A Gradient in Synaptic Strength and Plasticity among Motoneurons Provides a Peripheral Mechanism for Locomotor Control  Wei-Chun Wang, Paul Brehm  Current Biology  Volume 27, Issue 3, Pages 415-422 (February 2017) DOI: 10.1016/j.cub.2016.12.010 Copyright © 2017 Elsevier Ltd Terms and Conditions

Current Biology 2017 27, 415-422DOI: (10.1016/j.cub.2016.12.010) Copyright © 2017 Elsevier Ltd Terms and Conditions

Figure 1 Motoneuron-Muscle Paired Recordings Reveal Differences in Strength and Reliability of Motoneuron Outputs (A) Schematic of a 5 dpf larval zebrafish. Light gray shading marks the brain and spinal cord. Recordings were done at body segment 11–15. (B) Experimental preparation for paired recordings on a secondary motoneuron (SMn) and a fast muscle cell, filled with Alexa Fluor 488 and 568, respectively. (B1) Focal plane of the SMn cell body (arrowhead) and its main axon (asterisk). Dashed lines mark the boundaries of the spinal cord. (B2) A more superficial focal plane with the motoneuron axonal projections and boutons that overlap with the recorded muscle cell. (C–E) Schematics (left) and example traces (right) of whole-cell current-clamp recordings from a spinal motoneuron and whole-cell voltage-clamp recordings from a muscle cell (C, primary motoneuron [PMn]-fast muscle; D, SMn-fast muscle; E, SMn-slow muscle). The averaged waveform of motoneuron action potentials (APs; n = 10; firing at 1 Hz) is shown in black. The input resistance of the recorded motoneuron is indicated. Individual evoked endplate currents (EPCs; n = 10; gray) and their averaged waveform (colored) are temporally aligned to the peak of motoneuron action potentials to show the variability in amplitude. Note the different scales for EPC amplitude. The round-shaped EPCs in the example SMn-slow muscle recording are filtered inputs from neighboring muscle cells electrically coupled to the recorded muscle cell [17]. The red trace represents the averaged waveform of all ten individual current responses in (E), including responses to direct synaptic inputs (fast responses), electrically coupled inputs (slow responses), and failures. (F) Averaged EPC waveforms of the three example recordings in (C)–(E) are temporally aligned to the peak of motoneuron action potentials. (G) Quantal content of motoneuron outputs versus input resistance (Rin) of motoneurons. The number of recordings is indicated in parentheses. ∗∗p < 0.01 following Spearman’s rank test (ρ is shown). (H) Reliability of motoneuron action potentials eliciting EPCs versus motoneuron Rin. ∗∗p < 0.01 following Spearman’s rank test (ρ). See also Figure S1. Current Biology 2017 27, 415-422DOI: (10.1016/j.cub.2016.12.010) Copyright © 2017 Elsevier Ltd Terms and Conditions

Figure 2 Secondary Motoneuron-Elicited EPCs in Both Fast and Slow Muscles Potentiate during High-Frequency Motoneuron Firing (A–C) Top: an example trace of muscle EPCs in response to 100 Hz motoneuron firing for 4 s in PMn-fast muscle (A), SMn-fast muscle (B), and SMn-slow muscle (C) paired recordings. Bottom: group data showing the mean amplitude of evoked EPCs during 100 Hz motoneuron firing (200 ms/20 event bin; normalized to the first bin). Gray lines indicate individual recordings; the thicker line with shading indicates mean ± SD. The inset in (A) shows the first six PMn-elicited EPCs. (D) The potentiation ratio during 100 Hz motoneuron firing plotted against motoneuron input resistance (Rin). The dashed line is drawn at 1, representing the mean EPC amplitude of the first bin. ∗∗p < 0.01, one-way ANOVA with post hoc Tukey-Kramer test. (E–G) Evoked muscle EPCs in response to 1 Hz, 30 s motoneuron firing before (Pre) and after (Post) 100 Hz, 4 s motoneuron firing for PMn-fast muscle (E), SMn-fast muscle (F), and SMn-slow muscle (G) paired recordings. Top: example traces of the first ten individual evoked EPCs (gray) and their averaged waveforms (black). EPCs are temporally aligned to the peak of motoneuron action potentials. Bottom: group data showing the averaged amplitude of evoked EPCs in each 5 s/five event bin, normalized to the first “pre” 1 Hz bin (gray lines, individual recordings; thicker line with shading, mean ± SD). ∗p < 0.05, paired t test; “post” 1 Hz bins were compared to the last “pre” 1 Hz bin. (H) The potentiation ratio (EPC amplitude ratio of the first “post” 1 Hz bin and the last “pre” 1 Hz bin) plotted against motoneuron Rin. Figure descriptions are the same as in (D). See also Figures S2A and S2B. Current Biology 2017 27, 415-422DOI: (10.1016/j.cub.2016.12.010) Copyright © 2017 Elsevier Ltd Terms and Conditions

Figure 3 Secondary Motoneuron-Elicited EPCs Potentiate during Repeated Firing Using a Template Firing Pattern that Mimics Natural Swimming (A) Whole-cell current-clamp recording from a primary motoneuron (PMn) during a high-frequency fictive swim bout elicited by a brief electrical stimulus to the tail. This PMn firing pattern was used as template 1. Mean swim burst frequency and motoneuron firing frequency within a swim burst are indicated in gray and black, respectively. (B) Whole-cell current-clamp recording from a secondary motoneuron (SMn) during a light-elicited, lower-frequency fictive swim bout (template 2). (C and D) Paired patch recordings from a motoneuron (black; PMn in C and SMn in D) and a fast muscle cell (gray); 1 ms depolarizing currents were injected into the motoneuron under current clamp to fire action potentials with proper intervals mimicking the template firing pattern (template 1 for PMns; template 2 for SMns). (E and F) Repetitive templates 1 and 2 were applied to PMns (E) and SMns (F), respectively, to mimic motoneuron firing during repeated swim bouts. An example trace of elicited EPCs in response to 23 mimic swim bouts is shown in gray. EPCs of the first (blue) and last (red) swim bouts are expanded to show the amplitude change for individual EPCs. Black tics mark the timing of motoneuron action potentials. (G and H) The total area under EPC waveforms (charge transfer) corresponding to each mimic swim bout for PMn-fast muscle (G) and SMn-fast muscle (H) paired recordings (gray lines, individual recordings; thicker black line with shading, mean ± SD). The EPC area was normalized to the area of the first swim bout. See also Figures S2C and S3. Current Biology 2017 27, 415-422DOI: (10.1016/j.cub.2016.12.010) Copyright © 2017 Elsevier Ltd Terms and Conditions

Figure 4 Secondary Motoneurons More Reliably Elicit Fast Muscle Action Potentials and Muscle Contractions after Repeated Motoneuron Firing with the Template Pattern (A–C) Example traces of paired current-clamp recordings of a motoneuron and a muscle cell (A, PMn-fast muscle; B, SMn-fast muscle; C, SMn-slow muscle). Top: averaged waveform of motoneuron action potentials (n = 10). Bottom: voltage responses of muscle cells, temporally aligned to the peak of motoneuron action potentials (n = 10). Subthreshold endplate potentials (EPPs) are shown in gray; muscle action potentials are in red. Muscle voltage changes in (A)–(C) share the same scale. (D) The reliability of motoneuron firing eliciting muscle action potentials versus motoneuron input resistance (Rin). ∗∗p < 0.01 following Spearman’s rank test (ρ is shown). (E and F) Muscle voltage change in response to 1 Hz, 30 s motoneuron firing before (Pre) and after (Post) repeated template firing. Templates 1 and 2 were used for PMn-fast muscle (E) and SMn-fast muscle (F) paired recordings, respectively. Top: example traces of the first ten muscle voltage responses temporally aligned to the peak of motoneuron action potentials. Bottom: group data for the ratio of motoneuron action potentials able to elicit muscle action potentials (5 s/five event bin; 1, suprathreshold; 0, subthreshold; gray lines, individual recordings; thicker black line with shading, mean ± SD). (G) Experimental preparation for simultaneous patch recording of a motoneuron and video recording of elicited muscle contractions. The red rectangle marks the region for muscle length measurements in (H). (H) Right: the blue and yellow dots mark the two ends of a fast muscle cell at its relaxed and maximally contracted state, respectively, during a PMn-elicited contraction. Left: the time course of muscle length change during the contraction with blue and yellow arrowheads indicating the relaxed state and maximum contraction, respectively. (I and J) The amount of muscle shortening in response to 1 Hz, 8 s motoneuron firing (I, PMn; J, SMn) before (Pre) and after (Post) repeated template firing. Top: example traces of the muscle length time course (n = 8; gray lines) aligned temporally with their averaged waveform (black line). Bottom: group data showing the maximum muscle shortening, normalized to the original muscle length, in response to each motoneuron action potential (gray lines, individual recordings; thicker black line with shading, mean ± SD). See also Figure S4 and Movies S1 and S2. Current Biology 2017 27, 415-422DOI: (10.1016/j.cub.2016.12.010) Copyright © 2017 Elsevier Ltd Terms and Conditions