Articles in PresS. J Neurophysiol (August 20, 2008). doi:10.1152/jn.90786.2008
Propriospinal input to shoulder
Task-dependent modulation of propriospinal inputs to human shoulder
Lynley V Roberts1, Cathy M Stinear1, Gwyn N Lewis2,3,4, Winston D Byblow1 1
Movement Neuroscience Laboratory, University of Auckland, Auckland, New Zealand
2
Health and Rehabilitation Research Centre, Auckland University of Technology,
Auckland, New Zealand 3
Department of Physical Medicine and Rehabilitation, Northwestern University, Chicago,
USA 4
Sensory Motor Performance Program, Rehabilitation Institute of Chicago, Chicago, USA
Running Head: Propriospinal inputs to shoulder Keywords: human, shoulder, hand, propriospinal, motor evoked potential, transcranial magnetic stimulation,
Correspondence: Winston D. Byblow Movement Neuroscience Laboratory University of Auckland Auckland, New Zealand Phone 373 7599 x 84897 Fax 373 7043
[email protected]
Copyright © 2008 by the American Physiological Society.
1
Propriospinal input to shoulder
2
Abstract In the human upper limb a proportion of the descending corticospinal command may be relayed through cervical propriospinal premotoneurons. This may serve to co-ordinate movements involving multiple joints of the arm, such as reaching. The present study was conducted to determine if a shoulder stabilising muscle, infraspinatus (INF), is functionally integrated into the putative cervical propriospinal network, and if there is task-dependent modulation of the network. Fourteen healthy adults participated in this study. Responses in the right INF were evoked by transcranial magnetic stimulation (TMS) over the motor cortex and compared to responses conditioned by ulnar nerve stimulation. Interstimulus intervals were chosen to summate inputs at the level of the premotoneurons. Participants performed a forearm and shoulder muscle co-contraction task or a grip-lift task that also coactivated forearm and shoulder muscles. During the co-contraction task, INF motor evoked potentials (MEPs) were significantly facilitated by ulnar nerve stimulation at low intensities and suppressed at higher intensities. Only facilitation reached significance during the griplift task. We have shown for the first time that propriospinal pathways may connect the hand to the rotator cuff of the shoulder. The modulation of facilitation/suppression during the grip-lift task suggests that inhibition of propriospinal premotoneurons is downregulated in a task dependent manner to increase the gain in the feedback reflex loop from forearm and hand muscles as required.
Propriospinal input to shoulder
3
Introduction The ability to reach, grip and manipulate objects with the hand is a quintessential function in higher order primates. Stereotypical forelimb movements in the cat are coordinated by a disynaptic component of the corticospinal tract, known as the cervical propriospinal system (Alstermark et al. 2007). The contribution of these premotoneurons in the control of the upper limb in higher primates has been under debate (Lemon and Griffiths 2005; Maier et al. 1998; Nakajima et al. 2000). Recent experiments in the macaque monkey have revealed the propriospinal system is subject to robust inhibitory control (Alstermark et al. 1999; Isa et al. 2006). In humans the disynaptic pathway is also constrained by inhibitory interneurons, in which the function is unclear (Iglesias et al. 2007; Nicolas et al. 2001). Indirect evidence suggests the cervical propriospinal system may co-ordinate muscle synergies necessary for upper limb reaching in humans (Gracies et al. 1991; Pauvert et al. 1998; see Pierrot-Deseilligny and Burke 2005). Sensory inputs are integrated with descending cortical outputs via feedback loops that may update the motor command en route to motoneurons. Sensory feedback serves to focus corticomotor drive to propriospinal premotoneurons appropriate for the motor synergy (Burke et al. 1992; Mazevet and Pierrot-Deseilligny 1994). The gain of these reflexes relies on the balance of excitation between propriospinal premotoneurons and inhibitory interneurons (Iglesias et al. 2007; Nicolas et al. 2001). The propriospinal system can be studied indirectly in humans using transcranial magnetic stimulation (TMS) conditioned by peripheral nerve stimulation. With appropriate interstimulus intervals, paired TMS and peripheral stimulation of propriospinally-linked muscles leads to a facilitation of the motor evoked potential (MEP). Two characteristics
Propriospinal input to shoulder
4
identify the propriospinal neurons as the putative site of summation. First, the interstimulus interval (ISI) between cortical and peripheral stimulation that gives rise to facilitation is longer than that expected for a monosynaptic reflex. This is thought to reflect the additional distance for afferent input from peripheral stimulation to reach the more superiorly located premotoneurons in the spinal cord (Gracies et al. 1991; Pauvert et al. 1998). Second, peripheral conditioning stimuli facilitate MEPs elicited by weak TMS intensities, but suppress those evoked by stronger TMS intensities. This implies that weak cortical and weak peripheral inputs summate at propriospinal neurons to facilitate MEPs, whereas strong cortical or peripheral inputs may converge upon inhibitory interneurons that serve to suppress MEPs through a disynaptic pathway (Iglesias et al. 2007; Nicolas et al. 2001). Using these techniques, excitability of the propriospinal system linking forearm and arm muscles has been shown to be task dependent. Transmission through propriospinal neurons is increased under muscle fatiguing conditions (Martin et al. 2007), during bilateral movement tasks (Stinear and Byblow 2004b) and when making dexterous movements with the hand (Iglesias et al. 2007; Marchand-Pauvert and Iglesias 2008). Altered propriospinal activity has been implicated in abnormal muscle synergies in patients who have experienced stroke (Mazevet et al. 2003; Stinear and Byblow 2004a). To date, there is no evidence of propriospinal input to muscles of the rotator cuff complex. Since gripping and precise use of the hand in humans is accompanied by changes in the excitability of the rotator cuff muscles (Laursen et al. 1998; Sporrong et al. 1996; 1995; Sporrong and Styf 1999), sensory feedback from contracting hand muscles may project to the rotator cuff via propriospinal premotoneurons. The aim of this experiment was to assess the role
Propriospinal input to shoulder
5
of the putative propriospinal system in controlling infraspinatus (INF) and to assess if task differences in MEP ratios suggest task-dependent modulation of the propriospinal premotoneurons. INF was chosen as it is an important muscle in the rotator cuff group. It contracts during shoulder flexion and abduction to control the head of the humerus in the glenoid socket (Kronberg et al. 1990; Palmerud et al. 2000) and is routinely activated during upper limb reaching movements. INF was paired with flexor carpi ulnaris (FCU), a wrist flexor muscle that is also commonly activated during reaching and grasping. Our hypothesis was that sensory feedback from the ulnar nerve affects transmission of descending drive to INF via the cervical propriospinal system, and furthermore is modulated by task.
Methods Participants Fourteen healthy adults (age range 21 – 47 years, mean age 29 years, 3 male) with no history of upper limb musculoskeletal or neurological disorder completed the study. Subjects provided informed consent in accordance with the Declaration of Helsinki, and the study was approved by the local ethics committee. Electromyography Subjects were seated and surface electromyography (EMG) was recorded from the right INF muscle by adhesive electrodes (Ambu, Ballerup, Denmark) positioned 3 cm below the midpoint of the spine of the scapula and aligned with the direction of the underlying muscle fibres. Surface EMG was recorded from the right FCU using a bellytendon montage. EMG signals were amplified (Grass P511AC, Grass Instrument
Propriospinal input to shoulder
6
Division, West Warwick, RI), filtered with a bandwidth of 30 – 1000 Hz, sampled at a frequency of 2 kHz (National Instruments, Austin, TX) and displayed on a computer by custom LabView software.
Transcranial Magnetic Stimulation Single pulse TMS (Magstim 200, Dyfed, Wales) was applied over the left hemisphere to elicit MEPs in the right INF. A figure of eight coil (90-mm wing diameter) induced a current directed posterior to anterior in the underlying tissue. The stimulation site evoking the largest amplitude MEP in the right INF was located and marked on the scalp. Active motor threshold (AMT) for each task (described below) was determined as the minimum stimulus intensity that elicited a 100 µV MEP in the INF in five out of ten stimuli. Five stimulus intensities were then used during task performance, in increments of 2% maximal stimulator output (MSO) starting from task AMT. The order of stimulus intensity was randomized. At each intensity, twelve non-conditioned and twelve peripheral nerve conditioned trials were collected at a rate of 0.2 Hz. Conditioning Peripheral Nerve Stimulation A Digitimer DS7A constant current stimulator (Digitimer, Hertfordshire, UK) delivered a square wave, l ms duration pulse every 5 seconds via an adhesive electrode (Ambu, Ballerup, Denmark) to the ulnar nerve at the elbow. The anode was placed over the medial aspect of the arm just above the elbow. The intensity of the stimulus was set to 0.8 x motor threshold to preferentially stimulate group I sensory afferents (Nicolas et al. 2001).
Propriospinal input to shoulder
7
The ISI between peripheral nerve stimulation and TMS were calculated using indirect methods (Alexander and Harrison 2003; Roberts et al. 2008) and confirmed in pilot experiments. The calculations were as follows (rounded to nearest ms for variations in height/distance). Efferent conduction time from M1 to INF is 11 ms, indicative by MEP latency. Based on distance (≈ 0.24 m), efferent conduction time from INF alpha motoneurons ( MN) to surface EMG electrodes is 5 ms accounting for delays at the neuromuscular junction. Therefore, conduction time from M1 to INF MN is 6 ms. Based on distance (≈ 0.5 m), afferent conduction time from peripheral stimulation at elbow to INF MN is 10 ms accounting for a 1 ms synapse onto MN. Therefore, in order for peripheral stimulus and TMS to interact at the INF MN an ISI 4 ms would be required (10 – 6 = 4 ms). Since we are interested in the interaction of the afference the premotoneuronal propriospinal circuits, we have added an additional 3ms for central conduction time (see Pierrot-Deseilligny and Burke 2005), and estimated an ISI of ≈ 7 ms between the peripheral stimulus and TMS in order to facilitate summation at the level of propriospinal neurons. Therefore, at the beginning of each experiment ISIs of 6, 7 and 8 ms were explored with a TMS intensity of task AMT + 2%. The ISI that produced the largest MEP facilitation was considered optimal, and used for the remainder of the experiment. Motor Tasks Two motor tasks were employed to investigate the modulation of propriospinal premotoneurons. The first was a simple co-contraction task. Participants were seated with their hands on their lap. When verbally cued they flexed their right shoulder to an angle of 90 degrees with the elbow straight, activating the INF, and then ulnar deviated the wrist.
Propriospinal input to shoulder
8
FCU EMG activity during ulnar deviation triggered the stimulators when a threshold of 0.1 mV was reached. After 5 s they returned the hand to their lap. The second task was a grip-lift task. Participants were again seated with their hands on their lap. From this starting position, they were asked to reach forward, grasp and lift a purpose built device from a table directly in front of them, using the thumb and fourth and fifth digits of their right hand. Participants held the device at a constant height for approximately 3 s, and then returned it to the table. Stimulators were triggered 400 ms following the onset of vertical lift so that stimuli were delivered near the lift endpoint. At this time point, both INF and FCU were typically activated, and participants were lifting with shoulder flexion, while the elbow remained extended. For both tasks, trials were rejected online if INF rmsEMG was outside the range 0.01 - 0.15 mV in the 100 ms preceding stimulation.
Data Analysis EMG data were full-wave rectified off-line using custom software. The area of the rectified MEP in INF EMG was measured from the onset latency (11 - 18 ms depending on the individual) using the same analysis window for all trials in each participant. For each participant, the stimulus intensity that produced the greatest facilitation was identified for each task. Responses to four stimulus intensities were analysed: maximal facilitation (F), F - 2 %, F + 2 % and F + 4 %. Mean MEP area was expressed as a ratio (conditioned/non-conditioned) at each of these stimulus intensities. MEP ratio was analysed using a 2 x 2 repeated measures analysis of variance (ANOVA) with factors Task (co-contraction, grip-lift) and Intensity (F - 2 %, F, F + 2 %, F + 4 %). Facilitation or
Propriospinal input to shoulder
9
suppression with respect to non-conditioned trials was assessed with one-sample t-tests. Paired t-tests assessed the effect of task on facilitation and suppression phases. Post hoc tests were adjusted for multiple comparisons using standard procedures (Rom 1990). Background rmsEMG was calculated from 100 - 10 ms prior to the stimulus and averaged across stimulus intensities for both non-conditioned and conditioned responses. The relationship between background EMG and MEP ratio was assessed by regression analysis to determine if the level of pre-stimulus EMG influenced MEP modulation. Taskdependent modulation of corticomotor excitability was assessed by plotting stimulusresponse curves of INF MEP area from non-conditioned trials at F – 2%, F, F + 2% and F + 4%. For each task, the slope of the line of best fit from linear regression of MEP area was determined for each participant, and the effect of task assessed using a paired t-test. Regression analyses were performed to explore any relationship between background EMG non-conditioned MEP size and MEP ratio. Significance level was set at p = 0.05. Mean ± standard error are reported in the text.
Results During the co-contraction task of the right INF and FCU muscles, electrical stimulation of the ulnar nerve facilitated INF MEPs at TMS intensities just above task AMT. The intensity producing maximal facilitation (F) was 40.6 ± 4.9% MSO (103.14 ± 1.87 % Task AMT). Similarly for the grip-lift task, F was 39.9 ± 5.8 % MSO (102.85 ± 1.03 % Task AMT). During co-contraction INF MEPs were maximally suppressed at F + 2 % MSO. Both facilitation and suppression of INF MEPs were attenuated during the gripping and lifting task compared to co-contraction.
Propriospinal input to shoulder
10
Effects of TMS Intensity and Task on MEP ratio Figure 1 illustrates rectified EMG traces from a typical subject. TMS intensitydependent facilitation and suppression of INF MEPs for the co-contraction and grip-lift tasks can be seen. MEP ratios for the group (conditioned/non-conditioned expressed as percentage change) are shown in Figure 2. There was a main effect of Intensity on INF MEP area (F3,11 = 41.8, p < 0.001), and an interaction between Task and Intensity (F3,11 = 4.4, p < 0.01). For the cocontraction task, at maximum facilitation (F), conditioned MEPs were 26% larger than non-conditioned (p < 0.001). Conversely at maximum suppression (F + 2%), conditioned MEPs were 16% smaller than non-conditioned (p < 0.001). There were no effects of the conditioning stimulus at F – 2% or F + 4% (all p > 0.19). During the grip-lift task, conditioned INF MEPs at F were 18% larger than non-conditioned (p < 0.01). There were no effects at F – 2%, F + 2% or F + 4% (all p > 0.11). Comparison of F + 2% (maximum suppression) between tasks using paired t-tests revealed more suppression during cocontraction than for the grip-lift task (p < 0.01) (Figure 2). There were no differences between tasks at F – 2%, F or F + 4% (all p > 0.24). Rectified traces were superimposed and visually inspected. There was no obvious deviation in the initial rising phase of the MEP between non-conditioned and conditioned traces. Effects of Non-conditioned MEP Area and Background EMG on MEP Ratio Regression analysis revealed no effect of non-conditioned MEP area on MEP ratio for either task (Co-contraction, r2 = 0.04, p = 0.12; Grip-Lift r2 = 0.009, p = 0.63). Therefore, non-conditioned MEP area did not significantly influence the size of the MEP
Propriospinal input to shoulder
11
ratio. Similarly, background EMG was not significantly correlated to MEP ratio for either task (Co-contraction, r2 = 0.018 or Grip-Lift, r2 = 0.003, both p > 0.3). Background EMG rms values were 0.030 ± 0.003 mV for co-contraction and 0.043 ± 0.005 mV for grip-lift. Changes in INF MEP ratios most likely reflect changes in excitability due to conditioning rather than fluctuations in background EMG. Corticospinal Excitability A stimulus response curve of the mean non-conditioned INF MEPs from a typical subject is shown in Figure 3. This demonstrates larger MEPs during the grip-lift task and a steeper recruitment slope. For the group data, the slope was steeper during grip-lift (0.25 ± 0.03 mV) than during co-contraction (0.15 ± 0.18 mV) (F 1,13 = 7.07, p < 0.05). Regression analyses revealed no significant correlation between non-conditioned MEP amplitude and background EMG. For co-contraction r2 = 0.07, and for grip-lift r2 = 0.03, (all p > 0.67). The steeper slope of non-conditioned MEPs is not a function of increased EMG, but a task-related increase in corticomotor drive. Discussion This is the first report, to our knowledge, that the INF muscle may be integrated into the cervical propriospinal system. Afferent feedback from forearm and hand muscles supplied by the ulnar nerve appears to modulate excitability of INF via premotoneurons. Furthermore, the extent that this feedback modifies excitability depends on the task. We propose that common propriospinal interneurons may form synergies between forearm, hand and shoulder muscles that function to assist in stabilising the shoulder during use of the hand. Propriospinal Modulation between Hand and Shoulder
Propriospinal input to shoulder
12
During co-contraction of the INF and FCU, INF MEPs elicited by weak TMS intensities were facilitated by peripheral input, while INF MEPs elicited by stronger TMS intensities were suppressed. INF MEP area was unaffected by further changes in TMS output. This putative propriospinally-mediated response pattern is in line with previous studies of propriospinal control of more distal muscles using similar techniques (Iglesias et al. 2007; Nicolas et al. 2001; Stinear and Byblow 2004b). These results support our hypothesis that the INF muscle may be controlled by propriospinal networks and can respond to feedback from afferents within the ulnar nerve. This sensory information probably serves to fine-tune the activity of INF for shoulder stability during upper limb reaching. The observed facilitation and suppression of MEPs likely occurs at the premotoneuronal level because the ISI used is too long to effect a monosynaptic response. An ISI of 4 ms for monosynaptic, and 7 ms for propriospinal summation was estimated based on the work of others and tested in pilot studies (Alexander and Harrison 2003; Roberts et al. 2008). Our average ISI of 7 ms evoking maximal INF MEP facilitation was consistent with a propriospinal pathway. Furthermore, at this ISI, the low threshold pathway subserving facilitation gives way to inhibition with higher TMS intensities. This effect seems most likely if responses are mediated at the same interneuron (Nicolas et al., 2001). Facilitation and suppression in the current study were evoked at 6, 7 or 8 ms depending on the individual, which are similar to those reported for other peripheral nerve and muscle combinations (Iglesias et al. 2007; Nicolas et al. 2001; Stinear and Byblow 2004b). Finally, the initial portion of the rising MEP slope did not appear altered by conditioning similar to earlier studies (Mazevet et al. 1996; Nicolas
Propriospinal input to shoulder
13
et al. 2001), although this was not measured quantitatively in the present study. Overall the data support the idea that the observed facilitation and inhibition at F and F+2% respectively are not the result of monosynaptic inputs to the INF motoneurons but rather, mediated via interposed interneurons. Ulnar nerve afferents provide feedback to other shoulder muscles besides INF. Deltoid, triceps and biceps also share propriospinally mediated connections (Gracies et al. 1991). Electrical stimulation of the ulnar nerve evokes long latency, possibly transcortical, reflexes in shoulder girdle muscles (Alexander and Harrison 2003; Alexander et al. 2005) and in INF and deltoid (Roberts et al. 2008). The lack of monosynaptic feedback projections from the ulnar nerve to deltoid (Creange et al. 1992), INF (Roberts et al. 2008) or the shoulder girdle muscles (Alexander and Harrison 2003; Alexander et al. 2005) lends further support to the idea of a propriospinally-mediated site of summation and underscores the important role of descending corticomotor commands in modulating feedback from the hand. The pathways mediating peripheral afference from the hand were not examined directly. However we consider it likely that the peripheral stimulation acted upon large diameter fibres, due to the low stimulus intensity used for evoking responses (Lourenco et al. 2006). Although the grip-lift task stimulated cutaneous receptors in the hand whereas the co-contraction task did not, this is unlikely to have influenced INF MEPs since cutaneous stimulation to the palmer side of the fingers does not alter responses evoked by ulnar nerve stimulation (Iglesias et al. 2007). Task-dependent Modulation of Propriospinal Premotoneurons
Propriospinal input to shoulder
14
When performing the grip-lift task, both facilitatory and inhibitory effects of the conditioning stimulus were attenuated. MEPs were less facilitated at F, although this did not differ significantly from the co-contraction task. In contrast, inhibition at F + 2% was significantly different between tasks, with no inhibition in the grip-lift task. The contrasting responses between co-contraction and grip-lift tasks at F + 2% suggest differential modulation via the propriospinal system. Background EMG activity or nonconditioned MEP size are unlikely explanatory variables for effects due to task, since neither correlated with MEP ratio. Less INF suppression at stronger TMS intensities during grip-lift suggests that descending drive to propriospinal premotoneurons may be altered by task. Task-related changes in propriospinal transmission were recently described between hand and wrist muscles (Iglesias et al. 2007), where flexor carpi radialis (FCR) MEPs were recorded during voluntary contractions, gripping and pointing using similar methods to the present study. Task related modulations were assessed by increasing peripheral stimulus intensity rather than TMS intensity, inducing the typical pattern of facilitation followed by suppression of MEPs during voluntary contractions. Differential effects of task were only observed at intensities greater than 0.8 x MT, where FCR MEPs were facilitated during gripping compared to voluntary contractions and pointing (Iglesias et al. 2007). These findings, combined with those of the present study, suggest that forearm and hand muscle afferents have strong divergent projections to more proximal joints via propriospinal premotoneurons and that these projections are modulated during functional activities. Task-dependent Modulation of Corticomotor Excitability
Propriospinal input to shoulder
15
The INF recruitment curves derived from non-conditioned trials indicate greater excitability during the grip-lift task. Difference in background EMG between tasks may have contributed in some part to this difference, but is unlikely to be the sole contributor since there was no correlation between EMG level at time of stimulation and non-conditioned INF MEP size. The greater excitability of direct corticomotor projections during grip-lift may serve to stabilise the shoulder for this task. Task-related increases in INF excitability during gripping appear to occur in parallel with decreased corticospinal drive to inhibitory interneurons. Disfacilitation of the inhibitory pathway would allow peripheral afferents to excite propriospinal neurons more than inhibitory interneurons (Figure 4). Alternatively, the difference in INF MEPs between the two tasks may be due to different proportions of cortical command relayed via propriospinal interneurons. However, if this were the case greater facilitation during grip-lift should have been apparent at both F and F + 2%, but this was not observed. Therefore, we propose that disinhibition of inhibitory interneurons is a more likely mechanism contributing to the effect of task. Greater sensitivity to afference from the hand may allow feedback to increase indirect corticospinal drive to INF. Any change in the external environment signaled by hand afferents can then adjust co-ordination to maintain task performance. Concurrent activation of indirect propriospinal pathways in this manner likely provides adaptable movement control allowing integration of sensory information from the forearm and hand during purposeful movement.
Propriospinal input to shoulder
16
Acknowledgements: The authors would like to acknowledge the input of Dr Caroline Alexander in the formative stage of this work. Thanks also to Carolina Dillen and Andrew Stevenson, for their assistance with data collection, and to anonymous reviewers for their helpful comments.
Grants: LVR is supported by a University of Auckland Senior Health Research Scholarship
Propriospinal input to shoulder
17
References Alexander CM, and Harrison PJ. Reflex connections from forearm and hand afferents to shoulder girdle muscles in humans. Experimental Brain Research 148: 277-282, 2003. Alexander CM, Miley R, and Harrison PJ. Functional modulation of shoulder girdle stability. Experimental Brain Research 161: 417-422, 2005. Alstermark B, Isa T, Ohki Y, and Saito Y. Disynaptic pyramidal excitation in forelimb motoneurons mediated via C(3)-C(4) propriospinal neurons in the Macaca fuscata. Journal of Neurophysiology 82: 3580-3585, 1999. Alstermark B, Isa T, Pettersson LG, and Sasaki S. The C3-C4 propriospinal system in the cat and monkey: a spinal pre-motoneuronal centre for voluntary motor control. Acta Physiologica 189: 123-140, 2007. Burke D, Gracies JM, Mazevet D, Meunier S, and Pierrot-Deseilligny E. Convergence of descending and various peripheral inputs onto common propriospinallike neurones in man. Journal of Physiology 449: 655-671, 1992. Creange A, Faist M, Katz R, and Penicaud A. Distribution of heteronymous Ia facilitation and recurrent inhibition in the human deltoid motor nucleus. Experimental Brain Research 90: 620-624, 1992. Gracies JM, Meunier S, Pierrot-Deseilligny E, and Simonetta M. Pattern of propriospinal-like excitation to different species of human upper limb motoneurones. Journal of Physiology 434: 151-167, 1991. Iglesias C, Marchand-Pauvert V, Lourenco G, Burke D, and Pierrot-Deseilligny E. Task-related changes in propriospinal excitation from hand muscles to human flexor carpi radialis motoneurones. Journal of Physiology 582: 1361-1379, 2007. Isa T, Ohki Y, Seki K, and Alstermark B. Properties of propriospinal neurons in the C3-C4 segments mediating disynaptic pyramidal excitation to forelimb motoneurons in the macaque monkey. Journal of Neurophysiology 95: 3674-3685, 2006. Kronberg M, Nemeth G, and Brostrom LA. Muscle activity and coordination in the normal shoulder. An electromyographic study. Clinical Orthopaedics and Related Research 76-85, 1990. Laursen B, Jensen BR, and Sjogaard G. Effect of speed and precision demands on human shoulder muscle electromyography during a repetitive task. European Journal of Applied Physiology and Occupational Physiology 78: 544-548, 1998. Lemon RN, and Griffiths J. Comparing the function of the corticospinal system in different species: organizational differences for motor specialization? Muscle & Nerve 32: 261-279, 2005. Lourenco G, Iglesias C, Cavallari P, Pierrot-Deseilligny E, and Marchand-Pauvert V. Mediation of late excitation from human hand muscles via parallel group II spinal and group I transcortical pathways. Journal of Physiology 572: 585-603, 2006. Maier MA, Illert M, Kirkwood PA, Nielsen J, and Lemon RN. Does a C3-C4 propriospinal system transmit corticospinal excitation in the primate? An investigation in the macaque monkey. Journal of Physiology 511 ( Pt 1): 191-212, 1998. Marchand-Pauvert V, and Iglesias C. Properties of human spinal interneurones: normal and dystonic control. Journal of Physiology 586: 1247-1256, 2008.
Propriospinal input to shoulder
18
Martin PG, Gandevia SC, and Taylor JL. Muscle fatigue changes cutaneous suppression of propriospinal drive to human upper limb muscles. Journal of Physiology 580: 211-223, 2007. Mazevet D, Meunier S, Pradat-Diehl P, Marchand-Pauvert V, and PierrotDeseilligny E. Changes in propriospinally mediated excitation of upper limb motoneurons in stroke patients. Brain 126: 988-1000, 2003. Mazevet D, and Pierrot-Deseilligny E. Pattern of descending excitation of presumed propriospinal neurones at the onset of voluntary movement in humans. Acta Physiologica Scandinavica 150: 27-38, 1994. Mazevet D, Pierrot-Deseilligny E, and Rothwell JC. A propriospinal-like contribution to electromyographic responses evoked in wrist extensor muscles by transcranial stimulation of the motor cortex in man. Experimental Brain Research 109: 495-499, 1996. Nakajima K, Maier MA, Kirkwood PA, and Lemon RN. Striking differences in transmission of corticospinal excitation to upper limb motoneurons in two primate species. Journal of Neurophysiology 84: 698-709, 2000. Nicolas G, Marchand-Pauvert V, Burke D, and Pierrot-Deseilligny E. Corticospinal excitation of presumed cervical propriospinal neurones and its reversal to inhibition in humans. Journal of Physiology 533: 903-919, 2001. Palmerud G, Forsman M, Sporrong H, Herberts P, and Kadefors R. Intramuscular pressure of the infra- and supraspinatus muscles in relation to hand load and arm posture. European Journal of Applied Physiology 83: 223-230, 2000. Pauvert V, Pierrot-Deseilligny E, and Rothwell JC. Role of spinal premotoneurones in mediating corticospinal input to forearm motoneurones in man. Journal of Physiology 508 ( Pt 1): 301-312, 1998. Pierrot-Deseilligny E, and Burke D. The Circuitry of the Human Spinal Cord. Its Role in Motor Control and Movement Disorders. Cambridge: Cambridge University Press, 2005. Roberts L, Wellington J, Harrison A, and Alexander CM. Reflex connections of the infraspinatus muscle in healthy humans. New Zealand Society of Physiotherapists Biennial Conference Abstracts 2008. Rom DM. A sequentially rejective test procedure based on a modified Bonferroni inequality. Biometrika 77: 663 - 665, 1990. Sporrong H, Palmerud G, and Herberts P. Hand grip increases shoulder muscle activity, An EMG analysis with static hand contractions in 9 subjects. Acta Orthopaedica Scandinavica 67: 485-490, 1996. Sporrong H, Palmerud G, and Herberts P. Influences of handgrip on shoulder muscle activity. European Journal of Applied Physiology and Occupational Physiology 71: 485492, 1995. Sporrong H, and Styf J. Effects of isokinetic muscle activity on pressure in the supraspinatus muscle and shoulder torque. Journal of Orthopaedic Research 17: 546-553, 1999. Stinear JW, and Byblow WD. The contribution of cervical propriospinal premotoneurons in recovering hemiparetic stroke patients. Journal of Clinical Neurophysiology 21: 426-434, 2004a.
Propriospinal input to shoulder
19
Stinear JW, and Byblow WD. Modulation of human cervical premotoneurons during bilateral voluntary contraction of upper-limb muscles. Muscle & Nerve 29: 506-514, 2004b.
Figure Captions Figure 1. Averaged rectified MEPs from one representative subject during co-contraction (left column) and grip-lift (right column) are shown for the four TMS intensities analysed. Non-conditioned (solid line) and conditioned traces (dashed line) are superimposed. The small arrow indicates when TMS was delivered (all trials); the large arrow indicates time of peripheral nerve stimulation (conditioned trials only). The MEP area was calculated within the time window depicted between the dashed vertical lines. The ratio of the conditioned response to the non-conditioned response (C/NC) is given at the right of each set of traces. Note the facilitation and inhibition of conditioned responses during the cocontraction task compared to grip-lift task at F and F+2%.
Figure 2. Conditioned MEP ratios (expressed as percent change from non-conditioned MEP ratios) during the co-contraction task (grey bars) and grip-lift task (black bars). Bars represent group averages (n = 14). Asterisks indicate significant facilitation (p < 0.01) or inhibition (p < 0.001) relative to non-conditioned trials (one-sample t-tests) and difference between tasks at F + 2% (paired t-test) (p < 0.01). Error bars indicate SEM.
Propriospinal input to shoulder
20
Figure 3. Stimulus-response curves of a typical subject during the cocontraction (grey) and grip-lift (black) tasks. Each data point represents the mean INF MEP area from 12 non-conditioned trials. Slope was estimated using linear regression.
Figure 4. Schematic of putative propriospinal connections between hand and shoulder investigated in the present study. Left. During co-contraction, the propriospinal premotoneurons (PN, large open circle) projecting to INF motoneurons are subject to excitation via descending corticospinal neurons (black lines on left) and ascending input from ulnar nerve afferents. Stronger TMS intensities recruit higher threshold descending inputs (thin grey lines on right) which summate on inhibitory interneurons and inhibit the PN. Right. “Disfacilitation” of descending projections (grey line) to inhibitory interneurons during the gripping and lifting task. Reduced corticospinal drive to the inhibitory neuron reduces its response to peripheral afferent input allowing for greater excitation of infraspinatus alpha motoneurons via PN.
0.15
0.15
TMS Intensity 0.1
F – 2%
0.1
1.01
0.05
1.05
0.05
0.2
0.2
0
0 0
50
100
150
200
0
0.15
0.15
0.1
0.1
F
50
1.31
100
150
200
150
200
150
200
150
200
1.14
0.05
0.05
0.2
0.2
0
0 0
50
100
150
200
0
0.15
0.15
0.1
0.1
F + 2%
50
0.67
100
0.94
0.05
0.05
0.2
0.2 0
0 0
50
100
150
0
200
0.15
0.15
0.1
0.1
F + 4%
50
0.98
100
0.99
0.05
0.05
0
0 0
50
100
150
200 200 µV
0
20 ms
50
100
40
*
MEP Ratio (%Change)
30
*
20 10 0 -10 -20 * -30
*
-40 F-2%
F
F+2%
TMS Intensity (% MSO)
F+4%
2.5
MEP Area (mV s)
2
1.5
1
0.5
0 F - 2%
F
F + 2%
Stimulus Intensity (MSO% relative to F)
F + 4%
Corticospinal neurons
Corticospinal neurons
Inhibitory interneuron
Inhibitory interneuron
PN
PN
α MN
α MN
Infraspinatus
Ulnar Nerve Afferents
Infraspinatus
Ulnar Nerve Afferents
