- Email: [email protected]
Locomotor role of the corticoreticular–reticulospinal–spinal interneuronal system
Progress in Brain Research, Vol. 143 ISSN 0079-6123 Copyright ß 2004 Elsevier BV. All rights reserved
CHAPTER 24
Locomotor role of the corticoreticular– reticulospinal–spinal interneuronal system Kiyoji Matsuyama1,*, Futoshi Mori2, Katsumi Nakajima2, Trevor Drew3, Mamoru Aoki1 and Shigemi Mori2
2
1 Department of Physiology, Sapporo Medical University, School of Medicine, Sapporo 060-8556, Japan Department of Biological Control System, National Institute for Physiological Sciences, Okazaki 444-8585, Japan 3 Department of Physiology, University of Montreal, Faculty of Medicine, Montreal, QC H3C 3J7, Canada
Abstract: In vertebrates, the descending reticulospinal pathway is the primary means of conveying locomotor command signals from higher motor centers to spinal interneuronal circuits, the latter including the central pattern generators for locomotion. The pathway is morphologically heterogeneous, being composed of various types of inparallel-descending axons, which terminate with different arborization patterns in the spinal cord. Such morphology suggests that this pathway and its target spinal interneurons comprise varying types of functional subunits, which have a wide variety of functional roles, as dictated by command signals from the higher motor centers. Corticoreticular fibers are one of the major output pathways from the motor cortex to the brainstem. They project widely and diffusely within the pontomedullary reticular formation. Such a diffuse projection pattern seems well suited to combining and integrating the function of the various types of reticulospinal neurons, which are widely scattered throughout the pontomedullary reticular formation. The corticoreticular–reticulospinal–spinal interneuronal connections appear to operate as a cohesive, yet flexible, control system for the elaboration of a wide variety of movements, including those that combine goal-directed locomotion with other motor actions.
Introduction
cerebral cortex to the spinal cord (SC) (Drew et al., Chapter 25 of this volume). The CR axons originate primarily from the premotor (area 6) and primary motor (area 4) regions of the sensorimotor cortex, and descend with the corticospinal (CS) axons through the internal capsule and cerebral peduncle (Brodal, 1981). These CR axons, some of which arise as collaterals of CS axons, terminate bilaterally in the medial pontomedullary reticular formation (PMRF), from which long-descending RS axons originate (Kuypers, 1981). The pontine and medullary RS axons descend ipsilaterally and bilaterally, respectively, throughout the full length of the SC, terminating in the internuncial layer of the ventral
The neural control of locomotion in vertebrates involves continuous interactions between various kinds of neural subsystems which are widely distributed throughout the central nervous system (CNS) (Grillner, 1985; Mori et al., 1992; Grillner and Walle´n, S. Mori et al., Chapters 1 and 33 of this volume). Of these subsystems, the corticoreticular (CR) pathway, the reticulospinal (RS) pathway and the spinal interneuronal (IN) system as a whole form a continuous, anatomical system that links the *Corresponding author: Tel.: þ 81-11-611-2111 (ext. 2669); Fax: þ 81-11-644-1020; E-mail: [email protected]
239 DOI: 10.1016/S0079-6123(03)43024-0
240
horn (VH), where several different types of INs are located (Brodal, 1981; Holstege and Kuypers, 1982; Peterson, 1984). As reviewed by Armstrong (1986), the motor cortex participates in the regulation and modification of the basic locomotor pattern. The majority of motor cortical neurons, including corticofugal neurons such as CS and CR neurons, are modulated in phase with the rhythmic activities of different limb muscles during locomotion in cats walking on a treadmill, and they significantly increase their discharge frequency during voluntary gait modifications (Drew, 1993; Kably and Drew, 1998b). There is also evidence that when the CS axons are transected at the caudal medullary level, stimulation of the pyramids releases locomotion in mesencephalic cats, thus indicating that the corticobulbar axons, including CR axons, must be responsible for this effect (Shik et al., 1968). In addition, stimulation of the motor cortex excites RS neurons mono- and polysynaptically in anesthetized cats (Pilyavsky and Gokin, 1978). Several RS unit-recording studies have shown that many RS neurons exhibit rhythmic discharges during locomotion evoked by stimulation of the mesencephalic locomotor region (MLR) in decerebrate cats (Orlovsky, 1970b), or during fictive locomotion in decerebrate, paralyzed cats (Perreault et al., 1993). In intact cats walking on a treadmill (Drew et al., 1986; Matsuyama and Drew, 2000a), the majority of RS neurons exhibit phasic modulation that is correlated to rhythmic muscle activity and/or rhythmic limb movements. Furthermore, when gait modifications, such as walking on an inclined plane and stepping over an obstacle are required, RS neurons exhibit changes in discharge activity, which coincide with the changes in muscle activity (Matsuyama and Drew, 2000b; Prentice and Drew, 2001). These changes are associated with the postural adjustment associated with the gait modifications. Therefore, the RS system as a whole may play a role in modulating interlimb coordination in each locomotor cycle and in producing coordinated postural responses during locomotion. There is also clear evidence that many lumbar INs located in the intermediate region and VH of the SC show rhythmic discharges during MLR-evoked locomotion and fictive locomotion in decerebrate
cats (Feldman and Orlovsky, 1975; Huang et al., 2000) and fictive scratching in spinal cats (Berkinblit et al., 1978). It has also been demonstrated that stimulation of the ventrolateral funiculus (VLF) at the upper cervical SC, which activates pathways including the RS tract, evokes fictive forelimb stepping and rhythmical discharges of cervical INs in decerebrate cats with the lower thoracic SC transected (Yamaguchi, 1986; Terakado and Yamaguchi, 1990). These results suggest that the RS pathway is, in large part, responsible for the activation of spinal IN circuits that produce sustained stepping movements of the limbs, i.e., the central pattern generator (CPG) for locomotion (Grillner, 1985). In this chapter, we will provide detailed information on: (1) the cortical areal specificity of the CR projection; (2) the morphological heterogeneity of the RS pathway; and (3) the morphology and physiology of lumbar commissural INs (CINs). We will then discuss the possible functional role of the CR–RS–spinal IN system in the control of locomotion.
Cortical areal specificity of the CR projection Since the motor cortical areas are organized in a topographical fashion (Hassler and Muhs-Clement, 1964; Porter and Lemon, 1993), it is natural to think that the whole CR pathway is composed of multiple subcomponents with different cortical origins and functions. The organization of the CR pathway, however, is poorly understood, particularly with respect to the pattern of the projections from different cortical areas. Therefore, to advance understanding of the locomotor role of the CR pathway, we examined interareal differences of its projections within the brainstem (Matsuyama and Drew, 1997). In this study, we injected the anterograde neural tracer, phaseolus vulgaris-leucoagglutinin (PHA-L) into four physiologically identified subdivisions of the cat pericruciate cortex (fore- and hindlimb representations of area 4; areas 6a and 6a ) (Matsuyama and Drew, 1997). Generally, as described previously (Keizer and Kuypers, 1984; Rho et al., 1997), PHA-L labeling from area 6 was more numerous and dense throughout the PMRF than was that from area 4. From the quantitative analyses of the density and
241
distribution of PHA-L labeling in the brainstem, we could specify the characteristic pattern of CR projections from each of the four cortical areas (Matsuyama and Drew, 1997). The number of labeled pyramidal tract (PT) axons that originated from area 6 was much more numerous than that from area 4. The labeled axons from area 4 were of large diameter, however, and they projected caudally to the SC. In contrast, most of those from area 6 were of small diameter, and few descended to the SC. At the caudal pontine and medullary level, PT axons from area 6 decreased in number more rapidly than those from area 4, primarily because numerous fine PT axons from area 6 ran dorsally and directly entered the PMRF. Figure 1A shows that a large proportion of CR axons from area 6 are more likely to be direct CR axons, whereas those from area 4 arise primarily as collaterals from CS axons (see also Kably and Drew, 1998a). Comparison of the distribution of terminal labeling from all four cortical areas further revealed the gross termination pattern of each CR projection in the PMRF. Generally, the number of terminal endings from area 6 distributed at the pontomedullary level was much greater than that from area 4.
In addition, a high percentage of the terminal endings resulting from the area-6 injection at this level were restricted to the PMRF (area 6a , 53%; area 6a , 71%) whereas in the case of the area-4 injections, a much smaller percentage was distributed in this region (forelimb region of area 4, 32%; hindlimb region, 11%; Matsuyama and Drew, 1997). Conversely, there was a very large projection from area 4 to the pontine nuclei and other brainstem structures, and a correspondingly smaller projection from area 6. Within the PMRF, terminal endings from area 6, as well as the forelimb region of area 4, were distributed with an ipsilateral predominance at the rostral pons, but more bilaterally at the caudal pons and medulla. A smaller relative number of terminal endings from the hindlimb region of area 4 were distributed in the PMRF, and they were localized mainly in the medullary RF. These findings indicate that although CR projections from different cortical areas overlap extensively, each projection is organized in a specific fashion depending on the cortical area from which the CR axons originate. In contrast to fastigoreticular terminals which have many close appositions to the somata of reticular neurons (Homma et al., 1995), CR terminals
Fig. 1. Projections of CR axons. (A) Parasagittal view of the projections of two types of CR axon: 1 and 2 directly enter the PMRF, whereas 3 arise as an axon collateral from a CS axon. Vertical broken lines indicated by thick upward arrows (at a and b) correspond to the B levels of brainstem transverse section. (B) Transverse views of the arborization pattern of two other CR axons that originated in cortical areas 6a (a) and 4 (b; forelimb representation). These axons’ projections were reconstructed from serial transverse sections at the level of the caudal pons (a) and rostral medulla (b). The a axon is a direct CR one, and that in b is a collateral from a CS axon indicated by a large dot. Arrows show the axons’ trajectories, and small dots represent the location of their terminal swellings. Abbreviations: FTG, gigantocellular tegmental field; FTM, magnocellular tegmental field; IO, inferior olive; L, left side; PT, pyramidal tract; R, right side; SO, superior olive; TB, trapezoid body. (B panels reproduced, in part, from Matsuyama and Drew, 1997 with permission from Wiley-Liss.)
242
were rarely closely apposed to the somata in the PMRF (Matsuyama and Drew, 1997). This probably indicated that CR terminals contact primarily the dendritic processes of reticular neurons, including RS neurons. Reconstruction of single CR axons further revealed the morphological features of their termination in the PMRF (Matsuyama and Drew, 1997). As shown in Fig. 1A, most CR axons turned dorsally and then distributed terminal axons from the ventral to dorsal parts of the PMRF. Single CR axons usually projected extensively within the PMRF in the transverse plane (Fig. 1B), but with a relatively narrow rostrocaudal extent (see Fig. 1A), and their terminals are far less densely aggregated than those distributed in other brainstem nuclei such as the pontine nuclei (see also Matsuyama and Drew, 1997). Since RS neurons are scattered throughout the whole PMRF (Brodal, 1981; Holstege and Kuypers, 1982), this diffuse projection pattern of each CR axon seems to be well suited to recruit varying types of RS neurons with different functions and locations.
Morphological heterogeneity of RS pathway The medial PMRF from which the major RS pathways originate can be subdivided into four major nuclei according to their cytoarchitecture and location in the brainstem (Brodal, 1981): the nuclei reticularis pontis oralis (NRPo) and caudalis (NRPc) in the pons, and the nuclei reticularis gigantocellularis (NRGc) and magnocellularis (NRMc) in the medulla. Furthermore, according to Berman (1968), these four reticular nuclei are organized into two groups; the gigantocellular (FTG) and magnocellular tegmental fields (FTM). The FTG corresponds to three of the reticular nuclei (NRPo, NRPc, NRGc), while the FTM corresponds to the NRMc. In sharp contrast to such clear anatomical subdivisions, the PMRF is organized in a loose topographical fashion in relation to limb– trunk motor outputs, with a large overlap in the representation, and no clear segregation as is normally seen in the motor cortex (Peterson, 1984; Drew and Rossignol, 1990). Since the contribution of the PMRF to motor control depends highly on the projection pattern of the RS pathways, we have performed a series of
PHA-L tracing studies to characterize the morphology of RS axons originating from each reticular nucleus in the cat (Matsuyama et al., 1988, 1997, 1999a,b). The results of the PHA-L tracing studies demonstrated the following general characteristics of the RS pathways originating from the FTG and FTM. As described previously (Holstege and Kuypers, 1982), PHA-L-labeled RS axons from the pontine FTG (NRPo and NRPc) descend purely ipsilaterally, while those from the medullary FTG (NRGc) and the FTM descend almost evenly bilaterally. Axons that descend contralaterally cross the midline at the medullary level. There is a clear difference in the axonal organization between FTG-RS and FTM-RS pathways. The former RS pathway is composed mainly of thick axons (diameter 3–10 mm at the cervical SC) descending in the ventral (VF) and VLF funiculi, while the latter pathway is composed of numerous thin axons (diameter <3 mm at the cervical SC), which descend in the VF, VLF and the dorsolateral funiculus (DLF; Matsuyama et al., 1988, 1997, 1999a). Furthermore, FTG-RS axons terminate mainly in spinal IN (laminae VII–VIII) and axial MN nuclei at all levels of the SC, while FTMRS axons terminate widely in laminae VI–VIII and axial and limb MN nuclei. We have traced the trajectories of individual FTGRS axons in continuity at multiple segments, and characterized intricate details of the innervation pattern of these axons along the rostrocaudal extent (Matsuyama et al., 1997, 1999b). Generally, as shown in Fig. 2A, FTG-RS axons give off multiple axon collaterals along their descent with an intercollateral interval of 5 mm in both the cervical and lumbar enlargements (2–3 vs. 1–2 collaterals/segment at the cervical vs. lumbar SC). The diameter of the stem FTG-RS axons decrease as they descend caudally (Matsuyama et al., 1997, 1999b). The axon collaterals terminate primarily in lamina VIII and the adjacent lamina VII (Fig. 2B). The innervation pattern of each axon along the rostrocaudal extent is different from one stem axon to another. This indicates that the FTG-RS pathway is morphologically a heterogeneous system composed of various types of axons with different innervation patterns. Physiologically, FTG-RS neurons are also subdivided into several groups according to their
243
Fig. 2. Projections of RS axons. (A) Multisegmental innervation pattern of two single pontine RS axons in the cervical (a) and lumbar (b) enlargements. These axons originated in the left pontine FTG. The arborization patterns of three axon collaterals arising from the single stem axons were reconstructed from serial transverse sections (at C5–7 for a; L5–7 for b). (B) Fine arborization of a single axon collateral at the lumbar SC. This axon collateral is from the RS axon in Ab at the L6 level. Again, arrows indicate the trajectories of the axons and their successive collaterals. Abbreviations: VF, ventral funiculus; VII, VIII and IX, Rexed’s laminae VII, VIII and IX. (Reproduced, in part, from Matsuyama et al., 1999b with permission from Wiley-Liss.)
discharge patterns during locomotion (e.g., EMGrelated, locomotor-rhythm-related, unrelated, tonic, silent; Drew et al., 1986; Prentice and Drew 2001). Therefore, it is highly possible that the FTG-RS pathway is a heterogeneous, parallel-descending pathway composed of various types of neuronal elements with different functions and morphology. This heterogeneity might be related to the functional complexity of FTG-RS pathway. Since various types of RS neurons are intermingled within the FTG region, and since many individual RS axons branch profusely in the SC, stimulation of even a limited region of the FTG evokes a complex and mixed pattern of motor outputs (see also Drew and Rossignol, 1990). Despite such a high degree of heterogeneity, there are three organizing principles related to the collateral termination pattern of individual FTG-RS axons. First, each axon collateral arborizes in a limited rostrocaudal region with a thickness of <1 mm. Second, axon collaterals arising from 80% of
FTG-RS axons innervate only the unilateral gray matter, ipsilateral to their stem axons, whereas those from the remaining axons ( 20%) innervate the bilateral gray matter. Third, the termination areas of axon collaterals from a given FTG-RS axon are similar at each segmental level along the course of that axon, and different from one stem axon to another. Such commonality of the pattern of collateral termination along the rostrocaudal extent of parent axons is found throughout the cervical and lumbar SC. We therefore propose that a single RS neuron innervates a large number of INs at multiple segments throughout the full length of the SC, and of more functional importance, a single RS neuron may activate INs at similar locations at different SC levels. Although preliminary, we have also analyzed the morphological features of FTM-RS axons (Matsuyama et al., 1999a). In contrast to FTG-RS axons, most FTM-RS axons have small diameters (<3 mm), and they give off axon collaterals at a low occurrence frequency ( 1 collateral/segment at the
244
cervical SC level). The terminal axons of FTM-RS axons are widely distributed in laminae VI–VIII and additionally in lamina IX, whereas those of FTG-RS axons are distributed primarily in laminae VIII–VII. This suggests that the RS pathways are morphologically divided into at least two subcomponents: a FTG-RS pathway, composed of thick axons with high collateralization and a FTM-RS pathway, composed of thin axons with low collateralization. Several findings suggesting that both FTG-RS and FTM-RS pathways are highly implicated in the control of locomotion. Recall first that there are several specific areas identified as locomotion-inducing sites in decerebrate and thalamic cats: notably, the MLR in the midbrain (Shik et al., 1966), the subthalamic locomotor region (SLR) in the subthalamus (Orlovsky, 1969) and the cerebellar locomotor region (CLR) in the cerebellum (Mori et al., 1999). The most marked efferent projections from the former two regions are those to the PMRF in the brainstem (Bayev et al., 1988), and a large number of RS neurons receive excitatory inputs from these regions (Orlovsky, 1970a). In particular, the major efferents from the MLR project mainly to the FTM (Steeves and Jordan, 1984), and stimulation of this site activates FTM-RS neurons with a wide (10–100 m/s) range of axonal conduction velocity (Garcia-Rill and Skinner, 1987). In contrast, one of the major efferents from the CLR is the crossed fastigioreticular pathway that terminates in the PMRF (FTG>FTM; Homma et al., 1995). Stimulation of this site monosynaptically activates FTGRS neurons with a relatively fast (50–100 m/s) conduction velocity (Mori et al., 2000). Therefore, we suggest that both FTG-RS and FTM-RS pathways are essential for the generation and regulation of locomotor movements. This suggestion is strongly supported by the finding that lesions of the ventral and ventrolateral SC at low thoracic levels in the cat, which probably severs descending pathways including both RS pathways, cause pronounced locomotor deficits accompanying postural instability. These include an irregularity of the hindlimb’s step-cycle duration and a step-by-step inconsistency of interlimb coupling between homolateral fore- and hindlimbs (Brustein and Rossignol, 1998).
Morphology and physiology of lumbar commissural INs A wide variety of spinal INs involved in motor functions are distributed in the spinal internuncial layer. These can be classified into several groups according to the characteristics of their morphology and physiology (for review, see Scheibel and Scheibel, 1969; Jankowska, 1992). Most of these INs receive peripheral afferent signals mono- or polysynaptically and they project to axial and/or limb MNs, thereby contributing to a wide variety of spinal reflex pathways (Jankowska, 1992). During stepping and scratching, a large number of spinal INs, which are located in the lumbar enlargement, exhibit rhythmic activity related to that of hindlimb MNs (Feldman and Orlovsky, 1975; Berkinblit et al., 1978). The majority of the rhythmically modulated INs are located in the medial to lateral parts of the intermediate gray matter, as well as in the VH, i.e., in laminae VI–VII. Among these are Ia INs, which mediate Ia reciprocal inhibition of antagonists, and burst rhythmically in phase with their parent muscle during real stepping. By inhibiting antagonistic MNs, they may contribute to the generation of alternating activity of extensor and flexor muscles in each limb (Feldman and Orlovsky, 1975). The previous studies have provided important information about several of the characteristics of spinal INs, including their pattern of activity during locomotion, their location and their neural inputs. This information is essential for understanding the locomotor roles of spinal INs. Because the function of each IN is largely related to its structure and termination pattern, however, we think it is also important to further detail the morphology of spinal INs. For example, we have recently examined the fine morphology of RS-activated lumbar INs, and recorded their discharge characteristics during the generation of fictive locomotion (Matsuyama et al., 1999c). For this, we utilized an intra-axonal recording technique combined with an intra-axonal injection of neurobiotin (NB). The microelectrode was positioned in the VF at L4–7 segments in decerebrate, paralyzed cats. This technique allowed us to obtain both morphological and physiological characteristics for each IN.
245
In fully stained INs, the morphology revealed included their dendrites and collateral branching pattern. Their stem axons cross the midline, and their somata are located mainly in lamina VIII, and also in laminae VI–VII. The majority ( 80%) discharge rhythmically during MLR-evoked fictive locomotion. They always exhibit one discharge peak in each locomotor cycle, and their discharge is phase-related to the discharge of left- or right-side hindlimb extensor nerve activity. Such INs are activated orthodromically by stimulation of the FTG (NRPc or NRGc). They are particularly responsive at short latency ( 4 ms) to contralateral NRGc stimulation. Since 40% of NRGc-RS axons cross the midline and descend in the contralateral VF and VLF (Matsuyama et al., 1988), we think that the activity of lamina VIII INs is strongly controlled by descending inputs via the crossed FTG-RS pathway. Furthermore, because a majority of lamina VIII INs which received excitatory RS inputs showed locomotor-related rhythmic activity, we also suggest that these neurons may be subcomponents of the locomotor CPG for each hindlimb. The lamina VIII INs which we identified are divided into three types according to their axonal course. As shown in Fig. 3A, their stem axons cross the midline at the SC level where their own somata are located, and enter the contralateral VF. The axon then, either bifurcates rostrally or caudally, ascends or descends in the VF, giving off multiple axon collaterals ( 5 collaterals/segment) for a few segments. These collaterals terminated primarily in lamina VIII and adjacent VII, and some additionally terminate in hindlimb MN nuclei (lamina IX; see Fig. 3B). Therefore, the lamina VIII INs which we identified can be considered as CINs. This coincides well with the previous findings reported by Scheibel and Scheibel (1969), which emphasized that a high population ( 85%) of spinal INs located in lamina VIII are CINs. The preliminary anterograde neural tracing study using the tracer, biotinylated dextran amine (Matsuyama and Aoki, 2001), has revealed some aspects of the gross projection pattern of lamina VIII CINs. Their axons cross the midline through the anterior commissure forming thick axon bundles, and they course caudally and/or rostrally in the contralateral VF. These axons commonly give off axon
collaterals projecting primarily in laminae VIII–VII and additionally in lamina IX. The density of these axon projections is highest at the SC level where the tracer is injected, and they decrease rapidly with distance from the injection site. These findings suggest that lamina VIII CINs located on the left and right sides are tightly coupled to each other at the same SC level. Such mutual connection of CINs might be well suited for the intrasegmental integration and coordination of the locomotor rhythm generated on both sides of the SC, this being essential for the generation of reciprocal rhythmic movements around similar joints of the left and right limbs (Kjaerulff and Kiehn, 1997). Furthermore, as shown in Fig. 3C, a majority of lamina VIII CINs that we identified receives excitatory inputs via the crossed FTG-RS pathway. Thus, these CINs and the crossed RS pathway may form one of the essential neural systems responsible for the generation and regulation of the reciprocal locomotor rhythm.
Summary and comments It has been well established that the RS pathway is an all-encompassing descending pathway, which mediates locomotor command signals to spinal IN circuits from various higher locomotor centers including the CLR, MLR and SLR (Grillner, 1985; Mori et al., 1992, 2000). The RS system has also been suggested to play an important role in a number of additional motor behaviors, including orienting and postural movements (Peterson, 1984; Drew et al., 1986; Mori et al., 1992). These motor behaviors commonly require the simultaneously coordination of activity in the head and limbs and between different limbs. A large part of this coordinating activity is thought to be ‘hardwired’ and mediated, in part, by the diffuse pattern of projection of individual RS axons (Kuypers, 1981; Peterson, 1984). The findings in the PHA-L study strongly support this idea, because each RS axon has a common propensity to innervate an exceptionally large number of laminae VIII–VII INs throughout the full length of the SC (Matsuyama et al., 1997, 1999b). Therefore, assuming that a spinal IN circuit located at each segment generates a local locomotor activity, we can imagine that the RS system as a whole may provide the means to achieve an intersegmental
246
Fig. 3. Lumbar CINs and their descending control. (A) Axonal morphology of three types of lumbar lamina VIII commissural CINs. Filled circles, solid lines and horizontal short bars indicate the cells’ somata, stem axons and axon collaterals, respectively. Vertical chain and horizontal broken lines indicate the midline of the SC and the border of SC segments, respectively. (B) Innervation pattern of an exemplary lamina VIII CIN. Note the cell’s multiple axon collaterals at, or close to, the SC segment where its soma is located. The axon’s collaterals terminate primarily in laminae VIII and the adjacent VII, with a few also in lamina IX. (C) Summary of the neural connections of the CR pathway, crossed FTG-RS pathway and lamina VIII CINs. CR axons from cortical areas 4 and 6 terminate in the PMRF where the RS pathway, including its crossed FTG-RS component, originates. Crossed FTG-RS axons descend in the contralateral VF throughout the full length of the SC, and their collateral terminations are primarily in laminae VIII and the adjacent VII. The lamina VIII CINs, which receive excitatory inputs from the crossed FTG-RS pathway, innervate primarily the contralateral laminae VIII–VII, with some terminations also in lamina IX. The top drawing is based on a parasagittal section of the right-side cortex. Broken lines indicate boundaries between cytoarchitectonically defined areas, and dotted lines indicate the position of layer V. The three-hatched areas in the top drawing correspond to the PHA-L injection sites in the fore- and hindlimb representation of area 4, and in area 6a (see text for details). Abbreviations: CA, caudate nucleus; 4FL and 4HL, fore- and hindlimb representations of area 4; MCX, motor cortex; 3, area 3.
integration of locomotor activities generated by numerous local spinal IN circuits, which are distributed along the rostrocaudal axis of the SC, and located on both sides of the SC. This arrangement may facilitate an integration and coordination of interlimb and/or limb–trunk locomotor rhythmic activity, something, which is necessary for the elaboration of automatic and synergistic locomotor movements. Moreover, the PHA-L studies have demonstrated that the RS pathway is a heterogeneous, paralleldescending pathway composed of various types of axons with different arborization patterns in the SC
(Matsuyama et al., 1997, 1999b). Such morphological heterogeneity suggests that this pathway and its target spinal INs provide varying types of functional subunits, which have a wide variety of functional roles under various movement command signals from the higher motor centers such as the motor cortex. Furthermore, the CR pathways can also be considered as parallel-descending pathways that are composed of various CR axons arising from different cortical areas with different functions. Thus, the CR– RS–spinal IN system might be considered as a multisynaptic, parallel-descending system, composed
247
of a great variety of neural elements. Of these elements, CR neurons terminating in the medullary FTG, RS neurons with crossed axons and spinal lamina VIII CINs may constitute a functionally unified system responsible for the generation and regulation of reciprocal left side–right side locomotor movements (see Fig. 3C). Finally, what is the primary function of the overall CR–RS–spinal IN system in the control of locomotion? Recall the classical finding that decorticate animals can stand and walk spontaneously but purposelessly, using movement patterns that are stereotypical like those of an automaton (Schaltenbrand and Cobb, 1931). This strongly suggests that the neocortex plays an important role in the goal-directed, diversified locomotion of intact animals. The PHA-L study (Matsuyama and Drew, 1997) showed that each CR neuron projects widely and diffusely within the PMRF. This pattern is well suited to combining and integrating the function of the various types of RS neurons, which are scattered widely throughout the PMRF. Therefore, we suggest that the functionally united CR–RS–spinal IN system serves to provide the optimally flexible neural substrate required for the elaboration and refinement of the wide variety of locomotor patterns used in goal-directed locomotion when both self-induced movements are subject to external perturbations.
Acknowledgments We would like to thank Masahiro Mori and Chijiko Takasu for their excellent technical assistance. This study has been supported by Grants in Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan (11170253, 12680804 and 13035039).
Abbreviations CIN CLR CNS CPG CR CS EMG
commissural interneuron cerebellar locomotor region central nervous system central pattern generator corticoreticular corticospinal electromyogram
FTG FTM IN MLR MN NB NRGc NRMc NRPc NRPo PHA-L PMRF PT RS SC SLR VF VH VLF VI
gigantocellular tegmental field magnocellular tegmental field interneuron mesencephalic locomotor region motoneuron neurobiotinTM nucleus reticularis gigantocellularis nucleus reticularis magnocellularis nucleus reticularis pontis caudalis nucleus reticularis pontis oralis phaseolus vulgaris-leucoagglutinin pontomedullary reticular formation pyramidal tract reticulospinal spinal cord subthalamic locomotor region ventral funiculus ventral horn ventrolateral funiculus VII, VIII, IX, Rexed’s laminae VI, VII, VIII, IX
References Armstrong, D.M. (1986). Supraspinal contributions to the initiation and control of locomotion in the cat. Prog. Neurobiol., 26: 273–361. Bayev, K.V., Beresovskii, V.K., Kebkalo, T.G. and Savoskina, L.A. (1988). Afferent and efferent connections of brainstem locomotor regions: study by means of horseradish peroxidase transport technique. Neuroscience, 26: 871–891. Berkinblit, M.B., Deliagina, T.G., Feldman, A.G., Gelfand, I.M. and Orlovsky, G.N. (1978). Generation of scratching. I. Activity of spinal interneurons during scratching. J. Neurophysiol., 41: 1040–1057. Berman, A.L. (1968). The Brain Stem of the Cat: A Cytoarchitectonic Atlas with Stereotaxic Coordinates. University of Wisconsin Press, Madison. Brodal, A. (1981). Neurological Anatomy in Relation to Clinical Medicine. 3rd edition. Oxford University Press, New York, pp. 180–293. Brustein, E. and Rossignol, S. (1998). Recovery of locomotion after ventral and ventrolateral spinal lesion in the cat. I. Deficits and adaptive mechanisms. J. Neurophysiol., 80: 1245–1267. Drew, T. (1993). Motor cortical activity during voluntary gait modifications in the cat. I. Cells related to the forelimbs. J. Neurophysiol., 70: 179–199.
248 Drew, T. and Rossignol, S. (1990). Functional organization within the medullary reticular formation of intact unanesthetized cat. I. Movements evoked by microstimulation. J. Neurophysiol., 64: 767–781. Drew, T., Dubuc, R. and Rossignol, S. (1986). Discharge pattern of reticulospinal and other reticular neurons in chronic, unrestrained cats walking on a treadmill. J. Neurophysiol., 55: 375–401. Feldman, A.G. and Orlovsky, G.N. (1975). Activity of interneurons mediating reciprocal Ia inhibition during locomotion. Brain Res., 84: 181–194. Garcia-Rill, E. and Skinner, R.D. (1987). The mesencephalic locomotor region. II. Projections to reticulospinal neurons. Brain Res., 411: 13–20. Grillner, S. (1985). Neurobiological bases of rhythmic motor acts in vertebrates. Science, 228: 143–149. Hassler, R. and Muhs-Clement, K. (1964). Architektonischer Aufbau des sensomotorischen und parietalen Cortex der Katze. J. Hirnforsch., 6: 377–420. Holstege, G. and Kuypers, H.G.J.M. (1982). The anatomy of brain stem pathways to the spinal cord in cat. A labeled amino acid tracing study. In: Kuypers H.G.J.M. and Martin G.F. (Eds.), Descending Pathway to the Spinal Cord, Prog. Brain Res., Vol. 57, Elsevier, New York, pp. 145–175. Homma, Y., Nonaka, S., Matsuyama, K. and Mori, S. (1995). Fastigiofugal projection to the brainstem nuclei in the cat: an anterograde PHA-L tracing study. Neurosci. Res., 23: 89–102. Huang, A., Noga, B.R., Carr, P.A., Fedirchuk, B. and Jordan, L.M. (2000). Spinal cholinergic neurons activated during locomotion: localization and electrophysiological characterization. J. Neurophysiol., 83: 3537–3547. Jankowska, E. (1992). Interneuronal relay in spinal pathways from proprioceptors. Prog. Neurobiol., 38: 335–378. Kably, B. and Drew, T. (1998a). Corticoreticular pathways in the cat. I. Projection patterns and collateralization. J. Neurophysiol., 80: 389–405. Kably, B. and Drew, T. (1998b). Corticoreticular pathways in the cat. II. Discharge activity of neurons in area 4 during voluntary gait modifications. J. Neurophysiol., 80: 406–424. Keizer, K. and Kuypers, H.G.J.M. (1984). Distribution of corticospinal neurons with collaterals to lower brainstem reticular formation in cat. Exp. Brain Res., 54: 107–120. Kjaerulff, O. and Kiehn, O. (1997). Crossed rhythmic synaptic input to motoneurons during selective activation of the contralateral spinal locomotor network. J. Neurosci., 17: 9433–9447. Kuypers, H.G.J.M. (1981) Anatomy of descending pathways. In: Brooks, V.B. (Ed.), Handbook of Physiology, Sec. 1, The Nervous System, Vol. II, Motor Control, Part 1. American Physiological Society, Bethesda, pp. 597–666.
Matsuyama, K. and Aoki, M. (2001). Intraspinal projection pattern of lumbar commissural neurons of cats. Neurosci. Res., Suppl. 25: S157. Matsuyama, K. and Drew, T. (1997). Organization of the projections from the pericruciate cortex to the pontomedullary brainstem of the cat: a study using the anterograde tracer Phaseolus vulgaris-leucoagglutinin. J. Comp. Neurol., 389: 617–641. Matsuyama, K. and Drew, T. (2000a). Vestibulospinal and reticulospinal neuronal activity during locomotion in the intact cat: I. Walking on a level surface. J. Neurophysiol., 84: 2237–2256. Matsuyama, K. and Drew, T. (2000b). Vestibulospinal and reticulospinal neuronal activity during locomotion in the intact cat: II. Walking on an inclined plane. J. Neurophysiol., 84: 2257–2276. Matsuyama, K., Ohta, Y. and Mori, S. (1988). Ascending and descending projections of the nucleus reticularis gigantocellularis in the cat demonstrated by the anterograde neural tracer, Phaseolus vulgaris leucoagglutinin (PHA-L). Brain Res., 460: 124–141. Matsuyama, K., Takakusaki, K., Nakajima, K. and Mori, S. (1997). Multisegmental innervation of single pontine reticulospinal axons in the cervico-thoracic region of the cat: anterograde PHA-L tracing study. J. Comp. Neurol., 377: 234–250. Matsuyama, K., Drew, T., Mori, F. and Mori, S. (1999a). The cortico-reticulo-spinal system: organization of the corticoreticular projection and fine architecture of the reticulospinal pathway in the cat. In: Gantchev G.N., Mori S. and Massion J. (Eds.), Motor Control Today and Tomorrow. Academic Publishing House ‘Prof. M. Drinov’, Sofia, pp. 45–56. Matsuyama, K., Mori, F., Kuze, B. and Mori, S. (1999b). Morphology of single pontine reticulospinal axons in the lumbar enlargement of the cat: a study using the anterograde tracer PHA-L. J. Comp. Neurol., 410: 413–430. Matsuyama, K., Mori, F. and Mori, S. (1999c). Fine structure of a reticulospinal-interneuronal system for coordinating locomotor rhythm in cats. In: Gantchev N. and Gantchev G.N. (Eds.), From Basic Motor Control to Functional Recovery: Concepts, Theories and Models Present State and Perspective. Academic Publishing House ‘Prof. M. Drinov’, Sofia, pp. 14–18. Mori, S., Matsuyama, K., Kohyama, J., Kobayashi, Y. and Takakusaki, K. (1992). Neuronal constituents of postural and locomotor control systems and their interactions in cats. Brain Dev., 14 (Suppl.): S109–S120. Mori, S., Matsui, T., Kuze, B., Asanome, M., Nakajima, K. and Matsuyama, K. (1999). Stimulation of a restricted region in the midline cerebellar white matter evokes coordinated quadrupedal locomotion in the decerebrate cat. J. Neurophysiol., 82: 290–300.
249 Mori, S., Matsui, T., Mori, F., Nakajima, K. and Matsuyama, K. (2000). Instigation and control of treadmill locomotion in high decerebrate cats by stimulation of the hook bundle of Russell in the cerebellum. Can. J. Physiol. Pharmacol., 78: 945–957. Orlovsky, G.N. (1969). Spontaneous and induced locomotion of the thalamic cat. Biophysics, 14: 1154–1162. Orlovsky, G.N. (1970a). Connexions of the reticulo-spinal neurons with the locomotor regions of the brainstem. Biophysics, 15: 178–186. Orlovsky, G.N. (1970b). Work of the reticulo-spinal neurons during locomotion. Biophysics, 15: 761–771. Perreault, M.C., Drew, T. and Rossignol, S. (1993). Activity of medullary reticulospinal neurons during fictive locomotion. J. Neurophysiol., 69: 2232–2247. Peterson, B.W. (1984). The reticulospinal system and its role in the control of movement. In: Barnes C.D. (Ed.), Brainstem Control of Spinal Cord Function. No. 6 as Research Topics in Physiology. Academic Press, London, pp. 27–86. Pilyavsky, A.I. and Gokin, A.P. (1978). Investigation of the cortico-reticulo-spinal connections in cats. Neuroscience, 3: 99–103. Porter, R. and Lemon, R. (1993). Corticospinal Function and Voluntary Movement. No. 45 as Monographs of the Physiological Society. Oxford Science Publication, Oxford. Prentice, S.D. and Drew, T. (2001). Contribution of the reticulospinal system to the postural adjustments occurring during voluntary gait modifications. J. Neurophysiol., 85: 679–698.
Rho, M.J., Cabana, T. and Drew, T. (1997). Organization of the projections from the pericruciate cortex to the pontomedullary reticular formation of the cat: a quantitative retrograde tracing study. J. Comp. Neurol., 388: 228–249. Schaltenbrand, G. and Cobb, S. (1931). Clinical and anatomical studies on two cats without neocortex. Brain, 53: 449–488. Scheibel, M.E. and Scheibel, A.B. (1969). A structural analysis of spinal interneurons and Renshaw cells. No. 11 as UCLA Forum in Medical Sciences. In: Brazier M.A.B. (Ed.), The Interneuron. University of California Press, Los Angeles, pp. 159–208. Shik, M.L., Severin, F.V. and Orlovsky, G.N. (1966). Control of walking and running by means of electrical stimulation of the midbrain. Biophysics, 11: 755–765. Shik, M.L., Orlovsky, G.N. and Severin, F.V. (1968). Locomotion of the mesencephalic cat evoked by pyramidal stimulation. Biophysics, 13: 127–135. Steeves, J.D. and Jordan, L.M. (1984). Autoradiographic demonstration of the projections from the mesencephalic locomotor region. Brain Res., 307: 263–276. Terakado, Y. and Yamaguchi, T. (1990). Last-order interneurones controlling activity of elbow flexor motoneurons during forelimb fictive locomotion in the cat. Neurosci. Lett., 111: 292–296. Yamaguchi, T. (1986). Descending pathways eliciting forelimb stepping in the lateral funiculus: experimental studies with stimulation and lesion of the cervical cord in decerebrate cats. Brain Res., 379: 125–136.
CHAPTER 24
Locomotor role of the corticoreticular– reticulospinal–spinal interneuronal system Kiyoji Matsuyama1,*, Futoshi Mori2, Katsumi Nakajima2, Trevor Drew3, Mamoru Aoki1 and Shigemi Mori2
2
1 Department of Physiology, Sapporo Medical University, School of Medicine, Sapporo 060-8556, Japan Department of Biological Control System, National Institute for Physiological Sciences, Okazaki 444-8585, Japan 3 Department of Physiology, University of Montreal, Faculty of Medicine, Montreal, QC H3C 3J7, Canada
Abstract: In vertebrates, the descending reticulospinal pathway is the primary means of conveying locomotor command signals from higher motor centers to spinal interneuronal circuits, the latter including the central pattern generators for locomotion. The pathway is morphologically heterogeneous, being composed of various types of inparallel-descending axons, which terminate with different arborization patterns in the spinal cord. Such morphology suggests that this pathway and its target spinal interneurons comprise varying types of functional subunits, which have a wide variety of functional roles, as dictated by command signals from the higher motor centers. Corticoreticular fibers are one of the major output pathways from the motor cortex to the brainstem. They project widely and diffusely within the pontomedullary reticular formation. Such a diffuse projection pattern seems well suited to combining and integrating the function of the various types of reticulospinal neurons, which are widely scattered throughout the pontomedullary reticular formation. The corticoreticular–reticulospinal–spinal interneuronal connections appear to operate as a cohesive, yet flexible, control system for the elaboration of a wide variety of movements, including those that combine goal-directed locomotion with other motor actions.
Introduction
cerebral cortex to the spinal cord (SC) (Drew et al., Chapter 25 of this volume). The CR axons originate primarily from the premotor (area 6) and primary motor (area 4) regions of the sensorimotor cortex, and descend with the corticospinal (CS) axons through the internal capsule and cerebral peduncle (Brodal, 1981). These CR axons, some of which arise as collaterals of CS axons, terminate bilaterally in the medial pontomedullary reticular formation (PMRF), from which long-descending RS axons originate (Kuypers, 1981). The pontine and medullary RS axons descend ipsilaterally and bilaterally, respectively, throughout the full length of the SC, terminating in the internuncial layer of the ventral
The neural control of locomotion in vertebrates involves continuous interactions between various kinds of neural subsystems which are widely distributed throughout the central nervous system (CNS) (Grillner, 1985; Mori et al., 1992; Grillner and Walle´n, S. Mori et al., Chapters 1 and 33 of this volume). Of these subsystems, the corticoreticular (CR) pathway, the reticulospinal (RS) pathway and the spinal interneuronal (IN) system as a whole form a continuous, anatomical system that links the *Corresponding author: Tel.: þ 81-11-611-2111 (ext. 2669); Fax: þ 81-11-644-1020; E-mail: [email protected]
239 DOI: 10.1016/S0079-6123(03)43024-0
240
horn (VH), where several different types of INs are located (Brodal, 1981; Holstege and Kuypers, 1982; Peterson, 1984). As reviewed by Armstrong (1986), the motor cortex participates in the regulation and modification of the basic locomotor pattern. The majority of motor cortical neurons, including corticofugal neurons such as CS and CR neurons, are modulated in phase with the rhythmic activities of different limb muscles during locomotion in cats walking on a treadmill, and they significantly increase their discharge frequency during voluntary gait modifications (Drew, 1993; Kably and Drew, 1998b). There is also evidence that when the CS axons are transected at the caudal medullary level, stimulation of the pyramids releases locomotion in mesencephalic cats, thus indicating that the corticobulbar axons, including CR axons, must be responsible for this effect (Shik et al., 1968). In addition, stimulation of the motor cortex excites RS neurons mono- and polysynaptically in anesthetized cats (Pilyavsky and Gokin, 1978). Several RS unit-recording studies have shown that many RS neurons exhibit rhythmic discharges during locomotion evoked by stimulation of the mesencephalic locomotor region (MLR) in decerebrate cats (Orlovsky, 1970b), or during fictive locomotion in decerebrate, paralyzed cats (Perreault et al., 1993). In intact cats walking on a treadmill (Drew et al., 1986; Matsuyama and Drew, 2000a), the majority of RS neurons exhibit phasic modulation that is correlated to rhythmic muscle activity and/or rhythmic limb movements. Furthermore, when gait modifications, such as walking on an inclined plane and stepping over an obstacle are required, RS neurons exhibit changes in discharge activity, which coincide with the changes in muscle activity (Matsuyama and Drew, 2000b; Prentice and Drew, 2001). These changes are associated with the postural adjustment associated with the gait modifications. Therefore, the RS system as a whole may play a role in modulating interlimb coordination in each locomotor cycle and in producing coordinated postural responses during locomotion. There is also clear evidence that many lumbar INs located in the intermediate region and VH of the SC show rhythmic discharges during MLR-evoked locomotion and fictive locomotion in decerebrate
cats (Feldman and Orlovsky, 1975; Huang et al., 2000) and fictive scratching in spinal cats (Berkinblit et al., 1978). It has also been demonstrated that stimulation of the ventrolateral funiculus (VLF) at the upper cervical SC, which activates pathways including the RS tract, evokes fictive forelimb stepping and rhythmical discharges of cervical INs in decerebrate cats with the lower thoracic SC transected (Yamaguchi, 1986; Terakado and Yamaguchi, 1990). These results suggest that the RS pathway is, in large part, responsible for the activation of spinal IN circuits that produce sustained stepping movements of the limbs, i.e., the central pattern generator (CPG) for locomotion (Grillner, 1985). In this chapter, we will provide detailed information on: (1) the cortical areal specificity of the CR projection; (2) the morphological heterogeneity of the RS pathway; and (3) the morphology and physiology of lumbar commissural INs (CINs). We will then discuss the possible functional role of the CR–RS–spinal IN system in the control of locomotion.
Cortical areal specificity of the CR projection Since the motor cortical areas are organized in a topographical fashion (Hassler and Muhs-Clement, 1964; Porter and Lemon, 1993), it is natural to think that the whole CR pathway is composed of multiple subcomponents with different cortical origins and functions. The organization of the CR pathway, however, is poorly understood, particularly with respect to the pattern of the projections from different cortical areas. Therefore, to advance understanding of the locomotor role of the CR pathway, we examined interareal differences of its projections within the brainstem (Matsuyama and Drew, 1997). In this study, we injected the anterograde neural tracer, phaseolus vulgaris-leucoagglutinin (PHA-L) into four physiologically identified subdivisions of the cat pericruciate cortex (fore- and hindlimb representations of area 4; areas 6a and 6a ) (Matsuyama and Drew, 1997). Generally, as described previously (Keizer and Kuypers, 1984; Rho et al., 1997), PHA-L labeling from area 6 was more numerous and dense throughout the PMRF than was that from area 4. From the quantitative analyses of the density and
241
distribution of PHA-L labeling in the brainstem, we could specify the characteristic pattern of CR projections from each of the four cortical areas (Matsuyama and Drew, 1997). The number of labeled pyramidal tract (PT) axons that originated from area 6 was much more numerous than that from area 4. The labeled axons from area 4 were of large diameter, however, and they projected caudally to the SC. In contrast, most of those from area 6 were of small diameter, and few descended to the SC. At the caudal pontine and medullary level, PT axons from area 6 decreased in number more rapidly than those from area 4, primarily because numerous fine PT axons from area 6 ran dorsally and directly entered the PMRF. Figure 1A shows that a large proportion of CR axons from area 6 are more likely to be direct CR axons, whereas those from area 4 arise primarily as collaterals from CS axons (see also Kably and Drew, 1998a). Comparison of the distribution of terminal labeling from all four cortical areas further revealed the gross termination pattern of each CR projection in the PMRF. Generally, the number of terminal endings from area 6 distributed at the pontomedullary level was much greater than that from area 4.
In addition, a high percentage of the terminal endings resulting from the area-6 injection at this level were restricted to the PMRF (area 6a , 53%; area 6a , 71%) whereas in the case of the area-4 injections, a much smaller percentage was distributed in this region (forelimb region of area 4, 32%; hindlimb region, 11%; Matsuyama and Drew, 1997). Conversely, there was a very large projection from area 4 to the pontine nuclei and other brainstem structures, and a correspondingly smaller projection from area 6. Within the PMRF, terminal endings from area 6, as well as the forelimb region of area 4, were distributed with an ipsilateral predominance at the rostral pons, but more bilaterally at the caudal pons and medulla. A smaller relative number of terminal endings from the hindlimb region of area 4 were distributed in the PMRF, and they were localized mainly in the medullary RF. These findings indicate that although CR projections from different cortical areas overlap extensively, each projection is organized in a specific fashion depending on the cortical area from which the CR axons originate. In contrast to fastigoreticular terminals which have many close appositions to the somata of reticular neurons (Homma et al., 1995), CR terminals
Fig. 1. Projections of CR axons. (A) Parasagittal view of the projections of two types of CR axon: 1 and 2 directly enter the PMRF, whereas 3 arise as an axon collateral from a CS axon. Vertical broken lines indicated by thick upward arrows (at a and b) correspond to the B levels of brainstem transverse section. (B) Transverse views of the arborization pattern of two other CR axons that originated in cortical areas 6a (a) and 4 (b; forelimb representation). These axons’ projections were reconstructed from serial transverse sections at the level of the caudal pons (a) and rostral medulla (b). The a axon is a direct CR one, and that in b is a collateral from a CS axon indicated by a large dot. Arrows show the axons’ trajectories, and small dots represent the location of their terminal swellings. Abbreviations: FTG, gigantocellular tegmental field; FTM, magnocellular tegmental field; IO, inferior olive; L, left side; PT, pyramidal tract; R, right side; SO, superior olive; TB, trapezoid body. (B panels reproduced, in part, from Matsuyama and Drew, 1997 with permission from Wiley-Liss.)
242
were rarely closely apposed to the somata in the PMRF (Matsuyama and Drew, 1997). This probably indicated that CR terminals contact primarily the dendritic processes of reticular neurons, including RS neurons. Reconstruction of single CR axons further revealed the morphological features of their termination in the PMRF (Matsuyama and Drew, 1997). As shown in Fig. 1A, most CR axons turned dorsally and then distributed terminal axons from the ventral to dorsal parts of the PMRF. Single CR axons usually projected extensively within the PMRF in the transverse plane (Fig. 1B), but with a relatively narrow rostrocaudal extent (see Fig. 1A), and their terminals are far less densely aggregated than those distributed in other brainstem nuclei such as the pontine nuclei (see also Matsuyama and Drew, 1997). Since RS neurons are scattered throughout the whole PMRF (Brodal, 1981; Holstege and Kuypers, 1982), this diffuse projection pattern of each CR axon seems to be well suited to recruit varying types of RS neurons with different functions and locations.
Morphological heterogeneity of RS pathway The medial PMRF from which the major RS pathways originate can be subdivided into four major nuclei according to their cytoarchitecture and location in the brainstem (Brodal, 1981): the nuclei reticularis pontis oralis (NRPo) and caudalis (NRPc) in the pons, and the nuclei reticularis gigantocellularis (NRGc) and magnocellularis (NRMc) in the medulla. Furthermore, according to Berman (1968), these four reticular nuclei are organized into two groups; the gigantocellular (FTG) and magnocellular tegmental fields (FTM). The FTG corresponds to three of the reticular nuclei (NRPo, NRPc, NRGc), while the FTM corresponds to the NRMc. In sharp contrast to such clear anatomical subdivisions, the PMRF is organized in a loose topographical fashion in relation to limb– trunk motor outputs, with a large overlap in the representation, and no clear segregation as is normally seen in the motor cortex (Peterson, 1984; Drew and Rossignol, 1990). Since the contribution of the PMRF to motor control depends highly on the projection pattern of the RS pathways, we have performed a series of
PHA-L tracing studies to characterize the morphology of RS axons originating from each reticular nucleus in the cat (Matsuyama et al., 1988, 1997, 1999a,b). The results of the PHA-L tracing studies demonstrated the following general characteristics of the RS pathways originating from the FTG and FTM. As described previously (Holstege and Kuypers, 1982), PHA-L-labeled RS axons from the pontine FTG (NRPo and NRPc) descend purely ipsilaterally, while those from the medullary FTG (NRGc) and the FTM descend almost evenly bilaterally. Axons that descend contralaterally cross the midline at the medullary level. There is a clear difference in the axonal organization between FTG-RS and FTM-RS pathways. The former RS pathway is composed mainly of thick axons (diameter 3–10 mm at the cervical SC) descending in the ventral (VF) and VLF funiculi, while the latter pathway is composed of numerous thin axons (diameter <3 mm at the cervical SC), which descend in the VF, VLF and the dorsolateral funiculus (DLF; Matsuyama et al., 1988, 1997, 1999a). Furthermore, FTG-RS axons terminate mainly in spinal IN (laminae VII–VIII) and axial MN nuclei at all levels of the SC, while FTMRS axons terminate widely in laminae VI–VIII and axial and limb MN nuclei. We have traced the trajectories of individual FTGRS axons in continuity at multiple segments, and characterized intricate details of the innervation pattern of these axons along the rostrocaudal extent (Matsuyama et al., 1997, 1999b). Generally, as shown in Fig. 2A, FTG-RS axons give off multiple axon collaterals along their descent with an intercollateral interval of 5 mm in both the cervical and lumbar enlargements (2–3 vs. 1–2 collaterals/segment at the cervical vs. lumbar SC). The diameter of the stem FTG-RS axons decrease as they descend caudally (Matsuyama et al., 1997, 1999b). The axon collaterals terminate primarily in lamina VIII and the adjacent lamina VII (Fig. 2B). The innervation pattern of each axon along the rostrocaudal extent is different from one stem axon to another. This indicates that the FTG-RS pathway is morphologically a heterogeneous system composed of various types of axons with different innervation patterns. Physiologically, FTG-RS neurons are also subdivided into several groups according to their
243
Fig. 2. Projections of RS axons. (A) Multisegmental innervation pattern of two single pontine RS axons in the cervical (a) and lumbar (b) enlargements. These axons originated in the left pontine FTG. The arborization patterns of three axon collaterals arising from the single stem axons were reconstructed from serial transverse sections (at C5–7 for a; L5–7 for b). (B) Fine arborization of a single axon collateral at the lumbar SC. This axon collateral is from the RS axon in Ab at the L6 level. Again, arrows indicate the trajectories of the axons and their successive collaterals. Abbreviations: VF, ventral funiculus; VII, VIII and IX, Rexed’s laminae VII, VIII and IX. (Reproduced, in part, from Matsuyama et al., 1999b with permission from Wiley-Liss.)
discharge patterns during locomotion (e.g., EMGrelated, locomotor-rhythm-related, unrelated, tonic, silent; Drew et al., 1986; Prentice and Drew 2001). Therefore, it is highly possible that the FTG-RS pathway is a heterogeneous, parallel-descending pathway composed of various types of neuronal elements with different functions and morphology. This heterogeneity might be related to the functional complexity of FTG-RS pathway. Since various types of RS neurons are intermingled within the FTG region, and since many individual RS axons branch profusely in the SC, stimulation of even a limited region of the FTG evokes a complex and mixed pattern of motor outputs (see also Drew and Rossignol, 1990). Despite such a high degree of heterogeneity, there are three organizing principles related to the collateral termination pattern of individual FTG-RS axons. First, each axon collateral arborizes in a limited rostrocaudal region with a thickness of <1 mm. Second, axon collaterals arising from 80% of
FTG-RS axons innervate only the unilateral gray matter, ipsilateral to their stem axons, whereas those from the remaining axons ( 20%) innervate the bilateral gray matter. Third, the termination areas of axon collaterals from a given FTG-RS axon are similar at each segmental level along the course of that axon, and different from one stem axon to another. Such commonality of the pattern of collateral termination along the rostrocaudal extent of parent axons is found throughout the cervical and lumbar SC. We therefore propose that a single RS neuron innervates a large number of INs at multiple segments throughout the full length of the SC, and of more functional importance, a single RS neuron may activate INs at similar locations at different SC levels. Although preliminary, we have also analyzed the morphological features of FTM-RS axons (Matsuyama et al., 1999a). In contrast to FTG-RS axons, most FTM-RS axons have small diameters (<3 mm), and they give off axon collaterals at a low occurrence frequency ( 1 collateral/segment at the
244
cervical SC level). The terminal axons of FTM-RS axons are widely distributed in laminae VI–VIII and additionally in lamina IX, whereas those of FTG-RS axons are distributed primarily in laminae VIII–VII. This suggests that the RS pathways are morphologically divided into at least two subcomponents: a FTG-RS pathway, composed of thick axons with high collateralization and a FTM-RS pathway, composed of thin axons with low collateralization. Several findings suggesting that both FTG-RS and FTM-RS pathways are highly implicated in the control of locomotion. Recall first that there are several specific areas identified as locomotion-inducing sites in decerebrate and thalamic cats: notably, the MLR in the midbrain (Shik et al., 1966), the subthalamic locomotor region (SLR) in the subthalamus (Orlovsky, 1969) and the cerebellar locomotor region (CLR) in the cerebellum (Mori et al., 1999). The most marked efferent projections from the former two regions are those to the PMRF in the brainstem (Bayev et al., 1988), and a large number of RS neurons receive excitatory inputs from these regions (Orlovsky, 1970a). In particular, the major efferents from the MLR project mainly to the FTM (Steeves and Jordan, 1984), and stimulation of this site activates FTM-RS neurons with a wide (10–100 m/s) range of axonal conduction velocity (Garcia-Rill and Skinner, 1987). In contrast, one of the major efferents from the CLR is the crossed fastigioreticular pathway that terminates in the PMRF (FTG>FTM; Homma et al., 1995). Stimulation of this site monosynaptically activates FTGRS neurons with a relatively fast (50–100 m/s) conduction velocity (Mori et al., 2000). Therefore, we suggest that both FTG-RS and FTM-RS pathways are essential for the generation and regulation of locomotor movements. This suggestion is strongly supported by the finding that lesions of the ventral and ventrolateral SC at low thoracic levels in the cat, which probably severs descending pathways including both RS pathways, cause pronounced locomotor deficits accompanying postural instability. These include an irregularity of the hindlimb’s step-cycle duration and a step-by-step inconsistency of interlimb coupling between homolateral fore- and hindlimbs (Brustein and Rossignol, 1998).
Morphology and physiology of lumbar commissural INs A wide variety of spinal INs involved in motor functions are distributed in the spinal internuncial layer. These can be classified into several groups according to the characteristics of their morphology and physiology (for review, see Scheibel and Scheibel, 1969; Jankowska, 1992). Most of these INs receive peripheral afferent signals mono- or polysynaptically and they project to axial and/or limb MNs, thereby contributing to a wide variety of spinal reflex pathways (Jankowska, 1992). During stepping and scratching, a large number of spinal INs, which are located in the lumbar enlargement, exhibit rhythmic activity related to that of hindlimb MNs (Feldman and Orlovsky, 1975; Berkinblit et al., 1978). The majority of the rhythmically modulated INs are located in the medial to lateral parts of the intermediate gray matter, as well as in the VH, i.e., in laminae VI–VII. Among these are Ia INs, which mediate Ia reciprocal inhibition of antagonists, and burst rhythmically in phase with their parent muscle during real stepping. By inhibiting antagonistic MNs, they may contribute to the generation of alternating activity of extensor and flexor muscles in each limb (Feldman and Orlovsky, 1975). The previous studies have provided important information about several of the characteristics of spinal INs, including their pattern of activity during locomotion, their location and their neural inputs. This information is essential for understanding the locomotor roles of spinal INs. Because the function of each IN is largely related to its structure and termination pattern, however, we think it is also important to further detail the morphology of spinal INs. For example, we have recently examined the fine morphology of RS-activated lumbar INs, and recorded their discharge characteristics during the generation of fictive locomotion (Matsuyama et al., 1999c). For this, we utilized an intra-axonal recording technique combined with an intra-axonal injection of neurobiotin (NB). The microelectrode was positioned in the VF at L4–7 segments in decerebrate, paralyzed cats. This technique allowed us to obtain both morphological and physiological characteristics for each IN.
245
In fully stained INs, the morphology revealed included their dendrites and collateral branching pattern. Their stem axons cross the midline, and their somata are located mainly in lamina VIII, and also in laminae VI–VII. The majority ( 80%) discharge rhythmically during MLR-evoked fictive locomotion. They always exhibit one discharge peak in each locomotor cycle, and their discharge is phase-related to the discharge of left- or right-side hindlimb extensor nerve activity. Such INs are activated orthodromically by stimulation of the FTG (NRPc or NRGc). They are particularly responsive at short latency ( 4 ms) to contralateral NRGc stimulation. Since 40% of NRGc-RS axons cross the midline and descend in the contralateral VF and VLF (Matsuyama et al., 1988), we think that the activity of lamina VIII INs is strongly controlled by descending inputs via the crossed FTG-RS pathway. Furthermore, because a majority of lamina VIII INs which received excitatory RS inputs showed locomotor-related rhythmic activity, we also suggest that these neurons may be subcomponents of the locomotor CPG for each hindlimb. The lamina VIII INs which we identified are divided into three types according to their axonal course. As shown in Fig. 3A, their stem axons cross the midline at the SC level where their own somata are located, and enter the contralateral VF. The axon then, either bifurcates rostrally or caudally, ascends or descends in the VF, giving off multiple axon collaterals ( 5 collaterals/segment) for a few segments. These collaterals terminated primarily in lamina VIII and adjacent VII, and some additionally terminate in hindlimb MN nuclei (lamina IX; see Fig. 3B). Therefore, the lamina VIII INs which we identified can be considered as CINs. This coincides well with the previous findings reported by Scheibel and Scheibel (1969), which emphasized that a high population ( 85%) of spinal INs located in lamina VIII are CINs. The preliminary anterograde neural tracing study using the tracer, biotinylated dextran amine (Matsuyama and Aoki, 2001), has revealed some aspects of the gross projection pattern of lamina VIII CINs. Their axons cross the midline through the anterior commissure forming thick axon bundles, and they course caudally and/or rostrally in the contralateral VF. These axons commonly give off axon
collaterals projecting primarily in laminae VIII–VII and additionally in lamina IX. The density of these axon projections is highest at the SC level where the tracer is injected, and they decrease rapidly with distance from the injection site. These findings suggest that lamina VIII CINs located on the left and right sides are tightly coupled to each other at the same SC level. Such mutual connection of CINs might be well suited for the intrasegmental integration and coordination of the locomotor rhythm generated on both sides of the SC, this being essential for the generation of reciprocal rhythmic movements around similar joints of the left and right limbs (Kjaerulff and Kiehn, 1997). Furthermore, as shown in Fig. 3C, a majority of lamina VIII CINs that we identified receives excitatory inputs via the crossed FTG-RS pathway. Thus, these CINs and the crossed RS pathway may form one of the essential neural systems responsible for the generation and regulation of the reciprocal locomotor rhythm.
Summary and comments It has been well established that the RS pathway is an all-encompassing descending pathway, which mediates locomotor command signals to spinal IN circuits from various higher locomotor centers including the CLR, MLR and SLR (Grillner, 1985; Mori et al., 1992, 2000). The RS system has also been suggested to play an important role in a number of additional motor behaviors, including orienting and postural movements (Peterson, 1984; Drew et al., 1986; Mori et al., 1992). These motor behaviors commonly require the simultaneously coordination of activity in the head and limbs and between different limbs. A large part of this coordinating activity is thought to be ‘hardwired’ and mediated, in part, by the diffuse pattern of projection of individual RS axons (Kuypers, 1981; Peterson, 1984). The findings in the PHA-L study strongly support this idea, because each RS axon has a common propensity to innervate an exceptionally large number of laminae VIII–VII INs throughout the full length of the SC (Matsuyama et al., 1997, 1999b). Therefore, assuming that a spinal IN circuit located at each segment generates a local locomotor activity, we can imagine that the RS system as a whole may provide the means to achieve an intersegmental
246
Fig. 3. Lumbar CINs and their descending control. (A) Axonal morphology of three types of lumbar lamina VIII commissural CINs. Filled circles, solid lines and horizontal short bars indicate the cells’ somata, stem axons and axon collaterals, respectively. Vertical chain and horizontal broken lines indicate the midline of the SC and the border of SC segments, respectively. (B) Innervation pattern of an exemplary lamina VIII CIN. Note the cell’s multiple axon collaterals at, or close to, the SC segment where its soma is located. The axon’s collaterals terminate primarily in laminae VIII and the adjacent VII, with a few also in lamina IX. (C) Summary of the neural connections of the CR pathway, crossed FTG-RS pathway and lamina VIII CINs. CR axons from cortical areas 4 and 6 terminate in the PMRF where the RS pathway, including its crossed FTG-RS component, originates. Crossed FTG-RS axons descend in the contralateral VF throughout the full length of the SC, and their collateral terminations are primarily in laminae VIII and the adjacent VII. The lamina VIII CINs, which receive excitatory inputs from the crossed FTG-RS pathway, innervate primarily the contralateral laminae VIII–VII, with some terminations also in lamina IX. The top drawing is based on a parasagittal section of the right-side cortex. Broken lines indicate boundaries between cytoarchitectonically defined areas, and dotted lines indicate the position of layer V. The three-hatched areas in the top drawing correspond to the PHA-L injection sites in the fore- and hindlimb representation of area 4, and in area 6a (see text for details). Abbreviations: CA, caudate nucleus; 4FL and 4HL, fore- and hindlimb representations of area 4; MCX, motor cortex; 3, area 3.
integration of locomotor activities generated by numerous local spinal IN circuits, which are distributed along the rostrocaudal axis of the SC, and located on both sides of the SC. This arrangement may facilitate an integration and coordination of interlimb and/or limb–trunk locomotor rhythmic activity, something, which is necessary for the elaboration of automatic and synergistic locomotor movements. Moreover, the PHA-L studies have demonstrated that the RS pathway is a heterogeneous, paralleldescending pathway composed of various types of axons with different arborization patterns in the SC
(Matsuyama et al., 1997, 1999b). Such morphological heterogeneity suggests that this pathway and its target spinal INs provide varying types of functional subunits, which have a wide variety of functional roles under various movement command signals from the higher motor centers such as the motor cortex. Furthermore, the CR pathways can also be considered as parallel-descending pathways that are composed of various CR axons arising from different cortical areas with different functions. Thus, the CR– RS–spinal IN system might be considered as a multisynaptic, parallel-descending system, composed
247
of a great variety of neural elements. Of these elements, CR neurons terminating in the medullary FTG, RS neurons with crossed axons and spinal lamina VIII CINs may constitute a functionally unified system responsible for the generation and regulation of reciprocal left side–right side locomotor movements (see Fig. 3C). Finally, what is the primary function of the overall CR–RS–spinal IN system in the control of locomotion? Recall the classical finding that decorticate animals can stand and walk spontaneously but purposelessly, using movement patterns that are stereotypical like those of an automaton (Schaltenbrand and Cobb, 1931). This strongly suggests that the neocortex plays an important role in the goal-directed, diversified locomotion of intact animals. The PHA-L study (Matsuyama and Drew, 1997) showed that each CR neuron projects widely and diffusely within the PMRF. This pattern is well suited to combining and integrating the function of the various types of RS neurons, which are scattered widely throughout the PMRF. Therefore, we suggest that the functionally united CR–RS–spinal IN system serves to provide the optimally flexible neural substrate required for the elaboration and refinement of the wide variety of locomotor patterns used in goal-directed locomotion when both self-induced movements are subject to external perturbations.
Acknowledgments We would like to thank Masahiro Mori and Chijiko Takasu for their excellent technical assistance. This study has been supported by Grants in Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan (11170253, 12680804 and 13035039).
Abbreviations CIN CLR CNS CPG CR CS EMG
commissural interneuron cerebellar locomotor region central nervous system central pattern generator corticoreticular corticospinal electromyogram
FTG FTM IN MLR MN NB NRGc NRMc NRPc NRPo PHA-L PMRF PT RS SC SLR VF VH VLF VI
gigantocellular tegmental field magnocellular tegmental field interneuron mesencephalic locomotor region motoneuron neurobiotinTM nucleus reticularis gigantocellularis nucleus reticularis magnocellularis nucleus reticularis pontis caudalis nucleus reticularis pontis oralis phaseolus vulgaris-leucoagglutinin pontomedullary reticular formation pyramidal tract reticulospinal spinal cord subthalamic locomotor region ventral funiculus ventral horn ventrolateral funiculus VII, VIII, IX, Rexed’s laminae VI, VII, VIII, IX
References Armstrong, D.M. (1986). Supraspinal contributions to the initiation and control of locomotion in the cat. Prog. Neurobiol., 26: 273–361. Bayev, K.V., Beresovskii, V.K., Kebkalo, T.G. and Savoskina, L.A. (1988). Afferent and efferent connections of brainstem locomotor regions: study by means of horseradish peroxidase transport technique. Neuroscience, 26: 871–891. Berkinblit, M.B., Deliagina, T.G., Feldman, A.G., Gelfand, I.M. and Orlovsky, G.N. (1978). Generation of scratching. I. Activity of spinal interneurons during scratching. J. Neurophysiol., 41: 1040–1057. Berman, A.L. (1968). The Brain Stem of the Cat: A Cytoarchitectonic Atlas with Stereotaxic Coordinates. University of Wisconsin Press, Madison. Brodal, A. (1981). Neurological Anatomy in Relation to Clinical Medicine. 3rd edition. Oxford University Press, New York, pp. 180–293. Brustein, E. and Rossignol, S. (1998). Recovery of locomotion after ventral and ventrolateral spinal lesion in the cat. I. Deficits and adaptive mechanisms. J. Neurophysiol., 80: 1245–1267. Drew, T. (1993). Motor cortical activity during voluntary gait modifications in the cat. I. Cells related to the forelimbs. J. Neurophysiol., 70: 179–199.
248 Drew, T. and Rossignol, S. (1990). Functional organization within the medullary reticular formation of intact unanesthetized cat. I. Movements evoked by microstimulation. J. Neurophysiol., 64: 767–781. Drew, T., Dubuc, R. and Rossignol, S. (1986). Discharge pattern of reticulospinal and other reticular neurons in chronic, unrestrained cats walking on a treadmill. J. Neurophysiol., 55: 375–401. Feldman, A.G. and Orlovsky, G.N. (1975). Activity of interneurons mediating reciprocal Ia inhibition during locomotion. Brain Res., 84: 181–194. Garcia-Rill, E. and Skinner, R.D. (1987). The mesencephalic locomotor region. II. Projections to reticulospinal neurons. Brain Res., 411: 13–20. Grillner, S. (1985). Neurobiological bases of rhythmic motor acts in vertebrates. Science, 228: 143–149. Hassler, R. and Muhs-Clement, K. (1964). Architektonischer Aufbau des sensomotorischen und parietalen Cortex der Katze. J. Hirnforsch., 6: 377–420. Holstege, G. and Kuypers, H.G.J.M. (1982). The anatomy of brain stem pathways to the spinal cord in cat. A labeled amino acid tracing study. In: Kuypers H.G.J.M. and Martin G.F. (Eds.), Descending Pathway to the Spinal Cord, Prog. Brain Res., Vol. 57, Elsevier, New York, pp. 145–175. Homma, Y., Nonaka, S., Matsuyama, K. and Mori, S. (1995). Fastigiofugal projection to the brainstem nuclei in the cat: an anterograde PHA-L tracing study. Neurosci. Res., 23: 89–102. Huang, A., Noga, B.R., Carr, P.A., Fedirchuk, B. and Jordan, L.M. (2000). Spinal cholinergic neurons activated during locomotion: localization and electrophysiological characterization. J. Neurophysiol., 83: 3537–3547. Jankowska, E. (1992). Interneuronal relay in spinal pathways from proprioceptors. Prog. Neurobiol., 38: 335–378. Kably, B. and Drew, T. (1998a). Corticoreticular pathways in the cat. I. Projection patterns and collateralization. J. Neurophysiol., 80: 389–405. Kably, B. and Drew, T. (1998b). Corticoreticular pathways in the cat. II. Discharge activity of neurons in area 4 during voluntary gait modifications. J. Neurophysiol., 80: 406–424. Keizer, K. and Kuypers, H.G.J.M. (1984). Distribution of corticospinal neurons with collaterals to lower brainstem reticular formation in cat. Exp. Brain Res., 54: 107–120. Kjaerulff, O. and Kiehn, O. (1997). Crossed rhythmic synaptic input to motoneurons during selective activation of the contralateral spinal locomotor network. J. Neurosci., 17: 9433–9447. Kuypers, H.G.J.M. (1981) Anatomy of descending pathways. In: Brooks, V.B. (Ed.), Handbook of Physiology, Sec. 1, The Nervous System, Vol. II, Motor Control, Part 1. American Physiological Society, Bethesda, pp. 597–666.
Matsuyama, K. and Aoki, M. (2001). Intraspinal projection pattern of lumbar commissural neurons of cats. Neurosci. Res., Suppl. 25: S157. Matsuyama, K. and Drew, T. (1997). Organization of the projections from the pericruciate cortex to the pontomedullary brainstem of the cat: a study using the anterograde tracer Phaseolus vulgaris-leucoagglutinin. J. Comp. Neurol., 389: 617–641. Matsuyama, K. and Drew, T. (2000a). Vestibulospinal and reticulospinal neuronal activity during locomotion in the intact cat: I. Walking on a level surface. J. Neurophysiol., 84: 2237–2256. Matsuyama, K. and Drew, T. (2000b). Vestibulospinal and reticulospinal neuronal activity during locomotion in the intact cat: II. Walking on an inclined plane. J. Neurophysiol., 84: 2257–2276. Matsuyama, K., Ohta, Y. and Mori, S. (1988). Ascending and descending projections of the nucleus reticularis gigantocellularis in the cat demonstrated by the anterograde neural tracer, Phaseolus vulgaris leucoagglutinin (PHA-L). Brain Res., 460: 124–141. Matsuyama, K., Takakusaki, K., Nakajima, K. and Mori, S. (1997). Multisegmental innervation of single pontine reticulospinal axons in the cervico-thoracic region of the cat: anterograde PHA-L tracing study. J. Comp. Neurol., 377: 234–250. Matsuyama, K., Drew, T., Mori, F. and Mori, S. (1999a). The cortico-reticulo-spinal system: organization of the corticoreticular projection and fine architecture of the reticulospinal pathway in the cat. In: Gantchev G.N., Mori S. and Massion J. (Eds.), Motor Control Today and Tomorrow. Academic Publishing House ‘Prof. M. Drinov’, Sofia, pp. 45–56. Matsuyama, K., Mori, F., Kuze, B. and Mori, S. (1999b). Morphology of single pontine reticulospinal axons in the lumbar enlargement of the cat: a study using the anterograde tracer PHA-L. J. Comp. Neurol., 410: 413–430. Matsuyama, K., Mori, F. and Mori, S. (1999c). Fine structure of a reticulospinal-interneuronal system for coordinating locomotor rhythm in cats. In: Gantchev N. and Gantchev G.N. (Eds.), From Basic Motor Control to Functional Recovery: Concepts, Theories and Models Present State and Perspective. Academic Publishing House ‘Prof. M. Drinov’, Sofia, pp. 14–18. Mori, S., Matsuyama, K., Kohyama, J., Kobayashi, Y. and Takakusaki, K. (1992). Neuronal constituents of postural and locomotor control systems and their interactions in cats. Brain Dev., 14 (Suppl.): S109–S120. Mori, S., Matsui, T., Kuze, B., Asanome, M., Nakajima, K. and Matsuyama, K. (1999). Stimulation of a restricted region in the midline cerebellar white matter evokes coordinated quadrupedal locomotion in the decerebrate cat. J. Neurophysiol., 82: 290–300.
249 Mori, S., Matsui, T., Mori, F., Nakajima, K. and Matsuyama, K. (2000). Instigation and control of treadmill locomotion in high decerebrate cats by stimulation of the hook bundle of Russell in the cerebellum. Can. J. Physiol. Pharmacol., 78: 945–957. Orlovsky, G.N. (1969). Spontaneous and induced locomotion of the thalamic cat. Biophysics, 14: 1154–1162. Orlovsky, G.N. (1970a). Connexions of the reticulo-spinal neurons with the locomotor regions of the brainstem. Biophysics, 15: 178–186. Orlovsky, G.N. (1970b). Work of the reticulo-spinal neurons during locomotion. Biophysics, 15: 761–771. Perreault, M.C., Drew, T. and Rossignol, S. (1993). Activity of medullary reticulospinal neurons during fictive locomotion. J. Neurophysiol., 69: 2232–2247. Peterson, B.W. (1984). The reticulospinal system and its role in the control of movement. In: Barnes C.D. (Ed.), Brainstem Control of Spinal Cord Function. No. 6 as Research Topics in Physiology. Academic Press, London, pp. 27–86. Pilyavsky, A.I. and Gokin, A.P. (1978). Investigation of the cortico-reticulo-spinal connections in cats. Neuroscience, 3: 99–103. Porter, R. and Lemon, R. (1993). Corticospinal Function and Voluntary Movement. No. 45 as Monographs of the Physiological Society. Oxford Science Publication, Oxford. Prentice, S.D. and Drew, T. (2001). Contribution of the reticulospinal system to the postural adjustments occurring during voluntary gait modifications. J. Neurophysiol., 85: 679–698.
Rho, M.J., Cabana, T. and Drew, T. (1997). Organization of the projections from the pericruciate cortex to the pontomedullary reticular formation of the cat: a quantitative retrograde tracing study. J. Comp. Neurol., 388: 228–249. Schaltenbrand, G. and Cobb, S. (1931). Clinical and anatomical studies on two cats without neocortex. Brain, 53: 449–488. Scheibel, M.E. and Scheibel, A.B. (1969). A structural analysis of spinal interneurons and Renshaw cells. No. 11 as UCLA Forum in Medical Sciences. In: Brazier M.A.B. (Ed.), The Interneuron. University of California Press, Los Angeles, pp. 159–208. Shik, M.L., Severin, F.V. and Orlovsky, G.N. (1966). Control of walking and running by means of electrical stimulation of the midbrain. Biophysics, 11: 755–765. Shik, M.L., Orlovsky, G.N. and Severin, F.V. (1968). Locomotion of the mesencephalic cat evoked by pyramidal stimulation. Biophysics, 13: 127–135. Steeves, J.D. and Jordan, L.M. (1984). Autoradiographic demonstration of the projections from the mesencephalic locomotor region. Brain Res., 307: 263–276. Terakado, Y. and Yamaguchi, T. (1990). Last-order interneurones controlling activity of elbow flexor motoneurons during forelimb fictive locomotion in the cat. Neurosci. Lett., 111: 292–296. Yamaguchi, T. (1986). Descending pathways eliciting forelimb stepping in the lateral funiculus: experimental studies with stimulation and lesion of the cervical cord in decerebrate cats. Brain Res., 379: 125–136.