Wheat aleurone polyphenols increase plasma eicosapentaenoic acid in rats.

METHODS: These studies were designed to assess whether wheat polyphenols (mainly ferulic acid [FA]) increased the very-long-chain omega-3 fatty acids (VLC n-3) [eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA)] in rats. Wheat aleurone (WA) was used as a dietary source of wheat polyphenols. Two experiments were performed; in the first one, the rats were fed WA or control pellets (CP) in presence of linseed oil (LO) to provide alpha-linolenic acid (ALA), the precursor of VLC n-3. In the second one, the rats were fed WA or CP in presence of control oil (CO) without ALA. The concentrations of phenolic acid metabolites in urine were also investigated.
RESULTS: The urinary concentration of conjugated FA increased with WA ingestion (p<0.05). Plasma EPA increased by 25% (p<0.05) with WA in the CO group but not in the LO group. In contrast, there was no effect of WA on plasma DHA and omega-6 fatty acids (n-6). Finally, both n-3 and n-6 in the liver remained unchanged by the WA.
CONCLUSION: These results suggest that WA consumption has a significant effect on EPA in plasma without affecting n-6. Subsequent studies are required to examine whether these effects may explain partly the health benefits associated with whole wheat consumption.

Clinical and epidemiological studies suggest that whole-grain consumption is associated with beneficial health effects (1–4). The mechanisms of these effects are not fully identified. They may be partly attributed to the presence of <span class="Chemical">polyphenols in wheat aleurone (WA), the inner layer of the bran. Beyond their antioxidant properties, certain polyphenols increase the endogenous synthesis of the very-long-chain omega-3 fatty acids (VLC n-3) [EPA (eicosapentaenoic acid) and DHA (docosahexaenoic acid)] with no significant effect on omega-6 fatty acids (n-6) (5–7). VLC n-3 are known to be health protective (8, 9). Our previous studies in rats demonstrated that the anthocyanins present in black corn were responsible for this effect (7) and raised the question of whether other polyphenols, including those present in whole wheat, may have a similar effect. Ferulic acid (FA) is the most abundant polyphenol in whole wheat and represents about 90% of the total polyphenol content of wheat bran (10, 11). The hypothesis tested in this study was that the FA-rich WA increases VLC n-3 status. The concentrations of phenolic acid metabolites in urine were also investigated in order to evaluate the absorption of these bioactive compounds.

Materials and methods

The <span class="Species">rats were cared for in accordance with the European Council Directive 86/609/EEC on the care and use of laboratory animals (OJ L 358). The protocols were performed under license from the French Ministry of Agriculture (license no. A380727) and approved by the local animal ethics committee (license n°113_ LBFA-FO-01).

Animals and experimental design

Forty-eight male <span class="Species">Wistar rats (Charles River Laboratories, baseline body weight 75–100 g) were fed a standard diet (A04, SAFE Diets, France). The animals were acclimated 1 week before being distributed into the four experimental groups (Fig. 1). They were housed in individual cages under conditions of constant temperature, humidity, and a standard light-dark cycle (12 h/12 h). Tap <span class="Chemical">water and standard diet were provided ad libitum.

Fig. 1

Study design.

In experiment 1 (Exp. 1), the <span class="Species">rats (n=12/group) received daily, in addition to the standard diet, either control pellets (CP) with control oil (CO) or WA pellets with CO. The CO was palm oil which is free from alpha-linolenic acid (ALA), the precursor of VLC n-3 fatty acids, and served as an isoenergetic control matrix (Fig. 1).

In experiment 2 (Exp. 2), the <span class="Species">rats (n=12/group) received daily, in addition to the standard diet, either CP with <span class="Chemical">linseed oil (LO) or WA pellets with LO. LO was used to provide ALA (18:3n-3) (Fig. 1).

Both CO and LO were provided by gavage and brought 9 kcal/g. The amount of fat provided by the oils was adjusted to the recommendations of the American Institute of Nutrition Rodent Diets-93 (12). Thus, the quantity of <span class="Chemical">linseed oil used for gavage was 200 µL/day corresponding to 92 mg/day of ALA. The WA pellets were prepared by mixing 2 g of the WA layer flour containing 26.7 mg/g dry weight of total FA with 2.5 mL water and 12.5 mg commercial white sugar. The aleurone fraction was provided by Barilla G. e R. F.lli, Parma (Italy), and obtained as follows: durum wheat (Triticum turgidum L. subsp. Durum) kernels went through three subsequent steps of debranning. The material obtained was micronized and separated by a turbo separator into inner and outer parts of the aleurone layer. The outer part of the aleurone, characterized by small particles size (340–400 nm), was used for this study.

The WA pellets provided, in addition to the standard diet, a total of 53.4 mg of FA per day and per rat. The CP were prepared by mixing 1 g of standard A04 ground in flour with 2.5 mL <span class="Chemical">water and 12.5 mg refined commercial white sugar. Mass density of 1 g of the flour of standard A04 was equivalent to 2 g of the WA layer flour. The CP provided, in addition to the standard diet, a total of 8.4 mg of FA per day and per rat. The WA and CP were prepared daily. Table 1 shows the lipid composition of the WA and A04 used in the studies.

Table 1

Macronutrient and fatty acid composition of A04 diet and wheat aleurone

	A04 diet (SAFE)	Wheat aleurone
	g/100 g
Proteins	16.1	12.6
Total carbohydrates	59.9	61.8
Lipids	3.1	8.7
Fatty acids	mg/100 g lipid
Saturated fatty acids	620	955
C18:1n-9	400	1,211
C18:2n-6	1,310	3,240
C18:3n-3	110	326
C20:4n-6	0.0	0.6
C20:5n-3(eicosapentaenoic acid)	20	12
C22:5n-3	10	11
C22:6n-3(docosahexaenoic acid)	30	1.2

Macronutrient and <span class="Chemical">fatty acid composition of A04 diet and <span class="Species">wheat aleurone

In all the experiments body weights and food consumption were recorded weekly. At the end of Exp. 1 and 2, blood was collected on fasted <span class="Species">rats and plasma was stored at −80°C. Immediately after blood collection, the rats were euthanized and their liver sampled, weighed, frozen in liquid nitrogen, and stored at −80°C until analysis. Finally, the urinary concentration of phenolic acids and their metabolites was estimated in the urine samples collected overnight during the last week of treatment of the animals in Exp. 1.

Fatty acid analyses

Plasma and liver <span class="Chemical">lipids were extracted in hexane/isopropanol as described (5–7). Briefly, methylated fatty acids were extracted with hexane, separated, and quantified by GC using a 6850 Series gas chromatograph system (Agilent Technologies, Palo Alto, CA, USA). Methyl ester peaks were identified by comparing their retention time to those of a standard mixture. Saturated, mono-, and poly-unsaturated fatty acid levels were expressed as a percentage of total fatty acids. Total cholesterol and triglycerides were measured using standard methodology on a Synchron Clinical System LX20 (Beckman Coulter, Brea, CA, USA).

Polyphenols (phenolic acids) and metabolite analyses

<span class="Chemical">Phenolic acids in CP and WA, as well as phenolic metabolites in rat urine, were analyzed using an Accela UHPLC 1250 equipped with a linear ion-trap-mass spectrometer (LTQ XL, Thermo Fisher Scientific Inc., San Jose, CA, USA) fitted with a heated-electrospray ionization probe (H-ESI-II; Thermo Fisher Scientific Inc.). Separations were performed using a BlueOrchid-1.8 C18 column (50×2 mm) (Knauer, Berlin, Germany). Pure helium gas (99.999%) was used for collision-induced dissociation (CID). The HPLC-MSn conditions used for the detection are available in the Supplementary file.

Extraction of polyphenols (phenolic acids) compounds from the pellets (A04 and WA)

For the extraction of the free phenolic fraction, 500 mg of sample were mixed with 6 mL of <span class="Chemical">water. The sample was mixed for 2 min, left under agitation for 30 min at room temperature, and centrifuged at 9,200×g for 10 min. Finally, the supernatant was recovered and stored at −20°C until the UHPLC-MSn analysis. For the extraction of the bound phenolic fraction, the extraction residue was further digested with 1.5 mL of 2 mol/L sodium hydroxide at room temperature for 1 h. After alkaline hydrolysis, the pH of the mixture was adjusted to 3 by adding 1.35 mL of 3 mol/L citric acid. The samples were extracted with 6 mL of ethyl acetate and treated as the water extract. The ethyl acetate extracts were evaporated to dryness and the residue was dissolved in methanol. Bound fraction extracts were kept at −20°C in the dark prior to UHPLC-MSn analysis. The HPLC-MSn conditions used for the detection are available in the Supplementary file.

Urinary phenolic acid metabolites

Urine samples of <span class="Species">rats treated with CP or WA were filtered through a 0.45 µm nylon filter before UHPLC-MSn analysis. The HPLC-MSn conditions used for the detection are available in the Supplementary file.

Statistical analysis

Statistics were realized using the software Minitab version 15 (GrimmerSoft, Paris, France). The difference between each <span class="Chemical">fatty acid was compared with unpaired Student's t-test. The difference between each <span class="Chemical">phenolic acid metabolites concentration excreted in urine after the consumption of the WP or CP were compared by non-parametric Mann–Whitney test.

Results

Body weight and food consumption

In Exp. 1, body weight after 12 weeks (373±26 g vs. 367±13 g) and food intake (23.8±2.6 g/day vs. 22.1±1.4 g/day) were not different in the two groups. In Exp. 2, body weight after 12 weeks (376±23 g vs. 378±21 g) and food intake (24.1±1.0 g/day vs. 23.3±1.3 g/day) were not different in the two groups.

Phenolic acids content in WA and CP pellets

Several <span class="Chemical">phenolic acids have been identified and quantified in the WA and CP pellets. The sum of <span class="Chemical">phenolic acids was 28 mg/g for WA and 10 mg/g for CP, mostly in bound form (Table 2). FA was by far the most abundant compound as monomeric and dimeric forms.

Table 2

Phenolic acid content in WA and CPa

	WA, n=3		CP, n=3
	Free	Bound	Free	Bound
Compounds, mg/g of diet
p-Hydroxybenzoic acid	0.12±0.02	ND	ND	ND
p-Coumaric acid	0.29±0.02	0.70±0.07	0.07±0.01	1.1±0.06
Caffeic acid	ND	0.04±0.00	ND	0.07±0.00
Trans-ferulic acid	0.31±0.06	8.9±0.63	0.06±0.01	3.3±0.34
Ferulic acid isomer	ND	2.8±0.44	ND	1.1±0.25
Sinapic acid	ND	0.17±0.02	ND	0.11±0.03
Sinapic acid isomer	ND	0.04±0.01	ND	0.04±0.00
Dimeric ferulic acids	ND	13.5±2.2	ND	3.6±0.23
Trimeric ferulic acids	ND	1.2±0.00	ND	0.33±0.03

Each value is mean±SD.

WA=wheat aleurone pellets, CP=control pellets, ND=not detected.

<span class="Chemical">Phenolic acid content in WA and CPa

WA=<span class="Species">wheat aleurone pellets, CP=control pellets, ND=not detected.

Urinary concentration of polyphenols (phenolic acids) metabolites

Several <span class="Chemical">phenolic acid metabolites have been identified in the urines of rats receiving either CP or WA (Table 3). Six metabolites of FA were present in conjugated forms (as sulfate and glucuronide derivatives). The rats fed WA excreted more FA-O-sulfate compared to controls (p<0.05) whereas no significant difference was found for the other major metabolites of other phenolic acids, notably hippuric acid and enterolactone.

Table 3

Urinary concentration of main phenolic acid metabolites in rats fed WA and CPa

	WA, n=5	CP, n=5	p
Compounds, µmol/L
Hydroxyphenylpropionic acid-like	36.95±29.2	13.40±2.1	NS
Hippuric acid	2,031±506.9	1,361±258.8	NS
Hydroxyphenylpropionic acid-O-sulfate	52.23±25.1	27.95±3.3	NS
Ferulic acid-O-sulfate	481.94±318.5	69.28±27.2	0.04
Dihydroferulic acid-O-sulfate	16.04±7.8	5.70±2.7	NS
Enterolactone-O-glucuronide	43.83±30.3	13.13±5.1	NS

Values are means±SD, n=5 per group.

NS=non-significant (p>0.05).

WA=wheat aleurone pellets, CP=control pellets.

Hydroxyphenylpropionic acid-like, 3′-Hydroxyphenylpropionic acid, Methoxyhydroxybenzoic acid-O-sulfate and Ferulic acid-O-glucuronide were also analysed but not reported in the table.

Urinary concentration of main <span class="Chemical">phenolic acid metabolites in <span class="Species">rats fed WA and CPa

WA=<span class="Species">wheat aleurone pellets, CP=control pellets.

<span class="Chemical">Hydroxyphenylpropionic acid-like, 3′-Hydroxyphenylpropionic acid, Methoxyhydroxybenzoic acid-O-sulfate and Ferulic acid-O-glucuronide were also analysed but not reported in the table.

Plasma lipid concentrations

In Exp. 1 with CO, <span class="Chemical">cholesterol and triglycerides were not different in the two groups (Table 4). Regarding plasma fatty acids, only EPA was significantly higher among rats receiving WA compared with CP (Table 4). In Exp. 2 with LO; cholesterol, triglycerides, and fatty acid composition were not different in the two groups (Table 4). Liver fatty acids were not different in rats receiving WA or CP, regardless of which oil used for the treatment (data not shown).

Table 4

Effect of WA on plasma lipids and fatty acids in rats receiving the control or the linseed oilsa

	Experiment 1		Experiment 2
	CP, n=12	WA, n=12	CP, n=12	WA, n=12
Lipids, mmol/L
Total cholesterol	0.53±0.07	0.55±0.13	0.57±0.10	0.50±0.06
Triglycerides	0.84±0.30	0.74±0.15	0.90±0.29	0.83±0.25
Fatty acids,% of total fatty acids
C14:0	0.27±0.06	0.27±0.05	0.28±0.08	0.28±0.11
C16:0	18.8±1.5	18.7±1.7	17.7±1.1	17.5±1.5
C18:0	7.4±0.93	6.7±0.80	6.9±0.90	6.6±1.2
C18:1n-9	9.6±2.4	10.0±2.7	9.3±2.45	8.7±1.81
C18:2n-6	23.1±1.4	24.1±1.4	26.0±1.1	26.5±1.1
C18:3n-3	0.75±0.12	0.85±0.22	2.8±0.49	2.9±0.49
C20:4n-6	26.8±5.1	25.6±4.9	21.1±4.4	21.2±4.2
C20:5n-3 (EPA)	1.6±0.28	2.0±0.35*	4.0±0.69	4.1±0.70
C22:4n-6	0.29±0.04	0.25±0.04	0.18±0.03	0.16±0.03
C22:5n-3	0.78±0.14	0.79±0.09	1.3±0.32	1.4±0.33
C22:6n-3 (DHA)	4.8±0.41	4.6±0.42	4.8±0.75	4.8±0.67
Total	100.0	100.0	100.0	100.0

Values are means±SD, n=12 per group.

p<0.05. WA=wheat aleurone pellets, CP=control pellets.

Effect of WA on plasma <span class="Chemical">lipids and fatty acids in rats receiving the control or the linseed oilsa

p<0.05. WA=<span class="Species">wheat aleurone pellets, CP=control pellets.

Discussion

The main aims of this study were to test whether WA <span class="Chemical">polyphenols may influence the metabolism of VLC n-3 as shown with some other polyphenols in animals and humans (5–7). We actually show that WA had significant effects on VLC n-3 metabolism. However, only EPA and not DHA increased whereas n-6 was not modified. Finally, liver fatty acids were not modified by WA; and supplementation with the ALA-rich LO abolished the effect of WA on plasma EPA, as if a high intake of ALA, the precursor of EPA in the omega-3 fatty acids (n-3) pathway, saturates the VLC n-3 synthesis pathway and may hide the effect of WA.

The significant concentration of <span class="Chemical">FA-O-sulfate in the urine of rats fed with WA compared to the control confirmed the significant amount of FA provided by this treatment. However, the individual levels of the FA metabolites in urine presented huge variability between the rats fed with WA. This result can be explained by the low bioavailability of these compounds and also the metabolism occurs in the gastrointestinal tract. Indeed previous studies performed with rats fed with cereal-based meals or wheat bran reported a low bioavailability of FA, ranging from 2 to 5% (13, 14).

In Exp. 1, plasma EPA increased by 25% compared to controls which confirms that some <span class="Chemical">polyphenols can be involved in the metabolism of VLC n-3 in the absence of effect on n-6. Indeed, previous studies in humans suggested that wine polyphenols may increase plasma and tissue levels of EPA and DHA (5, 6). This was confirmed in a study reporting increased concentrations of EPA and DHA in rats after ingestion of anthocyanin-rich black corn (7). The lack of DHA increase in the present study was not totally unexpected. Compared with the synthesis of EPA from ALA, the synthesis of DHA from EPA involves additional steps, carbonyl chain elongation, desaturation, and beta-oxidation. It is theoretically conceivable that WA polyphenols stimulated the synthesis of EPA from ALA but not the synthesis of DHA from EPA as the synthesis of DHA from EPA is partly hormone-dependent (15, 16) contrary to the synthesis of EPA from ALA. Subsequent studies should test whether polyphenols with estrogen-like activity (lignans, for instance) may stimulate the two main steps of the n-3 pathway, EPA from ALA and DHA from EPA.

We found no difference between groups in the liver <span class="Chemical">fatty acids. These results were quite surprising because the liver is thought to play a central role in the synthesis of EPA (and eventually DHA) from ALA. The difference between liver and plasma in the effect of WA polyphenols on VLC n-3 suggests a mechanism other than stimulation of the n-3 pathway mainly in the liver. For instance, the implication of the gut microbiota can be suspected in accordance with a recent study in growing gilts fed flax-seed containing diet (17). The authors showed that the ileal flow of EPA and DHA exceeded their dietary intakes indicating net appearance of these fatty acids in the upper gut, and suggesting that the enteric microbiota could actually elongate and desaturate ALA. Subsequent studies are required to confirm that hypothesis.

In Exp. 2, we tested whether increased ALA (under the form of LO) intake can stimulate the synthesis of EPA compared with a CO not containing any ALA. Contrary to our expectations, WA had no effect on EPA in presence of dietary ALA.

Two explanations could be proposed. If the gut microbiota is really involved in the synthesis of EPA from ALA, it could be speculated that high ALA intake may have inhibited the growth of some specific bacteria in the gut (18) and therefore prevent EPA increase. Another possibility would be that too much substrate, that is, too much ALA in the diet, could have saturated and inhibited the whole n-3 pathway and again prevent EPA increase. Several concordant mechanisms could also be considered in future studies.

Subsequent studies are needed to evaluate the dose-effect, or whether there is a threshold effect, in the relation between WA and the n-3 pathway. Also it is not clear whether longer exposition to WA <span class="Chemical">polyphenols may influence the effects on VLC n-3. The low bioavailability of FA provided by aleurone can be also a limitation to this study. Finally, the biological mechanisms involved in the effect of the <span class="Chemical">polyphenols on VLC n-3 are not elucidated yet. More investigations on the PUFA gene expression pathways and transcription factors are needed, and the role of the gut microbiota is still unclear.

In conclusion, our data show that WA and probably the <span class="Chemical">polyphenols present in WA interferes with the metabolism of VLC n-3 without affecting the metabolism of n-6. Although the exact mechanisms are not clearly identified, our results provide new insights into the health benefits of <span class="Chemical">polyphenol-rich whole grains.