The data summarized here suggest the existence of a new central pathway for the hormonal regulation of pain. These data mainly collected in quail, a useful model in neuroendocrinology, demonstrate that numerous neurons in the superficial laminae of the spinal cord express aromatase (estrogen-synthase). Chronic and systemic blockade of this enzyme in quail alters nociception within days, indicating that the slow genomic effects of sex steroids on nociception classically observed in mammals also occur in birds and require aromatization of androgens into estrogens. However, by contrast with these slow effects, acute intrathecal inhibition of aromatase in restricted spinal cord segments reveals that estrogens can also control nociception much faster, within 1 min, presumably through the activation of a nongenomic pathway and in a manner that depends on an immediate response to fast activation/deactivation of local aromatase activity. This emergent central and rapid paracrine mechanism might permit instantaneous and segment-specific changes in pain sensitivity; it draws new interesting perspectives for the study of the estrogenic control of pain, thus far limited to the classical view of slow genomic changes in pain, depending on peripheral estrogens. The expression of aromatase in the spinal cord in other species and in other central nociception-related areas is also briefly discussed.
besides their crucial control of sexual differentiation during development and of reproductive physiology and behavior in adulthood, estrogens regulate a host of “nonreproductive” functions in the central nervous system, including protection against lesion-induced trauma, learning and memory, stress, mood, cognition, somatosensory perception, and pain sensitivity (81). The estrogenic regulation of pain has been demonstrated by numerous clinical observations and behavioral studies in the laboratory (20, 29). Differences in concentration and organizational/activational actions of sex steroids in male vs. female likely account for part of the sex differences in pain sensitivity (18, 40). Clinical observations have identified numerous pain conditions dependent on sex steroids and have found that their prevalence also varies as a function of sex (18, 40). Some interpretations suggest that the estrogenic regulation of both the somatosensory and pain systems could play an important role during reproduction by permitting the adequate central processing of reproductively relevant stimuli (e.g., genital stimulation during intercourse) (45, 48, 58); dysfunction of this evolutively appropriate regulation could, in turn, prevent proper reproduction (e.g., painful coitus or dyspareunia) (19, 58). The exact mechanisms through which sex hormones control pain appear to be highly diverse and complex, and they remain to be elucidated. In the present contribution, we summarized recent findings emphasizing the possible importance of central synthesis and action of estrogens in the regulation of pain through classical slow genomic pathways but also through fast nongenomic pathways only recently brought to the fore in pain research.
Estrogens exert their effects through at least two different cellular mechanisms. First, the genomic pathway implies the binding of estrogens to intracellular nuclear estrogen receptors (ERs; two isoforms in amniotes: ERα and β) controlling gene transcription (50, 77). This “classic” pathway alters neuronal physiology and behavior hours to days after the initiation of hormonal manipulation. Second, the nonclassic or nongenomic pathway requires the activation of a membrane estrogen-binding site not yet fully identified and localized but known to modify neuronal cell excitability within seconds through the activation of second messenger cascades (56, 82, 86). To date, the majority of studies of the estrogenic regulation of pain reported slow genomic effects occurring hours to days after hormonal treatment (2–4, 16, 21, 25, 28, 30, 31, 37, 51, 52, 57, 59, 60, 62, 64, 67, 69, 72, 76, 80, 85, 89–91, 97, 101, 108), but as we shall see rapid (nongenomic) effects of spinal aromatase on nociception, recent studies have demonstrated that estrogens can, in addition, act nongenomically, within seconds to minutes, on nociceptive behavior (39) and on response to adenosine in cultured dorsal root ganglion (DRG) cells (26, 27, 83).
Numerous areas involved in nociception in the nervous system contain ERα and/or β: DRG, dorsal horn of the spinal cord, periaqueductal gray, parabrachial nuclei, raphe nuclei, hypothalamus, limbic system, and several cortical areas (6, 17, 32, 54, 79, 93–95, 99, 103). Each of these areas is a possible substrate for the estrogenic regulation of pain. In the mammalian spinal dorsal horn, ERα and/or β are located mainly in lamina II and to some extent I, III, V, and X (e.g., 6, 93, 99). Basic research accordingly suggests that estrogens alter pain sensitivity by influencing, at least in part, spinal (anti-)nociceptive mechanisms within hours to days after hormonal replacement in gonadectomized subjects. In the spinal cord, 17β-estradiol (E2) increases the binding of tritiated muscimol to GABAA receptors and decreases the contents of substance P and neurokinin A (68, 73). These compounds actively mediate the transmission of nociceptive inputs from DRG to spinal cord (substance P and neurokinin A) or inhibit this transmission (GABA) (71). The colocalization of ERα with the preproenkephalin mRNA in the rat spinal cord and the demonstration that a bolus of estrogens enhances within 4 h the enkephalin gene expression in the spinal cord of ovariectomized female rats provide converging evidence for a control of nociception at the spinal level at least through genomic mechanisms (4, 5). Similarly, in the caudal trigeminal nucleus in male and female rats, the opioid receptor-like 1 (ORL1) mRNA is present in most ERα and/or β mRNA-containing neurons (41). Estrogens downregulate ORL1 in females, resulting in its differential expression between males and females at proestrus or after estrogen treatment when ovariectomized. Interestingly, extracellular recordings in the caudal trigeminal nucleus and behavioral tests indicate that high concentrations of estrogens induce pronociceptive effects of orphanin FQ on local NMDA-evoked response in proestrous females, whereas low concentrations of estrogens in males and ovariectomized females correlate with an opposite effect of orphanin FQ (i.e., antinociception) (42). This suggests that variations in orphanin effects and ORL1 expression at the trigeminal level due to fluctuations in plasmatic E2 concentration in females is partly responsible for sex difference in pain in rodents (41, 42).
THE SPINAL DORSAL HORN AS AN ESTROGEN PRODUCER: WHAT CAN A BIRD TELL US?
The enzyme aromatase or estrogen-synthase catalyzes the conversion of C19 androgens (e.g., testosterone, T) into estrogens (e.g., 17β-estradiol, E2). Although the ovary and, to a lesser extent, testicle are the major sources of estrogens released in the bloodstream, numerous tissues, including the brain, express aromatase. In the brain, the areas containing the enzyme are the hypothalamic and limbic systems, where locally formed estrogens control sex differentiation and reproductive behaviors (i.e., preoptic area, amygdala, and bed nucleus of the stria terminalis) (14, 23, 49, 74, 87, 88). These effects are largely exerted through the activation of nuclear estrogen receptors coexisting (i.e., expressed in a same area but not necessarily by the same cell) but rarely colocalized (i.e., expressed by the same cell) with aromatase (15, 70, 98, 100).
The expression of aromatase in the spinal cord, where ERα and β are expressed, was previously assessed in female rats (63). The absence of any detectable level of aromatase activity (AA) logically led to the conclusion that estrogens acting in the spinal cord exclusively arise from the periphery (i.e., gonads). The evolution of the spinal cord apparently occurred without major functional change across species (22, 75); homologies even include the expression of ERα in the superficial laminae of the dorsal horn in the rat, cat, quail (Fig. 1A), and ring dove (6, 32, 66, 99, 104)1 . By contrast with mammals, quail provide a useful model for the study of neural aromatase (13). The aromatase protein is easily detected by immunocytochemistry in developing and adult quail brains, and its distribution reliably correlates with the detection of AA and aromatase mRNA (8, 78). In rats and, especially in adults, immunocytochemistry reveals the presence of a less dense and less intense population of aromatase-immunoreactive (ARO-ir) cells in the forebrain and, for unexplained reasons, their distribution does not always match that of AA and aromatase mRNA (43, 49, 87). AA is also about 10 times higher in quail than in rats. On this basis, Japanese quail were chosen to reassess the presence of aromatase in the adult spinal cord as a first approach to study the hormonal regulation of nociception at the spinal level in vertebrates (33).
In adult male and female quail, ARO-ir somata and fibers were found in the spinal dorsal horn from the upper cervical segment to the lower caudal area, mostly in laminae I and II, with additional sparse cells being present in lamina III, the medial part of lamina V (now partly reinterpreted as the centrally located column of Terni, similar to the mammalian intermediolateral column), the lateral spinal nucleus, and lamina X (Fig. 1, A and B) (33). Radioenzyme assays, based on the measurement of tritiated water released from the conversion of tritiated androgens into estrogens by aromatase, confirmed the presence of substantial levels of AA throughout the rostrocaudal extent of the spinal cord (Fig. 1D). Interestingly, the mechanisms that control the spinal AA are different than what has been described in the limbic system. Indeed, spinal AA and the number of ARO-ir cells in five representative segments of the spinal cord were not different in sexually mature males or females and were not influenced in males by castration associated or not with a treatment with testosterone (Fig. 1, C and D). In the absence of a control of spinal AA by steroids, we previously proposed that peripheral, intrinsic, and descending inputs to spinal ARO-ir cells regulate AA (33). Interestingly, all ARO-ir neurons in laminae I–III and in the lateral spinal nucleus coexpress the receptor for substance P (neurokinin 1 receptor) and are in close apposition with substance P-immunoreactive fibers, presumably largely arising from nociceptive primary afferent fibers (36). Substance P was previously shown to regulate brain AA and could therefore modulate spinal AA on a segment-specific basis (1) [see also rapid (nongenomic) effects of spinal aromatase on nociception]. Taken together, these data demonstrated, for the first time, the presence a local estrogen production in presumed nociceptive neurons in the spinal cord of an amniote vertebrate; they clearly suggested an implication of spinal aromatase in the modulation of nociception by estrogens. Furthermore, given the important conservation of the spinal cord and of its expression of ERα throughout phylogeny, these data suggest that the presence of spinal aromatase in rats should be reassessed (see distribution of aromatase in other nociception-related areas and in other species).
SLOW GENOMIC EFFECTS OF AROMATIZATION ON NOCICEPTION
A hot-water test was found to be the easiest, most sensitive, and most reliable method to evaluate behavioral responsiveness to noxious stimuli in quail (38). Basically, a foot of the bird was immersed in a hot water bath and the latency of the foot withdrawal was recorded. Birds were found to react to this test in a very reproducible manner, and latencies were clearly related to the temperature of the water (Fig. 2A). At 54°C, intact male (noncastrated, sexually mature male quail exposed to a long-day photoperiod and thus having a high plasma concentration of T, as verified by the larger dimension of the cloacal gland) displayed a mean latency averaging 4 s (38). In a set of experiments designed to test the chronic and long-term effect of aromatization on the latency, castrated male quail were implanted with subcutaneous capsules that were empty (CX), filled with crystalline T (CX+T), or crystalline E2 (CX+E2) (37). Two weeks after implantation, the hot water test (54°C) revealed that CX birds displayed a markedly longer latency than intact subjects; conversely, E2 and T similarly restored the withdrawal latency to the baseline level typical of intact males (Fig. 2B). During a 10-day period, birds from each experimental group received a daily intraperitoneal injection of either a nonsteroidal aromatase inhibitor (vorozole), a nuclear estrogen receptor antagonist (tamoxifen), or their vehicle solution (polyethylene glycol). All birds were then submitted to the hot-water test (54°C) 24 h after the last intraperitoneal injection. Injections were discontinued, and birds were tested again 2 wk and 4 wk after the last injection to follow the elimination of the effect of the drugs. Twenty-four hours after the last injection, vorozole significantly increased the latency in CX+T birds but had no effect in the CX+E2 and CX subjects (Fig. 2C). The effect of vorozole totally disappeared 2 and 4 wk after the last injection (not shown here). The 10-day treatment with tamoxifen completely blocked the effects of both E2 and T on foot withdrawal latency (Fig. 2D). The effect of tamoxifen ceased in CX+T birds after 2 wk (not shown here). In contrast, after 2 wk, CX+E2 birds retained latency higher than the baseline and only reached the baseline 2 more wk later.
These behavioral studies using a paradigm similar to those previously used in rodents thus showed that T and E2 regulate nociception in quail similarly to what has been shown in rats or mice. But in addition, the blockade of the effects of T by vorozole and by tamoxifen revealed that the effects of androgens on nociception require, at least in part, conversion into estrogens. The time course of the observed changes in foot withdrawal latencies after the treatment with T or E2 reflects long-term effects of E2 on nociception, presumably mediated by the activation of nuclear estrogen receptors. The presence of ARO-ir neurons and of ERα-ir cells in the dorsal horn of the spinal cord suggests that these effects occur, at least in part, at the spinal cord level and, hence, that chronic and pathological disruption or hyperactivation of aromatase might result in long-term changes in pain sensitivity, hereby increasing risks of estrogen-dependent pain conditions. Further experiments using, for example, prolonged spinal implants of steroid-related drugs will be necessary to more fully assess this hypothesis. Interestingly, estrogen receptors seem predominant in sensory areas, while androgens appear to be preferentially expressed in motor and premotor areas (6, 32, 46, 53, 66, 93, 99, 102, 107); this differential distribution remains to be carefully verified (see Ref. 32), but it could partly explain how estrogens rather than androgens seem to regulate nociception.
RAPID (NONGENOMIC) EFFECTS OF SPINAL AROMATASE ON NOCICEPTION
Besides exerting slow genomic effects through ERα and β (50, 77), estrogens regulate cell functions much faster, within seconds, through the activation of membrane estrogen-binding sites linked to second messenger cascades (55, 56, 82, 84). In the nervous system, such effects of estrogens are often assumed to resemble those of neurotransmitters and neurosteroids (e.g., DHEA). One can thus suggest that rapid effects of estrogens require fast adaptation of the concentrations of estrogens in the vicinity of their site of action via a rapid control of local AA, followed by a fast catabolization of the locally formed estrogens to ensure a rapid, but also short-term, effect. On this basis and given the recent demonstration that aromatase can be rapidly regulated by phosphorylation in the brain (10–12), it was hypothesized that fast changes in spinal AA rapidly deplete local estrogen concentration and instantly prevent rapid effects of estrogens on the behavioral responsiveness to nociceptive stimuli. To test this hypothesis, we mimicked a rapid phosphorylation-dependent AA inhibition by using an acute intrathecal injection of aromatase inhibitors at the lumbar level in intact male quail, and we then tested the effects of this injection 1, 5, and 30 min later using the hot-water test (39).
Vorozole significantly increased the foot withdrawal latency 1 and 5 min after intrathecal injection, and this effect ended within 30 min (Fig. 3A). To further assess the notion that the responsiveness to noxious stimuli was indeed modulated by spinal aromatase, we tested with the same protocol the effects of 1,4,6-androstatriene-3,17-dione, a steroidal aromatase inhibitor structurally unrelated to vorozole; we obtained identical results (Fig. 3B). Finally, we coinjected vorozole and E2 in the spinal cord and measured foot withdrawal latency at the same intervals as before. Although subjects injected with vorozole alone displayed an increase in latency, those simultaneously treated with vorozole and E2 displayed nearly constant withdrawal latencies (Fig. 3C). The injection of vorozole directly into the brain of males did not affect the foot withdrawal latency from hot water, suggesting that these rapid effects of local aromatization might be specific to the spinal cord (not shown) (39).
Overall, this new effect of spinal aromatization on nociception suggests the existence of an emergent central mechanism for the hormonal regulation of pain. By contrast with the previous studies showing long-term genomic effects of estrogens on nociception, the present finding indicates that estrogens can rapidly modulate behavioral responsiveness to painful stimuli at the spinal level. This effect depends on rapid control of local aromatization of androgens into estrogens and thus suggests that minute-to-minute changes in spinal AA restricted to some spinal segments might result in topographically distinct alteration of nociception.
Earlier studies of the rapid effects of estrogens at the cell membrane in the hypothalamus suggest that the rapid estrogenic regulation of nociception mediated at the spinal level might result at least in part from a precise interaction of estrogens with the opioid and glutamate receptors repeatedly shown to play a critical role in the spinal processing of peripheral nociceptive inputs (39, 86).
In the present experiment, spinal AA was altered using exogenous inhibitors. However, recent studies suggest that various spinal components involved in nociception might directly regulate spinal aromatase either through slow control of the transcription of its gene (CYP19) or through fast activation/deactivation, involving kinases and phosphorylation of the aromatase protein. For instance, glutamate, a major neurotransmitter released by thin primary afferent fibers (PAF) in the superficial spinal laminae has been shown to rapidly regulate AA in the quail preoptic area and could thus exert a similar effect in the spinal cord (10, 61, 71). Substance P has been shown to alter CYP19 transcription in the brain and is also known to more rapidly regulate cell function through phosphorylation. Its presence in the spinal cord as a major neuropeptide released by nociceptive-specific PAF suggests that it could alter spinal AA in both a genomic and nongenomic fashion (1, 9, 24). Cyclooxygenase-2 (COX2) synthesizes prostaglandins, including PGE2, in the periphery but also in the spinal cord in response to chronic painful stimuli giving rise to inflammation (96, 105). Various studies in tumor cells have shown that PGE2 upregulates the concentration of aromatase mRNA, and interestingly, estrogens have been shown to stimulate COX-2 in the periphery (92, 109). These results suggest that estrogens and PGE2 locally formed in the spinal cord could enhance the activity of COX-2 and aromatase, respectively. The impact of such interaction on inflammatory pain should be investigated. Finally, given the segmental organization of the PAF terminals in the spinal dorsal horn, both slow and fast nongenomic control of spinal AA could result in slow but prolonged or, on the contrary, instant but short-lived alteration, respectively, of the estrogenic regulation of nociception in specific segments of the spinal cord. This possibility should be kept in mind considering that various pain conditions in women are regulated by estrogens and are only dependent on restricted segments of the spinal cord or medulla. For example, migraine largely depends on the caudal nucleus of the trigeminal system. Similarly, dyspareunia largely depends on spinal segments at the lumbosacral level. It would be interesting, for instance, to study whether exogenous control of spinal aromatase in very specific segments of the spinal cord could alter nociception and result into physiological, neurophysiological, and behavioral changes typically observed during dyspareunia or migraine. Numerous estrogen-dependent pain conditions are more prevalent perimenstrually or in postmenopausal women; after verifying the expression of aromatase in the human spinal cord, it would also be interesting to study whether spinal aromatization in women may fail to adequately counterbalance cyclical or postmenopausal changes in peripheral estrogen synthesis in specific segments of the spinal cord.
DISTRIBUTION OF AROMATASE IN OTHER NOCICEPION-RELATED AREAS AND IN OTHER SPECIES
The divergent side-to-side or schizocerate arrangement of laminae I–III characterizes the spinal cord of most avian groups, including galliformes, whereas in a few other avian species and in mammals, these laminae have conserved a more ancient dorsoventral arrangement (106). The peculiar schizocerate dorsal horn logically leads to the idea that spinal aromatase might be a specific trait of this variety of birds. However, functions of laminae I–III in both a variety of birds and in mammals appear to be closely identical, and estrogen receptors were observed in the same homologous areas in the spinal cord in quail (schizocerate), ring dove (leiocerate), and mammals. In addition, more recently, studies on the leiocerate passeriform manakin demonstrated the expression of aromatase mRNA in laminae I and II throughout the rostrocaudal extent of the spinal cord (44). In their previous studies, MacLusky et al. (63) did not detect AA in the rat spinal cord and concluded that estrogens regulating spinal function are formed in the periphery. The apparent lack of aromatase in the rat spinal cord could be due to the lack of sensitivity of the AA assay method available at that time. On the basis of the results gathered in quail and of the apparently good conservation of the spinal cord during evolution, we reassessed the presence of aromatase in the rat spinal cord using immunocytochemistry (Evrard HC, Balthazart J, Harada N, Erskine MS, unpublished observation). A careful observation of the sections under microscope revealed the presence of a dense immunoreactive staining in the laminae I and II of the dorsal horn, in the lateral spinal nucleus, around the central canal, and in the intermediolateral cell column. No qualitative differences were noted between males and females. This staining corresponds almost exactly to the previous observation made in quail, suggesting that the expression of aromatase, like the expression of estrogen receptors, is, in fact, a well-conserved character. However, a striking difference was that aromatase immunoreactivity was observed in fibers with punctate structures; however, by contrast, it was not observed in cell bodies with Japanese quail spinal cord, in which both fibers and cell bodies were immunoreactive for the anti-aromatase antibody. This difference might be important for our understanding of the difference in central estrogen synthesis/action in birds vs. mammals, and is currently under study.
As mentioned above, numerous brainstem nuclei involved in nociception express estrogen receptors: spinal trigeminal, solitary tract, parabrachial, raphe, and locus coeruleus nuclei. Like for the spinal cord and the limbic system, we recently found that these areas express aromatase immunoreactivity in adult quail (35). This finding suggests again that locally formed estrogens could alter the function of these nuclei, including the regulation of nociception. Homologous areas in rats contain less intense and less dense, but still distinct, aromatase immunoreactivity in fibers (axons and/or dendrites) but not in cell bodies similar to what has been observed in the rat spinal cord (Evrard HC, Harada N, Erskine MS, unpublished data).
In both humans and animals, an unresolved discrepancy distinguishes studies showing hyperalgesic effects of estrogens from those reporting antinociceptive, as well as pain conditions enhanced with estrogen-based treatment from those showing the reverse. This discrepancy was also observed in birds. Using a similar hot-water test in two different avian species, we found that androgens and estrogens increase behavioral responsiveness to a thermal stimulus in Japanese quail but reduce it in male house sparrows (37, 47). The understanding of such a difference will probably represent a key turn in the study of the estrogenic regulation of nociception. Various causal factors have already been suggested to explain these inconsistencies (species, strain, methodological approach, stimulation site, and type of stimulus), and we propose that the large diversity of central (and peripheral) sites where estrogen receptors are expressed and, moreover, where estrogens can be produced, represents a substrate for such divergence across studies.
In conclusion, the regulation of nociception by estrogens and the expression of nuclear estrogen receptors in the dorsal horn of the spinal cord have been demonstrated in a wide range of species, including birds, rodents, and primates. This suggests that the estrogenic regulation of pain might rely, at least in part, on a central neuroendocrine mechanism well conserved throughout phylogeny. Conservation often suggests the relative importance of a trait and, in the present context, it indicates that the estrogenic regulation of the spinal processes of nociception might be of nonnegligible importance in the physiology of healthy individuals and hence in patients suffering from pain conditions correlated with hormonal dysfunctions. This contribution reviewed various data mainly collected in a bird but with some correspondence to mammals, too. These data demonstrate that the enzyme aromatase is expressed in the spinal dorsal horn in males and females. They also suggest that spinal aromatase ensures a local source of estrogens that can regulate the spinal mechanism of nociception independently from gonadal estrogens and that can be modulated on a local, segment-specific basis.
The absence of sex difference in spinal AA and ARO-ir cell number suggests the absence of sex difference in spinal aromatase expression and effects. In addition, the fact that testosterone did not alter AA and ARO-ir cell number in males indirectly suggests that spinal AA in females does not vary across the ovarian cycle. To understand to what degree spinal aromatase might affect pain differentially in males and females, comparative studies are now necessary to determine whether the inhibition of spinal aromatase in females results in effects of a similar amplitude to the effects already observed in males. It would also be interesting to study whether the effects of spinal aromatase on nociception in females have a similar impact during the different periods of the ovulatory cycle, whether spinally formed estrogens coincidentally counterbalance the cyclical decrease of peripheral estrogen concentration, and whether local and rapid regulations of spinal AA via phosphorylation differ between males and females. One complexity of the study of the estrogenic regulation of pain and, maybe, of the sex difference in pain arises from the broad distribution of the estrogen-sensitive and also estrogen-synthesizing cells in the nervous system and in the peripheral tissues.
Finally, the present review emphasized that, besides exerting slow genomic effects, estrogens produced by and acting in the spinal cord induce a fast, presumably nongenomic change in responsiveness to painful stimuli. This constitutes a new central mechanism for the control of hormonal regulation; this mechanism remains to be fully characterized to understand its relative importance in the estrogenic regulation of nociception in males vs. females and its impact on physiology and behavior throughout life span in naturally behaving animals including, perhaps, humans.
The research reviewed in this paper was conducted by the author in the laboratories of Dr. J. Balthazart [University of Liège; National Institutes of Health (NIH) Grant MH-50388 to Drs. G. F. Ball and J. Balthazart and a grant from the Belgian Fund for Collective Research (FRFC 2.4562.05 to J. Balthazart) and Dr. M. S. Erskine (Boston University; NIH Grants MH-64187 and MH-01435). This work was also supported by the 2004 Award of the Institut Belge de la Douleur-UPSA-Belgisch Pijninstituut.
Present address of H. C. Evrard: Barrow Neurological Institute, 350 West Thomas Rd., Phoenix, AZ 85013.
The author thanks Drs. Erskine and Balthazart for support throughout the research reviewed in this report.
↵1 In mammals, paleognathes (“ancient birds”), and in few neognathes, including passeriformes (songbirds) and some columbiformes (e.g., ring dove and pigeon), Rexed's laminae I-VI are superposed dorsoventrally (“leiocerate” organization). By contrast, in the majority of neognathes, including galliformes (e.g., chicken and quail), the organization of the dorsal horn has evolved into a new side-to-side or “schizocerate” arrangement of laminae I-III (106). Despite this different arrangement, the homonymous in mammals, leiocerate and schizocerate birds appear to be functional homologous. Because what we report in quail could be interpreted as a peculiarity of the schizocerates, information on the distribution of estrogen receptor and aromatase for leiocerate birds were also provided when available.
- Copyright © 2006 the American Physiological Society