonumdaxm mo museumuu nemuwoaaaumo mo magnumOuonm .m shaman I 111131 is] 11.431... .1131... 0<3c<0 4<2m¢ ?IIIIIIIIIIIIIIIIIIIIII-00 if; .. 0.934..» 77 .uumunfi mm3 mfixmuno: ecu =o53 Honuo some Eouw uanmMMHm AHquUfiwfianm uo: mums mo>uoo umamumo one .Hmcou .ownoaem «0 mouse omumaomfia .mowuuomcmuu muou Hmcaam uuumm one oumum uumuafi emu me wanna owummumfio o>umc assauon mounoquMHo unmofiwHGMHm muonou mxmfiuoum< .uoumo eunucmum unomounou when so mmcHH :Hmu 0:9 .mm& some o>onm voumuwmnfi mum mamsfinm mo muunanc ~58 .coauuommmuu ouou Humane do wcfisoaaow :Ha owe one .oe .om .Auch maxmuno: uomucw "unnamed wafiamamm q emu um >uH>Huom o>uon umfioumu mam .Hmnou .oflnoaam mo moHnEmm CHE N ou L we mowmuo>m umomoueou mumm .cOHuuomnmuu once Hmcfinm HmUH>uou swan nouns was encuma mm>uo= owuwcummfihm we mwumnomfim mnfiowco on. on on .5 ow. on on E. .o -oa m. ‘0' u l I u loo“ loo 2 -oo. 0<.O¢uon omHvuoo no .Hooou .oHaono mo huH>Huoo :OHns cH mamaunm on common =oo=ouw Houmosawoexm: .:0Huuonnmuu ouoo HmnHem nouns one woman uoounH consume mousoHoMMHm unouaanme.ouooov omeuoumd .hHo>Huooemou .oHe\mumoA one wuss oH commouexo our Anny some once: one Amdzv ousmmoua HmHuouwm some you mosHm> Had .m.m.H names was oo=Ho> a: .5 H 2 «a H a: «e + 8 a H a: m H 3 e a as +| H o2 o H as awesome .2 H a: a H a: 1 H a: a... H 8 i H a: «o H R a H ANN o H a: 28¢ «S H a: o H a: i H m2 a H 3 i H a: o H 3 o H RN e H m2 oeofiam em .2: mm .22 em .2: a: .2: emu co on wowuuommmwu ouoo HmaHmm nouns mouson mHumwaoa uomuaH manomwii HounoeHuonxu .noHuooonmuu vuou HonHem nouns one snows: soon more: was unseeded HmHuouum some owmwo>< .e wands —A SYMPATHETIC NERVE ACTIVITY (% of control) Figure 9. 00 F 80 60 20 039 so no Effects of spinal cord transection on simultaneously recorded discharge rates of pairs of sympathetic nerves. Bars represent rates of discharge of nerves expressed as a percent of control (intact neuraxis) activity. Percentages were logarithmically transformed prior to analysis of variance, because the percentages were not normally distributed. Therefore coefficients of varia- tion (C.V.) from the analyses of variance are given (below) rather than the standard errors. S: splenic; R: renal; C: cardiac; *: significant difference between the two nerves; N.S.: not significant; S/R C.V.: 742; S/C C.V.: 282; R/C C.V.: 15%. 83 percent changes in rates of activity of renal and cardiac nerves (Figure 9). Sympathetic responses to hypercapnia. To assess the contribution of PaCO2 to differential generation of ongoing activity of the three nerves under study, the lungs were ventilated with a mixture of room air, 95% 02, and 52 CO2 for 5 to 10 min, making the animals (n - 11) hypercapnic and acidotic (Table 5). The levels of hypercapnia achieved did not cause significant changes in the rates of discharge of renal or cardiac nerves either prior to or following spinal cord transection (Figure 10). Splenic nerve activity increased during hypercapnia while the neuraxis was intact but did not change during this procedure after the spinal cord had been severed. The decreases in mean arterial pressure induced by the hypercapnia were statistically significant only in the group of cats in which renal nerve activity was recorded (Table 5). Heart rate was unaffected by inspiration of C02, both prior to and following transection of the spinal cord. Hyperventilation of the lungs with a mixture of room air and 1002 02 did not cause changes in the discharge rates of splenic, renal, or cardiac nerves. Sympathetic responses to increases or decreases in arterial pressure. Sensitivity of sympathetic activity to changes in blood pressure was evaluated in 9 cats to appraise the contribution of systemic arterial pressure to differing levels of ongoing activity of the three nerves. A transient 38 to 65 mmHg increase in arterial pressure was produced before spinal transection by intravenous 84 .Hscou .ULcounm consume Amuuue gouge cu massage cu summon seduce” deuce-«woman: ofi&flu000h QGS 00>uflc Oflfifihfiu MO .mueeoouoaae use «Hcamuosuoe summoned Houuuuuo coon cu mucououuuv unmouuueuuo mouoeov sequouo< .m.m.H memos one mosam> To +l mm +| an on us +| mm Anzaav A H.a- «H2. n.H mos Ameaao m4: ~.~.H n.am ~.~.H o.sm o.u H n.~m Am=aav «some 3 +| ~.N o.oo +l ¢.N c.mm +| a.” n.em szssv N8a.. mucmmuuomwz eo.o No.0 mo.o Ao.c No.6 +| od.~ +| o~.m +| o~.~ 2m +| +| m In III. s-l s e 7‘ h +l M F: O [N an «a +| ea F) +l mm In +l no Ammaao m .n mmn as t 100}- $5.:- 5-'°’ EH ‘ : L '455?‘ . CO 51:51:25 “‘ 0 “#5. 3" EH c'u'n‘u : ‘07 a; 20*- 15$ (sell ‘31:! 2100 ‘- ' m ' m - a . :ziu 7' CHemR CHenfl cum-n SPLINIC RENAL CARDIAC Effect of hemorrhage on sympathetic discharge before and after spinal cord transection. Format is the same as that of Figure 4. HEM: hemorrhage. The excitation of cardiac nerve activity was sig- nificantly greater than that of splenic nerve activity. 89 T4. The paravertebral sympathetic chains had been transected previously between T4 and L5 in these animals to isolate the renal and splenic nerves from influences of upper thoracic (T1 to T4) sympathetic preganglionic outflow. Isolation of this T4-L5 segment from the rest of the spinal cord did not cause changes in splenic or renal nerve activity (Figure 13). Fifteen to 60 min after isolation of the spinal cord segment, the dorsal roots from T4 to L5 were severed, thus depriving this isolated section of spinal cord of spinal afferent input.. No changes were observed in the rate of splenic nerve discharge. Renal nerve activity tended to be slightly increased, but this change was not statistically significant. Blood pressure was not affected by any of these procedures. Figure 13. 90 Bar Splenic 40- 20 OE 100— Renal \QS ./ é 01x ISO DRX SYMPATHETIC NERVE ACTIVITY (eta/sec) Ongoing discharge of sympathetic nerves following dorsal rhizotomy in 6 cats. Bars represent averages of 1 to 2 min samples of activity of splenic and renal nerves following C1 spinal cord transection (C1X), after transection of the spinal cord at T4 and L5 (ISO), and following dorsal rhizotomy from T4 to L5 (DRX). The thin lines on bars represent standard error. Activity of both nerves was unaffected by dorsal rhizotomy. Mean arterial pressure was 92 i 4, 91 i 4, and 90 i 4 mmHg following C1X, ISO, and DRX, respectively. 91 DISCUSSION These data indicate that the continuous electrical discharge recorded from splenic nerves is less dependent upon supraspinal pathways than is activity of renal or cardiac nerves. Although discharge of all three postganglionic nerves persisted after high cervical spinal cord transection, activity of renal and cardiac nerves decreased by more than fifty percent, whereas splenic nerve activity did not change significantly. The differences in dependence of the three nerves upon supraspinal inputs did not result from hypoxemia, hypercapnia, low levels of arterial pressure, or from tonic activity of afferent nerves entering the spinal cord through dorsal roots. The most valid comparisons possible in this investigation are the relative degrees of depression of ongoing activity produced in each of the nerves by cord transection. Limitations of multifiber recording techniques complicate comparison of discharge rates or absolute voltages among the nerves. Because spike counting is subject to frequency saturation, the discharge rates of the 3 nerves were compared in the intact state to be certain that one nerve was not more likely to be affected by quantitation errors than another. Despite the limitations of spike counting, this technique was considered preferable to voltage integration because, after spinal cord transection, renal and cardiac nerves often have only a few active fibers firing, and these fibers discharge asynchronously with widely different voltages. The values 92 obtained by voltage integration are greatly affected by the amplitudes of single spikes in few-fiber preparations and, therefore, are difficult to interpret. However, to provide a comparative analysis which did not include errors inherent to spike counting, results from a few animals were quantified by voltage integration. Even though limitations are inherent in either technique of analysis, both yielded the same relative results and detected a lesser decrease in activity of splenic than cardiac and renal nerves upon spinal cord transection. Paired comparisons of the effects of spinal cord transection on discharge of nerves recorded simultaneously also led to the conclusion that splenic nerve activity is less dependent upon supraspinal sources of excitation than is discharge of renal or cardiac nerves. Since tonically active spinal afferent neurons were not responsible for the differing amounts of depression of nerve activity caused by spinal transection, it is possible that supraspinal neurons directly influence sympathetic outflow differentially. Barman, s£_sl., (1984) have demonstrated nonuniformity of the temporal relationship between the discharge of medullary reticular formation neurons and that of different sympathetic nerves. The source of the differences among these three nerves could not be determined in this study. Since mean arterial pressure did not differ statistically among the groups of cats during any single time period, small differences in blood pressure were unlikely to have contributed to inequalities in discharge rates among the nerves. Furthermore, results of the paired comparisons yielded the same qualitative conclusions as 93 did the group comparisons. If the especially well-maintained spontaneous activity in splenic nerves were caused by hypoxia or by compromised perfusion of the spinal cord or sympathetic ganglia after cord transection, then increasing arterial perfusion pressure would result in a decrease of the anoxia-induced nerve activity.‘ However, activity of splenic and renal nerves did not respond to increases in arterial pressure produced by intravenous infusion of phenylephrine. Although cardiac nerve activity was inhibited by increased arterial pressure, this response probably was caused by stimulation of thoracic spinal afferent nerves rather than by increased spinal or ganglionic perfusion. Malliani, gp_s£., (1971) were able to abolish similar responses of upper thoracic preganglionic nerve activity in acutely spinalized cats by removal of the stellate ganglion to eliminate cardiopulmonary afferent pathways. In our study mean arterial pressure was kept above 70 mmHg, and the cats were ventilated with a mixture of room air and 1002 O2 (PaO2 . ca. 300 mmHg). Rohlicek and Polosa (1981) demonstrated that discharge of cervical sympathetic preganglionic nerves was not sensitive to arterial P02 between values of 40 and 400 mmHg. Although p02 of the spinal cord was not measured in our experiments, it is likely that spinal tissueAO2 tension is directly related to arterial oxygen content (Leniger-Follert, Lflbbers, and Wrabetz, 1975; Metzger and Heuber, 1977; Nair, Whalen, and Buerk, 1975). Furthermore, blood flow in the transected spinal cord is autoregulated between arterial pressures of 50 94 to 125 mmHg (Metzger and Heuber, 1977). Therefore, hypotension and hypoxia did not appear to be sources of ongoing sympathetic activity in the spinal cats, although these stimuli were capable of causing responses in these experiments. Because the hypotension produced by hemorrhage approached the lower limits of autoregulation of blood flow in our experhments, the excitatory responses of splenic and cardiac nerves to hemorrhage in spinal cats probably resulted from decreased vascular perfusion of spinal tissue during the hypotension. The greater responses of cardiac nerves may again have been caused by a specific effect of decreased influences from a population of cardiopulmonary pressoreceptors. Hypercapnia and acidosis are other possible sources of maintained discharge of sympathetic nerves in spinalized animals. Gootman and Cohen (1981) described COZ-induced increases in splanchnic nerve activity in spinal cats. In addition, Zhang, et al., (1982) reported a direct relationship between PaCO2 and discharge rate of sympathetic preganglionic nerves in acutely spinalized cats. However, since neither splenic, renal, nor cardiac nerve activity responded to increases in PaCOz, it could not have been responsible for the ongoing nerve activity or the differences among nerves observed in our study. The discrepancy between our data and those of Zhang, et a1. (1982) may relate to differences in the arterial pressure changes induced by hypercapnia in the two studies. Zhang, et al. (1982) reported hypercapnia-induced decreases in blood pressure, hypotension which could have contributed to 95 the increased nerve activity in that study. In contrast, hypercapnia usually did not cause significantly lower blood pressure in our study. In addition, Zhang, et a1. (1982) reported a sigmoidal relationship between Paco2 and firing frequency of cervical sympathetic neurons. The values for hypercapnic PaCO2 in our experiments were on the lower portion of the rising phase of this curve. Therefore, the significant increases in PaCO2 produced in our study were not sufficient to cause consistent excitation of the nerves. The values of PaCOZ for normocapnic cats in this investigation were similar to those reported elsewhere (Fink and Schoolman, 1963; Herbert and Mitchell, 1971). Activity of renal and cardiac nerves did not respond to increased tidal CO2 when the neuraxis was intact, perhaps because chemoreceptor afferent nerves were severed in our experiments. Excitatory neural responses to increased P8002 or asphyxia have been documented in cats with intact neuraxes (Cohen and Gootman, 1972; Preiss and Polosa, 1977), if peripheral chemoreceptor pathways remained intact. However, Priess and Polosa (1977) reported that rates of discharge in 382 of cervical sympathetic preganglionic neurons either decreased or did not change in response to increased end-tidal C02, even though carotid body chemoreceptors remained intact. Since this study was conducted in animals within 2 hr after spinal cord transection, the absence of sensitivity of sympathetic discharge to increased PaCOz or to increased blood pressure could be attributed to hyporeflexia (”spinal shock”), which has been observed in acutely 96 spinalized preparations (Mukherjee, 1957; Sherrington, 1906). However, spinal shock probably is not responsible for generating the ongoing sympathetic activity and is likely to be depressing it. If inhibitory influences of afferent or prepriospinal neurons were more powerful than excitatory influences, then the result of hyporeflexia could be increased basal nerve activity. If this were the case, one would expect sympathetic activity to decrease with time after spinal cord transection. However, nerve activity did not change significantly during the 2-hr period monitored after spinal cord transection. In addition, the duration of spinal shock was minimized by averting a large fall in blood pressure after spinal cord transection by phenylephrine infusion (Mukherjee, 1957). Activity of splenic and renal nerves did not change significantly following dorsal rhizotomy, indicating that spinal afferent nerve activity did not tonically excite or inhibit sympathetic discharge. Effects of dorsal rhizotomy on cardiac nerve activity were not assessed. Since cardiac and renal nerve activity reacted similarly throughout other aspects of the study, and a search for the source of splenic nerve hyperactivity was the impetus for this experiment, it was considered adequate to monitor only renal nerve activity as a contrast to splenic nerve activity. It is possible that the anesthetic used in this study (chloralose) may have affected some of the observed results. For example, chloralose does attenuate the respiratory response to COz-induced stimulation of 97 central chemoreceptors (Baylis and Monroe, 1964). Similarly, chloralose may have diminished the neural responses to CO2 in animals with intact neuraxes, which could account for the inconsistent effects of increased PaCO2 on sympathetic nerve activity in this study. However, no evidence suggests that chloralose produces unequal effects on the ongoing activity of different sympathetic nerves. Confirmation of the results of this investigation in decerebrate, unanesthetized cats awaits further investigation. Although nerve activity persisted after spinal cord transection, the functional consequences of the ongoing discharge are not known. In 5222! renal nerve activity can increase by two-fold after spinal cord transection (Taylor and Schramm, 1986). Under these conditions, the increased renal nerve activity contributes to decreased excretion of sodium and water (Osborn and Schramm, 1987). However, the increased- renal nerve activity appears to have little effect on renal vascular resistance due to the ability of the renal vascular bed to autoregulate renal blood flow (Osborn and Schramm, 1987). Two days after spinal transection at the sixth cervical segment, Ardell, g£_sl. (1982) observed minimal activity of renal and external carotid sympathetic nerves and failed to elicit a fall in blood pressure by intravenous administration of hexamethonium until one week later. This suggests that ongoing sympathetic activity does not support blood pressure in acute spinal animals. In contrast, data of Johnson, et a1. (1969) demonstrate that intravenous administration of hexamethonium can cause a 98 fall in blood pressure in acute spinal cats. In our study, blood pressure had to be supported artificially by intravenous infusions of phenylephrine, indicating that the observed ongoing discharge was not adequate to support blood pressure completely. The decreased heart rate in the spinal animals reflected the sparse cardiac sympathetic activity. The remaining renal nerve activity probably would not be adequate to cause renal vasoconstriction but could contribute to tonic control of sodium reabsorption and renin secretion (DiBona, 1982). The exceptional splenic nerve activity and splenic sympathetic reflexes (Weaver et al., 1983) in spinal animals would prevent pooling of blood in this capacitive reservoir and would facilitate extrusion of stored erythrocytes into the circulation (Greenway and Lister, 1974). Thus, the varying levels of tonic discharge to different visceral organs may contribute to cardiovascular support in the spinal animal. In summary, these experiments have identified another characteristic of splenic sympathetic nerves which distinguishes them from cardiac and renal nerves. Splenic nerves are not only more reactive to many excitatory inputs (Calaresu, et al., 1984; Weaver, 35 ‘31:, 1983, 1984) but also are less dependent upon tonic supraspinal excitation than are cardiac and renal nerves. This characteristic may be a general preperty of the innervation of splanchnic capacitive circulation, as mesenteric nerve activity also is minimally affected by spinal cord transection (Stein and Weaver, 1987). It was not possible to determine sources of external excitation driving splenic more than 99 cardiac or renal nerves in the spinal animals. Possible explanations for the lesser dependence of splenic nerves upon supraspinal excitation are that: 1) splenic nerves are more dependent upon excitation by local neural circuits in the spinal cord or sympathetic ganglia than are renal or cardiac nerves, 2) activity of splenic nerves may be generated by peripheral sources of tonic afferent nerve activity which is not of dorsal root origin, 3) pacemaker potentials in pre- or postganglionic neurons directed toward the spleen may generate the ongoing discharge in splenic nerves, 4) splenic nerves receive greater descending inhibitory or less descending excitatory influences than renal or cardiac nerves, or 5) components of neural circuits innervating the spleen are inherently more excitable (e.g., closer to threshold) than those innervating the heart or kidney. Possibly, fractionated bulbospinal outflow and spinal or ganglionic mechanisms all contribute to differences in the ongoing or reflex activity of splenic, renal, and cardiac nerves. CHARACTERISTICS OF ONGOING AND REFLEX DISCHARGE OF SINGLE SPLENIC AND RENAL SYMPATHETIC POSTGANGLIONIC FIBERS IN CATS Provisionally accepted for publication in Journal of Physiology INTRODUCTION Much of our current understanding of the way in which neural discharge to different organs can be affected dissimilarly by afferent inputs is based on experiments in which electrical discharge has been recorded from whole nerves innervating different organs (Karim, Kidd, Malpus, and Penna, 1971; Weaver, et al., 1984). For example, stimulation of splenic or cardiac spinal afferent nerves by algogenic substances such as capsaicin and bradykinin causes reflex excitation of splenic and renal sympathetic nerve activity and increased arterial pressure (Calaresu, Tobey, Heidemann, and Weaver, 1984; Weaver, Fry, Heckler, and Oehl, 1983). Although excitatory responses of multifiber renal nerve activity are significant, excitation of multifiber splenic nerve discharge is consistently of greater magnitude than that of renal nerve activity. Excitatory responses of splenic nerve activity may be greater than those of renal, because more of the splenic fibers are engaged in these excitatory reflexes. Alternatively, similar proportions of fibers in each nerve may be engaged in the reflexes, but the magnitude of excitation of each splenic fiber may exceed the magnitude of excitation of each renal fiber. Differential responses of multifiber splenic and renal nerve activity to cardiovascular pressoreceptor stimulation also have been reported (Tobey and Weaver, 1987). Activity of renal nerves was more inhibited than that of splenic nerves in response to phenylephrine- 100 101 induced increases in arterial pressure. In contrast, Ninomiya g£_sl. (Ninomiya, Nisimaru, and Irisawa, 1971) observed greater inhibition of splenic than renal multifiber nerve activity in response to pressoreceptor stimulation. However, Ninomiya g£_sl3 attempted to analyse responses of only the cardiac rhythmic components of the multifiber nerve activity. Therefore, these investigators may have underestimated contributions of the responses of non-rhythmic components of sympathetic nerve activity. For example, perhaps the non-rhythmic component of renal nerve activity responds much more to pressoreceptor stimulation than does the non-rhythmic component of splenic nerve activity. The best way to separate rhythmic from non-rhythmic components of sympathetic nerve activity is to test activity of individual fibers for cardiac rhythmicity. In this way, the potential heterogeneity of responses of individual fibers, with or without cardiac rhythmic activity, can be evaluated accurately. In addition to the non-uniform splenic and renal nerve responses to reflex inputs, ongoing activity of whole splenic nerves is less dependent upon supraspinal sources of tonic excitation than is ongoing activity of whole renal nerves (Heckler and Weaver, 1985). Again, one can only speculate whether a greater number of splenic than renal neurones remains active following spinal cord transection, or whether activity of each renal neurone is depressed relative to that of each splenic neurone. In this paper we describe ongoing and reflex activity of single fibers within splenic and renal sympathetic nerves. The purpose of the present study was to investigate the organization of differential splenic and renal sympathetic outflow by assessment of the heterogeneity 102 of reflex responses and ongoing activity of individual splenic and renal nerve fibers. Therefore, experiments were designed to allow: 1) correlation of ongoing activity of single splenic and renal fibers with peak systolic arterial pressure, 2) correlation of ongoing discharge of single fibers with integrated phrenic nerve discharge, 3) quantitation of reflex changes in single fiber activity caused by stimulation and unloading of systemic pressoreceptors, and 4) quantitation of reflex changes in activity of single fibers elicited . by selective chemical stimulation of splenic or intestinal afferent nerves before and after high cervical spinal cord transection. 103 METHODS Generalpprocedures. Experiments were done on 16 adult cats (3.0 :;0.2 kg) of either sex, anaesthetized with intravenously administered alpha-chloralose (80 mg/kg; Sigma Chemical Company, St. Louis, MO). Supplemental doses (20 mg/kg) were given when necessary throughout the experiments, as determined by assessment of the cat's palpebral reflex and response to paw pinch, the stability of blood pressure, and the size of the pupils. A tracheostomy tube was inserted, and cannulae were passed into the inferior vena cava via the femoral veins for delivery of drugs and solutions. Cannulae were passed into the thoracic aorta via the femoral arteries for monitoring arterial pressure and for withdrawal of blood for analysis of pH, PaOZ, and PaCO2 (Blood Gas Analyser, Model 165; Corning Medical, Medfield, MA). A catheter was passed through the urethra to allow continuous emptying of the urinary bladder. After careful assessment of the animals' plane of anaesthesia (as stated above), gallamine triethiodide (Flaxedil; Davis Geck, New York, NY) was administered, to ensure adequate muscle relaxation for surgical and experimental procedures, and the animals were respired with room air by a Harvard respirator. Initial and supplemental doses of gallamine were 5.0 mg/kg and 2.0 mg/kg, respectively. Oesophageal temperature was monitored and maintained at approximately 37°C. A solution of 52 dextrose w/v in half-strength normal saline (pH adjusted to 7.3) was infused slowly throughout the experiments to compensate for fluid loss. Blood gas composition was maintained (pH - 7.35-7.45, PaO2 > 85 mmHg) by administration of sodium 104 bicarbonate, changing respiratory rate or depth, or by addition of 1002 02 to the inspired air. A laminectomy was done to expose the first cervical segment of the spinal cord for later transection. Following a midline laparotomy, a few loops of the small intestine were placed gently in a small plastic dish. All vascular and neural connections between the intestine and central structures remained intact. The dish was then filled with normal saline and covered with plastic wrap to prevent cooling and dehydration of the intestine. The splenic artery was cannulated via the left gastric artery without obstructing blood flow to the spleen. Snares were placed around the splenic artery and vein central to the cannulation site, being careful to avoid damaging the splenic nerves. Vascular connections of the spleen with other organs were ligated and cut to provide isolation of the splenic vasculature. A pneumothorax was made to minimize artifacts in the neural recordings caused by movement associated with artificial respiration. Nerves to the kidney (7 cats) and spleen (9 cats) were identified close to the respective organs, dissected from surrounding tissue, and severed (Figure 14). The central ends of the nerves were desheathed, separated into small bundles, and placed on a small black plastic platform for further splitting. Nerve bundles subsequently were teased apart progressively and placed on fine bipolar platinum-iridium electrodes until ongoing activity of single fibers could be distinguished (amplifier bandwidth: 30 Hz to 3 kHz). Activity of single cells was discriminated from that of other active fibers by the uniform amplitude of action potentials, and occasionally also by spike duration, with a slope/height window discriminator (Frederick Haer, New 105 .on :OHusaom onHHom Bums Ha mN oH :Honmvouo mamuwouoHa od nuH3 ooHumounH Hanan ecu mo oonnwuoonm an zao>HuuoHom noumasaHum mums muoueouou Homouom .Amv sesame noumaomH haemanumm> one no humans one cunH Axmv auonhvmun mamquHUHS CL «0 aOHuuonnH mo kHo>HuooHom wouoasaHum owes mucumooou oHoonm .Auos oHuonumaamm Honey no anoHam aoum common munoamHHw aouw monHmuno owes mwcHououou woon sawsHm umouomwm .mo>uo: usouowmm HouoomH> mo :OHumasaHum HmoHaono mo avenues pom unoaowsmuum waHouooou Honounoz .ea «somea 106 20.h<._22.hm hzwmmn—n: ..<2....mwhz. 20....(4323fl hzwmwn—n: 0.2m4am .efi ouanm goose.» , cameo 50> Scoy/ Eco-no .c_acom .cc Scenes-no \ \ hzm2w02fiw mum mumnfim mo muoaaaz .momooamou canoanm can Honda :ooSuon mooaouowmfiv oumofivafi youcmo any ca maonshm mmm swung .muouuo vumvnmum voaooa museums“ mama Houuaoo can no mmafia sash .Ammv cowuumficw onaunnmaaaono mam .Amzv :oauooficw ovfimmsuqouuwa .on Houuaoo mo avowumm wsauov Amaocmm nosoav «snowman Hmwuouum was muonwm Aaonmn unwfiu uoqnsv owcoamm can Aamsmn umoa Honmov Hanan mo mama owumcomfiw omnuo>m mnu muaomoumou Hon comm .muonwm oauonumnahm uwaoanm was Hmaou mo moans owumnomfiv no mafivaHnn was nowadaaaaum acumooouommoun mumaoaaamoficumo was Hmfiuouum mo muoommm ma 0 ma. 0 12 ms dVW warm w th (D (DOB/BOXMS) All/\ILOV .1.an 3v 0.2maam a: 4wuo< .noaumasawum Hmcaumouna an aura coaumasafium ofinmanm an whoa vommuuoaa mm3 Mona“ aficoanm mfisu mo >ufi>fiuu< .ouswfiw on» no nonuou uswfiu Hosea anu :H coco: mum mcofiumunaamo .mmnfiomuu Houusou man we unwfiu ocu Ou ozone mum mamwuamuoa nowuom wou>Hmnm osu mo moomuu nowsmlummm .:Ha«x>wmua saw: muouoooou uwaoaam vow Hchumousfi mo coaumazaaum cu uncommon anaaxma can Houucoo mo moowuwm mafiuav ozonm ma hufi>fiuom umnau oawcfim Aaucma unwfiuv Hanan can Afloama uuuav aficoamm .mo>uwc uaouwmwm Hmumumfi> mo :oHumfisaHum HmUfiaocu Ou Mona“ oauocumnaxm decoy oao can owaoanm oao mo omumsomav mo momaoamou wo mvuooou QHLQMHonHHumo .ou ousmfim oo- m6 >1 on 53.3 .:=.¢ __ _ UZIbUh-z .E.‘ as.» . f f 6523 4