Abstract
In a two-year carcinogenicity study with administration of high doses of the partial nicotinic agonist varenicline (recently approved for smoking cessation), mediastinal hibernomas occurred in three male rats. To investigate potential mechanisms for partial and full nicotinic agonists to contribute to development of hibernomas, the effects of nicotine on rat brown adipose tissue (BAT) were studied. Male and female rats were administered nicotine at doses of 0, 0.3, and 1 mg/kg subcutaneously for fourteen days. Intrathoracic (mediastinal periaortic and mediastinal perithymic) BAT and interscapular BAT were examined microscopically, and determinations of uncoupling protein-1 (UCP-1) expression and norepinephrine (NE) content were made. Additionally, NE turnover was measured in mediastinal periaortic and perithymic BAT. Nicotine (1 mg/kg) administration resulted in decreased vacuolation only in mediastinal periaortic and mediastinal perithymic BAT of males and elevated UCP-1 in mediastinal periaortic BAT of males and females. Increased NE content occurred only in mediastinal periaortic BAT of males given 0.3 and 1 mg/kg doses, whereas NE turnover was decreased in both males and females given 1 mg/kg. Together, these data demonstrate that nicotine primarily affects mediastinal BAT in male rats, consistent with the gender and location of the hibernomas observed in the two-year carcinogenicity study.
Introduction
The main function of brown adipose tissue (BAT) is to create heat via the mechanism of nonshivering thermogenesis. Norepinephrine (NE) is considered the most important factor in the regulation of BAT thermogenesis (Cannon and Nedergaard, 2004). NE stimulates the β3-adrenergic receptor on brown adipocytes leading to lipolysis and the β-oxidation of free fatty acids (Arch et al., 1984; Chaudhry and Granneman, 1999). The increase in β-oxidation generates electron donors in the BAT mitochondria, which results in the pumping of protons out of the mitochondrial matrix. In addition, NE induces the expression of uncoupling protein-1 (UCP-1), an essential component of the nonshivering thermogenic response that is uniquely expressed in BAT (Nedergaard et al., 2001). Activation of UCP-1 drives protons back into the mitochondria, and the energy released from this proton-motive force generates heat (Figure 1). At the same time, promotion of mitochondrial β-oxidation in the brown adipocytes also results in superoxide formation (Echtay et al., 2002), a major cause of intracellular oxidative damage, which may be partially countered by UCP-1.
In addition to controlling BAT thermogenesis, NE controls the proliferation of BAT via β-adrenergic pathways (Himms-Hagen, 1985; Nagase et al., 1994; Tsukazaki et al., 1995). During adaptation to cold and increased thermogenesis, BAT undergoes hyperplasia (Bukowiecki et al., 1982). NE also suppresses apoptosis in BAT (Lindquist et al., 2000). Therefore, the combination of excessive oxidative stress associated with increased proliferation and decreased apoptosis may sensitize the BAT development of hibernomas.
Nicotine, in addition to its well-known psychoactive properties, has a number of peripheral effects, one of which is increased thermogenesis in rodents (Lupien and Bray, 1988). The proposed mechanism is via sympathetic nervous system stimulation leading to an increase in NE. This stimulation could be a direct effect of nicotine on nicotinic acetylcholine receptors (nAChR) modulating sympathetic stimulation or an indirect response to the acute decrease in body temperature caused by nicotine (Rezvani and Levin, 2004). Whatever the mechanism, ultimately, nicotine increases both NE turnover and binding of guanosine 5′-diphosphate (a marker of thermogenesis) to mitochondria in BAT within three hours of treatment and increases expression of UCP-1 in BAT from the rodent (Arai et al., 2001; Lupien and Bray, 1988).
The relevance of the nicotinic role in BAT metabolism has been of interest for novel nicotinic agents in smoking cessation. Varenicline is a partial nicotinic agonist with high affinity for α4β2 nAChR (Coe et al., 2005; Rollema et al., 2007) that is approved for this indication. Similar to nicotine, varenicline decreases body temperature in some species, for example, mouse and monkey; it has not been tested in the rat (unpublished data). A two-year oral carcinogenicity study was conducted in rats in which high doses of varenicline (1, 5, or 15 mg/kg) or a vehicle control were administered orally once per day. Although no statistically significant differences in tumor incidence were observed between control and treated groups, three hibernomas, a neoplasm of BAT, were observed in the mediastinum of male rats administered the two highest doses of varenicline. The one hibernoma observed in the 5 mg/kg group was benign, and the two hibernomas observed in the 15 mg/kg dose group were malignant (unpublished data). These occurrences were unlikely to be the result of genetic toxicity, as varenicline has been determined not to be genotoxic in vitro in Ames bacterial mutation, Chinese hamster ovary/hypoxanthine-guanine phosphoribosyl transferase, or human lymphocyte assays, or in vivo in rat bone marrow tests for cytogenetic aberration. Although reports of drug-induced hibernoma formation are very rare, this finding has some parallels with a similar study on the reversible α-adrenergic antagonist phentolamine, a different class of drug (Poulet et al., 2004). In that study, male and female rats were given 10, 50, or 150 mg/kg phentolamine mesylate orally, which resulted in the occurrence of hibernomas (largely localized to mediastinal BAT), with a much higher frequency in male rats compared with female rats (10:1). Phentolamine is believed to cause hibernomas in rats via an increase in NE from a chronic adaptive sympathetic response to the effects of phentolamine on blood pressure (Poulet et al., 2004), but this drug has never been linked with the occurrence of hibernoma in humans.
Based on the data summarized above, we hypothesized that nicotinic agents like varenicline, at exaggerated doses, could function similarly to nicotine and increase sympathetic stimulation and NE levels. This increase leads to elevated β-oxidation, proliferation, and increased UCP-1 expression with concurrent inhibition of apoptosis in BAT, which ultimately leads to the development of hibernomas. The objective of this study was to investigate the effect of high-dose nicotine on BAT of rats in an attempt to understand the site-specific location (mediastinum) and sex predilection (males only) for the occurrence of these tumors in the two-year varenicline study in rats. The parameters measured in this study were BAT vacuolation (an indicator of lipolysis), NE content and turnover, and UCP-1 messenger ribonucleic acid (mRNA) and protein expression in BAT as markers for the potential mechanism of hibernoma development in rats.
Materials and Methods
Animals
Thirty male (150–200 g) and thirty female (125–175 g) Sprague-Dawley rats (Crl:CD), aged from six to seven weeks, were obtained from Charles River Laboratories, Kingston, NY, USA. Rats were individually housed in suspended, stainless-steel cages and exposed to a twelve-hour light (6 a.m. to 6 p.m.)–twelve-hour dark cycle. They received standard commercial laboratory rodent chow (certified rodent diet 5002, PMI Feeds, Inc.) and reverse-osmosis purified drinking water from a municipal source (regulated by the Environmental Protection Agency) ad libitum throughout the study.
Treatments
Rats (ten/sex/group) were tagged for identification and randomized to receive nicotine 0.3 mg/kg (low dose), 1 mg/kg (high dose), or 0 mg/kg (control). Nicotine was dissolved in 0.9% saline and administered as a daily subcutaneous injection to the dorsum for fourteen days.
Rats (three/sex/group) were designated for measurement of BAT NE content and turnover. These rats were dosed acutely with a single injection (80 mg/kg at 1 mL/kg i.p.) of DL-α-methyl-para-tyrosine (Sigma-Aldrich, St. Louis, MO, USA; Cat. No. 7628) four hours after the last nicotine injection on Day 14. This protocol was approved by the institutional animal care committee, and the activities complied with or exceeded the Animal Welfare Act Regulation (Title 9 CFR, Parts 1 to 3) and the Association for Assessment and Accreditation of Laboratory Animal Care, International Standards as set forthby the Guide for the Care and Use of Laboratory Animals.
Evaluations
Rats were observed once daily for clinical signs prior to treatment initiation and on the last day of treatment. During the fourteen-day treatment period, rats were observed twice daily. Body weight was measured prior to treatment and on Days 1, 4, 7, 10, and 14.
Histopathology and UCP-1 Immunohistochemical Quantitation
For histopathology and quantitative analyses of UCP-1 protein and mRNA, rats (seven/sex/group) were euthanized approximately twenty-four hours following the last treatment with nicotine on Day 14. Subcutaneous (interscapular BAT) and intrathoracic (mediastinal) BAT from two areas (periaortic and perithymic) of rats (four/sex/group) were fixed in neutral-buffered formalin, dehydrated, and embedded in paraffin for histopathology and UCP-1 protein immunohistochemistry (IHC). Mediastinal periaortic, mediastinal perithymic, and interscapular BAT from the remaining three rats/sex/group were snap-frozen in liquid nitrogen and stored at −80°C for UCP-1 mRNA quantitation by reverse transcriptase polymerase chain reaction (RT-PCR).
Serial sections of mediastinal periaortic, mediastinal perithymic, and interscapular BAT were deparaffinized and dehydrated. For light microscopic histopathology, sections were stained with hematoxylin and eosin and examined for changes in the amount of cytoplasmic vacuolation. For UCP-1 IHC, the slides were stained with rabbit antibodies to UCP-1 followed by biotinylated anti-rabbit IgG and streptavidin AlexaFluor 488. Nuclear DNA was stained with DAPI. The amount of UCP-1 protein was quantified by analyzing cellular fluorescence using Laser Scanning Cytometry (LSC, CompuCyte Corp, Cambridge, MA, USA) and a quantitation method similar to that previously described by Pruimboom-Brees et al. (2005). Briefly, slides were visually inspected for positive staining under an epifluorescence microscope (Olympus BX51, Olympus America, Inc., Melville, NY, USA) using mercury lamp illumination. Then, using the iCyte model of the LSC system, AlexaFluor488 and DNA-associated DAPI fluorescence were excited with an argon ion laser (488 nm) and a violet diode laser (400 nm); emissions were detected and measured using standard band pass 515–545 nm and long pass 460–485 nm filters, respectively. Individual adipocytes were identified by DAPI fluorescence, and adipocyte perimeter was selected at 18 pixels outside the nuclear contour to include the cytoplasmic green fluorescence (contour analysis), to generate a mean integral fluorescence value of UCP-1/cell. Fluorescence of a minimum of 20,000 adipocytes per section of hibernoma was measured at 20X magnification. Negative control slides were used to determine the level of background fluorescence and to define gating prior to analysis.
UCP-1 mRNA RT-PCR Analysis
Total RNA was extracted from BAT using the RNeasy Lipid Tissue mini kit (Qiagen, Valencia, CA, USA; Cat. N0. 74804) and then reverse-transcribed to cDNA. Subsequently, cDNA was then amplified with the UCP-1 Taqman Gene Assay (Applied Biosystems, Foster City, CA, USA). RT PCR was performed on an ABI Prism 7900HT (Applied Biosystems, Foster City, CA, USA) instrument in conjunction with an automation robot. The baseline and threshold were set automatically with SDS2.2 software (Applied Biosystems, Foster City, CA, USA), and data were exported for analysis. mRNAs from the acidic ribosomal phosphoprotein P0 (Arbp) and glyc-eraldehyde-3-phosphate dehydrogenase (GAPDH) genes were also amplified and used as endogenous gene references. UCP-1 gene expression was normalized to the mean of reference genes Arbp and GAPDH, and the relative quantitation was expressed as fold change in gene expression.
Measurement of NE Content and Turnover
Part of nicotine’s ability to decrease weight gain in rats may be a result of increased thermogenesis in rat BAT, caused by sympathetic nervous system activation. Because Lupien and Bray (1988) have reported that a single nicotine injection in rats increased NE turnover in BAT after α-methyl-tyrosine administration, we looked at NE content and turnover to determine whether NE levels in BAT were altered and if so, whether the changes were associated with an increased turnover or decreased turnover. Rats assigned for NE content evaluation were euthanized two hours after the injection of DL-α-methyl-para-tyrosine, which blocks tyrosine hydroxylase and prevents reaccumulation of NE if NE stores are depleted because of neuronal or drug stimulation. Mediastinal periaortic, mediastinal perithymic, and interscapular BAT were collected, snap-frozen in liquid nitrogen, and stored at −80°C until NE content determination was carried out. NE turnover determination was carried out on the remaining tissue from mediastinal periaortic and mediastinal perithymic BAT, but this measurement was prevented in interscapular BAT owing to limitations in the quantity of this tissue. Measurement of NE content and turnover followed a modified method similar to that described previously by Lupien and Bray (1988) and Rollema et al. (2007). Briefly, BAT samples were thawed and weighed, then homogenized in 0.1 N perchloric acid at 4°C. Following centrifugation at 14,000 rpm for ten minutes, concentrations of NE and its major metabolite, 3-methoxy-4-hydroxyphenylglycol (MHPG), were measured in 10 μL of the supernatant by high-performance liquid chromatography (HPLC) with coulometric detection at 650 mV (Coulochem II ESA, Chelmsford, MA, USA), using an autosampler (717 Refrigerated Autosampler, Waters, Bedford, MA, USA). Analytes were separated over a C18 Beckman Ultrasphere 5μ ODS 4.6 mm × 25 cm column with a mobile phase consisting of 20 mM NaH2PO4, 9% methanol, 0.2 mM octanesulfonic acid, and 0.2 mM EDTA, pH 3.8, at 1 mL/min (LC-10AD pump, Shimadzu, Columbia, MD, USA). Absolute values of NE are represented as ng/g wet weight of tissue. NE turnover was calculated as the ratio ([MHPG]/[NE]) and expressed as percentage of control.
Statistical Analyses
Statistical significance was analyzed by one-way analysis of variance (ANOVA) using Bonferroni’s/Dunnett’s post-hoc test or Fisher’s probable least-squares difference test (N = 4–6).
Results
No significant changes in body weight were observed, and the behavior and appearance of the treated rats were normal and not different from the control rats during the course of the study.
Histopathology
In control animals, the mediastinal periaortic and mediastinal perithymic BAT had less-prominent vacuolation, characterized by the presence of very few large cytoplasmic vacuoles as compared with the interscapular BAT. BAT in females, regardless of the location, had very little vacuolation as compared to BAT of males. Treatment-related decreased vacuolation (decreased number of large vacuoles), an indication of increased lipolysis and fatty acid oxidation, was observed in BAT only from nicotine-treated males (1 mg/kg) compared with BAT from control males. In males given 1 mg/kg, a similar decrease in relative vacuolation was evident in interscapular, mediastinal periaortic, and mediastinal perithymic BAT (Table 1 and Figure 2). No decreased vacuolation was observed in nicotine-treated female rats at any of the BAT sites sampled (Figure 2). This lack of effect as seen in males was most likely a result of the minimal vacuolation (characterized by a lack of large vacuoles) present in these sites for control female rats (data not shown).
UCP-1 Protein Expression
At 1 mg/kg, nicotine increased UCP-1 expression in the mediastinal periaortic BAT of both male and female rats. LSC analysis of interscapular and mediastinal BAT sections indicated significantly increased UCP-1 expression (p < .05) in the mediastinal periaortic but not in the mediastinal perithymic or interscapular BAT (Figure 3).
UCP-1 Gene Expression
No significant changes in UCP-1 gene expression were measured in any of the three BAT areas sampled from male or female rats at either dose of nicotine, relative to endogenous reference gene expression (Figure 4).
NE Content of BAT
In control animals, the NE content was significantly greater in female than in male rats for the mediastinal periaortic and mediastinal perithymic BAT groups (Figure 5). In treated animals, a significant increase in content of NE was observed only in the mediastinal periaortic BAT of male rats dosed at both 0.3 and 1 mg/kg nicotine (p < .05; Figure 5). No other statistically significant increases in BAT NE levels were observed in either males or females. Decreased NE levels (p < .05) (Figure 5) occurred in interscapular BAT from males and mediastinal periaortic BAT from females treated with nicotine 1 mg/kg.
NE Turnover in BAT
Treatment of rats with nicotine resulted in a dose-related decrease in NE turnover (Figure 6). Compared with controls, NE turnover was reduced significantly in the mediastinal periaortic and mediastinal perithymic BAT of male rats administered the 1 mg/kg dose of nicotine (p < .05). Similarly, in female rats given the 1 mg/kg dose, NE turnover was reduced in mediastinal periaortic BAT (p < .05), but the decrease in the mediastinal perithymic BAT narrowly missed statistical significance. No decrease in NE turnover was observed in either sex from the 0.3 mg/kg dose group.
Discussion
Nicotine and other nicotinic agonists (e.g., varenicline) stimulate release of NE, which plays a role in BAT-mediated thermogenesis. The objective of this study was to evaluate the potential pharmacological responses of different BAT depots in male and female rats administered a high dose of nicotine daily for fourteen days. The pharmacological response of BAT after nicotine treatment was assessed by NE content and turnover, the amount of cytoplasmic vacuolation (an indicator of lipolysis), and UCP-1 levels (a key protein for thermogenesis). In control animals, there was a sexual dimorphism in NE content, with females exhibiting greater NE content in both mediastinal periaortic and mediastinal perithymic BAT than males.
Nicotine treatment increased NE content in BAT in a localized and gender-specific manner. Increases were observed only in the mediastinal periaortic BAT from males dosed with nicotine at 0.3 and 1 mg/kg. In contrast, NE content in interscapular BAT decreased at nicotine 1 mg/kg, consistent with previous findings (Lupien and Bray, 1988). In females, NE content was unchanged in interscapular and mediastinal perithymic BAT, but it was reduced in the mediastinal periaortic BAT at nicotine 1 mg/kg. This may be the result of a more intense and/or prolonged and gender-specific effect of nicotine in this BAT site compared with mediastinal perithymic and interscapular BAT. In control animals, sexual dimorphism was also observed in the decreased prominence of vacuolation of BAT at all sites in females as compared with males. Administration of nicotine (1 mg/kg) resulted in decreased vacuolation of BAT from all sites in males only, whereas no change in vacuolation occurred in females. In contrast, sexual dimorphism for NE turnover was not observed within the time frame of this study. Turnover was reduced in mediastinal periaortic and mediastinal perithymic BAT of males and in mediastinal periaortic BAT of females, with a trend toward decreased NE turnover in female mediastinal perithymic BAT, which narrowly missed significance. Similarly, a gender difference was not observed in UCP-1 protein levels, which were increased in the mediastinal periaortic BAT of both sexes with nicotine (1 mg/kg).
The BAT changes observed in this study are consistent with nicotine-activated, NE-mediated lipolysis secondary to increased β-oxidation of the fatty acid that occurs to a greater extent in the mediastinal periaortic BAT of males. Although in normal circumstances the elevation of UCP-1 would be expected to limit superoxide production, the apparent predisposition to increased NE content in mediastinal periaortic BAT of males could potentially overcome any protective effect of increased UCP-1, which could result in mediastinal periaortic BAT of males being more susceptible to oxidative radical injury and subsequent hibernoma formation than mediastinal periaortic BAT of females.
The reason for the sexual dimorphism in BAT pharmacology and hibernoma susceptibility in the rat is not fully understood. However, it is known that sexual dimorphism exists in the adrenergic control of brown adipocytes in the rat. Studies have shown that overfeeding in males causes the release of NE and subsequent activation of the β3-adrenergic receptor on brown adipocytes. In females, overfeeding results in body weight excess but a lower activation of thermogenesis (Rodriguez et al., 2001). This may be a result of a higher level of β3-adrenergic receptors in BAT of males leading to greater thermogenic capacity. In female rats, estradiol and progesterone reduce the density of the β3-adrenergic receptor and inhibit BAT thermogenesis (Malo and Puerta, 2001), which could be an explanation for the absence of hibernomas in female rats.
The reason for the differences in nicotinic pharmacology in the different BAT depots is also not clear. It is possible that cholinergic innervation is present in rat mediastinal BAT but not in BAT from the interscapular, cervical, or perirenal areas in this species (Giordano et al., 2004). This difference in innervation of various BAT depots could explain the discrete localization in the present study of the nicotine-mediated increase in NE content in male rat mediastinal periaortic BAT and potential for hibernoma observed at that BAT site in the varenicline study.
BAT, although present in most mammals, varies across species in its time of development, quantity, and function. In smaller mammals, such as rodents, BAT is present at birth, develops rapidly after birth, and is important in thermogenesis (Cannon and Nedergaard, 2004). In larger mammals, BAT depots are evident at birth but diminish rapidly afterward. In humans, BAT is present in the fetus, with the maximal amount (1% of body weight) present at birth (Lean and James, 1986). Unlike rodents, BAT in humans becomes devoid of mitochondria and loses its thermogenic capacity soon after birth (Sell et al., 2004). The rats in the varenicline study in which hibernomas were observed were exposed to high doses, far in excess of the therapeutic dose used for smoking cessation therapy in humans. The rare occurrence of hibernoma in rats is likely secondary to high-dose pharmacological stimulation of BAT via a nicotinic pathway and the result of a mechanism specific to this rodent species with thermogenically active brown fat.
Spontaneous hibernomas are a rare occurrence in both research and clinical settings, and their etiology is not well understood. Hibernomas arising in the thoracic cavity of rats have been reported (Coleman, 1980; Stefanski et al., 1987). In humans there are only 170 case reports of hibernomas described (Furlong et al., 2001). Clinical evidence suggests hibernomas in humans are slow growing and associated with sites of persistent brown fat in adults. Hibernomas in humans were considered benign and were almost always subcutaneous in location (Furlong et al., 2001).
In conclusion, these data are consistent with the hypothesis that during adrenergic-mediated thermogenesis, the mediastinal periaortic BAT of male rats is more susceptible to oxidative radical injury than in females, and chronic, high-level stimulation of NE release by excessive activation of nAChRs with nicotinic agonists in male rats may lead to development of hibernoma in the mediastinal periaortic BAT.
Footnotes
Acknowledgments
The authors gratefully thank Gwen Curry, Jeanne Wolfgang, Germaine Boucher, James Yan, Raoul Jamon, Kristine Hibbard, and Alane Kennedy (all of Pfizer Global Research and Development, Groton, Connecticut, USA) for their technical assistance in these studies. This study was sponsored by Pfizer Inc. Editorial support was provided by Christopher Grantham, PhD, of Envision Pharma Ltd, Horsham, UK, and was funded by Pfizer Inc.
Current address for Dominique J. Brees: Glaxo SmithKline, Ware, UK
Current address for Steven B. Sands: Johnson & Johnson Pharmaceutical Research and Development, San Diego, CA, USA
Current address for Michael R. Elwell: Covance Laboratories, Vienna, VA, USA