Rat lactotrophs isolated by fluorescence-activated cell sorting are electrically excitable

~#lee~~ar and Cellular E~docri~olo~, 51 (1987) 201-210 EIsevier Scientific Publishers Ireland, Ltd.

201

MCE 01660

Rat lactotrophs isolated by fluorescence-activated are electrically excitable

cell sorting

Guo-Guang Chen * , Paul A. St. John and Jeffery L. Barker Laborarmy o~~~ro~~.~sio[o~,

~~~~D~-N~~,

(Received

Key words: Prolactin;

Pituitary;

Flow cytometry;

7 October

Action

Building 36. Room tC02, Bethesda, MD 20892, U.S.A. 1986; accepted

potential;

Sodium

S February

conductance;

1987)

Calcium

conductance;

Electrophysiology

Summary

Lactotrophs (prolactin-containing cells) from the anterior pituitary of the adult female rat were labeled by a cell-surface reaction with anti-prolactin antibodies and then isolated by fluorescence-activated cell sorting (FACS). FACS-isolated pituitary cells were maintained in culture 1-9 days, after which their excitable membr~e properties were examined using the whole-cell patch recording technique. Comparisons between lactotrophs sorted at different laser-power settings on the FACS showed that although the electrical excitability of the sorted lactotrophs was acutely altered by exposure to high laser power, excitability comparable to unsorted cells was retained by the use of minimal laser power (10 mW). Recordings were made from 67 cells that had been isolated using the low laser power. These cells had resting membrane potentials ranging from - 22 mV to - 60 mV. More than 80% of the cells responded to depolarizing current injections with regenerative, full-amplitude, overshooting action potentials. Approximately 20% of these cells exhibited spontaneous 20-30 mV fluctuations in the level of resting membrane potential, the majority of which did not overshoot 0 mV. Both the action potentials and the spontaneous fluctuations involved Na+ and Ca ‘+ ion conductance mechanisms.

Introduction

The anterior lobe of the pituitary gland contains several different classes of cells that secrete different peptide hormones. A great deal is known about the effects of various hormones and neurotransmitters on secretion from the intact pituitary, both in vivo and in vitro. Relatively little information, however, is available concerning secretion * Present address: Dept. of Neuroscience, Albert Einstein College of Medicine, 1300 Morris Park Ave., Bronx, NY 10461, U.S.A. Address for correspondence: Paul A. St. John, Laboratory of Neurophysiology, NINCDS-NIH, Building 36, Room 2CO2. Bethesda, MD 20892, U.S.A. 0303-7207/87/$03.50

a 1987 Elsevier Scientific

Publishers

Ireland,

and its alteration at a cellular level. For example, what electrophysiological and biochemical properties underlie hormone secretion and its alteration? Do all the cells that secrete a particular hormone respond in the same way to secretagogues? Does a particular neurotransmitter act directly on the cells whose secretion is altered? Some of these questions have been addressed in studies of clonal pituitary cell lines such as GH, which secrete both growth hormone (GH) and prolactin (PRL), and AtT-20, which secrete adrenocorticotropic hormone (ACTH). The cells in these tumor lines, however, can be heterogeneous with respect to certain important properties. Moreover, the cytoplasmic and membrane properLtd

202

ties of cells in these lines have not yet been closely compared with those of primary pituitary cells. In order to learn more about secretion and its control in normally differentiated pituitary cells, we and others have begun to study cells dissociated from the pituitaries of adult rats and maintained in vitro. Several strategies have been taken in other laboratories to identify and, in some cases, isolate various classes of pituitary cells for study in vitro. We have used a method we recently developed (St. John et al., 1986b) to identify prolactin-containing cells from the anterior pituitary and to isolate these cells by fluorescence-activated cell sorting (FACS). In this preliminary report, we describe some excitable membrane properties of FACS-isolated prolactin-containing rat pituitary cells maintained in vitro. Materials and methods The procedures for preparing cells, labeling dissociated cells, and analyzing and isolating the cells by cell sorting were those described previously (St. John et al., 1986a, b). Briefly, pituitaries from non-pregnant female Sprague-Dawley rats (200-250 g; Taconic Farms, Germantown, NY) were removed immediately after decapitation and transferred to buffer A (calciumand magnesium-free Hanks’ balanced salt solution (Gibco, Grand Island, NY) containing 25 mM Hepes (pH 7.3) and 4 mg/ml bovine serum albumin (BSA; fraction V from Sigma Chemical Co., St. Louis, MO)). Anterior lobes were dissected free of intermediate and posterior lobes and minced gently. The tissue was then incubated with shaking at 36” C for 30 min in buffer A containing 2 mg/ml collagenase (Boehringer-Mannheim, Indianapolis, IN), followed by 30 min in buffer A containing 2 mg/ml collagenase, 1 mg/ml trypsin (type TRL3 from Cooper Biomedical, Malvern, PA) and 1 mg/ml deoxyribonuclease (type I, Sigma Chemical Co.). The partially dissociated tissue was centrifuged at 300 x g for 5 min at room temperature, the pellet was resuspended in buffer B (Hanks’ balanced salt solution with 1.3 mM calcium and 0.8 mM magnesium and containing 25 mM Hepes, pH 7.3, and 4 mg/ml BSA), which was used for all subsequent steps. Final dissociation was achieved by five pas-

sages through a Pasteur pipette with a fire-polished tip. The resulting cell suspension was rinsed by centrifugation (300 X g for 5 min) and resuspension. In order to remove red blood cells, cell suspensions containing approximately 10’ cells in 2 ml were layered onto 4 ml of buffer B containing 200 mg/ml BSA and centrifuged at speed No. 1 of a bench-top clinical centrifuge (approximately 50 X g) for 10 min, and the pituitary cells were recovered in the pellet. The final yield was normally 1.2-1.5 X 10” cells per pituitary, and viability, determined by exclusion of trypan blue or by staining with acridine orange and ethidium bromide (Parks et al., 1979; St. John et al., 1986a), was over 90%. Labeling was performed using rabbit antibodies against rat prolactin (NIADDK-anti-rPRL-IC-1) at a final dilution of 1 : 1000 in buffer B. Cell suspensions to be used for labeling of cell-surf&e antigens were incubated in cell culture medium (see below) for 2 h at 37 o C following dissociation and before further processing. Samples were then transferred to 4°C for all subsequent steps. Cells were centrifuged at 300 X g for 5 min, then resuspended at a density of l-2 X 10’ cells/ml in diluted primary antibody and incubated for 60 min. The cells were rinsed twice by centrifugation, then incubated in fluorescein- or rhoda~ne-conjugated secondary antibody (affinity purified F(ab’), goat anti-rabbit IgG diluted 1 : 50 in buffer B; antibody from Jackson Immunoresearch, Avondale, PA) for 60 min. After this incubation, cells were rinsed and held at 4” C until further use (at this temperature, labeling remained stable for at least 5-6 h after the final step). Cell suspensions or cultures to be used for immunocytochemical labeling of intracelhiur antigens were fixed in 4% formaldehyde in 100 mM phosphate buffer, pH 7.2, for 30-60 min at 4°C. Samples were rinsed 3 times in buffer B, made permeable by treatment for 5 min in buffer B containing 5 mg/ml saponin (Sigma Chemical Co.), and rinsed again before incubation in primary antibody. Incubations in primary and secondary antibodies were the same as described above for unfixed cells in suspension. Fluorescence microscopy was performed on a Leitz Dialux 22 equipped with a standard rhodamine filter set, a narrow-band fluorescein filter

203

set, and a 100 W mercury-arc lamp for epi-illumination. Samples were examined at final magnifications of 160 X to 1000 X with fluorescence objectives and phase-contrast optics. Cell sorting and analysis were performed on a FACS 440 (Becton-Dickinson, Mountain View, CA) using methods developed for analysis and sorting of mammalian neurons and primary pituitary cells (St. John et al., 1986a, b). Exciting light at 488 nm came from a 5 W argon ion laser. In each experiment, control samples were used to define a ‘threshold’ level of fluorescence intensity that was higher than that of 99% of unlabeled cells. Cells with fluorescence more intense than this threshold were considered ‘positive’; cells with intensities at least 10% higher than the threshold intensity were isolated by sorting. Sorting isolations were performed with the laser power set at either 200 mW or 10 mW (see Results). All experiments were performed with 70 pm nozzle tips and sorting rates of 2000-2500 cells/s. Cell culture Dissociated pituitary cells were placed in culture at an initial density of 2-3 x 104/cm2 in 35 mm Primaria (Falcon) tissue culture dishes. Culture medium, consisting of Ham’s F-10 (Gibco, Grand Island, NY), 2.5% fetal bovine serum (Gibco), 12.5% horse serum (Hazleton-Dutchland, Denver, PA), 2 mM glutamine (Gibco), and 1 mM sodium pyruvate (Gibco), was changed every 3-4 days. Electrophysiologv For electrophysiological recording experiments, culture medium was replaced with buffer containing 142 mM NaCl, 5.3 mM KCl, 2.0 mM CaCl,, 2.0 mM MgCl,, 5 mM glucose, and 10 mM Hepes (pH 7.2-7.4). In some experiments, 4 mM CoCl, was substituted for CaCl,. For sodium-free solutions, choline chloride was isosmotically substituted for NaCl. Osmolarity was maintained constant in all other modifications of the recording solution. All recordings were obtained at room temperature (22-24 o C). Whole-cell patch recordings were made from single cells under phase-contrast optics on the modified stage of an inverted microscope using either a homemade circuit (Smith et al., 1981) or

an Axoclamp amplifier system (Axon Instruments, Burlingame, CA). The recording electrodes were filled with a solution containing (in mM): 140 K+ gluconate, 2 MgCl,, 2.4 NaCl, 1.1 EGTA (resulting in a final intracellular concentration at or below 10 nM), 5 Hepes, adjusted to pH 7.2 with KOH and 310 mOsm with sucrose. Tip resistances of the electrodes in the recording medium were 4-8 MS2. Recordings under current-clamp were made through a bridge circuit that allowed simultaneous injection of current and measurement of voltage. Injected current and membrane voltage were digitized, acquired, and stored on a PDP 11/40 computer for off-line analysis.

1 a. control

1 b. anti-rPRL

Fig. 1. FACS analysis of cell-surface labeling of dissociated anterior pituitary cells by anti-prolactin antibodies. Figure shows correlated dual-parameter histograms for forward-angle light scatter (on a linear scale) and fluorescence intensity (on a logarithmic scale). Note that sample incubated with anti-PRL antibodies (b)contains approximately 20% of the cells (large arrow) with fluorescence more intense than that found on cells in a control sample that was incubated with secondary antibody alone (u).

204

205

Results Identification and isolation of lactotrophs In previous experiments, we found that antibodies against prolactin (PRL) bound to approximately 20-30% of the live cells dissociated from the anterior pituitary (St. John et al., 1986b). The binding could be detected both by fluorescence microscopy and by fluorescence-activated cell sorting (FACS). Binding was confined to the level of the cell surface (St. John et al., 1986a), and because the labeled cells excluded even lowmolecular-weight dyes, we concluded that the antibodies must have been bound to the cells’ outer surfaces. In the present experiments, labeling of live anterior pituitary cells by anti-PRL antibodies was examined almost entirely by FACS and was similar to that previously described. The fluorescence observed on the labeled cells ranged in intensity from just above that of cells in control samples (Fig. 1) to relatively high levels that were off the scale usually employed (not shown in Fig. 1). Anterior pituitary cells labeled on the cell surface by anti-PRL antibodies were isolated by cell sorting and placed in culture. After several days (Fig. 2), the cultures included both phasebright cells that contained pituitary hormones and did not appear to divide in culture and very flat, phase-dark cells that contained no detectable pituitary hormone (St. John, unpublished). Although the latter accounted for only a few percent of all sorted cells at the time of sorting, they appeared to multiply continuously in culture so that their number had increased dramatically after 4 days (St. John, unpublished).

Previous analyses have shown that the population isolated by this method is more than 90% pure with regard to cell-surface labeling and between 85% and 98% pure with regard to immunocytochemically detectable intracellular PRL (St. John et al., 1986b). The high purity of hormonal content among the sorted cells was confirmed in the present study. Cells labeled on the cell surface by anti-PRL antibodies, isolated by cell sorting, and placed in culture were found by immunocytochemistry to contain intracellular PRL in almost all (95-100%) of the cells (Fig. 2a-d). This was the case for cells studied in cultures ranging from 1 to 9 days old. No obvious differences in the percentage of cells labeled intracellularly were observed among cultures of different ages. In many cells, the anti-PRL immunoreactivity was found in a peri-nuclear distribution within the cell (Fig. 2g and h). As expected from previous observations (St. John et al., 1986b), cultures of cells not detectably labeled on the cell surface and isolated by sorting also included some PRL-containing cells (Fig. 2e and f). Electrophysiological recordings were made from cells that had been maintained in vitro for 1-9 days (see below). Cells from the anterior pituitary, whether sorted or unsorted, attached to tissue culture dishes within 1-2 h after plating and became progressively flatter with time in culture (see also Taraskevich and Douglas, 1977). In most experiments, recordings were made from cells that had been in culture for 1-4 days after isolation, since it was easier to obtain adequate recordings from the less-flattened cells. Initial FACS isolations were performed using conditions that had been developed for sorting

Fig. 2. Intracellular anti-PRL immunocytcchemistry in anterior pituitary cells isolated by cell sorting and maintained in culture. Dissociated cells were incubated with anti-PRL antibodies followed by FITC-conjugated secondary antibodies, then both labeled (‘surface-positive’) and not-detectably-labeled (‘surface-negative’) cells were isolated by cell sorting and cultured for 4 days. Cultures were fixed and made permeable, then intracellular PRL was detected by immunocytochemistry using anti-PRL antibodies followed by RhITC-conjugated secondary antibodies. Cultures include phase-bright cells that contain pituitary hormones and very flat, phase-dark cells that contain no detectable pituitary hormone (St. John, unpublished). o and b: Sorted and cultured ‘surface-positive’ cells labeled by anti-PRL antibodies and viewed by phase-contrast (a) and fluorescence (b) optics. c and d: Sorted and cultured ‘surface-positive’ cells exposed to secondary antibodies alone (control) and viewed by phase-contrast (c) and fluorescence (d ). e and f: Sorted and cultured ‘surface-negative’ cells labeled by anti-PRL antibodies and viewed by phase-contrast (e) and fluorescence (f). g and h: Sorted and cultured ‘surface-positive’ cells labeled by anti-PRL antibodies and viewed by phase-contrast (g) and fluorescence (h). Bar in b = 50 pm and applies to u-f. Bar in h = 25 pm and applies to g and h. The presence m sorted ‘surface-negative’ cultures of many cells that do not contain detectable PRL along with those that do indicates that the ubiquitous expression of intracellular PRL among the sorted ‘surface-positives’ was not simply induced by the culture conditions.

206

mammalian neurons (St. John et al., 1986a) and that had been used previously to isolate viable anterior pituitary cells (St. John et al., 1986b). These conditions included running the laser power at 200 mW. Although passage through the cell sorter under those conditions had no detectable effect on the viability of pituitary cells, it did appear to have at least one small effect on the electrophysiological properties of the sorted cells. Cells that had been isolated in this way displayed slightly diminished excitability in response to depolarizing current pulses (see below), compared to unsorted pituitary cells plated at the same time. Separate testing of each step involved in passage through the cell sorter established that the alteration of membrane excitability was caused by exposure to the laser beam. The laser-induced deficit could be avoided, however, by operating the laser at much lower power (10 mW). In this case, the action potentials of sorted cells were not distinguishably different from those of unsorted cells. Spontaneous electrical activity under resting conditions Most PRL-containing pituitary cells isolated by cell sorting (54 of 67) displayed little, if any, spontaneous membrane potential activity under the recording conditions used here (recordings in standard recording solution at room temperature). However, most were electrically excitable, as described below. Thirteen of 67 cells recorded with the whole-cell patch technique exhibited spontaneous fluctuations in membrane potential (Fig. 3). Two of the 13 cells generated full-amplitude (40-60 mV) action potentials that overshot 0 mV and typically lasted less than 100 ms (Fig. 3Al). The depolarizing phase of these potentials was relatively sluggish, with the threshold for triggering all-or-none rapid depolarization having about a 5 mV range near - 30 mV. Prominent after-hyperpolarizations were apparently absent despite the relatively depolarized resting potential. The frequency of action potentials recorded soon after breaking through the intra-pipette membrane seal was higher than later in the recording (Fig. 3A2). The initial pattern of activity appeared to be random, although clusters of action potentials sometimes

Al

A2

Bl

B2

Fig. 3. Spontaneous action potentials and voltage fluctuations of PRL-~ont~ning pituitary cells in culture. A: Spontaneous, overshooting action potentials without a clear pattern of activity were recorded using the whole-ceil patch recording technique in a cell that had been in culture for 2 days after FACS isolation. AI: Within 30 set after the start of recording (resting membrane potential, RMP = - 38 mV). Overshooting action potentials 50 mV in amplitude are evident. A2: After 2 min of recording, the action potentials have decreased in frequency and the RMP is -44 mV. The baseline membrane potential exhibits continual low-amplitude activity. 8: Spontaneous voltage fluctuations. Bl: Seconds after the start of recording from a 4-days-in-culture cell whose RMP is -25 mV, spontaneous. apparently random, 20 mV fluctuations in membrane potential. none of which overshoot 0 mV, are apparent. The activity looks action-potential-Ike because of the slow time base used for recording and illustration. 82: After 40 min recording the fluctuations are still present, though at diminished frequency and occasionally they overshoot 0 mV. The RMP is now -. 36 mV and considerable low-amplitude ‘noise’ in the trace is also obvious. C: Spontaneous voltage fluctuations observed in Na *-free recording medium. The ceil was recorded in choline-cont~~ng medium. Randomly occurring waves in membrane potential, similar to those illustrated in panel B, were present at the start of recording from a cell (RMP = - 24 mV) but by 8 min these have run down in amplitude, though not in frequency. D: Spontaneous voltage fluctuations observed in Ca”+-free recording medium. The cell was recorded in medium nominally free of Ca’+ and containing 4 mM Co”. DI: 20 mV fluctuations at about 1 Hz are present at an RMP of -32 mV. DZ: The fluctuations are abolished by sustained injection of hyperpolarizing current (between the arrows) that polarizes the cell to -60 mV. Removal of the hyperpolarizing current results in a rebound of the membrane voltage transiently to a more depolarized level than at rest, followed by a gradual repolarization to the resting level without further evidence of any spontaneous

activity.

were collected loosely into a burst. The baseline of cells expressing spontaneous overshooting action potentials was not smooth but fluctuated over

207

about a 5 mV range. This irregular baseline was characteristic of most of the cells recorded. Another form of fluctuating membrane potential activity distinguishable from action potentials and irregular baseline both in terms of amplitude and duration was recorded in the remaining 11 cells (Fig. 3B). The voltage fluctuations generated in these cells were usually 20-30 mV peak-to-peak and did not overshoot 0 mV. They lasted for hundreds of milliseconds to seconds and occurred at a random frequency, which decreased (as did the amplitude) as the recording continued (Fig. 3B2). Full-amplitude action potentials could be evoked by injection of depolarizing current in cells expressing spontaneous fluctuations. The rundown of such activity may have been due to the gradual increase in membrane potential seen in stable recordings (see below). Cells that displayed spontaneous fluctuations had the same range of membrane potential (- 25 mV to - 55 mV) as those whose baseline electrical activity was relatively quiet. The spontaneous oscillations in potential persisted, albeit somewhat attenuated in amplitude, in Na+-free medium containing choline (Fig. 3C) and in Ca2+-free medium containing 4 mM Co2+ (Fig. 3D). Sometimes a pacemaker-like oscillation was recorded in the latter medium (Fig. 3Dl). These oscillations were depressed in amplitude by application of Cd’+ (not shown). In Ca’+-free medium the oscillations were eliminated by hyperpolarizing the cell (Fig. 3D2). Electrical properties of isolated Iactotrophs Whole-cell patch recordings were used to measure resting membrane potentials and input resistances in sorted cells that exhibited relatively quiet baselines. Resting membrane potentials fell in a wide range (-22 to -60 mV), the mean being -41.2 (+ 9.90 SD) mV in 52 cells. Input resistance, measured at the resting potential over a lo-20 mV range, was 269.4 f 119.5 MO in 18 cells. More than 80% (55/67 cells examined) of the PRL-containing cells isolated by the above procedure generated overshooting action potentials in response to depolarizing current pulses of sufficient intensity (Fig. 4A). All cells, both those that generated dizershooting action potentials and exhibited considerable those that did not, delayed-type rectification at depolarized potentials

Fig. 4. Membrane properties of electrically quiet PRL cells. A: A full-amplitude action potential can be evoked in a quiet ccl (cultured for 3 days; resting membrane potential = -54 mV) by injecting depolarizing current. The regenerative response overshoots 0 mV by 15 mV and has an obvious shoulder on its repolarizing phase. B: Current-voltage (I-V) relations in an inexcitable cell cultured for 6 days. Abscissa, current pulses (100 ms) were applied directly to the cell (RMP = -36 mV) through the whole-cell patch electrode using a bridge circuit (see Methods). Ordinate, the magnitude of the membrane potential response measured at 10 ms (circles) and at 90 ms, near the end of the current pulse (squares), in response to current injections. Inset shows some of the actual voltage records. Note that a full-amplitude action potential is not triggered, that the instantaneous I-V relations are relatively ohmic, and that clear non-linear components comprise the ‘steady-state’ I-V curve. Thus, delayed forms of rectification at depolarizing and hyperpolarizing potentials are evident.

(Fig. 4B). In addition, many cells displayed anomalous rectification at hyperpolarizing potentials. Like the spontaneous fluctuations of membrane evoked action potentials in sorted potential, PRL-containing anterior pituitary cells appeared to include both Na+ and Ca2+ conductance mechanisms. Application of the Naf-channel blocking agent tetrodotoxin (TTX, 1 PM) by pressure ejection to cells in physiological recording medium blocked the regenerative, overshooting action potential in several cells tested (Fig. 5A). Substitution of choline chloride for sodium chloride in the recording medium produced the same effect (not shown). Similarly, application of 2 mM Co2+ (Fig. 5B) or 200 PM Cd’+ to cells in normal recording medium eliminated the action potential, as did close application of Ca’+-free medium (not shown). In some cells, remnants of regenerative depolarizing activity remained in Na+- or Ca’+free medium. Application of Cd’+ to cells in Na+-free medium completely abolished all detec-

208 A

B

:\ !

‘,

,(I‘,L_

3

BEFORE

;!

BEFORE

‘L+

I V

COCl*

I

J

v

-I

,

_r---

-_

SOmV

I

0.76

‘,

:

7_.

._

-I

nA

20

_ . .

30 mV 0.75 nA

Ins

action potentials in quiet Fig. 5. Naf- and Ca 2+-dependent PRL cells. A: An overshooting action potential evoked by depolarizing current injection is blocked by application of 1 PM TTX to a cell, cultured for 2 days, whose RMP was -50 mV. B: A regenerative, overshooting action potential evoked by depolarizing current injection is blocked by close application of 2 mM CoCl 2 to a cell cultured for 4 days (RMP = - 44 mv).

table active response, leaving only passive depolarization, while application of TTX to cells in Ca’+-free medium blocked all active membrane responses. Finally, thyrotropin-releasing hormone (TRH), which acts at nanomolar concentrations to trigger the release of prolactin from pituitary cells in vitro, was applied to 26 cells by pressure ejection. Since the electrical response to TRH in clonal prolactin-secreting (GH,) cells recorded with the whole-cell patch technique rapidly disappears with a half-time of 5-10 min after the start of the recording (Dufy et al., 1986), most of the cells in the present experiments were tested within 3-4 rain after initiation of the recording. Nevertheless, none of the cells displayed any change in membrane potential or membrane conductance in response to 50 nM TRH. Discussion The experiments presented that PRL-containing cells from itary of the adult female rat can sorting, cultured in isolation for

here demonstrate the anterior pitube isolated by cell a period of days,

and their electrical activity studied with whole-cell, patch-type recording techniques. The cells recorded were first labeled by indirect immunofluorescence using polyclonal antibodies against rat PRL followed by FITC-conjugated secondary antibodies. Cells labeled by this method were isolated by fluorescence-activated cell sorting and maintained in culture for several days. In the present series of experiments, as in a previous study (St. John et al., 1986b), the phenotype of the isolated cells was identified by immunocytochemistry. At least 95% of the cells in these cultures expressed immunocytochemically detectable PRL. Thus, although the particular cells studied electrophysiologically were not themselves identified by immunocytochemistry after recording, it is likely that almost all of them were lactotrophs. Approximately 80% of the isolated lactotrophs studied with patch-type recording methods were electrically excitable in that depolarizing current pulses of sufficient intensity evoked full-amplitude action potentials. Two types of experiments showed that these action potentials included both Ca2+-dependent and Nat-dependent components. First, elimination of one of these ions from the extracellular medium either diminished the amplitude or eliminated the action potential. Second, blockade of voltage-gated Ca*+ or Na+ conductance by application of Cd*+ or TTX, respectively, also diminished or blocked the action potentials. Approximately 20% of the cells examined in the present experiments displayed sizable spontaneous fluctuations of membrane potential under resting conditions. In two of these 13 cells, the fluctuations were full-amplitude action potentials, while in the remainder, the fluctuations did not overshoot 0 mV. Like the evoked action potentials observed in most of the cells, the spontaneous fluctuations appeared to involve both Ca*+ and Na+ conductance mechanisms. With the whole-cell patch recording technique used in these experiments, the fluctuations gradually disappeared after the first few minutes of recording. The electrophysiological properties of clonal pituitary tumor cells secreting PRL and/or growth hormone have been studied with high-resistance/ high-electrolyte-concentration microelectrodes and with low-resistance patch pipettes containing physiological levels of intracellular ions. Such re-

209

cordings have shown that cultures of clonal pituitary cells include cells exhibiting spontaneous action potential activity, cells that are quiet but electrically excitable, and cells that are quiet and electrically inexcitable (e.g., Dufy, 1983). Both Na+-dependent and Ca2+-dependent conductances are expressed, depending on the subclone of the GH, line. There have been relatively few studies of membrane excitability in primary anterior pituitary cells. Taraskevich and Douglas (1977; see also Douglas and Taraskevich, 1985) used extracellular suction electrodes to record from anterior pituitary cells of unidentified hormonal type in vitro, and found that most of the cells were electrically excitable. The action potentials primarily involved Ca*+ conductance mechanisms with little, if any, evidence for Na+ conductance, although the authors could not exclude the presence of some of the latter. Vincent et al. (1985) reported similar results from anterior pituitary cells enriched for lactotrophs or for somatotrophs. More recently, investigators have begun to study anterior pituitary cells of particular phenotypes (e.g., Lingle et al., 1986; Mason and Waring, 1986). Lingle et al. (1986) have used whole-cell patch-clamp recording techniques to examine acutely cultured anterior pituitary lactotrophs identified in a mixed population by reverse hemolytic plaque assay. They have described two depolarizing conductance mechanisms involving Ca2+ ions but none involving Na+. However, De Riemer and Sakmann (personal communication), using virtually the same approach, have reported the presence of TTX-sensitive Na+ currents in some plaque-identified cultured lactotrophs. We have found evidence for both Na+ and Ca2+ conductances in lactotrophs identified by cellsurface antibody binding. There are several possible explanations for the apparent inconsistencies between the present results and those of others who have recorded from anterior pituitary cells, including the source and preparation of cells. It is also possible that the different methods identify different subpopulations of lactotrophs (e.g., Frawley et al., 1985; Hoeffler et al., 1985; Boockfor et al., 1986; Luque et al., 1986; also see discussion in St. John et al., 1986b) and that the different subpopulations have

different electrophysiological properties. Finally, it cannot yet by ruled out that the methods used for identification have altered the electrophysiological properties of the cells. Additional experiments will be needed to clarify this issue. The ability of lactotrophs as studied here to generate action potentials in response to a sufficient stimulus and the spontaneous fluctuations of membrane potential observed in some of these cells may be integral to their cellular mechanisms for secretion. Action potentials would cause an influx of calcium ions, and this in turn would be expected to lead to increased secretion of hormone. Subthreshold fluctuations of membrane potential might also cause an influx of calcium. A constant train of such fluctuations or action potentials might lead to a steady-state concentration of intracellular calcium that was higher than in resting cells and, thus, a higher basal rate of hormone secretion. Exogenous hypothalamic hormones might act to alter secretion from lactotrophs by modifying the spontaneous fluctuations and/or the evoked action potentials, as well as by mobilizing intracellular sources of Ca2+ ions and second-messenger circuits (see Gershengorn, 1985). These possibilities will be addressed in future experiments. Electrical responses to TRH have not yet been detected among the PRL-containing cells studied by the present method, which labels only about half of all PRL cells (St. John et al., 1986b). Boockfor et al. (1986) have provided evidence that many of the PRL-secreting cells identified by reverse hemolytic plaque assay show little or no PRL-secretory response to TRH. It will be interesting to determine whether the cells identified by the present method are the TRH-insensitive cells recognized by Boockfor et al. (1986; also see discussion in Gershengorn, 1985). In summary, we have developed a strategy to isolate and culture lactotrophs from the rat pituitary using indirect immunofluorescence and fluorescence-activated cell sorting. Cells cultured for several days exhibit a spectrum of excitable membrane properties, but do not respond to TRH. Using this protocol, we should be able to examine in detail the excitability of lactotrophs and then relate PRL secretion to changes in membrane properties.

210

Acknowledgement We would like to thank Christopher R. Lingle for making his results available to us before publication. References Biales, B., Dichter, M.A. and Tischler, A. (1977) Nature 267, 172-174. Boockfor, F.R., Hoeffler, J.P. and Frawley, L.S. (1986) Neuroendocrinology 42, 64-70. Douglas, W.W. and Taraskevich, P.S. (1985) In: The Electrophysiology of the Secretory Cell, Eds.: A.M. Poisner and J.M. Trifaro (Elsevier, New York) pp. 65-92. Dufy, B. (1983) In: Current Methods in Cellular Neurobiology, Vol. 4, Eds.: J.L. Barker and J.F. McKelvy (Wiley, New York) pp. 49-79. Dufy, B., MacDermott, A. and Barker, J.L. (1986) Biochem. Biophys. Res. Commun. 137, 388-396. Frawley, LX, Boockfor, F.R. and Hoeffler, J.P. (1985) Endocrinology 116, 134-737. Gershengom, M.C. (1985) Recent Prog. Horm. Res. 41, 607-653. Hoeffler, J.P., Boockfor, F.R. and Frawley, L.S. (1985) Endocrinology 117, 187-195.

Hoeffler, J.P., Boockfor, F.R. and Frawley, L.S. (1985) Endocrinology 117, 187-195. Lingle, C.J., Sombati, S. and Freeman, M.E. (1987) J. Neurosci., in press. Luque, E.H, Munoz de Toro, M., Smith, P.F. and Neill, J.D. (1986) Endocrinology 118, 2120-2124. Mason, W.T. and Waring, D.W. (1986) Neuroendocrinology 43, 205-219. Ozawa, S. and Saito, T. (1980) Experientia 36, 1235-1236. Smith, T.G., Barker, J.L., Smith, B.M. and Colburn, T.R. (1981) In: Excitable Cells in Tissue Culture, Eds.: P.G. Nelson and M. Lieberman (Plenum, New York) pp. 111-136. St. John,

P.A., Kell, W.M., Mazzetta, J.S., Lange, G.n snd Barker, J.L. (1986a) J. Neurosci. 6, 1492-1512. St. John, P.A., Dufy-Barbe, L. and Barker, J.L. (19865) Endocrinology 119, 2783-2795. Taraskevich, P.S. and Douglas, W.W. (1977) Proc. Nat]. Acad. Sci. U.S.A. 74, 40644067. Thomer, M.O., Borges, J.L.C., Cronin, M.J., Keefer, D.A., Hellmann, P. Lewis, D., Dabney, L.G. and Quesenberry, P.J. (1982) Endocrinology 110, 1831-1833. Vincent, J.D., Israel, J.M. and Brigant, J.L. (1985) Neurochem. Int. 7. 1007-1016.