Ambient Temperature and 17β-Estradiol Modify Fos Immunoreactivity in the Median Preoptic Nucleus, a Putative Regulator of Skin Vasomotion

Abstract

Estrogen has pronounced effects on thermoregulation, but the anatomic sites of integration between the reproductive and thermoregulatory axes are unknown. In this study, we tested whether estradiol-17β (E₂) treatment would alter the activity of thermoregulatory brain regions responding to mild changes in ambient temperature (T_AMBIENT). Core and tail skin temperatures were recorded at the ambient temperatures of 20, 24, or 31 C in ovariectomized (OVX) rats with and without E₂. Neuronal activity was evaluated by counting the number of Fos-immunoreactive cells in the brains of rats killed 90 min after exposure to one of the three ambient temperatures. Of 14 brain areas examined, the median preoptic nucleus (MnPO) was the only site that exhibited increased Fos immunoreactivity at the high T_AMBIENT of 31 C. At 24 C, OVX rats exhibited increased numbers of MnPO Fos-immunoreactive cells, compared with OVX + E₂ rats. Interestingly, tail skin vasomotion and MnPO Fos expression were affected in a similar manner by T_AMBIENT and E₂ treatment. In the arcuate nucleus and anteroventral periventricular nucleus (AVPV), Fos immunoreactivity was highest at the low T_AMBIENT of 20 C, with inhibitory (arcuate nucleus) and stimulatory (AVPV) effects of E₂. No other areas responded to both T_AMBIENT and E₂ treatment. These results implicate the MnPO, the arcuate nucleus, and the AVPV as sites of integration between the reproductive and thermoregulatory axes. Combined with studies showing the importance of MnPO neurons in heat-defense pathways, the MnPO emerges as a likely site for E₂ modulation of thermoregulatory vasomotion.

Thermoregulation is altered by the levels of ovarian hormones throughout a woman's life span. Changes occur across the menstrual cycle, with oral contraceptive use, pregnancy, and most dramatically as a result of the menopause. During the menopausal transition and early postmenopausal phase, most women experience hot flushes, a syndrome consisting of the episodic activation of heat dissipation mechanisms (1). Hot flushes are secondary to estrogen withdrawal because they are triggered by conditions that lower serum estrogens and are effectively treated by estrogen replacement (2, 3). Hot flushes are temporally linked to pulses of LH in peripheral plasma (4), although clinical studies indicate that neither LH nor GnRH is required for hot flush symptoms to occur (5, 6). These data provide evidence of a central neural trigger for hot flushes related to the hypothalamic control of pulsatile GnRH secretion (7). Despite the high prevalence of this syndrome, little is understood about the etiology of hot flushes or the biological effects of estrogens on thermoregulation in general.

Tail skin vasodilatation is an important heat dissipation effector in rodents (8). Fluctuations in tail skin temperature are characteristic of normal thermoregulation within the thermoneutral zone, in which body temperature is regulated only by dry heat loss mechanisms such as active changes in skin vasomotion (9). In intact cycling rats, these large fluctuations in tail skin temperature are reduced on the night of proestrous (10), consistent with a shift in the thermoneutral zone with increasing serum estrogen (11). Of note, the average tail skin temperature is increased by ovariectomy, and this effect is reversed by treatment with estrogens (10–17). Estrogen treatment of ovariectomized (OVX) rats also alters the ambient temperature (T_AMBIENT) threshold for the initiation of skin vasodilatation, resulting in a shift in the thermoneutral zone (11). Moreover, increased heat escape behavior occurs in OVX rats, compared with estrogen-replaced controls (13). Thus, OVX rats demonstrate increased activation of heat dissipation effectors, compared with OVX rats with estrogen replacement. Although the mean increase in tail skin temperature induced by ovariectomy of rats does not mimic the episodic syndrome of flushing in women (12), studies of these phenomena are useful in understanding how estrogen modifies thermoregulatory heat-defense mechanisms from a basic science perspective.

Skin vasodilatation and other thermoregulatory effectors are hierarchically controlled by the hypothalamus, with the preoptic area playing a critical role (18, 19). Independent autonomic effectors involved in heat defense are regulated by preoptic neurons that integrate information from core body and skin temperature as well as other physiological conditions (19, 20). Previous studies have shown that estradiol-17β (E₂) modulates the electrical activity of warm-sensitive preoptic neurons in tissue slice preparations (21). Moreover, estrogen receptor-α mRNA is expressed in many nuclei in the rostral hypothalamus, including the median preoptic nucleus (MnPO), anteroventral periventricular nucleus (AVPV), and medial preoptic area (22). Based on these data, we hypothesized that neurons in the rostral hypothalamus/preoptic area could mediate the effects of estrogens on tail skin vasodilatation. However, it is not known whether E₂ modifies the activity of warm-sensitive thermoregulatory centers in vivo, and little information is available on the sites of integration between the reproductive and thermoregulatory axes.

In the present study, we determined whether E₂ treatment of OVX rats modified the activation of hypothalamic regions that respond to mild changes in ambient temperature. Fos immunohistochemistry was chosen as a marker of activity based on numerous studies showing increased hypothalamic Fos expression in response to a variety of reproductive (23) or thermoregulatory stimuli (24–27). Ambient temperatures were selected near the upper and lower boundaries of the thermoneutral zone, based on our previous studies (11). Additional core and tail skin recordings were performed to confirm the thermoregulatory state of the animals at these selected ambient temperatures. Because tail skin vasodilatation is elevated at warm ambient temperatures and increased in OVX rats without estrogen replacement (11, 13–16, 28), an important goal was to determine the effects of estrogen on Fos immunoreactivity in hypothalamic thermoregulatory nuclei that responded to high ambient temperatures.

Materials and Methods

Animal care procedures

Animal protocols were approved by the University of Arizona Animal Care and Use Committee and conformed to National Institutes of Health guidelines. Female Sprague Dawley rats (150–200 g, Harlan Sprague Dawley; Harlan Laboratories, Indianapolis, IN) were housed with free access to food and water on a 12-h light, 12-h dark schedule (lights on at 0700 h) in the Animal Care Facility at the University of Arizona. Because high levels of phytoestrogens in standard rat chow can affect tail skin temperatures (15, 29), rats were fed low-phytoestrogen rat chow (Harlan Teklad 2014; Harlan Laboratories, Houston, TX).

Rats were ovariectomized under general anesthesia and implanted ip with a Subcue temperature data logger (Canadian Analytical Technologies, Inc., Calgary, Alberta, Canada). Ten days later, rats were anesthetized with isoflurane gas and implanted with two SILASTIC capsules (sc; Dow Corning, Midland, MI) (30 mm, 1.57 mm inner diameter, 3.18 mm outer diameter; Dow Corning, Midland, MI) containing either E₂ in sesame oil (180 μg/ml, OVX + E₂ group) or saline (OVX group). The SILASTIC capsules were plugged on each end with 5-mm wood sticks and silicone glue, resulting in an effective capsule length of 20 mm. Similar SILASTIC capsules have been shown to deliver physiological concentrations of E₂ for 5 wk after implantation (30). In our hands, this treatment results in serum E₂ levels of 24.8 ± 0.6 pg/ml and suppression of serum LH when measured 3–4 wk after implantation (11). Moreover, we have previously shown that this E₂ treatment alters the thermoregulatory system in OVX rats (11). Saline was used as a vehicle control instead of sesame oil because of initial concerns that phytoestrogens in the sesame oil could exert an independent effect on thermoregulation. However, we have subsequently detected no difference in tail skin temperature or core temperatures of OVX rats with sesame oil capsules vs. saline capsules (10).

Experiment 1: measurements of core and tail skin temperatures of OVX + E₂ and OVX rats exposed to ambient temperatures of 20, 24, and 31 C

These experiments were performed between 0800 and 1500 h, 9–25 d after E₂ or saline capsule implantation. To minimize stress, rats were habituated to the recording procedures for at least three sessions before data acquisition. An IT-18 thermocouple (Physitemp Inc., Clifton, NJ) was taped 5–6 cm from the base of the tail. This thermocouple was protected with a Nalgene T-tube fitted over 1 in. of the rat's tail and polyethylene tubing over the wire. Rats were placed in a grid plastic cage (6 in. × 6 in. × 4 in.) with food and an attached water bottle. The thermocouple was threaded through the cage roof, taped to a swivel, and inserted into a TC4000 or QuadTemp data logger (Madgetech, Inc., Contoocook, NH). This experimental set-up allowed the rats freedom of movement, unlimited access to food and water, and the ability to regulate temperature by changes in posture and grooming. The plastic cages were placed in an environmental chamber (Forma model 3940; Thermo Scientific, San Jose, CA) with ambient temperatures (T_AMBIENT) set at 20, 24, or 31 C and humidity set at 50%. Core temperature (T_CORE) was recorded every 10–15 min and tail skin temperature (T_SKIN) was recorded every 5 sec. A separate thermocouple was placed in the environmental chamber to record T_AMBIENT. Data were recorded after 1 h of acclimation to the environmental chamber for each T_AMBIENT. In a first protocol, temperatures of 12 OVX + E₂ and 11 OVX rats were recorded for a 1-h session at each T_AMBIENT in a random order, with two recordings per day (11). In a second protocol, 11 OVX + E₂ and 11 OVX rats were recorded twice at each T_AMBIENT for 2–3 h, with one recording per day and 1 wk separating the two sessions at each T_AMBIENT. The results of these two protocols were not significantly different, and these data were pooled. In addition, the length of time after estrogen replacement did not affect the T_CORE or T_SKIN results, consistent with our recent demonstration that the effects of E₂ on the tail skin temperature in OVX rats during the light-phase are stable from 8 to 21 d after treatment (10).

Data analysis

Skin vasomotion was indirectly evaluated by the heat loss index (HLI) (9): HLI = (T_SKIN − T_AMBIENT)/ (T_CORE − T_AMBIENT).

During tail skin vasoconstriction, HLI decreases to a theoretical minimum of 0 due to T_SKIN approaching T_AMBIENT. During tail skin vasodilatation, HLI increases to a theoretical maximum of 1 due to T_SKIN approaching T_CORE (9). Average T_CORE and HLI were calculated for each rat at each T_AMBIENT, and these values were used to generate group averages. Differences between groups were tested by two-way ANOVA (T_AMBIENT × E₂) using SigmaPlot Software (Systat Software, Inc., San Jose, CA). If there were significant overall effects of T_AMBIENT or E₂, post hoc comparisons were made using Tukey's multiple comparisons tests (α = 0.05).

Experiment 2: analysis of Fos-immunoreactive (ir) cells in the brain nuclei of OVX + E₂ and OVX rats exposed to ambient temperatures of 20, 24, and 31 C

Fifty rats were ovariectomized and implanted with sc capsules as described above (OVX + E₂, n = 25; OVX, n = 25). To minimize stress, rats were habituated to the experimental procedures in at least five sessions over 10 d. Between 10 and 21 d after capsule implantation, each rat was placed in a plastic grid cage, with food and water, and this cage was placed inside an environmental chamber set at 20 C. After 1–3 h of acclimation at 20 C, the chamber temperature was either changed to 24 or 31 C or kept at 20 C. After 90 min, each rat was removed, administered an overdose of sodium pentobarbital (100 mg/kg, ip), and perfused via the ascending aorta with heparinized saline followed by 4% paraformaldehyde in 0.1 m PBS (pH 7.4). The rats were killed between 1000 and 1400 h. The brain was removed, postfixed for 1 h in 4% paraformaldehyde, and cryoprotected in ascending solutions of sucrose in PBS (pH 7.4) (10, 20, and 30%). Each brain was blocked using a rat brain matrix (Braintree Scientific, Braintree, MA), and frozen sections in the coronal plane (40 μm thickness) were then taken with a sliding microtome. Sections were stored at −20 C in cryoprotectant solution (31).

The Fos antiserum (AB-5, lot D00007099; EMD Biosciences Inc., La Jolla, CA) was generated in the rabbit against amino acids 4–17 of human c-Fos. This specificity of this antibody has been verified by preabsorption experiments (32, 33) and negative staining of sections from Fos knockout mice (32). Every 10th section was Nissl stained to aid with matching sections to plates of a rat brain atlas (34). Sections from the same levels of the brain were processed simultaneously for Fos immunohistochemistry. Free-floating brain sections were rinsed in 0.1 m PBS (pH 7.4) (3 × 10 min), treated with 0.3% H₂O₂ in PBS for 30 min, and rinsed again. Slices were then immersed for 60 min in blocking solution (3% normal goat serum in PBS with 0.3% Triton X-100) and incubated for 40 h at 4 C with a Fos antiserum in blocking solution (1:20,000, optimal concentration determined by prior titration). Slices were rinsed, incubated for 2 h with biotinylated goat antirabbit IgG (1:250; Vector Laboratories, Burlingame, CA) in blocking solution, and rinsed again. Sections were treated with a Vectastain Elite ABC kit for 90 min and rinsed, and then Fos immunoreactivity was visualized with a standard diaminobenzidine reaction (0.5 mg/ml diaminobenzidine, 0.01% H₂O₂ in PBS, 5–6 min). Control sections with omission of the primary antibody did not exhibit staining.

Fos-ir neurons were identified by dark brown nuclear staining and were counted in selected brain areas using an image-combining computer microscope consisting of a motorized stage (LUDL Electronics, Hawthorne, NY), Lucivid miniature monitor (MicroBrightfield, Colchester, VT), and Neurolucida software (MicroBrightfield). The borders of specific brain areas were manually digitized with the assistance of Nissl-stained sections and a rat brain atlas (34). We focused on brain areas of the rostral hypothalamus and adjacent septal regions based on their role in thermoregulation: AVPV, lateral preoptic area, medial preoptic area, medial preoptic nucleus, medial septal nucleus, MnPO, parastrial nucleus, septohypothalamic nucleus, and the ventrolateral preoptic area. Additional nuclei that have been previously implicated in thermoregulatory (dorsomedial nucleus, paraventricular nucleus) or reproductive processes (arcuate nucleus, ventromedial nucleus) were studied. The AVPV, lateral preoptic area, MnPO, medial septum, medial preoptic area, parastrial nucleus, and septohypothalamic nucleus were outlined in brain slices matched to plates 32 and 34 of the rat brain atlas (34); the medial preoptic nucleus in sections matched to plates 36 and 37; the paraventricular nucleus in sections matched to plates 46 and 48; and the arcuate nucleus, dorsomedial nucleus, and ventromedial nucleus to plates 57 and 59 (Fig. 1). The ventrolateral preoptic area was delineated by a 300- × 200-mm box combined with a 300- × 100-mm triangle (plate 34, Fig. 1) or a 500- × 300-mm box (plates 36 and 37, Fig. 1) based on previous studies (34–36).

Fig. 1.

Line drawings adapted from the Paxinos and Watson rat brain atlas (34) showing the location of areas analyzed for Fos immunoreactivity. The corresponding atlas plate is listed in the top center, and the distance from bregma is at the bottom left of each drawing. Analyzed regions (shaded areas) were outlined with the aid of Nissl-stained sections. Scale bar (upper left drawing), 500 μm (applies to all). 3V, Third ventricle; ac, anterior commissure; AHA, anterior hypothalamic area; ARC, arcuate nucleus; DMN, dorsomedial nucleus; f, fornix; LH, lateral hypothalamus; LPO, lateral preoptic area; LV, lateral ventricle; ME, median eminence; MPN, medial preoptic nucleus; MPO, medial preoptic area; MS, medial septal nucleus; mt, mammillothalamic tract; oc, optic chiasm; ot, optic tract; POA, preoptic area; PS, parastrial nucleus; PVN, paraventricular nucleus; RCh, retrochiasmatic area; SCH, suprachiasmatic nucleus; SHy, septohypothalamic nucleus; VLPO, ventrolateral preoptic area; VMN, ventromedial nucleus.

During data acquisition and analysis, experimenters were blinded to the experimental groups. Fos-ir cells were manually mapped and counted within the digitized border of each brain region. All brain regions were counted bilaterally (one to three sections per brain area) except for the MnPO, which is a midline nucleus. These data were averaged to give the number of cells per unilateral section, or in case of the MnPO, the number of cells per section. For simplicity, these data are reported as the number of cells per section. The number of cells per section was calculated for each rat, and these numbers were used to generate group means and se. Data were analyzed using SigmaPlot software with two-way ANOVA (T_AMBIENT × E₂). If there were significant overall effects for T_AMBIENT or E₂, post hoc comparisons were made using Tukey's multiple comparison test (α = 0.05).

Results

Thermoregulatory status of OVX + E₂ and OVX rats exposed to ambient temperatures of 20, 24, and 31 C

Core temperatures of OVX + E₂ rats were normothermic at the environmental temperatures of 20, 24, and 31 C (37.06 ± 0.05, 37.00 ± 0.06, and 37.03 ± 0.07, respectively, mean ± sem). The T_CORE of OVX rats was not significantly different from the OVX + E₂ groups at the ambient temperatures of 20 and 24 C (36.95 ± 0.05 and 36.90 ± 0.07 C). However, at the highest T_AMBIENT of 31 C, the average T_CORE of OVX animals was significantly increased (with respect to all other OVX and OVX + E₂ groups) to 37.30 ± 0.09 C (Fig. 2A).

Fig. 2.

Average T_CORE (A) and HLI (B) in OVX + E₂ and OVX rats exposed to ambient temperatures of 20, 24, or 31 C. A, At 20 and 24 C, the average core temperatures were not significantly different between the OVX + E₂ and OVX animals. At the T_AMBIENT of 31 C, the OVX rats (but not the OVX + E₂ rats) exhibited a significant increase in T_CORE compared with OVX rats at 20 and 24 C and OVX + E₂ rats. B, HLI was significantly increased from 20 to 24 C and from 24 to 31 C in both the OVX + E₂ and the OVX rats, reflecting increased skin vasodilatation at higher ambient temperatures. At all ambient temperatures, the OVX rats exhibited a significantly higher HLI than the OVX + E₂ rats (n = 22–23 rats/group). Values represent means ± sem. *, Significant difference between OVX + E₂ and OVX rats; #, significant effect of T_AMBIENT in OVX rats; +, significant effect of T_AMBIENT in OVX + E₂ rats.

In both OVX + E₂ and OVX rats, the HLI, a measure of tail skin vasomotion, was significantly increased from 20 to 24 C and from 24 to 31 C. At all three ambient temperatures, the HLI was significantly elevated in the OVX vs. OVX + E₂ rats (Fig. 2B).

The number of Fos-ir cells was significantly increased in the MnPO at the warm T_AMBIENT of 31 C and increased in OVX, compared with the OVX + E₂-treated rats at 24 C

Of 14 examined brain areas, the MnPO was the only area that exhibited significantly more Fos-ir cells in rats exposed to the T_AMBIENT of 31 C (Figs. 3 and 4). The number of MnPO Fos-ir cells was the most elevated at the highest T_AMBIENT in both OVX + E₂ and OVX rats. At the ambient temperature of 24 C, OVX rats had significantly greater numbers of Fos-ir cells in the MnPO, compared with OVX + E₂ rats. A similar elevation of Fos immunoreactivity in OVX animals was also observed at the T_AMBIENT of 20 C, but this effect did not reach significance (Fig. 4). Comparison of Figs. 4 and 2B shows that the number of MnPO Fos-ir cells and the HLI, reflecting skin vasomotion, were affected in a similar manner by both T_AMBIENT and E₂ treatment.

Fig. 3.

Photomicrographs of Fos immunoreactivity in the MnPO of OVX + E₂ rats exposed to the ambient temperatures of 24 C (A) and 31 C. (B). The MnPO was the only area (of 14 areas examined) that demonstrated significantly increased numbers of Fos-ir cells in response to the higher ambient temperature. 3V, Third ventricle; ac, anterior commissure. Scale bar in A, 50 μm and applies to both A and B.

Fig. 4.

A, Average number of Fos-ir cells per section in the MnPO in OVX + E ₂ and OVX rats exposed to ambient temperatures of 20, 24, and 31 C. Fos immunoreactivity was significantly increased in the MnPO of OVX + E ₂ rats at the T_AMBIENT of 31 C, compared with rats exposed to ambient temperatures of 20 and 24 C. At 24 C, Fos immunoreactivity was significantly elevated in the OVX rats, compared with the OVX + E₂ rats (n = 8 rats/group). Values represent means ± sem. B, Representative computer-assisted maps of Fos-ir cells in the MnPO (closed circles) and adjacent preoptic area (open circles) of OVX + E₂ (top panels) and OVX (bottom panels) rats at three ambient temperatures. This map is at the level of plate 34 from the Paxinos and Watson rat brain atlas (34). The boundaries of the MnPO (dotted lines) were drawn with the aid of adjacent Nissl-stained sections. 3V, Third ventricle; ac, anterior commissure. Scale bar in upper left, 100 μm for all maps. *, Significant difference between OVX + E₂ and OVX rats; +, significant effect of T_AMBIENT in OVX + E₂ rats.

The numbers of Fos-ir cells were significantly increased in the arcuate nucleus and AVPV at the low T_AMBIENT of 20 C, with inhibitory (arcuate nucleus) and stimulatory (AVPV) effects of E₂

In the arcuate nucleus, the number of Fos-ir cells was highest at the low T_AMBIENT of 20 C but only in the OVX group (Fig. 5). In E₂-treated rats, the number of Fos-ir cells in the arcuate nucleus was consistently low, regardless of the T_AMBIENT. At both 20 and 24 C, OVX rats exhibited significantly more Fos-ir cells in the arcuate nucleus than OVX + E₂ rats. The number of Fos-ir cells in the AVPV was also highest in rats exposed to the low T_AMBIENT of 20 C (Fig. 6). In contrast to both the MnPO and arcuate nucleus, the number of Fos-ir cells in the AVPV was highest in the OVX + E₂ treated rats. This effect was significant only in rats exposed to the low T_AMBIENT of 20 C.

Fig. 5.

A, Average numbers of Fos-ir cells per section in the arcuate nucleus in OVX + E₂ and OVX rats exposed to ambient temperatures of 20, 24, and 31 C. Increased numbers of Fos-ir cells were detected at the T_AMBIENT of 20 C in the arcuate nucleus of OVX rats, compared with OVX rats at 31 C. E₂ significantly reduced Fos-ir in the arcuate nucleus at the ambient temperatures of 20 and 24 C. Values represent means ± sem (n = 6–8 rats/group). B, Representative computer-assisted maps of Fos-ir cells in the arcuate nucleus (closed circles) and adjacent ventromedial nucleus (open circles) of OVX + E2 (top panels) and OVX (bottom panels) rats at three ambient temperatures. This map is at the level of plate 57 from the Paxinos and Watson rat brain atlas (34). The boundary of the arcuate nucleus (dotted line) was drawn with the aid of adjacent Nissl-stained sections. 3V, Third ventricle; ARC, arcuate nucleus. Scale bar in upper left, 100 μm for all maps. *, Significant difference between OVX and OVX + E₂ rats; #, significant effect of T_AMBIENT in OVX rats.

Fig. 6.

A, Average numbers of Fos-ir cells per section in the AVPV in OVX + E₂ and OVX rats exposed to ambient temperatures of 20, 24, and 31 C. E₂ treatment increased the number of Fos-ir cells in rats exposed to the T_AMBIENT of 20 C in the AVPV (n = 6–8 rats/group). Values represent means ± sem. B, Representative computer-assisted maps of Fos-ir cells in the AVPV (closed circles) and adjacent medial preoptic area (open circles) of OVX + E₂ (top panels) and OVX (bottom panels) rats at three ambient temperatures. This map is at the level of plate 32 from the Paxinos and Watson rat brain atlas (34). The boundary of the AVPV (dotted line) was drawn with the aid of adjacent Nissl-stained sections. 3V, Third ventricle. Scale bar in upper left, 100 μm for all maps. *, Significant difference between OVX and OVX + E₂ rats; +, significant effect of T_AMBIENT in OVX + E₂ rats.

Multiple areas exhibited increased numbers of Fos-ir cells at the T_AMBIENT of 20 C, without significant overall effects of E₂, and other areas exhibited no effect of either T_AMBIENT or E₂

The dorsomedial nucleus, lateral preoptic area, medial preoptic nucleus, medial septal nucleus, and parastrial nucleus exhibited significantly increased numbers of Fos-ir cells at the lowest T_AMBIENT of 20 C, compared with the highest T_AMBIENT of 31 C (Table 1). However, because two-way ANOVA revealed no overall effect of E₂ in these areas, post hoc tests for estrogen effects were not performed.

Table 1.

Number of Fos-ir cells in select brain areas in OVX and OVX + E₂ rats exposed to mild changes in T_AMBIENT

	Brain area	20 C		24 C		31 C
		OVX + E2	OVX	OVX + E2	OVX	OVX + E2	OVX
Significant increase in Fos-ir at 20 C (compared with 31 C), no effect of E2	Dorsomedial nucleus	5.0 ± 0.9	9.8 ± 1.8	4.1 ± 0.8	3.2 ± 0.7	3.2 ± 1.2	4.4 ± 1.4
	Lateral preoptic area	8.3 ± 1.9	11.7 ± 1.5	8.1 ± 1.9	6.5 ± 1.2	5.0 ± 0.6	5.3 ± 0.9
	Medial preoptic nucleus	7.4 ± 0.9	7.3 ± 1.0	6.7 ± 1.6	7.3 ± 2.0	4.4 ± 0.8	3.9 ± 0.6
	Medial septal nucleus	28.6 ± 3.8	30.2 ± 7.0	17.8 ± 1.8	23.8 ± 8.1	15.2 ± 3.5	14.2 ± 3.3
	Parastrial nucleus	7.1 ± 1.1	13.8 ± 2.1	9.9 ± 2.3	9.2 ± 2.7	6.6 ± 0.7	5.9 ± 1.1
No significant effect of TAMBIENT or E2	Medial preoptic area	26.7 ± 3.6	29.7 ± 4.5	19.1 ± 3.7	25.1 ± 6.0	21.4 ± 2.8	20.5 ± 2.8
	Septohypothalamic nucleus	15.6 ± 2.7	16.4 ± 4.2	11.7 ± 3.1	13.4 ± 4.2	12.3 ± 2.6	8.6 ± 1.4
	Ventrolateral preoptic area	3.0 ± 0.7	2.9 ± 0.3	1.7 ± 0.2	2.3 ± 0.6	2.5 ± 0.8	3.3 ± 0.6
	Ventromedial nucleus	3.0 ± 1.7	4.7 ± 0.7	3.8 ± 1.2	2.1 ± 0.5	1.9 ± 0.7	3.2 ± 1.0

Numbers represent mean ± sem (n = 5–8 rats/group).

There was no effect of either T_AMBIENT or E₂ on Fos immunoreactivity in the medial preoptic area, septohypothalamic nucleus, ventrolateral preoptic nucleus, and ventromedial nucleus (Table 1). Because of the low numbers of For-ir cells detected in the magnocellular and parvocellular regions of the paraventricular nucleus (one or less Fos-ir cell per section), statistical analysis was not performed on these areas.

Discussion

The reproductive axis modifies mammalian thermoregulation, as shown by the changes in body temperature during the menstrual cycle and pregnancy (37, 38), and by the occurrence of hot flushes in humans in response to gonadal hormone withdrawal (2, 3). These phylogenetically old, homeostatic systems are regulated at the level of the hypothalamus, but little is known about the anatomic sites of integration. To address this issue, we used Fos immunohistochemistry to examine the response of brain areas to mild changes in environmental temperature and to determine whether these responses are modified by E₂ treatment. Because the OVX rats without estrogen replacement exhibit increased activation of skin vasodilatation and an increased sensitivity to warm ambient temperatures (11, 39), we were particularly interested in the effect of E₂ on warm-responsive hypothalamic nuclei.

Mild ambient temperature stimuli were selected to avoid the confounding effects of stress and to minimize changes in core body temperature that could directly stimulate thermosensitive hypothalamic neurons. The selection of these ambient temperatures was based on our previous study documenting a shift in the thermoneutral zone from 26.1 to 31.9 C in OVX + E₂ rats to 21.9 to 29.7 C in OVX rats (11). Thus, in both groups, the T_AMBIENT of 20 C was below the thermoneutral zone, and 31 C was near the upper limit of thermoneutrality. The intermediate T_AMBIENT (24 C) was neutral for OVX rats but subneutral for the OVX + E₂ rats (11). In the present study, additional measurements of tail skin and core temperature were taken to confirm the effects of E₂ at these ambient temperatures. As expected, the heat loss mechanism of skin vasodilatation was increased at the higher ambient temperatures and decreased by E₂ treatment of OVX animals (11, 13). Moreover, at the high T_AMBIENT of 31 C, the average T_CORE was increased by approximately 0.3 C in the OVX but not in the OVX + E₂ rats. These data agree with the previous demonstration that E₂ replacement allows OVX rats to maintain their core temperatures closer to normothermic levels when exposed to a warm environment (11, 39).

Remarkably, the MnPO was the only site demonstrating increased numbers of Fos-ir cells at 31 C, compared with 20 and 24 C. The selectivity of this response diverges from previous studies in which male rats, when exposed to high ambient temperatures, exhibited increased Fos immunoreactivity in numerous areas in addition to the MnPO (25–27). This discrepancy could be a gender effect, but it is more likely due to the higher ambient temperatures used in the previous studies, which ranged from 33 to 47 C. Similarly, the lowest ambient temperature used in this study, although subneutral, was very mild compared with the cold environments (4–10 C) used to elicit Fos immunoreactivity in OVX female rats (40) and male rats (24, 26, 27, 41). Changes in core body temperature, increased activation of peripheral thermosensors, and/or recruitment of additional thermoeffectors or other physiological systems might contribute to the higher levels of Fos activation observed in the previous studies.

The MnPO contains warm-sensitive neurons (42), but the increased Fos immunoreactivity in the MnPO at 31 C in E₂-treated rats cannot be explained by higher body temperatures because the T_CORE of these animals was not significantly elevated. We also considered the possibility that activity in the MnPO could be modulated indirectly by other physiological circuits. MnPO neurons have been implicated in the control of osmotic balance (43), but the low Fos immunoreactivity in the magnocellular paraventricular nucleus (one or less Fos-ir cell per section) indicates that changes in this factor are unlikely the cause of increased Fos immunoreactivity in the MnPO at the highest ambient temperature. Changes in osmotic balance are also unlikely to explain the Fos activation at 31 C because the rats had free access to water and food. Alternatively, Fos immunoreactivity in the MnPO can be altered by sleep (44, 45), and sleep patterns are affected by both ambient temperature and estrogen status (46–50). However, neither ambient temperature nor estradiol treatment altered Fos immunoreactivity in the ventrolateral preoptic nucleus, an area that exhibits robust sleep-dependent Fos immunoreactivity (45). More importantly, previous studies have reported that baseline sleep patterns during the light phase in rats are not significantly altered by estrogen treatment (50) or by mild increases in ambient temperature (23–30 C) (51). These considerations suggest that sleep patterns in this study did not differ substantially across treatment groups and thus were unlikely to cause the observed effects in the MnPO. In contrast, the changes in ambient temperature would have been detected by warm-sensitive skin thermoreceptors that project via brain stem relays to the MnPO (52). Thus, it seems likely that the increase in MnPO Fos-ir at 31 C was due to changes in ambient temperature and not secondary to alterations in other homeostatic systems.

Our findings are consistent with recent studies implicating the MnPO as an important area for the control of body temperature through the dilatation and constriction of skin blood vessels (27, 41, 52–55). The MnPO is part of the neuronal circuitry regulating heat-defense responses (52), and it is the first site in the preoptic area to receive peripheral temperature information (52, 54). Modulation of MnPO neurons by local injections of pharmacological agents changes the activity of sympathetic fibers controlling tail vasomotion (55). The MnPO may influence skin vasomotion through projections to other regions of the hypothalamus (19, 52, 54) or to the rostral medullary raphe in which sympathetic vasoconstrictor neurons control tail vasomotion (41, 56, 57). In a separate proposed pathway, MnPO neurons activated by high ambient temperatures were shown to project to the rostral periaqueductal gray (27), a midbrain site that may convey signals for thermoregulatory vasodilatation (58, 59). In addition to its role in thermoregulatory vasomotion, the MnPO is implicated in the facilitation of heat-escape behavior during osmotic stimulation (60) and brown adipose tissue thermogenesis (19, 54, 61).

Given the important role of MnPO neurons in thermoregulatory vasomotion, our results suggest that this nucleus could mediate the effects of E₂ on skin vasomotion in OVX rats (10, 11, 13–16, 28). Of the 14 brain regions examined, the MnPO was the only area that increased Fos immunoreactivity in response to the high T_AMBIENT of 31 C. Importantly, the effects of T_AMBIENT and E₂ on MnPO Fos immunoreactivity occurred in parallel to their effects on tail skin vasomotion. Moreover, at the T_AMBIENT of 24 C, Fos immunoreactivity was significantly elevated in the MnPO of OVX rats, compared with OVX + E₂ rats. Using similar experimental conditions, we previously observed that 24 C was above the ambient temperature threshold for skin vasodilatation in OVX rats but below the threshold (subneutral) in OVX + E₂ rats (11). Thus, differential activation of MnPO neurons could underlie the lower ambient threshold for skin vasodilatation in the OVX rats.

In addition to the MnPO, the arcuate nucleus and AVPV were significantly affected by T_AMBIENT and E₂. Both areas exhibited increased Fos immunoreactivity at the lowest T_AMBIENT, with E₂ treatment having opposite effects. In rats exposed to 20 C, E₂ reduced Fos immunoreactivity in the arcuate nucleus and increased Fos immunoreactivity in the AVPV. These effects on Fos immunoreactivity might reflect a role of these areas in thermoregulation. Electrical stimulation of the medial basal hypothalamus (primarily the arcuate and ventromedial nuclei) influences the activity of thermosensitive neurons in the preoptic area (62). Pharmacological inhibition of the AVPV influences the activity of sympathetic nerve fibers supplying tail skin blood vessels (55). The arcuate nucleus and AVPV project to thermoregulatory areas in the preoptic/septal regions (63–65) including the MnPO (66). However, it is also possible that the Fos-ir changes in the arcuate nucleus and AVPV reflect the effects of ambient temperature on other physiological circuits. For example, the arcuate nucleus and the AVPV are key areas in the sex steroid regulation of reproduction (67–69) and cold environments interact with photoperiod to affect the reproductive axis (70). Alternatively, the changes in the number of Fos-ir cells in the arcuate nucleus could reflect activity in energy balance circuits. Food intake is increased during cold acclimation (71), decreased during warm acclimation (72), and decreased by systemic or hypothalamic E₂ treatment (73). Further studies are necessary to understand the physiological significance of the T_AMBIENT and E₂ modulation of Fos in the arcuate nucleus and AVPV.

In summary, Fos immunoreactivity was modified by both mild changes in T_AMBIENT and E₂ in three areas: the MnPO, arcuate nucleus, and AVPV. The arcuate nucleus has a well-established role in energy balance, and both the arcuate nucleus and the AVPV are key components of the reproductive neuroendocrine axis. In contrast, the MnPO has a demonstrated role in the thermoregulatory axis, receiving thermal information as a part of a heat-defense pathway (52). There was a close correspondence between MnPO activation and tail skin vasodilatation and an inhibitory effect of E₂ on both parameters. Based on these data, we hypothesize that the MnPO plays an important role in the E₂, modulation of thermoregulatory vasomotion.