Cerebral Cortex Advance Access published online on September 4, 2008
Cerebral Cortex, doi:10.1093/cercor/bhn143
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Interlaminar Differences of Intrinsic Properties of Pyramidal Neurons in the Auditory Cortex of Mice
1 University of Potsdam, Institute for Biochemistry and Biology, 14476 Potsdam, Germany, 2 Charité Universitäts Medizin Berlin, Center for Anatomy, Institute of Cell Biology and Neurobiology, 10115 Berlin, Germany
Address correspondence to Dr. Stefan Huggenberger, Zoological Institute II, University of Cologne, Weyertal 119, 50931 Köln, Germany. Email: st.huggenberger{at}uni-koeln.de.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsDiscussionConclusionsReferences |
|---|
Cortical information processing depends crucially upon intrinsic neuronal properties modulating a given synaptic input, in addition to integration of excitatory and inhibitory inputs. These intrinsic mechanisms are poorly understood in sensory cortex areas. We therefore investigated neuronal properties in slices of the auditory cortex (AC) of normal hearing mice using whole-cell patch-clamp recordings of pyramidal neurons in layers II/III, IV, V, and VI in the current- and voltage clamp mode. A total of 234 pyramidal neurons were included in the analysis revealing distinct laminar differences. Regular spiking (RS) neurons in layer II/III have significantly lower resting membrane potential, higher threshold for action potential generation, and larger Kir and Ih amplitudes compared with layer V and VI RS neurons. These currents could improve temporal resolution in the upper layers of the AC. Additionally, the presence of a T-type Ca2+ current could be an important factor of RS neurons in these upper layers to amplify temporally closely correlated inputs. Compared with upper layers, lower layers (V and VI) exhibit a higher relative abundance of intrinsic bursting neurons. These neurons may provide layer-specific transfer functions for interlaminar, intercortical, and corticofugal information processing.
Key Words: auditory processing • cortical layers • Ih • interspike interval • patch-clamp • T-current
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsDiscussionConclusionsReferences |
|---|
The anatomical similarities between auditory and other sensory cortices have led to the hypothesis that fundamental principles of neocortical structure and connectivity are common to all sensory (and other) cortical areas (Rockel et al. 1980
). However, there are several unusual and perhaps unique anatomical features of the auditory cortex (AC) that complicate attempts to define common rules for sensory neocortical organization. A major difference between the primary AC and other sensory cortices is the projection of many layer III neurons of the AC across the corpus callosum (Winer 1984, 1992). Moreover, small pyramidal neurons in lower layer III and layer IV appear to be the key neurons receiving thalamic afferents in the AC (Smith and Populin 2001), rather than spiny stellate neurons as in visual and somatosensory cortices (Jones 1975; Lund et al. 1979; Simons and Woolsey 1984; Linden and Schreiner 2003; Staiger et al. 2004).
The AC appears to perform sound analysis by combining spatially distributed coincident or time-correlated neuronal responses (Ehret 1997). Thus, temporal-coding is an essential task e.g. for the localization of sound sources and the processing of speech. Accordingly, the kinetics of inhibitory responses are faster in the AC (Buonomano 2000; Hefti and Smith 2002), and the AC may have a unique class of neurons that generate action potentials (APs) only transiently upon depolarization (Metherate and Aramakis 1999). Further unique features of the AC seem to be neurons of layer II/III that are connected to each other through 2 types of synaptic connections distinguished by release probability and synaptic strength (Atzori et al. 2001).
Interlaminar cortical information processing depends not only upon summation of excitatory and inhibitory inputs but also on intrinsic neuronal properties modulating these synaptic inputs (Metherate and Aramakis 1999). More recently, some neuronal properties were compared in layer II/III and layer V (Atzori et al. 2004). There were no obvious interlaminar differences in the passive properties, but neurons in layer II/III displayed a lower activation threshold compared with layer V neurons. Estimates suggested that the number of simultaneously active presynaptic neurons required to reach threshold must be twice as high in layer V neurons (Atzori et al. 2004).
Concerning the control of neuronal excitability, the hyperpolarization-activated inward current (Ih) is of particular interest. This current contributes to several basic physiological processes including resting membrane conductance, stabilization of resting potential and the limitation of temporal summation of excitatory input (Magee 1998, 1999). Therefore, the Ih is thought to represent a key modulator of neuronal excitability (Magee 1999; Berger et al. 2001, 2003). A further important factor of cortical excitability is the transient, low-voltage activated Ca2+ current (T-current; Jahnsen and Llinás 1984; Sutor and Zieglgänsberger 1987), which transiently augments the excitability of neurons and boosts inputs during a very distinct time frame (Deisz et al. 1991).
Despite the studies mentioned above, the data concerning intrinsic properties of the AC are rather rare. Therefore, we investigated intrinsic neuronal properties in slices of the AC of mice using whole-cell patch-clamp recordings. We compared firing behavior during current injections and in particular properties of pyramidal neurons in layers II (including upper layer III), IV (including lower layer III), V, and VI. The data obtained indicate profound interlaminar differences in intrinsic neuronal properties and in the proportion of different neuron types. These characteristics are discussed in the context of potential filtering and amplification mechanisms for intracortical (columnar) processing.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsDiscussionConclusionsReferences |
|---|
Preparation of Mouse AC Slices
Sixty-six NMRI and 10 Balb/C mice between days 25–70 after birth were used for the experiments. Mice possess adult-like hearing abilities approximately 18–20 days after birth (Song et al. 2006) and both strains do not have hearing impairments before the age of 3 months (Zheng et al. 1999; Drayton and Noben-Trauth 2006). The animals were anesthetized with ether and decapitated, as approved under the European Communities Council Directive and institutional animal care guidelines. The skull was opened dorsally and a tissue block, which contained the caudal half of the cerebrum, was removed. The specimens were incubated in ice-cold (about 4 °C) carbogenated artificial cerebrospinal fluid (ACSF) containing NaCl (124 mM), KCl (5 mM), NaH2PO4 (1.25 mM), MgSO4 (2 mM), D(+) glucose (10 mM), NaHCO3 (26 mM), and CaCl2 (2 mM), carbogenized with 95% O2/5% CO2; pH was 7.4 (Deisz 1999; Strauss et al. 2004). After 1–2 min, the block was cut in halves in the mid-sagittal plane and one half was mounted with cyanoacrylate glue on the bottom of a vibratome chamber. Frontal 300-µm slices of the area containing the AC were cut (Caviness and Frost 1980; Paxinos and Franklin 2001; Geissler and Ehret 2004) using a vibratome (TPI, St. Louis, MO). The second half was stored in ice-cold carbogenated ACSF and cut 10–15 min later. The slices were stored in beakers containing ACSF equilibrated continuously with carbogen at 32 °C for 1–2 h. Afterward, the beakers were stored at room temperature.
Electrophysiological Recordings
The slices were transferred into a submerged-type recording chamber continuously perfused with 1–2 ml/min carbogenated ACSF at 32 °C (TC-102 Temperature Controller, Medical Systems, Greenvale, NY). The AC was visually localized within the slice according to the coordinates given in recent literature (Paxinos and Franklin 2001; Geissler and Ehret 2004). Due to variations in the angle of cutting in the frontal plane it is not possible to distinguish primary and secondary areas of the AC. Due to these inaccuracies we refer to the area of recording as AC but most neurons (probably between 90% and 95%) were recorded in the primary auditory area (Paxinos and Franklin 2001; Geissler and Ehret 2004).
Pyramidal shaped neurons of different layers were selected and approached under visual control using infrared video microscopy with a 40x magnification water-immersion objective (Olympus, Hamburg, Germany). For somatic whole-cell patch-clamp recordings an EPC-9 amplifier together with the corresponding PULSE 8.6 software (HEKA Elektronik, Lambrecht, Germany) was used. Patch-clamp pipettes were pulled from borosilicate glass pipettes with an outer diameter of 1.5 mm and inner diameter of 0.86 mm (Science Products, Hofheim, Germany) on a P 87 micropipette puller (Sutter Instruments, Novato, CA). The pipettes were filled with a solution containing: K+ gluconate (120 mM), NaCl (20 mM), MgCl2 (1 mM), CaCl2 (1 mM), N-(2-Hydroxyethyl)-piperazine-N'-2-ethansulfonic acid (10 mM), ethylene glycol-bis-(β-aminoethyl ether)-N,N,N',N'-tetraacetic acid (11 mM), and Mg2+-ATP (2 mM); pH 7.2 adjusted with KOH (Strauss et al. 2004). All substances used were of analytical grade (Merck, Darmstadt, Germany). The resulting pipette resistance was in the range from 4 to 10 M
. A seal of 1 G or more was obtained and recordings were carried out only if the resulting series resistance was less than 20% of the passive membrane resistance (Rm, see below). Recordings were accepted only if changes of the initial access resistance remained within 20%. In current-clamp mode 400 ms current pulses were applied in 25 pA steps starting at –300 pA up to a maximum of 400 pA. During the voltage clamp experiments, the holding potential was set to –60 mV. Hyperpolarizing command potential to –140 mV for 2 s elicited inward rectifying currents. The selective antagonists for Kir (Ba2+) and Ih (ZD7288; Cs+ for both) were used to isolate individual currents. During these experiments, the slices were perfused at least 10 min with carbogenated ACSF including 4 mM CsCl, 1 mM BaCl2, or 0.05 mM ZD7288 (Tocris, Bristol, UK). For bath application of Ba2+ a modified ACSF was used containing no NaH2PO4 and an equimolar amount of MgCl2 instead of MgSO4. During recording, the shape of neurons and the position of pipettes were documented by taking digital images of the video signal (40x magnification; Fig. 1a). After completing recordings from a given neuron, the laminar position of the pipette (in relation to the rhinal fissure; Fig. 1b) was documented by taking digital images of the video signal (10x magnification water-immersion objective). Slices were then formalin-fixed and Nissl-stained to verify the topographical relations of recorded areas.
|
In some cases, 1% biocytin (Sigma-Aldrich Chemie GmbH, Taufkirchen, Germany) was added to the pipette solution. When biocytin containing pipette solutions were used, neurons were recorded for at least 10 min before the pipette was carefully withdrawn. Afterwards, the slices were individually stored for 1 h in carbogenated ACSF at room temperature before transfer to fixation solution (4% paraformaldehyde in 0.1 mM phosphate buffer). Fixed slices were stored at 4 °C, up to one week. For staining, slices were washed 2 h in 0.1 mM phosphate-buffered saline (PBS) and incubated 2 h in a mixture of 2 mg/ml bovine serum albumin in PBS (PBS-A), phenylhydrazine and 0.3% Triton X (1 ml/0.5 µl/30 µl, respectively). Afterwards, slices were washed in PBS (2 x 10 min, 1 x 20 min) and incubated with a horseradish peroxidase streptavidin complex (Perkin Elmer, Boston, MA) in PBS-A (1:200) at 4 °C over night. The next day slices were washed in PBS (2 x 15 min, 1 x 30 min) and preincubated for 15 min in a mixture of Tris(hydroxymethyl)aminomethane (50 mM in H2O)/3,3'-diaminobenzidine tetrahydrochloride (50 mg/ml)/imidazol (1 M) (1 ml/10 µl/10 µl, respectively, pH = 7.6). Staining reaction was started by adding 50 µl of 3% (NH4)2Ni(SO4) and 25 µl of 0.03% H2O2. Fifteen minutes later the slices were washed in PBS (2 x 15 min), mounted on coated glass slides and dehydrated by a standard alcohol series for embedding in Entellan (Merck).
Data Analysis
From current-clamp experiments, the following parameters were determined using the PULSE FIT 8.6 software (HEKA Elektronik): resting membrane potential (Em), membrane time constant (
m), membrane resistance (Rm), firing behavior (initial bursting or regular spiking [RS]), AP threshold (Ethr), AP amplitude (APamp), and AP width (APwidth). Rm was measured at the –100 pA current step. The m was determined by fitting an exponential function to the induced potentials at the –100 pA current step. The Ethr was calculated by the statistical mean between the lowest potential where APs were elicited and the highest potential where spiking did not occur. The activation threshold (Athr) was calculated by the difference of Ethr and Em. APamp was measured from Em to the peak of the first AP in the sweep of the lowest potential where APs were elicited. APwidth was measured at 
APamp of the same AP. The adaptation ratio of interspike intervals (ISIad) was measured in response to the lowest amplitude current pulse that resulted in firing of at least 6 APs (or an initial burst followed by at least 5 APs) throughout the 400 ms pulse. ISIad was calculated as the ratio of the first ISI and the last ISI. To further analyze ISIad, in some cases, positive 1000-ms current pulses were applied in 10 pA steps. To calculate the surface area (SA) of neurons, the membrane capacity (Cm) was determined using Rm and m and divided by the specific capacity of 1 µF/cm2 (Hille 1992).
Neurons included in this analysis met the following criteria: Em between –65 and –85 mV, APamp > 80 mV, APwidth < 2.6 ms, and firing patterns characteristic for pyramidal neurons (Agmon and Connors 1992; Metherate and Aramakis 1999; Rose and Metherate 2005).
Current traces (recorded during hyperpolarizations from –60 to –140 mV) were approximated by double-exponential fits (between 
5 ms after the capacitance peak and at 10 ms before the end of the current trace) to original traces (without leakage subtraction). These double-exponential fits yielded the magnitudes of Kir and Ih and the kinetic (first-order [fast] and second-order [slow] components) of Ih (Fig. 2). Amplitude of the latter was determined by back-extrapolating the current to the start of the voltage step (Sciancalepore and Constanti 1998; Koch and Grothe 2003; Strauss et al. 2004; Fig. 2b); the difference between the total current and Ih was taken as the magnitude of Kir. The validity of this approach was corroborated by pharmacological isolation of the 2 currents by Ba2+ or ZD7288 application (Fig. 2a,c). As a further measure of Ih, the amplitude of the so-called "sag" during a hyperpolarizing current step from 0 to –300 pA was determined (Fig. 2d). Furthermore, the current densities Kir/SA and Ih/SA were determined.
|
To analyze the interlaminar differences of pyramidal neurons in the AC, we took the peculiarities of the AC into account and segregated the layers according to their thalamic and intracortical input (Linden and Schreiner 2003
). Therefore, upper layer III was pooled with layer II (referred to layer II/III) and lower layer III with layer IV (referred to layer III/IV). Layers V and VI were handled separately. Values in the text are given as means ± SD. Univariate f-tests and chi-squared tests were used (SPSS, SPSS Inc., Chicago, IL) to test for significance. P-values smaller than 0.01 were taken to indicate significant differences (otherwise mentioned in the text).
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsDiscussionConclusionsReferences |
|---|
A total of 234 pyramidal neurons of the AC were analyzed. The measurements of all neurons (see below) did not show any correlation with the animal's age or strain so that we pooled these data. Based on the firing pattern to positive current injections, 2 basic response types were distinguished: RS neurons (n = 181, 77.4%) and intrinsically bursting (IB) neurons (n = 53, 22.6%; Fig. 3). Nine neurons were biocytin-labeled during recording and revealed pyramidal shape and dendritic spines (Fig. 1c,d).
|
Comparison of RS and IB Neurons
In IB neurons, depolarizing current pulses at resting potential evoked bursts consisting of a series of 2–5 APs superimposed on a slow envelope of depolarization (Fig. 3a,b). The ISI of the first 2 APs within the burst was nearly constant (< ±0.3 ms) when the depolarizing current was increased (Figs 3a, 4a). After the initial burst, most of these neurons (n = 48) showed sustained firing of APs for the time of the current injection (Fig. 3a). We refer these neurons as "latent bursters" (IBl) because some of these neurons did not generate bursts on application of small depolarizing current pulses but a single first AP followed by a short pronounced after-depolarization (an equivalent after-depolarization is shown in the last burst of trace in Fig. 3b). With larger depolarization, a burst with 2–5 APs occurred. Five IB neurons showed multiple bursts during a 400-ms current injection near threshold (Fig. 3b) which we refer to as repetitive IB neurons (IBr). Four of these IBr neurons were found in layer V (11.5% of neurons measured in layer V; Fig. 4b) and one in layer IV (1.2% of neurons measured in layer III/IV; Fig. 4b). Injecting current pulses >75 pA above Ethr evoked regular firing after the initial burst in these neurons (not shown in figures).
|
RS neurons responded to suprathreshold depolarizing current injection with a sustained and almost regular firing throughout the current duration (Fig. 3c,d), the AP frequency decreased slightly during a given constant current injection, though. Increasing the magnitude of injected current, increased the frequency of APs (Fig. 4a).
The adaptation ratio of the ISI (ISIad) during constant current injection was between 0.2 and 1 in the RS neurons and thus significantly higher compared with IB neurons (ISIad 0.04 to 0.21; Table 1; Fig. 4c). As illustrated in Figure 4c, ISIad of most RS neurons clustered between 0.45 and 0.65. According to this distribution (Fig. 4c), we refer to RS neurons with ISIad values smaller than 0.45 as type I RS or adapting RS neurons (RS I; Figs 3c, 4c), as opposed to type II RS neurons (RS II), having ISIad greater than 0.45 (Figs 3d, 4c). This segregation of pyramidal neuron types becomes clearer when the frequency corresponding to the ISI between the first 2 APs at different current injections is compared. The frequency of the first 2 APs is nearly constant in IB neurons, whereas the following AP frequencies (fAP) became faster when injected currents were increased (Fig. 4a; note that the second fAP is the reciprocal value of the first ISI following the burst). In contrast, the first fAP of RS I neurons increased significantly in relation to both the following AP frequencies and the first fAP of a lower suprathreshold current injection (Fig. 4a). In RS II neurons the first fAP increased only slightly relative to the following frequencies when currents were increased (Fig. 4a). It is important to note that RS I neurons could show a similar firing pattern as some IB neurons but the first ISI of RS I neurons decreased when depolarizing current was increased and the first AP of RS I neurons was followed by a less pronounced after-depolarization compared with IB neurons (Fig. 3c).
|
Considering the marked effects of Ih on neuronal excitability (e.g. Strauss et al. 2004
), we routinely measured the amplitude of the sag as a possible index for the magnitude of Ih (Fig. 2d). The sag ranged from 0.17 to 4.79 mV (1.31 ± 1.05 mV, n = 64) in RS neurons and from 0.21 to 3.97 mV (1.62 ± 0.96 mV, n = 40) in IB neurons, the differences were not significant. However, the limitations of this estimate (cf. Day et al. 2005; Aponte et al. 2006) could be demonstrated by measuring the sag during application of Ba2+ to block Kir channels. This revealed a significant increase of the sag in RS neurons by approximately 1260% (control: 1.1±0.79 mV, Ba2+ application: 13.87 ± 5.66 mV; n = 9, P < 0.001). Therefore, a direct evaluation of the magnitude of the Ih was carried out in voltage clamp (Fig. 2), revealing a comparable Ih magnitude (control: –140.3 ± 61.5 mV, Ba2+ application: –156.3 ± 34.9 mV; n = 9, P = 0.13).
Interestingly, even though Ih and Kir amplitudes were on average not significantly different between RS and IB neurons (Table 1), the ratio of both currents (Ih/Kir) was significantly higher in IB neurons (IB: 0.44 ± 0.26, RS: 0.37 ± 0.18, P < 0.0001). Moreover, the fast time constant of Ih, fast, was significantly higher in IB neurons (Table 1).
Ionic Mechanisms of ISI Adaptation in RS Neurons
The adaptation of the ISI of RS neurons may be due to activation of voltage-dependent K+-currents (Schwindt et al. 1988; Lorenzon and Foehring 1993). We therefore tested the effects of K+-current antagonists on ISIad (Fig. 4d; n = 18 with Cs+, n = 9 with Ba2+, and n = 7 with ZD7288; Sciancalepore and Constanti 1998; Bal and Oertel 2000; Koch and Grothe 2003; Robinson and Siegelbaum 2003; Strauss et al. 2004). Accordingly, the ISIad values are approximately 33% higher with unselectively blocked K+-current by Cs+ application than in the control (Fig. 4d; 0.75 ± 0.12 vs. 0.5 ± 0.12, P < 0.001, n = 18). This increase was caused by an increase of the first ISI, whereas subsequent ISIs (including the second ISI) during a trace of constant current application were essentially unaffected (Fig. 5a, c). Interestingly, the selective Ih antagonist ZD7288 had the same effect regarding ISIad as Cs+ application (Fig. 4d; 0.74 ± 0.08 vs. 0.5 ± 0.09, P < 0.01, n = 5), but Ba2+ had no effect on ISI (Fig. 4d).
|
A considerable fraction of adaptation remained during K+-current block with both Cs+ and ZD7288 and ISIad was still below 1 (n = 2, not shown in figures). We therefore evaluated a possible contribution of T-currents (Jahnsen and Llinás 1984
; Sutor and Zieglgänsberger 1987; Deisz et al. 1991) to the ISIad. To this end we applied the fairly selective T-current antagonist Ni2+ or depolarized the neurons by constant current injection (Em –65 mV) for steady-state inactivation of the T-current. Bath application of Ni2+ resulted in a marked increase of ISIad, from 0.48 ± 0.13 to 0.65 ± 0.16 (n = 20, P < 0.001). Constant current depolarization caused comparable alterations of ISIad from 0.58 ± 0.2 to 0.77 ± 0.05 (n = 33, P < 0.01). On average, the first ISI increased from 30.3 ± 10.5 ms to 40.4 ± 12 ms in Ni2+ (n = 19, P < 0.01) and from 31.1 ± 12 ms to 39.9±11.1 ms with depolarization (n = 38, P < 0.01). The ISI of the last APs was little affected under these conditions, for example, 67.7 ± 16.3 ms in control and 66.2 ± 13.7 ms in Ni2+, obvious from the nearly constant number of APs during the analyzed current pulses (Fig. 5). This indicates that the apparent ISIad of RS neurons consists of 2 distinct components, one with pharmacological properties indicative for Ih and the other with pharmacological and biophysical properties suggestive for the T-type Ca2+-current. Accordingly, it is likely that RS I neurons have a higher expression level of the channels to adjust ISIad. These evaluations are far from a detailed biophysical characterization of the underlying mechanisms but provide at least a useful tool to distinguish between functional different sets of pyramidal neurons. Interlaminar Comparisons
Previous studies of rodent sensory cortex areas (e.g. Silva et al. 1991; Agmon and Connors 1992) showed a high number of IB neurons in deeper layers. We therefore compared the distribution of RS and IB neurons in different layers: layer II/III had significantly the lowest (8.3%), whereas layer VI had the highest percentage of IB neurons (35.5%; Fig. 4b). However, the latter proportion was not significantly different from that in layers III/IV and V (27.6% and 28.6%; Fig. 4b). The distribution of the 2 RS neuron types (cf. Fig. 4c) was not significantly different in layer V and VI but layer II/III had significantly more RS II neurons (69.1%) than deeper layers and layers II/III and III/IV had significantly fewer RS I neurons (22.6% and 19.5%) compared with deeper layers (25.7% and 25.8%; Fig. 4b).
Interlaminar Comparison of RS Neurons
To further analyze the characteristics of RS neurons in the different cortical layers, we compared the 15 standardized measurements and found significant differences in 10 of these measurements (Table 2). Because data of all layers were tested at once for differences of each parameter (and not pair wise for the sake of clarity) by using univariate f-statistics we will describe trends of significantly different measurements below, the complete list of means and standard deviations is given in Table 2. RS neurons of layers II/III and III/IV had lower Em (–76.0 mV and –75.7 mV) than neurons in layers V and VI (–73.8 and –71.7 mV). Correspondingly, the latter had the smallest APamp (99.5 mV; Table 2). Rm had by far the smallest value in layer II/III RS neurons (55.1 M) compared with other layers (Table 2). Ethr was least negative in RS neurons of layer II/III (–49.2 mV), intermediate in layer III/IV (–53.4 mV), and lowest in layers V and VI (–54.2 mV and –54.9 mV; Table 2). Because layer II/III had the most negative Em but the least negative Ethr, Athr was highest in these neurons (26.8 mV) and lowest in neurons of layers V and VI (19.6 and 16.8 mV; Fig. 6a; Table 2). ISIad and APwidth were not significantly different among layers in RS neurons. m had largest values in layer V (10.45 ms) and other layers were not clearly different (3.2–7.4 ms; Table 2). SA was largest in layer II/III (11.5 x 103 µm2) and smallest in layers III/IV and VI (8.8 x 103 µm2 and 8.9 x 103 µm2; Table 2). RS neurons of layer II/III had the largest amplitudes of Kir and Ih (–847.0 pA and –315.6 pA) and neurons in layer III/IV and V had the smallest amplitude of these currents (cf. Table 2). Interestingly, the densities of these currents (Kir/SA, Ih/SA) were not significantly different between RS neurons of different layers but showed the same trend as the current amplitudes (layer II/III neurons had the largest densities and neurons in layer III/IV and V the smallest) except the high values of current densities in layer VI (–93.1 pA/µm2 and –28.5 pA/µm2) due to the small SA of these neurons (cf. Table 2). In addition, RS neurons of layer V had the slowest time course of Ih (fast 21.9 ms), RS neurons of the other layers exhibited much faster and indistinguishable time courses (9–11 ms). The slow time constant of Ih (slow) was comparable in all layers.
|
|
Figure 6b demonstrates the dependency of Rm on the sum of Kir and Ih amplitudes: higher Kir and Ih amplitudes were correlated with lower Rm values and vice versa.
The function of the fit is Rm = 164.12 e0.0008 (Kir+Ih) and its correlation coefficient r2 = 0.54. Furthermore, this plot reveals that the significantly lower Rm values in layer II/III (crosses in Fig. 6b) were correlated with their high Kir and Ih amplitudes. This relationship is further substantiated by the effects of Cs+ on Rm: in layer II/III RS neurons, Cs+ increased Rm by 633% ± 597% (n = 16), in layer III/IV by 437% ± 288% (n = 14; P < 0.05; layer V and VI neurons have not been tested with Cs+ so far). This indicates that the low values of Rm in layer II/III RS neurons were mediated by Cs+-sensitive currents.
15 of the 234 neurons (6.8%) measured had Ih amplitudes smaller than 50 pA. These neurons were excluded from our statistics regarding differences of Ih parameters. However, eleven of these 15 neurons were found in layers III/IV and V, which had the lowest Ih amplitudes compared with layers II/III and VI (Table 2). Accordingly, it is likely that the neurons with these low Ih amplitudes are not mere artifacts but may be examples of the smallest range of Ih amplitudes in the AC.
Interlaminar Comparison of IB Neurons
In IB neurons, interlaminar differences were less pronounced than in RS neurons. Only Ethr was significantly different (Table 2). Ethr was lowest in layer V neurons (–59.7 mV), intermediate in layers III/IV and VI (–55.7 and –54.4 mV) and highest in layer II/III (–48.0 mV).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsDiscussionConclusionsReferences |
|---|
Methodological Considerations
Although RS neurons were formerly categorized as adapting or nonadapting (Agmon and Connors 1992), we found a continuum of ISIad values (between 0.2 and 1) in these neurons which is similar to recent results (Metherate and Aramakis 1999; Rose and Metherate 2005). However, most of the neurons analyzed by Rose and Metherate (2005) had low adaptation ratios (<0.3), while our results suggest that most RS neurons are intermediate (between 0.45 and 0.7, RS II neurons; Fig. 4c). The reason for this discrepancy could be slightly different methods to calculate adaptation ratios. We used current pulses slightly above threshold because differences of ISIad were more pronounced in this range compared with larger depolarizing current steps as used before (Rose and Metherate 2005).
For some (approximately 25% of RS neurons) whole-cell recordings in this study, ISIad decreased over time (up to 13% in >10 min). An earlier paper (Metherate and Aramakis 1999) reported a similar decrease in some of their neurons and hypothesized that this finding probably reflects washout of an intracellular constituent, for example, Ca2+ buffer or free Ca2+ because ISIad is thought to involve spike-induced activation of Ca2+-dependent K+-currents (Schwindt et al. 1988; Lorenzon and Foehring 1993). Nevertheless, in our experiments, current-clamp measurements to analyze ISIad were conducted during the first 3 min after rupture of the cell-attached patch so that a possible washout should have been minimal. Moreover, the loss of intracellular constituents due to diffusion into the pipette might contribute to a time-dependent depression of the Ih referred to as "rundown" (Zhou and Lipsius 1993). To obtain a near physiological view, we measured the Ih amplitudes during the first minutes (<3 min) after rupture of the patch, when loss should have been minimal. The loss of intracellular constituents may also contribute to the effects observed during the application of some of the antagonists 10–20 min after control measurements. However, the reversibility of Cs+ effects and the insignificant effect of Ba2+ on ISIad (Fig. 4d) for instance indicates that the effects were due mostly to the established effects of the drugs. Moreover, experiments decreasing the T-current by DC depolarization immediately after rupture of the patch are consistent with perfusion results (Fig. 4d).
The characterization of Ih in our experiments relied on 2 established pharmacological methods, first an isolation of Ih by application of Ba2+ ions to obtain a reliable magnitude of Ih without contamination by Kir, and second by the well established Ih antagonist ZD7288. This together with estimates of Ih from voltage clamp as well as current-clamp experiments provides evidence for the presence of Ih throughout all layers of the AC, which is in line with results on the sensorimotor cortex (Strauss et al. 2004).
In this context, the approach considering the sag in current-clamp experiments to represent an index for the magnitude of Ih must be discussed. Some central nervous system neurons (e.g., thalamic relay neurons of the dorsal lateral geniculate nucleus, McCormick and Pape 1990; neurons of the central inferior colliculus, Koch and Grothe 2003) exhibit pronounced sags in the order of 20 mV together with large Ih. In RS neurons of the cortex, however, such a pronounced sag is absent despite a marked Ih (Strauss et al. 2004). The comparison of the relative magnitudes of the 2 hyperpolarization-activated currents (Ih and Kir) reveals that in layer II/III neurons of the AC the ratio of Ih:Kir is about 1:2.9 (see Table 1). In thalamic and inferior colliculus neurons, which exhibit a pronounced sag, this ratio is almost reversed (2.1:1 and 1.8:1 for thalamic and inferior colliculus neurons, respectively, estimated from published figures; McCormick and Pape 1990; Koch and Grothe 2003). Thus, in neurons with a pronounced sag, a rather small Kir seems to be dominated by the Ih. However, a large Kir may conceal the Ih (Aponte et al. 2006); hence the absence of a sag does not allow conclusions towards the magnitude of Ih. This could be demonstrated directly in AC neurons by application of Ba2+: the sag increased markedly, while the measurements of our voltage clamp experiments (cf. Fig. 2b) did not reveal significant differences comparing the Ih amplitudes in control conditions and during Ba2+ application, which is in line with previous reports (Day et al. 2005; Aponte et al. 2006). Aponte et al. (2006) for instance provided strong evidence for the presence of Ih despite no obvious sag in interneurons of the rat hippocampus.
Distribution of Neuron Types
In vitro studies of the AC (Metherate and Aramakis 1999; Hefti and Smith 2000, 2002) and the present experiments have identified at least 2 classes of pyramidal neurons, RS and IB neurons. The IB neurons have also been found in other cortical areas not only in layers IV and V (Connors et al. 1982; McCormick et al. 1985; Chagnac-Amitai and Connors 1989; Connors and Gutnick 1990; Silva et al. 1991; Agmon and Connors 1992) but also in layer II/III of the sensorimotor cortex of guinea pigs (Gutnick et al. 1982; Deisz 1996).
Despite this consensus concerning the presence of IB neurons in the neocortex, the percentage of these neurons is less clear. In the AC of the rat, a recent study found 10% IB neurons in layer V and none in layer II/III (Atzori et al. 2004). This percentage of IB neurons in layer V is similar to our results if only repetitive IB neurons in layer V are counted (IBr; 11.5%, Fig. 4b). In contrast, an earlier study (Hefti and Smith 2000) found 36.8% IB neurons in layer V of the AC in rats, which is similar to our percentages of all IB (IBr and IBl) neurons in layer V (Fig. 4b). The considerable variability regarding the number of IB neurons between our experiments and previous studies is probably due to different definitions and hence inclusion criteria to count as IB or RS neurons. For instance, Atzori et al. (2004) included only those neurons as IB that we refer to as IBr, whereas Hefti and Smith (2000) probably had also included "latent bursting neurons" (IBl). These IBl, share a constant ISI between the first 2 APs with IBr neurons, but full bursts may be precluded due to a variable control of burst generation by other conductances (e.g., Ih; Strauss et al. 2004). In any case, RS and IB neurons in the AC are not only distinguished by their firing pattern (burst recruitment, ISIad) but also by the relation of Ih and Kir. The significantly higher ratio of Ih versus Kir in IB neurons may denote weaker shunt currents in these neurons supporting burst activation (Day et al. 2005). Our finding, that Ih is not significantly different between RS and IB neurons, differs from studies in rat subiculum where bursters have larger Ih magnitudes and different temporal summation properties than RS neurons (Van Welie et al. 2006). It supports, however, the suggestion that the Ih is not responsible for the distinct firing pattern of pyramidal neurons (Van Welie et al. 2006).
The strong effect of subtle increases of depolarization on the ISI in RS I neurons is probably due to the marked voltage-dependent kinetic of the T-current becoming faster with larger depolarizations. However, a fraction of adaptation remained during T-current and K+-current block with both Ni2+ and Cs+ (Fig. 5d). Here, likely candidates contributing to the low ISIad may be M-currents.
Neurons that generate APs only early in response to current pulses and then cease firing (Lorenzon and Foehring 1993; Metherate and Aramakis 1999) were not found. Perhaps these phasic neurons are characteristic for the developing cortex because they were observed in rats during postnatal weeks 2 and 3 (on average, mice used for this study were 6 weeks old). Onset neurons, which fire a single AP at the beginning of a suprathreshold current pulse, are probably also characteristic for the developing AC (Metherate and Aramakis 1999; Kotak et al. 2005) because we did not find this neuron type in our study.
Potential Intrinsic Filter Mechanisms
The proportion of adapting RS (RS I) neurons in layers V and VI is significantly higher than in layer II/III and III/IV (Fig. 4b). Because the T-current seems to be a considerable factor for low ISIad values in these neurons (Fig. 4d) it is likely that temporally closely correlated inputs might recruit the "boosting" of EPSP by the T-current (Deisz et al. 1991), whereas later events of temporally dispersed input might be attenuated by the activity of Ca2+-dependent K+-currents (see above). Moreover, temporally closely correlated excitatory inputs may preferentially be strengthened by temporal summation limited by Ih (Magee 1998) because this current was an additional factor for low ISIad (Fig. 4d). Regarding the transfer function of such RS I neurons this firing behavior would be particularly suited to augment brief transient inputs, whereas tonic input would be less effectively conveyed, that is, such neurons might improve transfer of transients, so that synchronous excitatory input from upstream neurons would be improved as shown before by an in vivo study (DeWeese and Zador 2006).
The high Athr found in layer II/III compared with layer V RS neurons appears to be at variance with data of Atzori et al. (2004) who found the opposite differences in the AC of developing rats (postnatal weeks 3 and 4). The reason for this discrepancy is unclear but it should be noted that the resting membrane potential (e.g. Em: –58.9 mV, Atzori et al. 2004 vs. –76.0 mV in this study) differs considerably. In any case, we envisage that the high Athr in layer II/III RS neurons would cause a preferential transfer of high amplitude input and therefore improve the signal to noise ratio. Hence, these neurons might provide a higher probability of columnar transfer as opposed to a presumably smaller input from corticocortical and commissural fibers. In contrast, the low Athr in layer V and VI neurons might favor cortical (e.g. commissural) and corticofugal spread of the processed and filtered signal.
The large Kir and Ih amplitudes in layer II/III may effectively prevent temporal summation (Magee 1999; Koch and Grothe 2003; Day et al. 2005; Carr and Surmeier 2007), which, together with the higher Athr, would provide improved temporal resolution (Buonomano 2000; Hefti and Smith 2002). The low Rm in RS neurons of layer II/III allows synaptic potentials to be rapid and brief and thus contributes to the improvement of temporal resolution mentioned above (Golding et al. 1995; Bal and Oertel 2000). Moreover, the low Rm attenuates spread of single inputs and favors spread of temporally closely correlated inputs in the upper layers (Magee 1998; Golding et al. 1995; Bal and Oertel 2000; Ferragamo and Oertel 2002). Because Rm is Cs+-sensitive it corresponds to higher Kir and Ih amplitudes in layer II/III RS neurons (Fig. 6b; compare Day et al. 2005). But whether Kir and Ih are the only candidates influencing the low Rm in layer II/III is not clear and should be tested in future experiments.
RS neurons in layer II/III possess the largest values of SA. It remains to be seen by further morphological characterization of the neurons whether the SAs are indeed different between the layers or whether the capacitance measurements reflect only part of the neuronal surface, that is, the perisomatic area accessible with the clamp. Nevertheless, this result is in line with anatomical findings where pyramidal neurons of layers II and III in the AC of cats have larger dendritic fields (but smaller soma sizes) than those of layer V (Winer 1984, 1985, 1992; Winer and Prieto 2001). Therefore, it is likely that the size of dendritic fields and the number and length of dendrites, respectively, correlate with larger SAs of RS neurons in layer II/III rather than with somatic size. This would suggest that the capacitance measurements indeed detect a large proportion of the neuronal surface. In turn, assuming a comparable total resting conductance per area (cf. Table 2), this might account additionally for the low input resistance in layer II/III. Interestingly, layer II/III neurons have on average also the lowest resistance per area, that is, the highest conductance per area (Kir/SA and Ih/SA). This feature would cause very brief synaptic potentials per se, which, with a high density of fast Ih, would synergistically contribute to little temporal summation. Conversely, the particularly slow Ih in layer V neurons would tend to facilitate temporal summation (Magee 1998). Considering the crucial dependence of the time course of Ih on subunit composition of channels the time constant in layer V might suggest a relative low expression of HCN 1 subunits which exhibit fast Ih activation in comparison to subunits HCN 2 to 4 (Strauss et al. 2004; Stieber et al. 2005).
The functional implications of the significant differences between the parameters of neurons in layers V and VI (Table 2) demands further research because the connectivity of these layers is similar (Linden and Schreiner 2003). It would be interesting to correlate the connectivity of a given neuron with its intrinsic parameters.
The laminar differences of intrinsic properties of IB neurons in the AC of mice are less pronounced than in RS neurons (Table 2). Nevertheless, the differences between IB of different layers may be functionally relevant. The high Ethr of IB neurons in layer II/III would tend to suppress subthreshold activity. The lower Ethr of IB neurons in layer V as well as their higher percentage in layers V and VI (Fig. 4b) may facilitate corticocortical and corticofugal output of a computed and synchronized signal (Chagnac-Amitai and Connors 1989).
|
Conclusions |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsDiscussionConclusionsReferences |
|---|
The differences of intrinsic neuronal properties in the AC of mice govern the neuronal transfer functions of synaptic inputs pertinent to information processing, for example, suppression of signals below threshold, bandpass amplification and coincidence detection. Perhaps these layer-specific neuronal properties found here for the AC may be a more general feature present also in other (sensory) cortical areas.
|
Acknowledgments |
|---|
We thank Hartmut Niekisch and Martin Grunau, who assisted during some of the experiments as part of their practical courses. We are grateful to Sylvain Gigout for reviewing an earlier version of the manuscript. We are most grateful to Burkhard Micheel and his team of the Biotechnology group (University of Potsdam, Germany) for supplying the animals.
Conflict of Interest: none declared.
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsDiscussionConclusionsReferences |
|---|
Agmon A, Connors BW. Correlation between intrinsic firing patterns and thalamocortical synaptic responses of neurons in mouse barrel cortex. J Neurosci. (1992) 12:319–329.[Abstract]
Aponte Y, Lien CC, Reisinger E, Jonas P. Hyperpolarization-activated cation channels in fast-spiking interneurons of rat hippocampus. J Physiol. (2006) 574:229–243.
Atzori M, Flores-Hernández J, Pineda JC. Interlaminar differences of AP activation threshold in the auditory cortex of the rat. Hear Res. (2004) 189:101–106.[CrossRef][Web of Science][Medline]
Atzori M, Lei S, Evans DIP, Kanold PO, Phillips-Tansey E, McIntyre O, McBain CJ. Differential synaptic processing separates stationary from transient inputs to the auditory cortex. Nat Neurosci. (2001) 4:1230–1237.[CrossRef][Web of Science][Medline]
Bal R, Oertel D. Hyperpolarization-activated, mixed-cation current (Ih) in octopus cells of the mammalian cochlear nucleus. J Neurophysiol. (2000) 84:806–817.
Berger T, Larkum ME, Lüscher HR. High Ih channel density in the distal apical dendrite of layer 5 pyramidal cells increases bidirectional attenuation of EPSPs. J Neurophysiol. (2001) 85:855–868.
Berger T, Senn W, Lüscher HR. Hyperpolarization-activated current Ih disconnects somatic and dendritic spike initiation zones in layer V pyramidal neurons. J Neurophysiol. (2003) 90:2428–2437.
Buonomano DV. Decoding temporal information: a model based on short-term synaptic plasticity. J Neurosci. (2000) 20:1129–1141.
Carr DB, Surmeier DJ. M1 muscarinic receptor modulation of Kir2 channels enhances temporal summation of excitatory synaptic potentials in prefrontal cortex pyramidal neurons. J Neurophysiol. (2007) 97:3432–3438.
Caviness VS, Frost DO. Tangential organization of thalamic projections to the neocortex in the mouse. J Comp Neurol. (1980) 194:335–367.[CrossRef][Web of Science][Medline]
Chagnac-Amitai Y, Connors BW. Synchronized excitation and inhibition driven by intrinsically bursting neurons in neocortex. J Neurophysiol. (1989) 62:1149–1162.
Connors BW, Gutnick MJ. Intrinsic firing patterns of diverse neocortical neurons. Trends Neurosci. (1990) 13:99–104.[CrossRef][Web of Science][Medline]
Connors BW, Gutnick MJ, Prince DA. Electrophysiological properties of neocortical neurons in vitro. J Neurophysiol. (1982) 48:1302–1320.
Day M, Carr DB, Ulrich S, Ilijic E, Tkatch T, Surmeier DJ. Dendritic excitability of mouse frontal cortex pyramidal neurons is shaped by the interaction among HCN, Kir2, and Kleak channels. J Neurosci. (2005) 25:8776–8787.
Deisz RA. A tetrodotoxin-insensitive sodium current initiates burst firing of neocortical neurons. Neuroscience. (1996) 70:341–351.[CrossRef][Web of Science][Medline]
Deisz RA. GABAB receptor-mediated effects in human and rat neocortical neurones in vitro. Neuropharmacol. (1999) 38:1755–1766.[CrossRef][Web of Science][Medline]
Deisz RA, Fortin G, Zieglgänsberger W. Voltage dependence of excitatory postsynaptic potentials of rat neocortical neurons. J Neurophysiol. (1991) 65:371–381.
DeWeese MR, Zador AM. Non-gaussian membrane potential dynamics imply sparse, synchronous activity in auditory cortex. J Neurosci. (2006) 22:12206–12218.
Drayton M, Noben-Trauth K. Mapping quantitative trait loci for hearing loss in Black Swiss mice. Hear Res. (2006) 212:128–139.[CrossRef][Web of Science][Medline]
Ehret G. The auditory cortex. J Comp Physiol A. (1997) 181:547–557.[CrossRef][Medline]
Ferragamo MJ, Oertel D. Octopus cells of the mammalian ventral cochlear nucleus sense the rate of depolarization. J Neurophysiol. (2002) 87:2262–2270.
Geissler DG, Ehret G. Auditory perception vs. recognition: representation of complex communication sounds in the mouse auditory cortical fields. Eur J Neurosci. (2004) 19:1027–1040.[CrossRef][Web of Science][Medline]
Gutnick MJ, Connors BW, Prince DA. Mechanism of neocortical epileptogenesis in vitro. J Neurophysiol. (1982) 48:1321–1335.
Golding NL, Robertson D, Oertel D. Recordings from slices indicate that octopus cells of the cochlear nucleus detect coincident firing of auditory nerve fibers with temporal precision. J Neurosci. (1995) 15:3138–3153.[Abstract]
Hefti BJ, Smith PH. Anatomy, physiology, and synaptic responses of rat layer V auditory cortical cells and effects of intracellular GABAA blockade. J Neurophysiol. (2000) 83:2626–2638.
Hefti BJ, Smith PH. Distribution and kinetic properties of GABAergic inputs to layer V pyramidal cells in rat auditory cortex. J Assoc Res Otolaryngol. (2002) 4:106–121.[Web of Science][Medline]
Hille B. Ionic channels of excitable membranes (1992) Sunderland (MA): Sinauer Associates, Inc.
Jahnsen H, Llinás R. Electrophysiological properties of guinea-pig thalamic neurons: an in vitro study. J Physiol. (1984) 349:2055–2226.
Jones EG. Varieties and distribution of non-pyramidal cells in the somatic sensory cortex of the squirrel monkey. J Comp Neurol. (1975) 160:205–267.[CrossRef][Web of Science][Medline]
Koch U, Grothe B. Hyperpolarization-activated current (Ih) in the inferior colliculus: distribution and contribution to temporal processing. J Neurophysiol. (2003) 90:3679–3687.
Kotak VC, Fujisawa S, Lee FA, Karthikeyan O, Aoki C, Sanes DH. Hearing loss raises excitability in the auditory cortex. J Neurosci. (2005) 25:3908–3918.
Larkum ME, Waters J, Sakmann B, Helmchen F. Dendritic spikes in apical dendrites of neocortical layer 2/3 pyramidal neurons. J Neurosci. (2007) 27:8999–9008.
Linden JF, Schreiner CE. Columnar transformations in auditory cortex? A comparison to visual and somatosensory cortices. Cereb Cortex. (2003) 13:83–89.
Lorenzon NM, Foehring RC. The ontogeny of repetitive firing and its modulation by norepinephrine in rat neocortical neurons. Dev Brain Res. (1993) 73:213–223.[Medline]
Lund JS, Henry GH, MacQueen CL, Harvey AR. Anatomical organization of the primary visual cortex (area 17) in the cat. A comparison with area 17 of the macaque monkey. J Comp Neurol. (1979) 184:599–618.[CrossRef][Web of Science][Medline]
Magee JC. Dendritic hyperpolarization-activated currents modify the integrative properties of hippocampal CA1 pyramidal neurons. J Neurosci. (1998) 18:7613–7624.
Magee JC. Dendritic Ih normalizes temporal summation in hippocampal CA1 neurons. Nat Neurosci. (1999) 2:508–514.[CrossRef][Web of Science][Medline]
McCormick DA, Connors BW, Lighthall JW, Prince DA. Comparative electrophysiology of pyramidal and sparsely spiny stellate neurons of the neocortex. J Neurophysiol. (1985) 54:782–806.
McCormick DA, Pape HC. Properties of a hyperpolarization-activated cation current and its role in rhythmic oscillation in talamic relay neurons. J Physiol. (1990) 431:291–318.
Metherate R, Aramakis VB. Intrinsic electrophysiology of neurons in thalamorecipient layers of the developing rat auditory cortex. Dev Brain Res. (1999) 115:131–144.[CrossRef][Medline]
Paxinos G, Franklin KBJ. The mouse brain in stereotactic coordinates (2001) San Diego (CA): Academic Press.
Robinson RB, Siegelbaum SA. Hyperpolarization-activated cation currents: from molecules to physiological function. Annu Rev Physiol. (2003) 65:453–480.[CrossRef][Web of Science][Medline]
Rockel AJ, Hiorns RW, Powell TP. The basic uniformity in the structure of the neocortex. Brain. (1980) 103:221–244.
Rose HJ, Metherate R. Auditory thalamocortical transmission is reliable and temporally precise. J Neurophysiol. (2005) 94:2019–2030.
Schwindt PC, Spain WJ, Foehring RC, Stafstrom CE, Chubb MC, Crill WE. Multiple potassium conductances and their functions in neurons from cat sensorimotor cortex in vitro. J Neurophysiol. (1988) 59:424–449.
Sciancalepore M, Constanti A. Inward-rectifying membrane currents activated by hyperpolarization in immature rat olfactory cortex neurones in vitro. Brain Res. (1998) 814:133–142.[CrossRef][Web of Science][Medline]
Silva LR, Gutnick MJ, Connors BW. Laminar distribution of neuronal membrane properties in neocortex of normal and reeler mouse. J Neurophysiol. (1991) 66:2034–2040.
Simons DJ, Woolsey TA. Morphology of Golgi-Cox-impregnated barrel neurons in rat SmI cortex. J Comp Neurol. (1984) 230:119–132.[CrossRef][Web of Science][Medline]
Smith PH, Populin LC. Fundamental differences between the thalamocortical recipient layers of the cat auditory and visual cortices. J Comp Neurol. (2001) 436:508–519.[CrossRef][Web of Science][Medline]
Song L, McGee JA, Walsh EJ. Frequency- and level-dependent changes in auditory brainstem responses (ABRs) in developing mice. J Acoust Soc Am. (2006) 119:2242–2257.[CrossRef][Web of Science][Medline]
Staiger JF, Flagmeyer I, Schubert D, Zilles K, Kötter R, Luhmann HJ. Functional diversity of layer IV spiny neurons in rat somatosensory cortex: quantitative morphology of electrophysiologically characterized and biocytin labeled cells. Cereb Cortex. (2004) 14:690–701.
Stieber J, Stöckl G, Herrmann S, Hassfurth B, Hofmann F. Functional expression of the human HCN3 channel. J Biol Chem. (2005) 280:34635–34643.
Strauss U, Kole MHP, Bräuer AU, Pahnke J, Bajorat R, Rolfs A, Nitsch R, Deisz RA. An impaired neocortical Ih is associated with enhanced excitability and absence epilepsy. Eur J Neurosci. (2004) 19:3048–3058.[CrossRef][Web of Science][Medline]
Sutor B, Zieglgänsberger W. A low-voltage activated, transient calcium current is responsible for the time-dependent depolarizing inward rectification of rat neocortical neurons in vitro. Pflugers Arch. (1987) 410:102–111.[CrossRef][Web of Science][Medline]
Van Welie I, Remme MWH, van Hooft JA, Wadman WJ. Different levels of Ih determine distinct temporal integration in bursting and regular-spiking neurons in rat subiculum. J Physiol. (2006) 576:203–214.
Winer JA. The pyramidal cells in layer III of the cat primary auditory cortex (AI). J Comp Neurol. (1984) 229:476–493.[CrossRef][Web of Science][Medline]
Winer JA. Structure of layer II in cat primary auditory cortex (AI). J Comp Neurol. (1985) 238:10–37.[CrossRef][Web of Science][Medline]
Winer JA. The functional architecture of the medial geniculate body and the primary auditory cortex. In: The mammalian auditory pathway: neuroanatomy—Webster DB, Popper AN, Fay RR, eds. (1992) New York: Springer. 222–409.
Winer JA, Prieto JJ. Layer V in cat primary auditory cortex (AI): cellular architecture and identification of projection neurons. J Comp Neurol. (2001) 434:379–412.[CrossRef][Web of Science][Medline]
Zheng QJ, Johnson KR, Erway LC. Assessment of hearing in 80 inbred strains of mice by ABR threshold analyses. Hear Res. (1999) 130:94–107.[CrossRef][Web of Science][Medline]
Zhou Z, Lipsius SL. Effect of isoprenaline on If current in latent pacemaker cells isolated from cat right atrium: ruptured vs. perforated patch whole cell recording methods. Pflugers Arch. (1993) 423:442–447.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||