Changes in intra- and interlimb reflexes from forelimb cutaneous afferents after staggered thoracic lateral hemisections during locomotion in cats

cutaneous reflexes, interlimb coordination, locomotion, spinal cord injury 12 Support or grant information: This work was supported by a grant from the National Institutes of Health: 13 R01 NS110550 to AF, IAR and BIP. AF is a Fonds de Recherche-Santé Quebec (FRQS) Senior Research 14 Scholar. JA and JH were supported by FRQS doctoral scholarships and ANM by a FRQS postdoctoral 15 scholarship. 16 17 Author contributions: SM, IAR, BIP, and AF contributed to conception and design of the study. SM, CL, 18 AM, JA, SY, RA and JH conducted the research. SM organized the database and performed the data and 19 statistical analysis. SM and AF wrote the first draft of the manuscript. All authors contributed to manuscript 20 revision, read, and approved the final version. 21 22 Acknowledgments: We thank Philippe Drapeau for providing data acquisition and analysis software, 23 developed in the Rossignol and Drew laboratories at the Université de Montréal. We thank the Biostatistics 24 department of the Centre de Recherche du Centre Hospitalier Universitaire de Sherbrooke for statistical 25 assistance. 26 27 Data availability statement: The raw data supporting the conclusions of this article will be made available 28 by the authors, without undue reservation. 29 30 Ethics statement: The animal study was reviewed and approved by the Animal Care Committee of the 31 Université de Sherbrooke. 32 33 Corresponding author: 34 Alain Frigon, PhD 35 Email: [email protected] 36 ORCID ID: 0000-0002-9259-2706 37 Competing statement: The authors declare no competing financial interests. 38 39 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint

40 In quadrupeds, such as cats, cutaneous afferents from the forepaw dorsum signal external perturbations and 41 send signals to spinal circuits to coordinate the activity in muscles of all four limbs. How these cutaneous 42 reflex pathways from forelimb afferents are reorganized after an incomplete spinal cord injury is not clear. 43 Using a staggered thoracic lateral hemisections paradigm, we investigated changes in intralimb and interlimb 44 reflex pathways by electrically stimulating the left and right superficial radial nerves in seven adult cats and 45 recording reflex responses in five forelimb and ten hindlimb muscles. After the first (right T5-T6) and second 46 (left T10-T11) hemisections, forelimb-hindlimb coordination was altered and weakened. After the second 47 hemisection, cats required balance assistance to perform quadrupedal locomotion. Short-, mid- and long-48 latency homonymous and crossed reflex responses in forelimb muscles and their phase modulation 49 remained largely unaffected after staggered hemisections. The occurrence of homolateral and diagonal mid- 50 and long-latency responses in hindlimb muscles evoked with left and right superficial radial nerve stimulation 51 was significantly reduced at the first time point after the first hemisection, but partially recovered at the 52 second time point with left superficial radial nerve stimulation. These responses were lost or reduced after 53 the second hemisection. When present, all reflex responses, including homolateral and diagonal, maintained 54 their phase-dependent modulation. Therefore, our results show a considerable loss in cutaneous reflex 55 transmission from cervical to lumbar levels after incomplete spinal cord injury, albeit with preservation of 56 phase modulation, likely affecting functional responses to external perturbations. 57 58 Key points: 59 • Cutaneous afferent inputs coordinate muscle activity in the four limbs during locomotion when the forepaw 60 dorsum contacts an obstacle. 61 • Thoracic spinal cord injury disrupts communication between spinal locomotor centers located at cervical 62 and lumbar levels, impairing balance and limb coordination. 63 • We investigated cutaneous reflexes from forelimb afferents during quadrupedal locomotion by electrically 64 stimulating the superficial radial nerve bilaterally, before and after staggered lateral thoracic hemisections in 65 cats. 66 • We showed a loss/reduction of mid- and long-latency homolateral and diagonal reflex responses in 67 hindlimb muscles early after the first hemisection that partially recovered with left superficial radial nerve 68 stimulation, before being reduced after the second hemisection. 69 • Targeting cutaneous reflex pathways from forelimb afferents projecting to the four limbs could help develop 70 therapeutic approaches aimed at restoring transmission in ascending and descending spinal pathways. 71 72 73 74 75 76 77 78 79 80 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint

81 locomotion, inputs from the skin provide information on the external environment, such as characteristics 82 of the terrain and obstacles encountered (Rossignol et al., 2006; Pearcey & Zehr, 2019a; Frigon et al., 83 2021). For instance, in cats and humans, cutaneous afferents play an important role in modifying limb 84 trajectory and maintaining balance during locomotion when the foot/hindpaw dorsum , innervated by the 85 superficial peroneal (SP) nerve, contacts an obstacle during the swing phase, termed the stumbling 86 corrective reaction (Forssberg et al., 1977; Prochazka et al., 1978; Forssberg, 1979; Duysens & Loeb, 1980; 87 Wand et al., 1980; Schillings et al., 1996; Van Wezel et al., 1997; Zehr et al., 1997; Quevedo et al., 2005b, 88 2005a). In quadrupeds, such as cats, stimulating the dorsum of the forepaw or the superficial radial ( SR) 89 nerve also elicits a stumbling corrective reaction that alters forelimb trajectory to move it away and over a 90 simulated obstacle (Miller et al., 1977; Matsukawa et al., 1982; Drew & Rossignol, 1985, 1987; Shimamura 91 et al., 1990; Fuwa et al., 1991; Hurteau et al., 2018; Mari et al., 2023). In cats and humans, electrically 92 stimulating the SP and SR nerves evoke short-, mid- and long-latency inhibitory and/or excitatory cutaneous 93 reflex responses in muscles of the four limbs that are modulated with task and phase during locomotion 94 (Haridas & Zehr, 2003; Mari et al., 2023). Mid- and long-latency responses are thought to involve supraspinal 95 contributions, but also polysynaptic spinal pathways (Fuwa et al., 1991; LaBella et al., 1992; Pijnappels et 96 al., 1998; Christensen et al., 1999; Hiersemenzel et al., 2000; Frigon & Rossignol, 2008; Hurteau & Frigon, 97 2018; Duysens 2024). 98 Spinal cord injury (SCI) disrupts ascending and descending pathways that communicate between the 99 brain and spinal networks controlling arm/forelimb and leg/hindlimb movements, including those activated 100 and modulated by cutaneous inputs. The disruption of these pathways and associated sensorimotor deficits 101 in balance and limb coordination vary depending on the severity and level of the lesion (Barbeau et al., 2002; 102 Edgerton et al., 2004; Frigon & Rossignol, 2006; Rossignol & Frigon, 2011). To investigate how SCI affects 103 cutaneous reflex pathways projecting to the four limbs, we recently used a staggered thoracic lateral 104 hemisections paradigm in cats and evoked cutaneous reflex responses by stimulating the SP nerve (Mari et 105 al., 2024). Staggered thoracic hemisections disrupt direct communication between the brain/cervical cord 106 and lumbar locomotor networks, disrupting fore-hind coordination and balance in cats and rats (Jane et al., 107 1964; Kato et al., 1984, 1985; Stelzner & Cullen, 1991; Courtine et al., 2008; Van Den Brand et al., 2012; 108 Cowley et al., 2015; Audet et al., 2023). We reported a significant reduction in the presence of mid- (19-34 109 ms) and long-latency (35-60 ms) responses in muscles of all four limbs, especially in the forelimbs (Mari et 110 al., 2024). In humans, a few studies have shown changes in interlimb reflexes (i.e. from the arms to the legs 111 or from the legs to the arms) after spinal cord injury, but only at rest (Calancie, 1991; Calancie et al., 1996, 112 2002; Butler et al., 2016). Thus, we do not know how reflexes evoked by forelimb afferents are reorganized 113 after SCI during locomotion. 114 Therefore, the purpose of the present study was to investigate reflex responses evoked by stimulating the 115 SR nerve before and after staggered thoracic hemisections in the same animal during treadmill locomotion, 116 extending our previous findings with responses evoked by hindlimb cutaneous afferents (Mari et al., 2024). 117 We first assessed SR-evoked reflex responses in the intact state and then following a first lateral 118 hemisection at mid-thoracic (T5-T6) on the right side. We then assessed how SR-evoked reflex responses 119 changed following a second lateral hemisection at left T10-T11. The main finding was a loss/reduction in 120 homolateral and diagonal responses in hindlimb muscles after staggered thoracic hemisections, which 121 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint correlated with weakened coordination between the fore- and hindlimbs and impaired balance during 122 quadrupedal locomotion (Audet et al., 2023). 123 124

125 Ethical approval 126 All procedures were approved by the Animal Care Committee of the Université de Sherbrooke (Protocol 127 442-18) in accordance with policies and directives of the Canadian Council on Animal Care. We obtained the 128 current data set from seven adult purpose-bred cats (> 1 year of age at the time of experimentation), 3 129 females and 4 males, weighing between 3.4 kg and 6.5 kg, purchased from Marshall BioResources (North 130 Rose, NY, USA). Before and after experiments, cats were housed and fed (weight-dependent metabolic diet 131 and water ad libitum) in a dedicated room within the animal care facility of the Faculty of Medicine and Health 132 Sciences at the Université de Sherbrooke. We followed the ARRIVE guidelines 2.0 for animal studies 133 (Grundy, 2015; Percie Du Sert et al., 2020). The investigators understand the ethical principles under which 134 the journal operates and our work complies with this animal ethics checklist. In order to maximize the 135 scientific output of each animal, they were used in other studies to investigate different scientific que stions, 136 some of which have been published (Lecomte et al., 2022, 2023; Merlet et al., 2022; Audet et al., 2023; Mari 137 et al., 2023, 2024). 138 139 General surgical procedures 140 All surgeries (implantation and spinal lesions) were performed under aseptic conditions with sterilized 141 equipment in an operating room. Prior to surgery, cats were sedated with an intramuscular (i.m.) injection of 142 butorphanol (0.4 mg/kg), acepromazine (0.1 mg/kg), and glycopyrrolate (0.01 mg/kg). We th en injected a 143 mixture (0.05 ml/kg, i.m.) of diazepam (0.25 mg/kg) and ketamine (2.0 mg/kg) in a 1:1 ratio five minutes later 144 for induction. We shaved the animal’s fur (back, stomach, fore- and hindlimbs) and cleaned the skin with 145 chlorhexidine soap. Cats were anesthetized with isoflurane (1.5-3%) and O2 delivered with a mask and then 146 with a flexible endotracheal tube. The depth of anesthesia was confirmed by applying pressure to a paw (to 147 detect limb withdrawal) and by assessing the size and reactivity of pupils. Isoflurane concentration was 148 adjusted throughout the surgery by monitoring cardiac and respiratory rates . Body temperature was 149 maintained constant (37 ± 0.5°C) using a water-filled heating pad placed under the animal, an infrared lamp 150 placed ~50 cm over it and a continuous infusion of lactated Ringers solution (3 ml/kg/h) through a catheter 151 placed in a cephalic vein. At the end of surgery, we injected subcutaneously an antibiotic (cefovecin, 8 152 mg/kg) and a fast-acting analgesic (buprenorphine, 0.01 mg/kg). We also taped a fentanyl (25 µg/h) patch to 153 the back of the animal 2-3 cm rostral to the base of the tail for prolonged analgesia, which we removed 4-5 154 days later. After surgery, cats were placed in an incubator and closely monitored until they regained 155 consciousness. We administered another dose of buprenorphine ~7 hours after surgery. At the end of 156 experiments, cats were anaesthetized with isoflurane (1.5–3.0%) and O2 before receiving a lethal dose of 157 pentobarbital (120 mg/kg) through the left or right cephalic vein. Cardiac arrest was confirmed using a 158 stethoscope to determine the death of the animal. Spinal cords were then harvested for histological analysis 159 (Lecomte et al., 2022, 2023; Audet et al., 2023; Mari et al., 2024). 160 161 162 163 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint Staggered lateral hemisections 164 After collecting data in the intact state, we performed a lateral hemisection between the 5th and 6th 165 thoracic vertebrae (T5-T6) on the right side of the spinal cord. Before surgery, we sedated the cat with an 166 intramuscular injection of a cocktail containing butorphanol (0.4 mg/kg), acepromazine (0.1 mg/kg) and 167 glycopyrrolate (0.01 mg/kg) and inducted with another intramuscular injection (0.05 ml/kg) of ketamine (2.0 168 mg/kg) and diazepam (0.25 mg/kg) in a 1:1 ratio. We shaved the fur overlying the back and t he skin was 169 cleaned with chlorhexidine soap. The cat was then anesthetized with isoflurane (1.5 -3%) and O2 using a 170 mask for a minimum of 5 minutes and then intubated with a flexible endotracheal tube. Isoflurane 171 concentration was confirmed and adjusted throughout the surgery by monitoring cardiac and respiratory 172 rates, by applying pressure to the paw to detect limb withdrawal and by assessing muscle tone. Once the 173 animal was deeply anesthetized, an incision of the skin over and between the 5th and 6th thoracic vertebrae 174 (T5-T6) was made and after carefully setting aside muscle and connective tissue, a small laminectomy of the 175 corresponding dorsal bone was performed. Lidocaine (xylocaine, 2%) was applied topically followed by 2 -3 176 intraspinal injections on the right side of the cord. We then sectioned the spinal cord laterally from the midline 177 to the right using surgical scissors. We placed hemostatic material (Spongostan) within the gap before 178 sewing back muscles and skin in anatomical layers. In the days following hemisection, voluntary bodily 179 functions were carefully monitored. The bladder and large intestine were manually expressed if needed. 180 Once data were collected following the first hemisection (9-13 weeks), we performed a second lateral 181 hemisection between the 10th and 11th thoracic vertebrae (T10-T11) on the left side of the spinal cord using 182 the same surgical procedures and post-operative care described above. 183 184 Electromyography and nerve stimulation 185 To record the electrical activity of muscles (EMG, electromyography), we directed pairs of Teflon-186 insulated multistrain fine wires (AS633; Cooner Wire Co., Chatsworth, CA, USA) subcutaneously from two 187 head-mounted 34-pin connectors (Omnetics Connector Corp., Minneapolis, MN, USA). Two wires, stripped 188 of 1–2 mm of insulation, were sewn into the belly of selected forelimb/hindlimb muscles for bipolar 189 recordings. The head-mounted connectors were fixed to the skull using dental acrylic and four to six screws. 190 We verified electrode placement during surgery by stimulating each muscle through the appropriate head 191 connector channel to assess the biomechanically desired muscle contraction. During experiments, EMG 192 signals were pre-amplified (×10, custom-made system), bandpass filtered (30–1,000 Hz) and amplified (100–193 5,000×) using a 16-channel amplifier (model 3500; AM Systems, Sequim, WA, USA). EMG data were 194 digitized (5,000 Hz) with a National Instruments (Austin, TX, USA) card (NI 6032E), acquired with custom-195 made acquisition software and stored on computer. Five forelimb muscles were implanted bilaterally: biceps 196 brachii (BB, elbow and shoulder flexor), extensor carpi ulnaris (ECU, wrist dorsiflexor), flexor carpi ulnaris 197 (FCU, wrist plantarflexor), latissimus dorsi (LD, shoulder retractor), and the long head of the triceps brachii 198 (TRI, elbow and shoulder extensor). Ten hindlimb muscles were implanted bilaterally: biceps femoris anterior 199 (BFA, hip extensor), biceps femoris posterior (BFP, hip extensor and knee flexor), iliopsoas (IP, hip flexor), 200 lateral gastrocnemius (LG, ankle plantarflexor and knee flexor), medial gastrocnemius (MG, ankle 201 plantarflexor and knee flexor), sartorius anterior (SRT, hip flexor and knee extensor), semitendinosus (ST, 202 knee flexor and hip extensor), soleus (SOL, ankle plantarflexor), tibialis anterior (TA, ankle dorsiflexor), and 203 vastus lateralis (VL, knee extensor). 204 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint For bipolar nerve stimulation, pairs of Teflon-insulated multistrain fine wires (AS633; Cooner Wire Co., 205 Chatsworth, CA, USA) were passed through a silicon tubing. A horizontal slit was made in the tubing and 206 wires within the tubing were stripped of their insulation. The ends protruding through the cuff were knotted to 207 hold the wires in place and glued. The ends of the wires away from the cuff were inserted into four-pin 208 connectors (Hirose or Samtec) and fixed to the skull using dental acrylic. Cuff electrodes were directed 209 subcutaneously from head-mounted connectors to the left and right SR nerves at the wrist which are purely 210 cutaneous at these levels. 211 212 Experimental design 213 We collected EMG and kinematic data before and at different time points after staggered hemisections 214 during quadrupedal locomotion at the cat’s preferred treadmill speed (0.3-0.5 m/s). Cats KA and KI stepped 215 at 0.3 and 0.5 m/s, respectively, while the other five cats stepped at 0.4 m/s. The treadmill consisted of two 216 independently controlled belts 130 cm long and 30 cm wide (Bertec) with a Plexiglas separator (130 cm long, 217 7 cm high, and 1.3 cm wide) placed between the two belts to prevent limbs impeding each other. In the 218 intact, preoperative state, cats were trained for 2-3 weeks in a progressive manner, first for a few steps and 219 then for several consecutive minutes, using food and affection as rewards. Once cats could perform 3 -4 220 consecutive minutes, we started the experiments. During experiments, we delivered trains of electrical stimuli 221 consisting of three 0.2 ms pulses at 300 Hz using a Grass (West Warwick, RI, USA) S88 stimulator. At the 222 start of the experiment, we determined the motor threshold, defined as the minimal intensity that elicited a 223 small motor response in an ipsilateral flexor muscle (e.g., ST or TA) during the swing phase. We then set 224 stimulation intensity at 1.2 times the motor threshold to mainly activate large diameter Aβ cutaneous 225 afferents. A locomotor trial lasted 4-5 min and consisted of ∼60 stimuli delivered pseudo-randomly every 2–4 226 cycles. Stimuli were delivered at specific points of the stimulated forelimb (left or right) movement: mid-227 stance, the transition from stance-to-swing, mid-swing and the transition from swing-to-stance. At the start of 228 the experiment, stimulation delays were determined using real-time EMG to detect the onset of an extensor 229 burst in relation to stance and swing phases. We then set delays in relation to this extensor EMG so that 230 stimuli were delivered at the four desired time points. The timing of the stimuli was assessed during off-line 231 analysis and stimuli that did not fall in the desired phases were excluded. We characterized responses in 232 muscles of the stimulated forelimb (homonymous), the opposite forelimb (crossed), the hindlimb on the same 233 side (homolateral) and the diagonal hindlimb (diagonal). Figure 1A describes the timeline of data collection 234 in all seven cats at two time points after the first hemisection (H1T1 and H1T2 in 7 cats) and/or at 1-2 time 235 points after the second hemisection (H2T1 in 2 cats and H2T2 in 6 cats). Some cats only have one time 236 point after the second hemisection because they took longer to recover quadrupedal locomotion. No data 237 were collected for cat KI after the second hemisection due to technical issues with the impl ants. 238 239 Histology 240 After confirming euthanasia (i.e., no cardiac and respiratory functions), we harvested an approximately 2 241 cm long section of the spinal cord centered on the lesions. Segments of the dissected spinal cord were then 242 placed in a 25 ml 4% paraformaldehyde (PFA) solution (in 0.1 M phosphate‐buffered saline (PBS), 4°C). 243 After 5 days, we placed the spinal cord in a new PBS solution containing 30% sucrose for 72 h at 4°C, then 244 froze it in isopentane at -50°C for cryoprotection. The spinal cord was then sliced in 50 µm coronal sections 245 using a cryostat (Leica CM1860, Leica BioSystems Inc., Concord, ON, Canada) and mounted on gelatinized-246 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint coated slides. The slides were dried overnight and then stained with a 1% cresyl violet acetate solution for 247 12 min. We washed the slides for 3 min in distilled water before being dehydrated in successive baths of 248 ethanol (50%, 70% and 100%, 5 min each) and transferring them in xylene for 5 min. Dibutylphthalate 249 polystyrene xylene was next used to mount and dry the spinal cord slides before being scanned by a 250 Nanozoomer 2.0-RS (Hamamatsu Corp., Bridgewater, NJ, USA). We then performed qualitative and 251 quantitative analyses to estimate lesion extent using ImageJ by selecting the slide with the greatest 252 identifiable damaged area. Using the scarring tissue stained with cresyl violet acetate, we estimated lesion 253 extent by dividing the lesion area by the total area of the selected slice and expressed it as percentage. 254 Lesion extent estimations for individual cats after the first and second spinal lesions are shown in Figure 1B. 255 They ranged from 40.7% to 66.4% (49.2 ± 8.9%) and 33.5% to 53.7% (46.0 ± 7.6%) for the first and second 256 hemisections, respectively. 257 258 Reflex analysis 259 We described the reflex analysis in several of our publications (Hurteau et al., 2017, 2018; Hurteau & 260 Frigon, 2018; Merlet et al., 2020, 2021; Mari et al., 2023), and recently illustrated it in detail in (Mari et al., 261 2024). Briefly, for all locomotor sessions, EMG signals were low-pass filtered (250 Hz) to facilitate the 262 visualization of the EMG activity envelope. We first defined locomotor cycles from successive burst onsets of 263 an extensor from the stimulated forelimb, then separated them as stimulated (i.e., cycles with stimulation) or 264 control (i.e., cycles without stimulation) cycles. Sections where the cat stepped irregularly were removed 265 from analysis based on EMG and video data. Stimulated cycles were then sorted and divided into 4 266 subphases based on stance onset of the stimulated forelimb: swing-to-stance, mid-stance, stance-to-swing 267 and mid-swing. Control (C̅) cycles were averaged and rectified to provide a baseline locomotor EMG, an 268 indication of the excitability level of the motor pool at stimulation. We averaged the stimulated (S̅) cycles and 269 time normalized C̅ to S̅ cycle durations and superimposed them. To determine response onsets and offsets, 270 defined as prominent positive or negative deflections away from C̅, we set windows using previous studies 271 as guidelines (Duysens & Stein, 1978; Duysens & Loeb, 1980; Pratt et al., 1991; Loeb, 1993; Hurteau et al., 272 2017, 2018; Hurteau & Frigon, 2018; Mari et al., 2023) with 97.5% confidence intervals. We termed short-273 latency (7–18 ms; SLR) excitatory and inhibitory responses as P1 and N1 responses, respectively, based on 274 the terminology introduced by (Duysens & Loeb, 1980). Responses in the crossed, homolateral, and 275 diagonal limbs that had an onset ≤18 ms were classified as P1 or N1, as the minimal latency for spino -bulbo-276 spinal reflexes in the cat is 16-18 ms (Shimamura & Livingston, 1962; Shimamura et al., 1990). Mid-latency 277 (19–34 ms; MLR) excitatory and inhibitory responses were termed P2 and N2, respectively. Long-latency 278 (35–60 ms; LLR) excitatory and inhibitory responses were termed P3 and N3, respectively. The EMG of 279 reflex responses S̅ was then integrated and subtracted from the integrated C̅ in the same time window to 280 provide a net reflex value. This net reflex value was then divided by the integrated C̅ value to evaluate reflex 281 responses. This division helps identify if changes in reflex responses across the cycle are independent of 282 changes in C̅ activity (Matthews, 1986; Frigon & Rossignol, 2007, 2008, 2009; Hurteau et al., 2017, 2018; 283 Hurteau & Frigon, 2018; Mari et al., 2023, 2024). 284 285 Statistical analysis 286 We performed statistical tests with IBM SPSS Statistics V26 (IBM Corp., Armonk, NY, USA). We 287 quantified reflex responses in five forelimb muscles (BB, ECU, FCU, LD, and TRI) and in ten hindlimb 288 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint muscles (BFA, BFP, IP, LG, MG, SRT, SOL, ST, TA, and VL) when stimulation was delivered to the left or 289 right SR. To evaluate whether homonymous, crossed, homolateral and diagonal responses were modulated 290 by phase, we performed a one factor (phase) ANOVA on all responses (P1, P2, P3, N1, N2 and N3) in eac h 291 cat and state/time point. Because we have several responses within a given phase, we considered all 292 responses during a locomotor session as a population. In our statistical analysis, we used mixed models to 293 deal with incomplete data sets. For instance, reflex responses are sometimes absent after spinal lesions. 294 Response occurrence probabilities, defined as the fraction of evoked responses obtained out of all cats for 295 pooled SLR, MLR and LLR from the different states/time points were compared using a generalized linear 296 mixed model (GLMM) with a binomial distribution and a logit link (mixed logistic regression) in all four limbs. 297 The GLMM analysis was performed using state/time point as a fixed factor. We incorporated random 298 intercepts at two distinct levels to consider the hierarchical relationships present in our dataset. A random 299 intercept on individual cats at the upper level captured variability across cats. A random intercept on muscle 300 nested within cat at a lower level, acknowledging that the same muscle response data were repeatedly 301 measured within each cat to help us account for any correlation or non-independence of observations within 302 the same cat-muscle pair. Statistical significance for all tests was set at p < 0.05. 303 304

305 Recovery of quadrupedal locomotion and changes in fore-hind coordination after staggered 306 hemisections 307 We recently described changes in the quadrupedal locomotor pattern after staggered thoracic lateral 308 hemisections (right T5-T6 followed by left T10-T11), including six cats of the present study (Audet et al., 309 2023). Briefly, we showed that cats spontaneously recovered quadrupedal locomotion following both 310 hemisections but required balance assistance after the second one. After the first and second hemisections, 311 the coordination between the forelimbs and hindlimbs became weaker and displayed 2:1 patterns, where the 312 forelimbs performed two cycles within one hindlimb cycle. 313 After the first hemisection (H1), all seven cats regained quadrupedal locomotion on the treadmill within 314 one to two weeks. We were able to conduct reflex sessions for several consecutive minutes at the first 315 (H1T1) and second (H1T2) time points. After the second hemisection (H2), the six cats tested recovered 316 quadrupedal locomotion within two to five weeks. However, they required mediolateral balance assistance 317 during reflex sessions, which was provided by an experimenter holding the tail of the animal but without 318 providing weight support. As stated in the Methods, some cats only have one time point after the second 319 hemisection due to the longer recovery of quadrupedal locomotion and their ability to maintain it for reflex 320 testing. Only two cats, KA and JA, participated in reflex sessions at the first time point (H2T1), i.e., 321 approximately two weeks after the second hemisection. All six cats performed reflex sess ions eight weeks 322 later at the second time point (H2T2). 323 324 Cutaneous reflexes evoked by stimulating the superficial radial nerve before and after staggered 325 hemisections 326 To assess the reorganization of cutaneous reflex pathways from forelimb afferents after SCI, we 327 stimulated the left and right SR nerves before and after the first and second hemisections and recorded 328 reflex responses during quadrupedal treadmill locomotion in muscles of the four limbs. Figures 2-5 illustrate 329 examples from representative cats for homonymous, crossed, homolateral and diagonal responses in 330 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint selected muscles at four phases of the cycle (swing-to-stance transition, mid-stance, stance-to-swing 331 transition and mid-swing). We observed several changes in reflex responses after the first and second 332 hemisections, but for the responses shown in Figures 2-5, we only highlight the most noticeable 333 observations. Tables 1-3 provide the full details of reflex responses before and after staggered hemisections 334 for individual cats. For each state/time point, filled areas represent evoked responses and are optimized for 335 display according to the strongest response obtained in one of the four phases. If an area is not filled, it 336 means that the stimulated EMG did not deviate sufficiently from the baseline EMG to be defined as a 337 response. The scale is optimized per state/time point and differs across state/time points. This is to show the 338 pattern of evoked responses and its phase-dependent modulation at a given state/time point, as in our 339 recent study (Mari et al., 2024) 340 341 Homonymous responses in forelimb muscles. We stimulated the left and right SR nerves and recorded 342 homonymous responses in muscles of the left and right forelimbs, respectively, before and after staggered 343 hemisections (Fig. 2 and Table 1). We illustrate homonymous reflex responses in three muscles (ECU, TRI 344 and BB) bilaterally in representative cats. The ECU and TRI muscles are mostly ac tive during stance while 345 BB is active during swing and/or at the stance-to-swing transition. We observed that the burst profiles of the 346 three selected forelimb muscles remained similar across states (before and after hemisections) and time 347 points during the locomotor cycle. 348 In the left ECU (Fig. 2A, left panel), in the intact state, we observed homonymous P1/P2 responses at 349 mid-swing, and weak N1 responses in the other three phases followed by small P3 responses at mid-stance 350 and stance-to-swing. After the first hemisection, at H1T1 and H1T2, P1/P2 responses remained at mid-351 swing, but relatively strong P2 and/or P3 responses appeared at swing-to-stance and mid-stance. After the 352 second hemisection, at H2T1, P2/P3 responses were reduced at mid-swing and absent in the other phases. 353 At H2T2, P1/P2 responses recovered at mid-swing and returned at swing-to-stance and mid-stance. In the 354 right ECU (Fig. 2A, right panel), in the intact state, we observed P1 responses in all phases followed by N2 355 at swing-to-stance and mid-stance and P2 at mid-swing. At H1T1 and H1T2, P1 responses were prominent 356 in all phases followed by prominent P2 and/or P3 responses, except for stance-to-swing with only P1 at 357 H1T2. At H2T1, P1/P2 responses remained at swing-to-stance and mid-swing, with N1 at mid-stance and no 358 responses at stance-to-swing. At H2T2, we observed P1 responses at mid-swing and stance-to-swing, and 359 N1 responses at mid-stance, with no P2/P3 responses in all phases. 360 In the left TRI (Fig. 2B, left panel), we observed homonymous P1 and P2 responses at swing-to-stance 361 and P2 responses at mid-swing and mid-stance. At H1T1, P1/P2 and P2 responses remained at swing-to-362 stance and mid-stance, respectively, but P2 responses were lost at mid-swing. At H1T2, we N1 responses 363 followed by P3 responses appeared at swing-to-stance and mid-stance. At H2T2, we only observed P2 364 responses in all phases. In the right TRI (Fig. 2B, right panel), we observed P2/P3 responses in all phases 365 that followed N1 responses at swing-to-stance and mid-stance. At H1T1, N1 responses were lost or 366 weakened at swing-to-stance and mid-stance, respectively, but P2 responses remained. N2 responses 367 appeared at stance-to-swing and P1 responses at mid-swing, while P2/P3 remained in both phases. At 368 H2T1, a weak N1/P2 responses returned or remained stance-to-swing and mid-stance but excitatory 369 responses were lost or reduced at stance-to-swing and mid-swing. At H2T2, N1 responses were prominent 370 at swing-to-stance and mid-stance, but P2 responses were lost or reduced. At stance-to-swing and mid-371 swing prominent P2/P3 returned. 372 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint In the left BB (Fig. 2C, left panel), the most noticeable response was a prolonged N1 response at mid -373 swing in the intact state. At H1T1, this N1 response was lost and a prominent P3 response appeared. At 374 H1T2, the N1 response returned followed by a P3 response. These N1/P3 responses remained at H2T1 and 375 H2T2. The other noticeable change was at swing-to-stance, with the appearance of P1 responses at H1T2, 376 which remained at H2T1 and H2T2. At H2T1 and H2T2, P2/P3 responses also appeared. In the right BB 377 (Fig. 2C, right panel), in the intact state, the most noticeable response at mid-swing was an N2 response 378 that followed a brief P1 response. At H1T1, this N2 response was lost while the P1 response became 379 prominent. The N2 response returned at H1T2 following a brief P1. At H2T1 and H2T2, the N2 res ponse was 380 prominent and was followed by prominent P3 responses. In the other phases, the only noticeable changes 381 were the appearance of a prominent P1 responses at H1T1 at swing-to-stance, which disappeared at H1T2 382 before returning at H2T2. 383 To summarize, even though thoracic hemisections did not directly affect the pathways transmitting 384 cutaneous afferent inputs from the SR nerve to ipsilateral (homonymous) motor circuits in the cervical cord, 385 we observed several small reflex changes in forelimb muscles. 386 387 Crossed responses in forelimb muscles. We stimulated the left and right SR nerves and recorded crossed 388 responses in muscles of the right and left forelimbs, respectively, before and after staggered hemisections in 389 the four phases (Fig. 3 and Table 2). We defined the four phases according to the stimulated limb, but it is 390 important to consider the phase of the contralateral limb where the responses were recorded. 391 In the left ECU (Fig. 3A, left panel), we observed crossed P2 responses in all four phases in the intact 392 state, with the largest at mid-stance of the stimulated right forelimb. After the first and second hemisections, 393 P2 responses remained in all phases. In the right ECU (Fig. 3A, right panel), in the intact state, we observed 394 crossed P2 responses in all phases, with the largest at stance-to-swing and mid-swing of the stimulated left 395 forelimb. After the first hemisection, P2 responses were reduced at stance-to-swing and mid-swing and P1 396 responses appeared at swing-to-stance. P2 responses recovered at H1T2 and were observed in all phases. 397 P2 responses persisted after the second hemisection at H2T2, except at mid-stance where they were 398 absent. 399 In the left TRI (Fig. 3B, left panel), in the intact state, we observed crossed N1 responses during mid-400 swing followed by P3 responses, whereas at mid-stance and stance-to-swing, we observed P2/P3 401 responses. After the first hemisection, at H1T1 and H1T2, the response pattern was largely maintained, but 402 P2/P3 responses appeared at swing-to-stance at H1T1 and N2/P3 responses at H1T2. After the second 403 hemisection, at H2T2, we observed N2 responses at swing-to-stance and mid-swing, maintained P2/P3 404 responses at stance-to-swing and a weak P2/P3 response at mid-stance. In the right TRI (Fig. 3B, right 405 panel), in the intact state, we observed prolonged crossed N1 responses at mid-swing and swing-to-stance 406 and a P2 response at mid-stance. After the first and second hemisections, we observed prolonged N1 407 responses at swing-to-stance followed by P3 responses, but only at H1T2 and H2T1. At mid-stance, we 408 observed P2/P3 responses at H1T1 and then only P3 responses at H1T2, H2T1 and H2T2. At stance-to-409 swing, we observed P2 responses at H1T1, P2/P3 at H1T2, P3 at H2T1 and P2/P3 at H2T2. At mid-swing, 410 N1 or N2 responses were followed by P3 responses at all time points. 411 In the left BB (Fig. 3C, left panel), in the intact state, we observed crossed P2/P3 responses in all four 412 phases that peaked at mid-stance of the stimulated right forelimb, which was followed by N3 responses at 413 mid-stance only. After the first and second hemisection, P2/P3 responses were maintained in all phases 414 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint except at stance-to-swing where they were present at H1T1 and H1T2 but disappeared after the second 415 hemisection. In the right BB (Fig. 3C, right panel), in the intact state, we observed crossed P2/P3 responses 416 in all phases that peaked at mid-swing of the stimulated left forelimb. After the first and second hemisection, 417 P2/P3 responses were maintained in all phases with the strongest responses observed at mid-swing and 418 swing-to-stance. 419 To summarize, although we noted some changes in crossed reflex responses in forelimb muscles after 420 staggered hemisections, they were mostly similar to those observed in the intact state. 421 Table 1 summarizes homonymous and crossed reflex response patterns observed in all 5 forelimb 422 muscles bilaterally in 6 and 5 cats for the left and right SR nerve stimulations, respectively, before and after 423 staggered hemisections. Overall, the reflex response patterns that we observed in forelimb muscles 424 remained generally similar after hemisections, as did their phase-dependent modulation. 425 426 Table 1. Homonymous and crossed reflex responses before and after staggered hemisections. 427 Left SR nerve stimulation Homonymous responses in left forelimb Crossed responses in right forelimb Cat State BB ECU FCU LD TRI BB ECU FCU LD TRI AR Intact H1T1 H1T2 P1* P2* P1* P1* P2 P1* N2/P2* P1* P2 P1* P2* N1/P1* P2* P3* N1/P1* P2* P3* N1/P1* P2* P2* P2* P2* P1* P2* P1 P2* P1* P2* N1 x x P2 P3* P2* P2* P1* P2* P2* N1* N1* N1 P3* N1* P3* N1* N1* HO Intact H1T1 H1T2 H2T2 P2* P1* P2* P1* P2 P1* P2 N1/P1* N1/P1* P1* N2* N1/P1* N1* P2* P3* P1* P3* N1/P1* P3 N1/P1* P3* P1* P1* P1* P1* P2 P1* P2* P1* P1* P2 P2 P2* P2 P2* P2* P2* P2* P2* P3* P2* P2 P2 P3* P2* x N2 P3 P3 P3* P2 N2* N2* N2/P2* P3* JA Intact H1T1 H1T2 H2T1 H2T2 P1* N2* P1* P3* P1* N2 P3* P1* N2/P2* P3* N1/P1* P2* P3* N1/P1* P2* N1/P1* N1/P1* P1* N1/P1* N1* P3* N1/P1* N1/P1* P3* N1/P1* P3* N1/P1* P2* P3* P1* P2* P1* P1* P2* P1* P1* P1* P2* P1* P2* P3* P1* P2* P3* P1* P2* P3* P1* P2* P2* P2* P2 P2 P2* x P2* P3* P2* P3 P2 P3* P2* P3 P2* N1* P2* P2* P3* P2* P2* N1 P3* P2* P3* N1/P1* P2* P2* P3 N1* P2* N1* P2* P3* N1* P2* P3* N1* P3 N1* P2 P3 KA Intact H1T1 H1T2 H2T1 H2T2 P1* N2* P1* P2* P1 P2* P1 P2* P1* P2* N1/P1* P2* P3 N1/P1* P2* N1/P1* P2* P3* N1/P1* N1/P1* P2* P3 N1/P1* P2* N1/P1* P2* N1/P1* P2* P3* N1* N1/P1* P2* P1* N1/P1* P2* N1/P1* P2 - - P1* P2* P2* N1* P2* P3* N1 P3 P1* P2* P3* P2* - P2* P2* P2* P2* x P2* P3* P2* P2* P2* x P2 P2* P2* P2 x P2* - - x x P2* P3* x x KI Intact H1T1 H1T2 N1/P1* P1* P2* P1* P2* N1/P1* N1/P1* P2* P3* N1/P1* P2* P3* N1/P1* P2* P3* N1/P1* P3* N1/P1* P3* P1* P2* P1* P2* N1/P1* P2* N1/P1* P2* P3* N1/P1* P2* N1/P1* P2* P3* P2* P2 P2* P2* N3* P2* N3* P2* N3 N2/P2* N3* P2* N3* P2* N3* P2 P1* P2* P2* P2* P2* N3* P2 N3* TO Intact H1T1 H1T2 H2T2 P2* P2* P1* P1* N1 P2* P1* P2* P3* N1* P2* P1* N2* N1/P1* P2* P3* x N1/P1* P2* P3* N1* P3* P1* P2* P1* P1* P2* P1* P1* P2* N1/P1* P2 N1* P2 P3 P2* P2* x - x P2* P1* P2 P2 P2* P2 N3* P2* N3 P2* N3* P2* N3* N1* P2* N1* P2 P3* P2* P3* P2 P3* P2* N3* P2* P2 N3* x Right SR nerve stimulation Homonymous responses in right forelimb Crossed responses in left forelimb Cat State BB ECU FCU LD TRI BB ECU FCU LD TRI AR Intact H1T1 H1T2 H2T2 P1* P2* N1/P1* P2* P1* P2* P3* P1* P2* N1/P1* P2* P3* N1/P1* P2* P3* N1/P1* P2* N1/P1* P3* N1/P1* P2* P3* N1/P1* P2* N1/P1* P2* N1* P2* N1* P2* N1* P2* P3 N1 P2* P3 N1* P2* N1/P1* P2* P3* N1/P1* P2* P3* N1/P1* P2* P3* N1/P1* P2* N1* P2* P2* P2* P2* P3* N1* P2* P2* P2* P3* P2* P2* P2* P2* P2* N1* N1* N1* N1* N1* P2* P3* N1* P2* P3* N1* P2* P3* N1* P2* P3* GR Intact H1T1 H1T2 H2T2 N1/P1* N2* N1/P1* N1/P1* N1/P1* N1/P1* P2* N1/P1* P3* N1/P1* P3 N1* P3* N1/P1* P2* N1/P1* P3* N1/P1* P3* N1* P3* P1 P1* P2* P1 P1* P2 N1/P1* P2* N1/P1* P2* P3* N1/P1* P3* N1/P1* P3* P2* P2 P2 P2* P2* P2* P2 P2 P2* P2* P2 N2/P2* P2* P2* - - N2* P2* P2 N2/P2* JA Intact H1T1 H1T2 H2T1 H2T2 P1* N2* P1* P3* P1* N2* P3* P1 N2/P2* P3* P1* N2/P2* P3* P1* N1/P1* P2* P3* N1/P1* N1/P1* N1/P1* P2* P3* P1* P2* N1/P1* P2* N1/P1* N1/P1* N1/P1* P3* P1* P2* P1* P2* P1* P1* P2* P1* P2* P1* N2* P3* N1/P1* P2* N1/P1* P2* P1* P2* N1/P1* P2* P3* P2* N3* P2* N3* P2* N3 P2* N3* P2* N3 P2 P2* P2* P2* P2* P2* x P2* P2* P2* N1* P3* N1 P3* P2* N1 P3* N1 P3* P2 x N2* P2* P2* P3* KA Intact H1T1 H1T2 H2T1 H2T2 x x P1* P1* P2* P1* P2* P1* N2* P1* P2* P1* P2* N1/P1* P2* N1/P1* P2* N1/P1* N1/P1* P3* N1/P1* P3 N1/P1* P2* N1/P1* P2* P1* P2* P1* P1* x x P1* N2* P1* P2* P1* P2* P1* P2* P1* P2* x P2* x P2* P2* P2* x x P2* P2* x x x P2* P2* x N1/P1* P3* x - - x x x x x TO Intact H1T1 H1T2 H2T2 P1* N2/P2* P1* - P1* N1/P1* P2 N1* P2 N1/P1* P2* N1* P2* N1/P1* P2 P3* x P1* x N1/P1* P2* P1* P1* P2* P1* P2* N1* P2 N1 P2 P3* N1* P2* N1* P2 P3* x P2* P2* P1* P2* P3* P2* P2* P2* N1* P3* P2* N3* x P2* P2* P3* N1* P3 x x N1* P3* N1 P3 x x N1 P3 428 Homolateral responses in hindlimb muscles. Out of 10 hindlimb muscles, we show examples from three 429 selected muscles, SOL, VL and SRT, bilaterally in representative cats. The SOL and VL muscles are mostly 430 active during stance while SRT is active during swing and/or at the stance-to-swing transition. We define the 431 phases relative to the stimulated forelimb. However, after the first and/or second hemisections, cats 432 frequently performed two forelimb cycles within one hindlimb cycle (i.e. 2:1 fore-hind patterns). This means, 433 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint for example, that stimulation at mid-stance of one forelimb can elicit reflex response at different phases for 434 the hindlimb (see Discussion). 435 In the left SOL (Fig. 4A, left panel), in the intact state, we observed homolateral P2 responses at swing-436 to-stance and mid-swing of the stimulated forelimb that were followed by N3 responses at swing-to-stance 437 and mid-stance. After the first hemisection, we observed no responses at H1T1, but some recovery at H1T2, 438 with P2/N3 responses at swing-to-stance and mid-stance and P2 responses at mid-swing. Homolateral 439 responses were lost after the second hemisection. In the right SOL (Fig. 4A, right panel), homolateral 440 response patterns were similar in the intact state, with P2/N3 at swing-to-stance and mid-swing and N3 at 441 mid-stance. After the first and second hemisections, we observed no homolateral responses. 442 In the left VL (Fig. 4B, left panel), in the intact state, we observed homolateral P2 responses at mid-swing 443 of the stimulated forelimb and N3 responses at mid-stance. After the first and second hemisections, these 444 responses disappeared, with the exception of a weak N3 response at H1T2 swing-to-stance. In the right VL 445 (Fig. 4B, right panel), we observed homolateral P2 responses at swing-to-stance and mid-swing of the 446 stimulated forelimb. After the first and second hemisections, we observed no responses. 447 In the left SRT (Fig. 4C, left panel), in the intact state, we observed homolateral P2/P3 responses at 448 swing-to-stance, mid-stance and mid-swing of the stimulated left forelimb. After the first hemisection, P2/P3 449 responses remained at swing-to-stance and mid-stance at H1T1 but were visibly reduced at H1T2. After the 450 second hemisection, we observed no responses. In the right SRT (Fig. 4C, right panel), in the intact state, 451 we observed homolateral P2/P3 responses at swing-to-stance, mid-stance and stance-to-swing of the 452 stimulated right forelimb. After the first hemisection, P2/P3 responses remained in these phases at H1T1 but 453 were visibly reduced or disappeared at H1T2. After the second hemisection, we observed no responses at 454 H2T1 but P2/P3 responses returned at mid-stance and stance-to-swing at H2T2. 455 Table 2 summarizes homolateral reflex response patterns in all 10 hindlimb muscles bilaterally in 6 and 5 456 cats for the left and right SR nerve stimulations, respectively, before and after staggered hemisections. 457 Response patterns mostly consisted of P2 responses followed by N3 responses in extensors (BFA, LG, MG 458 and SOL), and P2/P3 responses in flexors (IP, SRT, ST and TA). Although the phase-dependent modulation 459 of responses generally remained after staggered hemisections, when responses were present, we observed 460 a loss in response occurrence in most muscles after the first and/or second hemisections. However , some 461 homolateral responses returned at H1T2 on the left side. 462 463 Table 2. Homolateral reflex responses before and after staggered hemisections. 464 Homolateral responses in left hindlimb (left SR nerve stimulation) Cat State BFA BFP IP LG MG SOL SRT ST TA VL AR Intact H1T1 H1T2 N2/P2* N3* x P2* N2/P2* x - P2* x x P2* N3* x x P2* N3* x P2* P2* N2* x x P2* x x - - - - x x P2* x x HO Intact H1T1 H1T2 H2T2 N3* x P1* P2* x x x - P2 x x x N3* x N3* x P2 N3* x N3* x P2 N3* x N3* x x x x x x - x x - x x P2* P2 N3 x N3* x JA Intact H1T1 H1T2 H2T1 H2T2 P2* - - x - P1* x - - - P2* P3* x P2* N3* x P2* N3* P2 P2 N3 x x P2* x P2* x x P2* N3* x P2* N3 x x P2* P2* P2* x x x x - - - x x x x x P2* x - x - KA Intact H1T1 H1T2 H2T1 H2T2 x x P2 N3* - - x - x - - - - - - - x x P2 N3* x x P2* N3 P2* N3* P2* N3* - - P2* N3 P2* N3* P2* N3 - - x x x x x x x P2* x x x x x x x x x P2* - - KI Intact H1T1 H1T2 x P2* N3* P2* x P2* N3* P2* N3* - - - N2* P2* N3* P2* N3* N2* P2* N3* P2* N3* N2* P2* N3* P2* N3* x P2* x x P1* P1* x x x x P2* N3* P2* N3* TO Intact H1T1 H1T2 H2T2 P2 N3* x P2* x P3* x - - N2/P2* x P2* x P2* N3* x N3* x P2* N3* x P2 N3 x P2* N3* x P2* N3* x P2* x x x x x x - x x x x N3* x x x .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint Homolateral responses in right hindlimb (right SR nerve stimulation) Cat State BFA BFP IP LG MG SOL SRT ST TA VL AR Intact H1T1 H1T2 H2T2 P2* N3 x x x x x x x P2* x P2* x P2 N3* N2* x x P2* N3* x x x P2* N3* N2* N2* x P2* P3* P2* x x x x x x P2* P3* x x - P2* - x x GR Intact H1T1 H1T2 H2T2 N2* - x - x x P2* x - - - - - x x x - - - - N2/P2* x P2 x P2* x P2* x - - x x P2* P2 P2 P2* N2/P2* x P2* - JA Intact H1T1 H1T2 H2T1 H2T2 x x x x - x x x x x x - - - - x x x x x x x x x x P2* N3* N2* N2 P3* x x P2* P2* P2* x P2* x x x x x P2* x x x - - x x x x KA Intact H1T1 H1T2 H2T1 H2T2 x x x - - x x x - - - - - - - x x x x x x P1* x - - x x x - - x N1/P1* P2* P3* N1/P1* P2* P3* P2* P1* P2* x x x - P2* x x x x x - x x - - TO Intact H1T1 H1T2 H2T2 P2* N3* x x x x x x x P2* N3* x x x P2* x x x P2* N3* x x x P2* N3* x x x N1* P3 x x x x x x x P2* x x x P2* x x x 465 Diagonal responses in hindlimb muscles. In the left SOL (Fig. 5A, left panel), in the intact state, we observed 466 diagonal N2 responses followed by P3 responses at stance-to-swing and mid-swing of the stimulated right 467 forelimb, with P2/N3 responses at mid-stance. After the first and second hemisections, we observed no 468 diagonal responses. In the right SOL (Fig. 5A, right panel), in the intact state, we observed diagonal N2 469 responses at mid-stance, stance-to-swing and mid-swing of the stimulated left forelimb followed by P3 470 responses at mid-stance and stance-to-swing. After the first hemisection, at H1T1 and H1T2, N2 responses 471 remained at in these three phases but P3 responses disappeared. After the second hemisection, we 472 observed no diagonal responses. 473 In the left VL (Fig. 5B, left panel), in the intact state, we observed diagonal P2 responses at swing-to-474 stance and mid-stance of the stimulated right forelimb, with N2 responses at mid-swing. After the first and 475 second hemisections, we observed no diagonal responses. In the right VL (Fig. 5B, right panel), in the intact 476 state, we observed diagonal P2 responses only at mid-stance of the stimulated left forelimb. After the first 477 and second hemisections, we observed no diagonal responses. 478 In the left SRT (Fig. 5C, left panel), in the intact state, we observed diagonal P2 responses in all phases 479 except at stance-to-swing of the stimulated right forelimb. After the first and second hemisections, we 480 observed no diagonal responses. In the right SRT (Fig. 5C, right panel), in the intact state, we observed 481 diagonal P2 responses in all phases except at swing-to-stance of the stimulated left forelimb. After the first 482 hemisection, we observed no responses at H1T1 but P2/P3 reappeared at H1T2 at swing-to-stance and mid-483 stance. After the second hemisection, we observed no diagonal responses. 484 Table 3 summarizes diagonal reflex response patterns in all 10 hindlimb muscles bilaterally in 6 and 5 485 cats for the left and right SR nerve stimulations, respectively, before and after staggered hemisections. 486 Response patterns consisted mostly of N2 followed by P3 responses in extensors (LG and SOL). In flexors 487 (BFP, IP, SRT, ST and TA), we observed P2/P3 responses. Similar to homolateral responses, we observed 488 a loss in response occurrence in most muscles after the first and/or second hemisections , although the 489 phase-dependent modulation remained if responses were present. We observed some return of diagonal 490 responses at the second time point after the first hemisection (H1T2) in right hindlimb muscles. 491 492 Table 3. Diagonal reflex responses before and after staggered hemisections. 493 Diagonal responses in right hindlimb (left SR nerve stimulation) Cat State BFA BFP IP LG MG SOL SRT ST TA VL AR Intact H1T1 H1T2 N2 x x P2* P3* x x P2* x x N2* P1* x N2 x x N2* x x P2* x x P2* x x P2* x x N2* - x .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint HO Intact H1T1 H1T2 H2T2 P2 x P1* P2* N1* P2* x x x P2* x P2* x N2* x x x P2* x P2* N3* x N2* x x N3* x x x x x x x x P2 x x x N2* x x x JA Intact H1T1 H1T2 H2T1 H2T2 x x P2 x - P3* x P2 x x P2* - - - - N2 P3* x P2 x x P3* x P2 x x N2* P3* N2 N2* x x P2* x P2* x x P3* x P2* x x x x x x - - x x x x KA Intact H1T1 H1T2 H2T1 H2T2 x x x - - x - x - - - - - - - P2* N3* x N2/P2* x x x x x - - N2* x x - - P2* x P2* x x x x P2* - x N2/P2* x x x x - x x - - KI Intact H1T1 H1T2 N2* P3 P1* P2* x x P1* P2* x - - - N2 P3 N2* x N2* P3 x P2* N2* P3* N2* x P2 P2* P2* - - - x x x N2 x N2* TO Intact H1T1 H1T2 H2T2 N2* P3* x x x x x x x P2* x P2* x P2* N3* x x x P2* N3* x P2 x x x N2* x P2* x x x x x x x P2 x x x P2* x x x Diagonal responses in left hindlimb (right SR nerve stimulation) Cat State BFA BFP IP LG MG SOL SRT ST TA VL AR Intact H1T1 H1T2 H2T2 x x x x P2* x - - P2* x x x N2* P3* N2* N2 x x N2* N2* x N2* P3* N2* N2* x P2* P2* P3* x x - - - P2* x x - P2* N2* N2* x GR Intact H1T1 H1T2 H2T2 N2/P2* x N2* x x x x x - - - - N2* x x x - - - - N2* x N2* x P2* x x x x x x x - x x x P2* x x - JA Intact H1T1 H1T2 H2T1 H2T2 - - - x - - x x - - P2* x x x x P3* x x x x P3* x x x x N2* P3* N2* N2* x N2* P2* x P3 x x x x x - - P3* x x x x x x - x - KA Intact H1T1 H1T2 H2T1 H2T2 x x x - - x x x - - - - - - - x x x x x x x x - - x x x - - x x x x x x x x x x x x x P2* P2* x x x - - TO Intact H1T1 H1T2 H2T2 P2 x x x P2* - - - P2* x x x N2/P2* P3 x x x P2* x x x N2/P2* N3/P3* x x x P2* x x x x x x - P2* x x x N2/P2* x x x 494 Staggered hemisections reduce the occurrence of mid- and long-latency responses in hindlimb 495 muscles 496 After complete or incomplete spinal lesions in cats, mid- and long-latency responses in hindlimb muscles 497 are generally reduced or abolished (Fuwa et al., 1991; LaBella et al., 1992; Frigon & Rossignol, 2008; Frigon 498 et al., 2009; Hurteau et al., 2017; Mari et al., 2024). Here, we investigated the probability of evoking reflex 499 responses in all four limbs before and after staggered hemisections by evaluating the distribution of SLRs 500 (N1/P1), MLRs (N2/P2) and LLRs (N3/P3). We did this by calculating the fraction of the total number of SLRs 501 or MLRs/LLRs separately, on recorded muscles for each limb across cats. For homolateral and diagonal 502 responses in the left and right hindlimbs, we only evaluated MLRs/LLRs because of infrequent occurrence of 503 SLRs. We excluded the H2T1 time point as only two cats were recorded. 504 We found no significant difference in response occurrence probability for left and right homonymous 505 SLRs (left, p = .441, GLMM; right, p = .925, GLMM) and MLRs/LLRs (left, p = .496, GLMM; right, p = .401, 506 GLMM) across states/time points (Fig. 6A). Similarly, the probability of evoking crossed SLRs (left, p = .085, 507 GLMM; right, p = .304, GLMM) and MLRs/LLRs (left, p = .095, GLMM; right, p = .225, GLMM) in left and right 508 forelimb muscles did not differ across states/time points (Fig. 6B). In contrast, we found a significant main 509 effect of state/time point on homolateral MLRs/LLRs occurrence probability in the left (p = 1.00 × 10-6, 510 GLMM) and right (p = 1.50 × 10-5, GLMM) hindlimbs (Fig. 6C). In the left hindlimb, homolateral responses 511 were 8.3 (p = 7.00 × 10-6) and 22.7 (p = 1.25 × 10-4) times more likely to be evoked in the intact state 512 compared to H1T1 and H2T2, respectively. They were also 5.2 (p = 3.73 × 10-4) and 14.4 (p = .001) times 513 more likely to be evoked at H1T2 compared to H1T1 and H2T2, respectively. In the right hindlimb, 514 homolateral responses were 10.0 (p = 1.00 × 10-4), 6.6 (p = 4.16 × 10-4) and 14.7 (p = 6.50 × 10-5) times 515 more likely to be evoked in the intact state compared to H1T1, H1T2 and H2T2, respectively. We found a 516 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint significant main effect of state/time point on diagonal MLRs/LLRs occurrence probability in the left (p = 7.12 517 × 10-7, GLMM) and right (p = 1.06 × 10-9, GLMM) hindlimbs (Fig. 6D). In the left hindlimb, diagonal 518 responses were 18.2 (p = 8.00 × 10-6), 9.3 (p = 1.46 × 10-4) and 58.8 (p = 7.00 × 10-6) times more likely to be 519 evoked in the intact state compared to H1T1, H1T2 and H2T2, respectively. They were also 6.3 (p = .036) 520 times more likely to be evoked at H1T2 compared to H2T2. In the right hindlimb, diagonal responses were 521 28.6 (p = 4.84 × 10-9), 7.2 (p = 1.00 × 10-5) and 100.0 (p = 3.00 × 10-5) times more likely to be evoked in the 522 intact state compared to H1T1, H1T2 and H2T2, respectively. They were also 4.0 (p = .009) and 13.8 (p = 523 .015) times more likely to be evoked at H1T2 compared to H1T1 and H2T2, respectively. 524 Therefore, overall, after the first and second hemisections, the probability of evoking homolateral and 525 diagonal MLRs/LLRs in hindlimb muscles with both left and right SR nerve stimulations was always lower 526 compared to the intact state. In addition, with stimulation of the left SR nerve, the occurrence probability of 527 evoking homolateral and diagonal MLRs/LLRS recovered after the first hemisection (at H1T2 compared to 528 H1T1), before seeing a drastic decrease after the second hemisection. 529 530

531 In the present study, we showed changes in reflex responses evoked by electrically stimulating 532 cutaneous afferents of the forepaw dorsum (SR nerve stimulation) during locomotion after staggered 533 hemisections, extending our recent study with reflex responses evoked by stimulating hindlimb cutaneous 534 afferents in the same animals and lesion paradigm (Mari et al., 2024). The main result of the present study 535 was a noticeable loss/reduction of mid- and long-latency homolateral and diagonal responses in hindlimb 536 muscles, after both the first and second hemisections. However, after the first hemisection, we observed a 537 partial recovery of these responses evoked by the left SR (contralesional) from the early to the late time 538 point, which then disappeared after the second hemisection. These changes in homolateral and diagonal 539 responses correlated with altered and weakened fore-hind coordination and impaired balance during 540 quadrupedal locomotion, as we recently reported (Audet et al., 2023), and also discussed in relation with 541 changes in reflex responses evoked by hindlimb cutaneous afferents (Mari et al., 2024). In the following 542 sections, we discuss changes in reflex responses evoked by forelimb cutaneous afferents, the putative 543 mechanisms and pathways involved and the functional significance of our results for locomotor 544 control/recovery after SCI. 545 546 Cutaneous reflexes from forelimb afferents before and after staggered lateral hemisections 547 In our staggered lateral hemisections paradigm, the first hemisection unilaterally disrupted direct 548 descending motor pathways from the brain and cervical cord to the lumbar cord, as well as ascending 549 pathways that carry somatosensory information from the hindlimbs to the cervical cord and then the brain. 550 The second hemisection on the left side then disrupted these direct pathways contralateral to the first 551 hemisection, generating a bilateral disruption. Lesion extent varied between animals (see Fig. 1B) and likely 552 contributed to inter-individual variability. As expected, short- (N1/P1), mid- (N2/P2) and/or long-latency 553 (N3/P3) responses in muscles of the homonymous and crossed forelimb remained after the first and second 554 hemisections with both left and right SR nerve stimulations (Fig. 6). These responses also retained their 555 phase-dependent modulation (Table 1). It is unlikely that the spinal lesions at T5-T6 and then T10-T11 556 damaged forelimb motoneuronal pools, which are located at C5-T2 spinal segments in cats (Sterling & 557 Kuypers, 1967; Fritz et al., 1986a, 1986b; Hörner & Kümmel, 1993). The SR nerve originates from the 558 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint brachial plexus, formed by the ventral branches of the last three cervical nerves and the first thoracic nerve , 559 and enters spinal segments C7-T1 (Ansón et al., 2013; Mencalha et al., 2014; Hakkı Nur et al., 2020). Thus, 560 the sensorimotor circuits responsible for homonymous and crossed reflex responses in forelimb muscles are 561 largely preserved after thoracic SCIs. 562 Although the pattern of reflex responses in homonymous and crossed forelimb muscles remained 563 generally similar after thoracic hemisections, we did observe several small changes ( Figs. 2 and 3). These 564 changes can involve supralesional circuit reorganization (Krupa et al., 2022) and/or the loss of inhibitory or 565 excitatory ascending pathways from lumbosacral segments (Sterling & Kuypers, 1967; Giovanelli Barilari & 566 Kuypers, 1969; Molenaar & Kuypers, 1978; English, 1985; Alstermark et al., 1987b; Dutton et al., 2006; 567 Reed et al., 2006). Long ascending propriospinal neurons make direct contact with motoneurons in the most 568 caudal cervical segments in rats (Reed et al., 2006, 2009; Brockett et al., 2013) and C3-C4 segments in cats 569 (English, 1985; Alstermark et al., 1987b, 2007), with the vast majority being excitatory (Miller et al., 1973; 570 Juvin et al., 2005; Reed et al., 2006; Brockett et al., 2013). These long projecting neurons play a role during 571 locomotion as their silencing can disrupt left-right coordination of the hindlimbs and forelimbs (Pocratsky et 572 al., 2020; Shepard et al., 2021). However, fore-hind coordination appears to mainly depend on short 573 propriospinal neurons because blocking their activity at thoracic levels, without disrupting transmission in 574 long propriospinal neurons, leads to independent rhythmic activity at cervical and lumbar levels in the in vitro 575 neonatal rat preparation (Ballion et al., 2001; Juvin et al., 2005). Moreover, in both fictive and real locomotion 576 studies in decerebrate cats with a complete thoracic transection performed at T10 -T13, reflex responses 577 evoked with SR stimulation were preserved and rhythmically modulated in forelimb motoneurons and 578 muscles (Shimamura et al., 1990; Fuwa et al., 1991; Seki & Yamaguchi, 1997). Thus, the neural circuits 579 modulating forelimb reflexes evoked by SR nerve afferents are mainly located at cervical and upper thoracic 580 segments. This can include cervical spinal locomotor CPGs (Yamaguchi, 1992, 2004; Kinoshita & 581 Yamaguchi, 2001) that interact with primary afferent inputs (Prochazka et al., 2002; Frigon & Rossignol, 582 2006; Frigon et al., 2021; Lalonde & Bui, 2021) and supraspinal signals (Shimamura & Livingston, 1963; 583 Shimamura et al., 1984; Brink et al., 1985; Alstermark et al., 1987a; Fleshman et al., 1988; Fuwa et al., 584 1991; Bretzner & Drew, 2005; Ni et al., 2014; Bazley et al., 2014; Duysens, 2024). However, ascending 585 pathways from the lumbosacral cord likely participate in part of the reflex response patterns and modulation, 586 particularly in the intact state. 587 In hindlimb muscles, the occurrence of homolateral and diagonal mid- and/or long-latency reflex 588 responses was considerably reduced after the first and second hemisections, with both left and right SR 589 nerve stimulations (Figs. 4-6). These responses are thus highly dependent on the integrity of descending 590 pathways running in the thoracic spinal cord. Thoracic lesions directly disrupt descending pathways from the 591 brain and cervical cord that generate homolateral and diagonal reflex responses (Miller et al., 1977; Haridas 592 & Zehr, 2003; Hurteau et al., 2018; Mari et al., 2023). Miller and colleagues (1977) proposed that long 593 descending propriospinal neurons, with axons mainly traveling in the ventrolateral spinal cord (Flynn et al., 594 2011), constitute the main pathways responsible for homolateral and diagonal responses, although a 595 contribution from short propriospinal neurons is also likely and cannot be excluded . Some of these pathways 596 project ipsilaterally and/or contralaterally (Côté et al., 2018; Laliberte et al., 2019). Additionally, thoracic 597 lesions disrupt brainstem pathways that release monoamines throughout the spinal cord and enhance 598 neuronal excitability (Noga et al., 2009, 2011). Studies have shown that serotonin facilitates hindlimb 599 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint cutaneous reflexes in spinal-transected cats (Barbeau & Rossignol, 1990) and rats (Maj et al., 1976; Nozaki 600 et al., 1977). 601 We observed a partial return of the occurrence of homolateral and diagonal reflex responses evoked by 602 the left SR nerve after the first hemisection, from the first to the second time point, consistent with ongoing 603 neuroplasticity (Fig. 6). Thus, over the course of 6-7 weeks, changes occurred in spinal circuits to facilitate 604 reflex transmission from cervical to lumbar levels. Numerous studies have shown that the spinal n eural 605 circuitry below an incomplete SCI undergoes substantial reorganization (Barriere et al., 2008; Frigon et al., 606 2009; Barrière et al., 2010; Martinez et al., 2011, 2012; Gossard et al., 2015), including after staggered 607 lateral hemisections (Jane et al., 1964; Kato et al., 1984, 1985; Stelzner & Cullen, 1991; Courtine et al., 608 2008; Van Den Brand et al., 2012; Cowley et al., 2015; Audet et al., 2023). After the first hemisection, the left 609 side of the spinal cord remains relatively intact, and plasticity can occur in descending pathways, which can 610 be beneficial or detrimental (Beauparlant et al., 2013; Fink & Cafferty, 2016; Shepard et al., 2023). This can 611 involve establishing new connections and/or reactivating latent ones, as well as reinforcing spared 612 descending propriospinal and supraspinal pathways through spontaneous collateral sprouting that target 613 both lesioned and unlesioned fibers (Edgerton et al., 2004; Cai et al., 2006; Maier & Schwab, 2006; Fenrich 614 & Rose, 2009; Basaldella et al., 2015; Higgin et al., 2020; Zavvarian et al., 2020). The return of neuronal 615 excitability at lumbar levels could have also facilitated responses evoked by descending pathways. 616 The recovery of homolateral and diagonal responses evoked by stimulating the left SR from the early to 617 the late time point after the first hemisection was lost after the second hemisection, which disrupted 618 descending pathways on the left side. This staggered thoracic lateral hemisections paradigm reveals 619

to the formation of new short connections to enable reflex transmission from cervical to lumbar 620 levels. Previous studies have shown that new propriospinal relays can form spontaneously to reroute the 621 supraspinal influences onto lumbar circuits (Bareyre et al., 2004; Courtine et al., 2008; Cowley et al., 2008, 622 2010; Zaporozhets et al., 2011; May et al., 2017). However, these new connections appear limited in 623 supporting cutaneous reflex transmission from cervical to lumbar segments. This is consistent with studies 624 using the in vitro neonatal rat brain stem-spinal cord preparation that required neurochemical excitation of 625 thoracic propriospinal neurons to generate hindlimb locomotion with brainstem electrical stimulation after 626 staggered thoracic lateral hemisection (Cowley et al., 2008; Zaporozhets et al., 2011). Figure 7 627 schematically presents a scenario explaining changes in reflex responses to the four limbs after staggered 628 hemisections with left and right SR nerve stimulation. 629 630 Functional considerations 631 Electrically stimulating the SR nerve mimics a mechanical contact of the forepaw dorsum, eliciting a 632 functional response consistent with a stumbling corrective or preventive reaction during the forelimb swing 633 and stance phases, respectively, as shown in intact and decerebrate cats (Miller et al., 1977; Matsukawa et 634 al., 1982; Drew & Rossignol, 1985, 1987; Shimamura et al., 1990; Fuwa et al., 1991; Hurteau et al., 2018; 635 Mari et al., 2023). In the present study and other studies, stimulating the SR nerve evoked reflex responses 636 in muscles of the four limbs in intact cats, consistent with a whole body functional response to an external 637 perturbation to alter limb trajectory and ensure dynamic balance (Haridas & Zehr, 2003; Hurteau et al., 2018; 638 Pearcey & Zehr, 2019a; Merlet et al., 2022; Mari et al., 2023). Our results indicate that functional responses 639 to SR nerve stimulation would have been appropriate in forelimb muscles to alter the trajectory of the 640 ipsilateral forelimb and reinforce support in the contralateral forelimb after staggered thoracic hemisections, 641 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint as reflex responses in forelimb muscles and their phase modulation were largely preserved . However, the 642 loss/reduction of homolateral and diagonal responses suggest an inability to coordinate the hindlimbs, which 643 likely would have affected the whole-body response to a real perturbation. 644 The loss/reduction of homolateral and diagonal responses also reflects a disruption of neural 645 communication between the brain/cervical cord and the lumbar cord, which cannot be compensated by new 646 connections, as discussed above. This affects sensorimotor functions that depend on this communication. 647 Indeed, after staggered thoracic hemisections, we recently showed that fore-hind coordination was altered 648 and weakened (Audet et al. 2023). Moreover, after the second hemisection, balance assistance was 649 required during treadmill locomotion. Although we can only speculate, the loss of cutaneous reflex 650 transmission from cervical to lumbar levels could have contributed to weakened interlimb coordination and 651 impaired postural control. 652 In humans, cutaneous afferents from the arm contribute to locomotor and/or postural control. For 653 instance, cutaneous stimulation at the wrist at the swing-to-stance transition during walking results in 654 increased ankle dorsiflexion (Haridas & Zehr, 2003). This could slow forward progression by reducing 655 propulsion from ankle plantarflexors, thereby minimizing a possible collision with an object with the 656 outstretched arm. Leg responses evoked from hand cutaneous nerves were facilitated with vision removed 657 during walking, but were restored when participants were asked to lightly touch a stable reference (Forero & 658 Misiaszek, 2015) that reinforced stability (Dickstein & Laufer, 2004; Forero & Misiaszek, 2013). Thus, 659 forelimb cutaneous afferents can assist with balance during walking. In cats, more weight is distributed to the 660 forelimbs for stability and propulsion after spinal lesions (Rossignol et al., 1999) and maintaining proper 661 sensorimotor interactions in cervical sensorimotor circuits is important, as shown in the present study . 662 Another factor to consider after single or staggered thoracic lateral hemisections is the emergence of 663 spatiotemporal left-right asymmetries between the hindlimbs (Kato et al., 1985; Martinez et al., 2011, 2013; 664 Audet et al., 2023), which can lead to walking instability (Dambreville et al., 2015; Huijben et al., 2018). 665 Recent studies using split-belt locomotion have shown that inducing left-right asymmetries reduced hindlimb 666 reflex responses in some muscles with SR or SP nerve stimulation (Hurteau et al., 2017; Hurteau & Frigon, 667 2018; Mari et al., 2023). Split-belt locomotion induces left-right asymmetries in sensory feedback from the 668 limbs, with increased loading for the limbs on the slow belt (Frigon et al., 2015; Park et al., 2019). The 669 hemisections also induce left-right asymmetries as the contralesional hindlimb spends a greater proportion of 670 the cycle supporting bodyweight (Martinez et al., 2011, 2012; Audet et al., 2023). This increased loading 671 generates asymmetric limb sensory feedback, which could play a role in modulating interlimb reflexes. 672 673 Methodological limitations 674 A limitation of the present study was that balance assistance was required after the second hemisection 675 to conduct reflex testing during locomotion, with an experimenter holding the tail for medio -lateral stability, 676 but without providing weight support, as discussed in our recent study (Mari et al., 2024). Without this 677 balance assistance, cats stumbled every few steps. Balance assistance likely facilitated reflex responses 678 and their phase-dependent modulation but without it, we could not have conducted reflex testing and 679 compared it to the other states (i.e. intact and following the first hemisection). Another limitation is that we 680 pooled reflex responses in hindlimb muscles according to the phase of the stimulated forelimb. As stated, 681 cats often performed 2:1 fore-hind patterns after the first and second hemisections, where a forelimb 682 performed two cycles within a hindlimb cycle. Because our stimulation is based on the forelimb cycle, it 683 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint means that after spinal lesions, some stimuli occurred in different phases of the hindlimb cycle. We 684 acknowledge this as a limitation but separating all responses so that both forelimb and hindlimb phases 685 corresponded would have required many more stimuli and several more minutes to complete a session. 686 These studies are difficult to perform in spinal cord-injured cats. Thus, for a given forelimb phase of 687 stimulation, we pooled all responses evoked in a hindlimb muscle, irrespective of the phase of the hindlimb. 688 This could have affected average responses, particularly if inhibitory and excitatory responses occurred at 689 the same latency in different hindlimb phases after hemisections. However, we investigated individual 690 responses and observed that most homolateral and diagonal responses were simply lost after hemisections. 691 692 Concluding remarks and clinical perspectives 693 In conclusion, the main finding of this study was the loss/reduction of homolateral and diagonal reflex 694 responses in hindlimb muscles from forelimb cutaneous afferents after staggered thoracic hemisections, 695 which correlated with weakened coordination between the fore- and hindlimbs and impaired balance. How 696 do our findings in the cat model relate to people with spinal cord injury? Our paradigm offers a substrate to 697 test the efficacy of therapeutic approaches (e.g. spinal cord stimulation, pharmacology) to restore neural 698 communication between the spinal locomotor networks controlling the arms/forelimbs and legs/hindlimbs. 699 Exploiting cervicolumbar connections, with rhythmic arm movements and/or primary afferent stimulation, 700 could facilitate locomotor rehabilitation after SCI. For instance, in cats (Harnie et al., 2024) and humans 701 (Frigon et al., 2004; Zehr et al., 2004; Hiraoka & Iwata, 2006; Loadman & Zehr, 2007; Javan & Zehr, 2008; 702 Dragert & Zehr, 2009; Hundza & Zehr, 2009; De Ruiter et al., 2010; Hundza et al., 2012; Massaad et al., 703 2014; Pearcey & Zehr, 2019b), rhythmic arm movements contribute and/or modulate reflexes in leg/hindlimb 704 muscles. In rats, quadrupedal locomotor training improved the quality of fore-hind coordination after a 705 thoracic spinal cord hemisection, with recovery correlating with an increased number of propriospinal 706 neurons just above and below the injury site (Shah et al., 2013). In humans with incomplete SCI, 707 coordinating rhythmic arm movements simultaneously with the legs facilitates corticospinal and/or 708 corticofugal drive (Zhou et al., 2017) to leg muscles and modulates cervicolumbar connectivity (Zhou et al., 709 2018). Combined arm and leg movements improved locomotor function. Applying transcutaneous spinal cord 710 stimulation at the cervical levels can also modulate the activity of lumbar networks (Barss et al., 2020; Parhizi 711 et al., 2021). Therefore, these results highlight the importance of activating forelimb sensory feedback, 712 engaging cervical locomotor networks and descending propriospinal pathways in the recovery of leg/hindlimb 713 locomotor movements after SCI. 714 715 716

717 718 Alstermark B, Isa T, Pettersson LG & Sasaki S (2007). The C3-C4 propriospinal system in the cat and monkey: 719 A spinal pre-motoneuronal centre for voluntary motor control. Acta Physiol 189, 123–140. 720 Alstermark B, Lundberg A, Pinter M & Sasaki S (1987a). Long C3-C 5 propriospinal neurones in the cat. Brain 721 Res 404, 382–388. 722 Alstermark B, Pinter M & Tantisira B (1987b). Branching and termination of C3-C 4 propriospinal neurones 723 in the cervical spinal cord of the cat. Neurosci Lett 74, 291–296. 724 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint Ansón A, Gil F, Laredo FG, Soler M, Belda E, Ayala MD & Agut A (2013). Correlative ultrasound anatomy of 725 the feline brachial plexus and major nerves of the thoracic limb. Vet Radiol Ultrasound 54, 185–193. 726 Audet J, Yassine S, Lecomte CG, Mari S, Soucy F, Morency C, Merlet AN, Harnie J, Beaulieu C, Gendron L, 727 Rybak IA, Prilutsky BI & Frigon A (2023). Spinal sensorimotor circuits play a prominent role in 728 hindlimb locomotor recovery after staggered thoracic lateral hemisections but cannot restore 729 posture and interlimb coordination during quadrupedal locomotion in adult cats. eneuro 10, 0191–730 23. 731 Ballion B, Morin D & Viala D (2001). Forelimb locomotor generators and quadrupedal locomotion in the 732 neonatal rat. Eur J Neurosci 14, 1727–1738. 733 Barbeau H, Fung J, Leroux A & Ladouceur M (2002). A review of the adaptability and recovery of locomotion 734 after spinal cord injury. In Progress in Brain Research, pp. 9–25. Elsevier Science. 735 Barbeau H & Rossignol S (1990). The effects of serotonergic drugs on the locomotor pattern and on 736 cutaneous reflexes of the adult chronic spinal cat. Brain Res 514, 55–67. 737 Bareyre FM, Kerschensteiner M, Raineteau O, Mettenleiter TC, Weinmann O & Schwab ME (2004). The 738 injured spinal cord spontaneously forms a new intraspinal circuit in adult rats. Nat Neurosci 7, 269–739 277. 740 Barrière G, Frigon A, Leblond H, Provencher J & Rossignol S (2010). Dual Spinal Lesion Paradigm in the Cat: 741 Evolution of the Kinematic Locomotor Pattern. J Neurophysiol 104, 1119–1133. 742 Barriere G, Leblond H, Provencher J & Rossignol S (2008). Prominent Role of the Spinal Central Pattern 743 Generator in the Recovery of Locomotion after Partial Spinal Cord Injuries. J Neurosci 28, 3976–744 3987. 745 Barss TS, Parhizi B & Mushahwar VK (2020). Transcutaneous spinal cord stimulation of the cervical cord 746 modulates lumbar networks. J Neurophysiol 123, 158–166. 747 Basaldella E, Takeoka A, Sigrist M & Arber S (2015). Multisensory Signaling Shapes Vestibulo-Motor Circuit 748 Specificity. Cell 163, 301–312. 749 Bazley FA, Maybhate A, Tan CS, Thakor NV, Kerr C & All AH (2014). Enhancement of Bilateral Cortical 750 Somatosensory Evoked Potentials to Intact Forelimb Stimulation Following Thoracic Contusion 751 Spinal Cord Injury in Rats. IEEE Trans Neural Syst Rehabil Eng 22, 953–964. 752 Beauparlant J, Van Den Brand R, Barraud Q, Friedli L, Musienko P, Dietz V & Courtine G (2013). Undirected 753 compensatory plasticity contributes to neuronal dysfunction after severe spinal cord injury. Brain 754 136, 3347–3361. 755 Bretzner F & Drew T (2005). Motor cortical modulation of cutaneous reflex responses in the hindlimb of the 756 intact cat. J Neurophysiol 94, 673–687. 757 Brink EE, Suzuki I, Timerick SJ & Wilson VJ (1985). Tonic neck reflex of the decerebrate cat: a role for 758 propriospinal neurons. J Neurophysiol 54, 978–987. 759 Brockett EG, Seenan PG, Bannatyne BA & Maxwell DJ (2013). Ascending and descending propriospinal 760 pathways between lumbar and cervical segments in the rat: Evidence for a substantial ascending 761 excitatory pathway. Neuroscience 240, 83–97. 762 Butler JE, Godfrey S & Thomas CK (2016). Interlimb reflexes induced by electrical stimulation of cutaneous 763 nerves after spinal cord injury. PLoS ONE 11, 1–18. 764 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint Cai LL, Courtine G, Fong AJ, Burdick JW, Roy RR & Edgerton VR (2006). Plasticity of functional connectivity in 765 the adult spinal cord. Philos Trans R Soc B Biol Sci 361, 1635–1646. 766 Calancie B (1991). Interlimb reflexes following cervical spinal cord injury in man. Exp Brain Res 85, 458–469. 767 Calancie B, Lutton S & Broton JG (1996). Central nervous system plasticity after spinal cord injury in man: 768 interlimb reflexes and the influence of cutaneous stimulation. Electroencephalogr Clin Neurophysiol 769 Mot Control 101, 304–315. 770 Calancie B, Molano MR & Broton JG (2002). Interlimb reflexes and synaptic plasticity become evident 771 months after human spinal cord injury. Brain 125, 1150–1161. 772 Christensen LOD, Morita H, Petersen N & Nielsen J (1999). Evidence suggesting that a transcortical reflex 773 pathway contributes to cutaneous reflexes in the tibialis anterior muscle during walking in man. Exp 774 Brain Res 124, 59–68. 775 Côté MP, Murray LM & Knikou M (2018). Spinal control of locomotion: Individual neurons, their circuits and 776 functions. Front Physiol 9, 1–27. 777 Courtine G, Song B, Roy RR, Zhong H, Herrmann JE, Ao Y, Qi J, Edgerton VR & Sofroniew MV (2008). 778 Recovery of supraspinal control of stepping via indirect propriospinal relay connections after spinal 779 cord injury. Nat Med 14, 69–74. 780 Cowley KC, MacNeil BJ, Chopek JW, Sutherland S & Schmidt BJ (2015). Neurochemical excitation of thoracic 781 propriospinal neurons improves hindlimb stepping in adult rats with spinal cord lesions. Exp Neurol 782 264, 174–187. 783 Cowley KC, Zaporozhets E & Schmidt BJ (2008). Propriospinal neurons are sufficient for bulbospinal 784 transmission of the locomotor command signal in the neonatal rat spinal cord. J Physiol 586, 1623–785 1635. 786 Cowley KC, Zaporozhets E & Schmidt BJ (2010). Propriospinal transmission of the locomotor command 787 signal in the neonatal rat. Ann N Y Acad Sci 1198, 42–53. 788 Dambreville C, Labarre A, Thibaudier Y, Hurteau MF & Frigon A (2015). The spinal control of locomotion and 789 step-to-step variability in left-right symmetry from slow to moderate speeds. J Neurophysiol 114, 790 1119–1128. 791 De Ruiter GC, Hundza SR & Zehr EP (2010). Phase-dependent modulation of soleus H-reflex amplitude 792 induced by rhythmic arm cycling. Neurosci Lett 475, 7–11. 793 Dickstein R & Laufer Y (2004). Light touch and center of mass stability during treadmill locomotion. Gait 794 Posture 20, 41–47. 795 Dragert K & Zehr EP (2009). Rhythmic arm cycling modulates Hoffmann reflex excitability differentially in 796 the ankle flexor and extensor muscles. Neurosci Lett 450, 235–238. 797 Drew T & Rossignol S (1985). Forelimb responses to cutaneous nerve stimulation during locomotion in 798 intact cats. Brain Res 329, 323–328. 799 Drew T & Rossignol S (1987). Responses of the Forelimb to Perturbations Applied during the Swing Phase of 800 the Step Cycle. In Motor Control, ed. Gantchev GN, Dimitrov B & Gatev P, pp. 171–175. Springer US, 801 Boston, MA. 802 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint Dutton RC, Carstens MI, Antognini JF & Carstens E (2006). Long ascending propriospinal projections from 803 lumbosacral to upper cervical spinal cord in the rat. Brain Res 1119, 76–85. 804 Duysens J (2024). The nine lives of cat reflexes. J PhysiolJP286594. 805 Duysens J & Loeb GE (1980). Modulation of ipsi- and contralateral reflex responses in unrestrained walking 806 cats. J Neurophysiol 44, 1024–1037. 807 Duysens J & Stein RB (1978). Reflexes induced by nerve stimulation in walking cats with implanted cuff 808 electrodes. Exp Brain Res 32, 213–224. 809 Edgerton VR, Tillakaratne NJK, Bigbee AJ, De Leon RD & Roy RR (2004). Plasticity of the spinal neural 810 circuitry after injury*. Annu Rev Neurosci 27, 145–167. 811 English AW (1985). Interlimb coordination during stepping in the cat: The role of the dorsal spinocerebellar 812 tract. Exp Neurol 87, 96–108. 813 Fenrich KK & Rose PK (2009). Spinal Interneuron Axons Spontaneously Regenerate after Spinal Cord Injury 814 in the Adult Feline. J Neurosci 29, 12145–12158. 815 Fink KL & Cafferty WBJ (2016). Reorganization of Intact Descending Motor Circuits to Replace Lost 816 Connections After Injury. Neurotherapeutics 13, 370–381. 817 Fleshman JW, Rudomin P & Burke RE (1988). Supraspinal control of a short-latency cutaneous pathway to 818 hindlimb motoneurons. Exp Brain Res 69, 449–459. 819 Flynn JR, Graham BA, Galea MP & Callister RJ (2011). The role of propriospinal interneurons in recovery 820 from spinal cord injury. Neuropharmacology 60, 809–822. 821 Forero J & Misiaszek JE (2013). The contribution of light touch sensory cues to corrective reactions during 822 treadmill locomotion. Exp Brain Res 226, 575–584. 823 Forero J & Misiaszek JE (2015). The amplitude of interlimb cutaneous reflexes in the leg is influenced by 824 fingertip touch and vision during treadmill locomotion. Exp Brain Res 233, 1773–1782. 825 Forssberg H (1979). Stumbling Corrective Reaction: A Phase-Dependent Compensatory Reaction During 826 Locomotion. J Neurophysiol 42, 36–953. 827 Forssberg H, Grillner S & Rossignol S (1977). Phasic gain control of reflexes from the dorsum of the paw 828 during spinal locomotion. Brain Res 132, 121–139. 829 Frigon A, Akay T & Prilutsky BI (2021). Control of Mammalian Locomotion by Somatosensory Feedback. 830 Compr Physiol 12, 2877–2947. 831 Frigon A, Barrière G, Leblond H & Rossignol S (2009). Asymmetric Changes in Cutaneous Reflexes After a 832 Partial Spinal Lesion and Retention Following Spinalization During Locomotion in the Cat. J 833 Neurophysiol 102, 2667–2680. 834 Frigon A, Collins DF & Zehr EP (2004). Effect of Rhythmic Arm Movement on Reflexes in the Legs: 835 Modulation of Soleus H-Reflexes and Somatosensory Conditioning. J Neurophysiol 91, 1516–1523. 836 Frigon A & Rossignol S (2006). Functional plasticity following spinal cord lesions. Prog Brain Res 157, 231–837 260. 838 Frigon A & Rossignol S (2007). Short-Latency Crossed Inhibitory Responses in Extensor Muscles During 839 Locomotion in the Cat. J Neurophysiol 99, 989–998. 840 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint Frigon A & Rossignol S (2008). Adaptive changes of the locomotor pattern and cutaneous reflexes during 841 locomotion studied in the same cats before and after spinalization. J Physiol 586, 2927–2945. 842 Frigon A & Rossignol S (2009). Partial denervation of ankle extensors prior to spinalization in cats impacts 843 the expression of locomotion and the phasic modulation of reflexes. Neuroscience 158, 1675–1690. 844 Frigon A, Thibaudier Y & Hurteau M-F (2015). Modulation of forelimb and hindlimb muscle activity during 845 quadrupedal tied-belt and split-belt locomotion in intact cats. Neuroscience 290, 266–278. 846 Fritz N, Illert M & Reeh P (1986a). Location of motoneurones projecting to the cat distal forelimb. II. 847 Median and ulnar motornuclei. J Comp Neurol 244, 302–312. 848 Fritz N, Illert M & Saggau P (1986b). Location of motoneurones projecting to the cat distal forelimb. I. Deep 849 radial motornuclei. J Comp Neurol 244, 286–301. 850 Fuwa T, Shimamura M & Tanaka I (1991). Analysis of the forelimb crossed extension reflex in thalamic cats 851 during stepping. Neurosci Res 9, 257–269. 852 Giovanelli Barilari M & Kuypers HGJM (1969). Propriospinal fibers interconnecting the spinal enlargements 853 in the cat. Brain Res 14, 321–330. 854 Gossard JP, Delivet-Mongrain H, Martinez M, Kundu A, Escalona M & Rossignol S (2015). Plastic changes in 855 lumbar locomotor networks after a partial spinal cord injury in cats. J Neurosci 35, 9446–9455. 856 Grundy D (2015). Principles and standards for reporting animal experiments in The Journal of Physiology 857 and Experimental Physiology. Exp Physiol 100, 755–758. 858 Hakkı Nur İ, Keleş H & Pérez W (2020). Origin and distribution of the brachial plexus of the Van cats. J Vet 859 Med Ser C Anat Histol Embryol 49, 251–259. 860 Haridas C & Zehr EP (2003). Coordinated Interlimb Compensatory Responses to Electrical Stimulation of 861 Cutaneous Nerves in the Hand and Foot during Walking. J Neurophysiol 90, 2850–2861. 862 Harnie J, Al Arab R, Mari S, Yassine S, Eddaoui O, Jéhannin P, Audet J, Lecomte C, Iorio-Morin C, Prilutsky BI, 863 Rybak IA & Frigon A (2024). Forelimb movements contribute to hindlimb cutaneous reflexes during 864 locomotion in cats. BioRxiv. 865 Hiersemenzel L-P, Curt A & Dietz V (2000). From spinal shock to spasticity. Neurology 54, 1574–1582. 866 Higgin D, Krupka A, Maghsoudi OH, Klishko AN, Nichols TR, Lyle MA, Prilutsky BI & Lemay MA (2020). 867 Adaptation to slope in locomotor-trained spinal cats with intact and self-reinnervated lateral 868 gastrocnemius and soleus muscles. J Neurophysiol 123, 70–89. 869 Hiraoka K & Iwata A (2006). Cyclic modulation of H-reflex depression in ipsilateral and contralateral soleus 870 muscles during rhythmic arm swing. Somatosens Mot Res 23, 127–133. 871 Hörner M & Kümmel H (1993). Topographical representation of shoulder motor nuclei in the cat spinal cord 872 as revealed by retrograde fluorochrome tracers. J Comp Neurol 335, 309–319. 873 Huijben B, Van Schooten KS, Van Dieën JH & Pijnappels M (2018). The effect of walking speed on quality of 874 gait in older adults. Gait Posture 65, 112–116. 875 Hundza SR, De Ruiter GC, Klimstra M & Zehr EP (2012). Effect of afferent feedback and central motor 876 commands on soleus H-reflex suppression during arm cycling. J Neurophysiol 108, 3049–3058. 877 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint Hundza SR & Zehr EP (2009). Suppression of soleus H-reflex amplitude is graded with frequency of rhythmic 878 arm cycling. Exp Brain Res 193, 297–306. 879 Hurteau M & Frigon A (2018). A spinal mechanism related to left–right symmetry reduces cutaneous reflex 880 modulation independently of speed during split-belt locomotion. J Neurosci 38, 10314–10328. 881 Hurteau M, Thibaudier Y, Dambreville C, Danner SM, Rybak IA & Frigon A (2018). Intralimb and Interlimb 882 Cutaneous Reflexes during Locomotion in the Intact Cat. J Neurosci 38, 4104–4122. 883 Hurteau M-F, Thibaudier Y, Dambreville C, Chraibi A, Desrochers E, Telonio A & Frigon A (2017). Nonlinear 884 Modulation of Cutaneous Reflexes with Increasing Speed of Locomotion in Spinal Cats. J Neurosci 885 37, 3896–3912. 886 Jane JA, Evans JP & Fisher LE (1964). An Investigation Concerning the Restitution of Motor Function 887 Following Injury to the Spinal Cord. J Neurosurg 21, 167–171. 888 Javan B & Zehr EP (2008). Short-Term Plasticity of Spinal Reflex Excitability Induced by Rhythmic Arm 889 Movement. J Neurophysiol 99, 2000–2005. 890 Juvin L, Simmers J & Morin D (2005). Propriospinal circuitry underlying interlimb coordination in 891 mammalian quadrupedal locomotion. J Neurosci 25, 6025–6035. 892 Kato M, Murakami S, Hirayama H & Hikino K (1985). Recovery of postural control following chronic bilateral 893 hemisections at different spinal cord levels in adult cats. Exp Neurol 90, 350–364. 894 Kato M, Murakami S, Yasuda K & Hirayama H (1984). Disruption of fore- and hindlimb coordination during 895 overground locomotion in cats with bilateral serial hemisection of the spinal cord. Neurosci Res 2, 896 27–47. 897 Kinoshita M & Yamaguchi T (2001). Stimulus time-locked responses of motoneurons during forelimb fictive 898 locomotion evoked by repetitive stimulation of the lateral funiculus. Brain Res 904, 31–42. 899 Krupa P, Siddiqui AM, Grahn PJ, Islam R, Chen BK, Madigan NN, Windebank AJ & Lavrov IA (2022). The 900 Translesional Spinal Network and Its Reorganization after Spinal Cord Injury. The Neuroscientist 28, 901 163–179. 902 LaBella LA, Niechaj A & Rossignol S (1992). Low-threshold, short-latency cutaneous reflexes during fictive 903 locomotion in the “semi-chronic” spinal cat. Exp Brain Res 91, 236–248. 904 Laliberte AM, Goltash S, Lalonde NR & Bui TV (2019). Propriospinal Neurons: Essential Elements of 905 Locomotor Control in the Intact and Possibly the Injured Spinal Cord. Front Cell Neurosci 13, 00512. 906 Lalonde NR & Bui TV (2021). Do spinal circuits still require gating of sensory information by presynaptic 907 inhibition after spinal cord injury? Curr Opin Physiol 19, 113–118. 908 Lecomte CG, Mari S, Audet J, Merlet AN, Harnie J, Beaulieu C, Abdallah K, Gendron L, Rybak IA, Prilutsky BI 909 & Frigon A (2022). Modulation of the gait pattern during split-belt locomotion after lateral spinal 910 cord hemisection in adult cats. J Neurophysiol 128, 1593–1616. 911 Lecomte CG, Mari S, Audet J, Yassine S, Merlet AN, Morency C, Harnie J, Beaulieu C, Gendron L & Frigon A 912 (2023). Neuromechanical Strategies for Obstacle Negotiation during Overground Locomotion 913 following Incomplete Spinal Cord Injury in Adult Cats. J Neurosci 43, 5623–5641. 914 Loadman PM & Zehr EP (2007). Rhythmic arm cycling produces a non-specific signal that suppresses Soleus 915 H-reflex amplitude in stationary legs. Exp Brain Res 179, 199–208. 916 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint Loeb GE (1993). The distal hindlimb musculature of the cat: interanimal variability of locomotor activity and 917 cutaneous reflexes. Exp Brain Res 96, 125–140. 918 Maier IC & Schwab ME (2006). Sprouting, regeneration and circuit formation in the injured spinal cord: 919 factors and activity. Philos Trans R Soc B Biol Sci 361, 1611–1634. 920 Maj J, Palider W & Baran L (1976). The effects of serotonergic and antiserotonergic drugs on the flexor 921 reflex of spinal rat: a proposed model to evaluate the action on the central serotonin receptor. J 922 Neural Transm 38, 131–147. 923 Mari S, Lecomte CG, Merlet AN, Audet J, Harnie J, Rybak IA, Prilutsky BI & Frigon A (2023). A sensory signal 924 related to left-right symmetry modulates intra- and interlimb cutaneous reflexes during locomotion 925 in intact cats. Front Syst Neurosci 17, 1199079. 926 Mari S, Lecomte CG, Merlet AN, Audet J, Yassine S, Eddaoui O, Genois G, Nadeau C, Harnie J, Rybak IA, 927 Prilutsky BI & Frigon A (2024). Changes in intra‐ and interlimb reflexes from hindlimb cutaneous 928 afferents after staggered thoracic lateral hemisections during locomotion in cats. J 929 PhysiolJP286151. 930 Martinez M, Delivet-Mongrain H, Leblond H & Rossignol S (2011). Recovery of hindlimb locomotion after 931 incomplete spinal cord injury in the cat involves spontaneous compensatory changes within the 932 spinal locomotor circuitry. J Neurophysiol 106, 1969–1984. 933 Martinez M, Delivet-Mongrain H, Leblond H & Rossignol S (2012). Incomplete spinal cord injury promotes 934 durable functional changes within the spinal locomotor circuitry. J Neurophysiol 108, 124–134. 935 Massaad F, Levin O, Meyns P, Drijkoningen D, Swinnen SP & Duysens J (2014). Arm sway holds sway: 936 Locomotor-like modulation of leg reflexes when arms swing in alternation. Neuroscience 258, 34–937 46. 938 Matsukawa K, Kamei H, Minoda K & Udo M (1982). Interlimb Coordination in Cat Locomotion Investigated 939 with Perturbation I. Behavioral and Electromyographic Study on Symmetric Limbs of Decerebrate 940 and Awake Walking Cats*. Exp Brain Res 46, 425–437. 941 Matthews PBC (1986). Observations of the automatic compensation of reflex gain on varying the pre-942 existing level on motor discharge in man. J Physiol 374, 73–90. 943 May Z, Fenrich KK, Dahlby J, Batty NJ, Torres-Espín A & Fouad K (2017). Following Spinal Cord Injury 944 Transected Reticulospinal Tract Axons Develop New Collateral Inputs to Spinal Interneurons in 945 Parallel with Locomotor Recovery. Neural Plast 2017, 1–15. 946 Mencalha R, Fernandes N, Sousa CADS & Abidu-Figueiredo M (2014). A cadaveric study to determine the 947 minimum volume of methylene blue to completely color the nerves of brachial plexus in cats. An 948 update in forelimb and shoulder surgeries. Acta Cir Bras 29, 382–388. 949 Merlet AN, Harnie J & Frigon A (2021). Inhibition and facilitation of the spinal locomotor central pattern 950 generator and reflex circuits by somatosensory feedback from the lumbar and perineal regions 951 after spinal cord injury. Front Neurosci 15, 1–14. 952 Merlet AN, Harnie J, Macovei M, Doelman A, Gaudreault N & Frigon A (2020). Mechanically stimulating the 953 lumbar region inhibits locomotor-like activity and increases the gain of cutaneous reflexes from the 954 paws in spinal cats. J Neurophysiol 123, 1026–1041. 955 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint Merlet AN, Jéhannin P, Mari S, Lecomte CG, Audet J, Harnie J, Rybak IA, Prilutsky BI & Frigon A (2022). 956 Sensory perturbations from hindlimb cutaneous afferents generate coordinated functional 957 responses in all four limbs during locomotion in intact cats. eNeuro 9, 0178–22. 958 Miller S, Reitsma DJ & Van Der Meché FGA (1973). Functional organization of long ascending propriospinal 959 pathways linking lumbo-sacral and cervical segments in the cat. Brain Res 62, 169–188. 960 Miller S, Ruit JB & Van Der Meché FGA (1977). Reversal of sign of long spinal reflexes dependant on the 961 phase of the step cycle in the high decerebrate cat. 128, 447–459. 962 Molenaar I & Kuypers HGJM (1978). Cells of origin of propriospinal fibers and of fibers ascending to 963 supraspinal levels. A HRP study in cat and rhesus monkey. Brain Res 152, 429–450. 964 Ni Y, Nawabi H, Liu X, Yang L, Miyamichi K, Tedeschi A, Xu B, Wall NR, Callaway EM & He Z (2014). 965 Characterization of Long Descending Premotor Propriospinal Neurons in the Spinal Cord. J Neurosci 966 34, 9404–9417. 967 Noga BR, Johnson DMG, Riesgo MI & Pinzon A (2009). Locomotor-Activated Neurons of the Cat. I. 968 Serotonergic Innervation and Co-Localization of 5-HT 7 , 5-HT 2A , and 5-HT 1A Receptors in the 969 Thoraco-Lumbar Spinal Cord. J Neurophysiol 102, 1560–1576. 970 Noga BR, Johnson DMG, Riesgo MI & Pinzon A (2011). Locomotor-activated neurons of the cat. II. 971 Noradrenergic innervation and colocalization with NEα1a or NEα2b receptors in the thoraco-972 lumbar spinal cord. J Neurophysiol 105, 1835–1849. 973 Nozaki M, Bell JA, Vaupel DB & Martin WR (1977). Responses of the flexor reflex to LSD, tryptamine, 5-974 hydroxytryptophan, methoxamine, and d-amphetamine in acute and chronic spinal rats. 975 Psychopharmacology (Berl) 55, 13–18. 976 Parhizi B, Barss TS & Mushahwar VK (2021). Simultaneous Cervical and Lumbar Spinal Cord Stimulation 977 Induces Facilitation of Both Spinal and Corticospinal Circuitry in Humans. Front Neurosci 15, 978 615103. 979 Park H, Latash EM, Molkov YI, Klishko AN, Frigon A, DeWeerth SP & Prilutsky BI (2019). Cutaneous sensory 980 feedback from paw pads affects lateral balance control during split-belt locomotion in the cat. J Exp 981 Biol 222, jeb198648. 982 Pearcey GEP & Zehr EP (2019a). We are upright-walking cats: Human limbs as sensory antennae during 983 locomotion. Physiology 34, 354–364. 984 Pearcey GEP & Zehr EP (2019b). Exploiting cervicolumbar connections enhances short-term spinal cord 985 plasticity induced by rhythmic movement. Exp Brain Res 237, 2319–2329. 986 Percie Du Sert N et al. (2020). The ARRIVE guidelines 2.0: Updated guidelines for reporting animal 987 research*. J Cereb Blood Flow Metab 40, 1769–1777. 988 Pijnappels M, Van Wezel BMH, Colombo G, Dietz V & Duysens J (1998). Cortical facilitation of cutaneous 989 reflexes in leg muscles during human gait. Brain Res 787, 149–153. 990 Pocratsky AM, Shepard CT, Morehouse JR, Burke DA, Riegler AS, Hardin JT, Beare JE, Hainline C, States GJR, 991 Brown BL, Whittemore SR & Magnuson DSK (2020). Long ascending propriospinal neurons provide 992 flexible, context-specific control of interlimb coordination. eLife 9, 1–24. 993 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint Pratt CA, Chanaud CM & Loeb GE (1991). Functionally complex muscles of the cat hindlimb IV. 994 Intramuscular distribution of movement command signals and cutaneous reflexes in broad, 995 bifunctional thigh muscles. Exp Brain Res 85, 281–299. 996 Prochazka A, Mushahwar V & Yakovenko S (2002). Activation and coordination of spinal motoneuron pools 997 after spinal cord injury. Prog Brain Res 137, 109–124. 998 Prochazka A, Sontag K-H & Wand P (1978). Motor reactions to perturbations of gait : proprioceptive and 999 somesthetic involvement. Neurosci Lett 7, 35–39. 1000 Quevedo J, Stecina K, Gosgnach S & McCrea DA (2005a). Stumbling corrective reaction during fictive 1001 locomotion in the cat. J Neurophysiol 94, 2045–2052. 1002 Quevedo J, Stecina K & McCrea DA (2005b). Intracellular analysis of reflex pathways underlying the 1003 stumbling corrective reaction during fictive locomotion in the cat. J Neurophysiol 94, 2053–2062. 1004 Reed WR, Shum-Siu A, Onifer SM & Magnuson DSK (2006). Inter-enlargement pathways in the ventrolateral 1005 funiculus of the adult rat spinal cord. Neuroscience 142, 1195–1207. 1006 Reed WR, Shum-Siu A, Whelan A, Onifer SM & Magnuson DSK (2009). Anterograde labeling of ventrolateral 1007 funiculus pathways with spinal enlargement connections in the adult rat spinal cord. Brain Res 1008 1302, 76–84. 1009 Rossignol S, Drew T, Brustein E & Jiang W (1999). Chapter 31 Locomotor Performance and Adaptation after 1010 Partial or Complete Spinal Cord Lesions in the Cat. In Progress in Brain Research, pp. 349–365. 1011 Elsevier. 1012 Rossignol S, Dubuc R & Gossard JP (2006). Dynamic sensorimotor interactions in locomotion. Physiol Rev 1013 86, 89–154. 1014 Rossignol S & Frigon A (2011). Recovery of locomotion after spinal cord injury: Some facts and mechanisms. 1015 Annu Rev Neurosci 34, 413–440. 1016 Schillings AM, Van Wezel BMH & Duysens J (1996). Mechanically induced stumbling during human treadmill 1017 walking. J Neurosci Methods 67, 11–17. 1018 Seki K & Yamaguchi T (1997). Cutaneous reflex activity of the cat forelimb during fictive locomotion. Brain 1019 Res 753, 56–62. 1020 Shah PK, Garcia-Alias G, Choe J, Gad P, Gerasimenko Y, Tillakaratne N, Zhong H, Roy RR & Edgerton VR 1021 (2013). Use of quadrupedal step training to re-engage spinal interneuronal networks and improve 1022 locomotor function after spinal cord injury. Brain 136, 3362–3377. 1023 Shepard CT, Brown BL, Van Rijswijck MA, Zalla RM, Burke DA, Morehouse JR, Riegler AS, Whittemore SR & 1024 Magnuson DS (2023). Silencing long-descending inter-enlargement propriospinal neurons improves 1025 hindlimb stepping after contusive spinal cord injuries. eLife 12, e82944. 1026 Shepard CT, Pocratsky AM, Brown BL, Van Rijswijck MA, Zalla RM, Burke DA, Morehouse JR, Riegler AS, 1027 Whittemore SR & Magnuson DSK (2021). Silencing long ascending propriospinal neurons after 1028 spinal cord injury improves hindlimb stepping in the adult rat. eLife 10, 1–24. 1029 Shimamura M, Kogure I & Fuwa T (1984). Role of joint afferents in relation to the initiation of forelimb 1030 stepping in thalamic cats. Brain Res 297, 225–234. 1031 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint Shimamura M & Livingston R (1962). Longitudinal conduction systems serving spinal and brainstem 1032 coordination-spino-bulbo-spinal reflex. J Physiol Soc Jpn 33, 225–233. 1033 Shimamura M & Livingston RB (1963). Longitudinal conduction systems serving spinal and brain-stem 1034 coordination. J Neurophysiol 26, 258–272. 1035 Shimamura M, Tanaka I & Fuwa T (1990). Comparison between spino-bulbo-spinal and propriospinal 1036 reflexes in thalamic cats during stepping. Neurosci Res 7, 358–368. 1037 Stelzner DJ & Cullen JM (1991). Do Propriospinal Projections Contribute to Hindlimb Recovery when All 1038 Long Tracts Are Cut in Neonatal or Weanling Rats? Exp Neurol 114, 193–205. 1039 Sterling P & Kuypers HGJM (1967). Anatomical organization of the brachial spinal cord of the cat. II. The 1040 motoneuron plexus. Brain Res 4, 16–32. 1041 Van Den Brand R, Heutschi J, Barraud Q, DiGiovanna J, Bartholdi K, Huerlimann M, Friedli L, Vollenweider I, 1042 Moraud EM, Duis S, Dominici N, Micera S, Musienko P & Courtine G (2012). Restoring voluntary 1043 control of locomotion after paralyzing spinal cord injury. Science 336, 1182–1185. 1044 Van Wezel BMH, Ottenhoff FAM & Duysens J (1997). Dynamic Control of Location-Specific Information in 1045 Tactile Cutaneous Reflexes from the Foot during Human Walking. J Neurosci 17, 3804–3814. 1046 Wand P, Prochazka A & Sontag K-H (1980). Neuromuscular Responses to Gait Perturbations in Freely 1047 Moving Cats*. Exp Brain Res 38, 109–114. 1048 Yamaguchi T (1992). Activity of Cervical Neurons during Forelimb Fictive Locomotion in Decerebrate Cats. 1049 Jpn J Physiol 42, 501–514. 1050 Yamaguchi T (2004). The central pattern generator for forelimb locomotion in the cat. In Progress in Brain 1051 Research, pp. 115–122. Elsevier. 1052 Zaporozhets E, Cowley KC & Schmidt BJ (2011). Neurochemical excitation of propriospinal neurons 1053 facilitates locomotor command signal transmission in the lesioned spinal cord. J Neurophysiol 105, 1054 2818–2829. 1055 Zavvarian M-M, Hong J & Fehlings MG (2020). The Functional Role of Spinal Interneurons Following 1056 Traumatic Spinal Cord Injury. Front Cell Neurosci 14, 00127. 1057 Zehr EP, Frigon A, Hoogenboom N & Collins DF (2004). Facilitation of soleus H-reflex amplitude evoked by 1058 cutaneous nerve stimulation at the wrist is not suppressed by rhythmic arm movement. Exp Brain 1059 Res 159, 382–388. 1060 Zehr EP, Komiyama T & Stein RB (1997). Cutaneous reflexes during human gait: Electromyographic and 1061 kinematic responses to electrical stimulation. J Neurophysiol 77, 3311–3325. 1062 Zhou R, Alvarado L, Kim S, Chong SL & Mushahwar VK (2017). Modulation of corticospinal input to the legs 1063 by arm and leg cycling in people with incomplete spinal cord injury. J Neurophysiol 118, 2507–2519. 1064 1065 1066 1067 1068 1069 1070 1071 1072 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint Tables and Figure legends 1073 Table 1. Homonymous and crossed reflex response before and after staggered hemisections. The 1074 table shows homonymous and crossed responses (P1, P2, P3, N1, N2 and N3) evoked in individual cats in 1075 left and right forelimb muscles in the intact state, and after the first (H1) and second (H2) hemisections at 1-2 1076 time points (T1 or T2). Bold responses with an asterisk indicate a significant phase modulation (one factor 1077 ANOVA, p < 0.05). x, No response. -, Non-implanted or non-analyzable (lost or excessive noise) muscle. BB, 1078 biceps brachii; ECU, extensor carpi ulnaris; FCU, flexor carpi ulnaris; LD, latissimus dorsi; TRI, triceps 1079 brachii. 1080 1081 Table 2. Homolateral reflex responses before and after staggered hemisections. The table shows 1082 homolateral responses (P1, P2, P3, N1, N2 and N3) evoked in individual cats in left and right hindlimb 1083 muscles in the intact state, and after the first (H1) and second (H2) hemisections at 1-2 time points (T1 or 1084 T2). Bold responses with an asterisk indicate a significant phase modulation (one factor ANOVA, p < 0.05). 1085 x, No response. -, Non-implanted or non-analyzable (lost or excessive noise) muscle. BFA, biceps femoris 1086 anterior; BFP, biceps femoris posterior; IP, iliopsoas; LG, lateral gastrocnemius; MG, medial gastr ocnemius; 1087 SOL, soleus; SRT, anterior sartorius; ST, semitendinosus; TA, tibialis anterior ; VL, vastus lateralis. 1088 1089 Table 3. Diagonal reflex responses before and after staggered hemisections. The table shows diagonal 1090 responses (P1, P2, P3, N1, N2 and N3) evoked in individual cats in left and right hindlimb muscles in the 1091 intact state, and after the first (H1) and second (H2) hemisections at 1-2 time points (T1 or T2). Bold 1092 responses with an asterisk indicate a significant phase modulation (one factor ANOVA, p < 0.05). x, No 1093 response. -, Non-implanted or non-analyzable (lost or excessive noise) muscle. BFA, biceps femoris anterior; 1094 BFP, biceps femoris posterior; IP, iliopsoas; LG, lateral gastrocnemius; MG, medial gastrocnemius; SOL, 1095 soleus; SRT, anterior sartorius; ST, semitendinosus; TA, tibialis anterior; VL, vastus lateralis. 1096 1097 Figure 1. Experimental chronology and estimation of lesions extent. (A) Chronology showing the first 1098 (T1) and second (T2) experimental time points after the first (H1) and second (H2) hemisections in all cats . 1099 (B) For the extent of the lesions, the black area represents the estimation as a percentage of total cross-1100 sectional area (reproduced with permission of (Mari et al., 2024)). Note that we only performed one lesion in 1101 cat KI. 1102 1103 Figure 2. Phase-dependent modulation of cutaneous reflexes evoked in homonymous forelimb 1104 muscles during locomotion before and following staggered hemisections. Each panel shows, from left 1105 to right, stance phases of the stimulated forelimb (empty horizontal bars) with its averaged rectified muscle 1106 activity normalized to cycle duration in the different states/time points, and homonymous reflex responses in 1107 representative cats for the left and right (A) extensor carpi ulnaris (ECU, cat KA), (B) triceps brachii (TRI, cat 1108 TO), and (C) biceps brachii (BB, cat JA). Reflex responses are shown with a post-stimulation window of 80 1109 ms in four phases in the intact state, and after the first (H1) and second (H2) hemisections at time points 1 1110 (T1) and/or 2 (T2). At each state/time point, evoked responses are scaled according to the largest response 1111 obtained in one of the four phases. The scale, however, differs between states/time points. The dotted 1112 vertical lines in the reflex responses indicate the N1/P1, N2/P2 and N3/P3 time windows. 1113 1114 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint Figure 3. Phase-dependent modulation of cutaneous reflexes evoked in crossed forelimb muscles 1115 during locomotion before and following staggered hemisections. Each panel shows, from left to right, 1116 stance phases of the stimulated forelimb (empty horizontal bars) and crossed forelimb (filled horizontal bars) 1117 with its averaged rectified muscle activity normalized to cycle duration in the different states/time points, and 1118 crossed reflex responses in representative cats for the left and right (A) extensor carpi ulnaris (LECU, cat 1119 GR; RECU, cat TO), (B) triceps brachii (LTRI, cat AR; RTRI, cat JA), and (C) biceps brachii (BB, cat JA). 1120 Reflex responses are shown with a post-stimulation window of 80 ms in four phases in the intact state, and 1121 after the first (H1) and second (H2) hemisections at time points 1 (T1) and/or 2 (T2). At each state/time point, 1122 evoked responses are scaled according to the largest response obtained in one of the four phases. The 1123 scale, however, differs between states/time points. The dotted vertical lines in the reflex responses indicate 1124 the N1/P1, N2/P2 and N3/P3 time windows. 1125 1126 Figure 4. Phase-dependent modulation of cutaneous reflexes evoked in homolateral hindlimb 1127 muscles during locomotion before and following staggered hemisections. Each panel shows, from left 1128 to right, stance phases of the stimulated forelimb (empty horizontal bars) and homolateral hindlimb (filled 1129 horizontal bars) with its averaged rectified muscle activity normalized to cycle duration in the different 1130 states/time points, and homolateral reflex responses in representative cats for the left and right (A) soleus 1131 (SOL, cat TO), (B) vastus lateralis (LVL, cat HO; RVL, cat TO), and (C) anterior sartorius (SRT, cat JA). 1132 Reflex responses are shown with a post-stimulation window of 80 ms in four phases in the intact state, and 1133 after the first (H1) and second (H2) hemisections at time points 1 (T1) and/or 2 (T2). At each state/time point, 1134 evoked responses are scaled according to the largest response obtained in one of the four phases. The 1135 scale, however, differs between states/time points. The dotted vertical lines in the reflex responses indicate 1136 the N1/P1, N2/P2 and N3/P3 time windows. 1137 1138 Figure 5. Phase-dependent modulation of cutaneous reflexes evoked in diagonal hindlimb muscles 1139 during locomotion before and following staggered hemisections. Each panel shows, from left to right, 1140 stance phases of the stimulated forelimb (empty horizontal bars) and diagonal hindlimb (filled horizontal 1141 bars) with its averaged rectified muscle activity normalized to cycle duration in the different states/time 1142 points, and diagonal reflex responses in representative cats for the left and right (A) soleus (LSOL, cat TO; 1143 RSOL, cat JA), (B) vastus lateralis (VL, cat TO), and (C) anterior sartorius (LSRT, cat TO; RSRT, cat JA). 1144 Reflex responses are shown with a post-stimulation window of 80 ms in four phases in the intact state, and 1145 after the first (H1) and second (H2) hemisections at time points 1 (T1) and/or 2 (T2). At each state/time point, 1146 evoked responses are scaled according to the largest response obtained in one of the four phases. The 1147 scale, however, differs between states/time points. The dotted vertical lines in the reflex responses indicate 1148 the N1/P1, N2/P2 and N3/P3 time windows. 1149 1150 Figure 6. Reflex response occurrence in all four limbs before and after staggered hemisections. 1151 Response occurrence probabilities are shown for short- (SLR) and mid-/long-latency (MLRs/LLRs) 1152 responses with stimulation of the left or right superficial radial nerve before (intact) and after the first (H1) and 1153 second (H2) hemisections at time points 1 (T1) and/or 2 (T2). Tables 1, 2 and 3 provide details on the 1154 number of pooled data for SLRs, MLRs and LLRs. Each filled circle represents the mean probability ± 1155 confidence interval (95%) in 5 forelimb or 10 hindlimb muscles pooled across cats for homonymous/crossed 1156 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint (A and B) and homolateral/diagonal (C and D) responses, respectively. If a significant main effect of 1157 state/timepoint was found (generalized linear mixed model), we compared states/time points. Asterisks 1158 indicate significant differences at p < 0.05*, p < 0.01** and p < 0.001***. When one state/time point was 1159 significantly different from two states/time points, the comparison starts with a longer horizontal line. 1160 1161 Figure 7. Schematic illustration of putative pathways and mechanisms contributing to cutaneous 1162 reflexes and their modulation before and after staggered hemisections. 1163 In the intact state, afferents from the left (A) and right (B) superficial radial (SR) nerves contact spinal 1164 interneurons that project to motoneurons within the hemisegment (homonymous responses) at cervical 1165 levels and commissural interneurons projecting contralaterally (crossed responses). SR nerve afferents also 1166 make contacts with propriospinal neurons that project to lumbar levels, terminating ipsilaterally (homolateral 1167 responses) or contralaterally via collateral projections at different segments (diagonal responses). The 1168 pathways responsible for short-latency responses (SLRs) are mainly confined to spinal circuits, including 1169 SLRs in hindlimb muscles. The pathways contributing to mid- and long-latency responses (MLRs/LLRs) 1170 transmit sensory information to supraspinal structures via long ascending projection neurons (propriospinal 1171 and/or dorsal lemniscal pathways) that project back to spinal circuits controlling the fore- and hindlimbs. After 1172 the first hemisection (on the right side), SLRs and MLRs/LLRs in forelimb muscles remain present although 1173 their response pattern can change due to the loss of inhibitory or excitatory ascending pathways from 1174 lumbosacral circuits. The occurrence of MLRs/LLRs in hindlimb decreases (dashed lines) due to disruptions 1175 in ascending and descending pathways to and from supraspinal structures. Spared supraspinal axons are 1176 potentially strengthened or sprout to form new connections to transmit descending signals. After the second 1177 hemisection (on the left side), direct ascending and descending pathways are both disrupted, and 1178 reorganization of short propriospinal neurons is required to relay signals through the lesions, although their 1179 ability to do so is limited, leading to a considerable loss in MLRs and LLRs in hindlimb muscles. 1180 1181 Figure Abstract 1182 Contacting an obstacle during locomotion activates cutaneous afferents to maintain balance and coordinate 1183 all four limbs. Spinal cord injuries disrupt neural communications between spinal networks controlling the 1184 fore- and hindlimbs, impairing balance and limb coordination. Cutaneous reflex pathways can be used to 1185 develop therapeutic approaches for restoring ascending and descending transmission to facilitate locomotor 1186 recovery. 1187 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint Forelimb reflexes Hindlimb reflexes T5-T6 T5-T6Intact T10-T11 T5-T6 T10-T11 Supraspinal structures Fore-hind coordination performanceSuperficial radial nerve stimulation Extensor Flexor P2 P1 P2 N1 P3 P2 P2 N3 8 weeks post-SCI Figure abstract .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint Figure 1 st 1 lesion nd 2 lesion 0 2 4 6 8 10 12 16 18 Time post injuries (weeks) Intact H1T1 H1T2 H2T2 H2T1 14 20 Cat ID AR TO KI KA JA HO GR TOKIKAJAHOGRAR st 1 lesion (right T5-T6) nd 2 lesion (left T10-T11) 53.1 46.9 42.266.443.547.440.751.4 33.5 40.9 49.4 51.7 53.7 A B .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint Left ECU Right BBLeft BB Right ECU Left forelimb homonymous responses Right forelimb homonymous responses STIM 10 ms 1 2 43 STIM 10 ms 2 4 1 3 STIM 10 ms 1 2 43 STIM 10 ms 2 4 1 3 A B 2 3 4 stance-to-swing mid-swing mid-stance 1 swing-to-stance Left TRI STIM 10 ms 1 2 43 C Right TRI STIM 10 ms 1 2 43 1 11 1 1 1 1 11 1 1 1 2 22 2 2 2 4 44 4 4 4 3 33 3 3 3 Intact H1T1 H1T2 H2T1 H2T2 Figure 2 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint Left ECU Right BBLeft BB Right ECU Left forelimb crossed responses Right forelimb crossed responses STIM 10 ms 1 2 43 STIM 10 ms 1 2 43 STIM 10 ms 2 4 1 3 STIM 10 ms 1 2 43 stance-to-swing mid-swing mid-stance swing-to-stance Left TRI STIM 10 ms 1 2 43 11 2 43 11 2 43 1 1 11 1 1 11 2 2 22 4 4 44 3 3 33 Right TRI STIM 10 ms 2 4 1 3 Intact H1T1 H1T2 H2T1 H2T2 2 3 4 1 Figure 3 A B C .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint Left SOL Right VLLeft VL Right SRTLeft SRT Right SOL Left hindlimb homolateral responses Right hindlimb homolateral responses STIM 10 ms 1 2 43 STIM 10 ms 1 2 43 STIM 10 ms 2 4 1 3 STIM 10ms 1 2 43 STIM 10 ms 1 2 43 STIM 10 ms 1 2 43 1 2 21 44 3 2 34 1 21 43 23 4 1 4 3 1 24 3 stance-to-swing mid-swing mid-stance swing-to-stance 1st cycle 2nd cycle 4 21 4 23 4 1 24 3 4 1 43 12 23 1 24 3 4 21 43 23 4 1 1 24 3 4 3 4 2 13 21 4 23 4 1 24 3 13 Intact H1T1 H1T2 H2T1 H2T2 2 3 4 1 Figure 4 A B C .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint Left SOL Right VLLeft VL Right SRTLeft SRT Right SOL STIM 10 ms 1 2 43 STIM 10 ms 1 2 43 STIM 10 ms 1 2 43 STIM 10 ms 2 4 1 3 STIM 10 ms 1 2 43 STIM 10 ms 2 4 1 3 stance-to-swing mid-swing mid-stance swing-to-stance 12 2 2 22 4 1 4 3 1st cycle 2nd cycle 3 21 44 22 3 12 243 13 21 43 23 31 1 22 43 4 Left hindlimb diagonal responses Right hindlimb diagonal responses 1 44 3 12 2 2 22 4 1 4 3 3 1 44 3 12 2 2 22 4 1 4 3 3 1 44 3 4 4 21 44 22 3 12 243 134 Intact H1T1 H1T2 H2T1 H2T2 2 3 4 1 Figure 5 A B C .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint Occurrence probability Occurrence probability SLR SLR 0 0.5 1.0 0 0.5 1.0 0 0.5 1.0 0 0.5 1.0 0 0.5 1.0 *** *** *** ** 0 0.5 1.0 0 0.5 1.0 ********* 0 0.5 1.0 Left SR stimulation Right SR stimulation Intact H1T1 H1T2 H2T2 ********* ** * * ********* Homonymous Crossed Homolateral Homolateral Diagonal Diagonal Crossed Homonymous A B C D MLR/ LLR MLR/ LLR MLR/ LLR MLR/ LLR Figure 6 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint Intact state After first hemisection MLR/LLRSLRMotoneuron Excitatory neurons Inhibitory neurons Propriospinal neurons Modulatory effects After second hemisectionA B Lumbar Thoracic Thoracic Cervical collaterals Supraspinal structures Supraspinal structuresSupraspinal structures Lumbar Cervical Supraspinal reflex control Supraspinal reflex control Lemniscal pathways Lemniscal pathways Propriospinal pathways Propriospinal pathways Ascending/Descending circuits reorganization Supraspinal structures Supraspinal structuresSupraspinal structures Left SR Left SR Left SR Right SR Right SR collaterals Right SR Figure 7 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 23, 2024. ; https://doi.org/10.1101/2024.04.23.590723doi: bioRxiv preprint