A Circumscribed Projection from the Nucleus of the Solitary Tract to The

Home , Hypoglossal nucleus, Nucleus ambiguus, Solitary tract

The Journal of Neuroscience, May 1989, g(5): 1668-l 682

A Circumscribed Projection from the Nucleus of the Solitary Tract to the Nucleus Ambiguus in the Rat: Anatomical Evidence for Somatostatin-284mmunoreactive Interneurons Subserving Reflex Control of Esophageal Motility

E. T. Cunningham, Jr.‘s21a and P. E. Sawchenko2 ‘Department of Neurosciences, University of California, San Diego, La Jolla, California 92093, and 2The Salk Institute for Biological Studies and The Clayton Foundation for Research-California Division, La Jolla, California 92037

Axonal transport and immunohistochemical methods were The nucleus of the solitary tract (NTS) occupies a pivotal po- used to investigate the anatomical and biochemical orga- sition in the central visceromotor system. Projections of the nization of projections from the nucleus of the solitary tract Vth, VIIth, IXth, and Xth cranial nerves terminate centrally (NTS) to the rostral, esophageal, part of the nucleus ambig- within the NTS, conveying sensoryinformation from the oral, uus (NA) in the rat. Discrete iontophoretic deposits of a ret- thoracic, and abdominal cavities. The NTS, in turn, projects to rogradely transported tracer, fluorogold, placed in the rostra1 virtually all levels of the neuraxis, so as to coordinate appro- NA labeled a column of cells within the NTS, termed the priate behavioral, neuroendocrine, and autonomic responses central part of the NTS (after Ross et al., 1985) situated just (seeSawchenko, 1983; Norgren, 1984, for reviews). medial to the solitary tract and extending from about 300 to Prominent among the projections of the NTS is an input to 1000 pm rostra1 to the obex. lontophoretic deposits of the the ventrolateral medulla (Morest, 1967; Cottle and Calaresu, anterogradely transported tracer, Phaseohs vulgaris-leu- 1975; Palkovits and Zaborszky, 1977; Loewy and Burton, 1978; coagglutinin (PHA-L), placed in the central part of the NTS Norgren, 1978; Ricardo and Koh, 1978; Becksteadet al., 1980; gave rise to dense and topographically restricted projections Sawchenkoand Swanson, 1982; Bieger, 1984; Stuesseand Fish, to the rostra1 NA. More caudal and ventral aspects of the NA 1984; Rosset al., 1985; Lovick, 1986; Dampney et al., 1987; did not receive prominent inputs from the central part of the Blessinget al., 1987) a region whose best-characterizedoutput NTS, and deposits that spared the central part of the NTS neurons include catecholaminergicneurons of the Al and Cl gave rise to only sparse projections to the rostra1 NA. Anti- cell groups and motor neurons of the vagus and glossopha- sera against somatostatin-28 (SS-28) stained cell bodies ryngeal nerves. Current evidence suggeststhat the catechol- within the central part of the NTS. In addition, a double- aminergic cell groups play distinctive roles in autonomic and/ labeling procedure, capable of colocalizing anterogradely or neuroendocrine controls of stress-and cardiovascular-related transported PHA-L and endogenous peptides to individual responses(Ross et al., 1981, 1983, 1984; Sawchenkoand Swan- fibers and/or terminals, demonstrated an appreciable num- son, 1982; Cunningham and Sawchenko, 1987, 1988; Morrison ber of SS-284mmunoreactive terminals within the rostra1 NA et al., 1988; seeCiriello et al., 1986, and Reis, 1988, for reviews). that arose from the NTS. Correspondingly, unilateral lesions By contrast, the function of vagal and glossopharyngealmotor that involved the central part of the NTS resulted in a marked neurons in the ventrolateral medulla has remained a topic of depletion of SS-28 immunoreactivity in the ipsilateral rostra1 debate, due primarily to conflicting data concerning the repre- NA. These data provide evidence for a discrete, partly so- sentation of the various viscera in the ambigual complex (see matostatinergic, projection from the central part of the NTS Biegerand Hopkins, 1987, for review). The selectivity, in terms to the rostra1 NA. Anatomical and physiological studies im- of the distribution, cells of origin, and chemical coding, of NTS plicating the central part of the NTS and the rostra1 NA in projections to each of these individual cell groups has not yet esophageal function suggest this pathway to be involved in been deciphered.The present study representsan initial attempt the reflex control of esophageal motility. to evaluate the specificity of projections from the NTS to the ventrolateral medulla, focusing first on inputs to the nucleus ambiguus (NA). Received Aug. 22, 1988; revised Oct. 17, 1988; accepted Oct. 18, 1988. Recently, Bieger and Hopkins (1987) employed retrograde Supported by NIH Grant HL-35 137, a M&night Foundation Scholar’s award tracing techniques to addressthe issueof viscerotopy within the (P.E.S.) and by a Grant-in-Aid of Research from Sigma Xi (E.T.C., Jr.). This work nucleus ambiguus(NA). Their results, in agreementwith many was conducted in part by the Clayton Foundation for Research-California Di- vision; P.E.S. is a Clayton Foundation Investiaator. We thank R. Benoit for of the degeneration studiesdone at the turn of the century (see generously providing antisera to somatostatin, anh R. Benoit and L. W. Swanson Molhant, 1914, and Lawn, 1966, for reviews), provide strong for having thoughtfully read and commented on the manuscript. We also thank evidence for apportioning the NA into 4 subdivisions basedon Carrie Norita, Daniel Aklufi, and Carlos Arias (technical), Kris Trulock (photographic), and Belle Wamsley (secretarial) for excellent assistance. cell size, packing density, and peripheral targets. Three of these Correspondence should be addressed to P. E. Sawchenko, The Salk Institute, subdivisions, or formations as they have beentermed, comprise P.O. Box 85800, San Diego, CA 92138. = Present address: The Johns Hopkins University, School of Medicine, Balti- the NA of the more classicaldescriptions, known for sometime more, MD 2 1205. to innervate branchiomeric derivatives in the head and neck. Copyright 0 1989 Society for Neuroscience 0270-6474/89/05 166% 15$02.00/O From rostra1 to caudal these subdivisions are the esophageal, The Journal of Neuroscience, May 1989, 9(5) 1669 or compact, formation (cf); the palatopharyngeal, or semicom- fibers, and terminals were mapped in the medulla using camera lucida pact, formation (scf); and the laryngeal, or loose, formation (If). techniques. Most animals that received PHA-L deposits also received 1 ~1 injec- The fourth subdivision, composedof scatteredcells lying ventral tions of a 5.0% aqueous suspension of True blue (Dr. Illing GmbH & to the branchiomeric subdivisions, was termed the external for- Co. KG) into the sheath surrounding the vagus nerve. These deposits mation (ef) and has been shown to innervate the heart (see were made at, or slightly rostra1 to, the level of the nodose ganglion and Bieger and Hopkins, 1987; Hopkins, 1987). retrogradely labeled cells in all subdivisions of the ambigual complex. Projections from the NTS to the NA have been suggestedby Zmmunohistochemistry. The available literature and our own material were screened in an effort to identify transmitter candidates in terminals numerous investigators. Retrograde transport studieshave de- in the rostra1 NA or in cell bodies in the central part of the NTS. scribed all aspectsof the caudal NTS as projecting to the ven- Prosomatostatin-derived peptides appeared to meet these criteria, and trolateral medulla (Bieger, 1984; Stuesseand Fish, 1984; Ross antisera directed against 3 of these, somatostatin- 14 (SS-14) somato- etal., 1985; Lovick, 1986; Blessingetal., 1987; Dampney, 1987), statin- (SS-28). and somatostatin-28 ,-, , (SS-28 ,_,,). were used to iden- and anterogradetracing studiesemploying the autoradiographic tify the d~ominant molecular form(s). ‘Thk sera ‘e’mployed were those designated S3 10, S309, and S320, respectively, of Benoit et al. (1982a, method have described relatively diffuse projections from the b), and the results of cross-adsorption studies attesting to the specificity medial part of the NTS to this region (Loewy and Burton, 1978; of these sera were compatible with those provided by this (Sawchenko Norgren, 1978; Ricardo and Koh, 1978; Becksteadet al., 1980; et al., 1988) and other (Morrison et al., 1982, 1983) laboratories. Spe- Sawchenkoand Swanson, 1982; Rosset al., 1985). Interestingly, cifically, medullary staining with each antiserum was abolished following preadsorption with 30 mM concentrations of the respective synthetic only those casesin which tracer depositsinvolved the region of immunogens and was at most modestly attenuated by equimolar con- the rostra1 aspect of the NA displayed labeling of a relatively centrations of the other 2 heterologous prosomatostatin-derived pep- circumscribed cluster of neurons in the medial part of the NTS tides. (Bieger, 1984; Stuesseand Fish, 1984; Rosset al., 1985) a cell Three routes of colchicine administration were used so as to give maximal enhancement ofperikaryal staining in the NTS. These included group termed the central part of the NTS by Rossand colleagues direct injection of a solution containing 50 pg colchicine in 25 bl saline (1985). This region has been implicated both anatomically into a lateral ventricle, a similar injection into the cisternum magnum, (Fryscak et al., 1984; Altschuler et al., 1988) and physiologically or direct application of a pledget of Gelfoam soaked in the same solution (Jean, 1972a, b, 1978; Amri et al., 1984; Bieger, 1984) as re- to the dorsal surface of the medulla. Each procedure was performed 48- ceiving esophagealsensory information (seeJean, 1984; Miller, 72 hr prior to perfusion. Combined anterograde tracing and immunohistochemistry. In order 1986, 1987, for reviews). to evaluate the degree of correspondence between SS-28- and PHA-L- The purposeof the presentstudy wasto investigate the precise stained fibers and terminals in the rostra1 NA following discrete deposits topography and transmitter specificity of the projection from of the tracer in the central part of the NTS, rabbit anti-SS-28 (serum the central part of the NTS to the ambigual complex. The results 309, 1:3000 dilution) and guinea pig anti-PHA-L (1: 1500) were applied concurrently to series of sections through the caudal medulla. A mixture provide support for the existence of an anatomically and bio- of rhodamine-conjugated goat anti-rabbit IgG (Tago, Inc., Burlington, chemically defined group of NTS interneurons subserving reflex CA) and fluorescein-conjugated goat anti-guinea pig IgG (Cooper control of esophagealmotility. Biomedical, West Chester, PA) was used to localize primary antisera. Control procedures to assure that the secondary antisera did not cross- react with an inappropriate primary antiserum or each other were neg- Materials and Methods ative (see Gerfen and Sawchenko, 1985). Photographic enlargements of Adult albino rats of the Sprague-Dawley strain were used in all exper- both SS-28- and PHA-L immunoreactivities were prepared of micro- iments. All surgery was carried out under chloral hydrate anesthesia graphs of sections taken through the rostra1 NA. These were photocopied (350 mg/kg, i.p.), using a direct approach to the caudal medulla after onto acetate and analyzed for colocalization by transilluminating su- reflection of the atlanto-occipital membrane. Empirically determined perimposed transparencies (see Gerfen and Sawchenko, 1985). stereotaxic coordinates were used for tracer and lesion placement. Lesion experiments. To evaluate whether the NTS is the primary Retrograde tracing experiments. Iontophoretic deposits of a 2.0% so- source of SS-28-immunoreactive fibers and terminals in the compact lution of the retrogradely transported fluorescent dye, fluorogold (FG, part of the NA, unilateral electrolytic lesions were made in the region Fluorochrome. Inc.: Schmued and Fallon. 1986) in 0.15 M NaCl were of the central part of the NTS. Monopolar tungsten microelectrodes (A- made into the’region of the rostra1 NA. Deposits were made over 10 M Systems, Inc.) were positioned in the medial NTS approximately 0.5 min through 10 pm (i.d.) glass micropipettes, using a constant-current mm rostra1 to the obex and a variable current source (model S48, Grass device (Model CS-3: Transkinetics. Inc.) set to deliver +2 UA nulses. Medical Instruments) was used to deliver a 1.O mA DC current for 30 with 7 set pulse and interpulse durations. After survival times bf 14: sec. After a survival time of 10 d, animals received bilateral injections 2 1 d, animals were perfused transcardially with 100 ml of 0.15 M NaCl, of True blue into the sheaths surrounding the vagus nerves. Three days followed immediately by 400 ml of 4.0% paraformaldehyde in 0.1 M later, the animals were perfused and prepared for localization of SS-28 sodium tetraborate buffer, pH 9.5. The brains were removed and al- immunoreactivity as described above. lowed to postfix for 6-8 hr in the final perfusate with 10% sucrose added. Five 1-in-5 series of 30-pm-thick sections were cut through the caudal Results medulla on a freezing microtome. One series was mounted and the extent of the injection site and the distribution of retrogradely labeled Retrograde tracing experiments cells were plotted using an X-Y digitizer (Minnesota Datametrics). This Eight depositsof FG were obtained that were centered over the same series was then stained with thionin and used to add nuclear rostra1part of the NA (Fig. 1). In each case,retrogradely labeled boundaries. cells were found in the medial, lateral, and commissural sub- Anterograde tracing experiments. Iontophoretic deposits of a 2.5% solution of Phaseolus vulgaris-leucoagglutinin (PHA-L, Vector Labo- divisions of the NTS, with one prominent cluster of loosely ratories; see Gerfen and Sawchenko, 1984) in 0.1 M sodium phosphate organized cellssituated ventrolateral and another far more com- buffer, pH 7.4, were made into the region of the central part of the NTS. pact grouping found ventromedial to the solitary tract (Figs. 1, Deposits were produced as described above, except that + 5 PA current 2). Those cells found ventrolateral to the solitary tract appear and 15 min injection time were used. After survival times of 7-14 d, to correspond to those describedby Loewy and Burton (1978) animals were perfused and prepared for histology as described above. One series was stained with thionin and used to make projection draw- in the cat and Rosset al. (1985) in the rat as projecting to the ings of nuclear boundaries. An adjacent series was processed for PHA-L ventrolateral medulla, and they occupy the ventrolateral and immunoreactivity following a conventional avidin-biotin immunope- ventral subnucleiof the NTS as describedby Kalia and Sullivan roxidase protocol (Gerfen and Sawchenko, 1984). Immunostained cells, (1982). Those cells found ventromedial to the solitary tract cor- 1670 Cunningham et al. l NTS-Ambigual Projections

Figure 1. Fluorescence photomicrographs to show retrograde labeling in the central part of the NTS (B) resulting from a deposit of FG centered in the rostra1 NA (A). Dark-field photomicrographs of the same sections are shown to provide a reference for cyto- and myeloarchitecture in the ventrolateral (A’) and dorsomedial (B’) medulla.

respond to a cell group described in the rat by Bieger (1984) Stuesseand Fish (1984), and Ross et al. (1985) as projecting to Anterograde tracing experiments. the region of the rostra1part of the NA. This cell group has not Fourteen animals received depositsof PHA-L in the central part been describedin Nissl material (seeKalia and Sullivan, 1982) of the NTS that resulted in denseand topographically restricted but appearsto lie in what has been termed the medial subnu- projections to the rostra1 part of the NA (Figs. 3, 4). In the cleus. Given that we found this cell group to be discriminable majority of these cases,very few other structures in the brain in Nissl-stainedmaterial and strikingly circumscribed with re- received inputs, with the exception of moderate projections to gard to its projections, we follow Ross et al. (1985), who dis- the dorsal motor nucleus of the vagus (Fig. 4) and the lateral tinguished it from the medial subnucleusand referred to it as parabrachial nucleus (not shown). However, these projections the central subnucleus,or part, of the medial NTS. may have arisen from cells outside of the central part of the In Nissl-stainedmaterial, the central part of the NTS formed NTS, as even the most restricted depositsinvolved a few such a column of cells immediately ventromedial to the solitary tract cells. In fact, 8 experimentsin which depositsof FG wereplaced and extending from about 300 to 1000 pm rostra1to the obex. in the lateral parabrachial nucleus failed to label cells in the It was composedof small, round to oval-shaped neurons, typ- central part of the NTS (Cunningham and Sawchenko,unpub- ically 6-10 pm in diameter. Virtually all of these cells were lished observations). retrogradely labeledby injections of FG centered in the rostra1 Projections from the central part of the NTS traversed the NA (Figs. 1, 2). medullary reticular formation as fairly straight, unbranched ax- Depositsof FG centered near but not within the rostra1part onswith regularly spacedvaricosities. In the region of the rostra1 of the NA (n = 20) failed to label cells in the central part of the part of the NA, thesefibers arborized extensively, giving rise to NTS, although other parts of the nucleus were labeled. In par- numerous terminal varicosities that appearedto wrap around ticular, numerouslabeled cellswere observed in the caudal half ambigual motor neurons(Fig. 4). Very sparseprojections were of the NTS, including its lateral, medial, and commissuralsub- seento other regionsof the ventrolateral medulla, including the divisions (E. T. Cunningham and P. E. Sawchenko,unpublished palatopharyngeal, laryngeal, and cardiac portions of the NA observations; see also Loewy and Burton, 1978; Ross et al., (Fig. 5). No projections were observed to the contralateral com- 1985). pact formation. ----

a--5------L;;

l .

---4-

/ /

Figure 2. Series of closely spaced line-drawings through the dorsomedial medulla to show the location and extent of retrogradely labeled neurons in the NTS following a discrete iontophoretic deposit of FG in the rostra1 part of the NA. The most rostra1 section (A) is about 1200 pm rostra1 to the obex, and more caudal sections are approximately 150 pm apart. Note the location of the central part of the NTS (c) in parts B-F. Note also the high density of retrogradely labeled neurons in this part of the nucleus. 1672 Cunningham et al. * NTS-Ambigual Projections

Figure 3. Bright-fieldphotomicrographs showing a representativePHA-L injectionsite in the centralpart of the NTS (A) in relationto anadjacent Nissl-stainedsection (B). Fluorescencephotomicrographs show the resultantterminal field in the rostra&compact, part of the NA (C) in relation to ambigualmotor neuronsin the samesection retrogradely labeled by an injectionof True bluein the vagusnerve at a highcervical level (C?.

Numerous depositsof PHA-L placed ventrolateral (n = 6) or cellular staining for SS-28within the central subnucleuswas far medial to the solitary tract (n = 19), but not involving the central more prominent than for either SS-14or SS-28,_,2,which stained part of the NTS, gave rise to sparseinputs to the rostra1part of at most 1 or 2 cells per section. the NA. There were, however, moderate to heavy projections to other aspectsof the ventrolateral medulla, including non- Combined anterograde tracing-immunohistochemical esophagealparts of the ambigual complex (Fig. 6), the facial staining studies nucleus, and regionsknown to contain catecholaminergicneu- The large proportion of neuronsin the central part of the NTS rons of the Al and Cl cell groups (not shown). that were retrogradely labeled following depositsof FG in the region of the rostra1NA, taken together with the presenceof at Somatostatin immunohistochemistry least some SS-28-immunoreactive neurons in this part of the Ofthe 3 antiseraevaluated, S309 (raisedagainst SS-28) provided NTS, is itself suggestivethat this pathway might be partly so- the most robust staining of fibers in the rostra1NA (Figs. 8, 9; matostatinergic.In addition, however, we have examined6 cases seealso Finley et al., 1981; Johanssonet al., 1984). Staining for that received PHA-L depositscentered in the central part of the SS-28 in other aspectsof the ambigual complex was far less NTS for the presenceof SS-28 immunoreactivity in antero- intense. Terminals in rostra1NA stained lessrobustly for SS- gradely labeled fibers and terminals. Dual staining for PHA-L 28,-,, and extremely weakly for SS-14. and SS-28 revealed strikingly similar distributions of the 2 Although both intraventricular and intracisternal pretreat- markers throughout the ambigual complex. Moreover, in each ments with colchicine were effective in enhancing perikaryal experiment an appreciablenumber of PHA-L-labeled fibers and staining for SS-28 in most parts of the NTS (seealso Finley et terminals were found to be SS-28 immunoreactive (Figs. 8, 9). al., 1981; Johanssonet al., 1984; Millhorn et al., 1987a; Saw- It was often the casethat staining for one or both markers was chenko et al., 1988) few cells were demonstrablein the central so densethat decisive resolution of individual varicosities was part of the nucleus. Direct application of colchicine-soaked obscured(Fig. 9). pledgetsto the dorsal surfaceof the medulla did, however, show as many as one-third of the cells in the central part of the NTS Lesion experiments to be SS-28 immunoreactive (Fig. 7). It should be noted that, Despite the fact that the dual-labeling studiessuggested that the even in the best cases,the majority of cells in the central part central part of the NTS constitutes the major sourceof SS-28- of the NTS were immunonegative for SS.Under all conditions, immunoreactive projections to the compact NA, the nonquan- Figure 4. Series of evenly spaced line drawings through the caudal brain stem to show the distribution of anterograde labeling following a discrete iontophorectic deposit of PHA-L in the central part of the NTS. The center of the injection site is show in D. Note the dense terminal field in the ventrolateral medulla, in the rostra1 part of the NA. The only other terminal field of any prominence is in the dorsal motor nucleus of the vagus nerve. Sections are arranged from rostra1 (A) to caudal (F) at approximately 600 Frn intervals. 1674 Cunningham et al. * NTS-Ambigual Projections

Figure 5. Seriesof fluorescencephotomicrographs through various rostrocaudallevels of the ambigualcomplex to showthe distributionof anterogradelabel (PHA-L) arisingfrom the centralpart of the NTS in relationto varioussubdivisions of the NA, labeledhere by an injectionof True blue in the vagusnerve (T&X). Note the denseand topographicallyrestricted projection to the compactformation (cf), while other aspects of the NA receiveat most sparseinputs.

titative nature of the method leaves open the possibility that a side were undamaged.In each of these cases,SS-28 immuno- significant portion of the somatostatinergic input might arise reactivity was markedly decreasedin the ipsilateral compact from sourcesother than the NTS. We therefore placed unilateral NA, while staining on the contralateral side was robust, and electrolytic lesionsin the medial NTS and examined for their comparableto that seenin intact controls (Fig. 10). effectson SS-28staining in the rostra1NA. In 3 animals,lesions destroyed the medial NTS from about 0.2 to 2.0 mm rostra1to Discussion the obex. In addition, a portion of the ventrolateral NTS, as The results of the present study provide new information on well as the gracile and hypoglossal nuclei, were consistently the topography and transmitter specificity of the projection from involved. The commissural NTS and the entire contralateral the NTS to the ambigual complex. Retrograde and anterograde The Journal of Neuroscience, May 1989, 9(5) 1675

Figure 6. Fluorescence photomicrographs of control injections of PHA-L not involving the central part of the NTS, and the resultant terminal field in the region of the rostra1 NA. Injections placed in the lateral part of the NTS (A) resulted in moderately dense terminal fields in the ventrolateral medulla but only sparse inputs to the compact formation (B and C). Similarly, injections placed in the medial part of the NTS (0) gave rise to appreciable anterograde labeling but only light inputs to the esophageal part of NA (E and F). tracing experiments defined a discrete,central, part of the medial subdivision of the NTS that gives rise to a denseand topograph- Topographic organization of NTS projections to the ambigual ically restricted projection to the rostra1NA. In addition, SS- complex 28 immunoreactivity was demonstrated in cells in the central Though numerousstudies have demonstratedprojections from part of the NTS and in their projections to the rostra1NA. These the NTS to the ventrolateral medulla, there exists little infor- results provide evidence for some measureof anatomical and mation asto the topographic or biochemical specificity of inputs biochemical dissociability of projections from the NTS to the to this region, particularly with respect to the subnuclear or- ventrolateral medulla and identify a potential substratefor the ganization of the NA. Noteworthy in this regardwere the results reflex control of esophagealmotility. of retrograde transport experiments suggestingthe existence of 1676 Cunningham et al. * NTS-Ambigual Projections

Figure 7. Bright-field photomicrographs to show SS-2%immunostaining in the central part of the NTS (A, low power, and C, high power) in relation to an adjacent Nissl-stained section (B, low power, and D, high power). Animals were pretreated with colchicine-soaked pledgets applied directly on the dorsal surface of the medulla 48-72 hr before perfusion.

a projection from the central part of the NTS to the rostra1part ventromedial to the solitary tract and extending from about 300 of the NA (Bieger, 1984; Stuesseand Fish, 1984; Rosset al., to 1000 pm rostra1 to the obex. In addition, the vast majority 1985). While these authors did not directly examine the pos- of cells in the central part of the NTS were found to project to sibility that the central part of the NTS might project to other the region of the rostra1 NA. Second, our anterograde tracing aspectsof the ventrolateral medulla or NA, they did describe studieshave provided the first demonstration of a discretepro- numerous retrogradely labeled cells in the central part of the jection from the central part of the NTS to the rostra1NA, and NTS following depositsof retrograde tracer in the region of the the resolution afforded by the PHA-L technique strongly sug- rostra1NA. geststhat anterogradely labeled fibers and terminals are inter- Similarly, anterograde tracing studiesemploying the autora- acting synaptically with esophagealmotor neurons. Third, we diographic method suggestedthe existence of widespreadpro- have shown that other aspectsof the NTS that do project to the jections from the NTS to the ventrolateral medulla, which in- ventrolateral medulla, including portions of the ambigual com- cluded the region of the NA (Loewy and Burton, 1978; Norgren, plex, project only sparselyto the rostra1NA. This suggeststhat 1978; Ricardo and Koh, 1978; Beckstead and Norgren, 1980; projections from the NTS to the various effector neuron pools Sawchenko and Swanson, 1982; Ross et al., 1985). Rossand in the ventrolateral medulla may well be anatomically and/or colleagues,in particular, showeda denseprojection to the rostra1 biochemically distinct and dissociable. part of the ambigual complex. In each of thesecases, however, tracer depositsin the NTS were fairly large, involving aspects Transmitter specificity of its medial, lateral, and commissural subdivisions, and re- Although numerous neuroactive agentshave been localized to sulted in relatively diffuse anterograde labeling throughout the fibers and cell bodies in the NTS (seeMaley and Elde, 1982; medullary reticular formation. Moreover, the resolution afford- Kalia et al., 1984; Yamazoe et al., 1984; Leslie, 1985; Palkovits, ed by the autoradiographic method did not provide information 1985; Miuraet al., 1987; and Diz et al., 1987, for reviews), none on the morphology of fibers and terminals, and could not, there- has been localized with regard to the central part of the nucleus. fore, resolve terminal fields form fibers-of-passagewith certain- Given the presenceof somatostatin-immunoreactive cell bodies ty. in the medial part of the NTS (Finley et al., 1981; Johansson The present study confirms the findingsof theseearlier studies et al., 1984; Millhorn et al., 1987a)and of a densesomatostatin- and extendsthem in several ways. First, our retrograde transport immunoreactive terminal field in the rostra1 part of the NA studieshave defined the extent of the central part of the NTS. (Finley et al., 1981; Johanssonet al., 1984), we began a search We found it to comprise a column of cells situated immediately for candidate constituents with somatostatin. The Journal of Neuroscience, May 1969, 9(5) 1677

Figure 8. Series of fluorescence photomicrographs through the rostral part of the NA to show neurons in the compact formation, retrogradely labeled by an injection of True blue in the vagus (73-x), in relation to fibers and terminals anterogradely labeled by an injection of PHA-L in the central part of the NTS and the endogenous SS-2%immunoreactive terminal field. Note the general similarity in the distribution of PHA-L and SS-28 immunostaining. Some examples of individual varicosities in which immunoreactivity for tracer and peptide were colocalized are indicated (arrows). 1676 Cunningham et al. a NTS-Ambigual Projections

Figure 9. Fluorescence photomicrographs through the rostra1 part of the NA to show the similaritv in the distribution of fibers and terminal fields anterogradely labeled from the central part of the NTS CPHA-L) and those dis- playing endogenous SS-28 immunoreactivity. Examples of varicosities stained for both markers are indicated (arrows).

Perikaryal stainingwas greatly enhancedin the NTS following denseterminal field in the NA, which appearsto arise in large application of colchicine-soaked pledgetsto the dorsal surface measurefrom the central NTS. Precedentsfor such disparities of the medulla. Even with this, however, somatostatin-immu- between perikaryal and associatedterminal labeling may be noreactive perikarya were difficult to demonstratein the central cited (Kuljis and Karten, 1986) and could be explained by rapid part of the NTS. This is problematic in light of the relatively axonal transport and/or relatively low rates of synthesis. The Journal of Neuroscience, May 1989, 9(5) 1679

Figure 10. Series of fluorescence photomicrographs through the ventrolateral medulla to show the effect of a unilateral electrolytic lesion of the NTS, which included the central subnucleus, on SS-28 staining in the compact part of NA. Ambigual motor neurons were labeled by bilateral injection of True blue in the vagus nerves (T&x). SS-28 immunoreactivity is markedly reduced ipsilateral to the lesion (LESION), whereas staining is unchanged on the contralateral (INTACT) side.

The use of well-characterized antiserathat preferentially rec- ognize 3 distinct prosomatostatin-derived peptides suggested Conclusions that the dominant molecular form expressedin the NTS-com- While the importance of the CNS in coordinating swallowing pact NA pathway is SS-28. Sera directed againstSS- 14 and SS- has been known for some time (Meltzer, 1899, 1907; seealso 28,-,, both stained perikarya and fibers in this system, but far Sumi, 1964; Doty, 1968; Miller, 1982, 1986, 1987; and Jean, lessrobustly than the one raised to SS-28. This contrasts with 1984, for reviews), it is only recently that distinct parts of the results obtained using these same sera in cortical structures, brain stemhave beenimplicated in establishingand maintaining where SS-28,_,, clearly appearsto predominate in fibers (Mor- specific components of this behavior. In particular, numerous rison et al., 1982, 1983), and suggestssome measure of regional studiesin sheep(Jean, 1972a, b, 1978; Amri et al., 1984) have specificity in the processingof prosomatostatin. identified 3 regions of the medulla from which swallowing-re- The combined anterograde transport-immunohistochemical lated activity may be recorded. The first of theseis in and around approach employed here is not a quantitative method (seedis- the lateral subdivision of the NTS, slightly rostra1to the obex. cussionin Levin et al., 1987). However, the similarity in density Most of the cells in this region display a burst of activity in the and distribution of PHA-L- and SS-28-labeledfibers, the extent initial phaseof swallowing,just prior to the contraction of the to which SS-28-immunoreactivity could be colocalized within pharyngeal musculature. A second group of cells is situated in anterogradely labeled fibers arising from the central part of the the ventrolateral medulla, in and around the NA, and shows NTS, and the resultsof the lesion experiments, all converge to increased neuronal activity during pharyngeal contraction, in support the conclusion that the pathway is, at least in part, the so-calledpharyngeal phase.As might be expected, many of somatostatinergic.Extrinsic (i.e., non-NTS) sourcesappear not these cells appear to be ambigual motor neurons innervating to contribute substantially to this pathway. It remains to be the pharynx, as they were able to be activated antidromically determined whether other candidate transmitters exist within by stimulating the cervical vagus nerve (Jean, 1972a). Finally, thoseperikarya of the central part of the NTS that failed to stain a third group of cells, whose activity increaseslate, during the for prosomatostatin-derived peptides. esophagealphase, is situated medial to the solitary tract, in the 1660 Cunningham et al. * NTS-Ambigual Projections

in the caudal medulla, including someaspects of the NTS (Mill- horn et al., 1987a, b). Thus, while the results identify SS-28 as a marker for circumscribed central NTS-rostra1 NA projection, its role in synaptic transmission,as well as those of any coex- isting principles, remain to be clarified. The present results,viewed in conjunction with other recent anatomical and biochemical findings, bring us closerto defining the circuitry for a relatively closed-loop reflex in the caudal medulla, in which SS-28-immunoreactive cells in the central part of the NTS serve as interneurons in the broadest senseof the term (Fig. 11; see Jankowska, 1987). In addition to ACh (e.g., Houser et al., 1983; Kimura et al., 1984), both galanin (GAL; Sawchenkoet al., 1987)and calcitonin gene-relatedpeptide (CGRP; Rosenfeld et al., 1983; Kawai et al., 1985) have beendescribed in cells in the compact part of the NA. Moreover, fibers and terminals immunopositive for GAL (Senguptaand Goyal, 1988) and CGRP (Rodrigo et al., 1985) have been described in the esophagus,and CGRP has been localized within motor end plates. To date, no transmitter or transmitter-candidate has been describedin esophagealafferents, although glutamate (seeTalman et al., 1984)and a number of peptides(see Figure 11. Schematicdrawing of a coronalhemisection through the Dockray and Sharkey, 1986) have been describedin vagal sen- medullato illustratethe proposedreflex arc. Esophagealsensory infor- mationis broughtinto the centralpart of the NTS via the vagusnerve sory systemsin general. Recent descriptions (Altschuler et al., (X). The centralpart of the NTS, in turn, sendsa dense,partly soma- 1988)of esophagealafferents ending preferentially in the central tostatinergic,projection to the rostral,or compact,part of the NA. The part of the NTS indicate that this issuemay now be approach- arc iscompleted by esophagealmotor neurons in the rostra1NA, which, able. in additionto (AC/z),have beenshow to containboth galanin(GAL) In summary, we have provided evidence for somatostatin- andcalcitonin gene-related peptide (CGRP). ergic interneurons projecting from the central part of the NTS to the rostral, compact, part of the NA. Together with anatom- generalregion of the central part of the NTS. Lesionsinvolving ical and physiological studiesimplicating the central part of the this third group preferentially disrupt (Jean, 1972b), while focal NTS and the rostra1 NA in esophagealfunction, these data glutamate stimulation preferentially elicits (Bieger, 1984), the provide support for a disynaptic, closed-looppathway subserv- esophagealphase of swallowing (seeMiller, 1986, for review). ing reflex control of esophagealmotility. Immunohistochemical Moreover, esophagealafferents have recently been shown to studies suggestthe biochemical specificity of various compo- terminate in the region of the central part of the NTS (Fryscak nents of this proposed loop to be more complex than at first et al., 1984; Altschuler et al., 1988). It would appear, therefore, expected. that the most likely role for a central NTS-rostra1 NA projection isthat ofan interneuron linking primary afferentswith the motor Appendix supply of the esophagus,and thus participating in the control of secondary esophagealperistalsis. We are aware of no more Abbreviations forfisures anatomically well-defined circuitry for a viscus-specificreflex AC/z, acetylcholine;AZ’, areapostrema; cc, centralcanal; CGRP, calcitonin gene-relatedpeptide; CA compact formation (nucleus ambiguus); in the mammalian brain stem. CN, cochlearnucleus; Cu, cuneate nucleus; dcst, dorsalspinocerebellar The presenceof SS, a peptide best known for its hypophys- tract; DMX, dorsalmotor nucleusof the vagus;dpy, decussationof the iotropic role in the inhibition of growth hormone secretion, in pyramidaltracts; EC, externalcuneate nucleus; ef externalformation esophagealreflex circuitry is somewhatunexpected, though its (nucleusambiguus); GAL, galanin;Gr, gracilenucleus; IO, inferior oli- widespread distribution in functionally unrelated parts of the vary nucleus;/x looseformation (nucleus ambiguus); LRN, lateralre- ticular nucleus;mlf; mediallongitudinal fasciculus; MV, medialvestib- brain (Finley et al., 1981; Johanssonet al., 1984) would seem ular nucleus;NTS, nucleusof the solitarytract; NT& centralpart of to rule out the existence of any unified system with which SS the NTS; NTS,, lateralNTS; NTS,,, medialNTS; PH, nucleusprepos- has an a priori association.In fact, even within the NTS, SS- itushypoglossi; PHA-L, Phaseolus vulgaris-leucoagglutinin;py,pyrami- 28 has beenlocalized in a projection from the commissuralpart dal tract; RPa, nucleusraphe pallidus; scf; semicompactformation (nucleusambiguus); SLN, superiorlaryngeal nerve; SNV, spinalnucleus of the nucleusto the neurosecretory hypothalamus (Sawchenko of the trigeminalnucleus; SS-28, somatostatin-28;stv, spinaltract of et al., 1988), a pathway ostensibly unrelated to swallowing. the trigeminalnucleus; SV, spinalvestibular nucleus; TB-X, True blue Physiologically, the majority of intracellular studieshave de- in the cervicalvagus; ts, solitarytract; VZZ,facial nucleus; VZZZn, eighth scribed SS as hyperpolarizing (seeSiggins and Groul, 1986, for nerve;X, nucleusX; XZZ,hypoglossal nucleus. review), and similar effectshave beendescribed for vagal motor neurons in the dorsal medulla (Oomura and Mizuno, 1986; G. References Siggins, personal commmunication). Somatostatin has, how- Altschuler,S. M., X. Bao,D. Bieger,D. A. Hopkins,and R. R. Miselis ever, been shown to enhance the excitatory effects of ACh in (1988) Viscerotopicrepresentation of the upperalimentary tract in the hippocampusand neocortex (Mantillas et al., 1986), and it the rat: Sensoryganglia and nuclei of the solitaryand spinal trigeminal could serve a similar, modulatory, function in the brain stem. tracts.J. Comb.-N&rol. (in press). Amri. M.. A. Car.and A. Jean (1984) Medullarvcontrol ofthe uontine This notion gains some support from recent reports of wide- swallowingneirones in sheep.Exp. Brain ReH.55: 105-l lo.- spreadcolocalization of SSwith enkephalin-like peptidesin cells Beckstead,R. M., J. R. Morse, and R. Norgren (1980) The nucleus The Journal of Neuroscience, May 1989, 9(5) 1661

of the solitary tract in the monkey: Projections to the thalamus and Jean, A. (1984) Control of the central swallowing program by inputs brain stem nuclei. J. Comp. Neurol. 190: 259-282. from the peripheral receptors. A review. J. Autonom. Nerv. Syst. 10: Benoit, R., P. Bohlen, N. Ling, A. Briskin, F. Esch, P. Brazeau, S.-Y. 225-233. Yina. and R. Guillemin (1982a) Presence of somatostatin 28,_,, in Johansson, O., T. Hokfelt, and R. P. Elde (1984) Immunohistochem- hypgthalamus and pancreas. P&c. Natl. Acad. Sci. USA 79:‘917- ical distribution of somatostatin immunoreactivity in the central ner- 921. vous system of the adult rat. Neuroscience 13: 265-339. Benoit, R., N. Ling, B. Alford, and R. Guillemin (1982b) Seven pep- Kalia, M., and J. M. Sullivan (1982) Brain stem projections of sensory tides derived from pro-somatostatin in rat brain. Biochem. Biophys. and motor components of the vagus nerve in the rat. J. Comp. Neurol. Res. Commun. 107: 944-950. 211: 248-264. Bieger, D. (1984) Muscarinic activation of rhombencephalic neurones Kalia, M., K. Fuxe, T. Hiikfelt, 0. Johansson, R. Lang, D. Ganten, C. controlling oesophageal peristalsis in the rat. Neuropharmacology 23: Cuello, and L. Terenius (1984) Distribution of neuropeptide im- 1451-1464. munoreactive nerve terminals within the subnuclei of the nucleus of Bieger, D., and D. A. Hopkins (1987) Viscerotopic representation on the tractus solitarius of the rat. J. Comp. Neurol. 222: 409-444. the upper alimentary tract in the medulla oblongata in the rat: The Kawai, Y., K. Takami, S. Shiosaka, P. C. Emson, C. J. Hillyard, S. nucleus ambiguus. J. Comp. Neurol. 262: 546-562. Girgis, I. MacIntyre, and M. Tohyama (1985) Topographic local- Blessing, W. W., S. C. Hedger, T. H. Joh, and J. 0. Willoughby (1987) ization of calcitonin gene-related peptide in the rat brain: An im- Neurons in the area postrema are the only catecholamine-synthesizing munohistochemical analysis. Neuroscience 15: 747-763. cells in the medulla or pons with projections to the rostra1 ventro- Kimura, H., P. L. McGeer, and J.-H. Peng (1984) Choline acetyl- lateral medulla (Cl-area) in the rabbit. Brain Res. 419: 336-340. transferase-containing neurons in the rat brain. In Classical Trans- Ciriello. J.. M. M. Caverson. and C. Polosa (1986) Function of the mitters and Transmitter Receptors in the CNS, Part II, Handbook of ventrblateral medulla in the control of the‘circulation. Brain Res. Chemical Neuroanatomy, Vol. 3, A. Bjorklund, T. Hiikfelt, and M. Rev. II: 359-391. J. Kuhar, eds., pp. 51-125, Elsevier, New York. Cottle, M. K., and F. R. Calaresu (1975) Projections from the nucleus Kuljis, R. O., and H. J. Karten (1986) Substance P-containing cells and tractus solitarius in the cat. J. Comp. Neurol. 161: 143-158. become progressively less detectable during retinotectal development Cunningham, E. T., Jr., and P. E. Sawchenko (1987) Projections from in the frog Rana pipiens. Proc. Natl. Acad. Sci. USA 83: 5736-5740. the C 1, C2, and C3 adrenergic cell groups to the paraventricular and Lawn, A. M. (1966) The localization, in the nucleus ambiguus of the supraoptic nuclei of the rat hypothalamus. Anat. Rec. 214: 27A. rabbit, of the cells of origin of motor nerve fibers in the glossopha- Cunningham, E. T., Jr., and P. E. Sawchenko (1988) Anatomical spec- ryngeal nerve and various branches of the vagus nerve by means of ificity of noradrenergic inputs to the paraventricular and supraoptic retrograde degeneration. J. Comp. Neurol. 127: 293-306. nuclei of the rat hypothalamus. J. Comp. Neurol. 274: 60-76. Leslie, R. A. (1985) Neuroactive substances in the dorsal vagal com- Dampney, R. A. L., J. Czachurski, K. Dembowsky, A. K. Goodchild, plex of the medulla oblongata: Nucleus of the tractus solitarius, area and H. Seller (1987) Afferent connections and spinal projections of postrema, and dorsal motor nucleus of the vagus. Neurochem. Int. the pressor region in the rostra1 ventrolateral medulla of the cat. J. 7. 191-211. Autonom. Nerv. Syst. 20: 73-86. Levin, M. C., P. E. Sawchenko, P. R. C. Howe, S. R. Bloom, and J. M. Diz, D. I., K. L. Barnes, and C. M. Ferrario (1987) Functional char- Polak (1987) Organization of galanin-immunoreactive inputs to the acteristics of neuropeptides in the dorsal medulla oblongata and vagus paraventricular nucleus with special reference to their relationship to nerve. Fed. Proc. 46: 30-35. catecholaminergic afferents. J. Comp. Neurol. 261: 562-582. Dockray, C. J., and K. A. Sharkey (1986) Neurochemistry of visceral Loewy, A. D., and H. Burton (1978) Nuclei of the solitary tract: afferent neurons. In Progress in Brain Research, Vol. 67: Visceral Efferent projections to the lower brain stem and spinal cord of the Sensation, F. Cervero and J. F. B. Morrison, eds., pp. 133-148, El- cat. J. Comp. Neurol. 181: 421-450. sevier, New York. Lovick, T. A. (1986) Projections from brainstem nuclei to the nucleus Doty, R. W. (1968) Neural organization of deglutition. In Handbook paragigantocellularis lateralis in the cat. J. Autonom. Nerv. Syst. 16: of Physiology. Alimentary Canal, Sec. 6, Vol. 4, C. F. Code, ed., pp. l-l 1. 1861-1902, Am. Physiol. Sot., Washington, D.C. Maley, B. E., and R. Elde (1982) Immunohistochemical localization Finley, J. C. W., J. L. Maderdrut, L. J. Roger, and P. Petrusz (198 1) of putative neurotransmitters within the feline nucleus tractus soli- The immunocytochemical localization of somatostatin-containing tarii. Neuroscience 7: 2469-2490. neurons in the rat central nervous system. Neuroscience 6: 2173- Mantillas, J. R., G. R. S&ins, and F. E. Bloom (1986) Somatostatin 2192. selectively enhances acetylcholine-induced excitations in rat hippo- Fryscak, T., W. Zenker, and D. Kantner (1984) Afferent and efferent campus and cortex. Proc. Natl. Acad. Sci. USA 83: 75 18-752 1. innervation of the rat esophagus. Anat. Embryol. 170: 63-70. Meltzer, S. J. (1899) On the causes of the orderly progress of the Gerfen, C. R., and P. E. Sawchenko (1984) An anterograde neuroan- *3-peristaltic movements in the oesophagus. Am. J. Physiol. 2: 266- atomical tracing method that shows the detailed morphology of neu- LIL rons, their axons and terminals: Immunohistochemical localization Meltzer, S. J. (1907) Deglutition through an oesophagus partly de- of an anterogradely transported plant lectin, Phaseolus vulguris-leu- prived of its muscularis, with demonstration. Proc. Sot. Exn. Biol. coagglutinin (PHA-L). Brain Res. 290: 219-238. Med. 4: 40-43. Gerfen, C. R., and P. E. Sawchenko (1985) A method for anterograde Miller, A. J. (1982) Deglutition. Physiol. Rev. 62: 129-184. axonal tracing of chemically specified circuits in the central nervous Miller, A. J. (1986) Neurophysiological basis of swallowing. Dysphagia system: Combined Phaseolus vulgaris-leucoagglutinin (PHA-L) tract 1~9l-lCm tracing and immunohistochemistry. Brain Res. 343: 144-150. Miller, A. J. (1987) Swallowing: Neurophysiologic control of the Hopkins, D. A. (1987) The dorsal motor nucleus of the vagus nerve esophageal phase. Dysphagia 2: 72-82. and the nucleus ambiguus: Structure and connections. In Cardiogenic Millhbm,D. E., T. H&kfelt,L. Terenius, A. Buchan, and J. C. Brown Reflexes, R. Hainsworth, P. N. McWilliam and D. A. S. G. Mary, (1987a) Somatostatin- and enkephalin-like immunoreactivities are eds., pp. 185-203, Oxford U. P., Oxford, UK. frequently colocalized in neurons in the caudal brain stem of rat. Exp. Houser. C. R.. G. D. Crawford. R. P. Barber. P. M. Salvaterra. and J. Brain Res. 67: 420-428. E. Vaughn ‘( 1983) Organization and morphological characteristics Millhorn, D. E., K. Seroogy, T. HBkfelt, L. C. Schmued, L. Terenius, of cholinergic neurons: An immunocytochemical study with a mono- A. Buchan, and J. C. Brown (1987b) Neurons of the ventral medulla clonal antibody to choline acetyltransferase. Brain Res. 266: 97-l 19. oblongata that contain both somatostatin and enkephalin immuno- Jankowska, E. (1987) Interneurons. In Encyclopedia of Neuroscience, reactivities project to nucleus tractus solitarii and spinal cord. Brain Vol. 1, G. Adelman, ed., pp. 541-542, Birkhluser, Boston. Res. 424: 99-108. Jean, A. (1972a) Localisation et activite des neurones dtglutiteurs Miura, M., K. Takayama, and J. Okada (1987) Study of possible bulbaires. J. Phvsiol. (Paris) 64: 227-268. transmitters in the solitary tract nucleus of the cat involved in the Jean, A. (1972b)- Effet ‘de lesions localisees du bulbe rachidien sur le carotid sinus baro- and chemoreceptor reflex. J. Autonom. Nerv. Syst. stade oesophagien de la dtglutition. J. Physiol. (Paris) 64: 507-5 16. 19: 179-188. Jean, A. (1978) Localisation et activite des motoneurones oesopha- Molhant, M. (19 14) Le nerf vague. Etude anatomique et experimentale. giens chez le mouton. J. Physiol. (Paris) 74: 737-742. Deuxieme partie. Le noyau ambigu; les connexions anatomique et la 1662 Cunningham et al. * NTS-Ambigual Projections

valeur fonctionelle du noyau central du vague et du noyau ambigu. Ross, C. A., D. A. Ruggiero, T. H. Joh, D. H. Park, and D. J. Reis Ntvraxe 13: 9-3 16. (1983) Adrenaline synthesizing neurons in the rostra1 ventrolateral Morest, D. K. (1967) Experimental study of the projections of the medulla: A possible role in tonic vasomotor control. Brain Res. 273: nucleus of the tractus solitarius and the area postrema in the cat. J. 356-361. Comp. Neurol. 130: 277-300. Ross, C. A., D. A. Ruggiero, T. H. Joh, D. H. Park, and D. J. Reis Morrison, J. H., R. Benoit, P. J. Magistretti, N. Ling, and F. E. Bloom (1984) Rostra1 ventrolateral medulla: Selective projections to the (1982) Immunohistochemical distribution of prosomatostatin relat- thoracic autonomic cell column from the region containing Cl ad- ed peptides in hippocampus. Neurosci. Lett. 34: 137-142. renaline neurons. J. Comn. Neurol. 228: 168-185. Morrison, J. H., R. Benoit, P. J. Magistretti, and F. E. Bloom (1983) Ross, C. A., D. A. Ruggiero, and D. J. Reis (1985) Projections from Immunohistochemical distribution of pro-somatostatin-related pep- the nucleus tractus solitarii to the rostra1 ventrolateral medulla. J. tides in cerebral cortex. Brain Res. 262: 344-35 1. Comp. Neurol. 242: 5 1 l-534. Morrison, S. F., T. A. Milner, and D. J. Reis (1988) Reticulospinal Sawchenko, P. E. (1983) Central connections of the sensory and motor vasomotor neurons of the rat rostra1 ventrolateral medulla: Relation- nuclei of the vagus nerve. J. Autonom. Nerv. Syst. 9: 13-26. ship to sympathetic nerve activity and the Cl adrenergic cell group. Sawchenko. P. E.. and L. W. Swanson (1982) The oraanization of J. Neurosci. 8(4): 1286-1301. noradrenkrgic pathways from the brainstem’to the paiaventricular Norgren, R. (1978) Projections from the nucleus of the solitary tract and supraoptic nuclei in the rat. Brain Res. Rev. 4: 275-325. in the rat. Neuroscience 3: 207-2 18. Sawchenko, P. E., E. T. Cunningham, Jr., and M. C. Levin (1987) Norgren, R. (1984) Central neural mechanisms oftaste. In The Nervous Anatomic and biochemical specificity in central autonomic pathways. System: Sensory Processes, V. B. Mountcastle sect. ed., F. E. Bloom, In Organization of the Autonomic Nervous System: Central and Pe- vol. ed., pp. 1087-l 124, American Physiological Society, Bethesda, ripheral Mechanisms, J. Ciriello, F. R. Calaresu, L. P. Renaud, and MD. C. Polosa, eds., pp. 267-281, Liss, New York. Oomura, Y., and Y. Mizuno (1986) Effect of somatostatin on the vagal Sawchenko, P. E., R. Benoit, and M. R. Brown (1988) Somatostatin motor neuron in the rat. Brain Res. Bull. 17: 397-401. 28-immunoreactive inputs to the paraventricular and supraoptic nu- Palkovits, M. (1985) Distribution of neuroactive substances in the clei: Principal origin from non-aminergic neurons in the nucleus of dorsal vagal complex of the medulla oblongata. Neurochem. Int. 7: the solitary tract. J. Chem. Neuroanat. I: 8 l-94. 213-219. Schmued, L. C., and J. H. Fallon (1986) Fluoro-gold: A new fluorescent Palkovits, M., and L. Zaborsky (1977) Neuroanatomy of central car- retrograde axonal tracer with numerous unique properties. Brain Res. diovascular control. Nucleus tractus solitarii: Afferent and efferent 377: 147-l 54. neuronal connections in relation to the baroreceptor reflex arc. Prog. Sengupta, A., and R. K. Goyal (1988) Localization of galanin im- Brain Res. 47: 9-34. munoreactivity in the opossum esophagus. J. Autonom. Nerv. Syst. Reis, D. J. (1988) The brain and hypertension. Arch. Neurol. 45: 180- 22: 49-56. 182. Siggins, G. R., and D. L. Gruol (1986) Mechanisms of transmitter Ricardo, J. A., and E. T. Koh (1978) Anatomical evidence of direct action in the vertebrate central nervous system. In The Nervous Sys- projections from the nucleus of the solitary tract to the hypothalamus, tem: Intrinsic Regulatory Systems of the Brain, V. B. Mountcastle, amygdala, and other forebrain structures in the rat. Brain Res. 1.53: sect. ed., F. E. Bloom, vol. ed., pp. l-l 14, American Physiological l-26. Society, Bethesda, MD. Rodrigo, J., J. M. Polak, L. Fernandex, M. A. Ghatei, P. Mulderry, and Stuesse, S. L., and S. E. Fish (1984) Projections to the cardioinhibitory S. R. Bloom (198 5) Calcitonin gene-related peptide immunoreactive region of the nucleus ambiguus of rat. J. Comp. Neurol. 229: 271- sensory and motor nerves of the rat, cat, and monkey esophagus. 278. Gastroenterology 88: 444-45 1. Sumi, T. (1964) Neuronal mechanisms in swallowing. Pfluegers Arch. Rosenfeld, M. G., J.-.J. Mermod, S. G. Amara, L. W. Swanson, P. E. 278: 467-477. Sawchenko, J. Rivier, W. W. Vale, and R. M. Evans (1983) Pro- Talman, W. T., A. R. Granata, and D. J. Reis (1984) Glutamatergic duction of a novel neuropeptide encoded by the calcitonin gene via mechanisms in the nucleus tractus solitarius in blood nressure control. tissue-specific RNA processing. Nature 304: 129-135. Fed. Proc. 43: 39-44. Ross, C. A., D. M. Armstrong, D. A. Ruggiero, V. M. Pickel, T. H. Yamazoe, M., S. Shiosaka, T. Shibasaki, N. Ling, K. Tateishi, E. Hashi- Joh, and D. J. Reis (198 1) Adrenaline neurons in the rostra1 ven- mura. T. Hamaoka. J. R. Kimmel. H. Matsuo. and M. Tohvama trolateral medulla innervate thoracic spinal cord: A combined im- (1984) Distribution of six neuropeptides in the ‘nucleus tractus sol- munocytochemical and retrograde transport demonstration. Neuro- itarii of the rat: An immunohistochemical analysis. Neuroscience 3. sci. Lett. 25: 257-262. 1243-1266.