Leptin differentially increases sympathetic nerve activity and its baroreflex regulation in female rats: role of oestrogen

Abstract

Key points

• Leptin increases sympathetic nerve activity (SNA) in males, which contributes to obesity-induced hypertension; however, whether leptin is equally effective in females is unknown.

• We report that leptin does increase SNA and heart rate in female rats; however, for lumbar and renal SNA, this action is only evident in pro-oestrus and in oestrogen-treated ovariectomized rats, but not in ovariectomized or dioestrus rats.

• Leptin increases SNA and heart rate similarly in male and pro-oestrus female rats; however, leptin increases arterial pressure only in males.

• Blockade of MC3/4 receptors in the paraventricular nucleus (PVN) with SHU9119 decreases SNA in leptin-treated pro-oestrus rats, suggesting that leptin increases SNA in part by increasing α-melanocyte-stimulating hormone drive of PVN presympathetic neurons.

• Our data establish sex differences in leptin's effects to increase SNA and arterial pressure, which emphasizes the need for enhanced recognition and investigation of sex differences in obesity-induced sympathoexcitation and hypertension.

Abstract

Obesity and hypertension are commonly associated, and activation of the sympathetic nervous system is considered to be a major contributor, at least in part due to the central actions of leptin. However, while leptin increases sympathetic nerve activity (SNA) in males, whether leptin is equally effective in females is unknown. Here, we show that intracerebroventricular (i.c.v.) leptin increases lumbar (LSNA) and renal (RSNA) SNA and baroreflex control of LSNA and RSNA in α-chloralose anaesthetized female rats, but only during pro-oestrus. In contrast, i.c.v. leptin increased basal and baroreflex control of splanchnic SNA (SSNA) and heart rate (HR) in rats in both the pro-oestrus and dioestrus states. The effects of leptin on basal LSNA, RSNA, SSNA and HR were similar in males and pro-oestrus females; however, i.c.v. leptin increased mean arterial pressure (MAP) only in males. Leptin did not alter LSNA or HR in ovariectomized rats, but its effects were normalized with 4 days of oestrogen treatment. Bilateral nanoinjection of SHU9119 into the paraventricular nucleus of the hypothalamus (PVN), to block α-melanocyte-stimulating hormone (α-MSH) type 3 and 4 receptors, decreased LSNA in leptin-treated pro-oestrus but not dioestrus rats. Unlike leptin, i.c.v. insulin infusion increased basal and baroreflex control of LSNA and HR similarly in pro-oestrus and dioestrus rats; these responses did not differ from those in male rats. We conclude that, in female rats, leptin's stimulatory effects on SNA are differentially enhanced by oestrogen, at least in part via an increase in α-MSH activity in the PVN. These data further suggest that the actions of leptin and insulin to increase the activity of various sympathetic nerves occur via different neuronal pathways or cellular mechanisms. These results may explain the poor correlation in females of SNA with adiposity, or of MAP with leptin.

Introduction

Obesity and hypertension are commonly associated, and activation of the sympathetic nervous system is considered to be a major contributor (Esler et al. 2006; Hall et al. 2010). However, two studies in humans revealed that while the basal level of muscle sympathetic nerve activity (MSNA) was directly correlated with indices of obesity in men, such an association was absent in women (Lambert et al. 2007; Tank et al. 2008).

Leptin, which is released from adipose tissue, has been identified as a significant mediator of obesity-induced sympathoexcitation (Hall et al. 2010; Greenfield, 2011), at least in males. However, whether leptin increases SNA in females has not been investigated. Thus, one explanation for the failure of MSNA to track with adiposity in female subjects may be that adipose-derived leptin does not increase SNA. In support, arterial blood pressure (AP) better correlates with leptin in men compared to women (Mallamaci et al. 2000; Allison et al. 2013). On the other hand, in females, the ability of leptin to increase SNA may depend on the level of oestrogen, which fluctuates cyclically and could weaken the overall correlation between leptin and SNA or AP. Indeed, abundant evidence suggests that oestrogen may amplify the effects of leptin. First, leptin and oestrogen receptors coexist in neurons in hypothalamic nuclei at which leptin increases SNA, such as the arcuate nucleus (ArcN) (Diano et al. 1998). Second, leptin and oestrogen share common signalling pathways (Gao & Horvath, 2008; Kelly & Qiu, 2010). Third, oestrogen amplifies leptin's anorexic effect (Clegg et al. 2006). Therefore, one aim of the present study was to test the hypothesis that the ability of leptin to increase SNA in females is enhanced by elevated oestrogen levels. To test this hypothesis, we first determined if the actions of leptin are greater in rats during pro-oestrus, when oestrogen levels are elevated, compared to dioestrus. We also determined if oestrogen replacement of ovariectomized (OVX) rats enhances leptin responses. A key feature of our experimental design was to test the effects of leptin on sympathetic nerves innervating several organs, including the lumbar (LSNA), renal (RSNA), and splanchnic (SSNA), since our prior work in males revealed that leptin's actions differed among these sympathetic nerves (Li et al. 2013). While oestrogen may also interact with hypothalamic insulin signalling pathways (Gonzalez et al. 2008; Pratchayasakul et al. 2014), in contrast to leptin, oestrogen does not enhance the effects of insulin to inhibit food intake (Clegg et al. 2006). Therefore, we also investigated if insulin's sympathoexcitatory effect varies with the oestrous cycle.

One site of interaction between oestrogen and leptin may be pro-opiomelanocortin (POMC) neurons in the ArcN. Indeed, arcuate POMC mRNA fluctuates with the rat oestrous cycle (Wise et al. 1990; Bohler et al. 1991). Moreover, both leptin and oestrogen excite POMC neurons (Kelly & Qiu, 2010). Finally, it is well established that a POMC product, α-melanocyte-stimulating hormone (α-MSH), activates pre-sympathetic neurons in the paraventricular nucleus of the hypothalamus (PVN) to increase SNA via binding to MC4 receptors (Zhang & Felder, 2004; Hall et al. 2010; Ye & Li, 2011). Therefore, our second aim was to test the hypothesis that the enhanced sympathoexcitatory response to leptin during pro-oestrus is mediated by increased α-MSH drive of PVN pre-sympathetic neurons. To test this hypothesis, we determined if blockade of PVN MC3/4 receptors (MC3/4R) following nanoinjection of SHU9119 decreases the sympathoexcitatory effect of leptin in rats more during pro-oestrus compared to dioestrus.

Methods

Animals and ethical approval

Experiments were performed using 70 female and 5 male Sprague-Dawley rats (13–17 weeks, Charles River Laboratories, Inc., Raleigh, NC, USA). All of the rats were acclimatized for ≥1 week before experimentation in a room with a 12/12 h light/dark cycle, with food (LabDiet 5001, Richmond, IN, USA) and water provided ad libitum. All procedures were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and approved by the Institutional (Oregon Health & Science University) Animal Care and Use Committee.

Experimental preparation

Vaginal epithelium cytology was examined daily; at least two 4–5 day cycles were determined to establish the oestrous cycle. Anaesthesia was induced and maintained with 2–5% isoflurane in 100% oxygen. Body temperature was maintained at 37 ± 1°C using a rectal thermistor and heating pad. A tracheal tube, femoral arterial (1) and venous (3) catheters, and stainless steel electrodes around the lumbar, splanchnic or renal sympathetic nerve were implanted, as previously described (Li et al. 2013; Cassaglia et al. 2014). The rat was then placed in a stereotaxic instrument (David Kopf Instruments, Tujunga, CA, USA), and, following a midline incision, the skull was prepared for insertion of a lateral intracerebroventricular (i.c.v.) cannula (for leptin, insulin or artificial cerebrospinal fluid (aCSF) infusions) or for PVN nanoinjections by burring a hole in the skull near the midline. After completion of surgery, isoflurane anaesthesia was slowly withdrawn over 30 min, and a continuous intravenous infusion of α-chloralose (50 mg kg⁻¹ loading dose over 30 min; 25 mg kg⁻¹ h⁻¹ maintenance dose; Sigma-Aldrich, St. Louis, MO, USA) was initiated and continued for the duration of the experiment. Throughout the experiment, artificial ventilation with 100% oxygen was maintained, and respiratory rate and tidal volume were adjusted to maintain expired CO₂ at 3.5–4.5%. Anaesthetic depth was regularly confirmed by the lack of a pressor response to a foot or tail pinch; if necessary, additional α-chloralose was administered i.v. After completion of surgery and the α-chloralose loading dose, rats were allowed to stabilize for ≥60 min before experimentation. After experimentation, in some rats, blood samples were withdrawn from the artery catheter for measurement of oestrogen levels. All rats were killed with an overdose of pentobarbital, and in some rats the uteri were harvested and weighed.

Data acquisition

Throughout the experiment, pulsatile arterial pressure (AP), mean arterial pressure (MAP) and heart rate (HR) were continuously recorded with Grass amplifiers (Model 79D, Grass Instrument Co., Quincy, MA, USA). Raw SNA, AP, MAP and HR were collected using a Biopac MP100 data acquisition and analysis system (Biopac Systems, Inc., Santa Barbara, CA, USA), sampling at 2000 Hz. SNA was band-pass filtered (100–3000 Hz) and amplified (×10,000). After data collection, post-mortem SNA was quantified and subtracted from values of SNA recorded during the experiment. The SNA signal was then rectified, integrated in 1 s bins and for the figures was normalized to basal values (% of control).

Measurement of baroreflex function

Baroreflex function was assessed as previously described (Cassaglia et al. 2014). Briefly, complete sigmoidal baroreflex curves were generated by first quickly lowering MAP to ∼50 mmHg by i.v. infusion of nitroprusside (1 mg ml⁻¹; 20 μl min⁻¹), and then by steadily and smoothly raising MAP to ∼175 mmHg over 3–5 min by both withdrawing the nitroprusside infusion and infusing phenylephrine at increasing rates (1 mg ml⁻¹; 1–35 μl min⁻¹). Baroreflex curves relating the changes in SNA and HR to the changes in MAP were constructed from data obtained during the MAP upswing from 50 to 175 mmHg. The sigmoidal baroreflex relationships between MAP and HR or SNA generated in each baroreflex test were fitted and compared using the Boltzmann equation: HR or SNA = (P1 − P2)/[1 + exp(MAP − P3)/P4] + P2. P1 is the maximum HR or SNA, P2 is the minimum HR or SNA, P3 is the MAP associated with the HR or SNA value midway between the maximal and minimal HR or SNA (BP50; denotes position of the curve on the x-axis), and P4 is the width, the coefficient used to calculate maximum baroreflex gain (BRG), –(P1 − P2)/(P4×4), which is an index of the maximum slope of the most linear part of the sigmoidal baroreflex curve. Absolute values of BRG and the mean ± SEM of the baroreflex maxima, minima and midpoints are depicted in the figures.

i.c.v. infusions

With a flat skull and using bregma as zero, single-barrelled glass pipettes drawn to a small tip were used for i.c.v. infusions. For Protocols 1 and 2, coordinates for positioning the i.c.v. cannulae were as follows (mm from bregma): 1.0 caudal, 1.4 lateral and 4.2 ventral. For Protocol 3, the i.c.v. cannulae were angled 20 deg in the anterior–posterior direction and positioned as follows (mm from bregma): 0.6 mm rostral, 1.4 lateral and 4.5 ventral. Leptin (R&D Systems, Minneapolis, MN, USA) and insulin (Eli Lilly, Indianapolis, IN, USA) were dissolved in aCSF, containing (in mmol l⁻¹): 128 NaCl, 2.6 KCl, 1.3 CaCl₂, 0.9 MgCl₂, 20 NaHCO₃, 1.3 Na₂HPO₄ and 2.0 dextrose (pH 7.4). At the end of the experiment, Alcian blue dye was infused i.c.v. using the same cannula, and correct pipette placement was confirmed via detection of the dye in the cerebroventricles post mortem (Li et al. 2013).

PVN nanoinjections

Nanoinjections into the PVN were conducted with a single-barrelled glass micropipette, as described in our previous study (Cassaglia et al. 2014). Briefly, with a flat skull and using bregma and the dorsal surface of the dura as zero, the micropipette (20–40 μm tip outer diameter) was positioned in the PVN using the following coordinates: 1.8–2.0 mm caudal, 0.5 mm lateral and 7.4–7.6 mm ventral. SHU9119 (Tocris Bioscience, Bristol, UK; 60 nl of 0.5 mm l⁻¹ in aCSF with 10% DMSO) or vehicle control (aCSF with 10% DMSO) was injected bilaterally into the PVN, with ∼2 min between injections, and each injection was conducted over approximately 5–10 s using a pressure injection system (Pressure System IIe, Toohey Company, Fairfield, NJ, USA). At the end of the experiment, ∼60 nl of 2.5% Alcian Blue was injected into the PVN using the same pipette and coordinates to identify injection sites using a standard anatomical atlas (Paxinos & Watson, 2007).

Experimental protocols

After stabilization, basal baroreflex function was established by producing ≥2 baroreflex curves, 30 min apart, with similar gains; the final control curve was used for data analysis. Thirty minutes later, one of the following protocols was performed.

Protocol 1. Do the sympathoexcitatory effects of leptin or insulin vary during the oestrous cycle?

After control measurements, baroreflex function was reassessed 1 and 2 h after i.c.v. administration of aCSF or leptin (3 μg in 3 μl, followed by 5 μg h⁻¹). This dose efficaciously increased SNA in previous dose–response studies in males (Li et al. 2013) and inhibits food intake in both males and females (Clegg et al. 2003). Insulin was infused i.c.v. in a separate group of rats at a dose (100 μU min⁻¹ at 0.6 μl min⁻¹) that has been shown to robustly increase SNA in both male and female rats (Muntzel et al. 1994;Pricher et al. 2008). These experiments were conducted in female rats on the first day of dioestrus or on pro-oestrus. To assess sex differences, we compared the leptin responses to previously reported responses of male rats receiving leptin i.c.v. (Li et al. 2013) and, for insulin, to responses of α-chloralose-anaesthetized male rats receiving the same dose of i.c.v. insulin in the present study.

Protocol 2. Does oestrogen enhance the effects of leptin?

Four days before experiments, animals were OVX bilaterally as described before (Goldman et al. 2009). Silastic capsules (10 mm; inner diameter, 0.62 inches; outer, 1.25 inches; Dow Corning Corp., Midland, MI, USA) containing 10% 17-β oestradiol (E2) and cholesterol (n = 4), or crystalline cholesterol (n = 5) were then implanted subcutaneously. We elected to examine responses to leptin only 4 days after OVX or OVX+E2 to avoid the complications of changes in body weight that can follow longer term treatments. Moreover, the consequences of OVX (and E2 treatment) can change with time (Goldman et al. 2009). After surgery, rats were provided with acetaminophen/codeine in the drinking water for 3 days. Body weight and food intake were monitored daily. On the day of the experiment, after instrumenting the rats for recordings of AP and LSNA, baseline measurements were made, leptin was injected i.c.v. as in Protocol 1, and basal measurements and baroreflex function were assessed 1 and 2 h later.

Protocol 3. Are the enhanced sympathoexcitatory effects of leptin during pro-oestrus mediated by increased α-MSH inputs to the PVN?

On the first day of dioestrus or on pro-oestrus, the rats were anaesthetized and instrumented for recordings of AP and LSNA. After stabilization, and control measurements were made, leptin or aCSF was infused i.c.v., as in Protocol 1. After 90 min, SHU9119 was nanoinjected bilaterally into the PVN. In the i.c.v. aCSF-treated group, the vehicle was also injected into the PVN at 1 h. Response values were the difference of the averages of 2 min bins for 10 min following SHU9119 from baseline values before injections.

Blood hormone analysis

In OVX rats and some intact rats, arterial blood was collected for analysis of serum E2 levels by radioimmunoassay (Goodman, 1978; Resko et al. 1980; Anglin & Brooks, 2003). After collection, serum was stored at –80°C until assayed by the OHSU Endocrine Technology and Support Core. Briefly, serum was first extracted with ether and separated on a 1.0 g Sephadex column. E2 levels were determined using an ultrasensitive antibody (sensitivity <5 pg). Inter- and intra-assay variations did not exceed 15%. All E2 values were determined in the same assay.

Statistical analysis

All data are presented as means ± SEM. Between-group differences were evaluated using a one-way or two-way ANOVA for repeated measures and the post hoc Newman–Keuls test, unless otherwise noted. P values < 0.05 were considered statistically significant.

Results

Variations in baseline measurements during the oestrous cycle

Serum E2 levels were elevated during pro-oestrus compared to dioestrus, and as previously reported in conscious rats (Goldman et al. 2009), baseline MAP was lower in pro-oestrus compared to dioestrus rats (Table1 and Figs 1 and 2). However, no differences in baseline measurements of HR, LSNA, SSNA or RSNA were observed within the cycle (Figs 1 and 2). Similarly, baseline baroreflex control of LSNA, SSNA and HR did not vary during the cycle (Fig. 3); however, as previously reported (Goldman et al. 2009), the gain, maximum and range (82 ± 9%, n = 5, dioestrus; 144 ± 15%, n = 5; unpaired t test, P < 0.05) of baroreflex control of RSNA were increased on pro-oestrus compared to dioestrus (Fig. 3).

Table 1.

Baseline measurements in dioestrus, pro-oestrus, OVX and OVX+E2 rats

	Dioestrus (n = 31)	Pro-oestrus (n = 30)	OVX (n = 5)	OVX+E2 (n = 4)
Body weight (g)	272 ± 3	270 ± 3	273 ± 12	254 ± 5
MAP (mmHg)	106.3 ± 1.7	98.9 ± 1.7†	112.8 ± 3.3	104.1 ± 2.9
HR (b.p.m.)	354.4 ± 6.4	350.0 ± 4.9	364.3 ± 7.7	345.4 ± 17.0
LSNA (μV s–1)	1.14 ± 0.17 (n = 22)	0.77 ± 0.13 (n = 22)	1.73 ± 0.49	0.98 ± 0.16
17β-Estradiol (pg ml–1)	34 ± 5* (n = 5)	82 ± 6† (n = 5)	21 ± 3*	59 ± 14†
Uterine weight (g)	0.48 ± 0.02* (n = 17)	0.88 ± 0.04* (n = 20)	0.24 ± 0.01*	0.70 ± 0.06*

The body weight, MAP, HR and LSNA data were culled from all measurements in this study. Plasma E2 was measured only in a subset of the rats.

^*P < 0.05 vs. all other groups

^†P < 0.05 vs. dioestrus and OVX rats.

Figure 1.

Baseline measurements in pro-oestrus and dioestrus rats

The left (dioestrus) and middle (pro-oestrus) panels illustrate representative recordings of arterial pressure (AP), LSNA, SSNA and RSNA. The right panel illustrates the mean ± SEM of baseline values culled from all dioestrus and pro-oestrus rats in this study. n values are given below bars. Values were compared using a t test. *P < 0.05 between groups.

Figure 2.

Leptin increases baseline values of MAP, HR and SNA: comparison between males and females during dioestrus and pro-oestrus

A, i.c.v. leptin increased MAP in male rats (n = 16), but not in female rats (pro-oestrus, n = 15; dioestrus, n = 16). B, basal HR in female rats is higher than in male rats. i.c.v. leptin increased HR similarly in male (n = 6) and female rats (pro-oestrus, n = 15; dioestrus, n = 16). C, i.c.v. leptin increased LSNA in male rats (n = 5), and in female rats in pro-oestrus (n = 7), but not in dioestrus females (n = 7). D, i.c.v. leptin increased RSNA in male rats (n = 5), and in female rats in pro-oestrus (n = 5), but not in female rats in dioestrus (n = 5). E, i.c.v. leptin increased SSNA in male (n = 6) and female rats (pro-oestrus, n = 4; dioestrus, n = 5). *P < 0.05 vs. control; †P < 0.05 vs. dioestrus; ^‡P < 0.05 vs. male. Note that MAP and HR values were combined from rats instrumented for measurements of LSNA, RSNA and SSNA. Data in male rats from Li et al. (2013).

Figure 3.

Leptin increases baroreflex control of LSNA and RSNA only during pro-oestrus, but leptin's effects on SSNA and HR are not influenced by the estrus cycle

The gain and maximum of baroreflex control of RSNA were elevated in pro-oestrus compared to dioestrus rats; baseline BP50 was reduced. However, no differences in baseline baroreflex control of LSNA, SSNA or HR were observed. Leptin enhances baroreflex control of LSNA (A, dioestrus, n = 7; B, pro-oestrus, n = 7) and RSNA (C, dioestrus, n = 5; D, pro-oestrus, n = 5) only during pro-oestrus. In contrast, the ability of leptin to increase baroreflex control of SSNA (E, dioestrus, n = 5; F, pro-oestrus, n = 4) and HR (G, dioestrus, n = 16; H, pro-oestrus, n = 15) did not vary significantly during the oestrous cycle. *P < 0.05 vs. control (CON); †P < 0.05 between groups.

The oestrous cycle differentially influences effects of i.c.v. leptin infusion on SNA and baroreflex function

i.c.v. aCSF infusion did not alter MAP, HR, LSNA or baroreflex control of LSNA or HR in four rats (pro-oestrus, n = 2; dioestrus, n = 2; Table2). While i.c.v. leptin did not alter MAP in females (Fig. 2), it enhanced SNA and its baroreflex regulation. However, the responses varied among the sympathetic nerves studied (Figs 2 and 3). The effects of leptin on LSNA and RSNA were similar: leptin increased baseline LSNA and RSNA and baroreflex control of LSNA and RSNA (gain, maximum and range (102 ± 17 to 198 ± 41%, LSNA; 144 ± 15 to 278 ± 24%, RSNA; P < 0.05)) in females during pro-oestrus, but not dioestrus. Conversely, leptin increased baseline SSNA and HR and enhanced baroreflex control of SSNA (by increasing gain, maximum and range (136 ± 25 to 280 ± 63%, dioestrus; 160 ± 31 to 367 ± 60%, pro-oestrus; P < 0.05)) and baroreflex control of HR (by shifting the curve upward without an increase in gain) in rats in both the dioestrus and the pro-oestrus state. Nevertheless, the increases in basal and baroreflex control of SSNA were produced more rapidly during pro-oestrus than during dioestrus.

Table 2.

Effects of i.c.v. aCSF on basal LSNA, MAP and HR (n = 4)

	Control	i.c.v. aCSF for 1 h	i.c.v. aCSF for 2 h
Basal LSNA (% of control)	100 ± 0	94 ± 5	93 ± 10
Basal MAP (mmHg)	104 ± 6	102 ± 7	103 ± 8
Basal HR (b.p.m.)	334 ± 8	346 ± 8	342 ± 7
LSNA max. (% of control)	196 ± 8	187 ± 4	206 ± 15
LSNA min. (% of control)	61 ± 8	62 ± 10	63 ± 10
LSNA BP50 (mmHg)	85 ± 5	87 ± 6	86 ± 9
LSNA width (mmHg)	10.8 ± 1.7	11.0 ± 2.9	12.1 ± 2.5
LSNA range (% of control)	135 ± 13	126 ± 11	144 ± 16
LSNA gain (% of control/mmHg)	3.4 ± 0.6	3.3 ± 0.6	3.3 ± 0.7
HR max (b.p.m.)	436 ± 6	442 ± 3	438 ± 8
HR min (b.p.m.)	305 ± 15	304 ± 17	301 ± 14
HR BP50 (mmHg)	120 ± 6	119 ± 6	124 ± 5
HR width (mmHg)	8.2 ± 0.5	8.3 ± 0.7	8.5 ± 0.4
HR range (b.p.m.)	132 ± 19	138 ± 18	137 ± 15
HR gain (b.p.m. mmHg–1)	4.0 ± 0.5	4.2 ± 0.3	4.0 ± 0.4

Comparison of baseline values and responses to leptin in females and males (data from Li et al. 2013)

Baseline HR was elevated in females compared to males; however, baseline MAP was not significantly different between sexes (Fig. 2). Leptin increased baseline LSNA, RSNA, SSNA and HR in pro-oestrus rats, and baseline SSNA and HR in dioestrus rats, similarly to responses in males (body weight, 410 ± 11 g, n = 16; Fig. 4). However, unlike in females, i.c.v. leptin increased MAP in male rats (Fig. 2).

Figure 4.

Leptin increases basal and baroreflex control of LSNA and HR in E2-treated OVX rats, but not in OVX rats

E2 treatment increased the baseline measurements of baroreflex gain of LSNA and HR (P < 0.05, t test). i.c.v. leptin increased basal LSNA (A) and HR (B) in E2-treated OVX rats, not in OVX rats. i.c.v. leptin did not alter baroreflex control of LSNA (C) or HR (D) in OVX rats (n = 5), but increased baroreflex control of LSNA (C) and HR (D) in E2-treated OVX rats (n = 4). *P < 0.05 vs. control (CON); †P < 0.05 between groups.

E2 treatment of OVX rats normalizes the effects of leptin

No difference in body weight gain following surgery was observed between the OVX and OVX+E2 groups (3.4 ± 2.9 and −1.0 ± 2.4 g, respectively). Similarly, overnight food intake did not differ between the two groups (16.7 ± 0.3 g, OVX; 16.1 ± 1.1 g, OVX+E2). However, OVX reduced uterine weight and lowered plasma E2 levels (Table1). E2 treatment of OVX rats elevated serum E2 levels similarly to pro-oestrus rats and prevented the decrease in uterine weight (Table1). The baseline values of LSNA, MAP and HR were not affected by OVX or OVX+E2 implantation (Table1). E2 treatment of OVX rats increased baseline gain of baroreflex control of LSNA and HR, without significantly affecting other baroreflex parameters (Fig. 4).

i.c.v. leptin administration did not alter basal LSNA or LSNA baroreflex curves in OVX rats; however, leptin increased basal LSNA, and LSNA reflex gain, maximum and range (152 ± 26 to 262 ± 54%, P < 0.05) in OVX rats treated with E2 (Fig. 4). In these rats, leptin also increased the LSNA reflex minimum after 2 h. Surprisingly, leptin failed to affect basal HR or HR baroreflex curves in OVX rats. E2 supplementation restored the effect of leptin on basal HR and HR reflex maximum (Fig. 4). Leptin did not change MAP in OVX rats with or without E2 supplement (113 ± 3 to 108 ± 3 mmHg, OVX; 104 ± 3 to 117 ± 5 mmHg, OVX+E2; P > 0.05).

Leptin increases LSNA during pro-oestrus via activation of PVN MC3/4R

In pro-oestrus (n = 4) and dioestrus (n = 4) rats receiving i.c.v. aCSF, bilateral PVN nanoinjection of aCSF had no effects on LSNA (−1.6 ± 1.1% of control), MAP (−2.1 ± 1.1 mmHg) or HR (−0.6 ± 3.9 b.p.m.) within 10 min. i.c.v. aCSF did not affect LSNA, MAP or HR in either pro-oestrus or dioestrus rats (Table3); however, i.c.v. leptin increased LSNA and HR in pro-oestrus rats, but increased only HR in dioestrus rats (Fig. 5 and Table3). In i.c.v. aCSF-infused pro-oestrus or dioestrus animals, bilateral nanoinjection of SHU9119 failed to significantly alter MAP, LSNA or HR (Fig. 5 and Table3). However, in pro-oestrus rats receiving leptin i.c.v., PVN SHU9119 significantly decreased LSNA, without altering MAP or HR (Fig. 5 and Table3). Nevertheless, LSNA remained significantly greater than initial control values. PVN SHU9119 had no effects in leptin-treated rats in dioestrus (Fig. 5). The histologically identified sites of PVN nanoinjections are depicted in Fig. 5D.

Table 3.

Effects of PVN nanoinjection of SHU9119 on basal LSNA, MAP and HR in dioestrus and pro-oestrus rats, treated with i.c.v. aCSF or leptin

	LSNA (% Control)			MAP (mmHg)			HR (b.p.m.)
	Baseline	90 min	100 min	Baseline	90 min	100 min	Baseline	90 min	100 min
Pro-oestrus i.c.v. aCSF (n = 4)	100 ± 0	105 ± 7	104 ± 5	91 ± 3	100 ± 4	101 ± 2	333 ± 8	347 ± 9	347 ± 9
Dioestrus i.c.v. aCSF (n = 4)	100 ± 0	94 ± 8	96 ± 7	115 ± 6	112 ± 10	111 ± 11	395 ± 19	377 ± 16	376 ± 15
Pro-oestrus i.c.v. leptin (n = 4)	100 ± 0	179 ± 16*	155 ± 24*†	103 ± 7	99 ± 7	97 ± 8	355 ± 19	382 ± 12*	378 ± 9*
Dioestrus i.c.v. leptin (n = 4)	100 ± 0	115 ± 8	110 ± 6	99 ± 6	98 ± 6	97 ± 5	331 ± 13	360 ± 18*	359 ± 187*

Baseline, before i.c.v. aCSF or leptin infusion; 90 min, before the PVN nanoinjection of SHU9119 (after 90 min of i.c.v. aCSF or leptin infusion); 100 min, 10 min after the PVN nanoinjection of SHU9119.

^*P < 0.05 vs. baseline;

^†P < 0.05 vs. before nanoinjection of SHU9119 (or 90 min of i.c.v. infusion).

Figure 5.

Bilateral nanoinjection of SHU9119 into the PVN decreases LSNA in pro-oestrus rats treated with i.c.v. leptin

A, representative recording of raw integrated LSNA. SHU9119 was bilaterally injected into the PVN (depicted by arrows) after 90 min of i.c.v. leptin infusion. B, raw traces of LSNA, before i.c.v. leptin (open triangles), and before (grey triangles) and after (black triangles) PVN nanoinjection of SHU9119; triangles indicate times of raw traces in integrated LSNA (A). C, effects of PVN nanoinjection of SHU9119 (mean values per 2 min period) on LSNA. The injections were conducted after 90 min of i.c.v. aCSF or leptin infusion. Open circles, pro-oestrus rats treated with i.c.v. aCSF (n = 4); filled circles, pro-oestrus rats treated with i.c.v. leptin (n = 4); open triangles, dioestrus rats treated with i.c.v. aCSF (n = 4); filled triangles, dioestrus rats treated with i.c.v. leptin (n = 4). *P < 0.05 vs. time zero; †P < 0.05 vs. before nanoinjection of SHU9119 (or 90 min of i.c.v. infusion). D, histological diagrams illustrating nanoinjection sites of SHU9119 in the PVN. Open circles, pro-oestrus rats treated with i.c.v. aCSF; filled circles, pro-oestrus rats treated with i.c.v. leptin; open triangles, dioestrus rats treated with i.c.v. aCSF; filled triangles, dioestrus rats treated with i.c.v. leptin. Diagrams were modified from the brain atlas of Paxinos & Watson (2007).

The effects of insulin do not vary with the oestrous cycle

As shown in Figs 6 and 7, and as reported previously (Pricher et al. 2008; Cassaglia et al. 2011), i.c.v. insulin increases basal and baroreflex control of LSNA in females, by increasing gain, the maximum and the range (147 ± 14 to 296 ± 9%, dioestrus; 116 ± 16 to 243 ± 36%, pro-oestrus; P < 0.05). Basal and baroreflex control of HR (gain, maximum and range (109 ± 6 to 159 ± 7 b.p.m., dioestrus; 101 ± 12 to 149 ± 5 b.p.m., pro-oestrus; P < 0.05)) were also increased by insulin (Figs 6 and 7). However, these effects did not differ between the dioestrus and pro-oestrus states. i.c.v. insulin did not alter MAP in female rats, during either pro-oestrus or dioestrus (Fig. 6). In males (body weight: 440 ± 11 g, n = 5; Figs 6 and 7), insulin increased basal and baroreflex control of LSNA and HR similarly to females. However, unlike females, insulin increased MAP in males (Fig. 6).

Figure 6.

Insulin increases baseline MAP, LSNA and HR: comparisons between male rats and female rats during pro-oestrus and dioestrus

A, i.c.v. insulin increased MAP in male rats (n = 5), but not in female pro-oestrus (n = 5) or dioestrus rats (n = 5). However, i.c.v. insulin increased basal HR (B) and LSNA (C) similarly in male (n = 5), and in female pro-oestrus (n = 5) and dioestrus rats (n = 5). *P < 0.05 vs. control.

Figure 7.

Insulin increases baroreflex control of LSNA similarly in male and female dioestrus and pro-oestrus rats

i.c.v. insulin increased baroreflex control of LSNA (A, C, E) and HR (B, D, F) similarly in female rats in dioestrus (n = 5; A, B) and pro-oestrus (n = 5; C, D) and in male rats (n = 5; E, F). *P < 0.05 vs. control (CON).

Discussion

The purpose of this study was to test the hypothesis that the ability of leptin to increase SNA in females is enhanced by oestrogen's facilitatory effects on α-MSH excitatory inputs to the PVN. Our major new findings are: (1) leptin increases basal and baroreflex control of RSNA and LSNA only during pro-oestrus, when oestrogen levels are high; (2) the ability of leptin to increase basal and baroreflex control of SSNA and HR did not vary significantly during the oestrous cycle; (3) the increases in LSNA, RSNA, SSNA and HR evoked by i.c.v. leptin are similar in male and pro-oestrus female rats – however, leptin increases MAP only in males; (4) E2 treatment of OVX rats amplified the effects of i.c.v. leptin to increase LSNA (and HR) and its baroreflex regulation; (5) bilateral PVN nanoninjection of the MC3/4R inhibitor SHU9119 decreased LSNA only in pro-oestrus rats receiving i.c.v. leptin – PVN SHU9119 did not have significant effects in dioestrus rats or in pro-oestrus rats receiving the aCSF vehicle; and (6) insulin's sympathoexcitatory effects did not vary during the oestrous cycle, and did not differ between males and females. Collectively these data support our hypothesis that leptin's stimulatory effects on SNA are enhanced by oestrogen, at least in part via an action to increase α-MSH drive of PVN presympathetic neurons. Interestingly, however, the synergism between leptin and oestrogen was evident for RSNA and LSNA, but not SSNA; insulin also did not exhibit a similar interaction with gonadal steroids, at least in the control of LSNA. Therefore, these data further suggest that the actions of leptin and insulin to increase the activity of various sympathetic nerves occur via different neuronal pathways or cellular mechanisms.

Variations in SNA and baroreflex function during the reproductive cycle

In women, increases in basal MSNA have been detected during the midluteal phase, when gonadal steroids are elevated (Minson et al. 2000; Carter et al. 2013); increases in plasma norepinephrine levels have also been observed (Goldstein et al. 1983). In the present study of anaesthetized rats, baseline LSNA, RSNA, SSNA and HR were not significantly elevated during pro-oestrus. This finding does not necessarily refute an effect of the acute surge in gonadal steroids to influence basal SNA in rats. It has been proposed that the inhibitory effects of oestrogen on SNA is opposed by excitatory effects of progesterone in women (Carter et al. 2013); these opposing actions may be balanced in the rat on the late afternoon of pro-oestrus, when our experiments were conducted. Alternatively, the brief increases in oestrogen or progesterone during pro-oestrus in rats (compared to the luteal phase in women) may be insufficient to trigger sustained changes in SNA. Moreover, any changes in SNA may be subtle enough to elude detection by between-group comparisons in anaesthetized animals. Finally, while oestrogen alone has been shown to decrease SNA (He et al. 1999; Saleh & Connell, 2000; Vongpatanasin et al. 2001), this action may also be neutralized by the oestrogen-induced rise in leptin (and its synergism with oestrogen to increase SNA) during the hormonal surge in rats and humans (Shimizu et al. 1997; Fungfuang et al. 2013a,b).

In agreement with several previous studies (De Meersman et al. 1998; Mohamed et al. 1999; Saleh & Connell, 1999; Saleh et al. 2000; Pamidimukkala et al. 2003), we found that E2 treatment of OVX rats enhances gain of baroreflex control of LSNA and HR. In parallel, MSNA baroreflex gain in women (Minson et al. 2000), and RSNA gain in rats (Goldman et al. 2009), increases when gonadal steroids are elevated during the reproductive cycle. Our present results confirm the oestrous-cycle dependency of the RSNA baroreflex in the rat (Goldman et al. 2009); however, baroreflex control of LSNA, SSNA and HR did not differ between pro-oestrus and dioestrus. The failure to observe a change in the HR baroreflex may seem to conflict with previous studies showing enhanced spontaneous baroreflex gain in rats (Goldman et al. 2009) and women (Saeki et al. 1997; Tanaka et al. 2003) during pro-oestrus and the midluteal phase, respectively. However, spontaneous baroreflex gain reflects primarily parasympathetic control of the heart, which has been shown to be increased by oestrogen (Saleh & Connell, 2000), and which may be counteracted by an opposing change in sympathetic control of HR. In support of this, other studies have failed to detect a change in baroreflex control of HR during the menstrual cycle (Minson et al. 2000; Kim et al. 2012).

Leptin increases SNA and baroreflex function in females: role of oestrogen

A major finding was that i.c.v. leptin increases basal LSNA and RSNA and the gain and maximum of baroreflex control of LSNA and RSNA in females, as in males (Li et al. 2013); however, these effects were evident during pro-oestrus, but not dioestrus. On the other hand, the effect of i.c.v. leptin to increase basal and the gain and maximum of baroreflex control of SSNA did not vary significantly during the cycle. Because OVX rats were unresponsive to leptin and elevation of E2 to pro-oestrus levels in OVX rats restored sympathoexcitatory responses, we conclude that the oestrogen surge is probably a major component of the enhanced LSNA and RSNA responses to leptin during pro-oestrus.

While these data suggest that E2 amplifies the effects of leptin to increase SNA, as it does leptin's anorexic effect (Clegg et al. 2006), it is noteworthy that leptin failed to increase LSNA or RSNA in dioestrus females. Moreover, the responses of pro-oestrus females were similar to males, with lower plasma E2 levels. These data raise the following question: why are dioestrus females selectively resistant to leptin? One possible explanation is that while plasma E2 levels may be lower, local brain E2 levels are higher in males. In support, aromatase, which facilitates the conversion of testosterone to E2, is higher in the hypothalamus of males compared to females (Roselli et al. 1984, 1985). Moreover, E2 activates POMC neurons in males, and testosterone's effect to activate POMC neurons in males requires aromatization to E2 (Qiu et al. 2007). Alternatively, while leptin has been shown to act in several hypothalamic sites to increase SNA in males (Marsh et al. 2003; Montanaro et al. 2005; Rahmouni & Morgan, 2007), leptin may act in fewer sites in females. Both of these hypotheses remain to be tested. Regardless of the mechanism, however, the failure of dioestrus females to respond to leptin may partially explain the finding that MSNA correlates better to indices of obesity in men compared to women (Lambert et al. 2007; Tank et al. 2008).

As in males (Li et al. 2013), in female rats, leptin increased basal HR and right-shifted the HR baroreflex curve without enhancing baroreflex gain, but the actions were similar during dioestrus and pro-oestrus. Given the lack of variation during the cycle, it may seem surprising that OVX abolished, and E2 replacement restored, leptin's effects on HR. In males, leptin increases HR solely via the suppression of cardiac parasympathetic activity (Li et al. 2013). In both males and females, oestrogen increases parasympathetic tone (Saleh & Connell, 1999, 2000; Saleh et al. 2000), and OVX markedly suppresses cardiac parasympathetic tone (Saleh & Connell, 2000; Goldman et al. 2009). Therefore, we speculate that OVX abolishes the effects of leptin on HR, because the significant reduction in oestrogen levels reduces the responsive reserve of parasympathetic tone, which then blunts leptin's ability to further inhibit it.

When we compared the leptin and insulin responses of pro-oestrus females to males (Li et al. 2013), we found remarkable consistency, with one exception: i.c.v. leptin and insulin increased arterial pressure in males, but not females. This result is in line with findings from obese rats fed a cafeteria diet, showing that while body weight and leptin increased in both sexes, hypertension occurred only in males (Plut et al. 2002). This finding may also explain the failure of arterial pressure to correlate with plasma leptin levels in women, in contrast to men (Mallamaci et al. 2000; Allison et al. 2013). Studies in humans have shown that vascular β-adrenergic receptors counteract α-adrenergic vasoconstriction in young women but not young men or post-menopausal women (Kneale et al. 2000; Hart et al. 2011), suggesting that the sex difference may be secondary to reduced noradrenaline-induced vasoconstriction in females.

In contrast to leptin, the insulin-induced increases in basal and baroreflex control of LSNA were not different between rats in dioestrus versus pro-oestrus. Moreover, leptin's and insulin's effects on baroreflex control of HR differ: the action of insulin is mainly on cardiac sympathetic activity (Siani et al. 1990; Berkelaar et al. 2013), whereas leptin acts via vagal efferents (Li et al. 2013). Indeed, leptin and insulin transform the HR baroreflex curve differently: leptin increases the baroreflex maximum and minimum, but not gain, whereas insulin increases the maximum and gain, but not the minimum. These data, along with a report that leptin and insulin activate distinct populations of POMC neurons in the ArcN (Williams et al. 2010), suggest that insulin and leptin increase SNA via different neurons and pathways. Moreover, our data showing that the effects of leptin in both males and females vary among the sympathetic nerves studied, coupled with information that the sympathoexcitatory effects of leptin and insulin on various nerves are mediated via different signalling pathways (Rahmouni et al. 2004, 2009), suggest that these metabolic hormones influence unique and potentially separate populations of neurons in the control of SNA and HR.

Leptin increases SNA in females: role of α-MSH inputs into the PVN

The sympathoexcitatory and cardiovascular effects of leptin are dependent on the central melanocortin system (Dunbar & Lu, 1999; Haynes et al. 1999), through stimulatory actions of α-MSH on MC4R in the PVN (Zhang & Felder, 2004; Ye & Li, 2011). Moreover, both oestrogen and leptin activate POMC neurons (Gao & Horvath, 2008; Brown et al. 2010; Kelly & Qiu, 2010). Therefore, we tested the hypothesis that the enhanced responses to leptin during pro-oestrus (and following E2 treatment) are mediated by α-MSH inputs to PVN. In support of this hypothesis, we found that blockade of PVN MC3/4R, with bilateral nanoinjections of SHU9119, in pro-oestrus leptin-infused but not dioestrus rats, decreased LSNA, without significantly altering MAP or HR. PVN SHU9119 had no effects in pro-oestrus rats receiving i.c.v. aCSF, suggesting that the excitatory effects of the pro-oestrus surge of oestrogen are insufficient to significantly elevate MC3/4R drive of basal LSNA.

The present study did not identify the site at which leptin and oestrogen interact during pro-oestrus to increase PVN MC3/4R drive of SNA. POMC neurons are found largely in the ArcN, with a few cells also situated in the nucleus tractus solitarius (Ellacott & Cone, 2004; Mercer et al. 2013). Because ArcN POMC neurons send a major projection to the PVN and ArcN lesions eliminate α-MSH immunoreactivity in the PVN (O'Donohue & Jacobowitz, 1980; Ellacott & Cone, 2004; Mercer et al. 2013), and because infusion of the present dose of leptin into the fourth cerebroventricle does not increase SNA (Li et al. 2013), the ArcN is likely to be the principal if not exclusive source of α-MSH drive of PVN pre-sympathetic neurons. Therefore, E2 could augment the effects of leptin either by increasing α-MSH actions in PVN or by enhancing the effects of leptin in the ArcN. Considerable data support the latter possibility, including evidence that oestrogen and leptin receptors coexist within the same brain sites/neurons, that oestrogen increases (and OVX decreases) hypothalamic leptin receptors, that oestrogen mimics leptin's actions on a cellular/molecular level, and that leptin and oestrogen share signalling pathways, specifically in POMC neurons (Kimura et al. 2002; Meli et al. 2004; Gao & Horvath, 2008; Kelly & Qiu, 2010; He et al. 2012).

While PVN SHU9119 significantly decreased LSNA in pro-oestrus rats given leptin, LSNA levels were not normalized. The residual sympathoexcitation may result from a genomic effect of MC3/4R activation, which is not rapidly reversed. Alternatively, other excitatory inputs, for example glutamatergic, may support the remaining leptin-induced increase in SNA. The PVN receives glutamatergic innervation from many sites at which leptin increases SNA, including the ventromedial hypothalamus, dorsomedial hypothalamus, ArcN and lateral hypothalamus (Ulrich-Lai et al. 2011); however, whether PVN glutamate is involved in the sympathoexcitatory response to leptin is currently unknown.

In addition to the POMC innervation, the other major arcuate projection to PVN expresses neuropeptide Y (NPY) and agouti-related protein (AgRP). We have recently reported that NPY inputs into the PVN tonically inhibit PVN presympathetic neurons, which are also excited by α-MSH (Cassaglia et al. 2014). Oestrogen decreases hypothalamic NPY/AgRP expression and release in association with a sensitization of the anorexic effect of leptin (Ainslie et al. 2001; Olofsson et al. 2009; Dhillon & Belsham, 2011) and directly inhibits NPY neuronal excitability (Roepke et al. 2011; Smith et al. 2013). As NPY and α-MSH inputs converge at PVN presympathetic neurons (Cassaglia et al. 2014), another potential mechanism by which oestrogen could amplify the excitatory effects of leptin is by reducing tonic NPY inhibition. However, further experiments are required to test this hypothesis.

In conclusion, leptin does increase SNA in females; however, for LSNA and RSNA, this action is only evident during pro-oestrus, probably due to an oestrogen-driven amplification. Given that adipose-derived leptin may be a major contributor to obesity-induced increases in SNA and AP (Hall et al. 2010), at least in males, this result may explain the failure of MSNA to correlate to indices of obesity in women (Lambert et al. 2007; Tank et al. 2008). In addition, because leptin increases MAP in males but not females, these data may also provide insight as to why the correlation between leptin and AP is not strong in women, unlike in men (Mallamaci et al. 2000; Allison et al. 2013). Collectively, our data in rats in conjunction with these earlier studies in humans (Mallamaci et al. 2000; Lambert et al. 2007; Tank et al. 2008; Allison et al. 2013) emphasize the need for enhanced recognition and investigation of the sex differences in obesity-induced sympathoexcitation and hypertension.

Glossary

aCSF

artificial cerebrospinal fluid

AgRP

agouti-related protein

AP

arterial blood pressure

ArcN

arcuate nucleus

BRG

baroreflex gain

E2

17-β oestradiol

HR

heart rate

i.c.v.

intracerebroventricular

LSNA

lumbar sympathetic nerve activity

MAP

mean arterial pressure

α-MSH

α-melanocyte-stimulating hormone

MSNA

muscle sympathetic nerve activity

NPY

neuropeptide Y

OVX

ovariectomized

POMC

pro-opiomelanocortin

PVN

paraventricular nucleus of the hypothalamus

RSNA

renal sympathetic nerve activity

SNA

sympathetic nerve activity

SSNA

splanchnic sympathetic nerve activity

Additional Information

Competing interests

The authors have no competing interests to declare.

Author contributions

All authors have read and approved the final submission. 1. Conception and design of the experiments: Z.S. and V.L.B. 2. Collection, analysis and interpretation of the data: Z.S. and V.L.B. 3. Drafting or revising the article: Z.S. and V.L.B.

Funding

This work was supported in part by NIH HL088552, by an American Heart Association Grant-in-Aid 12GRNT11550018, and a grant from the Medical Research Foundation of Oregon. The OHSU Endocrine Technology and Support Core is supported by the Oregon National Primate Research Center Core Grant P51 OD011092.