Abstract

The supraoptic nucleus (SON) of the hypothalamus regulates body water and salt balance via the release of the peptide hormones vasopressin (VP) and oxytocin from magnocellular neuroendocrine cells (MNCs). Excitatory synaptic transmission in the SON is largely mediated by glutamate receptors. One subclass, N-methyl-D-aspartate (NMDA) receptors has been implicated in the release of VP from MNCs in the SON. Previously, we have used an in vitro model of supraoptic neurons to confirm the presence of functional NMDA receptors on SON neurons. In the present study we used this same model to further characterize the ion channel properties of NMDA receptors on these neurons. Our results indicate that SON neurons in culture retain magnocellular neuroendocrine characteristics such as bipolar or simple multipolar morphology, large cell body diameters as well as immunoreactivity for vasopressin neurophysin II (VP-NPII). Single channel recordings made from membrane patches of SON cells with MNC morphology express functional NMDA receptor ion channels, with ohmic behavior and multiple unitary conductances. These findings suggest that NMDA channels on SON MNCs may consist of NMDAR1 with multiple NMDAR2 subunits and validate the use of this model for the molecular study of magnocellular neuroendocrine cells.

Introduction

The SON plays a major role in neuroendocrine regulation of body water and plasma osmolality (Hatton 1990). MNCs of the SON secrete VP, a nonapetide with a six-member ring and an amidated glycine at the carboxy terminus. VP in the blood is associated with a carrier protein, neurophysin II (North 1987). VP acts to both constrict blood vessels and also to increase water reabsorption in the kidney, in turn, raising blood pressure and increasing body water volume. These functions are important during hemorrhage, or with dehydration during which blood plasma becomes hypertonic. The regulation of the synthesis and release of VP from the SON is thought to be regulated, in part, by the activation the NMDA-type of glutamate receptors expressed on MNCs (Amaya et al. 1999; Hu and Bourque 1992; Meeker et al. 1999; Morsette et al. 1998; Swenson et al. 1998; Xu and Herbert 1998).

NMDA receptors, a-amino-3-hydroxy-5-methyl-4-isoxalone (AMPA) receptors and kainate receptors are subclasses of glutamate receptors. These receptors flux ions across cell membranes in response to binding of glutamate. These ionotropic receptors are used by the majority of excitatory synapses in the central nervous system as well as in the SON (Meeker et al. 1993; Hollmann and Heinemann 1994). NMDA receptors are distinguished from the other glutamate receptor subtypes in a number of significant ways. NMDA receptors are gated both by ligand binding, and voltage due to a voltage-dependant block of the ion pore by the divalent cation magnesium (Nowak et al. 1984, Ascher et al. 1988). They are also delineated from other glutamate receptors by their requirement for two coagonists; glutamate, aspartate or NMDA at one site and glycine at another distinct site (Johnson and Ascher 1987; Currás and Pallotta 1996). In addition to permeability to potassium and sodium, they also have a considerable permeability to calcium (Ascher and Nowak 1988), a property that has implications for learning and memory (Malenka and Nicoll 1999) and ischemic neuronal death as well as neurodegenerative diseases (Choi 1988; Choi and Rothman 1990; Lee et al. 1999; Meldrum and Garthwaite 1990).

NMDA receptors are expressed as heteroligomers of receptor subunits, of which five have been cloned: a principle NR1 subunit, which is necessary for channel function, and four NR2 subunits NR2A, NR2B, NR2C and NR2D (Ishii et al. 1993; Kutsuwada et al. 1992; Moriyoshi et al. 1991; Meguro et al. 1992; Monyer et al. 1992), which are thought to modify the functional properties of the receptor-ion channel complex such as magnesium block, single-channel conductance, and glutamate affinity (Monyer et al. 1994; Meguro et al. 1993; Stern et al. 1994; Sucher et al. 1996). The subunit stoichiometry of NMDA receptors is still under debate as different studies suggest either a tetrameric, or at least pentameric configuration (Laube et al. 1998; Hawkins et al. 1999).

While expression of the NR1 subunit is thought to be nearly ubiquitous throughout the brain (Kutsuwada et al. 1992; Monyer et al. 1992) and is found in the SON (Al-ghoul et al. 1997; van den Pol 1994), the expression of the different NR2 subunit-type mRNA has been shown to vary regionally, as well as developmentally (Hollman and Heinemann 1994; Monyer et al. 1994). Robust mRNA expression for all of the NR2 subunits has been shown in the SON except for NR2A subunit, which shows only negligible expression (Al-Ghoul et al. 1997).

Our lab has shown colocalization of the NR1 and NR2B with both vasopressin and oxytocin suggesting that the MNCs express this subunit (Currás-Collazo et al. 2000; Decavel and Currás 1997) as well as developmental changes in the expression of the NR1 and NR2B subunit proteins in the SON (Currás and Dao 1998). We have also seen robust expression of the NR2D protein in homogenates of postnatal SON (Pak and Currás, unpublished observations). SON punch cultures appear to be a convenient model system to study MNCs as they express NR1 and NR2B, and possibly NR2D receptor subunits (Currás et al. 1998; Currás and Dao 1998).

The role of the NMDA receptors in the hypothalamus has been downplayed by some, due in part, to results obtained in earlier studies using less than optimal conditions for activation of the NMDA receptors (Gribkoff and Dudek 1990; van den Pol et al. 1990). Subsequent studies have, however, demonstrated a role of NMDA receptors in neuroendocrine function. Intracellular recordings of MNCs from superfused explants of rat hypothalamus in vitro exhibit bursting typical of in vivo activity in response to NMDA and endogenous glutamate ligands, but not AMPA (Hu and Bourque 1992). Further evidence of the importance of NMDA receptors has been reported by Yang et al. (1995) who report a significant NMDA receptor-mediated component when stimulating afferents to the SON. Moreover, dehydration results in significant changes in expression levels of the NR1 and NR2B subunits. (Decavel and Currás 1997; Currás-Collazo and Dao 1999) Changes in expression levels were specific to the SON and another magnocellular neuroendocrine nucleus, the Parventricular nucleus, and in the case of NR2B, could be reversed by rehydration.

Activation of NMDA receptors may also result in increased VP release or production since receptor antagonists block osmotic-induced release of VP (Swenson et al. 1998) as well as increased levels of heterogeneous nuclear RNA, which is correlated with gene expression (Amaya et al. 1999). Direct stimulation of NMDA but not AMPA receptors also results in significant VP release (Meeker et al. 1999; Morsette et al 1998). Further evidence comes from studies using intracerebroventricular infusion of the NMDA channel blocker, MK-801. This treatment had suppressed plasma VP levels by 81% and 75% by rats following chronic water deprivation or saline injection respectively. The same effects were not seen with the non-NMDA antagonists CNQX and DNQX (Xu and Herbert 1998). These findings suggest that the activation of NMDA receptors in the SON can have an effect on both the expression and release of VP.

Previous studies performed in our laboratory using single channel recording of cultured SON neurons helped establish the presence of functional NMDA receptors on SON MNCs (Currás et al. 1998). Our goal in the present study is to gain further insight into the ion channel properties of NMDA receptors measured from patches pulled from cultured MNCs. The high resolution of single-channel patch-clamp recording allows inspection of the conductance states of single NMDA channels, a property that reveals their subunit composition. Our results indicate multiple conductance states and ohmic channel behavior or NMDA receptor channels on SON neurons. Comparisons of the NMDA channel properties between MNCs and other central neurons may help to elucidate how NMDA receptor channels function in neuroendocrine processes.

Methods

SON "punch" culture preparation

SON "punch" cultures were prepared as previously described (Currás et al. 1998). Briefly, punches of tissue were harvested from the area immediately surrounding the optic chiasm and tracts from fetal rats at embryonic day 16-18 using a 23-ga needle with a modified tip attached to a 1ml syringe. Punches were plated onto etched glass coverslips (Bellco) previously coated with 0.1mg/ml of poly-D-lysine and grown in minimal essential medium (MEM) with 20% fetal bovine serum and 20ug/ml gentamycin. Cultures were stored in a humidified incubator at 36oC with 5% CO₂ and fed every 2-3 days with a 50% medium replacement. After 10 days in vitro (DIV), MNCs in culture were identified on the basis of their large size (Sofroniew and Glassman 1981), and their bipolar or simple multipolar morphology.


Vasopressin-neurophysin II immunoreactivity in adult tissue sections and SON cultured neurons

Brains removed from adult female rats and SON punch cultures were fixed in 4% paraformaldehyde in 0.1 M phosphate buffered saline (PBS, pH 7.4) for 1 hour (coverslips) or 48 hours (sections). Fifty mm sections were cut on a vibratome and sections were washed for 1 hour in PBS. Sections and coverslips were then treated with 0.5% sodium borohydride in PBS for 30 minutes and washed in PBS. Permeabilization with Triton-X (0.3% in PBS) was then performed for 15-30 minutes. Sections and coverslips were incubated with the rabbit anti-rat primary antibody (1:1,000), recognizing Neurophysin II, (gift from Dr. Alan Robinson) for 72 hours in PBS containing 0.3% Triton-X and 0.02% azide. After washing in PBS, the sections were incubated with PBS containing a flouroscein-conjugated goat anti-rabbit IgG (1:50-1:200) for 1.5 hrs (sections) at room temperature. Coverlip cultures were incubated instead in biotinylated goat anti-rabbit IgG (1:200) for 1 hour followed by washes and incubation in avidin-HRP (ABC reagent, Vector Elite Kit) made in PBS for 1 hour at room temperature. After further washes in PBS the sections were mounted wet in Vectashield (Vector Laboratories). Coverslips were treated with solution containing 0.25 mg/ml diaminobenzidine (DAB) with 0.2% NiSO₄ and 0.01% H₂O₂ in Tris buffer, pH 7.6, then dehydrated in successively greater concentrations of alcohol, delipidized in xylene and mounted with DPX. Sections and coverslips in the control group were processed similarly but without the primary antibody. Sections and coverslips were visualized using a NIKON FXA upright microscope with bright-field and epifluorescece optics.


Single channel recordings

Cultured neurons on glass coverslips were placed in a recording chamber with a 1.5ml volume and continuously perfused with an external solution of (mM): NaCl (150), KCl (2.5), CaCl₂ (2) HEPES (10) pH 7.4 using NaOH; 290-300mOsm/kg at room temperature. Test solutions contained either 100 mM or 300 mM NMDA (Research Biochemicals), and 10 mM glycine in external bathing solution and were either added to the bath or applied with a U-tube (Krishtal and Pidoplichko 1980). Flow of solutions for the U-tube was gravity driven and controlled by a 2-way solenoid valve. The perfusion of the bath chamber flowed continuously (1-2 ml/min) so the test solution could quickly be removed from the location of the patch.

Patch pipettes were pulled from N51A borosilicate glass (Drummond Scientific) and coated with Sylgard-184 (Dow Corning) to reduce pipette capacitance. Pipettes with a resistance of 5-15 MOhms were used in recordings. Internal solution in the pipette included (mM): CsCl (145) EGTA(5) KCl (2.5) CaCl₂ (1) HEPES (10); pH 7.2-7.3 with CsOH, 260-270 mOsm. In some experiments electrodes were filled with an internal solution containing biocytin (0.5%) to allow the labeling of the cell from which patches were obtained. A gigaseal (gigaohm resistance) was formed between the pipette and the cell membrane. After a gigaseal was achieved, the patch of membrane under the pipette tip was broken by applying negative pressure, resulting in the whole-cell configuration. The cell was then voltage clamped at -60 to -40mV. For cells in which biocytin was used the cell was kept in the whole-cell configuration for approximately 5 minutes to allow transfer of biocytin into the cell by diffusion. The pipette was then pulled away from the cell to obtain an outside-out patch of the cell membrane. Single-channel currents were measured using an Axopatch 200 A amplifier (Axon Instruments) and stored on magnetic tape. PCLAMP software and a PC computer were used for analysis of recorded single-channel events.

During the initial minutes of each recording patches were challenged with 2-3 second pulses of external solution alone to test for the presence of mechanosensitive channels identified in acutely dissociated SON MNCs which would yield inward currents in the absence of NMDA (Oliet and Bourque 1993). Patches were then voltage-clamped at one or several potentials while NMDA was pulse applied by U-tube (300 mM NMDA) or bath applied (100 mM NMDA) in the presence 10 mM glycine to examine the NMDA single-channel activity in outside out patches. An interval of 15 seconds in between pulse applications of agonist was allowed for recovery of channel desensitization and return to the open state.

The single channel conductance of single channels was measured using a modified Ohm's law equation g= I/ V_m-E_i where g is the single channel conductance, I is the single channel current, V_m is the holding potential and E_i was the reversal potential which was set at 0, an approximation based on previous determinations (Currás et al. 1998).


Staining of recorded neurons

Coverslips containing neurons filled with biocytin were fixed with 4% Paraformaldehyde for 10 min-1 hr at room temperature and washed in PBS by dipping 8-10 times. Coverslips were incubated in avidin-HRP (ABC reagent, Vector Laboratories) made in PBS for 1 hr at room temperature. The staining was visualized with DAB solution and coverslips were dehydrated and mounted in the same manner as coverslips from immunocytochemistry experiments.