Abstract

Excitatory connections between neocortical layer 4 (L4) and L6 are part of the corticothalamic feedback microcircuitry. Here we studied the intracortical element of this feedback loop, the L4 spiny neuron-to-L6 pyramidal cell connection. We found that the distribution of synapses onto both putative corticothalamic (CT) and corticocortical (CC) L6 pyramidal cells (PCs) depends on the presynaptic L4 neuron type but is independent of the postsynaptic L6 PC type. L4 spiny stellate cells establish synapses on distal apical tuft dendrites of L6 PCs and elicit slow unitary excitatory postsynaptic potentials (uEPSPs) in L6 somata. In contrast, the majority of L4 star pyramidal neurons target basal and proximal apical oblique dendrites of L6 PCs and show fast uEPSPs. Compartmental modeling suggests that the slow uEPSP time course is primarily the result of dendritic filtering. This suggests that the dendritic target specificity of the 2 L4 spiny neuron types is due to their different axonal projection patterns across cortical layers. The preferential dendritic targeting by different L4 neuron types may facilitate the generation of dendritic Ca2+ or Na+ action potentials in L6 PCs; this could play a role in synaptic gain modulation in the corticothalamic pathway.

Introduction

In the whisker-to-barrel cortex pathway, granular L4 and infragranular L5B/L6A are the main target layers of thalamocortical projections arising from the ventral posteromedial nucleus (VPM) of the thalamus (Chmielowska et al. 1989; Meyer et al. 2010; Oberlaender et al. 2012). In turn, pyramidal cells in layers 5B and 6A provide direct projections back to both the posteromedial nucleus and VPM of the thalamus, respectively (White and Keller 1987; Killackey and Sherman 2003). Therefore, the thalamic nuclei do not only receive input from peripheral sensory receptors, but their activity is also regulated internally by a corticothalamic feedback circuit: A subset of L5B and L6A PCs project directly back to the thalamus (Hoogland et al. 1987,  1991; Bourassa et al. 1995; Veinante et al. 2000) while a second, indirect pathway acts via a monosynaptic intracortical connection from L4 to L5B/L6A thereby integrating synaptic input to these 2 thalamorecipient cortical layers (Lubke and Feldmeyer 2007; Feldmeyer 2012). The direct feedback pathway has been studied extensively both at the structural (Hoogland et al. 1987,  1991; Bourassa et al. 1995; Veinante et al. 2000; Guillery and Sherman 2002) and functional level (Golshani et al. 2001; Gentet and Ulrich 2004; Reichova and Sherman 2004; Temereanca and Simons 2004; Groh et al. 2008). However, for the indirect pathway, monosynaptic L4–L6A connections have rarely been reported: only one pair in cat visual cortex (Tarczy-Hornoch et al. 1999) and 3 pairs in mouse barrel cortex (Lefort et al. 2009) have been observed and their electrophysiological, especially morphological properties have not been characterized in detail.

To investigate monosynaptic excitatory L4–L6A connections, we made dual whole-cell patch-clamp recordings from presynaptic L4 spiny neurons (spiny stellate cell [SSC] or star pyramidal neuron) and postsynaptic CC or CT L6A PCs in acute rat brain slices of the primary somatosensory (barrel) cortex. Synaptically coupled cell pairs were filled with biocytin during recordings and morphologically reconstructed post hoc for a quantitative morphological analysis and the identification of putative synaptic contacts. Altogether, we recorded 20 monosynaptic L4–L6A excitatory connections for a correlated electrophysiological and morphological analysis. Our findings demonstrate that L4 SSCs exclusively establish synaptic contacts on the distal portion of apical dendrites of L6A PCs and elicit uEPSPs with a slow time course when recorded at the L6A soma. In contrast, L4 star pyramidal neurons (SPNs) preferentially target the basal and proximal apical oblique dendrites of L6A PCs and show fast uEPSPs. The preferential targeting of distal or proximal dendrites suggests that L4 SSCs and SPNs promote the generation of a Ca2+ spike in the tuft of apical dendrite or Na+ action potential (AP) in the axon of L6A PCs, respectively. A concerted activation of both dendritic compartments may serve to combine and integrate thalamocortical inputs with intracortical L6A PC inputs. Coincident activation of both distal and proximal synaptic inputs via L4 SSCs and SPNs, respectively, may modulate the sensitivity or gain (i.e., the rate of AP firing) of L6A PCs to synaptic signals from other cortical or subcortical regions such as the thalamus (Chance et al. 2002; Larkum et al. 2004). This synaptic gain modulation may be relevant for corticothalamic feedback signaling.

Materials and Methods

Slice Preparation

All experimental procedures involving animals were carried out in accordance with the German animal welfare act and the guidelines of the Federation of Laboratory Animal Science Associations (FELASA). The appropriate permissions for killing animals for brain slice experiments were obtained from the Northrhine-Westphalian Landesamt für Natur, Umwelt und Verbraucherschutz (LANUV). Wistar rats (18–28 days old; mean 21 ± 3 days) were anesthetized with isoflurane and decapitated. Oblique coronal slices (350 μm thick) of the primary somatosensory (barrel) cortex were cut at 45° to the midline in ice-cold extracellular solution containing 125 mM NaCl, 2.5 mM KCl, 25 mM glucose, 25 mM NaHCO3, 1.25 mM NaH2PO4, 1 mM CaCl2, and 5 mM MgCl2 (bubbled with 95% O2 and 5% CO2) using a vibrating microslicer (Microm HM 650 V, Walldorf, Germany). Slices were then transferred to a slice keeper and incubated at room temperature (22–24°C) in the same extracellular solution for at least 30 min before use.

Paired Recordings

After a sufficiently long incubation, slices were transferred to a recording chamber and continuously superfused (perfusion speed ∼5 mL/min) with an artificial cerebrospinal fluid containing 125 mM NaCl, 2.5 mM KCl, 25 mM glucose, 25 mM NaHCO3, 1.25 mM NaH2PO4, 2 mM CaCl2, and 1 mM MgCl2 (bubbled with 95% O2 and 5% CO2) at 32–33°C.

Putative postsynaptic L6A PCs in the barrel cortex were randomly selected, all of them being located in the middle or upper part of layer 6A. While we tried to choose L6A PCs that were mainly located underneath large L4 barrels, this was not always possible because of the low apparent connectivity of the L4–L6A connection (see below).

Whole-cell recordings from L6A PCs were then made using patch pipettes (resistance ∼5–8 MΩ) filled with intracellular solution containing 135 mM K-gluconate, 4 mM KCl, 10 mM HEPES, 10 mM phosphocreatine, 4 mM ATP-Mg, and 0.3 mM GTP (adjusted to pH 7.3 with KOH); the osmolarity of the solution was ∼300 mOsmol/L. Biocytin (Sigma, Munich, Germany) at a concentration of 5 mg/mL was added to the internal solution. After patching the postsynaptic L6A PC, putative presynaptic L4 neurons were stimulated in the loose-cell configuration using a “searching” patch pipette of 7–10 MΩ resistance filled with a modified high Na+ solution containing 105 mM Na-gluconate, 30 mM NaCl, 10 mM HEPES, 10 mM phosphocreatine, 4 mM ATP-Mg, and 0.3 mM GTP (adjusted to pH 7.3 with NaOH). When an AP was evoked by loose-seal stimulation, this was in most cases visible as a small deflection on the voltage trace. When the AP resulted in an uEPSP in the postsynaptic L6A PC at short latency (within 10 ms), the searching pipette was withdrawn. The presynaptic L4 neuron was then repatched with a new pipette filled with biocytin-containing intracellular solution. Then APs were elicited in the presynaptic neuron and uEPSPs were recorded in the postsynaptic neuron at the whole-cell (current clamp) mode (see Qi et al. 2015 for more details). Signals were amplified using an EPC10-triple patch-clamp amplifier (HEKA Elektronik, Lambrecht, Germany), filtered at 2.9 kHz and sampled at 10 kHz. During the recording series resistance was continuously monitored and compensated to 80%. Membrane potentials were not corrected for the liquid junction potential.

Synaptic Pharmacology

In some experiments aimed at determining the size of the α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) or N-methyl-d-aspartate (NMDA) component of the uEPSP, either 10 μM 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) or 50 μM (2R)-amino-5-phosphonovaleric acid (AP5) was added to the extracellular solution, respectively. As a control 10–20 sweeps were recorded from L4–L6 pairs in normal extracellular solution. After perfusing neocortical slices with an antagonist (CNQX or AP5)-containing extracellular solution consecutive sweeps were recorded until the uEPSP amplitude became stable. At the end of the experiment slices were again perfused with standard extracellular solution. As many as possible uEPSPs were recorded until a (partial) wash-out of the antagonist was achieved.

Data Analysis

Acquired data were stored on the hard disk of a Macintosh computer for off-line analysis using Igor (WaveMetrics, Lake Oswego, OR). Unitary EPSP amplitude, 20–80% rise time, latency and other properties were determined as described previously (Feldmeyer et al. 1999). In brief, all sweeps were aligned to their corresponding presynaptic AP peaks and averaged to generate the mean uEPSP. Then the uEPSP peak amplitude for each individual sweep was determined within a “peak search window” of 5 ms after the presynaptic AP and averaged over 1 ms; subsequently, a baseline potential measured within a window of similar duration just preceding the uEPSP was subtracted. The paired-pulse ratio was defined as the second uEPSP amplitude divided by the first uEPSP amplitude of the mean uEPSP elicited by paired APs. Failures were defined as events with amplitudes <1.5× the standard deviation (SD) of the noise within the baseline window. In order not to misclassify small responses as failures, care was taken to verify that the failure average was near zero. The coefficient of variation (cv) was calculated as the SD divided by the mean of uEPSP amplitude.

The resting membrane potential (Vrest) of postsynaptic neurons was measured immediately after establishing the whole-cell recording configuration. The membrane time constant was calculated as the average of the time constants obtained by fitting a single exponential function to the potential deflection in response to current injections from −50 to 50 pA in 10 pA steps; the input resistance was calculated as the slope of the linear fit to the current–voltage relationship.

Histological Procedures and Morphological Reconstructions

After recording, slices were fixed at 4°C for at least 24 h in 100 mM PBS, pH 7.4, containing 4% paraformaldehyde. Slices containing biocytin-filled neurons were then processed according to procedures we described previously (Marx et al. 2012).

Biocytin-labelled pairs of neurons were examined under the light microscope at high magnification to identify putative synaptic contacts. Representative pairs were photographed at low magnification to document the dendritic and axonal arborization. Subsequently, neurons were reconstructed with the aid of Neurolucida software (MicroBrightField, Colchester, VT) using an Olympus Optical (Hamburg, Germany) BX61 microscope. The reconstruction provided the basis for the quantitative morphological analysis. Potential synaptic contacts were identified as close apposition of a presynaptic axonal bouton and the postsynaptic dendrite in the same focal plane under the microscope with 100× objective and 10× eyepiece. The synapse-to-soma distance was calculated in the Neuroexplorer software (MicroBrightField) as the path length along the dendrite from the location of synaptic contact to the soma in 3-dimensional (3D) space.

Axonal and Dendritic Density Maps

Axonal and dendritic “length density” maps were constructed using the computerized 3D reconstructions (for details, see Lubke et al. 2003). The length of all axonal and dendritic branches was measured in a 50 × 50 × 50 μm cuboid, projected in the 2D plane yielding a raw density map and subsequently normalized to the maximum axonal and dendritic density, respectively. To align these maps with respect to the barrel centre, barrel borders were identified in the low power (4× objective) bright-field photomicrographs made from the acute brain slice. Spatial low-pass filtering of these maps was performed by 2D convolution with a Gaussian kernel (σ = 50 μm), and continuous 2D density functions were constructed using bicubic interpolation in Mathematica 4.1 (Wolfram Research, Champaign, IL). The presynaptic axonal and postsynaptic dendritic length density maps thus obtained were then multiplied to calculate the predicted “innervation domain” between L4 spiny neurons and L6A PCs. Spatial low-pass filtering of “innervation domain” maps was performed as described above.

Neuronal Simulations

To obtain quantitative estimates how the location of synaptic contacts affects the EPSP time course and amplitude compartmental modeling was performed in the NEURON simulation environment (Hines and Carnevale 1997). 3D somatodendritic reconstructions of representative L6A pyramidal cells and for comparison also L5B pyramidal cells (see Supplementary Material) were obtained using the Neurolucida system (Microbrightfield). ASCII files of these reconstruction files were imported directed into NEURON and subsequently converted to a NEURON-readable file format. Before file import and conversion, a shrinkage correction of the morphological reconstruction was made using published correction factors (Marx et al. 2012). In addition, a correction for the area of missing dendritic spines was included by adding 0.83 μm2 surface area per linear micrometer of length to dendritic compartments (Mainen and Sejnowski 1996). Dendritic branches were divided into small compartments; no compartment exceeded 50 μm in length. A stereotypic axon was attached to the soma (Mainen et al. 1995). Typical ion channels (e.g., Na+, K+, Ca2+ etc.) and their distributions in different somatodendritic compartments were adopted (Mainen and Sejnowski 1996). Hyperpolarization-activated cyclic nucleotide-gated (HCN) channels were not included in the present model because they were probably not activated during uEPSPs. The specific axial resistance, membrane resistance, and membrane capacitance were 150 Ωcm, 30,000 Ωcm2, and 0.75 μF/cm2, respectively. The time step for numerical simulations was 0.025 ms.

The standard modeling approach for conductance-based synapses was used. The excitatory synaptic conductance comprising both AMPA and NMDA components was modeled by double exponential functions with different time constants τ1 and τ2. Here, τ1 = 0.28 ms, τ2 = 1.38 ms, and τ1 = 0.56 ms, τ2 = 44.3 ms for AMPA and NMDA components, respectively. Furthermore, voltage- and Mg2+ dependence of NMDA component was also included in the simulation (Sarid et al. 2007).

Statistical Analysis

Tukey's test was used for statistical comparisons of different data sets. Correlation analysis was performed by calculating Pearson's linear correlation coefficients. For all data, mean ± SD was given.

Results

Morphological and Electrophysiological Characterization of L4–L6 Excitatory Connections

Monosynaptic excitatory connections between L4 SCCs or SPNs with L6A PCs were studied in acute slices of rat barrel cortex (Fig. 1A and Supplementary Table 1). Excitatory neurons were identified as L4 SPNs when they showed an apical dendrite projecting into layer 2/3. In 10 of 20 L4–L6A connections, we recorded exclusively “slow” uEPSPs with a long latency to the presynaptic AP, a very slow rise time of ∼6 ms and a slow decay time (Fig. 1B1). The remaining 10 connections showed “fast” uEPSPs with short latencies, fast rise and decay times (Fig. 1C1). All pairs with a presynaptic L4 SSC showed exclusively slow uEPSPs (n = 6), while most L4 SPN–L6A PC connections showed fast uEPSPs (n = 10 out of a total of 14). Only a minor fraction of the L4 SPN–L6A PC connections (4 out of a total of 14) showed also slow uEPSPs. No age-related difference in the axonal and dendritic domains of L4 SSCs and L4 SPNs, respectively, were observed so that data for each presynaptic cell type were pooled.

Figure 1.

Two types of electrophysiologically and morphologically characterized L4–L6 excitatory connections. (A) Schematic of the recording configuration. (B1, C1) Representative recordings of an L4 SSC–L6 PC (B1) or L4 SPN–L6 PC (C1) connection. Five consecutive uEPSPs (middle traces) recorded in the L6 PC elicited by APs (top trace) in a presynaptic L4 SSC or SPN; mean uEPSP waveform (bottom). (B2, C2) Photomicrograph superimposed by Neurolucida reconstructions of biocytin-filled cell pairs of L4 SSC–L6 PC (B2) and L4 SPN–L6 PC (C2); the axonal domain of both L6 PCs suggests that they are CC projecting pyramidal cells. Note that the apical dendrite of L4 SPN is indicated by an arrow. Presynaptic axons, blue; presynaptic somatodendritic domain, red; post-axons, green; post-somatodendritic domain, black. Barrels are indicated by dashed lines. Light-microscopically identified putative synaptic contacts are marked by light-blue dots or circles; right, contacts at higher magnification.

Figure 1.

Two types of electrophysiologically and morphologically characterized L4–L6 excitatory connections. (A) Schematic of the recording configuration. (B1, C1) Representative recordings of an L4 SSC–L6 PC (B1) or L4 SPN–L6 PC (C1) connection. Five consecutive uEPSPs (middle traces) recorded in the L6 PC elicited by APs (top trace) in a presynaptic L4 SSC or SPN; mean uEPSP waveform (bottom). (B2, C2) Photomicrograph superimposed by Neurolucida reconstructions of biocytin-filled cell pairs of L4 SSC–L6 PC (B2) and L4 SPN–L6 PC (C2); the axonal domain of both L6 PCs suggests that they are CC projecting pyramidal cells. Note that the apical dendrite of L4 SPN is indicated by an arrow. Presynaptic axons, blue; presynaptic somatodendritic domain, red; post-axons, green; post-somatodendritic domain, black. Barrels are indicated by dashed lines. Light-microscopically identified putative synaptic contacts are marked by light-blue dots or circles; right, contacts at higher magnification.

This differential uEPSP time course indicates that the 2 presynaptic L4 neuron types innervate different dendritic compartments of postsynaptic L6A PCs. Light microscopic examination revealed that L4 SSCs established putative synaptic contacts on apical dendritic tufts of L6A PCs (Figs 1B2 and 2A; see also Supplementary Fig. 1). In contrast, most synaptic contacts of L4 SPN–L6A PC connections were found on basal and proximal apical oblique dendrites (Figs 1C2 and 2A). For a few (4 out of 14) L4 SPN–L6A PC connections synaptic contacts were also found in the apical tuft and distal oblique collaterals (Fig. 2A). For L4 SSC–L6A PC connections the synapse-to-soma distance was always >200 µm; for L4 SPN–L6A PC connections contacts were mostly located within a 200 μm range (with the exception of 6 of 24 contacts; Fig. 2B). L4 SSC and SPN–L6A PC connections differed significantly from one another in the 20–80% EPSP rise time and synapse-to-soma distance (Fig. 2C), 2 parameters that showed a strong correlation (r = 0.84; P = 6.49 × 10−11, Pearson's linear correlation test; Fig. 2D). Except for the rise time and latency, the 2 L4–L6A connection subtypes did not differ in other uEPSP properties (Table 1) or postsynaptic neuronal membrane properties (Table 2).

Table 1

Properties of uEPSPs and synaptic contacts in L4–L6 excitatory connections

Presynaptic neuron SSC (n = 6) SPN (n = 14) P value 
20–80% Rise time (ms) 6.70 ± 2.08 (2.94–9.15) 2.51 ± 2.17 (0.77–7.18) 8.30 × 10−4 
Latency (ms) 3.77 ± 1.57 (2.16–6.14) 2.45 ± 1.22 (1.29–4.92) 0.05 
Decay time constant (ms) 37.5 ± 16.1 (25.2–69.1) 30.8 ± 10.9 (11.3–59.3) 0.29 
Amplitude (mV) 0.29 ± 0.16 (0.12–0.55) 0.31 ± 0.19 (0.07–0.74) 0.79 
Paired-pulse ratio 0.95 ± 0.28 (0.45–1.24) 1.03 ± 0.25 (0.64–1.38) 0.54 
Failure rate (%) 25.0 ± 16.4 (10.1–46.7) 28.2 ± 20.7 (1.7–66.7) 0.75 
Coefficient of variation 0.50 ± 0.11 (0.40–0.65) 0.58 ± 0.17 (0.33–0.83) 0.30 
Number of contacts 2.3 ± 1.4 (1–4) 1.7 ± 0.8 (1–4) 0.22 
Geometric distance of contacts (µm) 591 ± 137 (275–801) 201 ± 210 (25–732) 3.93 × 10−7 
Presynaptic neuron SSC (n = 6) SPN (n = 14) P value 
20–80% Rise time (ms) 6.70 ± 2.08 (2.94–9.15) 2.51 ± 2.17 (0.77–7.18) 8.30 × 10−4 
Latency (ms) 3.77 ± 1.57 (2.16–6.14) 2.45 ± 1.22 (1.29–4.92) 0.05 
Decay time constant (ms) 37.5 ± 16.1 (25.2–69.1) 30.8 ± 10.9 (11.3–59.3) 0.29 
Amplitude (mV) 0.29 ± 0.16 (0.12–0.55) 0.31 ± 0.19 (0.07–0.74) 0.79 
Paired-pulse ratio 0.95 ± 0.28 (0.45–1.24) 1.03 ± 0.25 (0.64–1.38) 0.54 
Failure rate (%) 25.0 ± 16.4 (10.1–46.7) 28.2 ± 20.7 (1.7–66.7) 0.75 
Coefficient of variation 0.50 ± 0.11 (0.40–0.65) 0.58 ± 0.17 (0.33–0.83) 0.30 
Number of contacts 2.3 ± 1.4 (1–4) 1.7 ± 0.8 (1–4) 0.22 
Geometric distance of contacts (µm) 591 ± 137 (275–801) 201 ± 210 (25–732) 3.93 × 10−7 

Bold italics indicate significant differences.

Table 2

Passive membrane properties of postsynaptic L6 pyramidal cells

Presynaptic neuron SSC (n = 6) SPN (n = 14) P value 
Vrest (mV) −71.7 ± 1.8 (−74.0 to −69.0) −74.2 ± 2.8 (−79.0 to −70.0) 0.06 
Input resistance (MΩ) 128.7 ± 36.2 (61.8–168.8) 174.1 ± 71.9 (105.4–306.6) 0.16 
Time constant (ms) 17.2 ± 2.5 (13.8–21.1) 24.3 ± 8.9 (15.8–44.5) 0.08 
Presynaptic neuron SSC (n = 6) SPN (n = 14) P value 
Vrest (mV) −71.7 ± 1.8 (−74.0 to −69.0) −74.2 ± 2.8 (−79.0 to −70.0) 0.06 
Input resistance (MΩ) 128.7 ± 36.2 (61.8–168.8) 174.1 ± 71.9 (105.4–306.6) 0.16 
Time constant (ms) 17.2 ± 2.5 (13.8–21.1) 24.3 ± 8.9 (15.8–44.5) 0.08 
Figure 2.

Geometric and functional properties of L4–L6 excitatory connections. (A) Top, pre- and postsynaptic somatodendritic domains of the 2 L4–L6 connection subtypes: SSC (n = 6), SPN (n = 14). Somatodendritic domains of L4 spiny neurons and L6 PCs are shown in red and black, respectively. Putative synaptic contacts are marked by light-blue dots. Somatodendritic domain of L4 SSCs/SPNs and L6 PCs were aligned with respect to the barrel center. Bottom, overlay of averaged uEPSPs of individual L4–L6 pairs for the 2 connection subtypes. uEPSPs were aligned to the peak of the presynaptic AP (dashed line). (B) Distribution of geometric distances of synaptic contacts from the L6 PC somata. (C) Histograms comparing the 20–80% rise time and the synapse-to-soma distance for the 2 L4–L6 connection subtypes; ***P < 0.001, Tukey's test. (D) Plots of synapse-to-soma distance versus 20–80% rise time of mean uEPSPs for the 2 L4–L6 connection subtypes. There is a highly significant linear correlation between rise time and the synapse-to-soma distance. Average values with standard deviations across all pairs within each group are indicated by larger circles. In panels BD, data from L4 SSC–L6 PC connections are given in light-gray, data from L4 SPN–L6A PC connections in dark-gray.

Figure 2.

Geometric and functional properties of L4–L6 excitatory connections. (A) Top, pre- and postsynaptic somatodendritic domains of the 2 L4–L6 connection subtypes: SSC (n = 6), SPN (n = 14). Somatodendritic domains of L4 spiny neurons and L6 PCs are shown in red and black, respectively. Putative synaptic contacts are marked by light-blue dots. Somatodendritic domain of L4 SSCs/SPNs and L6 PCs were aligned with respect to the barrel center. Bottom, overlay of averaged uEPSPs of individual L4–L6 pairs for the 2 connection subtypes. uEPSPs were aligned to the peak of the presynaptic AP (dashed line). (B) Distribution of geometric distances of synaptic contacts from the L6 PC somata. (C) Histograms comparing the 20–80% rise time and the synapse-to-soma distance for the 2 L4–L6 connection subtypes; ***P < 0.001, Tukey's test. (D) Plots of synapse-to-soma distance versus 20–80% rise time of mean uEPSPs for the 2 L4–L6 connection subtypes. There is a highly significant linear correlation between rise time and the synapse-to-soma distance. Average values with standard deviations across all pairs within each group are indicated by larger circles. In panels BD, data from L4 SSC–L6 PC connections are given in light-gray, data from L4 SPN–L6A PC connections in dark-gray.

For comparison, intralaminar L4–L4 (n = 10, including 2 reciprocally connected pairs), L6A–L6A (n = 5) and interlaminar L4-L2/3 (n = 2) (cf. Feldmeyer et al. 2002), L4-L5A (n = 2) (cf. Feldmeyer et al. 2005) excitatory connections were recorded under the same experimental conditions. Like most L4–L6A SPN connections, uEPSPs at these connections showed a short latency and fast rise time in marked contrast to the slow uEPSP time course in L4–L6A SSC connections (Supplementary Fig. 2). Other uEPSP properties (e.g., amplitude, paired-pulse ratio, decay time constant etc.) of these connections are not significantly different from that of the 2 L4–L6 connection subtypes (Supplementary Table 2).

Differential Dendritic Filtering of Synaptic Inputs in Model Neurons

The uEPSP time course of L4–L6A connections is tightly correlated with the location of their synaptic contacts: “slow” uEPSPs correspond to more distal synapses; while “fast” uEPSPs to more proximal synapses (Figs 2 and 3A). To illustrate the mechanism for the generation of 2 uEPSP waveforms, we constructed a simple multi-compartmental model neuron with a typical L6A PC somatodendritic morphology (Fig. 3B). Functionally identical synaptic conductances were injected at different locations on the basal to apical dendrites. Only voltage-gated Na+, K+, and Ca2+ channels were used in the model, all other conductances (e.g., HCN channels) were neglected because of the low probability of activation by one uEPSP. The simulation results showed that, as the synaptic location moves away from the soma, uEPSPs recorded at soma become slower and smaller while proximally generated uEPSPs are fast and large (Fig. 3C).

Figure 3.

Simulation of “slow” and “fast” uEPSPs as a function of the dendritic synapse-to-soma distance. (A) Two experimentally recorded uEPSP waveforms (slow vs. fast) of L4–L6 excitatory connections. (B) Left, morphology of the L6 PC used for compartmental modeling indicating the color-coded locations of synaptic input. Right, simulated somatic uEPSPs (middle black trace) were generated by injecting a synaptic conductance (bottom trace) into the apical tuft (black dot), which fitted well with the experimentally recorded slow uEPSP (black trace in the inset). If the same synaptic conductance was injected into the basal dendrite (red dot), simulated somatic uEPSP (middle red trace) fitted well with the experimentally recorded fast uEPSP at the rising phase (red trace in the inset). Simulated dendritic (top traces) and somatic (middle traces) uEPSP waveforms for other synaptic injection sites are also shown. (C) Calculated uEPSP amplitude at soma versus amplitude at injection site (top) and 20–80% rise time (bottom) as a function of the injection site (synapse) to soma distance for the uEPSP waveforms shown in B. Note the strong amplitude attenuation of locally generated dendritic uEPSPs compared with recorded somatic uEPSPs. Best linear and exponential fits are shown as dashed lines.

Figure 3.

Simulation of “slow” and “fast” uEPSPs as a function of the dendritic synapse-to-soma distance. (A) Two experimentally recorded uEPSP waveforms (slow vs. fast) of L4–L6 excitatory connections. (B) Left, morphology of the L6 PC used for compartmental modeling indicating the color-coded locations of synaptic input. Right, simulated somatic uEPSPs (middle black trace) were generated by injecting a synaptic conductance (bottom trace) into the apical tuft (black dot), which fitted well with the experimentally recorded slow uEPSP (black trace in the inset). If the same synaptic conductance was injected into the basal dendrite (red dot), simulated somatic uEPSP (middle red trace) fitted well with the experimentally recorded fast uEPSP at the rising phase (red trace in the inset). Simulated dendritic (top traces) and somatic (middle traces) uEPSP waveforms for other synaptic injection sites are also shown. (C) Calculated uEPSP amplitude at soma versus amplitude at injection site (top) and 20–80% rise time (bottom) as a function of the injection site (synapse) to soma distance for the uEPSP waveforms shown in B. Note the strong amplitude attenuation of locally generated dendritic uEPSPs compared with recorded somatic uEPSPs. Best linear and exponential fits are shown as dashed lines.

The mean amplitudes of “slow” and “fast” uEPSPs when measured at the L6A PC somata, were similar for our L4–L6A connections (Table 1). This is unexpected because the simulation suggests a substantial dendritic attenuation of distal inputs (Fig. 3). However, uEPSPs generated at synaptic contacts near the L6A PC apical tufts are likely to be strongly under-sampled because only very large uEPSPs will be resolved while small uEPSPs from distal synaptic sites will escape detection because of the strong dendritic filtering (Supplementary Fig. 3). This may account for the relatively low number of L4 SCC–L6A PC connections and explain why the average uEPSP amplitude of L4 SSC and SPN–L6A PC connections are similar. In addition, using specific AMPA and NMDA glutamate receptor antagonists (Supplementary Fig. 4), we could establish that “slow” and “fast” uEPSPs do not have a markedly different receptor composition at the postsynaptic site; the difference in the uEPSP time course is primarily the result of dendritic filtering during uEPSP propagation to the soma (Spruston et al. 1994).

Prediction of the Synaptic Location Based on Axodendritic Overlap

We calculated the potential innervation domain (see Materials and Methods) for L4 SSC–L6A PC connections using the average density maps of L4 SSC axons and L6A PC dendrites. The result suggests that the vast majority of synaptic contacts are established on the distal apical tuft of L6A PCs, in accordance with our paired recording data (Fig. 4A). The innervation domain for L4 SPN–L6A PC connections (Fig. 4B) have a proximal and a distal target region. The actual distribution of putative synaptic contacts for L4 SPN–L6A PC connections largely fulfills this prediction: the majority of L4 SPN–L6A PC synaptic contacts were located on the basal and proximal apical oblique dendrites near L6A PC somata; a few (n = 4 of 24) contacts were established on apical tuft dendrites. Thus, for both L4–L6A connection subtypes the axodendritic overlap appeared to be sufficient to describe the location of synaptic contacts at the cellular (Shepherd et al. 2005; Brown and Hestrin 2009; Yu et al. 2009) or subcellular level (da Costa and Martin 2009; Petreanu et al. 2009; Richardson et al. 2009).

Figure 4.

Density maps of L4–L6 excitatory connections. (A) Maps of axonal (A1) and dendritic (A2) length density of L4 SSCs and L6 PCs. Neuron somata were aligned to the barrel center. Predicted innervation domain (A3) of L6 dendrites by L4 axons were calculated as described in the Materials and Methods. Contours (thick white lines) enclosing 80 and 90% of the integrated density are shown superimposed. Positions of L4 SSC somata (red dots), L6 PC somata (black dots), putative synaptic contacts (light-blue dots) and barrel outlines (thin white dotted lines) are indicated symbolically. (B) Same as in A but for L4 SPN–L6 PC connections.

Figure 4.

Density maps of L4–L6 excitatory connections. (A) Maps of axonal (A1) and dendritic (A2) length density of L4 SSCs and L6 PCs. Neuron somata were aligned to the barrel center. Predicted innervation domain (A3) of L6 dendrites by L4 axons were calculated as described in the Materials and Methods. Contours (thick white lines) enclosing 80 and 90% of the integrated density are shown superimposed. Positions of L4 SSC somata (red dots), L6 PC somata (black dots), putative synaptic contacts (light-blue dots) and barrel outlines (thin white dotted lines) are indicated symbolically. (B) Same as in A but for L4 SPN–L6 PC connections.

To compare the axonal projection patterns of L4 SSCs and SPNs, we reconstructed additional biocytin-filled single L4 spiny neurons (10 for each neuron type). From the data it is apparent that the L4 SSC axonal domain is significantly more dense in layer 4 and supragranular layer 2/3 than in infragranular layers 5B and 6A (Supplementary Fig. 5A). Axons of L4 SPN have shown a lower degree of collateralization throughout layers 2–6. However, L4 SPN axons send more collaterals to infragranular layers 5B and 6A compared with L4 SSC axons (Supplementary Fig. 5B,C) which may explain the differential dendritic targeting of the 2 L4 spiny neuron types.

Pre- but not Postsynaptic Cell-type-Specific Formation of Synapses

We also tested whether L4 spiny neurons preferentially form synaptic contacts with one of the 2 major L6A PC types, that is, CC and CT neurons. These 2 types of L6A PCs could be distinguished reliably based on their axonal projection patterns: the axon of L6A CC PCs forms a dense plexus mainly in layers 5 and 6A of home column where the soma resides and sends a long-range horizontal projection in infragranular layers across several neighbor columns (Fig. 5A); while L6 CT PCs have a largely home-columnar organization of their axon with vertically oriented axonal collaterals that terminate at the layer 4/layer 2/3 border, occasionally giving rise to numerous short collaterals in layer 4 (Zhang and Deschênes 1997; Kumar and Ohana 2008; Velez-Fort et al. 2014). However, slow and fast uEPSPs were recorded independently of the L6A PC type and showed no significant differences in synaptic properties and contact locations (Fig. 5B,C). This indicates that the dendritic target region of postsynaptic L6A PCs is largely determined by the axonal projection pattern of the presynaptic L4 neuron type and is independent of the projection targets of these L6A neurons.

Figure 5.

Comparison of L4–L6 excitatory connections with either CC or CT L6 pyramidal cells as the postsynaptic neurons. (A1) Post-axonal structures of L4-L6 excitatory connections with postsynaptic corticocortical (CC) (left) and corticothalamic (CT) L6 PCs (right). The axonal arbors of L6 PCs are shown in green. Note the columnar confinement of CT axons. (A2) 2D maps of axonal length density of CC and CT L6 PCs. (B) Top, The somatodendritic arbors of L4 spiny neurons and L6 PCs are shown in red and black, respectively. Light-microscopically identified putative synaptic contacts are marked by light-blue dots. Bottom, Overlay of mean uEPSPs of individual L4–L6 pairs within their corresponding connection subtypes. Note that both fast and slow uEPSPs were recorded from postsynaptic CC or CT L6 PCs. (C) Histograms comparing the 20–80% rise time (C1), latency (C2) and the synapse-to-soma distance (C3) for CC and CT connection subtypes. Gray filled/open circles represent data from individual pairs (n = 14 for CC and n = 6 for CT). All parameters were not significantly (ns) different (P > 0.05, Tukey's test) between the 2 connection subtypes.

Figure 5.

Comparison of L4–L6 excitatory connections with either CC or CT L6 pyramidal cells as the postsynaptic neurons. (A1) Post-axonal structures of L4-L6 excitatory connections with postsynaptic corticocortical (CC) (left) and corticothalamic (CT) L6 PCs (right). The axonal arbors of L6 PCs are shown in green. Note the columnar confinement of CT axons. (A2) 2D maps of axonal length density of CC and CT L6 PCs. (B) Top, The somatodendritic arbors of L4 spiny neurons and L6 PCs are shown in red and black, respectively. Light-microscopically identified putative synaptic contacts are marked by light-blue dots. Bottom, Overlay of mean uEPSPs of individual L4–L6 pairs within their corresponding connection subtypes. Note that both fast and slow uEPSPs were recorded from postsynaptic CC or CT L6 PCs. (C) Histograms comparing the 20–80% rise time (C1), latency (C2) and the synapse-to-soma distance (C3) for CC and CT connection subtypes. Gray filled/open circles represent data from individual pairs (n = 14 for CC and n = 6 for CT). All parameters were not significantly (ns) different (P > 0.05, Tukey's test) between the 2 connection subtypes.

Discussion

We characterized monosynaptic excitatory connections between L4 spiny neurons and L6A PCs in rat barrel cortex which are part of the thalamo-cortico-thalamic neuronal feedback microcircuitry between sensory cortices and the thalamus. We found that axons of L4 SSCs exclusively innervate the distal apical tufts of L6A PCs and elicit “slow” uEPSPs in L6A somata. To our knowledge, this is the first recording, in any cortical region, of uEPSPs exclusively originating from the apical tuft region of a pyramidal cell. However, inhibitory interneurons such as Martinotti cells (Kapfer et al. 2007; Silberberg and Markram 2007; Murayama et al. 2009) or L1 interneurons (Jiang et al. 2013; Lee et al. 2015) are also likely to innervate the apical tufts of L5B PCs. Unlike L4 SSCs, the axons of L4 SPNs preferentially target the proximal basal or apical oblique dendrites of L6A PCs and show “fast” uEPSPs (Fig. 6).

Figure 6.

Innervation scheme between L4 spiny neurons and L6 pyramidal cells in the barrel cortex and its possible function. L4 SSC, SPN, and L6 PC are neurons located in the 2 major thalamic (VPM) innervation domains of the neocortex (gray shaded areas). SSCs form synapses in the apical tuft dendrite of L6 PCs closed to the Ca2+ spike initiation zone; SPNs innervate predominately proximal dendrites close to the axon initial segment where Na+ APs are generated. Thus, SSC and SPN may serve in the generation of Ca2+ and Na+ spikes, respectively. They may act synergistically thereby producing AP bursts that could serve to enhance the corticothalamic feedback and/or the corticocortical feedforward signaling. Same color code as in Figure 1.

Figure 6.

Innervation scheme between L4 spiny neurons and L6 pyramidal cells in the barrel cortex and its possible function. L4 SSC, SPN, and L6 PC are neurons located in the 2 major thalamic (VPM) innervation domains of the neocortex (gray shaded areas). SSCs form synapses in the apical tuft dendrite of L6 PCs closed to the Ca2+ spike initiation zone; SPNs innervate predominately proximal dendrites close to the axon initial segment where Na+ APs are generated. Thus, SSC and SPN may serve in the generation of Ca2+ and Na+ spikes, respectively. They may act synergistically thereby producing AP bursts that could serve to enhance the corticothalamic feedback and/or the corticocortical feedforward signaling. Same color code as in Figure 1.

Existence of Monosynaptic Excitatory Connections Between 2 Main Thalamorecipient Layers in Barrel Cortex

So far, synaptic connections between L4 and L6 excitatory neurons have been identified only indirectly in the barrel cortex (Staiger et al. 2000; Wirth and Luscher 2004; Hooks et al. 2011; Tanaka et al. 2011) or other primary sensory cortices (Briggs and Callaway 2001; Zarrinpar and Callaway 2006; Zhou et al. 2010; Velez-Fort et al. 2014). Here, we provide for the first time direct evidence for the existence of such connections and give a detailed, correlated description of their structural and functional properties. Previous data from paired (Beierlein and Connors 2002; Mercer et al. 2005) and single-cell recordings combined with the photo-release of caged glutamate (Hooks et al. 2011) have shown that the main excitatory input to L6 PCs comes from infragranular layers 5B and 6. However, a recent neuroanatomical study demonstrated that L6 corticothalamic PCs are likely to be strongly innervated by L4 neurons just above them (Tanaka et al. 2011), that is, from the same barrel column. Thus, excitatory L4–L6 connections are well placed to control AP firing of L6 PCs, thereby indirectly regulating the activity of thalamus.

Our preliminary data (Qi et al. 2014) demonstrate that L6A PCs also send monosynaptic connections both to L4 spiny neurons and to L4 interneurons through their upper-layer projecting axonal collaterals. According to the axonal projection pattern of L6A PCs, corticothalamic PCs are more likely to be the presynaptic neuron in these L6A-to-L4 excitatory connections. However, this needs to be verified in a future, more extensive study. Furthermore, we also found excitatory L6 PC–L4 interneuron connections (G. Qi and D. Feldmeyer, unpublished results). Whether L6 PCs preferentially form synapses with interneurons compared with spiny neurons in layer 4 as suggested for cat visual cortex (McGuire et al. 1984) will be the subject of future studies. Very recently, the existence of functional connections between L6 CT PCs and L4 excitatory or L4 fast-spiking cells was confirmed using the channelrhodopsin-2-assisted optical stimulation in mouse visual and barrel cortices (Kim et al. 2014). However, the morphology of presynaptic L6 CT PCs and the postsynaptic L4 neurons was not analyzed in this study.

Different Synaptic Connectivity Patterns of L4 SSCs and SPNs and Their Functional Implication for Whisker Signal Processing

L4 spiny neurons establish excitatory synapses onto pyramidal cells located in almost all layers of the barrel cortex while receiving excitatory synaptic input from other L4 spiny neurons and thalamic afferents (Feldmeyer et al. 1999; Lefort et al. 2009). Therefore, layer 4 acts as a “hub” linking the thalamus to extragranular layers in the cortex. L4 spiny neurons are divided into 2 distinct subtypes, SSCs and SPNs that have been suggested to play different functional roles: local sensory signal processing for SSCs and a function of a more global integration to SPNs (Staiger et al. 2004). It has been shown that excitatory inputs onto L4 spiny neurons distinguish between SSCs and SPNs in rat barrel cortex (Schubert et al. 2003) and monkey visual cortex (Yabuta et al. 2001). However, studies of neither intralaminar L4–L4 (Feldmeyer et al. 1999) nor short-range interlaminar L4–L2/3 and L4–L5A (Feldmeyer et al. 2002, 2005) connections found different synaptic innervation patterns for presynaptic SSCs and SPNs, respectively. In marked contrast, the interlaminar L4–L6 connection described here exhibits functional and structural differences in the subcellular neuronal connectivity depending on the presynaptic L4 spiny neuron type, or, more specifically its axonal projection pattern. However, during early postnatal development, this target specificity may not be as pronounced because the fraction of L4 SPNs is substantially higher at more immature stages (Vercelli et al. 1992; Callaway and Borrell 2011; our own unpublished data).

Our data indicate that the L4–L6A uEPSP time course is solely determined by dendritic filtering. However, like L5 and L2/3 PCs, L6A PCs possess distally located Ih channels (Ledergerber and Larkum 2010) which have been shown to convey a dendritic site independence of the EPSP time course, thereby effectively reducing the summation of distal synaptic input at the soma (Williams and Stuart 2000, 2002). However, the effect of Ih channels is small in L6A pyramidal cells (Ledergerber and Larkum 2010) and may even be absent when only small unitary synaptic connections are activated.

Pyramidal cells have a stereotypic dendritic structure: several basal dendrites emanating from the base of soma and one apical dendrite which can be further divided into trunk, oblique, and tuft dendrites projecting to the pial surface (Spruston 2008). It has been found that excitatory synaptic inputs to pyramidal basal dendrites initiate Na+ spikes in the axon initial segment while inputs to apical tuft and distal trunk dendrites evoke Ca2+ spikes (for review, see Major et al. 2013). Our data show that axons of L4 SSCs and SPNs rather selectively innervate the distal apical dendritic tufts and proximal basal dendrites of L6 PCs, respectively. Only following activation of both L4 neuron types these 2 independent dendritic compartments are functionally linked and can act in concert. A recent study demonstrated that the 2 thalamorecipient strata in the neocortex, layers 4 and 5B/6A, are activated in parallel and act independently from one another (Constantinople and Bruno 2013). L4–L6 connections may serve to integrate signaling between these 2 cortical strata receiving dominant thalamic input. Furthermore, the 2 dendritic target regions on the postsynaptic L6 PC are located close to the Ca2+ and Na+ spike generation zones (Ledergerber and Larkum 2010,  2012) (see Fig. 6), as has been shown for L5B PCs (Larkum and Zhu 2002; Larkum et al. 2009). Distal SSC and proximal SPN synapses onto L6 PCs could therefore act synergistically by simultaneously generating Ca2+ and Na+ spikes (Larkum 2013; Hay and Segev 2015). This may result in AP bursts in the L6 PC soma which in turn will increase the gain of corticothalamic feedback and/or the corticocortical feedforward signaling as has been suggested for visual cortex (Olsen et al. 2012). In addition, AP bursts have also been implicated in mechanisms involved in synaptic plasticity, for example, the “Hebbian” associative long-term potentiation at excitatory synapses in the hippocampus (Pike et al. 1999) and the “spike-timing-dependent” synaptic plasticity in the neocortex (Kampa et al. 2006, 2007). This suggests that L6A PCs integrate synaptic inputs in layer 4 similar to the way other L5B and L2/3 PCs integrate synaptic input in layer 1.

Supplementary Material

Supplementary material can be found at: http://www.cercor.oxfordjournals.org/

Funding

G.Q. was supported by a CSC-Helmholtz scholarship for PhD students. This work was supported by the DFG research group on Barrel Cortex Function (BaCoFun).

Notes

We thank Ted Abel, Karlijn van Aerde, Gabriele Radnikow for insightful comments on the manuscript, Arnd Roth for providing software to construct the density maps, and Werner Hucko for excellent technical assistance. Conflict of Interest: None declared.

References

Beierlein
M
Connors
BW
.
2002
.
Short-term dynamics of thalamocortical and intracortical synapses onto layer 6 neurons in neocortex
.
J Neurophysiol
 .
88
:
1924
1932
.
Bourassa
J
Pinault
D
Deschenes
M
.
1995
.
Corticothalamic projections from the cortical barrel field to the somatosensory thalamus in rats: a single-fibre study using biocytin as an anterograde tracer
.
Eur J Neurosci
 .
7
:
19
30
.
Briggs
F
Callaway
EM
.
2001
.
Layer-specific input to distinct cell types in layer 6 of monkey primary visual cortex
.
J Neurosci
 .
21
:
3600
3608
.
Brown
SP
Hestrin
S
.
2009
.
Intracortical circuits of pyramidal neurons reflect their long-range axonal targets
.
Nature
 .
457
:
1133
1136
.
Callaway
EM
Borrell
V
.
2011
.
Developmental sculpting of dendritic morphology of layer 4 neurons in visual cortex: influence of retinal input
.
J Neurosci
 .
31
:
7456
7470
.
Chance
FS
Abbott
LF
Reyes
AD
.
2002
.
Gain modulation from background synaptic input
.
Neuron
 .
35
:
773
782
.
Chmielowska
J
Carvell
GE
Simons
DJ
.
1989
.
Spatial organization of thalamocortical and corticothalamic projection systems in the rat SmI barrel cortex
.
J Comp Neurol
 .
285
:
325
338
.
Constantinople
CM
Bruno
RM
.
2013
.
Deep cortical layers are activated directly by thalamus
.
Science
 .
340
:
1591
1594
.
da Costa
NM
Martin
KA
.
2009
.
Selective targeting of the dendrites of corticothalamic cells by thalamic afferents in area 17 of the cat
.
J Neurosci
 .
29
:
13919
13928
.
Feldmeyer
D
.
2012
.
Excitatory neuronal connectivity in the barrel cortex
.
Front Neuroanat
 .
6
:
24
.
Feldmeyer
D
Egger
V
Lubke
J
Sakmann
B
.
1999
.
Reliable synaptic connections between pairs of excitatory layer 4 neurones within a single “barrel” of developing rat somatosensory cortex
.
J Physiol
 .
521
(Pt 1)
:
169
190
.
Feldmeyer
D
Lubke
J
Silver
RA
Sakmann
B
.
2002
.
Synaptic connections between layer 4 spiny neurone-layer 2/3 pyramidal cell pairs in juvenile rat barrel cortex: physiology and anatomy of interlaminar signalling within a cortical column
.
J Physiol
 .
538
:
803
822
.
Feldmeyer
D
Roth
A
Sakmann
B
.
2005
.
Monosynaptic connections between pairs of spiny stellate cells in layer 4 and pyramidal cells in layer 5A indicate that lemniscal and paralemniscal afferent pathways converge in the infragranular somatosensory cortex
.
J Neurosci
 .
25
:
3423
3431
.
Gentet
LJ
Ulrich
D
.
2004
.
Electrophysiological characterization of synaptic connections between layer VI cortical cells and neurons of the nucleus reticularis thalami in juvenile rats
.
Eur J Neurosci
 .
19
:
625
633
.
Golshani
P
Liu
XB
Jones
EG
.
2001
.
Differences in quantal amplitude reflect GluR4-subunit number at corticothalamic synapses on two populations of thalamic neurons
.
Proc Natl Acad Sci U S A
 .
98
:
4172
4177
.
Groh
A
de Kock
CP
Wimmer
VC
Sakmann
B
Kuner
T
.
2008
.
Driver or coincidence detector: modal switch of a corticothalamic giant synapse controlled by spontaneous activity and short-term depression
.
J Neurosci
 .
28
:
9652
9663
.
Guillery
RW
Sherman
SM
.
2002
.
Thalamic relay functions and their role in corticocortical communication: generalizations from the visual system
.
Neuron
 .
33
:
163
175
.
Hay
E
Segev
I
.
2015
.
Dendritic Excitability and Gain Control in Recurrent Cortical Microcircuits
.
Cereb Cortex
 .
25
:
3561
3571
.
Hines
ML
Carnevale
NT
.
1997
.
The NEURON simulation environment
.
Neural Comput
 .
9
:
1179
1209
.
Hoogland
PV
Welker
E
Van der Loos
H
.
1987
.
Organization of the projections from barrel cortex to thalamus in mice studied with Phaseolus vulgaris-leucoagglutinin and HRP
.
Exp Brain Res
 .
68
:
73
87
.
Hoogland
PV
Wouterlood
FG
Welker
E
Van der Loos
H
.
1991
.
Ultrastructure of giant and small thalamic terminals of cortical origin: a study of the projections from the barrel cortex in mice using Phaseolus vulgaris leuco-agglutinin (PHA-L)
.
Exp Brain Res
 .
87
:
159
172
.
Hooks
BM
Hires
SA
Zhang
YX
Huber
D
Petreanu
L
Svoboda
K
Shepherd
GM
.
2011
.
Laminar analysis of excitatory local circuits in vibrissal motor and sensory cortical areas
.
PLoS Biol
 .
9
:
e1000572
.
Jiang
X
Wang
G
Lee
AJ
Stornetta
RL
Zhu
JJ
.
2013
.
The organization of two new cortical interneuronal circuits
.
Nat Neurosci
 .
16
:
210
218
.
Kampa
BM
Letzkus
JJ
Stuart
GJ
.
2007
.
Dendritic mechanisms controlling spike-timing-dependent synaptic plasticity
.
Trends Neurosci
 .
30
:
456
463
.
Kampa
BM
Letzkus
JJ
Stuart
GJ
.
2006
.
Requirement of dendritic calcium spikes for induction of spike-timing-dependent synaptic plasticity
.
J Physiol
 .
574
:
283
290
.
Kapfer
C
Glickfeld
LL
Atallah
BV
Scanziani
M
.
2007
.
Supralinear increase of recurrent inhibition during sparse activity in the somatosensory cortex
.
Nat Neurosci
 .
10
:
743
753
.
Killackey
HP
Sherman
SM
.
2003
.
Corticothalamic projections from the rat primary somatosensory cortex
.
J Neurosci
 .
23
:
7381
7384
.
Kim
J
Matney
CJ
Blankenship
A
Hestrin
S
Brown
SP
.
2014
.
Layer 6 corticothalamic neurons activate a cortical output layer, layer 5a
.
J Neurosci
 .
34
:
9656
9664
.
Kumar
P
Ohana
O
.
2008
.
Inter- and intralaminar subcircuits of excitatory and inhibitory neurons in layer 6a of the rat barrel cortex
.
J Neurophysiol
 .
100
:
1909
1922
.
Larkum
M
.
2013
.
A cellular mechanism for cortical associations: an organizing principle for the cerebral cortex
.
Trends Neurosci
 .
36
:
141
151
.
Larkum
ME
Nevian
T
Sandler
M
Polsky
A
Schiller
J
.
2009
.
Synaptic integration in tuft dendrites of layer 5 pyramidal neurons: a new unifying principle
.
Science
 .
325
:
756
760
.
Larkum
ME
Senn
W
Luscher
HR
.
2004
.
Top-down dendritic input increases the gain of layer 5 pyramidal neurons
.
Cereb Cortex
 .
14
:
1059
1070
.
Larkum
ME
Zhu
JJ
.
2002
.
Signaling of layer 1 and whisker-evoked Ca2+ and Na+ action potentials in distal and terminal dendrites of rat neocortical pyramidal neurons in vitro and in vivo
.
J Neurosci
 .
22
:
6991
7005
.
Ledergerber
D
Larkum
ME
.
2010
.
Properties of layer 6 pyramidal neuron apical dendrites
.
J Neurosci
 .
30
:
13031
13044
.
Ledergerber
D
Larkum
ME
.
2012
.
The time window for generation of dendritic spikes by coincidence of action potentials and EPSPs is layer specific in somatosensory cortex
.
PLoS ONE
 .
7
:
e33146
.
Lee
AJ
Wang
G
Jiang
X
Johnson
SM
Hoang
ET
Lante
F
Stornetta
RL
Beenhakker
MP
Shen
Y
Zhu
JJ
.
2015
.
Canonical organization of layer 1 neuron-led cortical inhibitory and disinhibitory interneuronal circuits
.
Cereb Cortex
 .
25
:
2114
2126
.
Lefort
S
Tomm
C
Floyd Sarria
JC
Petersen
CC
.
2009
.
The excitatory neuronal network of the C2 barrel column in mouse primary somatosensory cortex
.
Neuron
 .
61
:
301
316
.
Lubke
J
Feldmeyer
D
.
2007
.
Excitatory signal flow and connectivity in a cortical column: focus on barrel cortex
.
Brain Struct Funct
 .
212
:
3
17
.
Lubke
J
Roth
A
Feldmeyer
D
Sakmann
B
.
2003
.
Morphometric analysis of the columnar innervation domain of neurons connecting layer 4 and layer 2/3 of juvenile rat barrel cortex
.
Cereb Cortex
 .
13
:
1051
1063
.
Mainen
ZF
Joerges
J
Huguenard
JR
Sejnowski
TJ
.
1995
.
A model of spike initiation in neocortical pyramidal neurons
.
Neuron
 .
15
:
1427
1439
.
Mainen
ZF
Sejnowski
TJ
.
1996
.
Influence of dendritic structure on firing pattern in model neocortical neurons
.
Nature
 .
382
:
363
366
.
Major
G
Larkum
ME
Schiller
J
.
2013
.
Active properties of neocortical pyramidal neuron dendrites
.
Annu Rev Neurosci
 .
36
:
1
24
.
Marx
M
Gunter
RH
Hucko
W
Radnikow
G
Feldmeyer
D
.
2012
.
Improved biocytin labeling and neuronal 3D reconstruction
.
Nat Protoc
 .
7
:
394
407
.
McGuire
BA
Hornung
JP
Gilbert
CD
Wiesel
TN
.
1984
.
Patterns of synaptic input to layer 4 of cat striate cortex
.
J Neurosci
 .
4
:
3021
3033
.
Mercer
A
West
DC
Morris
OT
Kirchhecker
S
Kerkhoff
JE
Thomson
AM
.
2005
.
Excitatory connections made by presynaptic cortico-cortical pyramidal cells in layer 6 of the neocortex
.
Cereb Cortex
 .
15
:
1485
1496
.
Meyer
HS
Wimmer
VC
Hemberger
M
Bruno
RM
de Kock
CP
Frick
A
Sakmann
B
Helmstaedter
M
.
2010
.
Cell type-specific thalamic innervation in a column of rat vibrissal cortex
.
Cereb Cortex
 .
20
:
2287
2303
.
Murayama
M
Perez-Garci
E
Nevian
T
Bock
T
Senn
W
Larkum
ME
.
2009
.
Dendritic encoding of sensory stimuli controlled by deep cortical interneurons
.
Nature
 .
457
:
1137
1141
.
Oberlaender
M
de Kock
CP
Bruno
RM
Ramirez
A
Meyer
HS
Dercksen
VJ
Helmstaedter
M
Sakmann
B
.
2012
.
Cell type-specific three-dimensional structure of thalamocortical circuits in a column of rat vibrissal cortex
.
Cereb Cortex
 .
22
:
2375
2391
.
Olsen
SR
Bortone
DS
Adesnik
H
Scanziani
M
.
2012
.
Gain control by layer six in cortical circuits of vision
.
Nature
 .
483
:
47
52
.
Petreanu
L
Mao
T
Sternson
SM
Svoboda
K
.
2009
.
The subcellular organization of neocortical excitatory connections
.
Nature
 .
457
:
1142
1145
.
Pike
FG
Meredith
RM
Olding
AW
Paulsen
O
.
1999
.
Rapid report: postsynaptic bursting is essential for “Hebbian” induction of associative long-term potentiation at excitatory synapses in rat hippocampus
.
J Physiol
 .
518
(Pt 2)
:
571
576
.
Qi
G
Radnikow
G
Feldmeyer
D
.
2015
.
Electrophysiological and morphological characterization of neuronal microcircuits in acute brain slices using paired patch-clamp recordings
.
J Vis Exp
 .
in press
.
Reichova
I
Sherman
SM
.
2004
.
Somatosensory corticothalamic projections: distinguishing drivers from modulators
.
J Neurophysiol
 .
92
:
2185
2197
.
Richardson
RJ
Blundon
JA
Bayazitov
IT
Zakharenko
SS
.
2009
.
Connectivity patterns revealed by mapping of active inputs on dendrites of thalamorecipient neurons in the auditory cortex
.
J Neurosci
 .
29
:
6406
6417
.
Sarid
L
Bruno
R
Sakmann
B
Segev
I
Feldmeyer
D
.
2007
.
Modeling a layer 4-to-layer 2/3 module of a single column in rat neocortex: interweaving in vitro and in vivo experimental observations
.
Proc Natl Acad Sci U S A
 .
104
:
16353
16358
.
Schubert
D
Kotter
R
Zilles
K
Luhmann
HJ
Staiger
JF
.
2003
.
Cell type-specific circuits of cortical layer IV spiny neurons
.
J Neurosci
 .
23
:
2961
2970
.
Shepherd
GM
Stepanyants
A
Bureau
I
Chklovskii
D
Svoboda
K
.
2005
.
Geometric and functional organization of cortical circuits
.
Nat Neurosci
 .
8
:
782
790
.
Silberberg
G
Markram
H
.
2007
.
Disynaptic inhibition between neocortical pyramidal cells mediated by Martinotti cells
.
Neuron
 .
53
:
735
746
.
Spruston
N
.
2008
.
Pyramidal neurons: dendritic structure and synaptic integration
.
Nat Rev Neurosci
 .
9
:
206
221
.
Spruston
N
Jaffe
DB
Johnston
D
.
1994
.
Dendritic attenuation of synaptic potentials and currents: the role of passive membrane properties
.
Trends Neurosci
 .
17
:
161
166
.
Staiger
JF
Flagmeyer
I
Schubert
D
Zilles
K
Kotter
R
Luhmann
HJ
.
2004
.
Functional diversity of layer IV spiny neurons in rat somatosensory cortex: quantitative morphology of electrophysiologically characterized and biocytin labeled cells
.
Cereb Cortex
 .
14
:
690
701
.
Staiger
JF
Kotter
R
Zilles
K
Luhmann
HJ
.
2000
.
Laminar characteristics of functional connectivity in rat barrel cortex revealed by stimulation with caged-glutamate
.
Neurosci Res
 .
37
:
49
58
.
Tanaka
YR
Tanaka
YH
Konno
M
Fujiyama
F
Sonomura
T
Okamoto-Furuta
K
Kameda
H
Hioki
H
Furuta
T
Nakamura
KC
et al
2011
.
Local connections of excitatory neurons to corticothalamic neurons in the rat barrel cortex
.
J Neurosci
 .
31
:
18223
18236
.
Tarczy-Hornoch
K
Martin
KA
Stratford
KJ
Jack
JJ
.
1999
.
Intracortical excitation of spiny neurons in layer 4 of cat striate cortex in vitro
.
Cereb Cortex
 .
9
:
833
843
.
Temereanca
S
Simons
DJ
.
2004
.
Functional topography of corticothalamic feedback enhances thalamic spatial response tuning in the somatosensory whisker/barrel system
.
Neuron
 .
41
:
639
651
.
Veinante
P
Lavallee
P
Deschenes
M
.
2000
.
Corticothalamic projections from layer 5 of the vibrissal barrel cortex in the rat
.
J Comp Neurol
 .
424
:
197
204
.
Velez-Fort
M
Rousseau
CV
Niedworok
CJ
Wickersham
IR
Rancz
EA
Brown
AP
Strom
M
Margrie
TW
.
2014
.
The stimulus selectivity and connectivity of layer six principal cells reveals cortical microcircuits underlying visual processing
.
Neuron
 .
83
:
1431
1443
.
Vercelli
A
Assal
F
Innocenti
GM
.
1992
.
Emergence of callosally projecting neurons with stellate morphology in the visual cortex of the kitten
.
Exp Brain Res
 .
90
:
346
358
.
White
EL
Keller
A
.
1987
.
Intrinsic circuitry involving the local axon collaterals of corticothalamic projection cells in mouse SmI cortex
.
J Comp Neurol
 .
262
:
13
26
.
Williams
SR
Stuart
GJ
.
2002
.
Dependence of EPSP efficacy on synapse location in neocortical pyramidal neurons
.
Science
 .
295
:
1907
1910
.
Williams
SR
Stuart
GJ
.
2000
.
Site independence of EPSP time course is mediated by dendritic I(h) in neocortical pyramidal neurons
.
J Neurophysiol
 .
83
:
3177
3182
.
Wirth
C
Luscher
HR
.
2004
.
Spatiotemporal evolution of excitation and inhibition in the rat barrel cortex investigated with multielectrode arrays
.
J Neurophysiol
 .
91
:
1635
1647
.
Yabuta
NH
Sawatari
A
Callaway
EM
.
2001
.
Two functional channels from primary visual cortex to dorsal visual cortical areas
.
Science
 .
292
:
297
300
.
Yu
YC
Bultje
RS
Wang
X
Shi
SH
.
2009
.
Specific synapses develop preferentially among sister excitatory neurons in the neocortex
.
Nature
 .
458
:
501
504
.
Zarrinpar
A
Callaway
EM
.
2006
.
Local connections to specific types of layer 6 neurons in the rat visual cortex
.
J Neurophysiol
 .
95
:
1751
1761
.
Zhang
ZW
Deschênes
M
.
1997
.
Intracortical axonal projections of lamina VI cells of the primary somatosensory cortex in the rat: a single-cell labeling study
.
J Neurosci
 .
17
:
6365
6379
.
Zhou
Y
Liu
BH
Wu
GK
Kim
YJ
Xiao
Z
Tao
HW
Zhang
LI
.
2010
.
Preceding inhibition silences layer 6 neurons in auditory cortex
.
Neuron
 .
65
:
706
717
.