Upregulation of excitatory neurons and downregulation of inhibitory neurons in barrel cortex are associated with loss of whisker inputs

Excitatory units in our study are pyramidal neurons and excitatory synapses in the barrel cortices. Their functions and processes/spines’ morphology are analyzed. In terms of the active intrinsic properties, Figure 2 illustrates the ability to convert excitatory inputs into digital spikes measured by inter-spike interval (ISI). These neurons after loss of whisker inputs appear to have higher ability to encode spikes (dark-blue trace in Figure 2A), compared with controls (dark-red). Figure 2B shows ISIs in pyramidal cells from whisker trimming (WT, open symbols) and control mice (filled). ISI values for spikes 1~2 up to 4~5 are 15.21±1.1, 27.16±2.6, 35.32±2 and 39.61±1.8 in the WT neurons (n=15); and 24.15±2.4, 34.9±2.2, 40.78±1.9 and 44.3±2.3 in the controls (n=16). ISI values for corresponding spikes in these two sources of neurons are statistically different (p<0.01). Therefore, the loss of whisker inputs enhances the capability of pyramidal neurons to convert excitatory inputs into digital spikes.

Figure 2C illustrates VGSC-mediated threshold potentials (Vts) at pyramidal neurons. Vts values for spikes 1 to 5 are 30.52±1.56, 40.19±1.1, 40.9±1.9, 39.62±1.88 and 39.56±1.39 in the WT neurons (open symbols; n=15), and are 35.52±1.56, 45.1±2, 44.61±1.68, 45.75±1.97 and 46.19±2.19 in the controls (filled, n=16). Vts values for corresponding spikes are significantly lower in the WT neurons than in the controls (p<0.01). Thus, a loss of whisker inputs reduces the threshold for firing spikes at barrel cortical pyramidal neurons.

Excitatory synaptic transmission was estimated by recording spontaneous excitatory postsynaptic currents (sEPSC) on pyramidal neurons. Figure 3 illustrates the effect of loss of whisker inputs on excitatory synaptic transmission. sEPSCs after whisker trimming appear higher (Figure 3B), compared to controls (Figure 3A). Figure 3C shows cumulative probability vs. sEPSC amplitudes from the WT neurons (open symbols, n=11) and the controls (filled, n=12). Figure 3D shows cumulative probability vs. inter-sEPSC intervals from the WT neurons (open symbols, n=11) and the controls (filled, n=12). Statistical analysis indicates that sEPSC amplitudes and frequencies (1/inter-sEPSC interval) are higher in the WT neurons than in the controls. Thus, the loss of whisker inputs enhances excitatory synaptic transmission including glutamate release probability and receptor responsiveness on the pyramidal neurons of the barrel cortices.

In terms of morphological changes of barrel cortical pyramidal neurons, we analyzed the densities of processes and spines, which reflected their capacity to receive the excitatory inputs. The processes and spines on pyramidal neurons were accounted from the images photographed by confocal microscope. Figure 4 shows the processes on apical and basal dendrites. The number of processes appears to change on the dendrites of pyramidal neurons from whisker-trimmed mice (bottom panels in Figure 4A), compared with those from controls (tops in Figure 4A). Statistical analyses in Figure 4B~D illustrate the number of processes on apical dendrites, primary and secondary processes on somata (basal dendrites) from the WT mice and the controls. The processes per 100 μm on apical dendrites are 6.8±0.54 in the WT neurons and 8.24±0.33 in the controls (Figure 4B; p<0.05, n=18). The primary processes of basal dendrites are 8.33±0.27 in the WT neurons and 7.46±0.24 in the controls (Figure 4C; p<0.05, n=18). The secondary processes from the basal dendrites are 15.78±0.54 in the WT neurons and 13.46±0.53 in the controls (Figure 4D; p<0.01, n=18). Thus, the area to receive synaptic inputs increases on the basal dendrites of pyramidal neurons, but decreases on their apical dendrites, after loss of whisker inputs.

Figure 5 illustrates the density of spines on apical and basal dendrites. The number of spines appears to increase on the dendrites of pyramidal neurons from the WT mice (bottom panels in Figure 5A), compared to those from the controls (tops in Figure 5A). Statistical analyses in Figure 5B-C show the number of spines on the apical and basal dendrites of pyramidal cells from the WT mice and the controls. Spines per 10 μm on apical dendrites are 9.1±0.16 in the WT neurons and 8.1±0.15 in the controls (p<0.001, n=13). Spines per 10 μm on basal dendrites are 7.15±0.29 in the WT neurons and 6.81±0.31 in the controls (p=0.35, n=12). Therefore, the sites for receiving excitatory presynaptic inputs increase on the apical dendrites of pyramidal neurons after loss of whisker inputs.

In summary, loss of whisker inputs upregulates the functions of excitatory neurons and synapses as well as the areas and sites of receiving synaptic inputs in the barrel cortices. We subsequently studied the influence of loss of whisker inputs on the GABAergic inhibitory neurons and synapses in the barrel cortices.

Inhibitory units in our study included GABAergic neurons and their output synapses in the barrel cortices. Inter-spike intervals and threshold potential were analyzed to indicate active intrinsic properties. Inhibitory synaptic transmission was evaluated by recording spontaneous inhibitory postsynaptic currents (sIPSC) on pyramidal neurons. The processes on GABAergic neurons were accounted from the images taken by a confocal microscope.

Figure 6 illustrates the capability of converting excitatory input into digital spikes measured by inter-spike interval (ISI). These GABAergic neurons after loss of whisker inputs appear to have lower ability to encode spikes (dark-blue trace in Figure 6A), compared to the controls (dark-red). ISI values for spikes 1~2 up to 4~5 are 17.43±0.64, 20.41±0.7, 22.37±0.67 and 24.77±0.97 in the WT neurons (open symbols in Figure 6B, n=15); and 14.88±0.98, 17.2±1, 18.46±1.0 and 19.3±0.99 in the controls (filled, n=16). ISI values for corresponding spikes in two sources of neurons are statistically different (p<0.01). Therefore, the loss of whisker inputs attenuates the capability of GABAergic neurons to convert excitatory inputs intro the digital spikes.

Figure 6C illustrates VGSC-mediated threshold potentials (Vts) at GABAergic neurons. Vts values for spikes 1 up to 5 are 33.1±0.91, 38.86±0.78, 40.1±0.89, 40.52±0.91 and 40.75±1.1 in the WT neurons (open symbols; n=15), and are 30.69±1.27, 34.15±1.26, 34.88±1.22, 36.23±1.72 and 36.88±1.56 in the controls (filled, n=16). Vts values for corresponding spikes are statistically higher in the WT neurons than the controls (p<0.01). Thus, loss of whisker inputs attenuates the active intrinsic properties of inhibitory neurons in the barrel cortices.

Figure 7 illustrates the effects of loss of whisker inputs on inhibitory synaptic transmission. sIPSCs after whisker trimming appear to be lower (Figure 7B), compared to the controls (Figure 7A). Figure 7C shows cumulative probability vs. sIPSC amplitudes from the WT neurons (open symbols, n=11) and the controls (filled, n=12). Figure 7D shows cumulative probability vs. inter-sIPSC intervals from the WT neurons (open symbols, n=11) and the controls (filled, n=12). Statistical analysis indicates that sIPSC amplitudes and frequencies (1/inter-sIPSC interval) are lower in the WT neurons than in the controls. Therefore, loss of whisker inputs attenuates the inhibitory synaptic transmission including GABA release probability and receptor responsiveness in the barrel cortices.

In terms of the morphological changes in the barrel cortical GABAergic neurons, we analyzed the densities of processes, which reflected the volume of receiving inhibitory inputs. The number of processes appears to decrease on GABAergic neurons from the WT mice (right panel in Figure 8A), compared with those from the controls (left in 8A). Statistical analyses in Figure 8B-C illustrate the number of primary and secondary processes on somata from the WT mice and controls. Primary processes are 5.27±0.25 in the WT neurons and 6.12±0.23 in the controls (p<0.05, n=15). Secondary processes are 10.13±0.5 in the WT neurons and 11.39±0.59 in the controls (p=0.2). Therefore, the main processes for receiving presynaptic inputs decrease on GABAergic neurons after loss of whisker inputs.

In order to analyze the activities of barrel cortical neurons and synapses relevant to the changes in whiskers’ experience in cell-specific manner, we need the mice whose cortical neurons are labeled by different markers. We cross-matched the mice from strains of C57(Thy1YFP)BL/6N (from He in IBP-CAS) and FVB-Tg(GADGFP)4570Swn/J (Jackson Lab, USA). Pyramidal neurons in C57 mice were genetically labeled by yellow fluorescent protein (YFP), in which the promoter was Thy1 on the upstream of YFP. GABAergic neurons in FVB mice were labeled by green fluorescent protein (GFP), in which the promoter was GAD on the upstream of GFP. Such cross-matched mice possess YFP-labeled pyramidal neurons and GFP-labeled GABAergic neurons in cerebral cortices (Figure 1). The mice in postnatal days 7 were divided into two groups that were whisker trimming on right side and control (intact whiskers), respectively. The whisker trimming was given every day for one week with no trimming the furs in the face of mice. During the operation, the mice were placed in home-made cages, in which their running and motion were restricted, but the extensions of their bodies and arms were allowed. The cares were taken including no stress and circadian disturbance to the mice. In addition, the mice with normal whisking and symmetric whiskers were selected for our experiments.

The cortical slices (400 μm) were prepared from the mice with whisker trimming and control. They were anesthetized by inhaling isoflurane and decapitated by guillotine. Slices were cut with a Vibratome in oxygenated (95% O₂ and 5% CO₂) artificial cerebrospinal fluid (ACSF), in which the concentrations (mM) of different elements were 124 NaCl, 3 KCl, 1.2 NaH₂PO₄, 26 NaHCO₃, 0.5 CaCl₂, 4 MgSO₄, 10 dextrose, and 5 HEPES, pH 7.35 at 4°C. The slices were held in the oxygenated ACSF (124 NaCl, 3 KCl, 1.2 NaH₂PO₄, 26 NaHCO₃, 2.4 CaCl₂, 1.3 MgSO₄, 10 dextrose, and 5 HEPES, pH 7.35) at 25°C for 2 hours. A slice was transferred to a submersion chamber (Warner RC-26G) that was perfused with the ACSF oxygenated at 31°C for whole-cell recording [41, 42, 57‐60]. Chemical reagents were from Sigma.

The neurons in the barrel cortical slices are showed GFP-labeling for GABAergic cells and YFP-labeling for pyramidal cells. These neurons in layers II-III were selected for whole-cell recordings under DIC-fluorescent microscope (Nikon FN-E600, Japan), in which the excitation wavelength was 488 nm. GABAergic neurons showed fast spiking without the adaptation in spike amplitude and frequency, typical properties for interneurons [41, 61‐64]. Cortical pyramidal neurons demonstrated regular spikes with the adaptation in their amplitudes and frequency.

Cortical neurons were recorded by an MultiClamp-700B amplifier under voltage-clamp for their synaptic activity and current-clamp for their active intrinsic properties. The electrical signals were inputted into pClamp-10 (Axon Instrument Inc, USA) for the data acquisition and analysis. The output bandwidth in this amplifier was 3 kHz. Pipette solution for studying excitatory events included (mM) 150 K-gluconate, 5 NaCl, 5 HEPES, 0.4 EGTA, 4 Mg-ATP, 0.5 Tris-GTP, and 5 phosphocreatine (pH 7.35; [65]. The solution to record inhibitory synapses contained (mM) 130 K-gluconate, 20 KCl, 5 NaCl, 5 HEPES, 0.5 EGTA, 4 Mg-ATP, 0.5 Tris–GTP and 5 phosphocreatine [66]. These pipette solutions were freshly made and filtered (0.1 μm). The osmolarity was 295~305 mOsmol and pipette resistance was 5~6 MΩ.

The functions of GABAergic neurons were assessed based on their active intrinsic properties and inhibitory outputs [67]. The functional status of their inhibitory outputs were evaluated by recording spontaneous IPSCs (sIPSC) under voltage-clamp on pyramidal neurons in the presence of 10 μM 6-Cyano-7-nitroquinoxaline-2,3-(1H,4H)-dione (CNQX) and 40 μM D-amino-5-phosphonovanolenic acid (D-AP5) in ACSF to block ionotropic glutamate receptors and to isolate IPSCs [66]. 10 μM bicuculline was washed into the slices at the end of experiments to test whether synaptic responses were mediated by GABA_AR, which did block sIPSCs in our experiments. The series and input resistances for all of the neurons were monitored by injecting hyperpolarization pulses (5 mV/50 ms), and calculated by voltage pulses versus instantaneous and steady-state currents. It is noteworthy that the pipette solution with the high concentration of chloride ions makes the reversal potential to be −42 mV. sIPSCs are inward when the membrane holding potential at −65 mV [66].

The functions of pyramidal neurons were assessed based on their active intrinsic properties and excitatory outputs [67]. The functional status of their excitatory outputs were evaluated by recording spontaneous EPSCs (sEPSC) under voltage-clamp on cortical pyramidal cells in presence of 10 μM bicuculline in ACSF to block ionotropic GABA receptors and isolate EPSCs. 10 μM CNQX and 40 μM DAP-5 were added into ACSF perfused into the slices at the end of experiments to test whether synaptic responses were mediated by GluR, which did block sEPSCs in our study. In addition, series and input resistances for all of these neurons were monitored by injecting hyperpolarization pulses (5 mV/50 ms), and calculated by voltage pulses vs. instantaneous and steady-state currents.

Action potentials at these cortical neurons were induced by injecting depolarization pulses, whose intensity and duration were changed based on the aim of experiments. The ability to convert excitatory inputs into sequential spikes was evaluated by inter-spike intervals (ISI) when depolarization pulses (200 ms in the duration and threshold for 10 ms pulse-induced spike for the intensity) were given [68]. Neuronal intrinsic properties in our study included spike threshold potential (Vts) and absolute refractory period (ARP). Vts were the voltages of spike-onsets [43, 63, 69‐71].

Data were analyzed if the recorded neurons had the resting membrane potentials negatively more than −60 mV, and action potential amplitudes more than 90 mV. The criteria for the acceptance of each experiment also included less than 5% changes in resting membrane potential, spike magnitude, and input resistance throughout each experiment. Input resistance was monitored by measuring cellular responses to hyperpolarization pulse at the same values as the depolarization that evoked action potentials. To estimate the effect of whisker trimming on neuronal spikes and synaptic transmission, we measured sEPSC, sIPSC ISI and Vts in the neurons from mice of control and whisker trimming. The differences in sEPSC, sIPSC, ISI and Vts were presented as mean±SE. The comparisons of these data were done by t-test.

The mice in one week after whisker trimming and controls were anesthetized by the intraperitoneal injection of sodium pentobarbital, and were perfused by 4% paraformaldehyde in 0.1 M phosphate buffer solution (PBS) from left ventricle/aorta until the body was rigid. The brains were quickly isolated and fixed in 4% paraformaldehyde PBS for additional 24 hours. Cortical tissues were sliced in the cross section of barrel cortex at 60 μm by a Vibratome. Sections were washed by PBS for 3 times, air-dried and cover-slipped. The images in the structures of YFP-labeled pyramidal neurons and GFP-GABAergic cells in the cross-sections of barrel cortices were photographed under a laser scanning confocal microscopy (Olympus FV-1000, Japan), in which their fluorescent markers were deconvoluted by 510 nm and 540 nm [72].

The structures of these neurons were analyzed by a commercialized software MetaMorph in Meta Imaging Series (ver. 6.1, Universal Imaging Cooperation in Molecular Device). As the brain tissues were sliced in series sections, the counting and analysis in cell structures were able to be done at least from two sections for each of barrels. The analyzed sections were chosen in a manner of one section from every two in order to prevent the influence of cells that crossed the neighboring sections on the analysis. In the analyses of dendrites, the primary processes (branches from somata) and secondary ones (branches from primaries) of pyramidal and GABAergic neurons were measured in each of barrel sections. In pyramidal neurons, the analyses of their dendrites included the apical and basal dendrites [41]. The spines were the protrusion extended from on the dendrites, which were accounted as spines per 10 μm.