A transcriptomic analysis of type I-III neurons in the bed nucleus of the stria terminalis

Abstract

The activity of neurons in the anterolateral cell group of the bed nucleus of the stria terminalis (BNST_ALG) plays a critical role in anxiety- and stress-related behaviors. Histochemical studies have suggested that multiple distinct neuronal phenotypes exist in the BNST_ALG. Consistent with this observation, the physiological properties of BNST_ALG neurons are also heterogeneous, and three distinct cell types can be defined (Type I–III) based primarily on their expression of four key membrane currents, namely I_h, I_A, I_T, and I_K(IR). Significantly, all four channels are multimeric proteins and can comprise of more than one pore-forming α subunit. Hence, differential expression of α subunits may further diversify the neuronal population. However, nothing is known about the relative expression of these ion channel α subunits in BNST_ALG neurons.

We have addressed this lacuna by combining whole cell patch clamp recording together with single cell reverse transcriptase polymerase chain reaction (scRT-PCR) to assess the mRNA transcript expression for each of the subunits for the four key ion channels in Type I-III neurons of the BNST_ALG. Here, cytosolic mRNA from single neurons was probed for the expression of transcripts for each of the α subunits of I_h (HCN1- HCN4), I_T (Ca_v3.1- Ca_v3.3), I_A (K_v1.4, K_v3.4, K_v4.1- K_v 4.3) and I_K(IR) (Kir2.1-Kir2.4).

An unbiased hierarchical cluster analysis followed by discriminant function analysis revealed that a positive correlation exists between the physiological and genetic phenotype of BNST_ALG neurons. Thus, the analysis segregated BNST_ALG neurons into 3 distinct groups, based on their α subunit mRNA expression profile, which positively correlated with our existing electrophysiological classification (Type I–III). Furthermore, analysis of mRNA transcript expression in Type I –Type III neurons suggested that, whereas Type I and III neurons appear to represent genetically homologous cell populations, Type II neurons may be further subdivided into three genetically distinct subgroups. These data not only validate our original classification scheme, but further refine the classification at the molecular level, and thus identifies novel targets for potential disruption and/or pharmacotherapeutic intervention in stress-related anxiety-like behaviors.

Introduction

Activation of the bed nucleus of the stria terminalis (BNST) has been argued to mediate stress responses (Fendt et al., 2003; Waddell et al., 2006) and anxiety-like behaviors to diffuse and/or unpredictable threats (Walker et al., 2003; Hammack et al., 2004), as well as stress-induced drug recidivism (Erb and Stewart, 1999; Shalev et al., 2001). Consistent with a role for the BNST in anxiety-like behavior, electrical stimulation of the anterolateral region produces many of the endocrine, cardiovascular and respiratory responses that are normally elicited by anxiogenic stimuli (Casada and Dafny, 1991). In addition, excitotoxic lesions of the BNST blocked the enhanced startle response elicited after central administration of the stress hormone, corticotropin-releasing factor (CRF) (Lee and Davis, 1997). Moreover, stressors and anxiogenic pharmacological agents, such as yohimbine, m-chlorophenylpiperazine, and caffeine increase the expression of transcription factors, such as the immediate early gene, c-fos, in the anterolateral BNST (Singewald et al., 2003; Day et al., 2004), further suggesting that increased activity of the BNST is associated with an increased state of anxiety.

The BNST is comprised of multiple subdivisions, each of which may contribute to different aspects of its function (Dong and Swanson, 2004). The anterolateral cell group of the BNST (BNST_ALG) is a heterogenous structure composed of the juxtacapsular, oval, rhomboid, and fusiform nuclei (Ju and Swanson, 1989). Based on its cytoarchitecture, chemoarchitecture, and projection patterns, the BNST_ALG also possesses distinct neuronal subpopulations, however 70–90% can be categorized as medium sized spiny GABAergic neurons (McDonald, 1983; Sun and Cassell, 1993). Despite this apparent homology, neurons of this region exhibit heterogenous physiological responses to several neurotransmitters, including opiates (Casada and Dafny, 1993; Dalsass and Siegel, 1990), norepinephrine (Casada and Dafany, 1993), acetylcholine (Casada and Dafany, 1993), oxytocin (Ingram et al., 1990), and serotonin (Levita et al., 2004; Guo et al., 2009).

Serotonergic modulation of neuronal activity in the BNST determines the developmental risk for stress-related psychopathology (Oler et al., 2009; Hariri and Holmes, 2006). Consistent with these observations, we have shown that chronic stress alters serotonin receptor mRNA expression in the BNST to favor excitation rather than inhibition (Hammack et al., 2009). However, in other brain regions, chronic stress has also been shown to regulate the expression of calcium-, potassium-, and calcium-dependent potassium channels (van Germert, et al., 2009; Chatterjee et al., 2009; Hjaersen et al., 2008; Xie et al., 1999). If we are to fully understand how chronic stress may regulate neuronal activity in regions such as the BNST_ALG, we must first determine the ion channel expression profile of neurons within this region.

To date, we have identified three physiologically distinct cell types (Type I–III) in the BNST_ALG based on their pattern of action potential firing and their differential expression of four subthreshold intrinsic membrane currents: 1) the hyperpolarization-activated nonspecific cation current (I_h), 2) the low-threshold calcium current (I_T), 3) the transient voltage-dependent potassium current (I_A), and 4) the inward rectifying current I_K(IR) (Hammack et al., 2007).

Type I and Type II neurons possess a prominent I_h current, whereas Type III neurons show little discernable I_h current (Hammack et al., 2007). The I_h channel is encoded by four channel subunits HCN1, HCN2, HCN3, and HCN4 (Ludwig et al., 1998; Santoro and Tibbs, 1999; Gauss and Seifert, 2000; Monteggia et al., 2000; Kaupp and Seifert, 2001). Conversely, Type II neurons are the only subtype that express a prominent I_T current, which is encoded by three genes: Ca_v3.1, Ca_v3.2, and Ca_v3.3 (Perez-Reyes, 2003). All BNST neurons appear to express some I_A current, mediated by voltage-dependent potassium (K_v) channels consisting of α and β subunits (Song, 2002), which are in turn encoded by several subfamilies of genes. Multiple K_v channel subunits can mediate I_A including K_v1.4, K_v3.4, and the K_v4 family (K_v4.1, K_v4.2 and K_v4.3) (Serodio et al., 1996; Coetzee et al., 1999). Similarly, the inward rectifying current I_K(IR) is also encoded by multiple genes of the Kir2 subfamily Kir2.1, Kir2.2, Kir2.3 and Kir2.4 (Pruss et al., 2005). What is not known is if, and/or how, differential expression of these gene products may contribute to the physiological diversity of Type I – III BNST_ALG neurons. Moreover, chronic stress may differentially regulate the expression of individual members of a single ion channel family. Hence, understanding the genetic phenotype of Type I – III neurons is a critical first step in understanding phow stress may selectively alter neuronal activity in the BNST and, hence, contribute to stress-induced psychopathology.

In the present study, we combined whole-cell patch clamp recordings with single-cell reverse transcription polymerase chain reaction (scRT-PCR) to identify the mRNA expression patterns of these four key ion channel subunits in physiologically defined BNST_ALG neurons.

Methods

Animals

All experiments were conducted on male Sprague-Dawley rats (Charles River, NC) at 28 to 42 days of age. All rats were housed five per cage and had unrestricted access to food and water. Care was taken to minimize the number of animals used; all procedures were done in accordance with policy guidelines set by the National Institutes of Health and were approved by the Emory University Institutional Animal Care and Use Committee (IACUC).

Electrophysiological Recordings

Preparation of BNST slices

BNST slices were obtained as previously described (Hammack et al., 2007; Guo and Rainnie, 2010). Briefly, rats were decapitated under deep isoflurane anesthesia (Abbott Laboratories, North Chicago, IL) and the brains rapidly removed and placed in ice-cold “cutting solution”, kynurenic acid– based artificial cerebrospinal fluid (ACSF), which contained (in mM): NaCl (130), KCl (3.5), KH2PO4 (1.1), MgCl2 (6.0), CaCl2 (1.0), NaHCO3 (30), glucose (10), and kynurenic acid (2). The glutamatergic antagonist kynurenic acid was included in the “cutting solution” to suppress any unwanted effects of glutamate release that may occur during tissue slicing. Divalent cation levels were also adjusted to reduce the probability of neurotransmitter release. A block of tissue containing the BNST was then mounted on the stage of a Leica VTS-1000 vibrating microtome (Leica Microsystems, Bannockburn, IL), and 350 μm coronal slices were cut. Slices were transferred to a holding chamber containing “cutting solution” at room temperature and gassed with a 95% O2–5% CO2 mixture for 1 h before being placed in oxygenated control ACSF containing (in mM): NaCl (130), KCl (3.5), KH2PO4 (1.1), MgCl2 (1.3), CaCl2 (2.5), NaHCO3 (30), and glucose (10). Experiments started a minimum of 0.5 h after the transfer of slices into the control ACSF to washout the action of kynurenic acid.

Visual identification of BNST_ALG neurons

Slices were placed in a Warner Series 20 recording chamber (Warner Instruments, Hamden, CT) mounted on the fixed stage of a Leica DM-LFS microscope (Leica Microsystems). Slices were fully submerged and continuously perfused at a rate of 1–2 ml/min with heated (32°C) and oxygenated ACSF. BNST_ALG neurons were visualized using infrared (IR) illumination and a 40X water-immersion objective (Leica Microsystems). Images were captured using an IR-sensitive charge-coupled device (CCD) digital camera (Orca ER, Hamamatsu, Tokyo, Japan), coupled to a Phoenix video frame grabber (Active Silicon, North Chelmsford, MA), and displayed on a computer monitor using Simple PCI 6.11 software (Compix, Sewickley, PA).

Recording procedures

For whole cell patch-clamp recording, autoclaved thin-walled borosilicate glass-patch electrodes (WPI, Sarasota, FL) were pulled on a Flaming/Brown micropipette puller (Model P-97, Sutter Instrument, Novato, CA). Patch electrodes had resistances ranging from 4 to 6 MΩ, when filled with a standard RNase free patch solution that contained (in mM): K-gluconate (138), KCl (2), MgCl2 (3) and HEPES (10). The patch-recording solution was adjusted to a pH of 7.3 with KOH and had a final osmolarity of 280 mOsm. Whole cell patch-clamp recordings were obtained using a MultiClamp 700B amplifier (Molecular Devices, Sunnyvale, CA), with a Digidata 1320A A–D interface, and pClamp 10 software (Molecular Devices). For all experiments, whole cell patch-clamp configuration was established only when the seal resistance was >1.5 GΩ. Neurons were excluded from analysis if they showed a resting membrane potential (Vm) more positive than −55 mV and/or had an action potential that did not overshoot +10 mV. Current-clamp signals were filtered at 5 kHz and digitized at 10–20 KHz. Series resistance was bridge balanced and monitored over the course of each experiment and was considered acceptable with change less than 15%. All electrophysiology data were analyzed with a custom made MATLAB 2009a script (Mathwork, Natick, MA).

Electrophysiological parameter definitions

Fast after-hyperpolarization (fAHP, in mV): the difference between the action potential threshold and the most negative membrane potential attained after the action potential, measured in response to the smallest depolarizing current step (Δ10 ms) evoking an action potential.

Input resistance (R_in, in MΩ): the difference between baseline (−60 mV) and the peak voltage response divided by the injected current (20–50 pA).

Membrane time constant (Tau, in ms): determined from the mono exponential curve best fitting the voltage deflection evoked in response to a 20–50 pA hyperpolarizing current step depending on the input resistance.

Resting membrane potential (RMP, in mV): the steady state membrane potential reached several minutes after breaking the seal, with no DC current applied.

First inter-spike-interval (ISI₁, in ms): the time interval between the threshold of the first action potential and the threshold of the second action potential in response to a 750 ms depolarizing current injection.

Last inter-spike-interval (ISI_Last, in ms): the time interval between the threshold of the penultimate action potential and the threshold of the last action potential in response to a 750 ms depolarizing current injection.

Spike adaptation ratio (dimensionless): the ratio of ISI₁ to ISI_Last.

I_h ratio: the difference between the steady state membrane potential at the end of a 750 ms hyperpolarizing current step and the most negative membrane potential at the beginning of the step, divided by the most negative membrane potential.

I_K(IR) ratio: The difference between the peak membrane potential at the beginning of the two smallest hyperpolarizing steps divided by the difference between the peak membrane potential at the beginning of the two largest hyperpolarizing steps, such that a cell with no observable inward rectification has a rectification ratio equal to 1.

Spike Threshold (in mV): the point at which the action potential was initiated. The point at which the membrane potential showed a greater than 10 fold change in the rate of rise (mV/ms) in response to the smallest depolarizing current step (10 ms) evoking an action potential.

Spike Amplitude (in mV): the difference between the threshold and the voltage at the peak of the action potential.

Spike half-width (in ms): measured at half height between threshold and peak action potential.

10–90% rise time (in ms): time to reach 90% of the peak amplitude of the action potential from 10% of the peak amplitude during the rising phase of the action potential.

90-10% decay time (in ms): time to reach 10% of the peak amplitude of the action potential from 90% of the peak amplitude during the falling phase of the action potential.

Histochemical Visualization of Recording Sites

To determine the location of recorded neurons in the BNST_ALG, biocytin (0.3%) was included in the patch recording solution and neurons were visualized using post hoc procedures as previously described (Levita et al., 2004). In brief, after recording, slices were placed in 4% paraformaldehyde overnight. The slices were then washed with 0.05 M phosphate buffer/0.15 M NaCl solution (PBS, 3×10 min), and re-sectioned to 70 μM using an OTS 4000 vibratome (FHC Inc, Bowdoinham, ME). Tissue sections were then placed in 0.6% H₂O₂ and 0.5% Triton X-100 in PBS for 1 h to quench endogenous peroxidase activity, washed in PBS (3×10 min) before being incubated in an avidin–biotin peroxidase complex for 2 h (ABC Vectorstain Elite kit; Vector Laboratories, Inc., Burlingame, CA). Slices were then washed in PBS (3×10 min), and visualized using 3,3-diaminobenzidine (DAB; Sigma-Aldrich, St. louis, MO) as the chromogen (DAB 0.7 mg/ml) and H₂O₂ 0.2 mg/ml in 0.06 M Tris buffer. The sections were placed in DAB until adequate staining had developed, and the reaction was stopped by transferring the sections to PBS. After two more washes in PBS the sections were mounted onto gelatinized slides, air dried, then dehydrated in an ascending series of alcohols, cleared in histoclear (2×3 min), and cover-slipped with Permount mounting medium (Fisher Scientific, New Jersey).

RNA Isolation and Reverse Transcription

RNA Isolation

Total RNA was isolated from BNST tissue by homogenizing each sample in Trizol (Invitrogen, Carlsbad, CA). The isolated RNA was then reverse transcribed using a cocktail containing 5 μl of 10XRT buffer, 10mM dNTP mix, 10X random hexanucleotide and Multiscribe RT 5U/ul and RNAase free water. The mixture was incubated in a thermal cycler at 25°C for 10 min and then at 37°C for 120 min, the resulting cDNA samples were stored at −20°C. All reagents were obtained from Applied Biosystems (Foster City, CA).

Qualitative PCR

The cDNA was amplified using 10X PCR buffer (Qiagen, Germantown, MD), 3mM MgCl₂ (Qiagen), 10mM dNTPs, 2.5U of Taq DNA Polymerase (Qiagen) and 100nM primers. PCR primers used for each of the ion channels were developed from GenBank sequences with commercially available Oligo software (IDT Tools, Coralville, IA, USA). The housekeeping gene 18S rRNA was used in all experiments as a positive control. All of the oligonucleotide primers used in this study are given in Table 1. Standard PCR was performed on a PTC-200 Peltier thermal cycler (MJ Research) using the following program: 94°C for 40 sec, 56°C for 40 sec and 72°C for 1 min for 40 cycles. PCR products were visualized by staining with ethidium bromide and separated by electrophoresis in a 1% agarose gel. Each experiment was repeated six times with different rat samples and each sample was assayed in triplicate to reduce inter-sample variation.

Table 1.

PCR Primer used in this study

Genes	Accession No.	PCR Product Size (bp)	Genes	Accession No.	PCR Product Size (bp)
18S rRNA	X01117	563	Kv3.4	X62841	676
HCN1	AF247450	405	Kv4.1	M64226	467
HCN2	AF247451	211	Kv4.2	S64320	265
HCN3	AF247452	335	Kv4.3	U42975	296/386
HCN4	AF247453	462	Kir2.1	X73052	413
Cav3.1	AF027984	407	Kir2.2	NM010603	358
Cav3.2	AF290213	256	Kir2.3	NM008427	204
Cav3.3	AF086827	556	Kir2.4	NM45963	234
Kv1.4	X16002	434	GAD67	M76177	426

Single-Cell RT-PCR

At the end of each recording session, the cell cytoplasm was aspirated into the patch recording pipette containing ~5 μl of RNase-free patch solution under visual control, by applying gentle negative pressure. The contents of the patch pipette were then expelled into a microcentrifuge tube containing 5 μl of the reverse transcription cocktail by applying positive pressure. The mRNA was reverse transcribed using 50U of Multiscribe RT and was performed in a final volume of 10 μl as described above. The cDNA was stored at −20°C before further processing. Subsequently, 3 μl of each RT product was amplified in triplicate in 0.2 ml thin walled PCR tubes. Here, 19 μl of PCR mixture I [1X Buffer, dNTP mix, 1 μg/μl primer V3 (dT)₂₄] and 0.05 U/μl Taq DNA Polymerase (Qiagen) was added to each tube for the first round of PCR. Cycling conditions for the first round of PCR were: 95°C for 3 min, 50°C for 2 min and 72°C for 3 min. The tubes were immediately put on ice for 1 min, and 19 μl of PCR mixture II was added with a composition the same as that of PCR mixture I but with primer V1 (dT)₂₄ replacing primer V3(dT)₂₄ (Kurimoto et al., 2006). A 20-cycle PCR amplification was then performed with the following cycling conditions: 95°C for 30 sec, 67°C for 1 min and 72°C for 3 min with a 6 sec extension per cycle. The total amount of cDNA amplified per sample was 123 μl. Amplified cDNA sample was quantified using a Nanodrop 1000 Spectrophotometer (Thermo Scientific, Wilmington, DE), and the concentration per sample ranged from 500–1000ng/μl, and had OD values ranging from 1.9–2.0. The amplified cDNA from each cell was analyzed for the expression of 18S rRNA as a positive control marker. The resultant cDNA was subjected to another amplification step using 2 μl of cDNA from each cell as a template and 100nM of each of the primers for GAD₆₇, I_h, I_T, I_A and I_K(IR) channel subunits. The PCR master mix for the final PCR consisted of 10X PCR Buffer, 3mM MgCl₂, 10mM dNTPs, 2.5U of Taq DNA Polymerase in a final volume of 20 μl. PCR was performed using a 10 min hot start at 95°C followed by a 40 cycle program (94°C for 40 sec, 56°C for 40 sec and 72°C for 1min). PCR products were visualized by staining with ethidium bromide and separated by electrophoresis in a 1% agarose gel.

Controls for the RT-PCR

PCR conditions were optimized using total RNA isolated from rat BNST_ALG so that a PCR product could be detected from (250pg–1 ng) of total RNA without contamination caused by non-specific amplification. For each PCR amplification, sterile water was used instead of cDNA as a control for contaminating artifacts. A control with no RT was also used in each amplification. Both the controls gave negative results throughout the study. All primers were intron-spanning to exclude amplification of genomic DNA. This possibility was further omitted by the fact that the cell nucleus was never harvested while isolating mRNA from single cell.

Quantitative PCR

Real-time PCR reactions were performed using an Applied Biosystems 7500 Fast-Real Time PCR system (Applied Biosystems, Foster City, CA). Here, 1 μl samples of unamplified cDNA obtained from the single (Type I, II and III BNST neurons) were combined with Taqman probes specific for 18S rRNA (assay ID, Hs99999901_s1), Ca_v3.3 (Rn01505210_g1) and K_v4.3 (Rn00709608_m1) and 1x Taqman universal PCR Master Mix (Applied Biosystems). The reaction for each cell sample was performed in triplicate, and using the 50 cycle thermal cycling program: cycle 1–20 min at 95°C; cycles 2 through 50– 95°C for 3 sec, followed by 60°C for 30 min.

Statistical analysis

Statistical analyses were carried out using SPSS 16.0 (SPSS, Chicago, IL). One way analysis of variance (ANOVA) was used to determine the differences of physiological properties among three cell types. An unsupervised, hierarchical cluster analysis using Ward’s method (Ward, 1963) and squared Euclidian distance was performed to classify cell types of BNST_ALG neurons. This method has been used successfully by previous studies to classify CCK cells in basolateral nucleus of amygdala (Jasnow et al., 2009), somatostatin cells of the sensorimotor cortex, and projection neurons and interneurons of the lateral amygdala (Halabisky et al., 2006; Sosulina et al., 2006). Briefly, this clustering method begins by separating each individual neuron into a cluster by itself. At each stage of the analysis, the neurons that are most similar, as measured by our electrophysiological parameters and ion channel mRNA expression profile, are grouped together to form another larger cluster. This process continues until all of the neurons are joined into a single cluster. Ward’s method minimizes the error sum of squares of any pair of clusters formed at a given step; this maximizes between-group differences and minimizes within-group differences. Before clustering, each electrophysiological property was log transformed and converted into standardized z-scores. This achieves a normal distribution and prevents variables with larger ranges from having a greater influence on the cluster solution than variables with small ranges. After the hierarchical cluster analysis was performed, the appropriate number of clusters was determined by examining within-group and between-group squared Euclidian distances. A good cluster solution should have large between-cluster differences and small within-cluster differences. A one-way analysis of variance (ANOVA) was used to detect statistically significant differences in electrophysiological parameters among the reported clusters. For all analyses, statistical significance was ascribed at p < 0.05. The parameters used in the cluster analysis were chosen based on intrinsic membrane currents, which are important in identifying the three types of neurons within the BNST_ALG. Membrane potential, I_h, ISI₁, I_K(IR), spike threshold and half width, together with the expression of the ion channel subunits HCN1-HCN4, Ca_v3.1-Ca_v3.3, K_v1.4, K_v3.4, K_v4.1-K_v4.3, and Kir2.1-Kir2.4 were used for the cluster analysis. Further analysis was conducted using a Discriminant Function Analysis (DFA) in order to determine the electrophysiological properties and genetic phenotype most strongly contributing to group membership. In addition, DFA builds a predictive model for group membership. Briefly, this analysis attempts to find linear combinations of the variables that best separate the groups of cells by maximizing the between-group variance. Finally, a leave-one-out cross-validation procedure was then used to validate the discriminant functions.

Results

Our previous study defined three electrophysiologically distinct subtypes of BNST_ALG neuron based on their expression of the membrane currents I_h, I_T, I_A, I_K(IR) (Hammack et al., 2007). However, the ion channels that mediate these currents can be formed by permutations of several different α subunits, and little is known about the relative expression of these ion channel subunits in BNST_ALG neurons. To address this critical knowledge gap, we harvested the cytoplasmic mRNA from physiologically identified BNST_ALG neurons and examined their expression of I_h, I_T, I_A, I_K(IR) channel subunits using single cell RT-PCR. For unsupervised cluster analysis and discriminant function analysis we used the following physiological parameters: membrane potential, I_h, ISI₁, I_K(IR), spike threshold, half-width; as well as the mRNA expression profile of the I_h, I_T, I_A, and I_(KIR) channel α subunits.

Three electrophysiologically distinct neuronal populations in the BNST_ALG

A total of 51 BNST neurons were included in this study, all of which were located in the boundaries of the BNST_ALG. For methodological reasons it is not possible to extract cytosolic RNA and also fill cells with biocytin. Hence, a schematic diagram of the position of a representative population of BNST_ALG neurons with similar biophysical properties is shown in Figure 1. As illustrated, the majority of neurons recorded were located within the boundaries of the oval nucleus (BNSTov) of the BNST, however, several cells were recorded that were located in the juxtacapsular (BNSTju) and anterolateral nuclei (BNSTal). For each of the 51 neurons recorded in this study, we conducted a series of standardized current-clamp protocols to determine their characteristic voltage response to transient depolarizing and hyperpolarizing current injections (Hammack et al., 2007). Based on their electrophysiological characteristics, Type I neurons accounted for 11%, Type II for 66%, and Type III neurons for 21% of the sample population. As shown in Table 2, the intrinsic membrane currents, I_h, I_T and I_K(IR), played a significant role in determining the subtypes of BNST_ALG neurons. Consistent with our previous study, Type I and II neurons had a significantly larger I_h current than Type III neurons, whereas Type III neurons displayed the largest I_(KIR) current. The I_T current was approximated by the duration of ISI₁, which was significantly shorter in Type II and Type III than in Type I neurons. However, each of these macroscopic currents is mediated by ion channels that are multimeric structures comprised of one or more functional subunits. Consequently, we next screened each of these BNST_ALG neurons for their expression of the specific ion channel α subunits.

Figure 1.

Schematic diagram showing the location of a representative sample of biocytin-labeled neurons visualized with histochemistry in the BNST_ALG. AC= anterior commissure. BNSTov= oval nucleus of the BNST. BNSTju=Juxtacapsular nucleus of the BNST. BNSTal= anterolateral nucleus of the BNST.

Table 2.

Electrophysiological properties of Type I, Type II and Type III neurons of the BNST_ALG

	Type I (n=6)	Type II (n=34)	Type III (n=11)
Percentage (%)	11	66	21
RMP (mV)	−61.0±1.753	−57.8±0.733	−67.3±1.291,2
Rin (MΩ)	266.7±52.2	297.1±21.9	195.8±38.6
Tau (ms)	25.4±2.97	22.7±1.24	19.7±2.19
Ih	0.042±0.0072,3	0.071±0.0031,3	0.010±0.0051,2
IK(IR)	1.82±0.283	2.21±0.123	3.75±0.211,2
Spike
Threshold (mV)	−42.8±1.303	−43.4±0.543	−38.3±0.961,2
Amplitude (mV)	74.8±2.88	73.2±1.213	78.4±2.122
Rise time (ms)	0.40±0.04	0.49± 0.01	0.43± 0.03
Decay time (ms)	1.01± 0.15	1.28± 0.06	1.17± 0.11
Half-width (ms)	1.01±0.10	1.21±0.04	1.06± 0.08
ISI1 (ms)	106.6±7.822,3	18.4±3.281	22.8±5.771
ISILast(ms)	127.9± 16.183	106.2±6.793	41.9±11.951,2
ISI1/ISILast	0.82±0.082,3	0.19±0.031,3	0.55±0.061,2
fAHP Diff	−7.64±1.01	−5.26±0.61	−5.48±0.81

Values are expressed as mean±SEM. Superscript following the SEM indicates cell type differences based on analysis of variance (ANOVA, p<0.05)

¹Type I,

²Type II and

³Type III.

Ion channel subunit mRNA expression in whole tissue and single BNST_ALG neurons

Isolated BNST_ALG tissue expressed mRNA transcripts for all of the ion channel subunits HCN1-4, Cav3.1-3, K_v1.4, K_v3.4, K_v4.1-3, and Kir2.1-4, and their expression pattern was qualitatively similar across all animals. Significantly, our RT-PCR protocol was sensitive enough to detect two splice variants of the K_v4.3 subunit in our tissue sample. As shown in Figure 2, the location of the bands for each of the 17 genes corresponds to the expected PCR fragment size based on the oligonucleotide primers used for this study (see Table 1). Although the whole tissue RT-PCR study revealed the extent of specific ion channel subunit expression in the BNST_ALG, it did not allow us to determine whether the electrophysiological properties of individual BNST neurons were associated with specific ion channel subunit expression. Therefore, we combined RT-PCR with whole cell patch clamp recordings to identify channel subunit expression in single BNST_ALG neurons.

Figure 2.

Gel picture showing the mRNA expression of I_h, I_T, I_A, I_K(IR) channel subunits and GAD₆₇, in the BNST_ALG. Total RNA was isolated and cDNA was amplified by PCR with primers specific for each of the channel subunits. The amplified products were electrophoresed in 1% agarose gel. M denotes a molecular weight marker.

Consistent with previous in situ hybridization studies (Cullinan et al., 1993) mRNA transcripts for the γ-aminobutyric acid (GABA) synthesizing enzyme, GABA-aminodecarboxylase 67 (GAD₆₇) was detected in all neurons tested, suggesting that Type I – III neurons are primarily GABAergic. Moreover, an initial examination of the ion channel subunit expression in individual neurons revealed that unlike whole tissue, mRNA transcripts for the K_v3.3, K_v4.1, Ca_v3.2 and Kir2.4 subunits were never detected in BNST_ALG neurons, suggesting that these transcripts may be preferentially expressed in the non-neuronal cell population.

We next screened each of the physiologically identified BNST_ALG neurons for the 16 channel subunits that together determine the expression of I_h, I_T, I_A and I_(KIR). Transcripts, for all of the remaining subunits of I_h, I_T, I_A, I_K(IR) channels were found to be differentially distributed across Type I–III neurons (Figure 3). Thus, an initial analysis of gene expression in individual neurons, using HCN channel subunit expression as a seed value, revealed that BNST_ALG neurons could be differentiated into three main subgroups, which showed a close correlation with the previously defined Type I–III physiological subtypes. The results of this analysis are summarized in Table 3. For example, one subgroup of neurons (n=5) were characterized by their expression of transcripts for HCN2 and HCN4 only, together with K_v4.2 and K_v4.3, and Kir2.1, but none of the other ion channel subunits. Significantly, this subgroup corresponded to 5/6 of the Type I neurons recorded in this study. Similarly, another subgroup of neurons (n=9) was characterized by their expression of transcripts for only the HCN4 subunit, together with Cav3.3, Kv4.2, and Kir2.1, 2.2 and 2.3 subunits, and these neurons corresponded to 9/11 of the Type III neurons recorded. Finally, three additional subgroups were identified in which all of the neurons expressed transcripts for HCN1 (n=34), but which were then subdivided based on their expression of at least one other HCN subunit. All of these neurons corresponded to Type II neurons.

Figure 3.

Representative figure showing the shapes, typical responses to hyperpolarizing and depolarizing current injections, and ion channel subunits mRNA expression of Type I–III BNST_ALG neurons.

Table 3.

Differential distribution of ion channel subunits in BNST_ALG neurons.

Cell Type	HCN				Cav			Kv						Kir			#
	1	2	3	4	3.1	3.2	3.3	1.4	3.4	4.1	4.2	4.3	2.1	2.2	2.3	2.4
I	0	+	0	+	0	0	0	0	0	0	+	+	+	0	0	0	5/6
II	+	+	+	+	+	0	0	0	0	0	+	0	+	0	+	0	15/34
	+	+	+	0	+	0	0	0	0	0	+	0	0	0	+/−	0	8/34
	+	+	0	0	+	0	+	0	0	0	0	+	0	0	+	0	11/34
III	0	0	0	+	0	0	+	0	0	0	+	0	+/−	+	+	0	9/11

Cells were first filtered according to their expression pattern for the HCN subunits. Subsequent filtering was determined by ≥ 80% expression (+), ~ 50% expression (+/−), and < 10% expression (0). Numbers in the far right column represent the number of cells for each subgroup that expressed all of the HCN channels indicated and ≥ 80% expression for two or more of the remaining α subunits.

Type I – III neurons also showed distinct distribution patterns for the I_T channel subunits. Hence, Ca_v3.3 mRNA was expressed in 100% of Type III neurons, and 47% of Type II neurons, whereas only 17% of Type I neurons expressed this subunit. Transcripts for Ca_v3.1 were expressed in 91% of Type II neurons, but were not expressed in either Type I or III neurons. Conversely, expression of the I_A channel subunits was similar across all three subtypes, with the following exceptions: transcripts for K_v1.4 were expressed only in Type II neurons (6%), whereas K_v3.4 transcripts were expressed in Type I (17%) and Type III neurons (9%), but not in Type II neurons. As shown in Table 4, transcripts for both K_v4.2 and K_v4.3 were found at high levels in all three subtypes of BNST_ALG neuron. Finally, transcripts for all three I_K(IR) channel subunits (Kir2.1-Kir2.3) were found in every BNST_ALG neuron tested irrespective of subtype, however, the individual subunits were not uniformly expressed. Hence, Kir2.1 expression was highest (83%) in Type I neurons compared to other two subtypes (29% in Type II, and 45% in Type III). Conversely, the expression of Kir2.2 and Kir2.3 was highest in Type III neurons (64 % and 72%) compared to Type II (20 %, and 62%) and Type I neurons (0%, and 17%), respectively.

Table 4.

Percentage of mRNA expression of ion channel subunits in BNST_ALG neurons

Membrane currents	Encoding genes for channel subunits	Percentage of expression
		Type I	Type II	Type III
Ih	HCN1	0	91	0
	HCN2	83	97	9
	HCN3	0	44	0
	HCN4	83	68	100
IT	Cav3.1	0	91	0
	Cav3.2	0	0	0
	Cav3.3	17	47	100
IA	Kv1.4	0	6	0
	Kv3.4	17	0	9
	Kv4.1	0	0	0
	Kv4.2	100	62	91
	Kv4.3	100	32	9
IK(IR)	Kir2.1	83	29	45
	Kir2.2	0	20	64
	Kir2.3	17	62	72
	Kir2.4	0	0	0

To ensure that the relative subunit expression profiles outlined above were due to genuine differences in transcript expression and not due to limits in our PCR detection sensitivity, we ran a quantitative RT-PCR comparison for two genes; Ca_v3.3 and K_v4.3, which showed opposite expression patterns in Type I–III neurons (see Table 4). As illustrated in Figure 4, a plot of relative fluorescence intensity as a function of PCR cycle number showed that the threshold C_t value for the Ca_v3.3 transcript in Type I – III BNST_ALG neurons was significantly different (Type III C_t = 33±0.15; Type II C_t = 36±0.23; and Type I C_t = 39±0.35). The expression pattern is consistent with the mRNA expression pattern observed for Ca_v3.3 gene transcripts using our qualitative PCR detection method, where Ca_v3.3 mRNA was expressed in 11/11 Type III neurons, 11/34 Type II neurons, and only in 1/6 Type I neurons. Similarly, for K_v4.3, the C_t values are significant (P<0.0001) for (Type I C_t=34±0.27; Type II C_t=36±0.12; Type III C_t=39±0.21). The percentage of expression pattern is also consistent and shown in the Table 4.

Figure 4.

Real-time RT-PCR measures Ca_v3.3 and K_v4.3 subunits in BNST_ALG neurons (Type I-III). Mean C_t values for Ca_v3.3 subunit in BNSTALG neurons are Type I (39±0.35), Type II (36±0.23) and Type III (33±0.15). The C_t values are significant among each other at *P<0.0001. Similarly, for K_v4.3, the C_t values are significant (*P<0.0001) for Type I (34±0.27), Type II (36±0.12) and Type III (39±0.21). The C_t values of reference gene 18S rRNA among three neurons are (Type I= 24.8±0.21; Type II= 24.6±0.29 and Type III=24.2±0.19)

Classification of BNST_ALG neurons using cluster analysis of ion channel subunit expression and physiological properties

To determine if the genetic subgroups identified above truly correlate with our previous physiological classification of BNST_ALG neurons, we ran an unbiased hierarchical cluster analysis, which incorporated all 16 genetic variables and 6 physiological variables for each neuron; using analytical techniques we have previously reported to determine group membership. This analysis yields the dendogram illustrated in Figure 5, and segregated the sample population into three distinct clusters, which is consistent with our previous observation. We next designated each neuron Type I – III based solely on their physiological properties as previously defined. Significantly, out of a total of 51 neurons, the largest dendrogram cluster (n=34) was comprised almost exclusively of Type II neurons. The remaining neurons also fell into two distinct clusters; one of which was exclusively comprised of Type I neurons (n=4), and the second which was comprised exclusively of Type III neurons (n=10). The unbiased cluster analysis did identify 2 Type I neurons and 1 Type III neuron that were clustered together with the Type II neurons. Nevertheless, this cluster analysis not only validated our original classification of BNST_ALG neurons into three distinct subtypes, but also showed a unique view into the genetic determinants of their physiological phenotype. Moreover, a subsequent discriminant function analysis (DFA) showed that the separation of BNST_ALG neurons into three distinct groups (Types I–III, Figure 6) was based on two discriminant functions. Both discriminant functions 1 and 2 significantly contributed to group membership (Wilks’ Lambda, p < 0.001, Eigenvalues = 33.7, 8.2, corresponding to 80.4% and 19.6% of the variance, respectively). The contribution of each variable to each discriminant function can be seen in the structure matrix of Table 5. Variables that are correlated with a particular discriminant function segregate cells most strongly along this function’s axis relative to other functions. The first discriminant function distinguished cells based on expression of Ca_v3.1, HCN1, K_v3.4, Kir2.2, HCN3 and spike threshold. The second function distinguished cells based on HCN2, K_v4.3, Ca_v3.3, K_v1.4, Kir2.1, K_v4.2, HCN4, Kir2.3, I_h I_ISI1, I_K(IR), action potential half width and membrane potential. The discriminant scores for each cell were determined for discriminant functions 1 and 2, and demonstrate that there is good separation of all three groups of BNST_ALG neurons in both discriminant functions (Figure 6).

Figure 5.

Hierarchical cluster analysis through dendogram. Individual neuron is represented by the intersection of branches on x-axis. Squared Euclidian distance between the group centroids is depicted by y-axis. Three groups were determined by analysis of within and between group distances. Type I, II and Type III BNST_ALG neurons are clustered into three groups.

Figure 6.

Separation of BNST_ALG neurons along canonical discriminant function 1 and 2. There is a good distribution of the cells into three distinct groups along both discriminant functions.

Table 5.

Structure Matrix coefficients

Variables	Discriminant Function 1	Discriminant Function 2
Cav3.1	0.397*	0.139
HCN1	0.208*	−0.037
Kv3.4	0.197*	−0.073
Kir2.2	−0.162*	−0.089
Spike Threshold	0.122*	0.017
HCN3	0.110*	0.038
HCN2	0.233	0.436*
ISI1	−0.236	0.414*
Ih	0.295	0.332*
IK(IR)	−0.065	−0.243*
Kv4.3	−0.048	0.231*
Half width	0.026	0.228*
Cav3.3	−0.033	−0.226*
RMP	0.157	0.223*
Kv1.4	0.179	0.195*
Kir2.1	0.097	−0.182*
Kv4.2	−0.067	−0.126*
HCN4	−0.061	−0.067*
Kir2.3	−0.051	0.066*

Pooled within-groups correlations between discriminating variables and standardized canonical discriminant functions. An asterisk (*) indicates the largest correlation between each variable and any discriminant function. The variable contributes most to the segregation of cells along that functions’ axis, relative to other discriminant functions.

Discussion

Here, we extend our previous report, which first characterized BNST_ALG neurons into three electrophysiologically distinct subtypes (Type I, Type II, and Type III), to show that each subtype also expresses a characteristic profile of transcripts for ion channel subunit mRNA that can be used to further differentiate the neuronal subpopulations. Using unbiased cluster- and discriminant functional analysis we show that BNST_ALG neurons display genetically distinct phenotypes that can explain subtle differences in physiologic function in the three subpopulations. For example, physiologically identified Type III neurons differ from Type I and Type II neurons by their expression of a prominent I_K(IR) current, minimal I_h current, and an apparent absence of I_T current. Consistent with this observation the present study has revealed that Type III neurons express all three Kir2.1-2.3 subunits, only the slow activating HCN4 subunit, and the Ca_v3.3 but not the Ca_v3.2 subunit. Significantly, the results of the current study not only validated our previous physiological classification of neurons in the BNST_ALG, but also suggested that three genetically distinct subtypes of Type II neurons may exist in the BNST_ALG, which could not be easily differentiated in our physiological recordings. These studies represent the first of their kind to examine the genetic fingerprint of BNST_ALG neurons. Our long-term goal is to use a genome-wide analysis to identify additional cell-specific gene expression profiles that may be modified by behavioral stress paradigms, or that could offer novel avenues for the pharmacologic manipulation of distinct subpopulations of BNST neurons.

I_h

In our electrophysiological recordings, we have shown that the I_h current in BNST_ALG neurons is bimodal; whereby some BNST neurons show a prominent, fast activating I_h, while others show a small, slow activating I_h (Hammack et al., 2007, Figure 3). Here, we showed that Type I and II neurons have a prominent, fast onset depolarizing sag in their voltage response to hyperpolarizing current injection, presumably mediated by a fast activating I_h current. In contrast, Type III neurons showed only a small, slow onset, depolarizing sag, presumably mediated by the activation of a slow I_h current (see Hammack et al., 2007 Figure 9). We hypothesized that the potentially different I_h kinetics among the cell types may be due to differing compositions of I_h channel subunits. The I_h channel subunits are encoded by four genes, HCN1- HCN4, and each subunit differs in its activation and inactivation kinetics (Robinson and Siegelbaum, 2003). Hence, HCN1 channels display the fastest activation kinetics, HCN2 have intermediate kinetics, HCN3 activates more slowly, and HCN4 shows the slowest kinetics (Santoro et al., 2000; Altomare et al., 2003; Mistrik et al., 2005). Previous in situ studies had shown that the BNST has high levels of expression for HCN1 mRNA, moderate levels for HCN3, and low levels of HCN2 and HCN4 mRNA expression (Monteggia et al., 2000). However, the distribution of HCN genes at the cellular level was not known. Here, we show that all four HCN genes are expressed by BNST neurons, however, consistent with our physiological observations transcripts for the channel subunit isoforms are differentially distributed across subtypes. Hence, Type I neurons that show an intermediate level of depolarizing sag express only HCN2 (83%) and HCN4 (83%). Type II BNST neurons, which show the pronounced depolarizing sag express high levels and combinations of all four HCN genes (HCN1-91%, HCN2-97%, HCN3-44% and HCN4-68%), and Type III neurons that lack a prominent depolarizing sag express only HCN4 mRNA. A similar consistency between HCN gene expression and I_h kinetics has been reported elsewhere in the brain. For example, hippocampal CA1 neurons strongly express HCN1 and exhibit a rapidly activating I_h current. In contrast, thalamocortical relay neurons express HCN2 and HCN4 and like Type I neurons exhibit a more slowly activating I_h current (Robinson and Siegelbaum, 2003).

I_T

The I_T current can be carried by channels encoded by the α₁ subunits of Ca_v3.1, Ca_v3.2 and Ca_v3.3 genes (Perez-Reyes 2003). I_T regulates action potential generation, as well as coordinating pacemaker and burst firing behavior (Huguenard and McCormick, 1992; Perez-Reyes, 1999). A prominent I_T current has been reported in neurons from multiple brain regions including the BNST (Rainnie, 1999), basal forebrain (Hoehn et al., 1993), temporal cortex (Sayer et al., 1993), piriform cortex (Magistretti and de Curtis, 1998), hippocampus (Zhang et al., 1993), amygdala (Kaneda and Akaike, 1989), reticular thalamus (Huguenard and Prince, 1992), hypothalamus (Akaike et al., 1989) and cerebellum (Kaneda et al., 1990). Previous in situ studies had shown that Ca_v3.1 mRNA was highly expressed in the BNST, whereas Ca_v3.2 and Ca_v3.3 showed only moderate to low levels of expression (Talley et al., 1999). However, these studies did not determine the relative distribution of the I_T channel subunits across different subdivisions of the BNST, nor could they determine the relative level of neuronal expression versus non-neuronal expression. In this study, we made the first foray into addressing this knowledge gap by characterizing the distribution of I_T channel subunits in the BNST_ALG and then examining cell specific expression patterns. Previous studies from our lab revealed that Type II neurons have a robust I_T current, whereas Type I and III neurons appear to lack I_T. We also reported that the I_T current in Type II neurons was completely blocked by 500 μM nickel, attenuated by 10 μM mibefridil, and had a time constant of decay of approximately 20 mS (Hammack et al., 2007). Consistent with these observations, in the present study we show that Type II cells predominantly express mRNA transcripts for Ca_v3.1 (91%), which has a time constant of decay of ~11ms, and an IC₅₀ for nickel of 250 μM (Klockner et al., 1999). Surprisingly, several Type III neurons were shown to express Ca_v3.3. It is not immediately apparent why no I_T current is observed in Type III neurons, however post-translational modification may prevent the functional expression of mature peptide in these neurons.

I_A

All three subtypes of BNST_ALG neuron possess an I_A current, suggesting that a common gene profile may be found in BNST neurons. Several genes encoding K_V channel subunits give rise to I_A currents, and these include K_v1.4, K_v3.4, K_v4.1, 4.2, and 4.3. However, in situ hybridization studies have reported low K_v1.4 expression in the lateral BNST (Lujan et al., 2003), and no K_v3.4 mRNA (Weiser et al., 1994). These observations were somewhat surprising since K_v1.4 and K_v3.4 subunits have been reported to be preferentially localized in axons where they are thought to regulate transmitter release (Vacher et al., 2008). Interestingly, the current study revealed that both Type I and Type III neurons express K_v3.4 mRNA transcripts, whereas Type II neurons appear to express K_v1.4 transcripts. Hence, release of neurotransmitter from Type II neurons may be differentially modulated compared to Type I and III neurons. Not only is the inactivation voltage of K_v1.4 channels (V_h= −62mV) more negative than that of K_v3.4 channel (V_h= −53mV), these channels also show a greater sensitivity to blockade by 4-aminopyridine (4-AP; ~10μM –vs- 300 μM). Moreover, the activity of K_v1.4 channels is thought to be tightly regulated by the activity of CaMKII and calcineurin. Significantly, calcineurin mRNA transcripts are almost exclusively found in Type II BNST neurons (Hazra and Rainnie, unpublished observation).

K_v4.1 - K_v4.3 transcripts are also reported to be differential distribution of throughout the brain (Serodio and Rudy, 1998). These authors reported that K_v4.1 expression in the CNS was very low compared with that of K_v4.2 and K_v4.3. The latter two transcripts were also shown to be differentially distributed, such that K_v4.2 was highly expressed in the caudate-putamen, hippocampus, reticular pontine nucleus, and several nuclei in the medulla, whereas K_v4.3 was expressed in the substantia nigra pars compacta, superior colliculus, Raphé nuclei and the amygdala. In the present study, we have shown that all four transcripts of K_v4 family mRNA are expressed in the BNST_ALG, however, our scRT-PCR experiments suggest that K_v4.1 transcripts may be preferentially expressed in glia or schwann cells. Interestingly, K_v4.2 and K_v4.3 transcripts were expressed in most BNST_ALG neurons.

Significantly, K_v4 proteins are believed to be somatodendritic I_A channels, where they are reported to play an important role in regulating dendritic excitability, back propagation of action potentials, integration of excitatory postsynaptic potentials (Hoffman et al., 1997; Ramakers and Storm, 2002), repetitive firing properties (Connor and Stevens, 1971) and action potential duration (for review see (Birnbaum et al., 2004). Our preliminary electron microscopic studies have shown that K_v4.2 is also localized in somatodendritic compartment of BNST_ALG neurons (Rainnie et al., unpublished observation). Consistent with these observations, we reported that the I_A current of BNST_ALG neurons has a dual exponential decay, with fast- and slow time constants (tau) of 17ms and 183 ms, respectively. Moreover, the I_A current was relatively insensitive to exogenous 4-aminopyridine (4-AP) application, with a significant block only occurring at concentrations > 5mM 4-AP (Hammack et al., 2007). It is notable, therefore, that a characteristic feature of the Kv4 family of I_A channels is their relative insensitivity to exogenous 4-AP (~ 5 – 9 mM) (Pak et al., 1991; Tseng et al., 1996) compared to Kv1.4 which is sensitive to low micromolar (~ 10 μM) concentrations (Stuhmer et al., 1989). In addition, Kv4.2 and 4.3 are reported to have multiple time constants of inactivation ranging from 15 ms to 370 ms depending on their assembly with chaperone proteins (Birnbaum et al., 2004). Together these data strongly suggest that the Kv4 family of I_A channel subunits is the principal mediators of the somatic I_A current in BNST_ALG neurons.

I_K(IR)

I_K(IR) play a pivotal role in the maintenance of the resting membrane potential as well as in the regulation of action potential duration (for a review, see (Nichols and Lopatin, 1997). To date, of the seven subfamilies of Kir channel that have been identified (Kir1- Kir7) based on their rectification properties and their regulation by intracellular messengers (Stanfield et al., 1994; Hibino et al., 2010) only Kir 2.0 channels are constitutively active with apparent time-dependent gating and strong inward rectification (Ishihara and Hiraoka, 1994; Stanfield et al., 1994). Moreover, the four isoforms of the Kir2 family of channels (Kir2.1 – Kir 2.4) are differentially distributed throughout the brain (Pruss et al., 2005). Hence, Kir2.1 is primarily expressed in the olfactory bulb, superior colliculus, neocortex, hippocampus, caudate putamen, and amygdala; Kir2.2 is extensively expressed in thalamic nuclei, brainstem, hypothamalus, basal ganglia and hippocampus; Kir2.3 is expressed in the olfactory bulb, basal ganglia, cortex and cerebellar purkinje cells; and Kir2.4 is almost exclusively expressed in brainstem motorneurons. In situ hybridization studies have also shown a differential expression of Kir2 channel mRNAs in the BNST (Pruss et al., 2005). Here, the expression of Kir2.1 mRNA was reported to be weak, whereas Kir2.2 and Kir2.3 mRNA are strongly expressed, and there is no detectable expression of Kir2.4 mRNA. In the present study, we show that all BNST_ALG neurons express at least one member of the Kir2 family of subunits, but only Type III neurons show robust mRNA expression for three channels; Kir2.3 (72%), Kir2.2 (64%) and Kir 2.1 (45%). Kir 2 channels are highly sensitive to blockade by extracellular barium (Day et al., 2005), and we have shown that barium induces a membrane depolarization and an increase in input resistance in BNST_ALG neurons (Hammack et al., 2007), suggesting that one or more Kir2 family of channels are functionally expressed in BNST_ALG neurons, and in particular Type III neurons. A similar barium-sensitive rectifying current has been reported in mouse frontal projection neurons that also express Kir2.3, Kir2.2 and Kir2.1 mRNA (Day et al., 2005).

Significance of diverse ion channel expression and modulation by neurotransmitters

Previous studies have shown that the activity of BNST neurons could be modulated by local release of neurotransmitters that are critically involved in stress, anxiety, and drug addiction, including norepinephrine (NE), dopamine (DA) and serotonin (5-HT) (Egli et al., 2005; Francesconi et al., 2009; Levita et al., 2004; Hammack et al., 2009). Hence, in the ventral BNST, increased NE release is thought to contribute to anxiety states elicited during protracted drug withdrawal (Aston-Jones and Harris, 2004). Our own studies have shown that local 5-HT release can bi-directionally regulate the excitability of BNST_ALG neurons, such that 5-HT_1A receptor activation functionally inhibits BNST neurons, whereas activation of 5-HT_2A/2C/7 receptors increases the excitability of these same neurons (Guo et al., 2009). Significantly, microinjection of 5-HT_1A receptor agonists into the BNST elicits an anxiolytic response in freely moving rats (Levita et al., 2004) and we have shown that this response can be attenuated by stress, which also increases the expression of 5-HT₇ receptor mRNA (Hammack et al., 2009).

Elsewhere in the brain these same neurotransmitter systems have been reported to directly modulate the activity of I_h, I_T, I_A and I_K(IR) channels, and we predict that neurotransmitter-induced modulation of ion channel activity will also occur in the BNST. For example, in rat dorsal root ganglion neurons activation of 5-HT₇ receptors mediates an increase in I_h (Cardenas et al., 1999). Similarly, 5-HT₇ receptor activation in neurons of the glomerulus enhances calcium influx through I_T channels (Lenglet et al., 2002). Conversely, 5-HT_2A receptor activation results in suppression of rhythmic firing activity in olivary neurons via the inhibition of I_T (Placantonakis et al., 2000). It is possible that the opposing actions of 5-HT receptor activation on I_T function result from coupling to distinct I_T channel subunits. Previously, we reported that Type I and II neurons express 5-HT₇ receptors, but not Type III neurons (Hammack et al., 2009). As shown in Table 3, all BNST neurons express the Ca_v3.3 I_T channel subunit but only Type II neurons co-express the Ca_v3.1 I_T channel subunit. If 5-HT₇ receptors were to exclusively couple to the Ca_v3.1 I_T channel subunit this would confer a mechanism by which 5-HT₇ receptor activation in the BNST could selectively modulate the activity of Type II neurons. Moreover, by using a molecular dissection approach we have identified several cell-specific expression patterns for the ion channel subunits, which can now be targeted for genetic manipulation by viral vectors expressing either dominant negative subunit constructs or siRNA (Jasnow et al., 2009). Moreover, we are using this same approach to map neuropeptide expression in BNST_ALG neurons, and we will look for any cross-correlation between neuropeptide expression and ion channel subunit expression patterns.

Mapping neuropeptide expression to genetic phenotype may be particularly informative with respect to Type II neurons. We have shown that three subpopulations of Type II neurons can be differentiated based on their transcriptome profiles. As illustrated in Figure 1, we primarily sampled neurons from the BNSTov, an area that expresses a diverse array of neuropeptides including corticotropin releasing factor (CRF), somatostatin, enkephalin, and neuropeptide Y (NPY; Walter et al., 1991; Swanson et al., 1989). Intriguingly, neuropeptides such as CRF and NPY are reported to have opposing actions on synaptic transmission (Kash & Winder, 2006), and anxiety-like behavior in response to stress (Nijsen et al., 2001; Ciccocioppo et al., 2003). We are currently investigating the possibility that the three transcriptome profiles may represent discrete markers for distinct subpopulations of peptidergic neurons in the BNST_ALG.

Recently we have developed a transgenic mouse line in which the expression of a green fluorescent protein (GFP) is driven by the promoter for corticotrophin releasing factor (Martin et al., 2010). These mice express GFP exclusively in the CRF-containing neurons of the BNST_ALG and central nucleus of the amygdala (CeA), as well as in the paraventricular nucleus of the hypothalamus (PVN). The CRF-GFP neurons of the BNST_ALG share many electrophysiological characteristics with Type III neurons of the rat BNST and we have studies in progress to compare and contrast the molecular phenotypes of these two cell populations. Finally, the identification of distinctive electrophysiological and molecular features for Type I-III BNST_ALG neurons will help to uncover their respective role(s) in the complex physiological and pathophysiological process in which they are involved, including anxiety and post-traumatic stress disorder.