- Email: [email protected]
Influence of dietary fats on c-Fos-like immunoreactivity in mouse hypothalamus
Brain Research 843 Ž1999. 184–192 www.elsevier.comrlocaterbres
Research report
Influence of dietary fats on c-Fos-like immunoreactivity in mouse hypothalamus Hongqin Wang, Len H. Storlien, Xu-Feng Huang
)
Metabolic Research Center, Department of Biomedical Science, UniÕersity of Wollongong, Wollongong, NSW 2522, Australia Accepted 27 July 1999
Abstract The hypothalamus is a brain region of major importance in regulation of energy balance via autonomic nervous control of both intake and expenditure. There is substantial evidence that diets high in saturated fats lead to obesity while diets equally high in polyunsaturated fats ŽPUFAs. do not. Using c-Fos as a marker, this study aimed to investigate hypothalamic neuronal response in mice fed high fat diets Ž58% of calories as fat. emphasising saturated, n y 3 or n y 6 polyunsaturated fatty acids, or a low fat Ž10% of calories. diet over periods of 1 and 7 weeks. In addition, a 4-week ‘‘reversal’’ intervention with n y 3 polyunsaturated or low fat diet was undertaken in saturated fat-fed mice. Food intake and body weight were measured over the feeding periods. At 1, 7 and 11 weeks mice were killed, epididymal fat pad were weighed and brains were removed for quantitation of hypothalamic c-Fos-like immunoreactive ŽFLI. neurons. Weight gain, and epididymal fat pad weight, were highest on the saturated fat diet and lowest on the n y 3 diet despite similar food intakes Žepididymal fat weight at week 7: saturated fat, 622 " 48 mg; n y 6 fat 423 " 69; low fat 387 " 10, n y 3 fat 225 " 26.. Compared to a low fat diet, FLI neurons in the dorsal part of lateral hypothalamic ŽdLH. area was dramatically increased by saturated fat feeding Žq367% at 1 week. while ventromedial hypothalamic ŽVMH. activity was decreased. In contrast with n y 6 and n y 3 feeding dLH FLI neuronal activity was unchanged but actually increased in the VMH. Paraventricular nucleus ŽPVN. FLI neurons increased in the high saturated group only at 7 and 11 weeks, after substantial fat accumulation. Substitution of saturated fat diet with the n y 3 diet partially reversed Ž48%. the increase in FLI neurons in PVN of saturated fat-fed mice, while it significantly increase FLI neurons in arcuate nucleus Žq400%.. In summary, this study demonstrates that dietary saturated fat modulates hypothalamic neuronal activity in a pattern Žhigh lateral, reduced ventromedial activity. consistent with its obesogenic effects. In contrast, diets equally high in PUFA Žparticularly of the n y 3 class. neither increase adiposity nor derange the lateralrmedial neuronal activity balance. q 1999 Elsevier Science B.V. All rights reserved. Keywords: Hypothalamus; Obesity; Saturated fat; n y 3 Polyunsaturated fat; n y 6 Polyunsaturated fat
1. Introduction Body adiposity is governed by a balance over time of energy intake and energy expenditure w2,26,29x. A high intake of dietary fat is considered to predispose obesity via effects on both the intake and expenditure sides of the energy balance equation w19,30x. However, all fats do not AbbreÕiations: Arc, Arcuate hypothalamic nucleus; DMH, Dorsomedial hypothalamic nucleus; FLI, c-Fos-like immunoreactive; HS, High saturated fat diet; HS-LF, High saturated fat diet followed by low fat diet; HS-n3, High saturated fat diet followed by ny3 polyunsaturated fat diet; LF, Low fat diet; LH, Lateral hypothalamic area; MPO, Medial preoptic nucleus; PUFA, Polyunsaturated fat; PVN, Paraventricular nucleus of hypothalamus; SCh, Suprachiasmatic nucleus of hypothalamus; VMH, Ventromedial hypothalamic nucleus ) C orresponding author. Fax: q 61-42-214096; e-m ail: [email protected]
impact on the energy balance equation in the same manner. Data have now accumulated which clearly link the obesogenic attributes of dietary fat to saturated fats while polyunsaturated fatty acids are relatively neutral or perhaps even protective w27x. A series of studies from Matsuo et al., as well as our group and others, have shown in rodents that compared to a polyunsaturated fat ŽPUFA.-enriched diet, one high in saturated fats leads to increased body fatness with a number of potential mechanisms identified w20–22,25,27x. Data from an increasing number of human studies is consistent with the animal literature w5,7,15x. Taken together, research is now starting to show a consistent pattern whereby saturated fat intake is closely associated with development of obesity with lower metabolic rate, decreased sympathetic nervous system activity and stiffer membranes with reduced binding capacities. The reduced beta-adrenergic
0006-8993r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved. PII: S 0 0 0 6 - 8 9 9 3 Ž 9 9 . 0 1 9 2 7 - 7
H. Wang et al.r Brain Research 843 (1999) 184–192
receptor binding, both centrally and peripherally w21,23x, is interesting particularly in relation to new evidence for improvement in insulin action and increased fat oxidation following treatment by a selective b3-adrenoceptor agonist in humans w31x. In contrast to the effects of saturated fats, PUFA intake is not, or only very weakly, associated with obesity and the pattern of changes in metabolic rate, etc., is opposite that of saturated lipids w27x. While peripheral mechanisms, such as changes in membrane lipid composition, may account for some of the effects of the dietary fat subtypes, it is equally likely that they are exerting their differential effects on metabolism via direct effects on the brain’s neuronal systems influencing energy balance. In particular, the hypothalamus plays a critical role in the regulation of energy homeostasis including energy intake and expenditure w2,3,11,14,20,26x. The dual centre hypothesis has been historically influential in focusing on the lateral hypothalamus as a feeding centre and the ventromedial area as a satiety centre. Recent, exciting developments have provided extensive new information on neuropeptides subtending this overall conceptual framework w4,9,32x. The aim of this study is to test the hypothesis that modulating the level and fatty acid profile of dietary fat will alter activity in hypothalamic neurons critically involved in the regulation of adiposity. Further, we hypothesise that those changes in neuronal activity will be anatomically discrete, directionally selective, and consistent with the obesogenic effects of saturated, but not polyunsaturated, fats. Neuronal activity was indexed by c-Fos-like immunoreactivity.
185
2. Materials and methods 2.1. Animals Three-week-old C57Blr6J male mice were obtained from Animal Resource Centre ŽPerth, Australia; the mouse genomic background was tested by allozyme electrophoresis and RAPD-PCR analysis.. They were housed individually in a temperature-controlled room Ž208C " 28C. with a 12:12-h light–dark cycle Žlights on at 0700 h.. Mice were given ad libitum access to tap water throughout the study. 2.2. Diet and experimental procedure Mice were fed standard laboratory chow for the first week to allow them to adapt to the new environment. They were subsequently randomly assigned to one of the groups shown in Fig. 1. In all mice fed a high fat diet, fat contributed 58% of calories. In mice fed a low fat diet, fat contributed 10% of calories. In mice fed a high saturated fat diet, edible tallow Ž528C melting point; Unilever, Australia. and safflower oil ŽMeadow Lea Foods, Australia. contributed equally to fat calories. In mice fed a high n y 3 PUFA diet, fish oil ŽEPA-28; from Yamanouchi Pharmaceutical, Japan. contributed all 58% of fat calories. In mice fed a high n y 6 PUFA diet, safflower oil contributed all 58% of calories. The saturated fat diet was thus high in saturated fat with a high n y 6rn y 3 PUFA ratio. The n y 3 diet was almost equally high in saturated fat but
Fig. 1. Experiment design and assignment of mice used in this study ŽPUFA: polyunsaturated fat diet..
H. Wang et al.r Brain Research 843 (1999) 184–192
186
Table 1 Composition of high saturated ŽHS., n y 3 polyunsaturated Ž n y 3., n y 6 polyunsaturated Ž n y 6., and low fat ŽLF. diets Ingredients
HS Žgrkg.
n y 3 Žgrkg.
n y 6 Žgrkg.
LF Žgrkg.
Gelatin Methionine Bran Minerals Vitamins Cornstarch Sucrose Oil Beef tallow Safflower oil Fish oil Casein Total energy Žkcalr100 g. Fat Ž%. Protein Ž%. Carbohydrate Ž%. Fatty acid composition Ž%. Ý Saturated Ý Monounsaturated Ý n y 6 polyunsaturated Ý n y 3 polyunsaturated
19 3 51 67 13 169 85
19 3 51 67 13 169 85
19 3 51 67 13 169 85
14 2.3 38 50 9.7 438 219
169.5 169.5 – 254 523 58 21 21
– – 339 254 523 58 21 21
– 339 – 254 523 58 21 21
– 41 – 188 365 10 22 68
31 27 40 0.3
24.5 22.8 11.4 42.0
with a high level of n y 3 PUFAs and very low n y 6rn y 3 ratio. The n y 6 diet was the most unsaturated but with a very high n y 6rn y 3 ratio. Detailed composition of the respective diets can be seen in Table 1. Diets were freshly made every week and stored at 48C. Mice were given food at 1600 h each day. Food consumption was measured daily and body weight was measured weekly. 2.3. Histology All mice were sacrificed by an overdose of sodium pentobarbitone anaesthesia Ž120 mgrkg i.p.. at a fixed time between 0700 and 1000 h to minimize variation due to circadian rhythm fluctuation in c-Fos protein. Care was taken not to stress the mice before induction of anaesthesia. The total carcass was weighed and epididymal adipose tissue dissected free and also weighed. The brains were fixed by perfusion through the left ventricle of the heart with 50 ml saline, followed by cold 4% paraformaldehyde made in 0.1 M phosphate buffer ŽPB, pH 7.4.. The brains were then removed and stored in 30% sucrose Žin PB. for two nights. The brains were frozen in liquid nitrogen and cryostat sectioned into 25 mm thickness coronal sections at y158C. Sets of consecutive sections were taken. The sections were used for the detection of c-Fos-like immunoreactive ŽFLI. neurons and for Nissl staining for accurate identification of neuroanatomical structures according to the mouse atlas of Franklin and Paxinos w8x. Sections were screened for visualisation of FLI neurons using an Olympus BX50 microscope. In regions where differences in FLI neurons were readily apparent, Fos-positive nuclei were counted following the area was outlined with a drawing tool of Magellan Com-
9.0 12.1 78.1 0.6
9.0 12.1 78.1 0.6
puter Program w13x. Every third section was counted for quantification of FLI neurons. The value represents an average of FLI neurons in a given area of one hemisphere in a single section. FLI neurons were based on relative size and dark stained nuclei. 2.4. Immunohistochemistry The detailed descriptions of technical procedures for immunohistochemistry were described previously w14x. Briefly, the brain sections were firstly washed in 0.8% HCl in 50% alcohol for 30 min for the abolition of endogenous peroxidase activity. The primary antibody, rabbit anti-c-Fos ŽArnel-Lab, US., was diluted at 1:8000 in PBS containing 0.1% Triton X-100 ŽPBS-Triton., and then incubated at room temperature overnight. After three washings in PBSTriton for 10-min each, sections were incubated in the biotinylated secondary antibody ŽBoehringer Mannheim, Germany., diluted at 1:250 in Tris–HCl for 1 h. Sections were then incubated in the avidin–biotin–peroxidase complex ŽSigma, US. diluted at 1:100 in PBS-Triton, for 2 h. Sections were washed in Tris–HCl for 30 min before reacting them with 0.05% diaminobenzidine tetrahydrochloride ŽSigma. in Tris–HCl ŽpH 7.4. and 0.007% hydrogen peroxide in the presence for 0.04% nickel ammonium sulphate. The specificity of the antisera has been tested previously w14x and the immunohistochemical reactions were tested by omission of the primary antisera. No peroxidase reaction occurred in these experiments. 2.5. Data analysis A microscope driven by an IBM compatible computer with the application of the Magellan Computer Program
H. Wang et al.r Brain Research 843 (1999) 184–192
187
Fig. 2. Energy intake Žkcalrday per mouse. in mice fed a high saturated ŽHS., n y 3 polyunsaturated Ž n y 3., n y 6 polyunsaturated Ž n y 6. or low fat ŽLF. diet; or mice fed a high n y 3 polyunsaturated ŽHS-n3. or low fat ŽHS-LF. diet for 4 weeks following 7-week high saturated fat diet.
w13x was used to obtain data concerning counting positively stained FLI neurons. Results are presented as means " S.E.M. The data were analysed with analysis of variance ŽANOVA.. Means were considered significantly different when P F 0.05.
3. Results 3.1. Energy intake, body weight gain, and epididymal fat storage The energy intake curve is shown in Fig. 2. Cumulative food intake over the 7 weeks of feeding showed the low-fat group to have a significantly lower Ž P F 0.05. intake than the three high-fat groups which did not differ. Body weight gain was highest in the high saturated fat diet group over the 7 weeks Ž11.7 " 0.6 g. compared to 11.1 " 1.0 g in the low-fat group, 10.0 " 0.6 g in the n y 6 PUFA group and 7.8 " 0.6 g in the n y 3 PUFA group Ž n y 3 group differs from all other groups, P F 0.01.. From weeks 7 to 11 the group which stayed on a high saturated fat diet gained a further 2.2 " 0.3 g Žvs. 1.3 " 0.0 g for the continuing low fat group. while the Table 2 Distribution of c-Fos-like immunoreactive neurons in the hypothalamus of mice fed a high saturated ŽHS., ny3 polyunsaturated Ž ny3., ny6 polyunsaturated Ž ny6. or low fat ŽLF. diet for 1 week HS Ž ns6. Arc DMH VMH dLH PVN SCh MPO U
2"1 6"1 1"0U 22"2U 5"1 5"1U 7"5
ny3 Ž ns6. U
0"0 10"3 5"1a 7"2 a 5"1 2"1a 5"1
ny6 Ž ns6.
LF Ž ns6.
1"0 5"1 5"1a 6"2 a 2"0 5"1U 4"1
4"1 8"1 3"1 6"2 3"1 2"1 5"1
P F 0.05 compared with low fat diet. P F 0.05 compared with high saturated fat diet.
a
group switched from high saturated fat to n y 3 PUFA lost 0.9 " 0.3 g. Interestingly, the group switch from high saturated fat to low fat continued to gain weight Ž1.7 " 0.4 g; weight change different from the saturated fat to n y 3 PUFA group, P F 0.05.. After 7 and 11 weeks of feeding, epididymal fat stores on the high saturated fat diet Ž622 " 48 mg, 814 " 73 mg. were almost double those on the low fat diet Ž387 " 10 mg, 433 " 43 mg; P F 0.01.. In contrast, the n y 6 PUFA group Ž423 " 69 mg. was equivalent to the low fat group at 7 weeks while the n y 3 PUFA group had accumulated significantly less Ž225 " 26 mg. epididymal fat. Shifting from a high saturated fat diet to low fat from weeks 7 to 11 saw no significant difference in epididymal fat storage. In contrast, a marked 65% decline in epididymal fat was seen after the switch from saturated to n y 3 PUFA diet during weeks 7 to 11. 3.2. Hypothalamic FLI neurons Present study examined FLI neurons in every third section of hypothalamus of mice fed either a saturated fat, n y 3 PUFA, n y 6 PUFA or low fat diet. Only those nuclei showing significant differences between groups are Table 3 Distribution of c-Fos-like immunoreactive neurons in the hypothalamus of mice fed a high saturated ŽHS., ny3 polyunsaturated Ž ny3., ny6 polyunsaturated Ž ny6. or low fat ŽLF. diet for 7 weeks
Arc DMH VMH dLH PVN SCh MPO U
HS Ž ns6.
ny3 Ž ns6.
ny6 Ž ns6.
LF Ž ns6.
1"0 12"1 15"3 27"2U 13"3U 5"1 12"2U
1"1 10"2 13"2 7"1a 3"0 a 1"0 a 7"1a
0"0 9"2 13"3 7"2 a 5"1 5"0 9"1
1"0 15"2 12"2 8"4 2"0 2"1 6"1
P F 0.05 compared with low fat diet. P F 0.05 compared with high saturated fat diet.
a
H. Wang et al.r Brain Research 843 (1999) 184–192
188
Table 4 Distribution of c-Fos-like immunoreactive neurons in the hypothalamus of mice fed a high saturated ŽHS. or low fat ŽLF. diet for 11 weeks, or mice fed a high ny3 polyunsaturated ŽHS-n3. or low fat ŽHS-LF. diet for 4 weeks following 7-week high saturated fat diet HS Ž ns6. Arc DMH VMH dLH PVN SCh MPO U
2"1 20"2 25"4 24"4U 21"0U 4"1 13"2U
LF Ž ns6. 0"0 14"1 13"4 9"2 3"1 4"1 5"1
HS-n3 Ž ns6. Ua
8"1 25"4 18"4 21"3U 11"3a 4"1 10"1
HS-LF Ž ns6. 6"2U a 22"3 16"6 34"2U 20"5U 4"0 12"2U
P F 0.05 compared with low fat diet. P F 0.05 compared with high saturated fat diet.
a
presented. Those areas, showing FLI neurons but not showing diet-induced differences, were medial tuberal nucleus, hypothalamic intermediate zone, supramammillary nuclei and medial preoptic area. 3.2.1. Lateral hypothalamus (LH) Mice fed a high saturated fat diet showed a dramatically increased FLI neurons in the dorsal Žabove a line through the fornix and perpendicular to the midline of the section.
part of the LH ŽdLH; Tables 2–4; Fig. 3.. Compared with a low fat diet, a high saturated fat diet increased FLI neurons in dLH by 367% in 1 week, 338% in 7 weeks and 267% in 11 weeks. In contrast, neither the n y 3 nor n y 6 PUFA diets altered dLH FLI neurons compared to the low fat group at any time point. 3.2.2. ParaÕentricular hypothalamic nucleus (PVN) The most distinctive feature of the PVN response to a high saturated fat diet was a delayed activation of FLI neurons ŽFigs. 4 and 5.. As it is shown in Fig. 5, there were no significant differences in the number of PVN FLI neurons between groups at 1 week ŽTable 2.. However, compared to the low fat group, a significant, 6.5 to 7-fold, increase in FLI neurons was observed after 7 and 11 weeks of a high saturated fat diet ŽTables 3 and 4.. In contrast, PVN FLI neurons Ž7 weeks. were only marginally increased in the n y 3 PUFA group while there was a 2.5-fold increase in the n y 6 PUFA group, the increase was still much less than that in the saturated fat group. Replacing a high saturated fat diet with a high n y 3 PUFA diet dramatically reduced Ž48%. FLI neurons. However, replacing a high saturated fat diet with a low fat diet had no significant effect.
Fig. 3. Photomicrographs showing c-Fos-like immunoreactive neurons in the dorsal part of lateral hypothalamic area ŽdLH. of the mice fed: ŽA. a high saturated ŽHS., ŽB. low fat ŽLF., ŽC. n y 3 polyunsaturated Ž n y 3., or ŽD. n y 6 polyunsaturated Ž n y 6. diet. DMH s dorsomedial hypothalamic nucleus; 3V s third ventricle; f s fornix.
H. Wang et al.r Brain Research 843 (1999) 184–192
189
Fig. 4. Photomicrographs showing c-Fos-like immunoreactive neurons in the paraventricular hypothalamic nucleus ŽPVN. of the mice fed a high saturated fat diet for 1 week ŽA., 7 weeks ŽB., 11 weeks ŽC., or low fat diet for 7 weeks ŽD.. Changing the diet from a high saturated fat to n y 3 polyunsaturated fat ŽF. reduces the number of c-Fos-like immunorective neurons in PVN, but not in a low fat ŽE. replacement.
3.2.3. Ventromedial hypothalamic nucleus (VMH) The VMH showed significant differences in FLI neurons only after 1 week of various diets ŽTable 2.. VMH FLI neurons were inhibited after 1 week of a saturated fat diet but activated equivalently by both PUFA diets. Changing diets from a high saturated fat diet into either low fat or high n y 3 PUFA had no significant influence on FLI neurons in the VMH.
Compared with a low fat diet, a high n y 6 PUFA diet significantly increased 2.5-fold FLI neurons in the SCh at both 1 and 7 weeks, similar to the mice fed a high saturated fat diet ŽTables 2 and 3.. Conversely, compared with a saturated fat diet, n y 3 PUFA diet had significant less FLI neurons in SCh in both 1 and 7 weeks, similar to the mice fed low fat diet. Changing the diet from a high
3.2.4. Medial preoptic nucleus (MPO) The MPO showed no difference in FLI neurons after 1 week of various diets. However, MPO showed a significant 2-fold increase in FLI neurons after 7 weeks and 2.6-fold after 11 weeks in the mice fed a saturated fat diet compared with low fat diet ŽTables 3 and 4.. Again, there were only marginal increases in MPO FLI neurons in the PUFA groups compared to the low fat-fed mice. Changing diets from a high saturated fat diet into either low fat or high n y 3 PUFA had no significant influence in FLI neurons in MPO. 3.2.5. Suprachiasmatic nucleus of hypothalamus (SCh) A distinctive feature of SCh was that n y 3 PUFA and n y 6 PUFA diets had opposing effects on this nucleus.
Fig. 5. The number of c-Fos-like immunoreactive neurons in the paraventricular nucleus of hypothalamus of the mice fed a high saturated ŽHS., ny3 polyunsaturated Ž ny3., ny6 polyunsaturated Ž ny6., or low fat ŽLF. diet; or mice fed a high ny3 polyunsaturated ŽHS-n3. or low fat ŽHS-LF. diet for 4 weeks following 7-week high saturated fat diet.
190
H. Wang et al.r Brain Research 843 (1999) 184–192
saturated fat to either high n y 3 PUFA or low fat diet had no significant affects on FLI neurons in SCh. 3.2.6. Arcuate hypothalamic nucleus (Arc) There were no differences in amount of FLI neurons in Arc between a saturated fat, n y 3 PUFA n y 6 PUFA, and low fat diet groups either at 1 or 7 weeks. However, changing of the diet from a high saturated fat to high n y 3 PUFA or low fat diet resulted in a large increase in FLI neurons in Arc Ž4- and 3-fold increases, respectively; Table 4.. 4. Discussion The present study was designed to examine hypothalamic neuronal response to high fat diets emphasising either saturated fat with a low n y 6rn y 3 PUFA ratio, high n y 3 PUFA or high n y 6 PUFA. Using c-Fos-like immunoreactivity as a marker, the most salient findings of this study were that: Ž1. compared to a low fat diet, a saturated fat diet increases FLI neurons primarily in the dLH and it does so in a time frame that precedes detectable increases in adiposity, Ž2. PUFA diets do not increase FLI neurons in the dLH, Ž3. a saturated fat diet decreases, while PUFA diet increases, FLI neurons in the VMH within 1 week of feeding, and Ž4. a saturated fat diet increased FLI neurons in the PVN but only after an increase in adiposity is detected. Overall, these results are consistent with direct effects of saturated fats to increase lateral, while decreasing ventromedial, hypothalamic neuronal activity. These effects would tune metabolism towards hyperphagia, decreased metabolic rate and weight gain. In contrast, equally high fat diets emphasising PUFAs neither increased lateral, nor decreased ventromedial, hypothalamic neuronal activity. The regulation of energy balance is complex, involving an integration of the internal milieu with environmental forces. The brain, and in particular the hypothalamus, acts as a final integrative centre and modulates energy homeostasis via effects on both intake and expenditure. Impressive strides have been made in the last few years in our understanding of both the neuroanatomy and key peptide regulators of energy balance. However, it is still true that while there is strong evidence of vigorous metabolic defence against energy deficit, the converse is not true. Diets high in saturated fats are particularly obesogenic in many species and strains. The current study offers new insights into mechanisms underlying this observation. The observation of dLH activation by a high saturated fat diet is critical. It is well documented that stimulation of the LH increases food intake and parasympathetic activity w12,18x. Neurons in the dLH send large numbers of descending projections to the dorsal motor nucleus of vagus, solitary nucleus, lateral parabrachial nucleus and lateroventral periaquiduct gray w33x. All these areas are involved in regulation of autonomic function. In addition, the neurons
in the dLH project to the central nucleus of amygdala, which is a key relay center regulating autonomic function. Lesions to the lateral hypothalamus result in a decrease in parasympathetic activity and, conversely, an increase in sympathetic activity, accompanied by decreased food intake and lower adiposity and increased oxygen consumption and core temperature w1,3,16x. The present study also showed that mice fed diets high n y 3 PUFA or n y 6 PUFA, like low fat diets, did not increase FLI neurons in the dLH. This suggests that saturated fat has a selective effect on the dLH, which may explain its obesogenic effects. It is interesting that the dLH showed this increase in FLI neurons within only 1 week on the saturated fat diet. It has been reported that C57 mice show no difference in the levels of blood glucose, insulin and leptin after 1 week on a high fat diet w30x. Taken together, these data are suggestive of a direct influence of saturated fat on dLH neuronal activity. It is also noted that mice fed the saturated fat diet for 1 week had the lowest number of FLI neurons in the VMH compared with all other groups. The VMH lesion-induced obesity syndrome has been well characterised w2x. Therefore, a less active VMH in the saturated fat diet mice would also contribute to development of obesity. Historically, the LH and VMH have been seen as key hypothalamic areas reciprocally controlling energy balance. It is important in the current results that saturated fat diet decreased FLI neurons in the VMH at an early stage. It is interesting to note the change in FLI neurons in various areas over time independent of diet. For example, FLI neurons in the VMH increased dramatically from week 1 to week 7 in all groups. The dLH increase was much smaller but again consistent over groups. However, in the PVN, the increase was restricted to the high saturated fat group. Historically, the lateral hypothalamus has been linked to the parasympathetic, and the ventromedial hypothalamus to the sympathetic nervous system. The balance between these two arms of the autonomic nervous system were then seen to tune metabolism with a high parasympatheticrsympathetic ratio linked to positive energy balance and a low ratio to weight loss. If the ratio of the FLI neurons in the dLH to VMH is then used as an index of parasympatheticrsympathetic balance, it can be seen that it is high at about 2r1 in the low fat group at 1 week of feeding Ži.e., in very young, 5-week-old mice. which makes sense in the rapid growth phase of early life. However, by 7 weeks Ž11-week-old mice., the ratio has dropped to 0.67r1. Interestingly, the ratio is, if anything, lower in the PUFA high-fat groups, but dramatically elevated in the high saturated fat group at 1 week Ž22r1. and still almost 2r1 at 7 weeks of feeding. This is entirely consistent with the very high rate of weight and fat accumulation in the high saturated fat group. It is also consistent with the low metabolic rate and reduced sympathetic nervous system activity previously reported for mice on a high saturated fat diet w21,22x.
H. Wang et al.r Brain Research 843 (1999) 184–192
In this study, we showed no significant change in FLI neurons in the PVN after 1 week on a saturated fat diet. However, there was a dramatic increase in FLI neurons in PVN after 7 and 11 weeks of saturated fat feeding; Ži.e., medium to long term w7 to 11 weeksx, but not short term w1 weekx, high saturated fat diet alters FLI neurons in PVN.. By 7 and 11 weeks, there was an increase in body fat storage and plasma leptin, which were not seen after 1 week of the saturated fat diet. Taken together, this suggests that increased PVN neuronal activation is not due to direct effects of saturated fat but may reflect an attempt by the central energy balance regulatory apparatus to limit fat stores. It is known from previous reports that the PVN– hindbrain pathway can inhibit feeding and reduce high fat diet-induced obesity w17x. This PVN–hindbrain feeding inhibitory pathway descends primarily from the PVN, turns caudally to take a longitudinal path passing through the hypothalamic perifornical area and innervates the dorsal motor nucleus of vagus nerve and solitary nucleus w10x. Therefore, FLI neurons in PVN in the high saturated fat diet-induced obese mice may act to reduce fat storage via negative feedback influenced by amount of fat stored. This is further supported by the demonstration that substituting saturated fat with n y 3 PUFA diet in already obese mice, which we have showed in this study to reduces visceral fat storage as well as FLI neurons in PVN. Therefore, it is suggested that n y 3 PUFA may firstly reduce fat storage followed by withdrawing feedback regulation to the PVN. There are distinctive patterns of changes between dietary groups in other hypothalamic areas that are worthy of comment. The present study showed that the MPO had a significant increase in FLI neurons after high saturated fat diet for 7 and 11 weeks, but not in any other dietary groups. It has previously been reported that the MPO galaninergic neurons respond to high fat diet as showing increased mRNA and protein production w18x. Therefore, it is possible that those activated MPO neurons are galaninergic. Second, changing from high-saturated fat to either a n y 3 or low fat diet induced a dramatic rise in FLI neurons in the arcuate nucleus. This area is selectively lesioned by monosodium glutamate administration w24x with a resultant endocrine syndrome which involves multiple defects including stunting, obesity and a high degree of metabolic efficiency. The arcuate nucleus is known to contain leptin receptors and monosodium glutamate-treated rats do not lose either weight or fat mass when treated with leptin w6x. The role of the arcuate area in the dramatic reversal of the high saturated fat effects by n y 3 Žand to a much lesser extent by a low fat diet. needs exploring in relation to possible direct fatty acid effects on arcuate leptin receptor activation and downstream signalling. Finally, there were differences between groups feed high n y 3 or n y 6 PUFAs with the n y 6 outcomes generally somewhere between the effects of a high saturated fat diet and the n y 3 diet group. The suprachiasmatic area ŽSCh. is one region of clear difference between
191
the two PUFA diet groups. Interestingly, the SCh is an NPY receptor-containing area where it has been shown that, at least under some conditions, NPY-immunoreactivity is inversely proportional to food intake w28x. Future work aimed at investigating the effects of dietary fats on NPY expression in SCh would appear warranted. In summary, this study showed that saturated fat diet can selectively, and may directly, activate LH, and inhibit VMH, neuronal activity. This pattern of hypothalamic activity may play a critical role in the strongly obesogenic properties of dietary saturated fat.
Acknowledgements We would like to thank ArProfessors Peter McLennan and Arthur Jenkins for helpful discussion and suggestions during this study. This study was supported by grants from the NHMRC of Australia.
References w1x L.L. Bernardis, L.L. Bellinger, The lateral hypothalamic area revisited: neuroanatomy, body weight regulation, neuroendocrinology and metabolism, Neurosci. Biobehav. Rev. 17 Ž1993. 141–193. w2x G.A. Bray, The nutrient balance hypothesis: peptides, sympathetic activity, and food intake, Ann. N. Y. Acad. Sci. 676 Ž1993. 223–241. w3x G.A. Bray, S. Inoue, Y. Nishizawa, Hypothalamic obesity. The autonomic hypothesis and the lateral hypothalamus, Diabetologia 20 Ž1981. 366–377. w4x L.A. Campfield, F.J. Smith, P. Burn, Strategies and potential molecular targets for obesity treatment, Science 280 Ž1998. 1383–1387. w5x C. Couet, J. Delarue, P. Ritz, J.M. Antione, F. Lamisse, Effect of dietary fish oil on body fat mass and basal fat oxidation in healthy adults, Int. J. Obes. 21 Ž1997. 637–643. w6x R. Dawson, M.A. Pelleymounter, W.J. Millard, S. Liu, B. Eppler, Attenuation of leptin-mediated effects by monosodium glutamate-induced arcuate nucleus damage, Am. J. Physiol. Ž1997. E202–E206. w7x E. Doucet, N. Almeras, M.D. White, J.P. Despres, C. Bouchard, A. Tremblay, Dietary fat composition and human adiposity, Eur. J. Clin. Nutr. 52 Ž1998. 2–6. w8x K.B.J. Franklin, G. Paxinos, The Mouse Brain in Stereotaxic Coordinates, 1st edn., Academic Press, San Diego, 1997. w9x J.M. Friedman, The alphabet of weight control, Nature 385 Ž1997. 119–120, news. w10x R.M. Gold, A.P. Jones, P.E. Sawchenko, Paraventricular area: critical focus of a longitudinal neurocircuitry mediating food intake, Physiol. Behav. 18 Ž1977. 1111–1119. w11x J.L. Halaas, K.S. Gajiwala, M. Maffei, S.L. Cohen, B.T. Chait, D. Rabinowitz, R.L. Lallone, S.K. Burley, J.M. Friedman, Weight-reducing effects of the plasma protein encoded by the obese gene, Science 269 Ž1995. 543–546, see comments. w12x R.E. Keesey, M.D. Hirvonen, Body weight set-points: determination and adjustment, J. Nutr. 127 Ž1997. 1875S–1883S. w13x X.F. Huang, I. Tork, G. Paxinos, Dorsal motor nucleus of the vagus nerve: a cyto- and chemoarchitectonic study in the human, J. Comp. Neurol. 330 Ž1993. 158–182. w14x X.F. Huang, H.Q. Wang, Altered c-fos expression in autonomic regulatory centers of genetically obese Žobrob. mouse brain, Brain Res. 799 Ž1998. 307–310.
192
H. Wang et al.r Brain Research 843 (1999) 184–192
w15x P.J. Jones, J.E. Ridgen, P.T. Phang, C.L. Birmingham, Influence of dietary fat polyunsaturated to saturated ratio on energy substrate utilization in obesity, Metabolism 41 Ž1992. 396–401. w16x M.D. Hirvonen, R.E. Keesey, Chronically altered body protein levels following lateral hypothalamic lesions in rats, Am. J. Physiol. 270 Ž1996. R738–R743. w17x A.L. Kirchgessner, A. Sclafani, PVN–hindbrain pathway involved in the hypothalamic hyperphagia-obesity syndrome, Physiol. Behav. 42 Ž1988. 517–528. w18x S.F. Leibowitz, Brain peptides and obesity: pharmacologic treatment, Obes. Res. 3 Ž1995. 573S–589S. w19x J.A. Levine, N.L. Eberhardt, M.D. Jensen, Role of nonexercise activity thermogenesis in resistance to fat gain in humans, Science 283 Ž1999. 212–214, see comments. w20x S. Lin, X.F. Huang, Altered hypothalamic c-Fos expression in diet-induced obese mice, Brain Res. Bull., in press. w21x T. Matsuo, H. Sumida, M. Suzuki, Beef tallow diet decreases beta-adrenergic receptor binding and lipolytic activities in different adipose tissues of rat, Metabolism 44 Ž1995. 1271–1277. w22x T. Matsuo, H. Sumida, M. Suzuki, Beef tallow diet decreases norepinephrine turnover rates in rat hypothalamus and cerebral cortex, Metabolism 44 Ž1995. 1377–1379. w23x T. Matsuo, M. Suzuki, Brain beta-adrenergic receptor binding in rats with obesity induced by a beef tallow diet, Metabolism 46 Ž1997. 18–22. w24x J.W. Olney, Status of monosodium glutamate revisited, Am. J. Clin. Nutr. 26 Ž1973. 683–685.
w25x D.A. Pan, A.J. Hulbert, L.H. Storlien, Dietary fats, membrane phospholipids and obesity, J. Nutr. 124 Ž1994. 1555–1565. w26x F. Rohner-Jeanrenaud, B. Jeanrenaud, Central nervous system and body weight regulation, Ann. Endocrinol. 58 Ž1997. 137–142. w27x L.H. Storlien, A.J. Hulbert, P.L. Else, Polyunsaturated fatty acids, membrane function and metabolic diseases such as diabetes and obesity, Clin. Nutr. Metab. Care 1 Ž1998. 559–563. w28x P.L. Selvais, C. Labuche, X.N. Nguyen, J.M. Ketelslegers, J.F. Denef, D.M. Maiter, Cyclic feeding behaviour and changes in hypothalamic galanin and neuropeptide Y gene expression induced by zinc deficiency in the rat, J. Neuroendocrinol. 9 Ž1997. 55–62. w29x L.H. Storlien, A.D. Kriketos, G.D. Calvert, L.A. Baur, A.B. Jenkins, Fatty acids, triglycerides and syndromes of insulin resistance, Prostaglandins, Leukotrienes Essent. Fatty Acids 57 Ž1997. 379–385. w30x R.S. Surwit, A.E. Petro, P. Parekh, S. Collins, Low plasma leptin in response to dietary fat in diabetes- and obesity-prone mice, Diabetes 46 Ž1997. 1516–1520. w31x C. Weyer, P.A. Tataranni, S. Snitker, E.J. Danforth, E. Ravussin, Increase in insulin action and fat oxidation following treatment with CL316.243, a highly selective beta 3-adrenoceptor agonist in humans, Diabetes 47 Ž1998. 1555–1561. w32x S.C. Woods, R.J. Seeley, D. Porte Jr., M.W. Schwartz, Signals that regulate food intake and energy homeostasis, Science 280 Ž1998. 1378–1383. w33x S.A. Zardetto, M.M. Moga, D.J. Magnuson, T.S. Gray, Lateral hypothalamic dynorphinergic efferents to the amygdala and brainstem in the rat, Peptides 9 Ž1988. 1121–1127.
Research report
Influence of dietary fats on c-Fos-like immunoreactivity in mouse hypothalamus Hongqin Wang, Len H. Storlien, Xu-Feng Huang
)
Metabolic Research Center, Department of Biomedical Science, UniÕersity of Wollongong, Wollongong, NSW 2522, Australia Accepted 27 July 1999
Abstract The hypothalamus is a brain region of major importance in regulation of energy balance via autonomic nervous control of both intake and expenditure. There is substantial evidence that diets high in saturated fats lead to obesity while diets equally high in polyunsaturated fats ŽPUFAs. do not. Using c-Fos as a marker, this study aimed to investigate hypothalamic neuronal response in mice fed high fat diets Ž58% of calories as fat. emphasising saturated, n y 3 or n y 6 polyunsaturated fatty acids, or a low fat Ž10% of calories. diet over periods of 1 and 7 weeks. In addition, a 4-week ‘‘reversal’’ intervention with n y 3 polyunsaturated or low fat diet was undertaken in saturated fat-fed mice. Food intake and body weight were measured over the feeding periods. At 1, 7 and 11 weeks mice were killed, epididymal fat pad were weighed and brains were removed for quantitation of hypothalamic c-Fos-like immunoreactive ŽFLI. neurons. Weight gain, and epididymal fat pad weight, were highest on the saturated fat diet and lowest on the n y 3 diet despite similar food intakes Žepididymal fat weight at week 7: saturated fat, 622 " 48 mg; n y 6 fat 423 " 69; low fat 387 " 10, n y 3 fat 225 " 26.. Compared to a low fat diet, FLI neurons in the dorsal part of lateral hypothalamic ŽdLH. area was dramatically increased by saturated fat feeding Žq367% at 1 week. while ventromedial hypothalamic ŽVMH. activity was decreased. In contrast with n y 6 and n y 3 feeding dLH FLI neuronal activity was unchanged but actually increased in the VMH. Paraventricular nucleus ŽPVN. FLI neurons increased in the high saturated group only at 7 and 11 weeks, after substantial fat accumulation. Substitution of saturated fat diet with the n y 3 diet partially reversed Ž48%. the increase in FLI neurons in PVN of saturated fat-fed mice, while it significantly increase FLI neurons in arcuate nucleus Žq400%.. In summary, this study demonstrates that dietary saturated fat modulates hypothalamic neuronal activity in a pattern Žhigh lateral, reduced ventromedial activity. consistent with its obesogenic effects. In contrast, diets equally high in PUFA Žparticularly of the n y 3 class. neither increase adiposity nor derange the lateralrmedial neuronal activity balance. q 1999 Elsevier Science B.V. All rights reserved. Keywords: Hypothalamus; Obesity; Saturated fat; n y 3 Polyunsaturated fat; n y 6 Polyunsaturated fat
1. Introduction Body adiposity is governed by a balance over time of energy intake and energy expenditure w2,26,29x. A high intake of dietary fat is considered to predispose obesity via effects on both the intake and expenditure sides of the energy balance equation w19,30x. However, all fats do not AbbreÕiations: Arc, Arcuate hypothalamic nucleus; DMH, Dorsomedial hypothalamic nucleus; FLI, c-Fos-like immunoreactive; HS, High saturated fat diet; HS-LF, High saturated fat diet followed by low fat diet; HS-n3, High saturated fat diet followed by ny3 polyunsaturated fat diet; LF, Low fat diet; LH, Lateral hypothalamic area; MPO, Medial preoptic nucleus; PUFA, Polyunsaturated fat; PVN, Paraventricular nucleus of hypothalamus; SCh, Suprachiasmatic nucleus of hypothalamus; VMH, Ventromedial hypothalamic nucleus ) C orresponding author. Fax: q 61-42-214096; e-m ail: [email protected]
impact on the energy balance equation in the same manner. Data have now accumulated which clearly link the obesogenic attributes of dietary fat to saturated fats while polyunsaturated fatty acids are relatively neutral or perhaps even protective w27x. A series of studies from Matsuo et al., as well as our group and others, have shown in rodents that compared to a polyunsaturated fat ŽPUFA.-enriched diet, one high in saturated fats leads to increased body fatness with a number of potential mechanisms identified w20–22,25,27x. Data from an increasing number of human studies is consistent with the animal literature w5,7,15x. Taken together, research is now starting to show a consistent pattern whereby saturated fat intake is closely associated with development of obesity with lower metabolic rate, decreased sympathetic nervous system activity and stiffer membranes with reduced binding capacities. The reduced beta-adrenergic
0006-8993r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved. PII: S 0 0 0 6 - 8 9 9 3 Ž 9 9 . 0 1 9 2 7 - 7
H. Wang et al.r Brain Research 843 (1999) 184–192
receptor binding, both centrally and peripherally w21,23x, is interesting particularly in relation to new evidence for improvement in insulin action and increased fat oxidation following treatment by a selective b3-adrenoceptor agonist in humans w31x. In contrast to the effects of saturated fats, PUFA intake is not, or only very weakly, associated with obesity and the pattern of changes in metabolic rate, etc., is opposite that of saturated lipids w27x. While peripheral mechanisms, such as changes in membrane lipid composition, may account for some of the effects of the dietary fat subtypes, it is equally likely that they are exerting their differential effects on metabolism via direct effects on the brain’s neuronal systems influencing energy balance. In particular, the hypothalamus plays a critical role in the regulation of energy homeostasis including energy intake and expenditure w2,3,11,14,20,26x. The dual centre hypothesis has been historically influential in focusing on the lateral hypothalamus as a feeding centre and the ventromedial area as a satiety centre. Recent, exciting developments have provided extensive new information on neuropeptides subtending this overall conceptual framework w4,9,32x. The aim of this study is to test the hypothesis that modulating the level and fatty acid profile of dietary fat will alter activity in hypothalamic neurons critically involved in the regulation of adiposity. Further, we hypothesise that those changes in neuronal activity will be anatomically discrete, directionally selective, and consistent with the obesogenic effects of saturated, but not polyunsaturated, fats. Neuronal activity was indexed by c-Fos-like immunoreactivity.
185
2. Materials and methods 2.1. Animals Three-week-old C57Blr6J male mice were obtained from Animal Resource Centre ŽPerth, Australia; the mouse genomic background was tested by allozyme electrophoresis and RAPD-PCR analysis.. They were housed individually in a temperature-controlled room Ž208C " 28C. with a 12:12-h light–dark cycle Žlights on at 0700 h.. Mice were given ad libitum access to tap water throughout the study. 2.2. Diet and experimental procedure Mice were fed standard laboratory chow for the first week to allow them to adapt to the new environment. They were subsequently randomly assigned to one of the groups shown in Fig. 1. In all mice fed a high fat diet, fat contributed 58% of calories. In mice fed a low fat diet, fat contributed 10% of calories. In mice fed a high saturated fat diet, edible tallow Ž528C melting point; Unilever, Australia. and safflower oil ŽMeadow Lea Foods, Australia. contributed equally to fat calories. In mice fed a high n y 3 PUFA diet, fish oil ŽEPA-28; from Yamanouchi Pharmaceutical, Japan. contributed all 58% of fat calories. In mice fed a high n y 6 PUFA diet, safflower oil contributed all 58% of calories. The saturated fat diet was thus high in saturated fat with a high n y 6rn y 3 PUFA ratio. The n y 3 diet was almost equally high in saturated fat but
Fig. 1. Experiment design and assignment of mice used in this study ŽPUFA: polyunsaturated fat diet..
H. Wang et al.r Brain Research 843 (1999) 184–192
186
Table 1 Composition of high saturated ŽHS., n y 3 polyunsaturated Ž n y 3., n y 6 polyunsaturated Ž n y 6., and low fat ŽLF. diets Ingredients
HS Žgrkg.
n y 3 Žgrkg.
n y 6 Žgrkg.
LF Žgrkg.
Gelatin Methionine Bran Minerals Vitamins Cornstarch Sucrose Oil Beef tallow Safflower oil Fish oil Casein Total energy Žkcalr100 g. Fat Ž%. Protein Ž%. Carbohydrate Ž%. Fatty acid composition Ž%. Ý Saturated Ý Monounsaturated Ý n y 6 polyunsaturated Ý n y 3 polyunsaturated
19 3 51 67 13 169 85
19 3 51 67 13 169 85
19 3 51 67 13 169 85
14 2.3 38 50 9.7 438 219
169.5 169.5 – 254 523 58 21 21
– – 339 254 523 58 21 21
– 339 – 254 523 58 21 21
– 41 – 188 365 10 22 68
31 27 40 0.3
24.5 22.8 11.4 42.0
with a high level of n y 3 PUFAs and very low n y 6rn y 3 ratio. The n y 6 diet was the most unsaturated but with a very high n y 6rn y 3 ratio. Detailed composition of the respective diets can be seen in Table 1. Diets were freshly made every week and stored at 48C. Mice were given food at 1600 h each day. Food consumption was measured daily and body weight was measured weekly. 2.3. Histology All mice were sacrificed by an overdose of sodium pentobarbitone anaesthesia Ž120 mgrkg i.p.. at a fixed time between 0700 and 1000 h to minimize variation due to circadian rhythm fluctuation in c-Fos protein. Care was taken not to stress the mice before induction of anaesthesia. The total carcass was weighed and epididymal adipose tissue dissected free and also weighed. The brains were fixed by perfusion through the left ventricle of the heart with 50 ml saline, followed by cold 4% paraformaldehyde made in 0.1 M phosphate buffer ŽPB, pH 7.4.. The brains were then removed and stored in 30% sucrose Žin PB. for two nights. The brains were frozen in liquid nitrogen and cryostat sectioned into 25 mm thickness coronal sections at y158C. Sets of consecutive sections were taken. The sections were used for the detection of c-Fos-like immunoreactive ŽFLI. neurons and for Nissl staining for accurate identification of neuroanatomical structures according to the mouse atlas of Franklin and Paxinos w8x. Sections were screened for visualisation of FLI neurons using an Olympus BX50 microscope. In regions where differences in FLI neurons were readily apparent, Fos-positive nuclei were counted following the area was outlined with a drawing tool of Magellan Com-
9.0 12.1 78.1 0.6
9.0 12.1 78.1 0.6
puter Program w13x. Every third section was counted for quantification of FLI neurons. The value represents an average of FLI neurons in a given area of one hemisphere in a single section. FLI neurons were based on relative size and dark stained nuclei. 2.4. Immunohistochemistry The detailed descriptions of technical procedures for immunohistochemistry were described previously w14x. Briefly, the brain sections were firstly washed in 0.8% HCl in 50% alcohol for 30 min for the abolition of endogenous peroxidase activity. The primary antibody, rabbit anti-c-Fos ŽArnel-Lab, US., was diluted at 1:8000 in PBS containing 0.1% Triton X-100 ŽPBS-Triton., and then incubated at room temperature overnight. After three washings in PBSTriton for 10-min each, sections were incubated in the biotinylated secondary antibody ŽBoehringer Mannheim, Germany., diluted at 1:250 in Tris–HCl for 1 h. Sections were then incubated in the avidin–biotin–peroxidase complex ŽSigma, US. diluted at 1:100 in PBS-Triton, for 2 h. Sections were washed in Tris–HCl for 30 min before reacting them with 0.05% diaminobenzidine tetrahydrochloride ŽSigma. in Tris–HCl ŽpH 7.4. and 0.007% hydrogen peroxide in the presence for 0.04% nickel ammonium sulphate. The specificity of the antisera has been tested previously w14x and the immunohistochemical reactions were tested by omission of the primary antisera. No peroxidase reaction occurred in these experiments. 2.5. Data analysis A microscope driven by an IBM compatible computer with the application of the Magellan Computer Program
H. Wang et al.r Brain Research 843 (1999) 184–192
187
Fig. 2. Energy intake Žkcalrday per mouse. in mice fed a high saturated ŽHS., n y 3 polyunsaturated Ž n y 3., n y 6 polyunsaturated Ž n y 6. or low fat ŽLF. diet; or mice fed a high n y 3 polyunsaturated ŽHS-n3. or low fat ŽHS-LF. diet for 4 weeks following 7-week high saturated fat diet.
w13x was used to obtain data concerning counting positively stained FLI neurons. Results are presented as means " S.E.M. The data were analysed with analysis of variance ŽANOVA.. Means were considered significantly different when P F 0.05.
3. Results 3.1. Energy intake, body weight gain, and epididymal fat storage The energy intake curve is shown in Fig. 2. Cumulative food intake over the 7 weeks of feeding showed the low-fat group to have a significantly lower Ž P F 0.05. intake than the three high-fat groups which did not differ. Body weight gain was highest in the high saturated fat diet group over the 7 weeks Ž11.7 " 0.6 g. compared to 11.1 " 1.0 g in the low-fat group, 10.0 " 0.6 g in the n y 6 PUFA group and 7.8 " 0.6 g in the n y 3 PUFA group Ž n y 3 group differs from all other groups, P F 0.01.. From weeks 7 to 11 the group which stayed on a high saturated fat diet gained a further 2.2 " 0.3 g Žvs. 1.3 " 0.0 g for the continuing low fat group. while the Table 2 Distribution of c-Fos-like immunoreactive neurons in the hypothalamus of mice fed a high saturated ŽHS., ny3 polyunsaturated Ž ny3., ny6 polyunsaturated Ž ny6. or low fat ŽLF. diet for 1 week HS Ž ns6. Arc DMH VMH dLH PVN SCh MPO U
2"1 6"1 1"0U 22"2U 5"1 5"1U 7"5
ny3 Ž ns6. U
0"0 10"3 5"1a 7"2 a 5"1 2"1a 5"1
ny6 Ž ns6.
LF Ž ns6.
1"0 5"1 5"1a 6"2 a 2"0 5"1U 4"1
4"1 8"1 3"1 6"2 3"1 2"1 5"1
P F 0.05 compared with low fat diet. P F 0.05 compared with high saturated fat diet.
a
group switched from high saturated fat to n y 3 PUFA lost 0.9 " 0.3 g. Interestingly, the group switch from high saturated fat to low fat continued to gain weight Ž1.7 " 0.4 g; weight change different from the saturated fat to n y 3 PUFA group, P F 0.05.. After 7 and 11 weeks of feeding, epididymal fat stores on the high saturated fat diet Ž622 " 48 mg, 814 " 73 mg. were almost double those on the low fat diet Ž387 " 10 mg, 433 " 43 mg; P F 0.01.. In contrast, the n y 6 PUFA group Ž423 " 69 mg. was equivalent to the low fat group at 7 weeks while the n y 3 PUFA group had accumulated significantly less Ž225 " 26 mg. epididymal fat. Shifting from a high saturated fat diet to low fat from weeks 7 to 11 saw no significant difference in epididymal fat storage. In contrast, a marked 65% decline in epididymal fat was seen after the switch from saturated to n y 3 PUFA diet during weeks 7 to 11. 3.2. Hypothalamic FLI neurons Present study examined FLI neurons in every third section of hypothalamus of mice fed either a saturated fat, n y 3 PUFA, n y 6 PUFA or low fat diet. Only those nuclei showing significant differences between groups are Table 3 Distribution of c-Fos-like immunoreactive neurons in the hypothalamus of mice fed a high saturated ŽHS., ny3 polyunsaturated Ž ny3., ny6 polyunsaturated Ž ny6. or low fat ŽLF. diet for 7 weeks
Arc DMH VMH dLH PVN SCh MPO U
HS Ž ns6.
ny3 Ž ns6.
ny6 Ž ns6.
LF Ž ns6.
1"0 12"1 15"3 27"2U 13"3U 5"1 12"2U
1"1 10"2 13"2 7"1a 3"0 a 1"0 a 7"1a
0"0 9"2 13"3 7"2 a 5"1 5"0 9"1
1"0 15"2 12"2 8"4 2"0 2"1 6"1
P F 0.05 compared with low fat diet. P F 0.05 compared with high saturated fat diet.
a
H. Wang et al.r Brain Research 843 (1999) 184–192
188
Table 4 Distribution of c-Fos-like immunoreactive neurons in the hypothalamus of mice fed a high saturated ŽHS. or low fat ŽLF. diet for 11 weeks, or mice fed a high ny3 polyunsaturated ŽHS-n3. or low fat ŽHS-LF. diet for 4 weeks following 7-week high saturated fat diet HS Ž ns6. Arc DMH VMH dLH PVN SCh MPO U
2"1 20"2 25"4 24"4U 21"0U 4"1 13"2U
LF Ž ns6. 0"0 14"1 13"4 9"2 3"1 4"1 5"1
HS-n3 Ž ns6. Ua
8"1 25"4 18"4 21"3U 11"3a 4"1 10"1
HS-LF Ž ns6. 6"2U a 22"3 16"6 34"2U 20"5U 4"0 12"2U
P F 0.05 compared with low fat diet. P F 0.05 compared with high saturated fat diet.
a
presented. Those areas, showing FLI neurons but not showing diet-induced differences, were medial tuberal nucleus, hypothalamic intermediate zone, supramammillary nuclei and medial preoptic area. 3.2.1. Lateral hypothalamus (LH) Mice fed a high saturated fat diet showed a dramatically increased FLI neurons in the dorsal Žabove a line through the fornix and perpendicular to the midline of the section.
part of the LH ŽdLH; Tables 2–4; Fig. 3.. Compared with a low fat diet, a high saturated fat diet increased FLI neurons in dLH by 367% in 1 week, 338% in 7 weeks and 267% in 11 weeks. In contrast, neither the n y 3 nor n y 6 PUFA diets altered dLH FLI neurons compared to the low fat group at any time point. 3.2.2. ParaÕentricular hypothalamic nucleus (PVN) The most distinctive feature of the PVN response to a high saturated fat diet was a delayed activation of FLI neurons ŽFigs. 4 and 5.. As it is shown in Fig. 5, there were no significant differences in the number of PVN FLI neurons between groups at 1 week ŽTable 2.. However, compared to the low fat group, a significant, 6.5 to 7-fold, increase in FLI neurons was observed after 7 and 11 weeks of a high saturated fat diet ŽTables 3 and 4.. In contrast, PVN FLI neurons Ž7 weeks. were only marginally increased in the n y 3 PUFA group while there was a 2.5-fold increase in the n y 6 PUFA group, the increase was still much less than that in the saturated fat group. Replacing a high saturated fat diet with a high n y 3 PUFA diet dramatically reduced Ž48%. FLI neurons. However, replacing a high saturated fat diet with a low fat diet had no significant effect.
Fig. 3. Photomicrographs showing c-Fos-like immunoreactive neurons in the dorsal part of lateral hypothalamic area ŽdLH. of the mice fed: ŽA. a high saturated ŽHS., ŽB. low fat ŽLF., ŽC. n y 3 polyunsaturated Ž n y 3., or ŽD. n y 6 polyunsaturated Ž n y 6. diet. DMH s dorsomedial hypothalamic nucleus; 3V s third ventricle; f s fornix.
H. Wang et al.r Brain Research 843 (1999) 184–192
189
Fig. 4. Photomicrographs showing c-Fos-like immunoreactive neurons in the paraventricular hypothalamic nucleus ŽPVN. of the mice fed a high saturated fat diet for 1 week ŽA., 7 weeks ŽB., 11 weeks ŽC., or low fat diet for 7 weeks ŽD.. Changing the diet from a high saturated fat to n y 3 polyunsaturated fat ŽF. reduces the number of c-Fos-like immunorective neurons in PVN, but not in a low fat ŽE. replacement.
3.2.3. Ventromedial hypothalamic nucleus (VMH) The VMH showed significant differences in FLI neurons only after 1 week of various diets ŽTable 2.. VMH FLI neurons were inhibited after 1 week of a saturated fat diet but activated equivalently by both PUFA diets. Changing diets from a high saturated fat diet into either low fat or high n y 3 PUFA had no significant influence on FLI neurons in the VMH.
Compared with a low fat diet, a high n y 6 PUFA diet significantly increased 2.5-fold FLI neurons in the SCh at both 1 and 7 weeks, similar to the mice fed a high saturated fat diet ŽTables 2 and 3.. Conversely, compared with a saturated fat diet, n y 3 PUFA diet had significant less FLI neurons in SCh in both 1 and 7 weeks, similar to the mice fed low fat diet. Changing the diet from a high
3.2.4. Medial preoptic nucleus (MPO) The MPO showed no difference in FLI neurons after 1 week of various diets. However, MPO showed a significant 2-fold increase in FLI neurons after 7 weeks and 2.6-fold after 11 weeks in the mice fed a saturated fat diet compared with low fat diet ŽTables 3 and 4.. Again, there were only marginal increases in MPO FLI neurons in the PUFA groups compared to the low fat-fed mice. Changing diets from a high saturated fat diet into either low fat or high n y 3 PUFA had no significant influence in FLI neurons in MPO. 3.2.5. Suprachiasmatic nucleus of hypothalamus (SCh) A distinctive feature of SCh was that n y 3 PUFA and n y 6 PUFA diets had opposing effects on this nucleus.
Fig. 5. The number of c-Fos-like immunoreactive neurons in the paraventricular nucleus of hypothalamus of the mice fed a high saturated ŽHS., ny3 polyunsaturated Ž ny3., ny6 polyunsaturated Ž ny6., or low fat ŽLF. diet; or mice fed a high ny3 polyunsaturated ŽHS-n3. or low fat ŽHS-LF. diet for 4 weeks following 7-week high saturated fat diet.
190
H. Wang et al.r Brain Research 843 (1999) 184–192
saturated fat to either high n y 3 PUFA or low fat diet had no significant affects on FLI neurons in SCh. 3.2.6. Arcuate hypothalamic nucleus (Arc) There were no differences in amount of FLI neurons in Arc between a saturated fat, n y 3 PUFA n y 6 PUFA, and low fat diet groups either at 1 or 7 weeks. However, changing of the diet from a high saturated fat to high n y 3 PUFA or low fat diet resulted in a large increase in FLI neurons in Arc Ž4- and 3-fold increases, respectively; Table 4.. 4. Discussion The present study was designed to examine hypothalamic neuronal response to high fat diets emphasising either saturated fat with a low n y 6rn y 3 PUFA ratio, high n y 3 PUFA or high n y 6 PUFA. Using c-Fos-like immunoreactivity as a marker, the most salient findings of this study were that: Ž1. compared to a low fat diet, a saturated fat diet increases FLI neurons primarily in the dLH and it does so in a time frame that precedes detectable increases in adiposity, Ž2. PUFA diets do not increase FLI neurons in the dLH, Ž3. a saturated fat diet decreases, while PUFA diet increases, FLI neurons in the VMH within 1 week of feeding, and Ž4. a saturated fat diet increased FLI neurons in the PVN but only after an increase in adiposity is detected. Overall, these results are consistent with direct effects of saturated fats to increase lateral, while decreasing ventromedial, hypothalamic neuronal activity. These effects would tune metabolism towards hyperphagia, decreased metabolic rate and weight gain. In contrast, equally high fat diets emphasising PUFAs neither increased lateral, nor decreased ventromedial, hypothalamic neuronal activity. The regulation of energy balance is complex, involving an integration of the internal milieu with environmental forces. The brain, and in particular the hypothalamus, acts as a final integrative centre and modulates energy homeostasis via effects on both intake and expenditure. Impressive strides have been made in the last few years in our understanding of both the neuroanatomy and key peptide regulators of energy balance. However, it is still true that while there is strong evidence of vigorous metabolic defence against energy deficit, the converse is not true. Diets high in saturated fats are particularly obesogenic in many species and strains. The current study offers new insights into mechanisms underlying this observation. The observation of dLH activation by a high saturated fat diet is critical. It is well documented that stimulation of the LH increases food intake and parasympathetic activity w12,18x. Neurons in the dLH send large numbers of descending projections to the dorsal motor nucleus of vagus, solitary nucleus, lateral parabrachial nucleus and lateroventral periaquiduct gray w33x. All these areas are involved in regulation of autonomic function. In addition, the neurons
in the dLH project to the central nucleus of amygdala, which is a key relay center regulating autonomic function. Lesions to the lateral hypothalamus result in a decrease in parasympathetic activity and, conversely, an increase in sympathetic activity, accompanied by decreased food intake and lower adiposity and increased oxygen consumption and core temperature w1,3,16x. The present study also showed that mice fed diets high n y 3 PUFA or n y 6 PUFA, like low fat diets, did not increase FLI neurons in the dLH. This suggests that saturated fat has a selective effect on the dLH, which may explain its obesogenic effects. It is interesting that the dLH showed this increase in FLI neurons within only 1 week on the saturated fat diet. It has been reported that C57 mice show no difference in the levels of blood glucose, insulin and leptin after 1 week on a high fat diet w30x. Taken together, these data are suggestive of a direct influence of saturated fat on dLH neuronal activity. It is also noted that mice fed the saturated fat diet for 1 week had the lowest number of FLI neurons in the VMH compared with all other groups. The VMH lesion-induced obesity syndrome has been well characterised w2x. Therefore, a less active VMH in the saturated fat diet mice would also contribute to development of obesity. Historically, the LH and VMH have been seen as key hypothalamic areas reciprocally controlling energy balance. It is important in the current results that saturated fat diet decreased FLI neurons in the VMH at an early stage. It is interesting to note the change in FLI neurons in various areas over time independent of diet. For example, FLI neurons in the VMH increased dramatically from week 1 to week 7 in all groups. The dLH increase was much smaller but again consistent over groups. However, in the PVN, the increase was restricted to the high saturated fat group. Historically, the lateral hypothalamus has been linked to the parasympathetic, and the ventromedial hypothalamus to the sympathetic nervous system. The balance between these two arms of the autonomic nervous system were then seen to tune metabolism with a high parasympatheticrsympathetic ratio linked to positive energy balance and a low ratio to weight loss. If the ratio of the FLI neurons in the dLH to VMH is then used as an index of parasympatheticrsympathetic balance, it can be seen that it is high at about 2r1 in the low fat group at 1 week of feeding Ži.e., in very young, 5-week-old mice. which makes sense in the rapid growth phase of early life. However, by 7 weeks Ž11-week-old mice., the ratio has dropped to 0.67r1. Interestingly, the ratio is, if anything, lower in the PUFA high-fat groups, but dramatically elevated in the high saturated fat group at 1 week Ž22r1. and still almost 2r1 at 7 weeks of feeding. This is entirely consistent with the very high rate of weight and fat accumulation in the high saturated fat group. It is also consistent with the low metabolic rate and reduced sympathetic nervous system activity previously reported for mice on a high saturated fat diet w21,22x.
H. Wang et al.r Brain Research 843 (1999) 184–192
In this study, we showed no significant change in FLI neurons in the PVN after 1 week on a saturated fat diet. However, there was a dramatic increase in FLI neurons in PVN after 7 and 11 weeks of saturated fat feeding; Ži.e., medium to long term w7 to 11 weeksx, but not short term w1 weekx, high saturated fat diet alters FLI neurons in PVN.. By 7 and 11 weeks, there was an increase in body fat storage and plasma leptin, which were not seen after 1 week of the saturated fat diet. Taken together, this suggests that increased PVN neuronal activation is not due to direct effects of saturated fat but may reflect an attempt by the central energy balance regulatory apparatus to limit fat stores. It is known from previous reports that the PVN– hindbrain pathway can inhibit feeding and reduce high fat diet-induced obesity w17x. This PVN–hindbrain feeding inhibitory pathway descends primarily from the PVN, turns caudally to take a longitudinal path passing through the hypothalamic perifornical area and innervates the dorsal motor nucleus of vagus nerve and solitary nucleus w10x. Therefore, FLI neurons in PVN in the high saturated fat diet-induced obese mice may act to reduce fat storage via negative feedback influenced by amount of fat stored. This is further supported by the demonstration that substituting saturated fat with n y 3 PUFA diet in already obese mice, which we have showed in this study to reduces visceral fat storage as well as FLI neurons in PVN. Therefore, it is suggested that n y 3 PUFA may firstly reduce fat storage followed by withdrawing feedback regulation to the PVN. There are distinctive patterns of changes between dietary groups in other hypothalamic areas that are worthy of comment. The present study showed that the MPO had a significant increase in FLI neurons after high saturated fat diet for 7 and 11 weeks, but not in any other dietary groups. It has previously been reported that the MPO galaninergic neurons respond to high fat diet as showing increased mRNA and protein production w18x. Therefore, it is possible that those activated MPO neurons are galaninergic. Second, changing from high-saturated fat to either a n y 3 or low fat diet induced a dramatic rise in FLI neurons in the arcuate nucleus. This area is selectively lesioned by monosodium glutamate administration w24x with a resultant endocrine syndrome which involves multiple defects including stunting, obesity and a high degree of metabolic efficiency. The arcuate nucleus is known to contain leptin receptors and monosodium glutamate-treated rats do not lose either weight or fat mass when treated with leptin w6x. The role of the arcuate area in the dramatic reversal of the high saturated fat effects by n y 3 Žand to a much lesser extent by a low fat diet. needs exploring in relation to possible direct fatty acid effects on arcuate leptin receptor activation and downstream signalling. Finally, there were differences between groups feed high n y 3 or n y 6 PUFAs with the n y 6 outcomes generally somewhere between the effects of a high saturated fat diet and the n y 3 diet group. The suprachiasmatic area ŽSCh. is one region of clear difference between
191
the two PUFA diet groups. Interestingly, the SCh is an NPY receptor-containing area where it has been shown that, at least under some conditions, NPY-immunoreactivity is inversely proportional to food intake w28x. Future work aimed at investigating the effects of dietary fats on NPY expression in SCh would appear warranted. In summary, this study showed that saturated fat diet can selectively, and may directly, activate LH, and inhibit VMH, neuronal activity. This pattern of hypothalamic activity may play a critical role in the strongly obesogenic properties of dietary saturated fat.
Acknowledgements We would like to thank ArProfessors Peter McLennan and Arthur Jenkins for helpful discussion and suggestions during this study. This study was supported by grants from the NHMRC of Australia.
References w1x L.L. Bernardis, L.L. Bellinger, The lateral hypothalamic area revisited: neuroanatomy, body weight regulation, neuroendocrinology and metabolism, Neurosci. Biobehav. Rev. 17 Ž1993. 141–193. w2x G.A. Bray, The nutrient balance hypothesis: peptides, sympathetic activity, and food intake, Ann. N. Y. Acad. Sci. 676 Ž1993. 223–241. w3x G.A. Bray, S. Inoue, Y. Nishizawa, Hypothalamic obesity. The autonomic hypothesis and the lateral hypothalamus, Diabetologia 20 Ž1981. 366–377. w4x L.A. Campfield, F.J. Smith, P. Burn, Strategies and potential molecular targets for obesity treatment, Science 280 Ž1998. 1383–1387. w5x C. Couet, J. Delarue, P. Ritz, J.M. Antione, F. Lamisse, Effect of dietary fish oil on body fat mass and basal fat oxidation in healthy adults, Int. J. Obes. 21 Ž1997. 637–643. w6x R. Dawson, M.A. Pelleymounter, W.J. Millard, S. Liu, B. Eppler, Attenuation of leptin-mediated effects by monosodium glutamate-induced arcuate nucleus damage, Am. J. Physiol. Ž1997. E202–E206. w7x E. Doucet, N. Almeras, M.D. White, J.P. Despres, C. Bouchard, A. Tremblay, Dietary fat composition and human adiposity, Eur. J. Clin. Nutr. 52 Ž1998. 2–6. w8x K.B.J. Franklin, G. Paxinos, The Mouse Brain in Stereotaxic Coordinates, 1st edn., Academic Press, San Diego, 1997. w9x J.M. Friedman, The alphabet of weight control, Nature 385 Ž1997. 119–120, news. w10x R.M. Gold, A.P. Jones, P.E. Sawchenko, Paraventricular area: critical focus of a longitudinal neurocircuitry mediating food intake, Physiol. Behav. 18 Ž1977. 1111–1119. w11x J.L. Halaas, K.S. Gajiwala, M. Maffei, S.L. Cohen, B.T. Chait, D. Rabinowitz, R.L. Lallone, S.K. Burley, J.M. Friedman, Weight-reducing effects of the plasma protein encoded by the obese gene, Science 269 Ž1995. 543–546, see comments. w12x R.E. Keesey, M.D. Hirvonen, Body weight set-points: determination and adjustment, J. Nutr. 127 Ž1997. 1875S–1883S. w13x X.F. Huang, I. Tork, G. Paxinos, Dorsal motor nucleus of the vagus nerve: a cyto- and chemoarchitectonic study in the human, J. Comp. Neurol. 330 Ž1993. 158–182. w14x X.F. Huang, H.Q. Wang, Altered c-fos expression in autonomic regulatory centers of genetically obese Žobrob. mouse brain, Brain Res. 799 Ž1998. 307–310.
192
H. Wang et al.r Brain Research 843 (1999) 184–192
w15x P.J. Jones, J.E. Ridgen, P.T. Phang, C.L. Birmingham, Influence of dietary fat polyunsaturated to saturated ratio on energy substrate utilization in obesity, Metabolism 41 Ž1992. 396–401. w16x M.D. Hirvonen, R.E. Keesey, Chronically altered body protein levels following lateral hypothalamic lesions in rats, Am. J. Physiol. 270 Ž1996. R738–R743. w17x A.L. Kirchgessner, A. Sclafani, PVN–hindbrain pathway involved in the hypothalamic hyperphagia-obesity syndrome, Physiol. Behav. 42 Ž1988. 517–528. w18x S.F. Leibowitz, Brain peptides and obesity: pharmacologic treatment, Obes. Res. 3 Ž1995. 573S–589S. w19x J.A. Levine, N.L. Eberhardt, M.D. Jensen, Role of nonexercise activity thermogenesis in resistance to fat gain in humans, Science 283 Ž1999. 212–214, see comments. w20x S. Lin, X.F. Huang, Altered hypothalamic c-Fos expression in diet-induced obese mice, Brain Res. Bull., in press. w21x T. Matsuo, H. Sumida, M. Suzuki, Beef tallow diet decreases beta-adrenergic receptor binding and lipolytic activities in different adipose tissues of rat, Metabolism 44 Ž1995. 1271–1277. w22x T. Matsuo, H. Sumida, M. Suzuki, Beef tallow diet decreases norepinephrine turnover rates in rat hypothalamus and cerebral cortex, Metabolism 44 Ž1995. 1377–1379. w23x T. Matsuo, M. Suzuki, Brain beta-adrenergic receptor binding in rats with obesity induced by a beef tallow diet, Metabolism 46 Ž1997. 18–22. w24x J.W. Olney, Status of monosodium glutamate revisited, Am. J. Clin. Nutr. 26 Ž1973. 683–685.
w25x D.A. Pan, A.J. Hulbert, L.H. Storlien, Dietary fats, membrane phospholipids and obesity, J. Nutr. 124 Ž1994. 1555–1565. w26x F. Rohner-Jeanrenaud, B. Jeanrenaud, Central nervous system and body weight regulation, Ann. Endocrinol. 58 Ž1997. 137–142. w27x L.H. Storlien, A.J. Hulbert, P.L. Else, Polyunsaturated fatty acids, membrane function and metabolic diseases such as diabetes and obesity, Clin. Nutr. Metab. Care 1 Ž1998. 559–563. w28x P.L. Selvais, C. Labuche, X.N. Nguyen, J.M. Ketelslegers, J.F. Denef, D.M. Maiter, Cyclic feeding behaviour and changes in hypothalamic galanin and neuropeptide Y gene expression induced by zinc deficiency in the rat, J. Neuroendocrinol. 9 Ž1997. 55–62. w29x L.H. Storlien, A.D. Kriketos, G.D. Calvert, L.A. Baur, A.B. Jenkins, Fatty acids, triglycerides and syndromes of insulin resistance, Prostaglandins, Leukotrienes Essent. Fatty Acids 57 Ž1997. 379–385. w30x R.S. Surwit, A.E. Petro, P. Parekh, S. Collins, Low plasma leptin in response to dietary fat in diabetes- and obesity-prone mice, Diabetes 46 Ž1997. 1516–1520. w31x C. Weyer, P.A. Tataranni, S. Snitker, E.J. Danforth, E. Ravussin, Increase in insulin action and fat oxidation following treatment with CL316.243, a highly selective beta 3-adrenoceptor agonist in humans, Diabetes 47 Ž1998. 1555–1561. w32x S.C. Woods, R.J. Seeley, D. Porte Jr., M.W. Schwartz, Signals that regulate food intake and energy homeostasis, Science 280 Ž1998. 1378–1383. w33x S.A. Zardetto, M.M. Moga, D.J. Magnuson, T.S. Gray, Lateral hypothalamic dynorphinergic efferents to the amygdala and brainstem in the rat, Peptides 9 Ž1988. 1121–1127.