Origin, Course and Terminations of the Rubrospinal Tract in the Pigeon (Colurnba livia) J. MARTIN WILD,ZJOHN B. CABOT,2' DAVID H. COHEN 2 ' AND H. J. KARTEN Department of Physiology, University of Virginia, School of Medicine, Charlottesville, Virginia 22908 and 3Departments of Psychiatry and Anatomy, SUNY,Stony Brook, School of Medicine, HeaEth Science Center, Stony Brook, New York 1 1 794
ABSTRACT The red nucleus and its spinal projections in the pigeon (Cob umba livia) have been studied using both normal and experimental material. The cytoarchitecture of the nucleus is described on the basis of Nissl-stained sections and reveals an organization generally similar to that of mammals. The large neurons (40-50 pm) tend to be located dorsomedially and ventrolaterally a t more caudal nuclear levels, while the small- and medium-sized neurons (1535 pm) predominate a t rostra1 levels. However, neurons of all sizes are present throughout the nucleus. Following lesions of the nucleus, the course of degenerating axons stained with the Fink-Heimer method has been traced throughout the brainstem and spinal cord. The rubrospinal tract crosses the midline, courses past the ventrocaudal aspect of the contralateral nucleus ruber, and then descends rostroventral and lateral to the nucleus tegmenti pontinus. In its caudal continuation the tract lies ventral to the brachium conjunctivum and the entering radix of the trigeminal nerve. I t then assumes a ventrolateral position in the caudal brainstem before shifting to a dorsolateral position in the lateral funiculus of the spinal cord. Within the spinal grey the rubrospinal tract terminates in laminae V, VI and to a lesser extent VII. The possibility of a topographical organization of the nucleus was investigated with injections of horseradish peroxidase into brachial, thoracic and lumbar spinal cord. Regardless of the level of injection, labelled neurons of all sizes were present throughout the contralateral nucleus ruber, indicating the absence of an obvious topography. The organization of the nucleus ruber and its projections have been studied extensively in mammals including, for example, rodents (Mizuno and Nakamura, '71; Mizuno et al., '73; Reid et al., '73; Brown, '74; Petrovicky, '74), carnovires (Edwards, '72), marsupials (Martin and Dom, '70; King et al., '71; Warner and Watson, '72; King et al., '74; Martin e t al., '741, ungulates (Otabe and Horowitz, '70), tree shrew (Murray et al., '76; Murray, '77) and primates (King et al., '71; Miller and Strominger, '73). (See Massion "671 for an earlier review.) In contrast, there have been few reports on the avian nucleus ruber save for the earlier non-experimental studies of Craigie ('28, '30) and Jungherr ('45). The description of the nucleus by these authors, however, was limited and within the context of a general J. COMP. NEUR. (1979) 187: 639-654.
treatment of the mesencephalic nuclei. A more recent report by Ovtscharoff and Gossrau ('72) has described the cytoarchitecture of the red nucleus of the hen (Gallus domesticus), and an avian rubrospinal tract has been briefly mentioned in reports by Zecha ('61) and Wold ('78). However, there are no studies providing a detailed description of the avian rubrospinal tract, and thus the present study was undertaken to describe its cells of origin, course and sites of termination in the pigeon (Columba livia) as part of a more ' This material was reported at the Seventh Annual Meeting of the k i e t y for Neuroscience, Annaheim, California (Wildet al., '77). 'Present addraw: Department of Neurobiology and Behavior, SUNY at Stony Brook, Stony Brook, New York 11794. 6Please address all reprint requesta to: Dr. David H. Cohen, Department of Neurobiology and Behavior, Graduate Biology Building, SUNY at Stony Brwk, Stony Brook, New York 11794.
639
640
J. M. WILD, J. B. CABOT, D. H. COHEN AND H. J. KARTEN
extensive investigation of the avian brainstem projections to the spinal cord. The general similarities of the avian and mammalian rubrospinal systems are also discussed. MATERIALS AND METHODS
Experimental animals All normal and experimental material was obtained from White Carneaux pigeons (CoEumba livia) ranging in age from two months to four years and weighing 400-700 gm.They were obtained from the Palmetto Pigeon Plant, Sumter, South Carolina. Cytoarchitectonic analysis The cytoarchitecture of the nucleus ruber was based upon light microscopic examination of normal material cut in transverse, sagittal and horizontal planes. In three animals 50-, 75- and 100-pm thick, serial transverse sections were prepared from celloidin embedded brains, and in an additional three animals 25-pm thick serial sections in transverse, sagittal and horizontal planes were examined with paraffin embedded material. All sections were stained with cresylechtviolet, and drawings of the nucleus were made from every second or third section with the aid of a Leitz drawing tube. In all cases the birds were sacrificed with an overdose of sodium pentobarbital and perfused through the left ventricle or internal carotid artery with avian saline followed by 10%formalin. The calvaria were then removed and the brains blocked stereotaxically (Karten and Hodos, '67) and embedded in either paraffin or celloidin. Anterograde degeneration Stereotaxically guided electrolytic lesions were placed in the nucleus ruber of 15pigeons. Electrodes were inserted via several different trajectories to control for damage to various adjacent nuclei and fiber tracts. Thus, in addition to an ipsilateral vertical penetration in the transverse plane of Karten and Hodos ('67), electrode penetrations were made from the contralateral telencephalon or at different angles through the ipsilateral telencephalon to avoid damage to the cerebellum and to the nuclei of Darkschewitsch and Cajal. Additional controls included lesions of the cerebellum, posterior commissure, nuclei of Darkschewitsch and Cajal, pre-rubral fields, lateral tegmentum, telencephalon and diencephalon.
The birds were sacrificed a s described above, the calvaria removed and the brains blocked stereotaxically in either transverse, sagittal or horizontal planes. The brains and spinal cords were removed and either embedded in a gelatin-albumin casement and cut frozen a t 25 fim, or postfixed in 20% sucrose for 24 hours and then cut frozen a t 33 pm. Serial brainstem sections a t intervals of 200 pm were stained according to the FinkHeimer method I, and each adjacent section was stained with cresylechtviolet. Selected sections in transverse, sagittal and horizontal planes were charted with either an electronic pantograph or with the aid of a camera lucida. Spinal cord material from upper cervical levels and the brachial and lumbar enlargements was processed in a similar manner.
Horseradish peroxidase (HRP) studies Nine pigeons were anesthetized with sodium pentobarbital (35 mg/kg, i.p.1, the trachea intubated, and constant-volume ventilation applied. Depending on whether the HRP injection was t o be in brachial, thoracic or lumbar cord, the skin overlying the appropriate portion of the vertebral column was incised, the underlying ligaments retracted, and a laminectomy performed. To ensure stability during the remainder of the procedures, all animals were immobilized with pancuronium bromide (0.12 mg/kg/hr, i.m.1. Following this, a single injection of 1-2p l of a 33-50%solution of HRP (Sigma Type VI or Boehringer Mannheim) was made with a Hamilton syringe inserted into the right lateral funiculus. The injection was completed in approximately 30 minutes, and the needle was left in place for an additional 30 minutes. Five pigeons received an injection in the brachial cord, two in the thoracic cord and two in the lumbar cord. Following a survival period of 48 hours, the birds were sacrificed with an overdose of sodium pentobarbital (35 mg/kg, i.p.1 and perfused through an internal carotid artery with 300 ml of 2% dextran in avian saline followed by 500 ml of fixative. The perfusate was a cold solution of 4%glutaraldehyde, 5%sucrose and 1.2%CaCl (4mM) in 0.1 M Tris-HC1buffer a t pH 7.6. The calvaria were removed, the brains blocked stereotaxically in either transverse or sagittal planes, and the brains and spinal cords removed for overnight refrigeration in the fixative with 10%sucrose added. Serial sections were cut frozen a t 50 pm, collected in 0.1 M phosphate buffer a t pH 7.4,
AVIAN RUBROSPINAL TRACT
and then incubated for 30 minutes in 3,3‘-diaminobenzidine, H,Oz and 0.1 M phosphate buffer a t pH 7.4 (LaVail et al., ’73).After rinsing in distilled water, the sections were transferred to phosphate buffer from which they were mounted on chrome-alum coated slides and air-dried. Alternate sections were lightly counterstained with cresylechtviolet. RESULTS
Cytoarchitecture The nucleus ruber in transverse plane a t three rostrocaudal levels is illustrated in figure 1. Both the mediolateral and rostrocaudal extents of the nucleus are 1.0-1.5mm, and its conformation varies from round or oval a t the caudal pole (fig. 1A) to approximately triangular a t central and rostral levels (figs. lB,C). The nucleus is bordered medially by the fibers comprising the rootlets of NIII, ventromedially by the area ventralis of Tsai, ventrolaterally by the nucleus mesencephalicus profundus, pars ventralis and dorsally and dorsolaterally by the medial reticular formation. The right panels of figure 1 show that the nucleus is composed of generally polygonal neurons of varying size. The large neurons (40-50 pm in diameter) have a large, centrally located nucleus with a prominent nucleolus and typically have course, intensely staining Nissl bodies. These neurons are preferentially located dorsomedially (A1 cell group, fig. 1) and ventrolaterally (A2 cell group, fig. 1) a t more caudal nuclear levels, while small- and medium-sized neurons (15-35p m in diameter) are relatively less basophilic and are more prominent rostrally in the nucleus (B cell group, fig. 1).The medium-sized neurons generally have finer Nissl substance and tend to be surrounded by perineural satellites, a finding in agreement with previous reports (e.g., Miller and Strominger, ’73; King et al., ’71). While the large neurons are considered as comprising a magnocellular division (A1 and A2 cell groups) and the small- and mediumsized neurons a parvocellular division (B cell group), it should be emphasized that neurons of all sizes are present throughout the entire rostrocaudal extent of the nucleus. In particular, a t more rostral levels distinct subdivisions are less obvious. Course and terminations of rubralprojections Control experiments Lesions of the cerebellum, posterior com-
64 1
missure, nuclei of Darkschewitsch and Cajal, telencephalon or diencephalon failed to produce any evidence of contributions to the rubrospinal tract. Perhaps one of the more controversial projections of the nucleus ruber is the ipdateral “rubroolivary” tract (Edwards, ’72). Though dense degeneration was observed in the ipsilateral inferior olive, particularly in its dorsomedial portion, virtually identical patterns were also seen following lesions dorsal and anterior to the nucleus ruber. A similar pattern of degeneration was found following lesions of the nuclei of Darkschewitsch and Cajal. In such control cases there was no evident injury to the nucleus ruber, and no indication o f degeneration within t h e rubrospinal tract. Similar qualifications pertain to a possible “rubrofacial” projection (Courville, ’66). Lesions of nucleus ruber The optimal survival time for demonstrating degenerating axons of the rubrospinal tract appeared to be four to six days. Prior to that time only a few of the large axons show any argyrophilia. The efferents of the nucleus cross the midline and brush past the ventrocaudal aspect of the contralateral nucleus ruber (figs. ZA, 3A,B). On the basis of chartings of horizontal sections it appeared that recurrent collateral branches of the crossed projection enter the opposite nucleus ruber (fig. 3A). However, the proximity of the crossed fibers to the opposite ruber makes it likely that fibers of the opposite rubrospinal tract were also damaged, and the argyrophilic axons seen in the contralateral nucleus might reflect retrograde degeneration. Autoradiographic material following a large mesencephalic 3H-leucine injection a t the level of the nucleus (MiganiWall, Cohen and Ebbesson, unpublished observations) supports such an argument, since there was no evidence for a projection t o the contralateral nucleus ruber. After decussating, a compact bundle of degenerating fibers sweeps laterally, anteroventral to the brachium conjunctivum (figs. 2B, 3A,B). The descending tract then passes ventral t o the entering radix of the trigeminal nerve and traverses the rhombencephalon a t the ventrolateral aspect of the brainstem (figs. 2C, 3 0 . Within the rhombencephalon an extremely dense terminal field was noted in a restricted region of the rostral ventrolateral medulla (vea, figs. 2D, 3C) and a limited num-
642
J. M. WILD, J. B. CABOT, D. H. COHEN AND H. J. KARTEN
ber of degenerating fibers were observed a t the base of the nuclei cuneatus-gracilis (fig. 2E). A small stream of degenerating axons and terminals was also consistently noted within the plexus of Horsely (PHI, lying on the ventral aspects of the nucleus of the descending tract of the trigeminus (figs. 2D,E, 3B,C), and this degeneration in the plexus formed a column in direct continuity with the spinal laminar terminal fields. No clear evidence of projections upon the motor nucleus of the facial nerve was observed, and no degeneration was seen in the inferior olive contralatera1 t o the lesion. As discussed above, the dense terminal degeneration observed within the ipsilateral inferior olive could not be definitively attributed to damage to the nucleus ruber. At the caudal rhombencephalic levels the degenerating fibers sweep dorsomedially to the ascending spinocerebellar tract and enter the spinal cord as a well-defined bundle within the dorsal portion of the lateral funiculus (fig. 2F). At cervical, brachial and thoracic spinal levels the rubrospinal tract is separated from the outer margins of the cord by the fibers of the dorsal spinocerebellar tract, while a t lum-
bar levels this separation disappears, probably signifying the accumulating contributions of the ascending fibers from the dorsal magnocellular column, Clarke’s column (Leonard and Cohen, ’751, a t higher spinal levels. As indicated in figure 4,the terminations of the rubrospinal tract within the spinal grey are confined to three laminae (laminar definitions according t o Leonard and Cohen, ’75). These terminal fields are primarily in laminae V, VI and to a lesser extent the dorsolateral aspect of VII. Degenerating axons could be seen streaming into the grey matter with an orientation parallel to the dorsal horn and just ventral to it (fig. 5). Most axons appeared to enter lamina V, with relatively fewer penetrating more ventrally into lamina VI. A few degenerating fibers could be seen running dorsomedially, dorsal to Clarke’s column, but the majority of terminal and preterminal degeneration was confined to the medial portions of laminae V and VI. This pattern of termination was similar a t all spinal levels examined. Cells o f origin of the rubrospina.1 tract Figure 6 illustrates the distributions of rubral neurons labelled following HRP injec-
Abbreviations
AL, ansa lenticularis BCP, hrachium conjunctivum afferentes Cbd, tractus spinocerebellaris dorsalis cc, canalis centralis ClC, Clarke’s column CoTe, commissura tectalis CP, commissura posterior CuEx, nucleus cuneatus externus dh, dorsal horn EM, nucleus ectomammillaris fd, funiculus dorsalis fv, funiculus ventralis FLM, fasciculus longitudinalis medialis GCt, substantia grisea centralis ICo, nucleus intercollicularis Imc, nucleus isthmi, pars magnocellularis 10, nucleus isthmo-opticus IP. nucleus interpeduncularis Ipc, nucleus isthmi, pars parvocellularis LoC, locus coeruleus Mc, nucleus magnocellularis M U , nucleus mesencephali lateralis, pars dorsalis MPv, nucleus mesencephali profundus, pars ventralis MV, nucleus motorius nervi trigemini NIII, nervus oculomotorius NVI, nervus abducens NVIIIc, nervus octavua, pars cochlearis NIX, nervus glossopharyngei NX, nervus vagus nVII, nucleus nervi facialis
nX, nucleus motorius dorsalis nervi vagi 01, nucleus olivaris inferior OS, nucleus olivaris superior PH, plexus of Horsley PL, nucleus pontis lateralis PM, nucleus pontis medialis PrV, nucleus sensorius principalis nervi trigemini RdxV, radix nervi trigemini RdXVIIIv, radix nervus octavus, pars vestibularis Rgc, nucleus reticularia gigantocellularis Rpc, nucleus reticularis parvocellularis RPgc, nucleus reticularis pontis caudalis, pars gigantocellularis RST, tractus rubrmpinalis Ru, nucleus ruber S, nucleus solitarius Scd, nucleus subcoeruleus dorsalis SG, substantia gelatinma Rolandi (trigemini) SpL, nucleus spiriformis lateralis TeOp, tectum opticum TPC, nucleus tegmenti pedunculi-pontinus, pars compacta TPD, nucleus tegmenti pedunculi-pontinus, pars dissipata TTD, nucleus et tractus descendens nervi trigemini ventrIV (V), ventriculus yea, ventrolateral medullary area VeD, nucleus vestibularis descendens VeM, nucleus vestibularis medialis vh, ventral horn Ve Vm, nucleus vestibularis, para ventromedialis
Fig. 1 Nissl-stained transverse sections through nucleus ruber (Ru) a t caudal (A), middle (B), and rostra1 (C) levels. The numbers a t the upper left of each panel (A3.5, A4.0, A4.5) refer to the AP coordinates of Karten and Hodos ('67). A low power photograph (scale: 250 pm) is shown a t the left of each panel, with a higher power photograph (scale: 100 pm) of nucleus ruber a t the right. A , and A, designate, respectively, the dorsomedial and ventrolateral magnocellular regions of the nucleus; B designates the parvocellular region. See Abbreviations.
2 w
J. M. WILD, J. B. CABOT, D. H. COHEN AND H. J. KARTEN
AVIAN RUBROSPINAL TRACT
645
646
J. M. WILD, J. B. CABOT, D. H. COHEN AND H. J. KARTEN
B
A
C
w
P3.5
Fig. 2 Schematic illustrations in the transverse plane of degenerating fibers of the rubrospinal tract (RST) following a lesion of nucleus ruber (Ru), indicated by the blackened area in panel A. The numbers at the lower right of panels A-E (A3.5, A1.5, A1.O, P1.75, P3.5) refer to the AP coordinates of Karten and Hodos 037). Panel F shows a schematic section from high cervical spinal cord. See Abbreviations.
tions in the lateral funiculus at the brachial and lumbar enlargements. Though not illustrated, a similar pattern is found subsequent to lateral funicular injections a t thoracic levels. Three major observations emerge from these experiments. First, all retrogradely labelled neurons are contralateral to the spinal
injection site, confirming the totally crossed nature of the spinal projections of the nucleus ruber. Second, labelled neurons of all sizes were prominent throughout the rostrocaudal extent of the nucleus, indicating that the cells of origin of the rubrospinal tract are not restricted either to a particular cell size or sub-
AVIAN RUBROSPINAL TRACT
647
Iv
C
Fig. 3 Schematic illustrations in the horizontal plane of degeneratingfibera of the rubraspinal tract (RST) following a lesion of nucleus ruber (Ru), indicated by the blackened area in panel A. A is most dorsal and C most ventral. See
Abbreviations.
648
J. M. WILD, J. B. CABOT, D. H. COHEN AND H. J. KARTEN
Brachial
Lumbar
Fig. 4 Schematic illustrations of degeneration of the rubrospinal tract (RST)in brachial and lumbar spinal cord. The laminae of the spinal grey (I-IX) are according to Leonard and &hen (‘75). See Abbreuiations.
AVIAN RUBROSPINAL TRACT
649
650
J. M. WILD, J. B. CABOT, D. H. COHEN AND H. J. KARTEN
N. RUBER Brachial Injection
Lumbar lnjection
Fig. 6 Schematic illustration of the distributions of retrogradely labelled neurons in nucleus ruher following injections of HRP in the contralateral lateral funiculus at brachial and lumbar levels. For each panel the upper, middle and lower schematics represent, respectively, caudal, intermediate, and rwtral levels of the nucleus. The nuclear subdivisions are as in figure 1,and each dot indicates a labelled cell.
AVIAN RUBROSPINAL TRACT
division of the nucleus. Third, there did not appear to be differential patterns of labelled neurons as a function of the level of the spinal injection. That is, the distribution of labelled cells over the mediolateral and rostrocaudal extents of the nucleus was similar following lumbar, thoracic, or brachial injections. Thus, any topography which might exist would seem t o be subtle and include substantial overlap. DISCUSSION
The purpose of this study was (a) to describe the cytoarchitecture of the nucleus ruber in the pigeon, (b) to describe the course and sites of termination of the rubrospinal tract, and (c) t o investigate the topographic organization of the nuclear projections upon the spinal cord.
Cytoarchitecture Examination of the Nissl-stained sections of the nucleus revealed a gradation of cell sizes from 15-50 km. Some general accounts of the avian red nucleus either do not describe small cells (Ariens Kappers e t al., '60; Kuhlenbeck, '751, or state that a magnocellular nucleus is "typical" of avian forms (Craigie, '28). The reason for such discrepancies is not clear and is not easily resolved on the basis of species differences. Jungherr ('45), for instance, does not mention small cells in the nucleus ruber of the chicken, but a more recent study of the hen (Ovtscharoff and Gossrau, '72) does present histochemical and ultrastructural observations on small as well as medium and large cells. Zecha ('61) has also reported the presence of small cells in the nucleus ruber of the pigeon. On the basis of the predominant groupings of different sized neurons, the red nucleus of the pigeon may be subdivided into magnocellular and parvocellular portions, with the large cells being located dorsomedially and ventrolaterally at more caudal levels and the small- and medium-sized cells being more prominent at rostral nuclear levels. Such findings are in general agreement with those of other studies of the avian red nucleus (Zecha, '61; Ovtscharoff and Gossrau, '72). Again, it should be emphasized that cells of all sizes can be found throughout the entire rostrocaudal extent of the nucleus, such that strict boundaries between subdivisions are often difficult to discern. In mammals the nucleus ruber is, with some exceptions (Pompeiano and Brodal, '57; King et al., '711, also frequently described
651
as consisting of rostral parvocellular and caudal magnocellular divisions, but their relative proportions and cytoarchitectonic boundaries vary considerably. In monkeys the magnocellular and parvocellular divisions are clearly demarcated and have different descending projections, the magnocellular division mainly giving rise to the spinal projection and the parvocellular division t o the brainstem projection (Poirier and Bouvier, '66; Kuypers and Lawrence, '67). In the cat and rat, however, it has been reported that rubral neurons of all sizes project to the spinal cord (Pompeiano and Brodal, '57; Flumerfelt and Gwyn, '74; Petrovicky, '74), while in the opossum it has been claimed that the small neurons are intrinsic and that the large neurons give rise to the descending projections (King et al., '74; Martin et al., '74). In the present study, neurons representing the entire size spectrum were found to give rise to spinally projecting axons. This, however, does not preclude the possibility of there being neurons intrinsic to the nucleus, since in each case there were many cells of all sizes which were not retrogradely labelled.
Rubrospinal tract As in the cat and rat, the rubrospinal tract of the pigeon is composed of fibers originating from large, medium and small cells, and i t travels at least as far as lumbar levels of the spinal cord. Its brainstem course resembles that in several mammalian species, including primates (Miller and Strominger, '73; Murray and Haines, '75). However, its precise position in the lateral funiculus more closely resembles the position of the tract in rat, cat, possum and opossum (Petras, '67; Waldron and Gwyn, '69; Martin and Dom, '70; Warner and Watson, '72) than in primates, where it is displaced ventrally and ventrolaterally to the corticospinal tract (Kuypers et al., '62; Murray and Haines, '75). The laminar distribution of the terminals of rubrospinal axons within the spinal grey also resembles the pattern described in mammals. Based upon the cytoarchitectonic description of the pigeon spinal cord of Leonard and Cohen ('751, the present study shows terminal and preterminal degeneration of rubrospinal axons within the medial and lateral portions of spinal laminae V, VI and the dorsolateral portions of VII, a distribution in general agreement with that of carnivores (Nyberg-Hansen
652
J. M. WILD, J. 9. CABOT, D. H. COHEN AND H. J. KARTEN
and Brodal, '64; Petras, '67; Kostyuk and Skibo, '751, rodents (Brown, '741, marsupials (Martin and Dom, '70; Warner and Watson, '72) and primates (Kuypers et al., '62; Murray and Haines, '75). Organization of the cells of origin of the ru brospinal tract
The nucleus ruber of mammals is often reported to be topographically organized. In the cat Pompeiano and Brodal ('57) found that in the caudal three-fourths of the nucleus, dorsal and dorsomedial regions project t o cervical spinal cord, ventral and ventrolateral regions to lumbar cord, and intermediate regions to thoracic cord. Physiological studies (Padel e t al., '72;Eccles et al., '75; Ghez, '75) have confirmed this topography in the cat. Petrovicky ('74) suggested a similar pattern of projections in the rat, except that the most medial region of the nucleus was believed t o project to brainstem motor nuclei. In the tree shrew and lesser bushbaby (Murray and Haines, '75; Murray et al., '76) as well, dorsomedial regions project to cervical levels and ventrolateral regions more caudally. In the present study, however, a topographical organization of the nucleus was not evident. There were as many rubral neurons labelled from lumbar as from thoracic or brachial injections, and there were no consistently unlabelled regions following injections at cervical, thoracic or lumbar levels. Thus, it would appear that rubral cells of all sizes have spinal projections and that these cells do not have an obvious topographic organization. While these findings do not exclude the possibility of a more subtle topography, they a t least suggest more overlap in the distribution of spinal projections than in many mammals. Branching of rubrospinal axons with terminations a t multiple levels has recently been described in the cat (Shinoda et al., '771, and the absence of an obvious topography of the avian nucleus ruber could reflect particularly extensive branching of such axons in this vertebrate class. Brainstem projections Although the brainstem terminal fields of rubral efferents were not the primary focus of the present study, some comment is warranted. The silver degeneration results suggested that the contralateral plexus of Horsley receives a rubral projection. A continuous stream of degenerating fibers was found throughout its extent, and this degeneration
was in direct continuity with that in the spinal cord. Although little is known about the afferentation and efferentation of this region, the present results suggest it might be similar to the intermediate spinal grey. It should be noted, however, that preliminary data using autoradiographic and HRP techniques (Gold and Cohen, unpublished observations) suggest that this projection probably arises from peri-rubral neurons. As mentioned previously, a controversial brainstem projection of the mammalian nucleus ruber is that to the ipsilateral inferior olive. This has been reported in the monkey, cat and opossum (Walberg, '56; Courville, '66; Courville and Otabe, '74; Miller and Strominger, '73; Martin e t al., '741, and it was observed in the present material as well, However, Martin et al. ('74) declined to make definitive statements concerning the rubral origin of olivary degeneration, since tegmental lesions not involving t h e nucleus were likewise found to yield degeneration within the ipsilateral inferior olive. The results of the present investigation support such a position. A crossed projection to the motor facial nucleus has also been reported in a variety of mammals (Courville, '66; Mizuno and Nakamura, '71; Edwards, '72; Miller and Strominger, '73; Mizuno et al., '73; Martin et al., '741, but such a projection was not seen in the present study. Finally, there have been several reports of a descending projection upon the deep cerebellar nuclei (Brodal and Gogstad, '54; Hinman and Carpenter, '59; Courville and Brodal, '66; Martin e t al., '74). While degenerating fibers in the cerebellum were evident in the present material, they were not described since the rubral lesions often damaged axons of the nucleus spiriformis medialis which is known to project upon the cerebellum (Karten and Finger, '76).However, recent data based upon cerebellar HRP injections does in fact document a rubrocerebellar projection (Brecha and Karten, unpublished observations). In summary, the organization of the avian nucleus ruber and its projections seems generally comparable to that of many mammals. Furthermore, the pattern of rubral afferentation appears basically similar in the two vertebrate classes. In mammals rubral afferents arise principally from the sensorimotor cortex and from the cerebellar interpositus and dentate nuclei (Massion, '67). A corresponding pattern of afferentation is found in the bird.
AVIAN RUBROSPINAL TRACT
An anterior region of the Wulst projects via a basolateral branch of the septomesencephalic tract to various diencephalic nuclei, the nucleus ruber, the medial reticular formation, t h e medial pontine nucleus, t h e nuclei cuneatus-gracilis and eventually to the cervical spinal cord (Karten, '71). Karten et al. ('78) have recently reported that this region of the anterior Wulst in the owl receives a precise somatosensory representation of the contralateral toes. Moreover, the intermediate and lateral cerebellar nuclei project via the brachium conjunctivum to the contralateral nucleus ruber (Karten, '64). These afferent and efferent connections of the nucleus ruber thus suggest a fundamental similarity in organization of a significant component of the motor pathways over diverse species with widely differing motor behaviors. ACKNOWLEDGMENTS
The research reported here was supported by NSF Grant BNS 75-20537 to D. H. Cohen and NIH Grant NS-12087 to H. J. Karten. J. M. Wild and J. B. Cabot were supported by a grant from the Alfred P. Sloan Foundation to the University of Virginia Neuroscience Program. J. M. Wild was a Fulbright-Hays Scholar during the period of this research. We are indebted to Mrs. Doris Hannum for her processing of the histological material and to Mrs. Nancy Richardson for her secretarial assistance. LITERATURE CITED Ariens Kappers, C. U.,G. C. Huber and E. C. Crosby 1960 The Comparative Anatomyof the Nervous System of Vertebrates, Including Man. Hafner, New York. Brcdal, A., and A. C. Gogstad 1954 Rubro-cerebellar connection. An experimental study in the cat. Anat. Rec., 118: 455-486. Brown, L. T. 1974 Rubrospinal projections in the rat. J. Comp. Neur., 154: 169-177. Courville, J. 1966 Rubrobulbar fibers to the facial nucleus and the lateral reticular nucleus (nucleus of the lateral funiculus). An experimental study in the cat with silver impregnation methods. Brain Res., 1: 317-337. Courville, J., and A. Brodal 1966 Rubro-cerebellar connections in the cat. An experimental study with silver impregnation methods. J. Comp. Neur., 126: 471-485. Courville, J., and S. Otabe 1974 The rubro-olivary projection in the macaque: An experimental study with silver impregnation. J. Comp. Neur., 158: 479-494. Craigie, E. H. 1928 Observations on the brain of the humming bird fchrysolnmpis nosquitus Linn. and Chlorostilbon caribaeus Lawr.). J. Comp. Neur., 45: 377-481. 1930 Studies on the brain of the kiwi fApteryr australis). J. Comp. Neur., 49: 223-357. Eccles, J. C., T. Rantucci, P. Scheid and H. Taborikova 1975 Somatotopic studies on red nucleus: Spinal projection level and reapective receptive fields. J. Neurophysiol., 38: 965-980.
653
Edwards, S. B. 1972 The ascending and descending projections of red nucleus in the cat: a n experimental study using an autoradiographic tracing method. Brain Res., 48: 45-63. Flumerfelt, B. A,, and D. G. Gwyn 1974 The red nucleus of the rat: its organization and interconnections. J. Anat. (London), 118: 374 (Abstract). Ghez, C. 1975 Input-output relations of the red nucleus in the cat. Brain Res., 98: 93-108. Hinman, A,, and M. B. Carpenter 1959 Efferent fiber projections of the red nucleus in the cat. J. Comp. Neur., 113: 61-82. Jungherr, E. 1945 Certain nuclear groups of the avian mesencephalon. J. Comp. Neur., 82: 55-75. Karten, H.J. 1964 Projections of the cerebellar nuclei of the pigeon (Columba liuiai. Anat. Rec., 148: 297-298 (Abstract). 1971 Efferent projections of the Wulst of the owl. Anat. Rec., 169: 353 (Abstract). Karten, H. J., andT. E. Finger 1976 A direct thalamo-cerebellar pathway in pigeon and catfish. Brain Res., 102: 335-338. Karten, H. J., and W. Hodm 1967 A Stereotaxic Atlas of the Brain of the Pigeon (Columha livia). John Hopkins Press, Baltimore. Karten, H. J., M. Konishi and J. Pettigrew 1978 Somatosensory representation in the anterior Wulst of t h e owl fSpeotyto cunicularisl. Neurosci. Abstr., 4: 554 (Abstract). King, J. S., M. H. Bowman and G. F. Martin 1971 The red nucleus of the opossum (Didelphis marsupialis uirginiana): A light and electron microscopic study. J. Comp. Neur., 143: 157-184. King, J. S., R. M. Dom and G. F. Martin 1974 Anatomical evidence for an intrinsic neuron in the red nucleus. Brain Res., 67: 317-323. King, J. S., R. C. Schwyn and C. A. Fox 1971 The red nucleus in the monkey fMacaca mulatta): A Golgi and an electron microscopic study. J. Comp. Neur., 142: 75-108. Kostyuk, P. G., and G. G. Skibo 1975 An electron microscopic analysis of rubrospinal tract termination in the spinal cord of the cat. Brain Res., 85: 511-516. Kuhlenbeck, H. 1975 The Central Nervous System of Vertebrates. Vol. 4, Spinal Cord and Deuterencephalon. S. Karger, Basel. Kuypers, H. G. J. M., W. R. Fleming and J. W. Farinholt 1962 Subcortical projections in the rhesua monkey. J. Comp. Neur., 118: 107-137. Kuypers, H. G. J. M., and D. G . Lawrence 1967 Cortical projections to the red nucleus and the brain stem in the rhesus monkey. Brain Res., 4: 151-188. LaVail, J. H., K. R. Winston and A. Tish 1973 A method based on retrograde intraaxonal transport of protein for identification of cell bodiea of origin of axons terminating within the CNS. Brain Res., 58: 470-477. Leonard, R. B., and D. H. Cohen 1975 A cytoarchitectonic analysis of the spinal cord of the pigeon fColumba liuia). J. Comp. Neur., 163: 159-180. Martin, G. F., and R. Dom 1970 The rubrospinal tract of the opossum (Didelphis uirginianai. J. Comp. Neur., 138: 19-30. Martin, G. F., R. Dom, S. Katz and J. S. King 1974 Theorganization of projection neurons in the opossum red nucleus. Brain Res., 78: 17-34. Massion, J. 1967 The mammalian red nucleus. Physiol. Rev., 47: 383-436. Miller, R. A,, and N. L. Strominger 1973 Efferent connections of the red nucleus in the brainstem and spinal cord of t h e rhesus monkey. J. a m p . Neur., 152: 327-346.
654
J.
M. WILD, J. B. CABOT, D. H. COHEN AND H. J. KARTEN
Mizuno, N., A. Mcchizuki, R. Akimoto, N. Matsushima and Y. Nakamura 1973 Rubrobulbar projections in the rabbit. A lieht and electron microscoDic studv. J. ComD. Neur., l i 7 : 267-280. Mizuno, N., and Y. Nakamura 1971 Rubral fibers t o the facial nucleus in t h e rabbit. Brain Res., 28: 545-549. Murray, H.M. 1977 Rubrobulbar projections in the tree shrew (Tupaia). Neurosci. Abstr., 3: 59 (Abstract). Murray, H.M., and D. E. Haines 1975 The rubrospinal tract in a prosimian primate (Galago senegalensis]. Brain Behav. and Evol., 12: 311-333. Murray, H. M., D. E. Haines and I. Cote 1976 The rubrospinal tract of the tree shrew (Tupaia glis). Brain Res., 116: 317-322. Nyberg-Hansen, R., and A. Brodal 1964 Sites and modes of termination of rubrospinal fibers in the cat. An experimental study with silver impregnation methods. J. Anat. (London), 98: 235-253. Otabe, J. S.,and A. Horowitz 1970 Morphology and cytoarchitecture of the red nucleus of the domestic pig (Sus scrofa). J. Comp. Neur., 138: 373-389. Ovtscharoff, W., and R. Gossrau 1972 Histochemistry and ultrastructure of the red nucleus of hens (Gallus domesticud. Histochemie, 30: 73-81. Padel, Y., J. Armand and A. M. Smith 1972 Topography of rubrospinal units in thecat. Exp. Brain Res., 14: 363-371. Petras, J. M. 1967 Cortical, tectal and tegmental fiber connections in the spinal cord of the cat. Brain Res., 6: 275-324. Petrovicky, P. 1974 The origin and topographical organ-
ization of the rubrospinal tract in t h e rat. Folia Morphol.,
22: 314-316. Poirier, L. J., and G. Bouvier 1966 The red nucleus and its efferent nervous pathways in the monkey. J. Comp. Neur., 128: 223-244. Pompeiano, O., and A. Brodal 1957 Experimental demonstration of a somatotopical origin of rubrospinal fibers in the cat. J. Comp. Neur., 108: 225-251. Reid, J. M., D. G. Gwyn and B. A. Flumerfelt 1973 A cytoarchitectonic and Golgi study of the red nucleus i n the rat. J. Comp. Neur., 162: 337-362. Shincda, Y.,C. Ghez and A. Arnold 1977 Spinal branching of rubrospinal axons in the cat. Exp. Brain Res., 30: 203-218. Walberg, F. 1956 Descending connections to t h e inferior olive. An experimental study in the cat. J. Comp. Neur., 106: 77-174. Waldron, H.A,, and D. G . Gwyn 1969 Descending nerve tracts in the spinal cord of the rat. I. Fibers from the midbrain. J. Comp. Neur., 137: 143-154. Warner, G., and C. R. R. Watson 1972 The rubrospinal tract in a diprodont marsupial OXchosaurus uulpecula). Brain Res., 41: 180-183. Wild, J. M.,J. B. Cabot, D. H. Cohen and H. J. Karten 1977 The avian rubrospinal tract: Its cells of origin, course and sites of termination. Neurosci. Abstr., 3: 280 (Abstract). Wold, J. E. 1978 The vestibular nuclei in the domestic hen (Gallus domesticus). IV. The projection to the spinal cord. Brain, Behav. and Evol., 15: 41-62. Zecha, A. 1961 Bezit een vogel een fasciculus rubro-bulbo-spinalis? Ned. J. Geneesk., 105: 2373.
ABSTRACT The red nucleus and its spinal projections in the pigeon (Cob umba livia) have been studied using both normal and experimental material. The cytoarchitecture of the nucleus is described on the basis of Nissl-stained sections and reveals an organization generally similar to that of mammals. The large neurons (40-50 pm) tend to be located dorsomedially and ventrolaterally a t more caudal nuclear levels, while the small- and medium-sized neurons (1535 pm) predominate a t rostra1 levels. However, neurons of all sizes are present throughout the nucleus. Following lesions of the nucleus, the course of degenerating axons stained with the Fink-Heimer method has been traced throughout the brainstem and spinal cord. The rubrospinal tract crosses the midline, courses past the ventrocaudal aspect of the contralateral nucleus ruber, and then descends rostroventral and lateral to the nucleus tegmenti pontinus. In its caudal continuation the tract lies ventral to the brachium conjunctivum and the entering radix of the trigeminal nerve. I t then assumes a ventrolateral position in the caudal brainstem before shifting to a dorsolateral position in the lateral funiculus of the spinal cord. Within the spinal grey the rubrospinal tract terminates in laminae V, VI and to a lesser extent VII. The possibility of a topographical organization of the nucleus was investigated with injections of horseradish peroxidase into brachial, thoracic and lumbar spinal cord. Regardless of the level of injection, labelled neurons of all sizes were present throughout the contralateral nucleus ruber, indicating the absence of an obvious topography. The organization of the nucleus ruber and its projections have been studied extensively in mammals including, for example, rodents (Mizuno and Nakamura, '71; Mizuno et al., '73; Reid et al., '73; Brown, '74; Petrovicky, '74), carnovires (Edwards, '72), marsupials (Martin and Dom, '70; King et al., '71; Warner and Watson, '72; King et al., '74; Martin e t al., '741, ungulates (Otabe and Horowitz, '70), tree shrew (Murray et al., '76; Murray, '77) and primates (King et al., '71; Miller and Strominger, '73). (See Massion "671 for an earlier review.) In contrast, there have been few reports on the avian nucleus ruber save for the earlier non-experimental studies of Craigie ('28, '30) and Jungherr ('45). The description of the nucleus by these authors, however, was limited and within the context of a general J. COMP. NEUR. (1979) 187: 639-654.
treatment of the mesencephalic nuclei. A more recent report by Ovtscharoff and Gossrau ('72) has described the cytoarchitecture of the red nucleus of the hen (Gallus domesticus), and an avian rubrospinal tract has been briefly mentioned in reports by Zecha ('61) and Wold ('78). However, there are no studies providing a detailed description of the avian rubrospinal tract, and thus the present study was undertaken to describe its cells of origin, course and sites of termination in the pigeon (Columba livia) as part of a more ' This material was reported at the Seventh Annual Meeting of the k i e t y for Neuroscience, Annaheim, California (Wildet al., '77). 'Present addraw: Department of Neurobiology and Behavior, SUNY at Stony Brook, Stony Brook, New York 11794. 6Please address all reprint requesta to: Dr. David H. Cohen, Department of Neurobiology and Behavior, Graduate Biology Building, SUNY at Stony Brwk, Stony Brook, New York 11794.
639
640
J. M. WILD, J. B. CABOT, D. H. COHEN AND H. J. KARTEN
extensive investigation of the avian brainstem projections to the spinal cord. The general similarities of the avian and mammalian rubrospinal systems are also discussed. MATERIALS AND METHODS
Experimental animals All normal and experimental material was obtained from White Carneaux pigeons (CoEumba livia) ranging in age from two months to four years and weighing 400-700 gm.They were obtained from the Palmetto Pigeon Plant, Sumter, South Carolina. Cytoarchitectonic analysis The cytoarchitecture of the nucleus ruber was based upon light microscopic examination of normal material cut in transverse, sagittal and horizontal planes. In three animals 50-, 75- and 100-pm thick, serial transverse sections were prepared from celloidin embedded brains, and in an additional three animals 25-pm thick serial sections in transverse, sagittal and horizontal planes were examined with paraffin embedded material. All sections were stained with cresylechtviolet, and drawings of the nucleus were made from every second or third section with the aid of a Leitz drawing tube. In all cases the birds were sacrificed with an overdose of sodium pentobarbital and perfused through the left ventricle or internal carotid artery with avian saline followed by 10%formalin. The calvaria were then removed and the brains blocked stereotaxically (Karten and Hodos, '67) and embedded in either paraffin or celloidin. Anterograde degeneration Stereotaxically guided electrolytic lesions were placed in the nucleus ruber of 15pigeons. Electrodes were inserted via several different trajectories to control for damage to various adjacent nuclei and fiber tracts. Thus, in addition to an ipsilateral vertical penetration in the transverse plane of Karten and Hodos ('67), electrode penetrations were made from the contralateral telencephalon or at different angles through the ipsilateral telencephalon to avoid damage to the cerebellum and to the nuclei of Darkschewitsch and Cajal. Additional controls included lesions of the cerebellum, posterior commissure, nuclei of Darkschewitsch and Cajal, pre-rubral fields, lateral tegmentum, telencephalon and diencephalon.
The birds were sacrificed a s described above, the calvaria removed and the brains blocked stereotaxically in either transverse, sagittal or horizontal planes. The brains and spinal cords were removed and either embedded in a gelatin-albumin casement and cut frozen a t 25 fim, or postfixed in 20% sucrose for 24 hours and then cut frozen a t 33 pm. Serial brainstem sections a t intervals of 200 pm were stained according to the FinkHeimer method I, and each adjacent section was stained with cresylechtviolet. Selected sections in transverse, sagittal and horizontal planes were charted with either an electronic pantograph or with the aid of a camera lucida. Spinal cord material from upper cervical levels and the brachial and lumbar enlargements was processed in a similar manner.
Horseradish peroxidase (HRP) studies Nine pigeons were anesthetized with sodium pentobarbital (35 mg/kg, i.p.1, the trachea intubated, and constant-volume ventilation applied. Depending on whether the HRP injection was t o be in brachial, thoracic or lumbar cord, the skin overlying the appropriate portion of the vertebral column was incised, the underlying ligaments retracted, and a laminectomy performed. To ensure stability during the remainder of the procedures, all animals were immobilized with pancuronium bromide (0.12 mg/kg/hr, i.m.1. Following this, a single injection of 1-2p l of a 33-50%solution of HRP (Sigma Type VI or Boehringer Mannheim) was made with a Hamilton syringe inserted into the right lateral funiculus. The injection was completed in approximately 30 minutes, and the needle was left in place for an additional 30 minutes. Five pigeons received an injection in the brachial cord, two in the thoracic cord and two in the lumbar cord. Following a survival period of 48 hours, the birds were sacrificed with an overdose of sodium pentobarbital (35 mg/kg, i.p.1 and perfused through an internal carotid artery with 300 ml of 2% dextran in avian saline followed by 500 ml of fixative. The perfusate was a cold solution of 4%glutaraldehyde, 5%sucrose and 1.2%CaCl (4mM) in 0.1 M Tris-HC1buffer a t pH 7.6. The calvaria were removed, the brains blocked stereotaxically in either transverse or sagittal planes, and the brains and spinal cords removed for overnight refrigeration in the fixative with 10%sucrose added. Serial sections were cut frozen a t 50 pm, collected in 0.1 M phosphate buffer a t pH 7.4,
AVIAN RUBROSPINAL TRACT
and then incubated for 30 minutes in 3,3‘-diaminobenzidine, H,Oz and 0.1 M phosphate buffer a t pH 7.4 (LaVail et al., ’73).After rinsing in distilled water, the sections were transferred to phosphate buffer from which they were mounted on chrome-alum coated slides and air-dried. Alternate sections were lightly counterstained with cresylechtviolet. RESULTS
Cytoarchitecture The nucleus ruber in transverse plane a t three rostrocaudal levels is illustrated in figure 1. Both the mediolateral and rostrocaudal extents of the nucleus are 1.0-1.5mm, and its conformation varies from round or oval a t the caudal pole (fig. 1A) to approximately triangular a t central and rostral levels (figs. lB,C). The nucleus is bordered medially by the fibers comprising the rootlets of NIII, ventromedially by the area ventralis of Tsai, ventrolaterally by the nucleus mesencephalicus profundus, pars ventralis and dorsally and dorsolaterally by the medial reticular formation. The right panels of figure 1 show that the nucleus is composed of generally polygonal neurons of varying size. The large neurons (40-50 pm in diameter) have a large, centrally located nucleus with a prominent nucleolus and typically have course, intensely staining Nissl bodies. These neurons are preferentially located dorsomedially (A1 cell group, fig. 1) and ventrolaterally (A2 cell group, fig. 1) a t more caudal nuclear levels, while small- and medium-sized neurons (15-35p m in diameter) are relatively less basophilic and are more prominent rostrally in the nucleus (B cell group, fig. 1).The medium-sized neurons generally have finer Nissl substance and tend to be surrounded by perineural satellites, a finding in agreement with previous reports (e.g., Miller and Strominger, ’73; King et al., ’71). While the large neurons are considered as comprising a magnocellular division (A1 and A2 cell groups) and the small- and mediumsized neurons a parvocellular division (B cell group), it should be emphasized that neurons of all sizes are present throughout the entire rostrocaudal extent of the nucleus. In particular, a t more rostral levels distinct subdivisions are less obvious. Course and terminations of rubralprojections Control experiments Lesions of the cerebellum, posterior com-
64 1
missure, nuclei of Darkschewitsch and Cajal, telencephalon or diencephalon failed to produce any evidence of contributions to the rubrospinal tract. Perhaps one of the more controversial projections of the nucleus ruber is the ipdateral “rubroolivary” tract (Edwards, ’72). Though dense degeneration was observed in the ipsilateral inferior olive, particularly in its dorsomedial portion, virtually identical patterns were also seen following lesions dorsal and anterior to the nucleus ruber. A similar pattern of degeneration was found following lesions of the nuclei of Darkschewitsch and Cajal. In such control cases there was no evident injury to the nucleus ruber, and no indication o f degeneration within t h e rubrospinal tract. Similar qualifications pertain to a possible “rubrofacial” projection (Courville, ’66). Lesions of nucleus ruber The optimal survival time for demonstrating degenerating axons of the rubrospinal tract appeared to be four to six days. Prior to that time only a few of the large axons show any argyrophilia. The efferents of the nucleus cross the midline and brush past the ventrocaudal aspect of the contralateral nucleus ruber (figs. ZA, 3A,B). On the basis of chartings of horizontal sections it appeared that recurrent collateral branches of the crossed projection enter the opposite nucleus ruber (fig. 3A). However, the proximity of the crossed fibers to the opposite ruber makes it likely that fibers of the opposite rubrospinal tract were also damaged, and the argyrophilic axons seen in the contralateral nucleus might reflect retrograde degeneration. Autoradiographic material following a large mesencephalic 3H-leucine injection a t the level of the nucleus (MiganiWall, Cohen and Ebbesson, unpublished observations) supports such an argument, since there was no evidence for a projection t o the contralateral nucleus ruber. After decussating, a compact bundle of degenerating fibers sweeps laterally, anteroventral to the brachium conjunctivum (figs. 2B, 3A,B). The descending tract then passes ventral t o the entering radix of the trigeminal nerve and traverses the rhombencephalon a t the ventrolateral aspect of the brainstem (figs. 2C, 3 0 . Within the rhombencephalon an extremely dense terminal field was noted in a restricted region of the rostral ventrolateral medulla (vea, figs. 2D, 3C) and a limited num-
642
J. M. WILD, J. B. CABOT, D. H. COHEN AND H. J. KARTEN
ber of degenerating fibers were observed a t the base of the nuclei cuneatus-gracilis (fig. 2E). A small stream of degenerating axons and terminals was also consistently noted within the plexus of Horsely (PHI, lying on the ventral aspects of the nucleus of the descending tract of the trigeminus (figs. 2D,E, 3B,C), and this degeneration in the plexus formed a column in direct continuity with the spinal laminar terminal fields. No clear evidence of projections upon the motor nucleus of the facial nerve was observed, and no degeneration was seen in the inferior olive contralatera1 t o the lesion. As discussed above, the dense terminal degeneration observed within the ipsilateral inferior olive could not be definitively attributed to damage to the nucleus ruber. At the caudal rhombencephalic levels the degenerating fibers sweep dorsomedially to the ascending spinocerebellar tract and enter the spinal cord as a well-defined bundle within the dorsal portion of the lateral funiculus (fig. 2F). At cervical, brachial and thoracic spinal levels the rubrospinal tract is separated from the outer margins of the cord by the fibers of the dorsal spinocerebellar tract, while a t lum-
bar levels this separation disappears, probably signifying the accumulating contributions of the ascending fibers from the dorsal magnocellular column, Clarke’s column (Leonard and Cohen, ’751, a t higher spinal levels. As indicated in figure 4,the terminations of the rubrospinal tract within the spinal grey are confined to three laminae (laminar definitions according t o Leonard and Cohen, ’75). These terminal fields are primarily in laminae V, VI and to a lesser extent the dorsolateral aspect of VII. Degenerating axons could be seen streaming into the grey matter with an orientation parallel to the dorsal horn and just ventral to it (fig. 5). Most axons appeared to enter lamina V, with relatively fewer penetrating more ventrally into lamina VI. A few degenerating fibers could be seen running dorsomedially, dorsal to Clarke’s column, but the majority of terminal and preterminal degeneration was confined to the medial portions of laminae V and VI. This pattern of termination was similar a t all spinal levels examined. Cells o f origin of the rubrospina.1 tract Figure 6 illustrates the distributions of rubral neurons labelled following HRP injec-
Abbreviations
AL, ansa lenticularis BCP, hrachium conjunctivum afferentes Cbd, tractus spinocerebellaris dorsalis cc, canalis centralis ClC, Clarke’s column CoTe, commissura tectalis CP, commissura posterior CuEx, nucleus cuneatus externus dh, dorsal horn EM, nucleus ectomammillaris fd, funiculus dorsalis fv, funiculus ventralis FLM, fasciculus longitudinalis medialis GCt, substantia grisea centralis ICo, nucleus intercollicularis Imc, nucleus isthmi, pars magnocellularis 10, nucleus isthmo-opticus IP. nucleus interpeduncularis Ipc, nucleus isthmi, pars parvocellularis LoC, locus coeruleus Mc, nucleus magnocellularis M U , nucleus mesencephali lateralis, pars dorsalis MPv, nucleus mesencephali profundus, pars ventralis MV, nucleus motorius nervi trigemini NIII, nervus oculomotorius NVI, nervus abducens NVIIIc, nervus octavua, pars cochlearis NIX, nervus glossopharyngei NX, nervus vagus nVII, nucleus nervi facialis
nX, nucleus motorius dorsalis nervi vagi 01, nucleus olivaris inferior OS, nucleus olivaris superior PH, plexus of Horsley PL, nucleus pontis lateralis PM, nucleus pontis medialis PrV, nucleus sensorius principalis nervi trigemini RdxV, radix nervi trigemini RdXVIIIv, radix nervus octavus, pars vestibularis Rgc, nucleus reticularia gigantocellularis Rpc, nucleus reticularis parvocellularis RPgc, nucleus reticularis pontis caudalis, pars gigantocellularis RST, tractus rubrmpinalis Ru, nucleus ruber S, nucleus solitarius Scd, nucleus subcoeruleus dorsalis SG, substantia gelatinma Rolandi (trigemini) SpL, nucleus spiriformis lateralis TeOp, tectum opticum TPC, nucleus tegmenti pedunculi-pontinus, pars compacta TPD, nucleus tegmenti pedunculi-pontinus, pars dissipata TTD, nucleus et tractus descendens nervi trigemini ventrIV (V), ventriculus yea, ventrolateral medullary area VeD, nucleus vestibularis descendens VeM, nucleus vestibularis medialis vh, ventral horn Ve Vm, nucleus vestibularis, para ventromedialis
Fig. 1 Nissl-stained transverse sections through nucleus ruber (Ru) a t caudal (A), middle (B), and rostra1 (C) levels. The numbers a t the upper left of each panel (A3.5, A4.0, A4.5) refer to the AP coordinates of Karten and Hodos ('67). A low power photograph (scale: 250 pm) is shown a t the left of each panel, with a higher power photograph (scale: 100 pm) of nucleus ruber a t the right. A , and A, designate, respectively, the dorsomedial and ventrolateral magnocellular regions of the nucleus; B designates the parvocellular region. See Abbreviations.
2 w
J. M. WILD, J. B. CABOT, D. H. COHEN AND H. J. KARTEN
AVIAN RUBROSPINAL TRACT
645
646
J. M. WILD, J. B. CABOT, D. H. COHEN AND H. J. KARTEN
B
A
C
w
P3.5
Fig. 2 Schematic illustrations in the transverse plane of degenerating fibers of the rubrospinal tract (RST) following a lesion of nucleus ruber (Ru), indicated by the blackened area in panel A. The numbers at the lower right of panels A-E (A3.5, A1.5, A1.O, P1.75, P3.5) refer to the AP coordinates of Karten and Hodos 037). Panel F shows a schematic section from high cervical spinal cord. See Abbreviations.
tions in the lateral funiculus at the brachial and lumbar enlargements. Though not illustrated, a similar pattern is found subsequent to lateral funicular injections a t thoracic levels. Three major observations emerge from these experiments. First, all retrogradely labelled neurons are contralateral to the spinal
injection site, confirming the totally crossed nature of the spinal projections of the nucleus ruber. Second, labelled neurons of all sizes were prominent throughout the rostrocaudal extent of the nucleus, indicating that the cells of origin of the rubrospinal tract are not restricted either to a particular cell size or sub-
AVIAN RUBROSPINAL TRACT
647
Iv
C
Fig. 3 Schematic illustrations in the horizontal plane of degeneratingfibera of the rubraspinal tract (RST) following a lesion of nucleus ruber (Ru), indicated by the blackened area in panel A. A is most dorsal and C most ventral. See
Abbreviations.
648
J. M. WILD, J. B. CABOT, D. H. COHEN AND H. J. KARTEN
Brachial
Lumbar
Fig. 4 Schematic illustrations of degeneration of the rubrospinal tract (RST)in brachial and lumbar spinal cord. The laminae of the spinal grey (I-IX) are according to Leonard and &hen (‘75). See Abbreuiations.
AVIAN RUBROSPINAL TRACT
649
650
J. M. WILD, J. B. CABOT, D. H. COHEN AND H. J. KARTEN
N. RUBER Brachial Injection
Lumbar lnjection
Fig. 6 Schematic illustration of the distributions of retrogradely labelled neurons in nucleus ruher following injections of HRP in the contralateral lateral funiculus at brachial and lumbar levels. For each panel the upper, middle and lower schematics represent, respectively, caudal, intermediate, and rwtral levels of the nucleus. The nuclear subdivisions are as in figure 1,and each dot indicates a labelled cell.
AVIAN RUBROSPINAL TRACT
division of the nucleus. Third, there did not appear to be differential patterns of labelled neurons as a function of the level of the spinal injection. That is, the distribution of labelled cells over the mediolateral and rostrocaudal extents of the nucleus was similar following lumbar, thoracic, or brachial injections. Thus, any topography which might exist would seem t o be subtle and include substantial overlap. DISCUSSION
The purpose of this study was (a) to describe the cytoarchitecture of the nucleus ruber in the pigeon, (b) to describe the course and sites of termination of the rubrospinal tract, and (c) t o investigate the topographic organization of the nuclear projections upon the spinal cord.
Cytoarchitecture Examination of the Nissl-stained sections of the nucleus revealed a gradation of cell sizes from 15-50 km. Some general accounts of the avian red nucleus either do not describe small cells (Ariens Kappers e t al., '60; Kuhlenbeck, '751, or state that a magnocellular nucleus is "typical" of avian forms (Craigie, '28). The reason for such discrepancies is not clear and is not easily resolved on the basis of species differences. Jungherr ('45), for instance, does not mention small cells in the nucleus ruber of the chicken, but a more recent study of the hen (Ovtscharoff and Gossrau, '72) does present histochemical and ultrastructural observations on small as well as medium and large cells. Zecha ('61) has also reported the presence of small cells in the nucleus ruber of the pigeon. On the basis of the predominant groupings of different sized neurons, the red nucleus of the pigeon may be subdivided into magnocellular and parvocellular portions, with the large cells being located dorsomedially and ventrolaterally at more caudal levels and the small- and medium-sized cells being more prominent at rostral nuclear levels. Such findings are in general agreement with those of other studies of the avian red nucleus (Zecha, '61; Ovtscharoff and Gossrau, '72). Again, it should be emphasized that cells of all sizes can be found throughout the entire rostrocaudal extent of the nucleus, such that strict boundaries between subdivisions are often difficult to discern. In mammals the nucleus ruber is, with some exceptions (Pompeiano and Brodal, '57; King et al., '711, also frequently described
651
as consisting of rostral parvocellular and caudal magnocellular divisions, but their relative proportions and cytoarchitectonic boundaries vary considerably. In monkeys the magnocellular and parvocellular divisions are clearly demarcated and have different descending projections, the magnocellular division mainly giving rise to the spinal projection and the parvocellular division t o the brainstem projection (Poirier and Bouvier, '66; Kuypers and Lawrence, '67). In the cat and rat, however, it has been reported that rubral neurons of all sizes project to the spinal cord (Pompeiano and Brodal, '57; Flumerfelt and Gwyn, '74; Petrovicky, '74), while in the opossum it has been claimed that the small neurons are intrinsic and that the large neurons give rise to the descending projections (King et al., '74; Martin et al., '74). In the present study, neurons representing the entire size spectrum were found to give rise to spinally projecting axons. This, however, does not preclude the possibility of there being neurons intrinsic to the nucleus, since in each case there were many cells of all sizes which were not retrogradely labelled.
Rubrospinal tract As in the cat and rat, the rubrospinal tract of the pigeon is composed of fibers originating from large, medium and small cells, and i t travels at least as far as lumbar levels of the spinal cord. Its brainstem course resembles that in several mammalian species, including primates (Miller and Strominger, '73; Murray and Haines, '75). However, its precise position in the lateral funiculus more closely resembles the position of the tract in rat, cat, possum and opossum (Petras, '67; Waldron and Gwyn, '69; Martin and Dom, '70; Warner and Watson, '72) than in primates, where it is displaced ventrally and ventrolaterally to the corticospinal tract (Kuypers et al., '62; Murray and Haines, '75). The laminar distribution of the terminals of rubrospinal axons within the spinal grey also resembles the pattern described in mammals. Based upon the cytoarchitectonic description of the pigeon spinal cord of Leonard and Cohen ('751, the present study shows terminal and preterminal degeneration of rubrospinal axons within the medial and lateral portions of spinal laminae V, VI and the dorsolateral portions of VII, a distribution in general agreement with that of carnivores (Nyberg-Hansen
652
J. M. WILD, J. 9. CABOT, D. H. COHEN AND H. J. KARTEN
and Brodal, '64; Petras, '67; Kostyuk and Skibo, '751, rodents (Brown, '741, marsupials (Martin and Dom, '70; Warner and Watson, '72) and primates (Kuypers et al., '62; Murray and Haines, '75). Organization of the cells of origin of the ru brospinal tract
The nucleus ruber of mammals is often reported to be topographically organized. In the cat Pompeiano and Brodal ('57) found that in the caudal three-fourths of the nucleus, dorsal and dorsomedial regions project t o cervical spinal cord, ventral and ventrolateral regions to lumbar cord, and intermediate regions to thoracic cord. Physiological studies (Padel e t al., '72;Eccles et al., '75; Ghez, '75) have confirmed this topography in the cat. Petrovicky ('74) suggested a similar pattern of projections in the rat, except that the most medial region of the nucleus was believed t o project to brainstem motor nuclei. In the tree shrew and lesser bushbaby (Murray and Haines, '75; Murray et al., '76) as well, dorsomedial regions project to cervical levels and ventrolateral regions more caudally. In the present study, however, a topographical organization of the nucleus was not evident. There were as many rubral neurons labelled from lumbar as from thoracic or brachial injections, and there were no consistently unlabelled regions following injections at cervical, thoracic or lumbar levels. Thus, it would appear that rubral cells of all sizes have spinal projections and that these cells do not have an obvious topographic organization. While these findings do not exclude the possibility of a more subtle topography, they a t least suggest more overlap in the distribution of spinal projections than in many mammals. Branching of rubrospinal axons with terminations a t multiple levels has recently been described in the cat (Shinoda et al., '771, and the absence of an obvious topography of the avian nucleus ruber could reflect particularly extensive branching of such axons in this vertebrate class. Brainstem projections Although the brainstem terminal fields of rubral efferents were not the primary focus of the present study, some comment is warranted. The silver degeneration results suggested that the contralateral plexus of Horsley receives a rubral projection. A continuous stream of degenerating fibers was found throughout its extent, and this degeneration
was in direct continuity with that in the spinal cord. Although little is known about the afferentation and efferentation of this region, the present results suggest it might be similar to the intermediate spinal grey. It should be noted, however, that preliminary data using autoradiographic and HRP techniques (Gold and Cohen, unpublished observations) suggest that this projection probably arises from peri-rubral neurons. As mentioned previously, a controversial brainstem projection of the mammalian nucleus ruber is that to the ipsilateral inferior olive. This has been reported in the monkey, cat and opossum (Walberg, '56; Courville, '66; Courville and Otabe, '74; Miller and Strominger, '73; Martin e t al., '741, and it was observed in the present material as well, However, Martin et al. ('74) declined to make definitive statements concerning the rubral origin of olivary degeneration, since tegmental lesions not involving t h e nucleus were likewise found to yield degeneration within the ipsilateral inferior olive. The results of the present investigation support such a position. A crossed projection to the motor facial nucleus has also been reported in a variety of mammals (Courville, '66; Mizuno and Nakamura, '71; Edwards, '72; Miller and Strominger, '73; Mizuno et al., '73; Martin et al., '741, but such a projection was not seen in the present study. Finally, there have been several reports of a descending projection upon the deep cerebellar nuclei (Brodal and Gogstad, '54; Hinman and Carpenter, '59; Courville and Brodal, '66; Martin e t al., '74). While degenerating fibers in the cerebellum were evident in the present material, they were not described since the rubral lesions often damaged axons of the nucleus spiriformis medialis which is known to project upon the cerebellum (Karten and Finger, '76).However, recent data based upon cerebellar HRP injections does in fact document a rubrocerebellar projection (Brecha and Karten, unpublished observations). In summary, the organization of the avian nucleus ruber and its projections seems generally comparable to that of many mammals. Furthermore, the pattern of rubral afferentation appears basically similar in the two vertebrate classes. In mammals rubral afferents arise principally from the sensorimotor cortex and from the cerebellar interpositus and dentate nuclei (Massion, '67). A corresponding pattern of afferentation is found in the bird.
AVIAN RUBROSPINAL TRACT
An anterior region of the Wulst projects via a basolateral branch of the septomesencephalic tract to various diencephalic nuclei, the nucleus ruber, the medial reticular formation, t h e medial pontine nucleus, t h e nuclei cuneatus-gracilis and eventually to the cervical spinal cord (Karten, '71). Karten et al. ('78) have recently reported that this region of the anterior Wulst in the owl receives a precise somatosensory representation of the contralateral toes. Moreover, the intermediate and lateral cerebellar nuclei project via the brachium conjunctivum to the contralateral nucleus ruber (Karten, '64). These afferent and efferent connections of the nucleus ruber thus suggest a fundamental similarity in organization of a significant component of the motor pathways over diverse species with widely differing motor behaviors. ACKNOWLEDGMENTS
The research reported here was supported by NSF Grant BNS 75-20537 to D. H. Cohen and NIH Grant NS-12087 to H. J. Karten. J. M. Wild and J. B. Cabot were supported by a grant from the Alfred P. Sloan Foundation to the University of Virginia Neuroscience Program. J. M. Wild was a Fulbright-Hays Scholar during the period of this research. We are indebted to Mrs. Doris Hannum for her processing of the histological material and to Mrs. Nancy Richardson for her secretarial assistance. LITERATURE CITED Ariens Kappers, C. U.,G. C. Huber and E. C. Crosby 1960 The Comparative Anatomyof the Nervous System of Vertebrates, Including Man. Hafner, New York. Brcdal, A., and A. C. Gogstad 1954 Rubro-cerebellar connection. An experimental study in the cat. Anat. Rec., 118: 455-486. Brown, L. T. 1974 Rubrospinal projections in the rat. J. Comp. Neur., 154: 169-177. Courville, J. 1966 Rubrobulbar fibers to the facial nucleus and the lateral reticular nucleus (nucleus of the lateral funiculus). An experimental study in the cat with silver impregnation methods. Brain Res., 1: 317-337. Courville, J., and A. Brodal 1966 Rubro-cerebellar connections in the cat. An experimental study with silver impregnation methods. J. Comp. Neur., 126: 471-485. Courville, J., and S. Otabe 1974 The rubro-olivary projection in the macaque: An experimental study with silver impregnation. J. Comp. Neur., 158: 479-494. Craigie, E. H. 1928 Observations on the brain of the humming bird fchrysolnmpis nosquitus Linn. and Chlorostilbon caribaeus Lawr.). J. Comp. Neur., 45: 377-481. 1930 Studies on the brain of the kiwi fApteryr australis). J. Comp. Neur., 49: 223-357. Eccles, J. C., T. Rantucci, P. Scheid and H. Taborikova 1975 Somatotopic studies on red nucleus: Spinal projection level and reapective receptive fields. J. Neurophysiol., 38: 965-980.
653
Edwards, S. B. 1972 The ascending and descending projections of red nucleus in the cat: a n experimental study using an autoradiographic tracing method. Brain Res., 48: 45-63. Flumerfelt, B. A,, and D. G. Gwyn 1974 The red nucleus of the rat: its organization and interconnections. J. Anat. (London), 118: 374 (Abstract). Ghez, C. 1975 Input-output relations of the red nucleus in the cat. Brain Res., 98: 93-108. Hinman, A,, and M. B. Carpenter 1959 Efferent fiber projections of the red nucleus in the cat. J. Comp. Neur., 113: 61-82. Jungherr, E. 1945 Certain nuclear groups of the avian mesencephalon. J. Comp. Neur., 82: 55-75. Karten, H.J. 1964 Projections of the cerebellar nuclei of the pigeon (Columba liuiai. Anat. Rec., 148: 297-298 (Abstract). 1971 Efferent projections of the Wulst of the owl. Anat. Rec., 169: 353 (Abstract). Karten, H. J., andT. E. Finger 1976 A direct thalamo-cerebellar pathway in pigeon and catfish. Brain Res., 102: 335-338. Karten, H. J., and W. Hodm 1967 A Stereotaxic Atlas of the Brain of the Pigeon (Columha livia). John Hopkins Press, Baltimore. Karten, H. J., M. Konishi and J. Pettigrew 1978 Somatosensory representation in the anterior Wulst of t h e owl fSpeotyto cunicularisl. Neurosci. Abstr., 4: 554 (Abstract). King, J. S., M. H. Bowman and G. F. Martin 1971 The red nucleus of the opossum (Didelphis marsupialis uirginiana): A light and electron microscopic study. J. Comp. Neur., 143: 157-184. King, J. S., R. M. Dom and G. F. Martin 1974 Anatomical evidence for an intrinsic neuron in the red nucleus. Brain Res., 67: 317-323. King, J. S., R. C. Schwyn and C. A. Fox 1971 The red nucleus in the monkey fMacaca mulatta): A Golgi and an electron microscopic study. J. Comp. Neur., 142: 75-108. Kostyuk, P. G., and G. G. Skibo 1975 An electron microscopic analysis of rubrospinal tract termination in the spinal cord of the cat. Brain Res., 85: 511-516. Kuhlenbeck, H. 1975 The Central Nervous System of Vertebrates. Vol. 4, Spinal Cord and Deuterencephalon. S. Karger, Basel. Kuypers, H. G. J. M., W. R. Fleming and J. W. Farinholt 1962 Subcortical projections in the rhesua monkey. J. Comp. Neur., 118: 107-137. Kuypers, H. G. J. M., and D. G . Lawrence 1967 Cortical projections to the red nucleus and the brain stem in the rhesus monkey. Brain Res., 4: 151-188. LaVail, J. H., K. R. Winston and A. Tish 1973 A method based on retrograde intraaxonal transport of protein for identification of cell bodiea of origin of axons terminating within the CNS. Brain Res., 58: 470-477. Leonard, R. B., and D. H. Cohen 1975 A cytoarchitectonic analysis of the spinal cord of the pigeon fColumba liuia). J. Comp. Neur., 163: 159-180. Martin, G. F., and R. Dom 1970 The rubrospinal tract of the opossum (Didelphis uirginianai. J. Comp. Neur., 138: 19-30. Martin, G. F., R. Dom, S. Katz and J. S. King 1974 Theorganization of projection neurons in the opossum red nucleus. Brain Res., 78: 17-34. Massion, J. 1967 The mammalian red nucleus. Physiol. Rev., 47: 383-436. Miller, R. A,, and N. L. Strominger 1973 Efferent connections of the red nucleus in the brainstem and spinal cord of t h e rhesus monkey. J. a m p . Neur., 152: 327-346.
654
J.
M. WILD, J. B. CABOT, D. H. COHEN AND H. J. KARTEN
Mizuno, N., A. Mcchizuki, R. Akimoto, N. Matsushima and Y. Nakamura 1973 Rubrobulbar projections in the rabbit. A lieht and electron microscoDic studv. J. ComD. Neur., l i 7 : 267-280. Mizuno, N., and Y. Nakamura 1971 Rubral fibers t o the facial nucleus in t h e rabbit. Brain Res., 28: 545-549. Murray, H.M. 1977 Rubrobulbar projections in the tree shrew (Tupaia). Neurosci. Abstr., 3: 59 (Abstract). Murray, H.M., and D. E. Haines 1975 The rubrospinal tract in a prosimian primate (Galago senegalensis]. Brain Behav. and Evol., 12: 311-333. Murray, H. M., D. E. Haines and I. Cote 1976 The rubrospinal tract of the tree shrew (Tupaia glis). Brain Res., 116: 317-322. Nyberg-Hansen, R., and A. Brodal 1964 Sites and modes of termination of rubrospinal fibers in the cat. An experimental study with silver impregnation methods. J. Anat. (London), 98: 235-253. Otabe, J. S.,and A. Horowitz 1970 Morphology and cytoarchitecture of the red nucleus of the domestic pig (Sus scrofa). J. Comp. Neur., 138: 373-389. Ovtscharoff, W., and R. Gossrau 1972 Histochemistry and ultrastructure of the red nucleus of hens (Gallus domesticud. Histochemie, 30: 73-81. Padel, Y., J. Armand and A. M. Smith 1972 Topography of rubrospinal units in thecat. Exp. Brain Res., 14: 363-371. Petras, J. M. 1967 Cortical, tectal and tegmental fiber connections in the spinal cord of the cat. Brain Res., 6: 275-324. Petrovicky, P. 1974 The origin and topographical organ-
ization of the rubrospinal tract in t h e rat. Folia Morphol.,
22: 314-316. Poirier, L. J., and G. Bouvier 1966 The red nucleus and its efferent nervous pathways in the monkey. J. Comp. Neur., 128: 223-244. Pompeiano, O., and A. Brodal 1957 Experimental demonstration of a somatotopical origin of rubrospinal fibers in the cat. J. Comp. Neur., 108: 225-251. Reid, J. M., D. G. Gwyn and B. A. Flumerfelt 1973 A cytoarchitectonic and Golgi study of the red nucleus i n the rat. J. Comp. Neur., 162: 337-362. Shincda, Y.,C. Ghez and A. Arnold 1977 Spinal branching of rubrospinal axons in the cat. Exp. Brain Res., 30: 203-218. Walberg, F. 1956 Descending connections to t h e inferior olive. An experimental study in the cat. J. Comp. Neur., 106: 77-174. Waldron, H.A,, and D. G . Gwyn 1969 Descending nerve tracts in the spinal cord of the rat. I. Fibers from the midbrain. J. Comp. Neur., 137: 143-154. Warner, G., and C. R. R. Watson 1972 The rubrospinal tract in a diprodont marsupial OXchosaurus uulpecula). Brain Res., 41: 180-183. Wild, J. M.,J. B. Cabot, D. H. Cohen and H. J. Karten 1977 The avian rubrospinal tract: Its cells of origin, course and sites of termination. Neurosci. Abstr., 3: 280 (Abstract). Wold, J. E. 1978 The vestibular nuclei in the domestic hen (Gallus domesticus). IV. The projection to the spinal cord. Brain, Behav. and Evol., 15: 41-62. Zecha, A. 1961 Bezit een vogel een fasciculus rubro-bulbo-spinalis? Ned. J. Geneesk., 105: 2373.
