Dueling Enigmas: Neurosteroids and Sigma Receptors in the Limelight

See allHide authors and affiliations

Science's STKE  28 Nov 2000:
Vol. 2000, Issue 60, pp. pe1
DOI: 10.1126/stke.2000.60.pe1


Neurosteroids can be positive or negative regulators of neurotransmitter receptor action, depending on the receptor and the chemical structure of the neurosteroid. This Perspective by Gibbs and Farb is one of two on the subject of neurosteroids. The authors address the possible role of σ receptors in mediating neurosteroid action and describe how the regulation of inhibitory and excitatory ion channels by neurosteroids has implications for the role of these molecules in learning and memory, nociception, and excitotoxicity.

Nearly 60 years have passed since Hans Selye first described the rapid effects of steroids on brain excitability (1), yet the molecular basis for steroid effects on the nervous system remains a mystery. As a relative newcomer, the "σ receptor" was first described in 1976 (2), but its role in the brain has remained unclear. Now, it appears that these two long-standing enigmatic stories may be converging.

Endogenous central nervous system (CNS) steroids, or "neurosteroids," (Fig. 1) exert acute modulatory effects on a variety of neurotransmitter receptors, including glutamate- and γ-aminobutyric acid (GABA) gated ion channels. Neurosteroids can potentiate or inhibit both the type A GABA (GABAA) and N-methyl-D-aspartate (NMDA) glutamate receptors, with some neurosteroids acting as positive modulators and others as negative modulators of receptor function. Evidence indicates that positive and negative modulators act through separate mechanisms at both GABAA and NMDA receptors, rather than sharing a common binding site (3, 4). The modulatory effects of neurosteroids on GABAA and NMDA receptors are rapid and can be observed in excised membrane patches (5-7) and heterologous expression systems (8, 9). Moreover, the site of action of neurosteroid modulators at the GABAA and NMDA receptors is associated with the external surface of the plasma membrane, as neurosteroids are ineffective when added intracellularly (3). These findings suggest that neurosteroids either bind directly to NMDA and GABAA receptors or interact with some ubiquitous factor that is closely associated with the receptor at the plasma membrane. Modulatory effects of neurosteroids have also been reported on neuronal nicotinic acetylcholine receptors (10), glycine receptors (11), nonNMDA glutamate receptors (12), and 5-HT3 serotonin receptors (13), suggesting that the capacity for modulation by neurosteroids may be a near-universal characteristic of ligand-gated ion channels.

Fig. 1.

Biosynthetic pathways of major neurosteroids. Question marks denote conversions that have been hypothesized but not demonstrated.

Perhaps the strongest candidates for a role as endogenous neuromodulators are allopregnanolone and its C-3β isomer pregnanolone, which enhance GABAA receptor function at nanomolar concentrations, making them the most potent steroid modulators of neurotransmission identified to date (14). Pregnanolone is endogenous in human brain but not rat brain and is similar in other respects to allopregnanolone. Pregnanolone and allopregnanolone levels decrease in the cerebral spinal fluid of patients suffering from unipolar major depression, and improvement in symptoms following treatment with selective serotonin reuptake inhibitors is accompanied by an increase in allopregnanolone and pregnanolone levels (15).

Meanwhile, the σ receptor, which was initially confused with opiate and NMDA receptors, was eventually shown to be a distinct entity (16). However, the σ receptor has remained a binding site in search of a function. The initial confusion is easy to understand because σ receptors bind a bewildering variety of pharmacologically important compounds, including phencyclidine (PCP), haloperidol, dextromethorphan, ifenprodil, (+)pentazocine, and progesterone. The cloning of the σ1 receptor (17) has only deepened the mystery. Despite evidence suggesting that σ receptors interact with heterotrimeric GTP-binding proteins (G proteins) (18), the σ1 receptor turns out to bear no resemblance to G protein-coupled receptors, or to any other known receptor, but instead exhibits 66% sequence similarity to none else than a fungal sterol isomerase (17). Pharmacological evidence indicates that there are likely additional types of σ receptors (18, 19), but these have not been cloned.

A number of findings suggest that σ receptors mediate or at least influence the effects of systemically administered neurosteroids. Progesterone is a moderately high-affinity ligand for the σ1 receptor (17), and a number of other steroids have been shown to bind with somewhat lower affinity [see (19) for review]; although, to the best of our knowledge, pregnenolone sulfate (PS) has not been tested. Responses of rat CA3 pyramidal neurons to microiontophoretically applied NMDA in vivo are enhanced by systemic injection of either the neurosteroid dehydroepiandrosterone (DHEA) or σ1 "agonists," such as (+)pentazocine and L-687,384, and this enhancement of the NMDA response is reversed by injection of either progesterone or the σ receptor "antagonist," NE-100 (20). Similarly, NE-100 inhibits the ability of systemically administered PS or DHEAS to relieve scopolamine- (21) or dizocilpine-induced (22) deficits of learning and memory and to ameliorate conditioned fear stress (23).

The steroid connection is particularly intriguing. There is considerable evidence that steroids play a role in the nervous system which goes far beyond the modulation of gene expression mediated by conventional steroid receptors. It is clear that steroids can be produced endogenously within the CNS. Brain steroid levels do not correlate well with those in blood, and pregnenolone, DHEA, and their respective sulfates persist in brain following adrenalectomy and gonadectomy. The enzymes for steroid biosynthesis from cholesterol are present in brain (24), and steroid synthesis from mevalonolactone has been demonstrated in glial cultures (25).

Considerable recent attention has focused on PS, an abundant neurosteroid, which acts as a positive modulator of excitatory NMDA receptors (12), and a negative modulator of inhibitory GABAA receptors (26). This inverse modulation of inhibitory and excitatory receptors makes PS an attractive candidate to play a regulatory role as an endogenous neuromodulator that exacerbates or sensitizes nervous tissue to an excitotoxic insult. PS enhances memory and cognitive performance in a variety of behavioral assays (27, 28), which correlates with its ability to potentiate NMDA receptor activation. Moreover, decreased levels of PS in the hippocampus of aged rats correlate with cognitive impairment, which can be transiently reversed by intraperitoneal or intrahippocampal injection of PS (29). Like PS, DHEA sulfate (DHEAS), another sulfated neurosteroid, enhances NMDA receptor function while inhibiting GABAA receptors (3, 4, 30) and produces a similar behavioral profile, enhancing learning and memory (31) (Fig. 2).

Fig. 2.

Sulfation regulates the direct modulation of ligand-gated ion channels by neurosteroids. This figure illustrates how the direct modulatory effects of pregnenolone and pregnanolone are altered by the addition of a sulfate group at the C-3 position. Positive modulation is indicated by red shading and a thick arrow. Negative modulation is indicated by blue shading and a thin arrow. Unshaded receptors represent the absence of a modulatory effect.

At both NMDA and GABAA receptors, the presence or absence of a sulfate group at the C-3 position appears to be an important determinant of neurosteroid activity. For example, addition of the sulfate group (Fig. 3) converts pregnenolone, which is inactive at both NMDA and GABAA receptors, into PS. Similarly, addition of a sulfate group converts pregnanolone or allopregnanolone from a positive modulator of the GABAA receptor into pregnanolone sulfate or allopregnanolone sulfate, negative modulators of both GABAA and NMDA receptors (4). Because steroid sulfatase and sulfotransferase enzymes are present in the CNS (32-34), it is likely that the addition and removal of the sulfate group is a control point for neurosteroid modulation of neurotransmitter receptors. This hypothesis is supported by the finding that inhibition of steroid sulfatase enhances learning and memory (35). Additionally, the presence of a charged sulfate group reduces the membrane permeability of steroids, raising the possibility that sulfated steroids could be concentrated and stored in vesicles, although stimulated release of neurosteroids has not been demonstrated.

Fig. 3.

The addition of sulfate to steroids is catalyzed by steroid sulfotransferases with the use of a sulfate group donated by 3′-phosphoadenosine-5′-phosphosulfate (PAPS). Sulfate groups may be removed by hydrolysis catalyzed by steroid sulfatases.

DHEAS increases while pregnenolone sulfate decreases the NMDA-induced release of [3H]norepinephrine from hippocampal slices. Both effects are antagonized by the σ1 receptor ligands progesterone and haloperidol. Pertussis toxin blocks the effect of DHEAS and reverses the effect of pregnenolone sulfate, suggesting that Gi/o plays a role in coupling σ1 and NMDA receptors (36). σ1 receptor immunoreactivity has been reported to translocate closer to the plasma membrane after stimulation by the membrane-permeable σ1 agonist (+)pentazocine (37). An unanswered question is how charged sulfated steroids, which would not be expected to readily cross the lipid bilayer, can reach σ1 receptors that are located on the endoplasmic reticulum (18).

Because the only known relative of the σ1 receptor, the yeast sterol C-8 and C-7 (C-8/C-7) isomerase, participates in the synthesis of the fungal analog of cholesterol, it has been suggested that the σ1 receptor plays a similar enzymatic role in mammalian cholesterol biosynthesis (38) (Fig. 4). However, no C-8/C-7 isomerase activity has been demonstrated for the σ1 receptor. This could be due to the absence of a necessary cofactor, but mammals have a sterol C-8/C-7 isomerase, which has no amino acid sequence similarity to either the yeast enzyme or the σ1 receptor. Although a few ligands bind to both the σ receptor and the mammalian isomerase with nanomolar affinity (39), the absence of a correlation between the rank order of potencies of a series of ligands indicates that the σ receptor and C-8/C-7 isomerases are nonidentical but may overlap in function, as might be expected if the σ1 receptor is a CNS variant of the isomerase. Moreover, although cholesterol biosynthesis is clearly a prerequisite for neurosteroid formation, it is difficult to understand how such a role could explain the functional interaction between σ ligands and exogenously administered neurosteroids such as DHEAS and PS, which are "below" cholesterol on the steroid biosynthetic pathway. Nevertheless, it is possible that the σ1 receptor plays some other enzymatic role in the neurosteroid pathway.

Fig. 4.

Two postulated roles for σ1 receptors in neurosteroid action. (A) Postulated role of intracellular σ1 receptors in mediating modulation of NMDA receptors by neurosteroids via a pertussis toxin-sensitive Gi/o protein (36). (B) Postulated role for σ1 receptor as a neural-specific C-8/C-7 sterol isomerase. The figure shows the formation of cholesterol in mammals by the action of a sterol C-8/C-7 isomerase and Δ 24-reductase on zymosterol, followed by the action of Δ5-dehydrogenase and Δ7-reductase on lathosterol (38).

The effects of PS on NMDA receptor-mediated synaptic activity in rat hippocampal cultures or ionic currents induced by receptors expressed in Xenopus oocytes occur in the low micromolar range. Significant effects are observed with as low as 1 or 2 µM steroid (9, 12, 40); however, this is above the average brain levels of PS. If sulfated steroids are packaged and released, then their synaptic concentrations could be orders of magnitude above average brain regional levels, and various research groups are actively investigating this possibility. However, it is also possible that there are other as-yet-unidentified sulfated neurosteroids that bind to glutamate receptors with higher affinity. Receptors in vivo or the NDMA signal transduction cascade may be more sensitive to PS, yielding responses from nanomolar concentrations of PS that are more consistent with its endogenous concentration. Factors that may influence steroid sensitivity include posttranslational modification of receptors, the presence of certain receptor subtypes, or the response of neural networks.

The therapeutic potential for neurosteroids acting at NMDA receptors is provocative. Pregnanolone hemisuccinate, a synthetic analog of pregnanolone sulfate, allosterically inhibits NMDA receptor activity in an insurmountable but non-voltage-dependent manner, distinguishing it from the high-affinity voltage-dependent noncompetitive inhibitors of the NMDA receptor. Behavioral experiments in rat or mouse show that sedation, hypnosis, and stupor occurs with increasing doses of pregnanolone hemisuccinate. At subsedative doses, pregnanolone hemisuccinate is neuroprotective and antagonizes infarct size induced by middle cerebral artery occlusion in rat. It is anticonvulsant against NMDA-induced seizures in mouse and analgesic in the mouse formalin footpad test for chronic pain (41-43).

Perhaps σ1 receptors play an enzymatic role in the synthesis of some high-affinity neurosteroid modulator of the NMDA receptor. Another possibility is that the σ1 receptor is a relative of the C-8/C-7 sterol isomerase that has been evolutionarily co-opted to play a role in the transduction of neurosteroid effects on NMDA receptors. There are various ways in which this could occur; the σ1 receptor could interact with NMDA receptors to modulate the direct effects of steroids on the NMDA receptor, or the σ1 receptor could itself act as a neurosteroid receptor that is coupled directly or indirectly to NMDA receptors.

As provocative as the link between neurosteroids and σ receptors appears to be, there is reason for caution. σ receptors almost seem to have been designed to mislead pharmacologists. Their ability to bind a wide variety of structurally diverse substances, many of which have important effects mediated by other receptor systems, has repeatedly illustrated the perils of being too hasty to conclude that a high affinity binding site for a drug mediates its pharmacological effects. It remains unclear whether the σ receptor is in fact a receptor in the physiological sense, or whether it plays some enzymatic or other role in neurosteroid signal transduction. A clear definition of the role of σ receptors may have to await the availability of genetic knockouts.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.

Stay Connected to Science Signaling

Navigate This Article