The effect of TNFα on food intake and central insulin sensitivity in rats
Annette D. de Kloeta,b,⁎, Gustavo Pacheco-Lópezc, Wolfgang Langhansc, Lynda M. Brownd
aProgram in Neuroscience, University of Cincinnati, Cincinnati, Ohio, United States
bDepartment of Psychiatry and Behavioral Neuroscience, University of Cincinnati, Cincinnati, Ohio, United States
cPhysiology and Behavior Laboratory, Institute of Food, Nutrition and Health, ETH Zurich, Schwerzenbach, Switzerland
dDepartment of Nutrition, University of North Carolina at Greensboro, Greensboro, North Carolina, United States
a b s t r a c ta r t i c l e i n f o
Received 5 October 2010
Received in revised form 4 November 2010
Accepted 29 November 2010
Tumor necrosis factor
Circulating and tissue levels of the proinflammatory cytokine tumor necrosis factor α (TNFα) are elevated in
obesity. TNFα interferes with insulin signaling in many tissues and also plays a causal role in the anorexia that
accompanies severe challenges to the immune system. The interactions between TNFα and insulin in the
control of eating are less well known. The present study evaluated the role of TNFα in the central nervous
system control of food intake by insulin in adult male Long Evans rats. We first determined the ability of
several doses of TNFα injected into the 3rd cerebral ventricle (i3vt) to reduce food intake in male rats.
Subsequently, we assessed the ability of a subthreshold dose of TNFα to modulate the effect of i3vt insulin on
food intake in male rats fed a low-fat chow or a high-fat (HF) diet. TNFα administered i3vt dose-dependently
reduced food intake in rats fed a standard low-fat chow diet. Moreover, a low, sub-threshold dose of TNFα
diminished the reduction in food intake by insulin in rats maintained on a chow diet, but enhanced insulin
action in rats maintained on a HF diet. These data suggest that the interaction of TNFα with central insulin
varies with nutritional and/or dietary conditions.
© 2010 Elsevier Inc. All rights reserved.
The incidence of obesity and related co-morbidities affects
individuals of all ages and both sexes and has become a world-wide
epidemic. Consequently, it is essential that therapeutic approaches to
overcome this crisis be developed, and numerous pharmacological
targets are currently under investigation. In this regard, it has become
increasingly clear that obesity is accompanied by low-grade inflam-
mation, suggesting a link between metabolism and the immune
system [1–5]. Both macrophages and adipocytes within adipose tissue
secrete numerous cytokines, and as adipose tissue expands, the levels
of proinflammatory cytokines also increase, both within the adipose
tissue itself as well as systemically . This inflammation has been
hypothesized to play a causal role both in adiposetissue expansion via
direct actions of cytokines on adipose tissue [4,7] and in the
development of obesity-related co-morbidities, such as insulin
resistance, through insulin-desensitizing actions at the insulin
receptor in several tissues .
Tumor necrosis factor α (TNFα) is one such proinflammatory
cytokine and a major mediator of innate immune reactions. As obesity
increases, TNFα levels in plasma and in adipose tissue also increase,
and this has been implicated in many symptoms of the metabolic
syndrome [3,6]. In peripheral tissues, TNFα reduces insulin sensitivity
by impairing insulin signal transduction [8,9]. On the other hand,
TNFα also mediates the cachexia that accompanies cancer and severe
infections, and there is evidence that TNFα acts within the central
nervous system, particularly within the hypothalamus, to contribute
to this catabolic condition [10–15].
The findings that TNFα levels directly correlate with the level of
adiposity, and that TNFα acts within the CNS to promote negative
energy balance, are reminiscent of what is known about the
pancreatic hormone insulin. It is impossible to write about the central
actions of insulin without mention of the work of Steve Woods, the
person celebrated in this Festschrift. His studies were seminal in
elucidating how insulin communicates the level of energy stored in
the body to the brain and how it helps control food intake and defend
a particular level of adiposity [16–22]. A more recent revelation is that
many different factors influence the sensitivity of the brain to insulin.
For example, when animals are maintained on a high-fat (HF) diet,
brain insulin sensitivity decreases while the expression of inflamma-
tory markers, including TNFα, within the CNS increases [23–25].
Given that TNFα is well-known to impair systemic insulin signaling
[8,9] and that both insulin and TNFα act within the hypothalamus to
control eating, it is reasonable to hypothesize that TNFα and insulin
interact at the level of the hypothalamus. Some evidence in fact
suggests that this may be the case [15,26,27].
The studies presented here were inspired by our discussions with
SteveWoods.We investigated theabilityofi3vt administered TNFα to
modify the effects of insulin on food intake in male rats fed a low-fat
Physiology & Behavior 103 (2011) 17–20
⁎ Corresponding author. Program in Neuroscience, University of Cincinnati, 2170
East Galbraith Road ML 0503, Cincinnati, OH 45237, United States. Tel.: +1 513 558
5866; fax: +1 513 297 0966.
E-mail address: email@example.com (A.D. de Kloet).
0031-9384/$ – see front matter © 2010 Elsevier Inc. All rights reserved.
Contents lists available at ScienceDirect
Physiology & Behavior
journal homepage: www.elsevier.com/locate/phb
or a high-fat diet. As TNFα impairs insulin signaling in peripheral
tissues, one logical hypothesis is that a low dose of TNFα would
similarly reduce brain insulin sensitivity, resulting in less suppression
of food intake in response to centrally-administered insulin. On the
other hand, the fact that both TNFα and insulin, when administered
individually into the brain, reduce food intake, might suggest the
converse hypothesis that the two peptides would have an additive or
even synergistic effect when administered in combination.
Adult Long Evans rats (Harlan, Indianapolis, IN) were individually
housed in tub cages with bedding and maintained on a 12-hour light,
12-hour dark cycle (lights on at 0100 h) and given ad libitum access to
water and food unless otherwise noted. Rats weighed 415–515 g at
the start of the experiments and were fed either low-fat rodent chow
(Harlan Teklad [LM485], Indianapolis, IN; 3.1 kcal/g; ~5% fat) or a
high-fat (HF) diet (Research Diets, New Brunswick, NJ; 4.54 kcal/g;
~41% fat). All procedures were approved by the University of
Cincinnati Institutional Animal Care and Use Committee.
2.2. Stereotaxic surgery
Seven days after arrival, rats were anesthetized with 0.1 ml/100 g
body weight intraperitoneal (ip) injections of a ketamine (70 mg/kg)/
xylazine (2 mg/kg) mixture. Subsequently, 22-gauge guide cannulas
(Plastics One Inc., Roanoke, VA) were stereotaxically implanted into
the 3rd cerebral ventricle (i3vt). Briefly, bregma and lambda were
situated at the same vertical coordinate and cannula tips were
positioned on the midline, 2.2 mm posterior to bregma and 7.5 mm
ventral to the dura mater. The cannulas were then fixed to the skull
using dental acrylic and anchor screws. Obturators extending 0.5 mm
beyond the cannula tract were inserted. When all rats had returned to
their pre-surgical body weights, cannula placement was confirmed by
i3vt infusion of 10 μg of NPY in 1 μl normal saline 3 h prior to the onset
of dark and monitoring the rats' food intake over a 60-min period.
Animals that did not eat at least 2 g of chow within 60 min were
excluded from the study.
Prior to the experiments, stock solutions of murine TNFα (R&D
Systems, Minneapolis MN) were made at a concentration of 10 ng/μl
in 0.2% BSA/PBS. On experimental days, the murine TNFα stock
solutions were diluted to the appropriate concentrations (0, 0.3, 1, 3
and 10 pg/ μl, equivalent to: 0, 17.6, 58.2, 176.5 and 588.2 pM) in 0.2%
BSA/PBS and delivered i3vt in a volume of 1 μl. Similarly, insulin
(100 U/ml; Novolin R Regular human insulin, Novo Nordisk, Prince-
ton, NJ) was diluted to the appropriate concentration of 10 U/ml
(equivalent to 60 μM) in saline on the day of the experiment.
2.4. Experiment 1
A cohort of male rats maintained on chow received each dose of
TNFα (0, 0.3, 1, 3 and 10 pg i3vt) in an individually randomized order
with3 non-injectiondays occurringbetweensuccessiveinjections.On
each experimental day, food hoppers were removed from cages and
weighed 3 h prior to the onset of dark. At 45 min prior to the onset of
the dark, body weights were recorded and rats received a 1 μl i3vt
injection of vehicle or vehicle plus TNFα which was slowly infused
with a Hamilton Syringe. Food hoppers were returned at the onset of
dark and weighed at 4 and 24 h after infusion.
2.5. Experiment 2
A separate group of rats from the same cohort as in Experiment 1
was fed chow or the HF diet for 2 weeks before receiving i3vt
cannulas. After recovery from the surgery and verification of cannula
placement, there were 4 experimental treatments with at least 3 non-
injection days occurring between treatment days. Each rat received
each treatment in an individually randomized order. On experimental
days, rats were fasted 4 h prior to the dark and received two 1 μl i3vt
injections, spaced 5 min apart, just before dark: vehicle (saline)/
vehicle (BSA/PBS), insulin (10 mU)/vehicle, vehicle/ TNFα (1 pg,
based on Experiment 1), or insulin/TNFα. Food intake was assessed
after 4 and 24 h and body weight was assessed after 24 h.
2.6. Data analysis
Datawere analyzedusing Statistica(StatSoft, Tulsa,OK, USA). Food
intakes were assessed using a repeated measures analysis of variance
(ANOVA). Main effects and interactions were assessed with Student–
3.1. i3vt TNFα dose-dependently reduces food intake in rats
In Experiment 1, young adult male rats were used to determine an
optimal sub-threshold dose of TNFα for use in the subsequent
experiment. As depicted in Fig. 1, i3vt TNFα dose-dependently
reduced food intake at both 4 and 24 h in rats maintained on low-
fat chow. Both the 3 and the 10 pg doses effectively reduced food
intake at the 24-h time-point and the 10 pg dose was also effective at
the 4-h time-point (pb0.05; Fig. 1). Based on this, a dose of 1 pg of
TNFα was selected as a sub-threshold dose for use in Experiment 2.
3.2. TNFα modulates central sensitivity to insulin's anorectic actions
The TNFα alone (1 pg) had no effect on food intake or body weight
on rats maintained on either the chow or the HF diet, indicating that
the dose was indeed sub-threshold (Fig. 2A–C). Consistent with
previous studies by Woods and colleagues, insulin alone at a dose of
10 mU reduced 24-h food intake (pb0.05; Fig. 2B) and body weight
(pb0.05; Fig. 2C) in rats maintained on the chow diet, but failed to do
so in rats maintained on the HFD [17,24,28]. The combination of
insulin plus TNFα blunted insulin's ability to reduce 24-h food intake
and body weight in rats maintained on chow and enhanced its ability
to reduce food intake and body weight in rats maintained on the HFD
Fig. 1. Mean 4- and 24-h food intake (kcal) of male Long Evans rats maintained on chow
and administered TNFα i3vt at the indicated doses. *=pb0.05. n=6/group. Bars=1
A.D. de Kloet et al. / Physiology & Behavior 103 (2011) 17–20
The present experiments represent an initial foray to consider how
TNFα interacts with factors that act in the brain to regulate energy
balance, in particularinsulin.TNFα bindstotwo distinctreceptors,the
type 1(TNFR1 or p55) and type 2 (TNFR2 or p75) TNFα receptors ,
both of which are found in the brain . Activationof these receptors
then mediates TNFα's physiological actions such as the initiation of
proinflammatory responses and stimulation of nitric oxide (NO), as
well as numerous other functions . With regard to energy balance,
stimulation of TNFα receptors plays a well-established role in the
cachexia that accompanies certain cancers and severe infections, and
it is clear that TNFα's actions within the central nervous system,
particularly within the hypothalamus, mediate this catabolism [10–
15]. Consistent with this, the first finding of this set of experiments
was that i3vt TNFα dose-dependently reduces food intake in adult
male Long Evans rats.
In addition to its catabolic role in the brain, another characteristic
of TNFα is that it acts at various peripheral tissues to impair insulin
signal transduction . Less clear, however, is whether TNFα interacts
with insulin in the brain. In this regard, another finding of these
experiments was that a subthreshold dose of i3vt TNFα (1 pg;
determined in Experiment 1) reduces sensitivity to the anorectic
actions of i3vt insulin in rats fed a normal low-fat chow diet, while it
enhances insulin action in rats fed a HF diet. One implication is that
when rats are maintained on a relatively low-fat diet, TNFα causes
central insulin resistance. Conversely, when rats are fed a HF diet,
central TNFα enhances the catabolic action of insulin.
The ability of a low-dose of TNFα to reduce insulin's anorectic
actions in rats maintained on chow is consistent with reports by
Romanatto et al. and Moraes et al. in which a low subthreshold dose of
TNFα (0.035 pg) administered into the lateral cerebral ventricle
attenuated the ability of insulin as well as of leptin to inhibit food
intake in male rats maintained on a chow diet [15,26,27]. Conversely,
a higher dose of TNFα (350 pg) that did reduce food intake by itself
did not interact with insulin's anorectic action [15,26,27].
There is substantial evidence supporting an interaction between
hypothalamic inflammation and insulin sensitivity [3,5,23]. Increases
in hypothalamic proinflammatory signaling, as occurs with the
consumption of a HF diet or with the administration of a sub-
threshold dose of TNFα, impairs the sensitivity to the anorexigenic
actions of insulin . Moreover, TNFα stimulates nitric oxide (NO)
production and Moreas et al. ascertained that this mediates, at least in
part, TNFα's insulin desensitizing actions in the brain. In this regard, a
low dose of TNFα (0.035 pg) stimulates NOS catalytic activity in the
hypothalamus in rats and also reduces sensitivity to insulin-induced
anorexia in rats, wild-type mice and neuronal nitric oxide synthase
(nNOS) knockout mice but fails to do so in inducible nitric oxide
synthase (iNOS) deficient mice, suggesting that nitric oxide (NO)
produced by iNOS in response to TNFα reduces central insulin
sensitivity . It is possible that a similar mechanism underlies the
insulin-desensitizing actions of TNFα the present studies.
Another important finding of the present studies is that TNFα has
divergent effects on the brain's sensitivity to insulin based on the
nutritional or dietary conditions. When exposed to a HF diet, the
interaction between TNFα and insulin actually becomes positive. In
this regard, and also consistent with our results, an increase in
hypothalamic inflammation has also been determined to reduce food
intake . Administration of TNFα into the lateral cerebral ventricle
induces the expression of hypothalamic neuropeptides that inhibit
food intake, including pro-opiomelanocortin (POMC) and corticotro-
phin-releasinghormone(CRH) to a greaterextentthanneuropeptides
that stimulate appetite . Perhaps when rats are fed a HF diet and
subsequently administered the sub-threshold dose of TNFα in
combination with insulin, these mechanisms contribute to the
enhanced hypophagia. Another possible explanation for these
differences based on dietary conditions relates to the role of reactive
oxygen species (ROS) in energy balance . Although, ROS are
implicated in peripheral insulin resistance , insulin's ability to
induce ROS in the hypothalamus is a critical step for promoting
catabolism . Exposure to a HF diet not only reduces insulin's
ability to reduce food intake but also reduces it's ability to stimulate
ROS . On the other hand, TNFα is a potent stimulator of ROS.
Perhaps, this increase in ROS induced by TNFα is then able to
reestablish insulin's anorectic action on a HF diet.
In summary, TNFα dose-dependently reduces food intake in male
rats fed a standard chow diet. Moreover, when administered directly
into the brain, a low, sub-threshold dose of TNFα diminishes insulin-
induced anorexia in rats maintained on a chow diet, but enhances
insulin action in rats maintained on a HF diet, indicating that the
interaction between TNFα and insulin in the brain depends on the
This work was supported in part by NIH grants DK078201 and F31
NS068122 (AdK) and the Albert J. Ryan Foundation (AdK).
Fig. 2. Mean 4-h caloric intake (A), 24-h caloric intake (B), and change in body weight
(C) of male Long Evans rats 24 h after receiving vehicle or insulin (10 mU) i3vt,
followed 5 min later by TNFα (1 pg) or vehicle i3vt. Rats were maintained on chow or
HF diet for 2 weeks prior to the start of the experiment. *=pb0.05. n=6/group.
A.D. de Kloet et al. / Physiology & Behavior 103 (2011) 17–20
 Xu H, Barnes GT, Yang Q, Tan G, Yang D, Chou CJ, et al. Chronic inflammation in fat
plays a crucial role in the development of obesity-related insulin resistance. J Clin
 Weisberg SP, McCann D, Desai M, Rosenbaum M, Leibel RL, Ferrante AW. Obesity is
associated with macrophage accumulation in adipose tissue. J Clin Invest
 Hotamisligil GS. Inflammation and metabolic disorders. Nature 2006;444:860–7.
 Neels JG, Pandey M, Hotamisligil GS, Samad F. Autoamplification of tumor necrosis
factor-alpha: a potential mechanism for the maintenance of elevated tumor
necrosis factor-alpha in male but not female obese mice. Am J Pathol 2006;168:
 Thaler JP, Schwartz MW. Minireview: inflammation and obesity pathogenesis: the
hypothalamus heats up. Endocrinology 2010;151:4109–15.
 Hotamisligil GS, Arner P, Caro JF, Atkinson RL, Spiegelman BM. Increased adipose
tissue expression of tumor necrosis factor-alpha in human obesity and insulin
resistance. J Clin Invest 1995;95:2409–15.
 Schreyer SA, Chua Jr SC, LeBoeuf RC. Obesity and diabetes in TNF-alpha receptor-
deficient mice. J Clin Invest 1998;102:402–11.
 Hotamisligil GS, Murray DL, Choy LN, Spiegelman BM. Tumor necrosis factor alpha
inhibits signaling from the insulin receptor. Proc Natl Acad Sci USA 1994;91:
 Hotamisligil GS. Inflammatory pathways and insulin action. Int J Obes Relat Metab
Disord 2003;27(Suppl 3):S53–5.
 Vassalli P. The pathophysiology of tumor necrosis factors. Annu Rev Immunol
 Grossberg AJ, Scarlett JM, Marks DL. Hypothalamic mechanisms in cachexia.
Physiol Behav 2010;100:478–89.
 Asarian L, Langhans W. A new look on brain mechanisms of acute illness anorexia.
Physiol Behav 2010;100:464–71.
 Langhans W. Signals generating anorexia during acute illness. Proc Nutr Soc
 Porter MH, Hrupka BJ, Altreuther G, Arnold M, Langhans W. Inhibition of TNF-
alpha production contributes to the attenuation of LPS-induced hypophagia by
pentoxifylline. Am J Physiol Regul Integr Comp Physiol 2000;279:R2113–20.
 Romanatto T, Cesquini M, Amaral ME, Roman EA, Moraes JC, Torsoni MA, et al.
TNF-alpha acts in the hypothalamus inhibiting food intake and increasing the
respiratory quotient–effects on leptin and insulin signaling pathways. Peptides
 Figlewicz DP, Sipols AJ, Seeley RJ, Chavez M, Woods SC, Porte Jr D. Intraventricular
insulin enhances the meal-suppressive efficacy of intraventricular cholecystokinin
octapeptide in the baboon. Behav Neurosci 1995;109:567–9.
 Ikeda H, West DB, Pustek JJ, Figlewicz DP, Greenwood MR, Porte Jr D, et al.
Intraventricular insulin reduces food intake and body weight of lean but not obese
Zucker rats. Appetite 1986;7:381–6.
 Woods SC. The control of food intake: behavioral versus molecular perspectives.
Cell Metab 2009;9:489–98.
 Woods SC, Chavez M, Park CR, Riedy C, Kaiyala K, Richardson RD, et al. The
evaluation of insulin as a metabolic signal influencing behavior via the brain.
Neurosci Biobehav Rev 1996;20:139–44.
 Woods SC, D'Alessio DA. Central control of body weight and appetite. J Clin
Endocrinol Metab 2008;93:S37–50.
 Woods SC, Lotter EC, McKay LD, Porte Jr D. Chronic intracerebroventricular
infusion of insulin reduces food intake and body weight of baboons. Nature
 Woods SC, Porte Jr D. The role of insulin as a satiety factor in the central nervous
system. Adv Metab Disord 1983;10:457–68.
 De Souza CT, Araujo EP, Bordin S, Ashimine R, Zollner RL, Boschero AC, et al.
Consumption of a fat-rich diet activates a proinflammatory response and induces
insulin resistance in the hypothalamus. Endocrinology 2005;146:4192–9.
 Chavez M, Riedy CA, Van Dijk G, Woods SC. Central insulin and macronutrient
intake in the rat. Am J Physiol Regul Integr Comp Physiol 1996;271:R727–31.
 Posey KA, Clegg DJ, Printz RL, Byun J, Morton GJ, Vivekanandan-Giri A, et al.
Hypothalamic proinflammatory lipid accumulation, inflammation, and insulin
resistance in rats fed a high-fat diet. Am J Physiol Endocrinol Metab 2009;296:
 Amaral ME, Barbuio R, Milanski M, Romanatto T, Barbosa HC, Nadruz W, et al.
Tumor necrosis factor-alpha activates signal transduction in hypothalamus and
modulates the expression of pro-inflammatory proteins and orexigenic/anorex-
igenic neurotransmitters. J Neurochem 2006;98:203–12.
 Moraes JC, Amaral ME, Picardi PK, Calegari VC, Romanatto T, Bermudez-Echeverry
M, et al. Inducible-NOS but not neuronal-NOS participate in the acute effect of
TNF-alpha on hypothalamic insulin-dependent inhibition of food intake. FEBS Lett
 Woods SC, D'Alessio DA, Tso P, Rushing PA, Clegg DJ, Benoit SC, et al. Consumption
of a high-fat diet alters the homeostatic regulation of energy balance. Physiol
 Locksley RM, Killeen N, Lenardo MJ. The TNF and TNF receptor superfamilies:
integrating mammalian biology. Cell 2001;104:487–501.
 Kinouchi K, Brown G, Pasternak G, Donner DB. Identification and characterization
of receptors for tumor necrosis factor-alpha in the brain. Biochem Biophys Res
 Jaillard T, Roger M, Galinier A, Guillou P, Benani A, Leloup C, et al. Hypothalamic
reactive oxygen species are required for insulin-induced food intake inhibition: an
NADPH oxidase-dependent mechanism. Diabetes 2009;58:1544–9.
 Gao D, Nong S, Huang X, Lu Y, Zhao H, Lin Y, et al. The effects of palmitate on
hepatic insulin resistance are mediated by NADPH Oxidase 3-derived reactive
oxygen species through JNK and p38MAPK pathways. J Biol Chem 2010;285:
A.D. de Kloet et al. / Physiology & Behavior 103 (2011) 17–20