Biocytin

Animal studies were approved by the Northwestern University Animal Care and Use Committee. In vivo stereotaxic injections of retrograde tracers (fluorescent microspheres, Lumafluor, Durham, NC) or AAV viruses encoding channelrhodopsin-2 (AAV-ChR2-Venus) were performed as described (Anderson et al., 2010 (link); Hooks et al., 2013 (link)). Retrograde tracer injections were made in either the contralateral M1 or ipsilateral M2, S1, and S2 of 6- to 7-week-old mice, to label corticocortical projection neurons in M1; brain slices were prepared 3–7 days later and imaged as described below. Viral injections were made in the motor thalamus of 3- to 4-week-old mice, targeting the ventrolateral nucleus, and optogenetic experiments in brain slices were performed ∼3 weeks later.
Brain slice preparation and electrophysiology were performed as previously described (Weiler et al., 2008 (link); Anderson et al., 2010 (link)). Whole-cell patch-electrode recordings were made from neurons in 0.3-mm-thick brain slices containing M1. Data were sampled at 10 kHz (most experiments) or 40 kHz (for intrinsic electrophysiology measurements) and filtered at 4 kHz. For optogenetic experiments, recordings in L4 (and other layers) were generally targeted to neurons with ‘pyramidal’ somata. For experiments aimed at characterizing intrinsic and morphological properties, recordings were targeted to L4 neurons with any soma shape or size except those suggestive of common types of interneurons, particularly basket cells (Apicella et al., 2012 (link)). Data acquisition was controlled by Ephus software (www.ephus.org) (Suter et al., 2010 (link)).
Standard electrophysiological stimulus protocols were delivered to assess intrinsic properties, as previously described (Suter et al., 2013 (link)). For each cell, after measuring the resting membrane potential, current was injected as needed to set the membrane potential to −70 mV, and then stimulus protocols were delivered to measure electrophysiological properties. Spike-frequency accommodation (SFA) ratio was calculated as the ratio of the third to fifth inter-spike interval in the first trace containing ≥6 spikes. Current threshold was defined as the amplitude of the current step that was sufficient to evoke one or more action potentials.
For group analyses of electrophysiological properties (Figure 7, Table 1), statistical comparisons were performed by pooling neurons into laminar groups corresponding to L4, L2/3, and L5A. Based on the results of the circuit analyses (Figures 1–6), we defined ‘L4’ as a thin zone centered on 0.33 (in units of normalized cortical depth) and spanning 0.05 the cortical thickness (i.e., depth range 0.305–0.355). We defined ‘L2/3’ as the laminar zone spanning 0.14–0.26, and ‘L5A’ as the zone 0.37–0.42. These laminar zones were separated by small gaps (0.045 between L2/3 and L4, and 0.015 between L4 and L5A), which reduced (but did not necessarily eliminate) the likelihood that some neurons were wrongly classified due to slice-to-slice variability in layer thicknesses. A small number of neurons thus fell outside these groups and were excluded from group analyses (but not from the plots; all data are plotted as circles in Figure 7).
Glutamate uncaging and laser scanning photostimulation (glu-LSPS) were performed as previously described (Weiler et al., 2008 (link); Wood et al., 2009 (link); Wood and Shepherd, 2010 (link); Shepherd, 2012 (link)), using 3- to 5-week-old mice. As described in ‘Results’, in one set of experiments, we acquired sets of input maps for L2/3 neurons in M1 or S1; in another set, we further analyzed a subset of glu-LSPS mapping data from a previous study (Weiler et al., 2008 (link)). Temporal windowing was used to detect photostimulation sites where the postsynaptic neuron's dendrites were directly stimulated (defined as excitatory events arriving within 7 msec post-stimulus) (Schubert et al., 2001 (link)), and these sites were excluded from analysis (shown in the figures as black pixels).
Optogenetic photostimulation in brain slices was performed as previously described (Kiritani et al., 2012 (link); Hooks et al., 2013 (link)), exploiting the retained photoexcitability of ChR2-expressing long-range axons in slices (Petreanu et al., 2007 (link)) and using conditions (in particular, tetrodotoxin and 4-aminopyridine in the bath solution) that isolate monosynaptic inputs (Petreanu et al., 2009 (link)). Responses to blue-LED photostimulation were sampled in voltage-clamp mode for each neuron, and multiple neurons were recorded per slice. Traces were analyzed to determine the average response amplitude in a 50-msec post-stimulus window. For the set of neurons recorded in the same slice, responses were normalized to the strongest response, resulting in a normalized laminar profile for each slice. Profiles from different slices and animals were pooled for group analyses.
Imaging and morphological reconstructions were performed as previously described (Suter et al., 2013 (link)), by acquiring two-photon image stacks of neurons that had been biocytin filled during slice recordings, fixed, and processed for fluorescent labeling. Three-dimensional reconstructions of axons and dendrites were manually traced (Neurolucida, MBF Bioscience, Williston, VT) and further analyzed using custom Matlab routines (Source code 1) to quantify dendritic and axonal length density, as previously described (Shepherd et al., 2005 (link)).
Images of expression patterns of molecular markers were obtained from the Allen Mouse Brain Atlas (http://mouse.brain-map.org) (Lein et al., 2007 (link)).
Many analyses involved plotting a parameter of interest as a function of cortical depth, providing a laminar profile of that parameter. To facilitate comparisons across slices, we converted from absolute cortical depth (distance from pia) to a normalized scale, with pia defined as zero and the cortex–white matter border defined as one. To the extent that the thicknesses of individual cortical layers vary as a constant fraction of the total cortical thickness, this normalization procedure is assumed to reduce some of the slice-to-slice variability; for example, due to small differences in slice angle.
Unless noted otherwise, statistical comparisons were performed using non-parametric tests (rank-sum or signed-rank tests, as appropriate) with significance defined as p < 0.05. For the group analyses shown in Figure 7 and Table 1, significance was defined as p < 0.05/3 (multiple-comparison correction).

BDA (molecular weight 3000) or biocytin was injected in vivo into the pallium of two species of cichlids of both sexes, N. brichardi (n = 5; standard length: 30–49 mm), A. nigrofasciata (n = 5; standard length: 45–55 mm) and zebrafish D. rerio (n = 2; standard length: 30 and 35 mm). Fish were anesthetized by immersion in water containing 150–180 mg/L MS222 and set in a device for physical restraint. Water containing 70–80 mg/L MS222 was perfused through the gill for aeration and to maintain the anesthetic condition. A dorsal portion of the cranium was opened to expose the brain. For BDA injections, a glass microelectrode (tip diameters: 12–16 µm) filled with 0.75% BDA solution in 0.05 M Tris-HCl-buffered saline (TBS; pH 7.4) was driven into the pallium with a manipulator (MN-3; Narishige). BDA was injected iontophoretically with square current pulses (+5 µA, 0.5 Hz, 50% duty cycle) passed through the electrodes at three to six places of the pallium each for 5 min with a stimulator (SEN-3301; Nihon Kohden, Japan). For biocytin injections, crystals of biocytin were inserted with a minute insect pin into three to six places of the pallium. After the injection, the cranial opening was closed with either plastic wrap (small fish) or dental cement (Ostron II; GC Dental Products, Japan). Postoperative fish were maintained in aquaria for 21–30 h. The fish were then deeply anesthetized with MS222 (over 200 mg/L) and perfused through the heart with 2% PFA and 1% glutaraldehyde in 0.1 M phosphate buffer (PB), pH 7.4. The brains were removed from the skull and post-fixed in the same fixative at 4 °C for 6–8 h.
The fixed brains were cryo-protected by immersion in 0.1 M PB containing 20% sucrose at 4 °C. Cryo-protected brains were embedded in 5% agarose (type IX, ultra-low gelling temperature) containing 20% sucrose and frozen in n-hexane at −60 °C. Then, frontal sections were cut at a thickness of 40 µm on a cryostat and mounted on gelatin-coated glass slides. The sections were dried and washed once with 0.05 M TBS containing 0.1% Tween 20 (TBST) and twice with TBS. To quench non-specific peroxidase activities, sections were steeped in methanol containing 0.3% H₂O₂ and washed three times with TBS and once with 0.03% TBST. Sections were then incubated with a solution of avidin-biotin-peroxidase complex (1:100; VECTASTAIN Elite ABC Standard Kit, Vector Laboratories) overnight. Afterwards, sections were incubated for one hour with 0.05% 3,3’-diaminobenzidine solution in 0.1 M PB containing 0.04% nickel ammonium sulfate and 0.01% H₂O₂. The reaction was stopped by washing four times with TBS, and the sections were counterstained with 0.05–0.1% cresyl violet, dehydrated, and coverslipped with Permount (Fisher Scientific).