Kappa-Opioid Receptor Blockade Ameliorates Obesity Caused by Estrogen Withdrawal via Promotion of Energy Expenditure through mTOR Pathway

Weight gain is a hallmark of decreased estradiol (E2) levels because of menopause or following surgical ovariectomy (OVX) at younger ages. Of note, this weight gain tends to be around the abdomen, which is frequently associated with impaired metabolic homeostasis and greater cardiovascular risk in both rodents and humans. However, the molecular underpinnings and the neuronal basis for these effects remain to be elucidated. The aim of this study is to elucidate whether the kappa-opioid receptor (k-OR) system is involved in mediating body weight changes associated with E2 withdrawal. Here, we document that body weight gain induced by OVX occurs, at least partially, in a k-OR dependent manner, by modulation of energy expenditure independently of food intake as assessed in Oprk1−/−global KO mice. These effects were also observed following central pharmacological blockade of the k-OR system using the k-OR-selective antagonist PF-04455242 in wild type mice, in which we also observed a decrease in OVX-induced weight gain associated with increased UCP1 positive immunostaining in brown adipose tissue (BAT) and browning of white adipose tissue (WAT). Remarkably, the hypothalamic mTOR pathway plays an important role in regulating weight gain and adiposity in OVX mice. These findings will help to define new therapies to manage metabolic disorders associated with low/null E2 levels based on the modulation of central k-OR signaling.


Introduction
Decreased levels of estradiol (E2) in postmenopausal women or in preclinical animal models are associated with hyperphagia, reduced energy expenditure, and weight gain [1][2][3][4][5], while E2 replacement therapy prevents weight gain and metabolic dysfunctions in postmenopausal women and ovariectomized (OVX) rodents by decreasing feeding and increasing energy expenditure [3,6,7]. Increased body mass during menopause in women is characterized by a marked increase in abdominal fat, which implies a much greater cardiovascular risk [8,9]. Data gleaned over the last decade have demonstrated that by a combination of central and peripheral effects, exerted via classical estrogen receptors (α and β) and the G-protein coupled estrogen receptor (GPR30), E2 can influence energy balance, peripheral tissue insulin sensitivity, inflammation, and cardiovascular risk [3,7,10,11]. In terms of energy balance and fat distribution, estrogen receptors (ERs) are widely expressed in the hypothalamus [12], and mice with a global or brain-specific targeted disruption of ER-alpha (ERα) are obese because of hyperphagia and hypometabolism [13,14]. Interestingly, E2 displays a nucleus-specific action within the hypothalamus to modulate different compartments of energy balance and metabolic homeostasis by acting in specific hypothalamic nuclei, such as the arcuate nucleus (ARC) and the ventromedial hypothalamus (VMH) [7,[14][15][16][17][18].
Similar to ER, it is known that the kappa-opioid receptor (k-OR) is widely expressed in the brain and that functional interaction between gonadal steroids and the opioid system exists [19,20], as shown by the fact that Kisspeptin/Neurokinin B/Dynorphin (KNDy) neurons located in the ARC mediate the E2-dependent decrease in body weight (BW) and increase in body temperature [21,22]. Further support of the relevant role of k-OR, in terms of energy balance, derives from the following: (i) genetic silencing of k-OR prevents high-fat diet-induced weight gain [23], (ii) k-OR -signaling is involved in ghrelin-and melanin-concentrating hormone (MCH)-induced food intake (FI) and in the anorexic effect of nicotine, as well as in its effects on energy expenditure (EE) [24][25][26], and (iii) k-OR signaling influences energy homeostasis during calorie restriction [27]. Hence, k-OR has been proposed as a promising drug target in obesity [28]. Of note, most of the studies done in relation to k-OR in energy and metabolic homeostasis were done in male animals, while their effects in females are largely unknown.
The fact that the biological effects exerted through k-OR, such as analgesia, nociception, learning, or addiction, are sex-specific and estrogen-dependent [29] make it critical to study whether k-OR is involved in mediating body weight changes associated with E2 withdrawal (e.g., in menopause or OVX), a topic that despite its relevance has not received enough attention. In this work, we disrupted k-OR signaling by genetic and pharmacological approaches to study its effects on OVX-induced weight gain. Our data show that genetic or pharmacological disruption of k-OR alleviated OVX-induced weight gain in a foodindependent manner. This effect was associated with increased UCP1 immunostaining in BAT and white adipose tissue (WAT). Finally, since phosphorylated forms of mTOR/p70S6K signaling were upregulated, we addressed whether mTOR pathways could be mediating the physiological effects observed. As a result, we identified a potential mechanism based on p70S6K activation for k-OR-dependent regulation of OVX-induced adiposity.

Genetic Deletion of k-OR Ameliorates OVX-Induced Weight Gain in Female Mice
In keeping with previous studies [30], OVX WT mice gained more weight compared to the SHAM group (p < 0.01) ( Figure 1A) in an FI-independent manner ( Figure 1A). Interestingly, the genetic disruption of k-OR alleviated OVX-induced weight gain in a food-independent manner ( Figure 1B). animals. Values are expressed as mean ± SEM (n = 8-12 per group). Two-way ANOVA (factors: OVX and time) was used to analyze BW gain evolution. Annotation indicates significant effect of a = OVX, b = time, c = interaction OVX x time. Unpaired t-test or Mann-Whitney test were used to compare the two groups (SHAM and OVX), indicating significant differences compared to SHAM p < 0.01 (**), p < 0.001 (***).
In OVX WT mice, the cumulative BW gain was accompanied by an increase in fat mass (p < 0.01) in detriment of lean mass (p < 0.001) ( Figure 1C), whereas this observation was absent in OVX Oprk1−/−mice ( Figure 1D). These results suggest that OVX-induced adiposity and weight gain are largely mediated by k-OR signaling.

k-OR Animals Respond Differently to E2 Withdrawal in Terms of Circulating Luteinizing Hormone (LH) and Lipids
The absence of estrogens is characterized by elevated serum luteinizing hormone (LH). As expected in WT, LH significantly increased after 11 weeks of OVX (p < 0.001) compared to control group (Table 1). Similar results were obtained when we measured glucose (p < 0.001) and cholesterol (p < 0.05). Of note, basal (SHAM group) triglyceride (TG) values were lower in mutant mice than WT suggesting that lipid metabolism is affected by the absence of k-OR (Table 1). Table 1. Serum physiological parameters in SHAM and OVX, and female mice at 11 weeks after surgery. Values are expressed as mean ± SEM. t-test or Mann-Whitney test were performed to evaluate differences between groups indicating significant differences compared to SHAM (*). p < 0.05 (*), p < 0.01 (**), p < 0.001 (***). LH, luteinizing hormone; TG, triglycerides.

Global Inhibition of k-OR Normalized the Reduced EE after OVX
To investigate the role of the k-OR system on the metabolic disturbances induced by OVX, we monitored WT and Oprk1−/−mice using indirect calorimetry chambers. We detected a reduced energy expenditure during the dark phase in WT OVX animals (p < 0.05) (Figure 2A,C). In contrast, no significant changes in energy expenditure were detected in the different experimental settings in mutant mice subjected to OVX ( Figure 2B,D). Finally, we observed that the locomotor activity was significantly reduced during the dark phase in both WT (p < 0.05) and Oprk1−/−mice (p < 0.01) after OVX ( Figure 2E,F), whereas no changes in respiratory exchange ratio (RER) were detected in WT or Oprk1−/−mice following OVX ( Figure 2G,H). Taken together, these data clearly indicate that k-OR-mediated effects in weight gain following OVX are associated with energy expenditure. . Unpaired t-test or Mann-Whitney test were used to compare the two groups (SHAM and OVX), indicating significant differences compared to SHAM (*). p < 0.05 (*), p < 0.01 (**).

The Role of k-OR System on BAT and WAT
Based on the hypothesis that the lack of response to E2 withdrawal-induced metabolic changes in the global KO mice are food-independent and may be explained by increased thermogenesis, we studied in more detail BAT function and browning in an OVX mice model. We measure rectal and interscapular BAT temperature in a 23 • C ( Figure S1A), cold (4 • C) (Figure S1B), and thermoneutral (30 • C) environment ( Figure 3). The reason for testing at different temperatures is that at environmental temperatures below thermoneutrality, possible effects of thermogenic agents can be masked by a compensatory decrease in thermoregulatory thermogenesis, i.e., the heat production occurring in order to defend the body temperature. We found increased BAT thermogenic capacity in OVX Oprk1−/−mice since BAT temperature was higher than in OVX WT (p < 0.05) ( Figure 3A,B) with a significant reduction in BAT mass (p < 0.05) ( Figure 3C,D). OVX Oprk1−/−mice also displayed elevated EE (p < 0.05) independently of LA ( Figure 3E,F) compared to OVX WT. According to the higher BAT temperature, uncoupling protein (UCP1) levels were increased in BAT of OVX Oprk1−/−mice compared to OVX WT (p < 0.05) ( Figure 3G). It is known that both activation of thermogenesis in BAT and the induction of beige/ brite adipocytes in the WAT, a process known as browning, have a relevant impact on total energy balance in both rodents and humans [31][32][33][34]. We checked the browning in gonadal (gWAT) and subcutaneous inguinal (siWAT) white adipose tissue ( Figure 4A) and measured fat depots after 11 weeks of OVX surgery ( Figure 4B). Compared to WT mice, mice lacking a k-OR system showed a clear tendency for greater browning activity, especially in siWAT (p = 0.06) ( Figure 4C,D) under estrogen deficiency conditions.

Central Pharmacological Inhibition of k-OR Recapitulates the Effect of Oprk1 Genetic Ablation
Next, we chronically administrated i.c.v. PF-04455242, a specific antagonist of k-OR [35], during one week in OVX WT and Oprk1−/−mice (to discard any unspecific effects through the other opioid receptors: mu and delta). Chronic i.c.v. infusion of antagonist in OVX WT mice significantly reduced BW (p < 0.05) ( Figure 5A), while no changes were observed in Oprk1−/−animals ( Figure 5B). Of note, the reduced BW observed was independent of FI ( Figure 5A,B). We noted a tendency to a reduction in siWAT depots (p < 0.07) after central blockade of the k-OR receptor ( Figure 5C). Central administration of the k-OR antagonist also significantly increased UCP1-positive immunostaining in BAT (p < 0.05) ( Figure 5E) and siWAT browning (p < 0.05) ( Figure 5G) in WT mice. As expected, these k-OR antagonist-induced effects were absent in Oprk1−/−mice ( Figure 5D,F,H). Non-statistically significant changes were observed in gWAT ( Figure S2).

mTOR Signaling Mediates the k-OR-Dependent Effects on Body Weight and Adiposity in an E2-Withdrawal Model
Mechanistically, we previously identified the mTOR pathway as a mediator of some of the central effects of estradiol in energy balance [18]. Thus, we studied whether it could also be involved in the k-OR-dependent mechanisms preventing OVX-induced adiposity. We found that after one week of i.c.v. administration of k-OR antagonist (PF-04455242) in WT OVX mice ( Figure 6A), the mTOR signaling pathway was activated in the medio-basal hypothalamic area (MBH) since phosphorylated levels of mTOR, p70S6K, and S60 were increased ( Figure 6B). Then, to evaluate the functional role of k-OR-mediated activation on thermogenesis, we expressed in the MBH a constitutively activated form of S6K (CAS6K) in OVX WT mice ( Figure 7A,B). The accuracy of the viral injections was corroborated by the expression of GFP ( Figure 7C). Under OVX condition, we observed a significant reduction of BW ( Figure 7D) and a decrease in WAT mass and adipocyte size in gWAT and siWAT ( Figure 7E,F), which was accompanied by elevated UCP1 levels, mainly in the gWAT (p < 0.01) ( Figure 7G).

Discussion
Compelling evidence has demonstrated that ovarian hormone depletion (e.g., in conditions of OVX or menopause) induces a marked increase in weight gain, visceral adiposity, and cardiovascular risk, while estrogen replacement improves these parameters. Yet, the molecular underpinnings and neuronal basis of those actions remain largely unknown. This work demonstrates for the first time that k-OR signaling plays a key role in mediating the OVX-induced impairment of energy homeostasis. In detail, Oprk1−/−mice were resistant to weight gain and adiposity induced by OVX. This effect was independent of FI and relied on a higher EE.
Several studies have demonstrated the putative connection between the opioid system via k-OR and BAT thermogenesis since mice lacking k-OR were resistant to diet-induced obesity by increasing EE [23]. Moreover, we have recently reported that k-OR mediates nicotine-induced increases in energy expenditure [26], whereas others have documented the role of k-OR in mediating the hypothermic response to calorie restriction [27]. In this context, our current data provide novel evidence for the important role of k-OR in alleviating OVXinduced weight gain through a food-intake-independent mechanism, which prompted us to study the possible involvement of BAT-activation and browning as potential mechanisms mediating the beneficial effect of both genetic silencing and pharmacological blockade of k-OR. The influence of k-OR on WAT browning was consistent in three of the models of k-OR inhibition presented here: global KO, central pharmacological blockade, and targeted constitutive activation of p70S6K in MBH. Therefore, the induction of browning in the WAT could explain, at least to some extent, the resistance of body weight gain induced by OVX. The data obtained regarding BAT is not as clear-cut in experiments carried out below thermoneutrality. Before ruling out the involvement of BAT, we decided further explore, under thermoneutral conditions, to uncover the effect of thermogenic agents that can be masked by a compensatory decrease in thermoregulatory thermogenesis to defend body temperature. Using this approach, we found an increase in BAT temperature associated with an increase in UCP1 levels. Further studies in the context of a Scholander-type experiment are needed to further clarify this issue, including measuring the temperature dependence of the effects of ovarian function deficiency and Oprk1-silencing on food intake; body-, BAT-, and tail-temperature; temperature preference; and energy expenditure.
Data gleaned over the last few years in rodents and humans have uncovered the existence of sex differences in k-OR-system expression and function. Of note, most of the available data regarding sex differences were published in relation to pain nociception and mood; meanwhile, there is a paucity of data exploring their effects in energy homeostasis in females-hence the relevance of the current report [29,36].
Although not addressed in our study, it is tenable that the link between E2 and k-OR signaling may be associated with estrogen-responsive genes, and further studies to elucidate this question would be worthy. In fact, it is known that E2 down-regulates the prodynorphin mRNA expression in the anterior pituitary of OVX rats [37].
Data gleaned by us and others in recent years provided evidence indicating that hypothalamic AMPK in SF1 neurons in the VMH play a key role in energy balance by influencing sympathetic nervous system (SNS) activation, which leads to BAT thermogenesis and browning of WAT. Although the relevance of this mechanism is beyond any doubt and, in fact, has been shown to mediate some central effects exerted by multiple key hormones on energy balance [38], there is evidence that additional mechanisms are also involved. We provide here evidence of an alternative k-OR-dependent regulation of OVX-induced weight gain and adiposity. We documented the molecular underpinning mediating these interactions, uncovering the involvement of the mTOR-p70S6K pathway.
In an E2-withdrawal model, the upregulation of phosphorylated forms of mTOR, p70S6K, and S6 under pharmacological inhibition of k-OR strongly suggest that mTOR is a k-OR downstream signaling pathway as it has been previously observed in other animal models and contexts [24,37,38]. Likewise, in an OVX model, we escape from E2 action, while inhibition of k-OR signaling or activation of p70S6K in the MBH promotes loss of BW and browning, thereby keeping with previous findings on the effects of mTOR in energy homeostasis [39][40][41]. However, recent studies support the idea that mTOR-related signaling pathways are important players for thermogenesis and browning, and the precise role of mTOR in heat production is controversial, probably due to the use of different experimental models and systems. Differences between brown and beige adipocytes lie in the distribution, origin, and UCP1 expression under non-stimulated conditions. In the basal state, beige adipocytes express a very low level of the thermogenic gene program compared to brown adipocytes. The beige cell's capacity to switch between energy storage and energy dissipation strongly depends on the type of stimulus that it receives [39]. Based on distinctive features that brown and beige adipocytes possess, it is an interesting matter to identify potential mechanisms and regulators of these two adipocytes in energy homeostasis. In this line, our study provides evidence of k-OR/mTOR/p70S6K-mediated effects on WAT depots in an OVX mice model. Considering all these data together, this mechanism seems interesting as a drug target for diseases characterized by E2 privation.
These findings further our understanding of how OVX causes weight gain and paves the way for the search for novel specific drug targets for tackling obesity in conditions of ovarian hormone withdrawal, such as menopause, which affects a rather large fragment of the population, especially considering the lengthening of life expectancy. An obvious candidate for this could be the use of selective k-OR antagonists, an area in which recent developments make that a particularly feasible approach in disease states linked to alteration in estrogen levels [42].

Animal Procedures
Female 8-to 10-week-old C57/BL6 wild type and Oprk1−/−mutant mice (Jackson Laboratory, Bar Harbor, ME, USA, strain name B6.129S2-Oprk1tmqKff/J) were housed under conditions of controlled temperature (23 • C) and illumination (12 h light/12 h dark cycle) [25,43]. They were allowed ad libitum access to water and standard chow (16% proteins, 60% carbohydrates, 3% fat, Harlan Research Laboratories, Indianapolis, IN, USA) under specific-pathogen-free (SPF) conditions. All animals were allowed to acclimate to their new surroundings for one week before they underwent experimental procedures or other manipulations.
Adult animals were bilaterally OVX, or SHAM operated under ketamine-xylazine anesthesia [44,45]. Care of all animals was within institutional animal care committee guidelines, and all procedures were reviewed and approved by the Ethics Committee of the University of Santiago de Compostela (protocol number 15005AE/12/FUN01/FIS02/CDG4), in accordance with EU normative for the use of experimental animals. At the end of each experimental setting, animals were killed by decapitation, and the tissues were removed rapidly, frozen immediately on dry ice, and stored at −80 • C until analysis. The surgical procedure was performed by modifying a protocol previously described in rats [44]. The BW and FI were monitored for 11 weeks. The whole-body composition was measured using nuclear magnetic resonance spectroscopy (NMR) imaging (EchoMRI, Houston, TX, USA), as previously described [43,46]. Animals were monitored in a custom 12-cage indirect calorimetry, FI, and locomotor activity (LA) monitoring system (TSE LabMaster, TSE Systems, Bad Homburg, Germany) for 72 h [47,48]. Data collected from the last 48 h were used to calculate all metabolic parameters. The tracking of metabolic parameters was carried out at basal (23 • C), cold (4 • C), and thermoneutrality (30 • C) conditions. Body temperature was recorded with a rectal probe connected to a digital thermometer (BAT-12 Microprobe-Thermometer; Physitemp; NJ, USA). The interscapular temperature was assessed and was visualized using a high-resolution infrared camera (E60bx: Compact-Infrared-Thermal-Imaging-Camera; FLIR; West Malling, Kent, UK) and analyzed with a FLIR-Tools specific software package [49].

Experimental Setting 2: Effect of the Central Pharmacological Inhibition of k-OR on the Metabolic Changes Induced by OVX
To assess the chronic central effects of the k-OR selective antagonist, PF-04455242, WT, and Oprk1−/−animals were subjected to OVX 21 days after the surgery PF-04455242 (3.4 nmol/day) or vehicle (VH) were delivered i.c.v. via brain infusion kit 3 (1-3 mm) for 7 days through a catheter tube connecting the i.c.v. cannula to an osmotic minipump (model 1007D, Alzet Osmotic Pumps; DURECT, Cupertino, CA, USA), as previously described [43,46,50]. The mini pump was inserted in a subcutaneous pocket on the dorsal surface. The incision was closed with sutures, and mice were kept warm until full recovery. Body weight, food intake, lipid metabolism, and thermogenesis/browning were analyzed.

Experimental Setting 3: Effect of Constitutive Activation of MBH p70S6K on OVX-Induced Adiposity
Plasmid pRK7-HA-S6K1-F5A-E389-R3A was courteously gifted by Dr. Clémence Blouet. Cloning and package in Ad5-CMV-GFP adenoviruses were performed at Viral Vector Production Unit (Universitat Autónoma de Barcelona, Barcelona, Spain). To elucidate the contribution of central mTOR signaling on OVX-induced metabolic changes, we bilaterally injected in the medio-basal hypothalamic area (MBH) null adenoviruses (Ad-Null) or adenoviruses encoding a constitutively active form (Ad-CAS6K) [24,51] to OVX mice. Twenty-one days after OVX, the two groups (n = 12 mice per group) were subjected to stereotaxic surgery (±0.4 mm lateral, −1.5 mm antero-posterior, and −5.8 mm dorso-ventral). We monitored the BW and chow FI for 2 weeks and, afterward, evaluated the impact of p70S6K activation on peripheral lipid metabolism.

Serum Measurements
Luteinizing hormone was measured by specific RIA [52]. Serum metabolic markers such as triglycerides, cholesterol, and glucose were also assayed using a specific enzymatic colorimetric protocol (SPINREACT, Girona, Spain) following the manufacturer's instructions and expressed in mg/dL.

Hematoxylin and Eosin Staining
Paraffin-embedded sections (BAT, gWAT, and siWAT) of 4 µm were cut with a microtome and stained using a standard Hematoxylin/Eosin Alcoholic procedure according to the manufacturer's instructions (BioOptica, Milan, Italy). Sections were then rinsed with distilled water and dried for 30 min at 37 • C. They were then mounted with permanent (non-alcohol, non-xylene-based) mounting media [53].

Immunofluorescence
Female mice were anesthetized with ketamine-xylazine and perfused intracardially with saline (0.9% NaCl) followed by 4% PFA in PBS (pH 7.4). Fixed brains were immersed in 30% sucrose and 0.01% sodium azide in PBS at 4 • C for 2 days. Next, 3 sets of coronal sections (30-µm-thick) were cut in a freezing microtome Leica CM1850 UV (Wetzlar, Germany) and stored at −20 • C in cryo-protectant. One set of free-floating sections from each animal was washed in TBS 0.1 M and incubated in blocking solution (2% donkey serum + 0.3% Triton X-100) in TBS 0.1 M for 60 min. Then, sections were incubated with rabbit antibody against green fluorescent protein (GFP) (Abcam ab290) in blocking solution overnight at 4 • C. For GFP visualization, we used Cy3 donkey anti-rabbit (Jackson ImmunoResearch Labs, 711-165-162). Sections were then washed and coverslipped with Fluorogel mounting solution. Images were captured with fluorescence stereo microscopes Leica M205 and camera CCD Leica 7000 T.

Statistical Analysis
The results are expressed as mean values ± SEM. GraphPad Prism (8.0) software was used for the data analysis. We used unpaired t-test and paired t-test analyses in beforeafter studies. Two-way ANOVA was conducted to analyze BW weight gain evolution. In cases that normality and homoscedastic test were not passed, we used a non-parametric Mann-Whitney test. Data were considered statistically significant at p < 0.05.