OPINION published: 13 January 2021 doi: 10.3389/fnhum.2020.622533 Novel Non-invasive Strategy for Spinal Neuromodulation to Control Human Locomotion Tatiana Moshonkina 1 , Alexander Grishin 1 , Irina Bogacheva 1 , Ruslan Gorodnichev 2 , Alexander Ovechkin 3,4 , Ricardo Siu 3 , V. Reggie Edgerton 5,6,7 and Yury Gerasimenko 1,8* 1 Pavlov Institute of Physiology, Russian Academy of Sciences, St. Petersburg, Russia, 2 Velikie Luki State Academy of Physical Education and Sports, Velikiye Luki, Russia, 3 Kentucky Spinal Cord Injury Research Center, Frazier Rehab Institute, University of Louisville Health, University of Louisville, Louisville, KY, United States, 4 Department of Neurological Surgery, University of Louisville, Louisville, KY, United States, 5 Department of Neurobiology, and Neurosurgery, University of California, Los Angeles, Los Angeles, CA, United States, 6 Institut Guttmann Hospital de Neurorehabilitació, Institut Universitari adscrit a la Universitat Autònoma de Barcelona, Barcelona, Spain, 7 Faculty of Science, The Center for Neuroscience and Regenerative Medicine, University of Technology Sydney, Ultimo, NSW, Australia, 8 Department of Physiology and Biophysics, University of Louisville, Louisville, KY, United States Keywords: neuromodulation, paralysis, transcutaneous electrical spinal cord stimulation, humans, motor pools INTRODUCTION It is well-documented that neural control of stepping and standing can be generated in mammals Edited by: within the spinal neuronal networks. Having a high level of automaticity, these locomotor- Yury Ivanenko, related neuronal circuits can produce a “stepping” movement pattern in the absence of input Santa Lucia Foundation (IRCCS), Italy from the brain and/or peripheral afferent inputs. Recent observations have provided important Reviewed by: insight into the properties of the spinal and supraspinal circuitry that are involved in movement Sukhvinder Kalsi-Ryan, control. We have shown that the spinal circuitry in mice, rats, cats, and humans can be University Health Network neuromodulated to regain sensorimotor function after complete paralysis (Gerasimenko et al., (UHN), Canada 2008). We have also shown that with epidural spinal cord stimulation at the lumbar level, Kristine C. Cowley, University of Manitoba, Canada full weight-bearing standing, and voluntary movements of the lower limb can be recovered in humans with complete paralysis (Angeli et al., 2014). Altering spinal cord excitability enables *Correspondence: voluntary movements after chronic complete paralysis in humans. Recent breakthrough studies Yury Gerasimenko gerasimenko@infran.ru reported that chronically paralyzed individuals regained the over-ground walking with balance assistance through interleaved continous lumbosacral (L1-S1) epidural stimulation and task Specialty section: specific locomotor training (Angeli et al., 2018; Gill et al., 2018). This article was submitted to We have developed a novel method of non-invasive transcutaneous spinal cord stimulation Motor Neuroscience, (scTS) which can modulate the excitability of spinal circuitry via electrodes placed on the skin a section of the journal overlying the cervical, lower thoracic, lumbosacral, and coccygeal vertebrae using a special form Frontiers in Human Neuroscience of electrical pulses delivered at a high frequency (Gerasimenko et al., 2015a). This methodology Received: 28 October 2020 was able to neuromodulate the spinal locomotor networks such that involuntary stepping-like Accepted: 14 December 2020 movements were induced in non-injured subjects when their legs were placed in a “gravity-neutral” Published: 13 January 2021 apparatus (Gerasimenko et al., 2015b). In addition, our initial results with scTS show that this Citation: strategy can facilitate individuals having motor complete paralysis to generate rhythmic stepping Moshonkina T, Grishin A, Bogacheva I, patterns and non-repetitive voluntary movements (Gerasimenko Y. P. et al., 2015). The novel Gorodnichev R, Ovechkin A, Siu R, finding in this and ongoing studies is that specifically configured multisite stimulation can produce Edgerton VR and Gerasimenko Y a more robust response when compared to the single site stimulation. Based on these findings, (2021) Novel Non-invasive Strategy for Spinal Neuromodulation to Control we developed a three-by-three multielectrode transcutaneous array that allows multiple sites to Human Locomotion. be modulated, thus, providing subject-specific options for controlling posture and locomotor Front. Hum. Neurosci. 14:622533. behavior (Gerasimenko et al., 2015a). We observed that the effectiveness of inducing of involuntary doi: 10.3389/fnhum.2020.622533 stepping movements in a non-injured subject with legs placed in a “gravity-neutral” position during Frontiers in Human Neuroscience | www.frontiersin.org 1 January 2021 | Volume 14 | Article 622533 Moshonkina et al. Non-invasive Neuromodulation to Control Locomotion spinal cord stimulation at one level (T11) with 3 interconnected Meanwhile, unilateral stimulation at T11 with a frequency of electrodes (A,B,C) located at midline (B) and laterally (A and C) 30 Hz applied during the swing phase increased walking speed to the spinal cord vs. stimulation at 2 levels (T11 + L1) with due to reduction of the stance phase duration. Additionally, the electrodes (T11-ABC) + (L1-ABC) was different. The amplitude stimulation induced an increase in the amplitude of movements of knee displacement and surface electromyographic (sEMG) in the hip joint and the lifting of the knee as well as foot activity of leg muscles were significantly higher with multi-site elevation. This was accompanied by an increase of sEMG activity stimulation at 2 levels than at one level (Gerasimenko et al., in BF and TA muscles (Figure 1B). Combined spatiotemporal 2015b). Our preliminary data reveal that use of the multielectrode stimulation at L1 and T11 applied to stance and swing phases, surface array can fine-tune the control of the locomotor behavior. correspondingly did not cause a change in the duration of the Here we introduce a new strategy of spinal neuromodulation stepping cycle and the phases of the step, however, it changed using the continuous stimulation of locomotor circuitry and the the kinematic characteristics of movements. In the stance phase, rhythmic stimulation of motor pools. the amplitude of movements in the hip joint increased. The amplitude of movements in the hip joint increased in the swing phase as well, but it also increased the lifting of the knee and COMBINED ACTIVATION OF LOCOMOTOR foot elevation (Figure 1B). Additional continuous stimulation CIRCUITRY AND MOTOR POOLS applied along the midline at T11 with frequency of 30 Hz facilitated the effect of rhythmic spatiotemporal stimulation. This study is based on the concept of differential modulation Thus, the data obtained shows that specifically configured multi- of neuronal networks projecting to specific interneurons that site scTS is able to selectively facilitate the activation of the coordinate the levels of recruitment of different combinations motor pools of the lower extremities and control their activity of motor pools throughout a step cycle. A novel approach to regulate the phases of the stepping cycle. of spatiotemporal spinal stimulation through rhythmic scTS coupled with continuous scTS to promote locomotor neural plasticity by activating regional spinal networks in a manner DISCUSSION similar to that observed during normal gait has been suggested. The general idea is the use of continuous spinal stimulation Our experience with multi-electrode epidural stimulation in to activate the locomotor networks in combination with mice, rats, and humans (Gerasimenko et al., 2008; Harkema rhythmic targeted activation of flexor and extensor motor pools et al., 2011; Angeli et al., 2014) indicate the potential to regain of leg muscles in different phases of step cycle to further overground load-bearing stepping, as well as voluntary control enable a stepping-like behavior. Recently it was shown that of lower limb movements, using non-invasive neuromodulatory spatiotemporal epidural stimulation of flexor and extensor motor strategies. The main issue is to determine the potential pools of left and right legs can facilitate swing or stance phases of scTS applied to different sites of the spinal cord can during the stepping cycle in chronic motor incomplete. Severely provide spatiotemporal specificity of the locomotor network paralyzed (unable to walk over-ground) individuals recovered in a way that facilitates phase-specific flexor-extensor motor over-ground walking with balance assist (Wagner et al., 2018). pool populations. Early, we demonstrated that transcutaneous Our experimental paradigm included rhythmic scTS with electrical stimulation of rostral and caudal areas of lumbosacral two cathodes at each T11 and L1 delivered 2.5 cm lateral to enlargement resulted in a selective topographical recruitment the midline of the spine, and continuous scTS at T11 or of proximal and distal leg muscles based on their threshold L1 applied along the midline between the spinous processes intensity, maximal slope of the recruitment curves, and (Figure 1A). ScTS was carried out with the electrical stimulator plateau point intensity and magnitude (Sayenko et al., 2018). BioStim-5 (Cosyma INC). Stimulation was delivered using 2.5- Our data are generally consistent with previous reports and cm-diameter round gel adhesive electrodes (Syrtenty Premium myotomal maps of the spinal cord and lumbosacral roots TENS) as cathodes, and two interconnected 5 × 9 cm2 self- (Stewart, 1992; Ivanenko et al., 2006). It is well-known that adhesive electrodes (Axelgaard, ValuTrode Cloth) placed over the lumbosacral enlargement not only contains motor neuron the iliac crests as anodes. The stimulation waveform consisted of pools projecting to proximal and distal leg muscles, but monophasic rectangular 1.0-ms pulses at a frequency of 15 and also encompasses neuronal networks controlling locomotion 30 Hz, each pulse filled with a carrier frequency of 5 kHz. and standing. We demonstrated that epidural stimulation In non-injured subjects (N = 6, 23.7 ± 2.3 years), stance and of the spinal rostral segments (L2) is more effective for swing phases were detected by sensors-gyroscopes. Initiation of inducing rhythmic movements, whereas stimulation of more hip extension was the trigger for activation of extensor pools caudal segments (S1–S2) allows for greater postural control (L1) during stance phase, whereas the initiation of hip flexion (Angeli et al., 2014). During combined locomotor-specific was triggered for activation of flexor motor pools (T11) during scTS over the T11 at 30 Hz, with the postural-specific scTS swing phase (Grishin et al., 2020). During walking on treadmill, over the L1 at 15 Hz, we observed the interplay of various stimulation applied unilaterally at L1 during the stance phase stimulation characteristics in generation of continuous and with a frequency of 15 Hz increased the amplitude of movements alternating weight-bearing, and facilitation of body-weight in the hip joint, and significantly increased sEMG activity of the transitions allowing effective stepping motions to be performed extensor muscles of the thigh and flexor muscles of the shank. (Sayenko et al., 2018). These results suggest that stimulation Frontiers in Human Neuroscience | www.frontiersin.org 2 January 2021 | Volume 14 | Article 622533 Moshonkina et al. Non-invasive Neuromodulation to Control Locomotion FIGURE 1 | Rhythmic and continues spinal cord transcutaneous stimulation paradigm. (A) Schematic localization of stimulation electrodes. Midline electrodes at T11 and L1 generate continuous stimulation at 30 Hz. Lateral electrode at T11 generate rhythmic stimulation at 30 Hz during swing phase and lateral electrode at L1 generate rhythmic stimulation at 15 Hz during stance phase. (B) Surface EMG activity of distal [tibialis ant. (TA) and medial gastrocnemius (MG)] and proximal [vastus lat. (VL) and biceps femoris (BF)] muscles as well as displacement of terminal poit (Fing) during walking on the treadmill without stimulation, with rhythmic stimulation of L1 at stance phase, with rhythmic stimulation of T11 at swing phase, and combined L1 and T11 stimulation applied to stance and swing phase, correspondingly. of multiple spinal sites related to postural and locomotor findings are compatible with work of others demonstrating the circuitries activation might be complementary and thus can be critical role of commissural neurons located in rostral lumbar a viable strategy to facilitate more effective stepping-like and segments supporting locomotor rhythm generation in response postural movements. to bulbospinal activation of locomotion in vitro neonatal rat Given the numerous observations of the different sources and spinal cord (Cowley et al., 2009). techniques to gain access to the neurons that generate locomotor Shik hypothesized the role of propriospinal system in rhythms, it is not surprising that modulation of these neurons, initiation of locomotion (Shik, 1997). It has been demonstrated commonly referred to as the locomotor central pattern generator, that microstimulation of stepping strip in the dorsolateral can be modulated electrically to different physiological states funiculus (DLF) elicited stepping movements in mesencephalic from multiple spinal segments, as well as multiple supraspinal cats (Kazennikov et al., 1983). According to these authors, sites. Although the rostral lumbar segments generate more robust neurons responding to stimulation of the stepping strip send pattern generation, neuromodulatory effects of stimulating at their axons into the ventrolateral funiculus (VLF) near the S1 was not expected to give the enhancement of the locomotor gray matter. It has been suggested that excited fibers of DLF rhythms derived from the fictive locomotor patterns observed can activate hindlimb stepping indirectly through DLF -VLF in the rodent neonatal preparation as reported in series of propriospinal loops (Shik, 1997). Our data are consistent with experiments by Lev-Tov and colleagues (Lev-Tov et al., 2000). suggestion about activation the locomotor-related neuronal For example, changes in the endogenous levels of cholinergic network through DLF. We have demonstrated that after chronic components in the sacral area with ascending projections via the local lesion of the dorsolateral column it was impossible to induce ventral funiculus (Finkel et al., 2014) and Alpha-1 adrenoceptor the locomotor activity in the cat by epidural stimulation of the agonists in this region have been observed to modulate the fictive spinal cord (Gerasimenko et al., 2002).Recently, neurotechnology locomotor output (Gabbay and Lev-Tov, 2004). that targets specific motor pools for restoration of walking We have also shown that the rostral lumbar segments in humans with spinal cord injury (SCI) was demonstrated. are the key controllers of hindlimb locomotor rhythmicity The authors delivered spatiotemporal epidural stimulation to in the adult spinal rat (Gerasimenko et al., 2019). We specific flexor/extensor motor pools during specific phases of the observed that the rats with spinal cord transections at T8 locomotor cycle. Using this targeted neurotechnology, paralyzed and S1 remained capable of generating coordinated hindlimb chronic motor incomplete individuals were able to walk over- locomotion when receiving epidural stimulation at L2. In ground with balance assist (Wagner et al., 2018). The technology contrast, minimal locomotion was observed when S1 stimulation we describe here is able to target these motor pools similarly to was delivered after spinal cord transections at T8 and L2. These restore walking without the need of surgical implantation. Frontiers in Human Neuroscience | www.frontiersin.org 3 January 2021 | Volume 14 | Article 622533 Moshonkina et al. Non-invasive Neuromodulation to Control Locomotion Here we demonstrated that non-invasive preferential AUTHOR CONTRIBUTIONS activation of spinal structures at specific segments is possible. According to the results presented on the Figure 1B, it The concept and design of this study were developed by YG. is clear that lateral stimulation at T11 engaged the neural TM, AG, IB, RG, RS, and AO were involved in data acquisition, circuits controlling flexion, whereas lateral stimulation at L1 analysis, and interpretation. The manuscript draft written by YG primarily recruited the circuits controlling extension during and TM was critically revised and approved by all authors. stepping. Spatiotemporal T11+L1 stimulation enhanced generation of motor patterns and enabled control of leg FUNDING movements. We suggest that this non-invasive technology could be effective for neuromuscular control of postural and This research was funded in part by National Institutes of locomotor function in post-stroke subjects and in individuals Health Grant 1R01 NS102920-01A1. The study was supported by with SCI. National Technology Initiative Foundation (project #400-230). REFERENCES instrument upravleniya lokomotsiyey cheloveka [A system for detecting the phases of the stepping cycle and stimulating the spinal cord as a tool for Angeli, C. A., Boakye, M., Morton, R. A., Vogt, J., Benton, K., Chen, Y., et al. controlling human locomotion]. Med. Tekh. 323, 10–14. (2018). Recovery of over-ground walking after chronic motor complete spinal Harkema, S., Gerasimenko, Y., Hodes, J., Burdick, J., Angeli, C., Chen, cord injury. N. Engl. J. Med. 379, 1244–1250. doi: 10.1056/NEJMoa1803588 Y., et al. (2011). Effect of epidural stimulation of the lumbosacral Angeli, C. A., Edgerton, V. R., Gerasimenko, Y. P., and Harkema, S. J. spinal cord on voluntary movement, standing, and assisted stepping (2014). Altering spinal cord excitability enables voluntary movements after motor complete paraplegia: a case study. Lancet 377, 1938–1947. after chronic complete paralysis in humans. Brain 137, 1394–1409. doi: 10.1016/S0140-6736(11)60547-3 doi: 10.1093/brain/awu038 Ivanenko, Y. P., Poppele, R. E., and Lacquaniti, F. (2006). Spinal cord maps of Cowley, K. C., Zaporozhets, E., Joundi, R. A., and Schmidt, B. J. (2009). spatiotemporal alpha-motoneuron activation in humans walking at different Contribution of commissural projections to bulbospinal activation of speeds. J. Neurophysiol. 95, 602–618. doi: 10.1152/jn.00767.2005 locomotion in the in vitro neonatal rat spinal cord. J. Neurophysiol. 101, Kazennikov, O. V., Shik, M. L., and Iakovleva, G. V. (1983). Stepping movements 1171–1178. doi: 10.1152/jn.91212.2008 caused by stimulation of the cat spinal cord dorsolateral funiculus. Biull. Eksp. Finkel, E., Etlin, A., Cherniak, M., Mor, Y., Lev-Tov, A., and Anglister, L. (2014). Biol. Med. 96, 8–10. doi: 10.1007/BF00839814 Neuroanatomical basis for cholinergic modulation of locomotor networks by Lev-Tov, A., Delvolvé, I., and Kremer, E. (2000). Sacrocaudal afferents induce sacral relay neurons with ascending lumbar projections. J. Comp. Neurol. 522, rhythmic efferent bursting in isolated spinal cords of neonatal rats. J. 3437–3455. doi: 10.1002/cne.23613 Neurophysiol. 83, 888–894. doi: 10.1152/jn.2000.83.2.888 Gabbay, H., and Lev-Tov, A. (2004). Alpha-1 adrenoceptor agonists generate a Sayenko, D. G., Rath, M., Ferguson, A. R., Burdick, J. W., Havton, L. A., “fast” NMDA receptor-independent motor rhythm in the neonatal rat spinal Edgerton, V. R., et al. (2018). Self-assisted standing enabled by non-invasive cord. J. Neurophysiol. 92, 997–1010. doi: 10.1152/jn.00205.2004 spinal stimulation after spinal cord injury. J. Neurotrauma 36, 1435–1450. Gerasimenko, Y., Gorodnichev, R., Moshonkina, T., Sayenko, D., Gad, doi: 10.1089/neu.2018.5956 P., and Reggie Edgerton, V. (2015a). Transcutaneous electrical spinal- Shik, M. (1997). Recognizing propriospinal and reticulospinal systems of initiation cord stimulation in humans. Ann. Phys. Rehabil. Med. 58, 225–231. of stepping. Motor Control. 1, 310–313. doi: 10.1123/mcj.1.4.310 doi: 10.1016/j.rehab.2015.05.003 Stewart, J. D. (1992). Electrophysiological mapping of the segmental anatomy of Gerasimenko, Y., Gorodnichev, R., Puhov, A., Moshonkina, T., Savochin, A., the muscles of the lower extremity. Muscle Nerve 15, 965–966. Selionov, V., et al. (2015b). Initiation and modulation of locomotor circuitry Wagner, F. B., Mignardot, J. B., Le Goff-Mignardot, C. G., Demesmaeker, output with multisite transcutaneous electrical stimulation of the spinal cord in R., Komi, S., et al. (2018). Targeted neurotechnology restores walking in noninjured humans. J. Neurophysiol. 113, 834–842. doi: 10.1152/jn.00609.2014 humans with spinal cord injury. Nature 563, 65–71. doi: 10.1038/s41586-018- Gerasimenko, Y., Preston, C., Zhong, H., Roy, R. R., Edgerton, V. R., and Shah, 0649-2 P. K. (2019). Rostral lumbar segments are the key controllers of hindlimb locomotor rhythmicity in the adult spinal rat. J. Neurophysiol. 122, 585–600. Conflict of Interest: YG researcher on the study team, hold shareholder interest in doi: 10.1152/jn.00810.2018 NeuroRecovery Technologies and Cosyma. He holds certain inventorship rights Gerasimenko, Y., Roy, R. R., and Edgerton, V. R. (2008). Epidural on intellectual property licensed by the regents of the University of California to stimulation: comparison of the spinal circuits that generate and control NeuroRecovery Technologies and its subsidiaries. VE holds shareholder interest locomotion in rats, cats and humans. Exp.Neurol. 209, 417–425. and certain inventor rights in NeuroRecovery Technologies and SpineX and doi: 10.1016/j.expneurol.2007.07.015 holds certain inventorship rights on intellectual property licensed by The Regents Gerasimenko, Y. P., Lu, D. C., Modaber, M., Zdunowski, S., Gad, P., Sayenko, D. of the University of California to NeuroRecovery Technologies and its subsidiaries. G., et al. (2015) Noninvasive reactivation of motor descending control after paralysis. J. Neurotrauma 32, 1968–1980. doi: 10.1089/neu.2015.4008 The remaining authors declare that the research was conducted in the absence of Gerasimenko, Y. P., Makarovskii, A. N., and Nikitin, O. A. (2002). any commercial or financial relationships that could be construed as a potential Control of locomotor activity in humans and animals in the absence conflict of interest. of supraspinal influences. Neurosci. Behav. Physiol. 32, 417–423. doi: 10.1023/A:1015836428932 Copyright © 2021 Moshonkina, Grishin, Bogacheva, Gorodnichev, Ovechkin, Siu, Gill, M. L., Grahn, P. J., Calvert, J. S., Linde, M. B., Lavrov, I. A., Strommen, Edgerton and Gerasimenko. This is an open-access article distributed under the J. A., et al. (2018). Neuromodulation of lumbosacral spinal networks enables terms of the Creative Commons Attribution License (CC BY). The use, distribution independent stepping after complete paraplegia. Nat. Med. 24, 1677–1682. or reproduction in other forums is permitted, provided the original author(s) and doi: 10.1038/s41591-018-0175-7 the copyright owner(s) are credited and that the original publication in this journal Grishin, A. A., Bobrova, E. V., Reshetnikova, V. V., et al. (2020). Sistema is cited, in accordance with accepted academic practice. No use, distribution or detektirovaniya faz shagatel’nogo tsikla i stimulyatsii spinnogo mozga kak reproduction is permitted which does not comply with these terms. Frontiers in Human Neuroscience | www.frontiersin.org 4 January 2021 | Volume 14 | Article 622533