The Remarkable Me Effects Of Pregs

There were significant overall ME effects for intrahippocampal injection of PREGS, DHEAS, and corticosterone (Fig. 1; 8). By far the most potent action was exerted by PREGS, quantities between 1016 g or 2.4 x 1018 moles/mouse and 1012 g or 2.4 x 1014 moles/mouse, giving mean trials to criterion which were significantly lower than the vehicle controls (p < 0.01). Multiplying moles of PREGS/mouse by Avogadro's number, 6.02 x 1023 molecules/mole, it was calculated that 1.45 x 106 molecules of PREGS were sufficient to cause significant ME on intrahippocampal injection. DHEAS and corticosterone showed significant ME only at much higher concentrations and over smaller concentration ranges than PREGS. Clearly, PREGS was more potent than the other steroids tested and its effects extended over a greater range of concentrations.

When PREGS was tested in the amygdala, mammillary bodies, septum, and caudate nucleus (Fig. 1), there were significant ME effects for amygdala, septum, and mammil-lary bodies, but not for the caudate. Two-way analysis of variance (ANOVA) showed that the dose-response curves for mammillary bodies and septal injection did not differ significantly from each other, but that both were significantly different from the curves obtained for hippocampus and amygdala.

The retention test scores on intra-amygdalar injection of extremely dilute solutions of PREGS showed that between 15 and 145 molecules of PREGS produced ME. The closely similar results obtained with solutions prepared in two separate laboratories by different

Fig. 1. Effects of post-training intraparenchymal injection of steroids on retention of FAAT in male mice. The mean and SEM for trials to criterion are shown for 15 animals at each dose indicated. Means differing from vehicle alone at p < 0.01 (*) or at p < 0.05 (**) based on Dunnett's t-tests are indicated. The shaded areas are the mean ± SEM for trials to criterion for vehicle controls. Mice were anesthetized with methoxyflurane, placed in a stereotactic instrument, and a hole was drilled through the skull over each injection site after deflecting the scalp. Mice were trained 48 h after surgery. Four training trials were given using an intertrial interval of 30 s, a warning buzzer of 55 DB, and foot-shock intensity of 0.30 mA. Immediately after training, the test solution containing vehicle or PREGS was injected over a period of 60 s into the target structure. One wk later, T-maze training was resumed until each mouse made five avoidance responses in 6 consecutive training trials. Retention was measured by the number of trials required for each mouse to meet this criterion; the fewer trials required, the greater the retention of learning. The mice in this study were weakly trained so that ME effects on retention of T-maze FAAT could be detected readily. The site of injection then was confirmed histologically using a mouse brain stereotaxic atlas. (See ref. 8 for further details.)

Fig. 1. Effects of post-training intraparenchymal injection of steroids on retention of FAAT in male mice. The mean and SEM for trials to criterion are shown for 15 animals at each dose indicated. Means differing from vehicle alone at p < 0.01 (*) or at p < 0.05 (**) based on Dunnett's t-tests are indicated. The shaded areas are the mean ± SEM for trials to criterion for vehicle controls. Mice were anesthetized with methoxyflurane, placed in a stereotactic instrument, and a hole was drilled through the skull over each injection site after deflecting the scalp. Mice were trained 48 h after surgery. Four training trials were given using an intertrial interval of 30 s, a warning buzzer of 55 DB, and foot-shock intensity of 0.30 mA. Immediately after training, the test solution containing vehicle or PREGS was injected over a period of 60 s into the target structure. One wk later, T-maze training was resumed until each mouse made five avoidance responses in 6 consecutive training trials. Retention was measured by the number of trials required for each mouse to meet this criterion; the fewer trials required, the greater the retention of learning. The mice in this study were weakly trained so that ME effects on retention of T-maze FAAT could be detected readily. The site of injection then was confirmed histologically using a mouse brain stereotaxic atlas. (See ref. 8 for further details.)

dilution procedures and tested under blinded conditions gave confidence in the validity of the results.

CAN THE MEMORIAL PATH OF PREGS IN THE NEUROLABYRINTH BE TRACED?

The finding that so few molecules of PREGS can enhance post-training memory processes when injected into the amygdala in mice established PREGS as the most potent memory enhancer yet reported. In terms of sensitivity, ME caused by PREGS resembles a pheromone-mediated process. For example, the 70 kDa sex-inducing glycoprotein pheromone of the multicellular green flagellate Volvox carteri, synthesized and released by sperm cells, exerts its biological response fully at a concentration of 6 x 10-17 M (36 molecules/mL) (11). In both of the aforementioned instances, there must occur great amplification of the effects of a very small number of molecules.

In the case of Volvox carteri, it was proposed that externally applied pheromone accumulates in the extracellular matrix, whence it enters surface-lying somatic cells. After several hours, a plethora of a protein is released into the extracellular matrix from which is cleaved a greatly amplified quantity of homologue of the original pheromone. The latter then exerts sex-inducing activity in the small number of internally-contained deep-lying asexual reproductive cells. Although the system as a whole requires external contact with only relatively few molecules of pheromone to achieve sex induction, concentrations several orders of magnitude higher are needed for sex induction by direct application of the pheromone to the reproductive cells. Can a mechanism be envisioned by which PREGS similarly might exert an effect in the amygdala, whereby post-training infusion of a small number of molecules of PREGS results in subsequent release from contiguous oligodendrocytes and astrocytes onto neurons of the activated circuits of much greater quantities of PREGS than originally were injected?

Even in minimally active states of mammalian nervous systems, neurons, non-neu-ronal cells (e.g., glial, ependymal, and endothelial cells), and the extracellular matrix participate in many mutually shaping interactions that range from physical forces they exert on each other (12) to exchanges, often through gap junctions, of varieties of trophic and/or inhibitory substances. Much data support the notion that throughout the nervous system there exist gap junctionally coupled astrocytic syncytial networks, enabling rapid passage among the cells of many molecules of up to approximately 1 kDa (e.g., 13). These interactions are greatly enhanced by nerve activity, during which and for a period after which, members of the system derive from each other and from blood-borne sources various substances required to achieve recovery from preceding activity and to undergo appropriate plastic changes (14). Such interactions among components of a healthy biological system are best described by paraphrase of an old dictum for a healthy social system (15): from each according to its capacities, to each according to its requirements, and together for the benefit of the system as a whole.

Mice used in the ME experiments were weakly trained. Conditions of training are such that, at most, within 2-3 d after training, the vehicle-treated controls show the same numbers of trials to criterion as those observed at initial training of naive animals. Posttraining administration of ME substances to such mice results in dose-related decreases by comparison with controls in numbers of trials to criterion at 7 d after initial training. Such substances generally have been found to exert overall physiologically excitatory or depolarizing effects in the systems under study. When during training the strengths of stimuli and numbers of trials are increased sufficiently above those given to the weakly trained animals, maximal retention of learning can be achieved for at least 7 d and sometimes even for many months thereafter. ME is not noted in such strongly trained mice upon post-training administration of substances that are effective in weakly trained animals.

In brief, neural excitation results in increased permeability to cations, thus decreasing the potential across membranes (depolarization). Inward Na+ current usually is responsible for most of the observed depolarization, but influx of extracellular Ca2+ also takes place by entry through voltage-gated Ca2+ channels. When there is Ca2+ influx into a nerve terminal, neurotransmitter release is facilitated. Increase in cytosolic Ca2+ also occurs by release from intracellular Ca2+ binding entities. The opening of K+ channels is activated by the changes in membrane potential. Resultant outward K+ currents then serve to repolarize the cells and, in some instances, to produce hyperpolarization before the Ca2+ balance is restored via the action of Ca2+-Mg2+ ATPase, mitochondrial uptake, and by rebinding of Ca2+; and the K+ channels are closed. The action of Na+-K+ ATPase restores the monocation balance.

The time-sequence coordination of the latter events is such that there are pulsatile localized increases in pools of cytosolic free Ca2+, one or more of which reflect accurately the amounts and durations of depolarization to which membranes are subjected. The patterns of such Ca2+ transients may have great informational content, their extent and frequency possibly encoding experiential information (16). The free Ca2+ either directly or via its interaction with Ca2+ binding proteins, of which calmodulin, parvalbumin, troponin C, S-100 and calbindin are examples, releases cascades of many intracellular processes including activation of genes, some of which processes continue after free Ca2+ is reduced to the resting level.

Concurrently with the aforementioned, chemical modifications (e.g., phosphorylation or dephosphorylation) of the Ca2+ binding proteins and allosteric effects exerted by noncovalent binding of substances to them alter their affinities for Ca2+ (17). Relatedly, changes occur in activities of enzymes involved in metabolism of the cyclic nucleotides, cAMP and cGMP, changing their turnovers, relative amounts, and intracellular distributions. There occur changes in activities of protein kinases and phosphatases that act on specific substrate proteins to establish new cell balances. Phosphorylation controls the activity of many enzymes and the conformational states of nonenzymatic proteins. The control of the degrees of protein phosphorylation is the "principal mechanism by which all cell functions in eukaryotic cells are controlled by extracellular signals. Processes as diverse as metabolism, membrane transport and permeability, secretion, contractility, the transcription and translation of genes, cell division and fertilization, neurotransmission, and even memory are all regulated by this versatile posttranslational modification" (18).

Dose-related ME produced by post-training treatment in weakly trained mice causes graded enhancement and/or prolongation of effects of the learning experience that outlast early brief events such as depolarization and the consequential ionic displacements. For example, when mice were trained under conditions giving poor retention in vehicle-injected controls, administration of DHEAS at 2, 30, or 60 min gave significant improvements (p < 0.01, p < 0.01, and p < 0.05, respectively). When injected at 90 and 120 min, DHEAS had no effect (19). Perseverative neuronal activity consisting of reverberation of impulses in neuronal circuits that outlasts the stimulating events is not what makes memory-related processes amenable to enhancement long after the first neural events during training have ceased (20).

Calmodulin is the most ubiquitously occurring Ca2+-binding protein and the most extensively studied. Among the changes occurring when free cytosolic Ca2+ becomes elevated following stimulation is an increase in Ca2+-calmodulin, which activates many cascade-initiating targets in a coordinated manner so that interlocking cascades throughout the entire activated system result rapidly in effective restorative processes and in adaptive plastic changes. Many substances, including steroids, are known to interact allosterically with calmodulin to regulate activation of calmodulin-dependent targets (17,21,22).

Among the targets activated by Ca2+-calmodulin is the nitric oxide (NO)-producing enzyme, NO synthase. NO, a gaseous free radical formed from the guanidino group of arginine, diffuses freely and rapidly among the matrix, vascular, glial, and neuronal components of the activated neural regions. Notably, NO activates soluble guanylate cyclase and increases rates of turnover and contents of cGMP. NO mediates the stimulation of cGMP formation by glutamate and other NMDA receptor agonists and by a variety of other excitatory influences that act on brain and other tissues. Indeed, the formation of NO from L-arginine is a widespread transduction mechanism for the stimulation of the soluble guanylate cyclase (23-25). Substances with overall excitatory effects, among which are the known memory enhancers, have cGMP as an intracellular messenger for their action (26). In vivo evidence has been adduced supporting the NO-cGMP relationship in rat cerebellum, increases in NO resulting in increases in extracellular content of cGMP (27). By action on presynaptic terminals, cGMP can produce long-term enhancement of transmitter release (28) and, therefore, probably long-term increases in efficacy of information transmittal in activated nerve circuits.

In many instances, it was found that in the effective dose ranges, progressive increases of ME substances first increased responses to a maximum, beyond which decreasing responses were observed until a level was reached at which no significant effects were seen over the controls (29). The decrements observed at higher doses of ME substances possibly may have been attributable to incoordinations induced in management of intracellular free Ca2+ and/or calmodulin, leading to chaotogenic effects on release of reaction cascades (30).

The inverted U-shaped dose-response curve previously described usually covers a 2-5-fold dose range. However, in the case of injection of PREGS into limbic system structures, the curves extended over the enormously greater dose ranges of 4 to 6 orders of magnitude (Fig. 1). The latter differentiates the PREGS effects from those of the usual ME substances and suggests that the dose-response curve for PREGS may be a composite of at least two effects. At the upper dose range, PREGS may act like any other excitation-enhancing substance, and at the lower range, in a uniquely special, possibly pheromone-like sensitive fashion. Although to date there has not been identified a specific cellular binding protein with sufficiently high affinity for PREGS to qualify as a saturable high-affinity receptor for PREGS, it has not been ruled out that such an entity exists. The shape of the left-hand portion of the dose-response curve for the amygdala (in Fig. 1) is not inconsistent with such a possibility.

Let us posit that in a manner somewhat analogous to that discussed previously for the flagellate Volvox carteri, once binding with an appropriate very high affinity entity occurs, the PREGS administered may participate in activation of a cascade that gives rise to a massive synthesis and/or mobilization of PREGS and its subsequent release. The higher concentrations of PREGS released then could act via much lower affinity, but synergistic, mechanisms to achieve amplification of the effects of the small number of molecules of PREGS initially acting on a presumptive high affinity receptor in a few cells. What might be a mechanism by which such events could occur?

It may be that in weakly trained animals, the post-training "window of opportunity" for ME substances to exert their action is the time during which the effects of initially suboptimally increased levels of Ca2+-calmodulin can be driven to higher levels by increasing cytosolic free Ca2+; or activities of NO synthase and/or guanyl cyclase may be increased directly. In the latter regard, it is particularly interesting that PREGS

not only is a negative modulator of the GABA receptor complex and a positive modulator of the NMDA receptor complex, thereby raising excitability of the system as a whole, but that PREGS also greatly amplifies the increase in cytosolic free Ca2+ that occurs during stimulation of NMDA receptors (31-35). PREGS also may be a direct activator of soluble guanylate cyclase (36). Thus, PREGS could exert a synergistic amplification of effects of excitatory neural transmission at much lower concentrations than would be expected from action of a given amount of PREGS on any one of the aforementioned systems alone, molar potencies in in vitro measurements made individually on these systems seeming too low to support an important in vivo role for PREGS. Enhancement at multiple sites of the pathway leading from neural excitation to increased metabolism of cGMP might help explain the powerful effects of PREGS on retention of FAAT upon intracerebroventricular or intrahippocampal injection. However, the astonishing results upon intra-amygdalar injection forced a search for an additional amplificatory mechanism by which a minute amount of externally supplied PREGS might result in release of PREGS readily made from preexisting stores of PREG or PREGS, or by synthesis of the latter from cholesterol, cholesterol sulfate, or other precursors.

Biosynthetic pathways for PREG, PREGS, DHEA, and DHEAS in brain may differ from those in extracerebral tissues. The latter substances, found in the brain at higher levels than in blood, largely are localized in astrocytes and oligodendrocytes, and exist therein in the free form and as sulfates and lipoidal esters (37). As yet uncharacterized bound forms of related substances are present in organic solvent extracts of brain, but not of adrenals or testes, that give rise to additional large amounts of PREG and DHEA by processes that are known to produce ketones from hydroperoxides or peroxides (38,39). To date, it has not been technically possible to determine whether or not PREGS and DHEAS also are increased by such treatment (V. V. K. Prasad and S. Lieberman, personal communication). I conjecture that this is likely to take place.

NO, formed in graded amounts on nerve activity, could react with steroidal peroxides or hydroperoxides in a manner similar to that with which NO reacts with superoxide anion to form peroxynitrite (40), to give substances that are not free radicals themselves, but that have great oxidizing power. This could result in enzymatic and/or nonenzymatic cleavage of side chains of cholesterol and cholesterol sulfate to form PREG, PREGS, DHEA, and DHEAS either successively or independently of each other as in the following: cholesterol (or cholesterol sulfate) ^ PREG (or PREGS) ^ DHEA (or DHEAS) and/or cholesterol (or cholesterol sulfate) ^ PREG (or PREGS) and DHEA (or DHEAS). At least a portion of the PREG and PREGS in brain probably comes from cholesterol and cholesterol sulfate via side-chain cleavage catalyzed by cytochrome P450scc; both the enzyme and its mRNA are found in brain. However, it is doubtful that the ketonic DHEA and DHEAS, indigenously formed in brain, arise from PREG or PREGS by side-chain cleavage by the "classical" action of cytochrome P450c17 (17a-hydroxylase/17,20 lyase), because neither the latter enzyme nor its mRNA have been detected in adult brain (7).

HYPOTHESIS: AN AMPLIFICATORY MECHANISM WITH PHEROMONE-LIKE SENSITIVITY PROPOSED FOR THE ME EFFECT OF PREGS

One variant of several potentially experimentally testable scenarios is outlined below. A mechanism is proposed by which a minute quantity of PREGS,externally introduced into the amygdala post-training to mice that have undergone FAAT, attaches with high affinity to receptors on a few glial cells. This induces release of much additional indigenously generated PREGS along the entire extents of nerve circuits that had been activated during the training experience, but which circuits already were electrophysiologically inactive at the time that PREGS was administered. Such a release of PREGS serves as a "now-print" signal, the PREGS acting on all of the neural elements of the participating circuits via relatively low-affinity processes previously described, metaphorically facilitating the "soldering" together of elements of the newly activated circuits by synaptic plastic changes more firmly than would occur otherwise in weakly trained animals. This is reflected in enhanced retention of learning by comparison with controls receiving vehicle alone after training.

Intra-amygdalar administered PREGS associates with astrocytes and oligodendrocytes by binding to a receptor complex that is highly represented on their membranes and in their cytoplasm. This receptor complex consists of an inducible transcription factor (iTF) bound to an inhibitory protein (BP) that keeps the iTF from entering the nucleus (patterned somewhat after NF-kB; see ref. 41). PREGS has a higher affinity for iTF than does the BP, displacing the latter and forming a PREGS-iTF complex that enters from the membrane into the cytoplasm. PREGS has a much higher affinity for iTF than for cytoplasmic PREGS metabolizing enzymes and, therefore, is stabilized in the PREGS-iTF complex, which enters the nucleus readily. In the nucleus, the PREGS-iTF complex or iTF and PREGS separately, after dissociation of the complex, facilitate the transcription of genes coding for enzymes involved in PREGS biosynthesis and in production of iTF and BP, resulting in increases in the relevant mRNAs in the cytoplasm. In this manner, a few molecules of externally administered PREGS, associating with the surfaces of only one or a few cells, lead to enhanced production of PREGS and of its production and control machinery. A portion of the newly formed plethoric supply of PREGS is liberated onto adjacent neuronal elements, and some is transmitted rapidly via gap junctions to neighboring nonneural cells, wherein the process in the cells originally contacted by the exogenously administered PREGS is repeated, except that the PREGS interacts directly with the iTF-BP complex in the cytoplasm. Thus, rapidly and for a brief period, the entire glial blanket surrounding recently activated neural circuitry liberates more PREGS and enhances achievement of maximal connectivities at all levels of the recently activated circuitry.

The aforementioned, at present admittedly a fanciful hypothesis, can be falsified or supported by experiment. At the outset, one could search for an enzyme or enzymes that form PREGS and DHEAS from cholesterol sulfate and DHEAS from PREGS using as cosubstrate peroxynitrite or oxyradicals. Were positive results obtained, tools currently available would permit a rational search to be made for the posited iTF and iTF-binding proteins. Taking another tack, it should be possible to visualize PREGS itself immuno-cytochemically, at light and electron microscopic levels, in pretraining and posttraining states with and without post-training intracerebral injection of ME-effective amounts of PREGS. The latter would be far below the level of detection by such methods. Elevations in PREGS content produced as suggested might be sufficient to allow PREGS to be noted in brains of PREGS-treated animals in greater amounts than in the vehicle controls, at least for a period of 90 min after training, particularly in the amygdalar region.

Was this article helpful?

0 0

Post a comment