Complex motor skills in vertebrates require specialized upper motor neurons with precise action potential (AP) firing. To examine how diverse populations of upper motor neurons subserve distinct functions and the specific repertoire of ion channels involved, we conducted a thorough study of the excitability of upper motor neurons controlling somatic motor function in the zebra finch. We found that robustus arcopallialis projection neurons (RAPNs), key command neurons for song production, exhibit ultranarrow spikes and higher firing rates compared to neurons controlling non-vocal somatic motor functions (AId neurons). Pharmacological and molecular data indicate that this striking difference is associated with the higher expression in RAPNs of high threshold, fast-activating voltage-gated Kv3 channels likely containing Kv3.1 subunits. The spike waveform and Kv3.1 expression in RAPNs mirror properties of Betz cells, specialized upper motor neurons involved in fine digit control in humans and other primates but absent in rodents. Our study thus provides evidence that songbirds and primates have convergently evolved the use of Kv3.1 to ensure precise, rapid AP firing in upper motor neurons controlling fast and complex motor skills.

Animal subjects: All of the work described in this study was approved OHSU’s Institutional Animal Care and Use Committee (Protocol #: IP0000146) and is in accordance with NIH guidelines. Zebra finches (Taeniopygia guttata) were obtained from our own breeding colony. All birds used were male and > 120 days post hatch. Birds were sacrificed by decapitation and their brains removed. For electrophysiology experiments brains were bisected along the midline, immersed in ice-cold cutting solution, and processed as described below. For in situ hybridization experiments brains were cut anterior to the tectum and placed in a plastic mold, covered with ice-cold Tissue-Tek OCT (Sakura-Finetek; Torrance, CA), and frozen in a dry ice/isopropanol slurry and processed as described below.

In situ hybridization: To compare mRNA expression levels for KCNC1, KCNC2, KCNC3, KCNC4, KCNMA1, KCNQ2, KCNQ3, KCNA1, KCNA2, and KCNA6 across RA and AId, brains sections (thickness = 10 μm) were cut coronally on a cryostat and mounted onto glass microscope slides (Superfrost plus; Fisher Scientific, Hampton, NH, USA), briefly fixed, and stored at -80°C. For each brain, every 10^th slide was fixed and stained for Nissl using an established cresyl violet protocol. Slides were examined under a bright-field microscope to identify sections containing the core region of RA and AId as previously defined (Nevue et al., 2020). In situ hybridization was conducted using an established protocol (Carleton et al., 2014). Briefly, slides were hybridized under pre-optimized conditions with DIG-labeled riboprobes synthesized from BSSHII-digested cDNA clones obtained from the ESTIMA: Songbird clone collection (Replogle et al., 2008). Specific clones corresponded to GenBank IDs CK302978 (KCNC1; Kv3.1), DV951094 (KCNC2; Kv3.2), DV953393 (KCNC3; Kv3.3), CK308792 (KCNC4; Kv3.4), DV954467 (KCNMA1; BK), FE737967 (KCNA1; Kv1.1), FE720882 (KCNA2; Kv1.2), FE733881 (KCNA6; Kv1.6), DV954380 (KCNQ2; Kv7.2), and CK316820 (KCNQ3, Kv7.2). After overnight hybridization, slides were washed, blocked, incubated with alkaline phosphatase conjugated anti-DIG antibody (1:600; Roche, Basal, Switzerland) and developed overnight in BCIP/NBT chromogen (Perkin Elmer; Waltham, MA, USA). Slides were coverslipped with VectaMount (Vector, Newark, CA, USA) permanent mounting medium, and then digitally photographed at 10X under bright field illumination with a Lumina HR camera mounted on a Nikon E600 microscope using standardized filter and camera settings. Images were stored as TIFF files and analyzed further using the FIJI distribution of ImageJ (Schindelin et al., 2012). We note that high-resolution parasagittal images depicting expression of KCNC1, KCNC2, KCNA1, KCNA6, KCNQ2, and KCNMA1 in RA of adult male zebra finches are available on the Zebra Finch Expression Brain Expression Atlas (ZEBrA; www.zebrafinchatlas.org). All probes were evaluated for specificity by examining their alignment to the zebra finch genome and avoiding probes with significant cross alignments to other loci (as detailed previously (Lovell et al., 2020)).

For each gene, we quantified both expression levels based on labeling intensity (i.e. average pixel intensity) and the number of cells expressing mRNA per unit area. We measured the average pixel intensity (scale: 0-256) in a 200 x 200 µm window placed over each target area in the images of hybridized sections. To normalize signal from background we subtracted an average background level measured over an adjacent control area in the intermediate arcopallium that was deemed to have no mRNA expression. The expression ratio was calculated as RA_OD/AId_OD where values greater than 1 are more highly expressed in RA and values less than 1 are more highly expressed in AId. We also quantified the number of labeled cells in each arcopallial region by first establishing a threshold of expression 2.5X above the background level. Standard binary filters were applied and the FIJI ‘Analyze Particles’ algorithm was used to count the number of labeled cells per 200 µm². 

Slice preparation for electrophysiology experiments: Frontal (180 μm for current clamp and 150 μm for voltage clamp) slices were cut on a vibratome slicer (VT1000, Leica) in an ice-cold cutting solution containing (in mM): 119 NaCl, 2.5 KCl, 8 MgSO₄, 16.2 NaHCO₃, 10 HEPES, 1 NaH₂PO₄, 0.5 CaCl₂, 11 D-Glucose, 35 Sucrose pH 7.3-7.4 when bubbled with carbogen (95% O2, 5% CO₂; osmolarity ~330-340 mOsm). Slices were then transferred to an incubation chamber containing artificial cerebral spinal fluid (aCSF) with (in mM): 119 NaCl, 2.5 KCl, 1.3 MgSO₄, 26.2 NaHCO₃, 1 NaH₂PO₄, 1.5 CaCl₂, 11 D-Glucose, 35 Sucrose pH 7.3-7.4 when bubbled with carbogen (95% O2, 5% CO2; osmolarity ~330-340 mOsm) for 10 min at 37°C, followed by a room temperature incubation for ~30 min prior to start of electrophysiology experiments. 

Patch clamp electrophysiology: RA and AId could be readily visualized in via infra-red differential interference contrast microscopy (IR-DIC) (Fig. 2B). Whole-cell patch-clamp recordings were performed at room temperature (~24°C) unless otherwise indicated. For experiments performed at 40°C, the bath solution was warmed using an in-line heater (Warner Instruments, Hamden, CT). The temperature for these experiments varied up to ± 2°C.

Slices were perfused with carbogen-bubbled aCSF (1-2ml/min) and neurons were visualized with an IR-DIC microscope (Zeiss Examiner.A1) under a 40x water immersion lens coupled to a CCD camera (Q-Click; Q-imaging, Surrey, BC, Canada). Whole-cell voltage- and current-clamp recordings were made using a HEKA EPC-10/2 amplifier controlled by Patchmaster software (HEKA, Ludwigshafen/Rhein, Germany). Data were acquired at 100 kHz and low-pass filtered at 2.9 kHz. Patch pipettes were pulled from standard borosilicate capillary glass (WPI, Sarasota, FL, USA) with a P97 puller (Sutter Instruments, Novato, CA). All recording pipettes had a 3.0 to 6.0 MΩ open-tip resistance in the bath solution. Electrophysiology data were analyzed offline using custom written routines in IGOR Pro (WaveMetrics, Lake Oswego, OR, USA). 

For current clamp recordings, intracellular solutions contained (in mM): 142.5 K-Gluconate, 21.9 KCl, 5.5 Na2-phosphocreatine, 10.9 HEPES, 5.5 EGTA, 4.2 Mg-ATP and 0.545 GTP, pH adjusted to 7.3 with KOH, ~330-340 mOsm. Synaptic currents were blocked by bath applying Picrotoxin (100 μM), DL-APV (100 μM), and CNQX (10 μM) (Tocris Bioscience) for ~3 min prior to all recordings. To initiate current clamp recordings, we first established a giga-ohm seal in the voltage clamp configuration, set the pipette capacitance compensation (C-fast), and then set the voltage command to -70 mV. We then applied negative pressure to break into the cell. Once stable, we switched to the current clamp configuration. Experiments in current clamp were carried out within a 15-min period. AP half-width was defined as the width of the AP halfway between threshold (when the rate of depolarization reaches 10 V/s) and the AP peak. The maximum depolarization and repolarization rates were obtained from phase plane plots generated from averaged spontaneous APs. We noted that the resting membrane potential tended to hyperpolarize to the same degree (~10 mV) in both RA and AId after positive current injections during these current clamp recordings (Alexander, Mitry, Sareen, Khadra, & Bowie, 2019). Recordings in which the resting membrane potential deviated by > 10 mV were discarded. We note that recordings were not corrected for a calculated liquid junction potential of +9 mV. 

Estimated current clamp measurements of membrane capacitance (C_m) were calculated from the measured membrane time constant (t_m; fit with a double exponential at the onset of a negative current injection) and input resistance (R_in; calculated slope of V-I plot) using the equation:

t_m(ms) = R_in(MΩ) • C_m(pF)

For voltage clamp recordings we attempted to limit the speed and space clamp error by 1) Cutting thinner slices (~150 µm) to eliminate more processes, 2) decreased the intracellular K+ concentration to decrease the driving force and 3) compensated the series resistance to 1MΩ. The intracellular solutions contained the following (in mM): 75 K-Gluconate, 5.5 Na2-phosphocreatine, 10.9 HEPES, 5.5 EGTA, 4.2 Mg-ATP, and 0.545 GTP, pH adjusted to 7.3 with KOH, adjusted to ~330-340 mOsm with Sucrose. Rs was compensated to 1 MΩ, the uncompensated Rs = 7.7 ± 0.4 MΩ (Mean ± SE; N=12 RAPNs and AId neurons). We ensured exclusion of interneurons by briefly observing the AP waveform in current clamp prior to switching to voltage clamp (Supplementary Fig. 2; (Zemel et al., 2021)). In order to isolate K+ currents, slices were exposed to bath applied CdCl2 (100 μM), TTX (1 μM), Picrotoxin (100 μM), CNQX (10 μM), and APV (100 μM) for ~5 min prior to running voltage clamp protocols. After protocols were applied, TEA (500 μM) was bath applied and the same voltage-clamp protocols were repeated after K+ currents were eliminated. K+ currents were isolated by subtracting the TEA-insensitive current traces from the initial traces. Capacitive currents generated during voltage-clamp recordings were eliminated by P/4 subtraction. Recordings were not corrected for a measured liquid junction potential of +12 mV. 

Morphology: For these experiments Biocytin (4mg/ml; Sigma, St. Louis, MO, USA) and Alexafluor 488 (Life Technologies, Carlsbad, CA, USA) were included in the intracellular solution used in current clamp experiments. Upon entering the whole cell current clamp configuration, the current was set to 0 pA and the cell was held in for 20 to 30 minutes at room temperature to allow for complete filling. The electrode was then slowly removed at a diagonal angle as the fluorescence was monitored to determine when the electrode detached. Upon complete separation of the electrode from the cell, the slice was placed in 4% paraformaldehyde overnight. The slice was then washed in PBS with 0.25% Triton-100x (3x for 10 minutes), blocked in a PBS solution containing 1% skim milk (1 hr) and then stained in a PBS solution containing 1% skim milk and avidin conjugated to Alexafluor 594 (1:200, Life Technologies, Carlsbad, CA, USA) for 2 hrs. The slice was then washed in PBS with 0.25% Triton-100x (3x for 10 minutes) before a final wash in PBS followed by mounting on a cover slip. Images were captured with a Zeiss LSM 980 Airyscan 2 confocal microscope. 

Maximum projections of the images were made and analyzed in FIJI-Image J (Schindelin et al., 2012) to calculate the 2-D soma area, dendritic complexity (D. A. Sholl, 1956), and estimates spine density (as measured from averaged 30 µm stretches from multiple tertiary branches in each cell) shown in Fig. 1. In a subset of these neurons (3 RAPNs and 2 AId neurons) we used the recently developed, open-source software, ShuTu (Jin et al., 2019), to reconstruct the morphologies in 3-D (see examples in the Supplementary Fig. 1 A-B; Video1-2). 3-D renderings were expanded 2-fold in the Z-direction to allow for better visualization of cellular processes. Despite the planar appearance of reconstructed neurons, previous imaging work in RAPNs and hippocampal neurons suggest these dendrites tile equally across planes (Spiro et al., 1999) and have been disproportionately compressed in the z-axis as a result of tissue processing (Pyapali, Sik, Penttonen, Buzsaki, & Turner, 1998) respectively.  We used ShuTu’s automatic reconstruction script to trace high-contrast neurites filled with biocytin and to estimate their diameters. For low-contrast, broken or occluded processes, we manually corrected the reconstruction in ShuTu’s GUI. The reconstructions were stored in SWC format. We used ShuTu’s GUI to manually trace all visible dendritic spines. We annotated spines with thin necks and mushroom-like shapes using single SWC points projected away from the dendrite and following the spine’s neck direction. For spines with curved necks, irregular shapes or filopodia-like spines, we used multiple SWC points tracing their entire extent. We estimated spine areas in two ways. We treated spines annotated by a single SWC point as spheres attached by cylindrical necks with an average radius of 0.5 pixels (0.066 microns). For spines with multiple SWC points, we used the trapezoidal cylinders defined by every pair of consecutive SWC point. We also added a dome cap to the last SWC point. To obtain a continuous profile of spine density along each dendritic branch (Supplementary Fig. 1E), we used a rolling window of width 20 SWC points (window length fluctuated between 5 and 39 µm, with an average of 17 µm for all cells) and of step size 1 SWC point. From the continuous spine density profile, we computed the average spine density of each dendritic segment (Supplementary Fig. 1C). We excluded from the density estimation of Supplementary Fig. 1C segments branching off the soma with fewer than 10 spines and any segment shorter than 20 microns. We estimated the total dendritic area and length of each cell (Supplementary Table 1) from the reconstructions by constructing trapezoidal cylinders for each pair of consecutive SWC point. Axonal area and length were estimated in the same way, although we were only able to connect a small fraction of axon cable to the soma for 3 neurons. For 2 neurons we were not able to trace any portion of the axon (Supplementary Table 1).

We estimated the surface area of the soma of each neuron by triangulating its 3-D structure. We first used ShuTu’s GUI (Jin et al., 2019) to manually trace binary masks of the somas in all slices. Next, we traced the contours of the masks with edge detecting filters. We then took a fixed fraction of 20 equally spaced points in each contour in order to generate polygonal contours.  All polygonal contours in the end had the same number of edges. Next, we applied a 3-point moving average filter along each sequence of vertex in the Z direction to smooth out rough edges. We then calculated the areas of the triangles formed by the edges and vertices of each adjacent pair of polygonal contours. We added to this number the surface areas of the first and last non-zero binary masks to represent the caps. We also corrected for the areas of the surface patches where the dendrites attach to the soma. Since these surface patches are small, we approximated them by the cross-sectional area of the first dendrite SWC point connected to the soma.

We estimated the volumes of the dendrites using 3-D binary masks which we constructed from the SWC structure. First, we re-scaled the SWC along the Z-axis to match the scale in the XY plane. Next, we interpolated the space between each connected pair of SWC points by trapezoidal cylinders with radii equal to the SWC points’ radii. The binary mask assigns ones to voxels intersecting either a SWC point or a trapezoidal cylinder and zeros otherwise. The volume of the dendrite is the fraction of active voxels in the mask multiplied by the volume of the voxel.

We estimated the volumes of the somas using an adaptation of our method for the surface areas.  Starting from the polygonal contours representing the Z-stack, we used the polygons’ vertices and baricenters to split the soma’s volume into a collection of tetrahedrons. The vertices of an edge in a slice, the vertices of the closest edge in the next slice, and the baricenters of the two contours define 4 tetrahedrons for which the volume can be easily calculated. The total volume of the soma is the sum of the volumes of the individual tetrahedrons.

We approximated the volumes of the spines directly from the SWC structure. Since the spines are small, we estimated their volumes to be approximately the sum of the volumes of the individual SWC points.

Comparative genomics of the KCNC/Kv3 (Shaw-related) potassium channel genes: To identify the full set of genes comprising this gene family in zebra finches, we first retrieved all genes annotated as KCNCs (voltage gated channel subfamily C members) from the latest RefSeq database for zebra finch (Annotation Release 106; GCF_003957565.2 assembly). We next retrieved the similarly annotated genes in other selected songbird and non-songbird avian species, observing the immediate synteny and correct cross-species BLAST alignments. To verify the correct orthology of avian genes to corresponding members of this gene family in mammals, we conducted cross-species BLAST searches, noting the top-scoring reciprocal cross-species alignments as well as conserved synteny as orthology criteria.  To build cladistic trees for evolutionary inferences, we also included (as outgroups to birds and mammals) representative extant organisms from selected branches of major vertebrate groups, as appropriate, including non-avian sauropsids (crocodiles, turtles, lizards), amphibians, and bony fishes.

Pharmacologic compounds: DL-APV, Picrotoxin, CNQX, XE991 and iberiotoxin were purchased from Tocris Biosciences (Bristol, United Kingdom). TEA and 4-AP were purchased from Sigma -Aldrich (St. Louis, Missouri, USA). α-DTX was purchased from Alomone labs (Jerusalem, Israel). AUT5 was provided as a gift from Autifony Therapeutics (Stevenage, United Kingdom).

Electrophysiology data were analyzed offline using IgorPro software (Wave-metrics). Statistical analyses were performed using Prism 4.0 (GraphPad). In situ hybridization data was analyzed using FIJI Image J (NIH). Morphology analysis was done using ShuTu software (http://personal.psu.edu/dzj2/ShuTu/).