Reverse engineering the lordosis behavior circuit.

Reverse engineering takes the facts we know about a device or a process and reasons backwards to infer the principles underlying the structure-function relations. The goal of this review is to apply this approach to a well-studied hormone-controlled behavior, namely the reproductive stance of female rodents, lordosis. We first provide a brief overview on the considerable amount of progress in the analysis of female reproductive behavior. Then, we propose an analysis of the mechanisms of this behavior from a reverse-engineering perspective with the goal of generating novel hypotheses about the properties of the circuitry elements. In particular, the previously proposed neuronal circuit modules, feedback signals, and genomic mechanisms are considered to make predictions in this manner. The lordosis behavior itself appears to proceed ballistically once initiated, but negative and positive hormonal feedback relations are evident in its endocrine controls. Both rapid membrane-initiated and slow genomic hormone effects contribute to the behavior's control. We propose that the value of the reverse-engineering approach is based on its ability to provide testable, mechanistic hypotheses that do not emerge from either traditional evolutionary or simple reductionistic perspectives, and several are proposed in this review. These novel hypotheses may generalize to brain functions beyond female reproductive behavior. In this way, the reverse-engineering perspective can further develop our conceptual frameworks for behavioral and systems neuroscience.

Identification of subtypes of muscarinic receptors that regulate Ca2+ and K+ channel activity in sympathetic neurons.

Many different G protein-coupled receptors modulate the activity of Ca2+ and K+ channels in a variety of neuronal types. There are five known subtypes (M1-M5) of muscarinic acetylcholine receptors. Knockout mice lacking the M1, M2, or M4 subtypes are studied to determine which receptors mediate modulation of voltage-gated Ca2+ channels in mouse sympathetic neurons. In these cells, muscarinic agonists modulate N- and L-type Ca2+ channels and the M-type K+ channel through two distinct, G-protein mediated pathways. The fast and voltage-dependent pathway is lacking in the M2 receptor knockout mice. The slow and voltage-independent pathway is absent in the M1 receptor knockout mice. Neither pathway is affected in the M4 receptor knockout mice. Muscarinic modulation of the M current is absent in the M1 receptor knockout mice, and can be reconstituted in a heterologous expression system using cloned channels and M1 receptors. Our results using knockout mice are compared with pharmacological data in the rat.

Assignment of muscarinic receptor subtypes mediating G-protein modulation of Ca(2+) channels by using knockout mice.

There are five known subtypes of muscarinic receptors (M(1)-M(5)). We have used knockout mice lacking the M(1), M(2), or M(4) receptors to determine which subtypes mediate modulation of voltage-gated Ca(2+) channels in mouse sympathetic neurons. Muscarinic agonists modulate N- and L-type Ca(2+) channels in these neurons through two distinct G-protein-mediated mechanisms. One pathway is fast and membrane-delimited and inhibits N- and P/Q-type channels by shifting their activation to more depolarized potentials. The other is slow and voltage-independent and uses a diffusible cytoplasmic messenger to inhibit both Ca(2+) channel types. Using patch-clamp methods on acutely dissociated sympathetic neurons, we isolated each pathway by pharmacological and kinetic means and found that each one is nearly absent in a particular knockout mouse. The fast and voltage-dependent pathway is lacking in the M(2) receptor knockout mice; the slow and voltage-independent pathway is absent from the M(1) receptor knockout mice; and neither pathway is affected in the M(4) receptor knockout mice. The knockout effects are clean and are apparently not accompanied by compensatory changes in other muscarinic receptors.

Disruption of the m1 receptor gene ablates muscarinic receptor-dependent M current regulation and seizure activity in mice.

Muscarinic acetylcholine receptors are members of the G protein-coupled receptor superfamily expressed in neurons, cardiomyocytes, smooth muscle, and a variety of epithelia. Five subtypes of muscarinic acetylcholine receptors have been discovered by molecular cloning, but their pharmacological similarities and frequent colocalization make it difficult to assign functional roles for individual subtypes in specific neuronal responses. We have used gene targeting by homologous recombination in embryonic stem cells to produce mice lacking the m1 receptor. These mice show no obvious behavioral or histological defects, and the m2, m3, and m4 receptors continue to be expressed in brain with no evidence of compensatory induction. However, the robust suppression of the M-current potassium channel activity evoked by muscarinic agonists in sympathetic ganglion neurons is completely lost in m1 mutant mice. In addition, both homozygous and heterozygous mutant mice are highly resistant to the seizures produced by systemic administration of the muscarinic agonist pilocarpine. Thus, the m1 receptor subtype mediates M current modulation in sympathetic neurons and induction of seizure activity in the pilocarpine model of epilepsy.

Estrogen modulation of two subpopulations of ß-endorphin neurons in ovariectomized guinea pigs distinguished by peripherally injected fluorogold.

ß-endorphin released by neurons in the arcuate nucleus affects the output of several neuroendocrine systems and estrogen levels modulate the production and secretion of ß-endorphin. We used intraperitoneal injection of fluorogold to retrogradely label the cell bodies of neurons that project outside the blood-brain-barrier in conjunction with immunohistochemistry for ß-endorphin to dual label the subpopulation of ß-endorphin neurons that project to the median eminence or other sites of access to the peripheral circulation. We found that some identified ß-endorphin neurons in the arcuate nucleus of ovariectomized guinea pigs sequestered fluorogold. Approximately 7% of ß-endorphin-containing cells co-localized with fluorogold. The effect of estrogen on the number of identified ß-endorphin cells was examined. A single estradiol benzoate injection to ovariectomized guinea pigs 24 h prior to sacrifice dramatically decreased the total number of ß-endorphin cells identified in the rostral, medial and the caudal portions of the arcuate nucleus. Also, a significantly smaller percentage of fluorogold-filled cells was found to contain ß-endorphin immunoreactivity in the estrogen-treated group. These data suggest that a subpopulation of ß-endorphin neurons has access to the peripheral circulation and may alter the output of neurosecretory terminals at the level of the median eminence. Furthermore, estrogen affects this subpopulation and the general population of ß-endorphin neurons in the arcuate nucleus in a similar manner.

Effects of estrogen on the number of neurons expressing beta-endorphin in the medial basal hypothalamus of the female guinea pig.

The distribution pattern of immunoreactive beta-endorphin neurons was studied in female guinea pigs that were ovariectomized, and one week later were injected with 25 micrograms estradiol benzoate or oil. The animals (5 from each group) were perfused after 24 hours with 4% paraformaldehyde. The locations of beta-endorphin cells and fibers were determined using avidin-biotin immunohistochemistry on free-floating vibratome sections. beta-endorphin-immunoreactive fibers were distributed widely throughout specific regions of the rostral forebrain, similar to what has been described in other species. beta-endorphin cell bodies were found in the arcuate nucleus and in adjacent ventrolateral areas throughout the rostrocaudal extent of the basal hypothalamus. Cells immunoreactive to beta-endorphin were also present in the caudal part of the ventromedial nucleus of the hypothalamus. The number of beta-endorphin neurons was quantified in anatomically matched sections through the rostral, medial and caudal basal hypothalamus of estradiol benzoate- and oil-treated guinea pigs. Analysis of variance revealed that the number of immunoreactive beta-endorphin cells was significantly increased in all regions of the basal hypothalamus of estrogen-treated guinea pigs as compared to vehicle-treated animals (P < 0.01). These data indicate that in the guinea pig, the number of neurons expressing beta-endorphin is increased in the arcuate nucleus 24 hours after estrogen treatment.

Estrogen suppresses mu-opioid- and GABAB-mediated hyperpolarization of hypothalamic arcuate neurons.

The effects of estrogen on the response of hypothalamic arcuate neurons to mu-opioid and GABAB agonists were investigated. Intracellular recordings were made from arcuate neurons in slices prepared from ovariectomized guinea pigs that were pretreated with estrogen or vehicle. Estrogen shifted the dose-response curve to the mu-opioid agonist DAMGO (Tyr-D-Ala-Gly-MePhe-Gly-ol) by 3.4-fold; the EC50 for DAMGO was 240 +/- 25 nM in estrogen-treated females versus 70 +/- 12 nM in the controls. The maximal hyperpolarization induced by DAMGO was equivalent in neurons from both groups. The Ke for the naloxone antagonism of the DAMGO response was similar in both groups, which would indicate that the affinity of the mu-receptor was unchanged. To explore where in the receptor/G-protein/K+ channel cascade estrogen may be acting to attenuate the mu-opioid-mediated hyperpolarization, the response to the GABAB agonist baclofen was also tested. Estrogen treatment also shifted the dose-response curve for the baclofen-induced hyperpolarization by 3.3-fold without altering the maximum hyperpolarization; the EC50 shifted from 11.0 +/- 4.0 microM to 36.0 +/- 5.0 microM. All of the neurons were identified after linking the intracellular biocytin with streptavidin-FITC, and a subpopulation of cells in both groups were immunoreactive for beta-endorphin. We conclude that estrogen decreases the functional coupling of the mu-opioid and GABAB receptors to the inwardly rectifying K+ channel possibly through an action on the G-protein.

Neurons in the rat arcuate nucleus are hyperpolarized by GABAB and mu-opioid receptor agonists: evidence for convergence at a ligand-gated potassium conductance.

Both gamma-aminobutyric acid (GABA) and the endogenous opioid peptides have pervasive effects on neuroendocrine function. This study examined the effects of selective activation of GABAB and/or mu-opioid receptors on neurons of the arcuate nucelus (ARC) of the rat hypothalamus using intracellular recording of cells in a hypothalamic slice. Some recorded neurons were filled with biocytin allowing subsequent identification and immunocytochemical evaluation for the presence of beta-endorphin. ARC neurons exhibited a broad array of active and passive conductances. Tyr-D-Ala-Gly-MePhe-Gly-ol (DAGOL), a mu-opioid receptor agonist, inhibited spontaneous firing, hyperpolarized 68% of ARC cells in a dose-dependent manner and increased cell conductance. Baclofen, a GABAB receptor agonist, hyperpolarized all cells tested. The reversal potentials for both the DAGOL- and baclofen-induced currents were near that of a potassium conductance. Maximal activation by either of the agonists blocked the effects of the other agonist. Identified beta-endorphin cells were inhibited by both DAGOL and baclofen. The results of these in vitro studies suggest that GABAB and mu-opioid receptors are coupled to the same set of potassium channels and that these channels directly and powerfully inhibit most ARC cells, including beta-endorphin neurons. We propose that convergence of inhibitory influences at the ligand-gated potassium conductance described here may be an important site of interaction for opioidergic, GABAergic and other putative neurotransmitter systems in the control of neuroendocrine circuits by the ARC.

Membrane properties and response to opioids of identified dopamine neurons in the guinea pig hypothalamus.

The electrophysiological properties and opioid responsiveness of the dopamine-containing neurons in the arcuate nucleus of the guinea pig hypothalamus were examined. Dopamine-containing neurons, identified immunocytochemically by the presence of tyrosine hydroxylase, had a mean length-to-width profile of 14.9 +/- 4.4 x 11.5 +/- 3.1 microns (N = 14). The Na+ action potential of these neurons was of short duration, and induction of repetitive firing (20-50 Hz) caused an afterhyperpolarization of 6-9 mV in amplitude, with a decay half-time of approximately 1.5 sec. Dopamine-containing cells exhibited a low threshold spike, which induced 1-4 Na+ action potentials. This potential had a threshold close to -65 mV, could not be induced without prior hyperpolarization and was not sensitive to TTX. Dopamine-containing neurons also exhibited a time- and voltage-dependent inward current at potentials negative to -70 mV, and Cs+ blocked this conductance. The mu-opioid agonist Tyr-D-Ala-Gly-mePhe-Gly-ol hyperpolarized (14 +/- 3 mV) dopamine neurons via induction of an outward current (93 +/- 44 pA near the resting membrane potential) which had a reversal potential similar to that expected for a selective potassium conductance. TTX (1 microM) did not block the opioid effects. These results show that dopamine neurons of the arcuate nucleus differ in their intrinsic conductances and their responsiveness to opioids from other CNS dopaminergic neurons. Furthermore, opioid activation of a potassium conductance resulted in a direct hyperpolarization of dopamine neurons of the arcuate nucleus, and we suggest that this mechanism may underlie the effects of opioids on dopamine-mediated prolactin release.

A method for immunocytochemical identification of biocytin-labeled neurons following intracellular recording.

Because of the large number of cell phenotypes in the nervous system, it has been difficult to characterize each as to specific electrophysiological properties. We have developed a technique that allows the identification of central and peripheral nervous system neurons following intracellular recording. We use electrodes that contain 2% biocytin to do current- and voltage-clamp recordings; the recorded neurons are revealed with streptavidin-fluorescein isothiocyanate labeling and identified through immunohistochemical staining for specific antigens. Presently, we report on the use of this technique to identify four cell types--dopamine, beta-endorphin, vasopressin and oxytocin--in the hypothalamus of the mammal. This technique should have widespread applicability for electrophysiologists.

Opioids hyperpolarize beta-endorphin neurons via mu-receptor activation of a potassium conductance.

Intracellular recordings were made from hypothalamic arcuate (ARC) neurons with biocytin-filled electrodes under current- and voltage-clamp in slices prepared from ovariectomized guinea pigs which were pretreated with estradiol. Forty-three neurons were identified after linking the intracellular biocytin with streptavidin-FITC and subsequently were examined for beta-endorphin immunoreactivity. Ten of these neurons were immunoreactive for beta-endorphin. beta-Endorphin neurons displayed the following passive membrane properties: RMP:-56 +/- 2 mV; Rin: 439 +/- 66 M omega; tau: 17.5 +/- 2.4 ms; and often fired spontaneously (5.9 +/- 2.2 Hz). These membrane characteristics were not different from identified neurons in the ARC that were not immunoreactive for beta-endorphin. beta-Endorphin neurons exhibited instantaneous inward rectification and time-dependent rectification. The mu-opioid agonist Tyr-D-Ala-Gly-MePhe-Gly-ol (DAGO) decreased spontaneous firing, induced membrane hyperpolarization (12 +/- 2 mV; range 6-22 mV) and decreased the Rin (38 +/- 4%) of the beta-endorphin neurons. These effects of DAGO were blocked by the opioid antagonist naloxone (1 microM) and were not blocked by 1 microM TTX. DAGO-responsive cells were unaffected by either kappa- or delta-receptor opioid agonists. These results indicate that mu-receptors may be autoreceptors on ARC beta-endorphin neurons and that activation of opioid mu-receptors hyperpolarizes beta-endorphin neurons via an increase in K+ conductance. Therefore, opioid peptides may modulate opioid tone through an 'ultra-short loop' feedback control mechanism.

Opioids act at mu-receptors to hyperpolarize arcuate neurons via an inwardly rectifying potassium conductance.

Intracellular recordings were made from 48 hypothalamic arcuate (ARC) neurons under current- and voltage-clamp in slices prepared from female guinea pigs which had been ovariectomized and pretreated with estradiol. Twenty ARC neurons were silent (RMP: -62 +/- 2 mV) and 28 cells were spontaneously active (7.3 +/- 1.1 Hz; threshold -57 +/- 1 mV). The input resistance (Rin), determined in the potential range between -60 and -80 mV, was 358 +/- 30 M omega (n = 38) and ARC neurons showed inward rectification at potentials negative to the equilibrium potential for potassium. The selective mu-opioid agonist Tyr-D-Ala-Gly-MePhe-Gly-ol (DAGO) was applied by pressure pipette application at concentrations of 10 or 20 microM. DAGO decreased spontaneous firing and it hyperpolarized 26 of 31 neurons (9.6 +/- 0.8 mV; range 3-21 mV). Concomitant with the hyperpolarization, DAGO caused a decrease in Rin of 32 +/- 3, and the reversal potential, measured from current-voltage plots, was -94 +/- 2 mV. These effects were mimicked by bath concentrations of 0.5-1.0 microM DAGO. In voltage clamp, DAGO caused an outward current to flow at -60 mV (range 50-185 pA, n = 6). This current reversed at -92 +/- 2 mV (n = 6) and exhibited inward rectification. An additional 6 ARC neurons were tested with DAGO in varying extracellular concentrations of K+ (2.5, 5 and 10 mM) and the reversal potential for the effect of DAGO shifted by 58 mV per decade change in extracellular K+ concentration. DAGO decreased spontaneous postsynaptic potentials in some cells, but TTX (1 microM) had no effect on the ability of DAGO to hyperpolarize the membrane. The hyperpolarization and decrease in Rin induced by DAGO were blocked by the opioid antagonist naloxone (100 nM-1 microM). DAGO responsive cells were unaffected by a kappa-opioid agonist (trans-(+/-)-3,4-dichloro-N-methyl-N-[2-(1- pyrrolidinyl)cyclohexyl]benzeneacetamide methanesulphonate; U50,488H), however, 2 of 5 cells also were hyperpolarized by a selective delta-receptor opioid agonist (Tyr-D-Pen-Gly-Phe-D-Pen; DPDPE). The effects of DPDPE, but not DAGO, were blocked by a delta-antagonist (ICI 174,864; 1 microM). The present results indicate that activation of ARC mu-receptors leads to an increase in an inwardly rectifying potassium conductance and a subsequent hyperpolarization of most ARC neurons. We suggest that this mu-receptor-induced hyperpolarization of ARC neurons may underlie the opioid inhibition of reproductive events in the mammal.

Opioid inhibition of spontaneously active neurons of the rat arcuate nucleus in vitro.

The effects of opioid agonists were determined on single-unit activity recorded from the arcuate nucleus (ARC) in perfused, coronal slices of hypothalamus taken from proestrous rats. The selective, mu-receptor agonist Tyr-D-Ala-Gly-MePhe-Gly-ol enkephalin (DAGO) produced a concentration-dependent decrease in the firing rate of 70-78% of the units tested. The concentration of DAGO that induced maximal inhibition of firing was approximately 0.5 microM. This inhibition of firing frequency occurred irrespective of cell location, firing pattern or baseline firing frequency. The effect of DAGO was antagonized by the opioid antagonist naloxone (0.1 microM). The selective, kappa-receptor agonist, trans-(+)-3,4 dichloro-N-methyl-[2-(1-pyrrolidinyl) cyclohexyl] benzeneacetamide methane sulfonate (U50,488H) did not decrease the firing rate in cells which did respond to DAGO. Blockade of synaptic activity decreased the level of spontaneous activity but did not prevent the inhibitory action of DAGO. These data support the hypothesis that opioids, through activation of mu-receptors, inhibit neuronal activity in the arcuate nucleus. Furthermore, the opioid inhibition occurs, in part, via a direct postsynaptic action.

Pulsatile infusion of luteinizing hormone-releasing hormone induces precocious puberty (vaginal opening and first ovulation) in the immature female guinea pig.

Previously we have hypothesized that an increase in luteinizing hormone-releasing hormone (LHRH) due to hypothalamic maturation is the key factor controlling the onset of puberty. This led to the working hypothesis that precocious puberty would be induced if LHRH is administered with an appropriate protocol. Thus, effects of pulsatile infusion of LHRH on the onset of first vaginal opening and first ovulation in immature female guinea pigs were studied. Luteinizing hormone-releasing hormone in hourly pulses of either 5 ng or 50 ng was infused through a chronically implanted jugular catheter for 9-29 days starting at 20 days of age. For the control experiment saline was infused in a similar manner. Infusion of 5 ng LHRH/h resulted in significantly earlier (P less than 0.001) ages at first vaginal opening (24.7 +/- 0.9 days) and at first ovulation (28.8 +/- 0.9 days) compared to saline controls (first vaginal opening 53.3 +/- 6.8 days; first ovulation 55.2 +/- 6.5 days). Infusion with a 10-fold higher LHRH dose (50 ng/h) also advanced the age at first vaginal opening (25.3 +/- 0.7 days), but precocious ovulation was no longer induced (53.7 +/- 5.3 days). Interestingly, LHRH infusion with the high dose resulted in a prolonged period of vaginal opening and cornification without ovulation. These results indicate that 1) pulsatile infusion of a small amount of LHRH with a constant frequency induces precocious puberty in a laboratory rodent, and 2) infusion of LHRH with a dose higher than the effective dose for the induction of early puberty results in a persistent estrous anovulatory syndrome. Therefore, the present study not only supports our hypothesis that an increase in endogenous LHRH release is responsible for the onset of puberty, but also further suggests that excessive release of LHRH or abnormal patterns of LHRH release may be involved in the etiology of the anovulatory persistent estrus syndrome.

Posterior hypothalamic lesions advance the onset of puberty in the female rhesus monkey.

The effects of experimental lesions in the posterior hypothalamus and the anterior hypothalamus on menarche and first ovulation were examined in nonhuman primates. With the aid of x-ray ventriculography, bilateral lesions were made by passing a radiofrequency current through a thermister electrode in the posterior hypothalamus (n = 7) or the anterior hypothalamus (n = 6) of female rhesus monkeys at 18 months of age. Four animals that received sham lesions as well as four normal females of a similar age served as controls. All animals were caged individually and examined daily for vaginal bleeding and sex skin color change. Developmental changes in gonadotropins, ovarian steroids, body weight, and nipple size were monitored throughout the experiments. The time of first ovulation was determined by laparoscopic observation of the newly formed corpus luteum and by the level of circulating progesterone. Histological examination confirmed that the bilateral lesions in the hypothalamus were approximately 2-3 mm in diameter and overlapped midline. Primary sites of posterior hypothalamic lesions included the premamillary area and the posterior nucleus, while the infundibular nucleus and the median eminence were entirely spared. The posterior lesions encroached upon the mamillary nuclei caudally in most cases and upon the ventromedial nucleus rostrally in some cases. Primary sites of anterior hypothalamic lesions included the medial preoptic area, the periventricular preoptic nucleus, and the anterior hypothalamic nucleus. Partial lesions of the diagonal bundle of Broca, the medial preoptic nucleus, and the paraventricular nucleus were also detected. Posterior hypothalamic lesions advanced the ages at menarche (22.2 +/- 1.3 months; P less than 0.001) and first ovulation (40.7 +/- 2.7 months; P less than 0.05) compared to those of control animals (menarche, 30.3 +/- 3.1; first ovulation, 51.2 +/- 3.3 months). The body weight at menarche of these lesioned animals (2.62 +/- 0.11 kg) was smaller (P less than 0.05) than that of controls (3.14 +/- 0.20 kg), but the body weight at first ovulation of lesioned animals (4.36 +/- 0.28 kg) was not different from that of controls (4.57 +/- 0.13 kg). Hormonal and physical changes during maturation, i.e. an increase in circulating estradiol and growth in nipple size before menarche and first ovulation, occurred earlier in the lesioned animals and the growth spurt before first ovulation not only began earlier but also attained mature levels several months earlier than that in control animals.(ABSTRACT TRUNCATED AT 400 WORDS)