Thyroid hormone metabolism may depend on dietary fat

The effect of dietary fat level and composition on the activities of enzymes involved in thyroid hormone metabolism: thyroid peroxidase (TPO) and hepatic type I deiodinase (DI) were investigated. Male Wistar rats weighing on average 277 g (SEM=4.23 g) received different levels (w/w 5%-LF, 10%-MF, 20%-HF) and types of dietary fat (sunfl ower oil group S, rape seed oil R and palm oil P) over a three weeks. TPO rose with fat intake in group R and declined in groups S and P. Hepatic DI activity was not affected by dietary fat composition, but was infl uenced by fat level, decreasing as fat intake increased. The infl uences of dietary fat level and composition on thyroid physiology are interdependent. TPO and DI activity seem to respond in a differentiated manner to changes in the amount and type of fatty acids consumed.

Thyroid peroxidase is a key enzyme in TH biosynthesis catalysing iodide oxidation, iodination of thyrosine residues of thyroglobulin and coupling of iodothyronines (Dunn and Dunn, 2001).Hepatic DI catalyses both outer and inner ring iodothyronine deiodination yielding metabolically active T 3 and inactive reverse triiodothyronine (rT 3 ) (Bianco et al., 2002).
Knowledge concerning the infl uence of dietary fat on thyroid activity could assist in the development of nutritional recommendations useful in the treatment of not only thyroid disorders but also in the prevention and treatment of atherosclerosis and obesity.The aim of this study was to investigate the impact of dietary fat composition and amount on the activities of key enzymes involved in thyroid hormone metabolism, TPO and DI.

Experimental design
Male Wistar rats (n=54) of average weight 277 g (SEM 4.23 g) were obtained from the animal husbandry facility at the Medical Research Center of the Polish Academy of Sciences (Warsaw, Poland).The animals were divided into groups (6 rats each) fed on diets differing in fat composition (sunfl ower oil -group S, rape seed oil -group R, palm oil -group P) and level (w/w): 5% -low fat diets (LF), 10% -medium fat (MF), 20% -high fat (HF).Diet compositions are given in Table 1 and the fatty acid content of dietary fats is shown in Table 2.The animals were housed individually in standard conditions.Feed and water were given ad libitum.After 3 weeks the rats were sacrifi ced and blood was collected by cardiac puncture.The thyroid and liver were excised, weighed, immediately frozen in liquid nitrogen and held at -80 o C prior to further analysis.All procedures were approved by the Local Animal Care and Use Committee in Warsaw.

Chemical analysis
Thyroid peroxidase activity was determined in a thyroidal microsomal fraction as described previously (Rosołowska-Huszcz et al., 2001).Iodothyronine deiodinase activity was measured in a liver microsomal fraction according to Nauman et al. (1984) with T 4 as a substrate.T 3 -I 125 (spec.activity 1.2 µCi/μg) was obtained from Orion laboratories (Hungary).
Protein content were determined by the method of Lowry et al. (1951) with bovine serum albumin as a standard.

Statistical analysis
The data were analysed using two-way analysis of variance.Signifi cant differences between means at the level P=0.05 were followed by post hoc least signifi cant difference test (LSD).Correlations were estimated by calculation of the Pearson correlation coeffi cient.Statistica 6.0 software was used.

RESULTS
Dietary fat level signifi cantly infl uenced body weight gain only in groups S and R (Table 3).Body weight gain was directly related to the intake of fat (r=0.51,P<0.001), PUFA (r=0.68,P<0.0001) and C18:2 (r=0.68,P<0.0001).Daily feed intake expressed in g per day was infl uenced neither by fat level nor its composition (Table 3).Effi ciency of feed energy (body weight gain divided by energy intake) depended both on the type (P<0.000002) and amount (P<0.04) of dietary fat (Table 3).The profi le of fatty acid intake was highly differentiated between groups (Table 4).
Table 3. Body weight gain (BWG), feed intake (FI) and effi ciency of feed energy (FE) for rats fed diets containing 5 (LF), 10 (MF) or 20% (HF) (w/w) of sunfl ower (S), rape seed (R) or palm (P) oil for three weeks.Values expressed as the mean with standard error for six animals Dietary fat composition signifi cantly affected plasma concentrations of C16:0, C18:0, C18:1 and C18:2.Plasma cholesterol and TAG concentrations were infl uenced neither by fat level nor its composition (Table 5).
The effects of the level and type of fat on TPO activity (Figure 1A) were interrelated (ANOVA, effect of interaction: P<0.0001).In rats fed the LF diet, TPO activity was signifi cantly lower in groups S and R than in group P. In rats   1,2 explanations see Table 3 given the MF diet, TPO activity was lower in group R than in S and P. Contrary to these fi ndings, in rats consuming the HF diet, TPO activity was higher in group R than in S and P. In all dietary groups TPO activity did not differ signifi cantly in rats fed the LF and MF diets.In groups S and P, TPO activity was lower in rats fed the HF diets than in rats given LF and MF diets, whereas in group R TPO activity in animals given the HF diet exceeded that observed in those fed the LF and MF diets.Thyroid peroxidase activity was inversely related to plasma TAG concentrations (r=-0.30,P<0.04).Deiodinase activity (Figure 1B) was not signifi cantly affected by dietary fat composition, but showed an inverse relationship to fat level (P<0.0001).Significantly higher DI activity was observed in groups fed the LF diet than in animals receiving MF and HF diets, however in groups S and R only.Deiodinase activity was directly related to carbohydrate intake (r=0.50,P<0.002) and plasma C22:6 concentration (r=0.35,P<0.04), while it was inversely related to fat (r=-0.56,P<0.0001), C18:1 (r=-0.52,P<0.001) and C18:2 (r=-0.38,P<0.02) intakes.

DISCUSSION
The dietary fats used in this study, although all of plant origin, vary widely with regard to the fatty acids (saturated, monounsaturated, polyunsaturated) they contain and also in the plasma fatty acid profi le they produce when consumed.Overall, body weight gain by rats was dependent on dietary fat level and composition.However, in animals fed palm oil, any differences in body weight gain between those receiving various amounts of fat did not reach statistical signifi cance.The direct relationship between body weight gain and the intake of polyunsaturated fatty acids and linoleic acid, as well as the lower body weight gain observed in animals fed on the diet richest in saturated fatty acids, agree with the results of a similar study (Ellis et al., 2002).However, contrary to the fi ndings of Ellis et al. (2002), and despite using similar levels of dietary fat, we observed the effect of fat composition in rats on both low and high fat diets.
Thyroid peroxidase and hepatic DI activities were found to be modulated by dietary fat.However, the effect of dietary fat on the activity of these two enzymes was shown to be different.While TPO activity was altered both by fat level and composition, DI activity depended only on the amount of fat consumed.
We identifi ed a clear interdependence between fat composition and level in modulating TPO activity.In rats fed the least fat (LF), TPO activity was lower in the groups receiving sunfl ower and rape seed oils than in those given palm oil.With the MF diet, TPO activity was lower in rats given rape seed oil than in those receiving sunfl ower oil and with the HF diet this relationship was reversed, with TPO activity highest in subgroup R.This suggests that FA in rape seed oil promote TPO activity whereas those of palm and sunfl ower oils reduce it.Among the fats used in this study, rape seed oil contains the highest amount of C18:1 and C18:2 acids, sunfl ower oil is richest in C18:2, and palm oil is richest in C16:0 and C18:0.TPO activity might therefore be stimulated by the consumption of polyunsaturated n-3 FA and/or monounsaturated n-9 C18:1, while it is reduced by saturated and polyunsaturated n-6 FA.Our results corroborate data obtained by others indicating stimulation of HPT axis activity by n-3 FA.A raised plasma concentration of TSH was found in rats fed diets enriched in C22:6 compared with animals on a diet supplemented with AA or a standard diet (Clandinin et al., 1998).Stimulating effects of n-3 polyunsaturated FA have also been observed for other elements of HPT axis activity such as transthyretin expression in the brain (Puskas et al., 2003) and thyrocyte proliferation (Makino et al., 2001).However, other results also suggest the involvement of PUFA n-6 in the stimulation of thyroid activity.The endogenous ligand of peroxisome proliferator activated receptor γ (PPARγ) -15-deoxy-Δ 12,14 -PGJ 2 (15d PGJ 2 ) -a product of C20:4 metabolism (Forman et al., 1995) was shown to facilitate the synthesis of thyroglobulin in a functional rat epithelial cell line (Kasai et al., 2000).Taking into consideration the similar mode of regulation of thyroglobulin and TPO gene transcription (Espinoza et al., 2001), 15dPGJ 2 might represent an intermediate in the signalling pathway by which PUFA affect TPO level.
The direct relationship between DI activity and carbohydrate intake observed in our study is in accordance with the stimulation of DI expression and activity by glucose (Bianco et al., 2002).The inverse relationship between DI activity and fat intake agrees with previous results showing a decrease in 5' deiodinating DI activity in rats fed on high fat diets (Kahl et al., 1998) and an increase in plasma rT 3 observed in men consuming high levels of fat (Otten et al., 1980).DI expression in the liver has an obligatory requirement for THR, either TRβ or TRα1 (Amma et al., 2001).Bonilla et al. (2001) found that a diet containing high levels of coconut oil (38% w/w) decreased expression of THR in the liver.Thus, an increase in the amount of fat consumed could affect DI expression by diminishing THR stimulation.PPAR have been also shown to modulate THR activity by competing for retinoic acid RXR receptors that heterodimerize with both PPAR and T 3 receptors (Miyamoto et al., 1997).
These results prompt us to conclude that the amount of fat consumed and its composition infl uence thyroid activity affecting thyroid hormone synthesis, and its deiodination.

Figure 1 .
Figure 1.Thyroid peroxidase (TPO [mU/mgP] -panel A) and iodothyronine deiodinase (DI [pmol T 3 /min/ mgP] -panel B) activities in rats fed diets containing 5% (LF), 10% (MF) or 20% (HF) (w/w) sunfl ower oil (group S), rape seed oil (group R) or palm oil (group P) for three weeks.Values are expressed as the mean with standard error for six animals.Different a, b letters indicate signifi cant differences within groups S, R and P (P<0.05);different A, B letters indicate signifi cant differences between groups S, R and P (P<0.05)

Table 4 .
Intake of FA, mg/day /100 g of fi nal body weight 1,2 explanations as for the Table1

Table 5 .
Plasma FA, TAG and cholesterol concentrations