Bifidobacterium breve with α-linolenic acid alters the composition, distribution and transcription factor activity associated with metabolism and absorption of fat

This study focused on the mechanisms that fatty acid conjugating strains - Bifidobacterium breve NCIMB 702258 and Bifidobacterium breve DPC 6330 - influence lipid metabolism when ingested with α-linolenic acid (ALA) enriched diet. Four groups of BALB/c mice received ALA enriched diet (3% (w/w)) either alone or in combination with B. breve NCIMB 702258 or B. breve DPC 6330 (109 CFU/day) or unsupplemented control diet for six weeks. The overall n-3 PUFA score was increased in all groups receiving the ALA enriched diet. Hepatic peroxisomal beta oxidation increased following supplementation of the ALA enriched diet with B. breve (P < 0.05) and so the ability of the strains to produce c9t11 conjugated linoleic acid (CLA) was identified in adipose tissue. Furthermore, a strain specific effect of B. breve NCIMB 702258 was found on the endocannabinoid system (ECS). Liver triglycerides (TAG) were reduced following ALA supplementation, compared with unsupplemented controls (P < 0.01) while intervention with B. breve further reduced liver TAG (P < 0.01), compared with the ALA enriched control. These data indicate that the interactions of the gut microbiota with fatty acid metabolism directly affect host health by modulating n-3 PUFA score and the ECS.

Several studies support a communication axis between the gut microbiota and adipose tissue, which influence the development of metabolic alterations associated with obesity. Conventionally raised mice have long been known to have > 40% more fat mass than their isocaloric-fed germ-free counterparts 1 and more recently, it was shown that transplantation of the gut microbiota from obese mice to germ-free mice resulted in a greater increase in total body fat mass accumulation compared with those colonised with gut microbiota isolated from lean donors 2 . The resistance of germ-free mice to high fat diet induced body weight gain and fat mass accumulation has also been described, suggesting that the gut microbiota may promote fat storage 3 .
Furthermore, the gut microbiota play a role in the low grade inflammation associated with obesity and related metabolic disorders [4][5][6][7][8] . Gut microbiota-derived lipopolysaccharide (LPS) may represent a cause for the onset and progression of inflammation (metabolic endotoxemia) associated with insulin resistance, type 2 diabetes and high fat diet induced increased fat mass. The endocannabinoid system (ECS) interconnects obesity, inflammation and the gut microbiota in the regulation of energy homeostasis, appetite 9 and gut barrier function 10 via the Dietary supplementation with B. breve NCIMB 702258 and B. breve DPC 6330 reduced liver TAG levels. No differences in food intake, body weight gain, visceral fat mass or serum TAG levels were observed between the groups ( Table 1). Analysis of total liver TAG showed that ALA supplementation, either alone (ALA-CON) or in combination with either B. breve strain (ALA + NCIMB 702258/ALA + DPC 6330) resulted in reduced liver TAG levels, compared with the unsupplemented control (CON) (P < 0.01; Table 1). Furthermore, ALA + NCIMB 702258 and ALA + DPC 6330 exhibited reduced liver TAG, compared with ALA-CON (P < 0.01; Table 1).
Unexpectedly, the n-3 PUFA score was significantly increased in all ALA enriched groups compared with CON, irrespective of microbial supplementation (P < 0.05; Table 2). It was found that hepatic peroxisomal beta oxidation was significantly increased following supplementation of the ALA enriched diet with B. breve (both strains), compared with CON and ALA-CON (P < 0.05; Table 2). This result correlated with significantly reduced hepatic levels of CNLA (P < 0.05) and c9t11CLA (P < 0.05) following supplementation with B. breve compared with CON and ALA-CON (Table 2). Interestingly, CNLA (P < 0.05) and c9t11CLA (P < 0.05) were increased in adipose tissue following supplementation with both B. breve strains compared with CON and ALA-CON (Table 3).
Strain specific effects of B. breve on endocannabinoid and congener levels in liver and adipose tissue. To understand the dietary and microbial impact of ALA and B. breve on the ECS, liver and adipose tissue levels of endocannabinoids (EC) and their congeners were determined. Hepatic levels of AEA, oleoylethanolamide (OEA) and 2-AG were reduced in all ALA enriched dietary groups, compared with CON (P < 0.05; Supplementary Table 1 (displaying significance) and Fig. 1). Furthermore, both B. breve strains reduced AEA and N-palmitoylethanolamide (PEA) levels in liver, compared with ALA-CON (P < 0.05; Supplementary Table 1 and Fig. 1), while B. breve DPC 6330 also reduced hepatic OEA levels, compared with ALA-CON (P < 0.05; Supplementary Table 1 and Fig. 1).
B. breve supplementation had a profound effect on EC and congener levels in adipose tissue. Both strains increased AEA and docosahexaenoylethanolamide (DHEA) levels here compared with CON and ALA-CON (P < 0.05; Supplementary Table 1 and Fig. 1). Interestingly, a greater effect of B. breve NCIMB 702258 on adipose tissue EC levels was identified, as this strain also increased PEA (P < 0.05) and OEA (P < 0.05) levels compared with CON and ALA-CON (Supplementary Table 1 and Fig. 1).
Impact on mRNA gene expression levels of enzymes involved in n-3 PUFA metabolism within liver and uptake of fatty acids to liver following dietary enrichment with ALA either alone or supplemented with B. breve NCIMB 702258 or B. breve DPC 6330. The mRNA expression of enzymes involved in n-3 PUFA metabolism within liver (Fig. 2(A)) and genes involved in the transport and uptake of fatty acids to the liver (Fig. 2(B)) were determined using real-time PCR. An increase in hepatic mRNA expression of Δ -6-desaturase was found in all ALA enriched groups, compared with CON (P < 0.01; Fig. 2(A)). Hepatic mRNA expression of elongation of very long chain fatty acids (ELOVL)-5 was reduced in ALA-CON compared with CON (P < 0.05; Fig. 2(A)). Hepatic mRNA expression of sterol regulatory element binding protein (SREBP)-1c was increased following supplementation with ALA, compared with CON (P < 0.05; Fig. 2(B)). The CON group exhibited increased hepatic mRNA expression of fatty acid synthase (FAS), compared with all ALA enriched groups (P < 0.01; Fig. 2(B)). Hepatic mRNA expression of ELOVL-5 was elevated in ALA + DPC 6330 (P < 0.05; Fig. 2(A)), compared with ALA-CON. ALA + NCIMB 702258, ALA + DPC 6330 and CON demonstrated significantly reduced hepatic mRNA expression of Δ -5-desaturase, compared with ALA-CON (P < 0.01; Fig. 2(A)).

(B)).
Impact on mRNA gene expression levels of enzymes involved in the transport and uptake of fatty acids to the ileum following dietary enrichment with ALA either alone or supplemented with B. breve NCIMB 702258 or B. breve DPC 6330. ALA + NCIMB 702258 and ALA + DPC 6330 decreased ileal mRNA gene expression of FABP-2, compared with ALA-CON (P < 0.05; Fig. 3). Although not significant, the expression of FABP-1 and CD-36 were increased in ALA + DPC 6330 (Fig. 3). The expression of fatty acid transport protein (FATP)-4 and diacylglycerol O-acyltransferase (DGAT)-2 were unchanged between all groups (Fig. 3).

Discussion
This study investigated the impact of dietary ALA enrichment with or without fatty acid conjugating B. breve supplementation on fat composition/distribution and the mechanisms through which commensal gut microbes alter lipid metabolism. The work presented herein tested the hypothesis that the alterations in host tissue fatty acid composition by B. breve could be linked to modulation of fat absorption processes in the small intestine and/ or metabolism of fatty acids to longer-chain unsaturated derivatives in the liver.   Previously, we reported that dietary supplementation with linoleic acid and B. breve NCIMB 702258 significantly increased bioactive c9t11CLA concentrations in the liver of mice 16 , while dietary supplementation with B. breve NCIMB 702258 together with ALA increased tissue n-3 LC-PUFA in mice, compared with those that  did not receive the strain 17 . More recently, we demonstrated an increase in brain DHA associated with B. breve NCIMB 702258 supplementation in mice 18 .

Ileum Gene Symbol Forward Primer Reverse Primer
In general, ALA is a poor precursor for n-3 LC-PUFA, particularly DHA, first due to the efficiency for ALA to undergo beta-oxidation in the mitochondria 40 , rendering this fatty acid less available for desaturation and elongation and second because biosynthesis of DHA requires a crucial step in the peroxisome for partial beta-oxidation 40 . Therefore, any other event that increases desaturation and peroxisomal beta-oxidation would also favour DHA biosynthesis. Our recent data demonstrate that dietary CLA, in a specific range of ratio with ALA, significantly increased DHA biosynthesis (unpublished data), rendering some food, for example CLA enriched cheese, an unexpected source of DHA. Furthermore, associations between a reduced n-3 PUFA score following consumption of a CLA enriched cheese, decreased plasma AEA and an improved blood lipid profile in hypercholesterolaemic patients have been described 44 , thus exposing a link between n-3 PUFA, CLA, the ECS and adiposity. Interestingly, dietary CLA in a model of fatty liver was associated with increased hepatic PEA and OEA 45 . Based on these data, we were prompted to evaluate whether B. breve, possibly by increasing CLA formation, could modify the tissue fatty acid profile (as before), but this time identify whether EC and their congener biosynthesis were also affected. While CLA was increased in adipose tissue following supplementation of the ALA enriched diet with both B. breve strains, no change to the CLA concentration was observed in the liver, most likely due to increased hepatic peroxisomal beta-oxidation to CD16:2. This was demonstrated by the increased ratio of CD16:2: c9t11CLA, however we did not detect a subsequent increase in hepatic DHA concentration. Furthermore, we demonstrated that dietary ALA supplementation increased liver and adipose tissue ALA and EPA with decreased arachidonic acid levels in the adipose tissue. Highly polyunsaturated n-3 fatty acids replace arachidonic acid as an eicosanoid substrate in cell membranes 46 , resulting in decreased production of arachidonic acid-derived proinflammatory eicosanoids such as prostaglandin E 2 and leukotriene B 4 47 thus, increased tissue EPA could be beneficial in a variety of chronic inflammatory settings such as non-alcoholic fatty liver disease (NAFLD) and IBD. The overall hepatic n-3 PUFA score was unsurprisingly increased in all ALA enriched dietary groups, compared with the control. B. breve NCIMB 702258 in combination with the ALA enriched diet increased EPA and DHA in adipose tissue, while B. breve DPC 6330 increased DHA only in the adipose tissue, compared with the ALA enriched control group.
Analysis of the EC revealed a decrease in hepatic AEA, OEA and 2-AG in all ALA enriched groups, compared with the control. Furthermore, both B. breve strains further reduced AEA and PEA in the liver, compared with the ALA enriched control group. In contrast, further supplementation of the ALA enriched diet with B. breve NCIMB 702258 specifically increased AEA, PEA, OEA and DHEA in the adipose tissue. These data suggest that the effect of B. breve NCIMB 702258 on the ECS may be similar to those triggered by inflammation. From our experimental conditions, intake of B. breve NCIMB 702258 triggered a cascade linked to activation of toll-like receptors (TLRs) 48 , but we cannot rule out the possibility that some of the effects might be mediated by CLA, such as the increase in adipose tissue PEA and OEA, as previously demonstrated in the liver of rats fed CLA 45 . Interestingly, feeding of both strains increased CLA in adipose tissue. However, only B. breve NCIMB 702258 enhanced AEA, PEA, OEA and DHEA adipose tissue levels, which suggest that even among strains of the same species, different metabolic responses occur on the ECS. The increased adipose tissue AEA could be related to TLR activation as previously demonstrated in lymphocytes, where fatty acid amide hydrolase activity was downregulated 49 . Therefore, these data suggest that concurrent effects of these nutritional factors may result in changes to the ECS. We have demonstrated an ALA enriched dietary effect on the hepatic mRNA expression of Δ -6-desaturase. This enzyme is involved in the biosynthetic pathway of long-chain, highly unsaturated PUFA 50 and so may partly explain the increased EPA concentrations in the liver and adipose tissue. Beyond docosapentaenoic acid formation, Δ -6-desaturase also metabolises tetracosapentaenoic acid (C24:5n-3), a precursor to DHA, however since Δ -6-desaturase is used by two substrates in the same pathway of n-3 PUFA metabolism, there is intrinsic potential for competitive substrate inhibition to occur 51 . Previous studies revealed that high dietary ALA content up-regulates hepatic Δ -6-desaturase gene expression levels in different models 42,52 . The hepatic mRNA expression of ELOVL-5 was not found to be increased for all ALA enriched dietary groups. One recent study demonstrating the effect of ALA supplementation on Δ -6-desaturase and ELOVL-5 expression in the common carp showed that after 42 days, while Δ -6-desaturase activity decreased, there was an increase in ELOVL-5 expression 42 . This suggests that while the activity of one rate-limiting enzyme involved in the metabolism of n-3 PUFA increases, there may not be a corresponding increase in the activity of the other at the same time, thus the sequence of enzyme activity is time dependent.
Supplementation of B. breve DPC 6330 in an ALA enriched diet had a significant effect on increasing the mRNA expression of genes involved in the transport and uptake of fatty acids to the distal small intestine. The mRNA expression of FABP-1 and CD-36 were increased in the ileum following supplementation of the ALA enriched diet with B. breve DPC 6330. FABP-1 and CD-36 glycoprotein, which play a role in cellular adhesion and in the regulation of fatty acid transport and uptake, have previously been shown to be increased in the intestine following in vivo exposure of human intestinal mucosa to the probiotic species Lactobacillus plantarum WCFSI 53 . Furthermore, the role of the microbiota in stimulating dietary fatty acid absorption in the intestinal epithelium have previously been demonstrated 54 .
Hepatic TAG levels were reduced following 6 weeks of dietary intervention with the ALA enriched diet. It has been reported that EPA may be a promising novel therapy to decrease hepatic TAG 55,56 and may therefore be of importance for the treatment of NAFLD. Coincidently, all ALA dietary groups were associated with increased hepatic EPA. In addition, all ALA enriched dietary groups had reduced hepatic FAS mRNA expression and lower SCD-1 expression whereby the decrease in SREBP-1c regulated enzymes had a significant effect on fatty acid synthesis in the liver 57 . Supplementation of the ALA enriched diet with both B. breve strains decreased hepatic TAG levels compared with the ALA only group. These data demonstrate a potential for administered B. breve to reduce liver TAG levels in the host. Hepatic TAG levels, the mRNA expression of CD-36 and FABP-1 were all decreased following further supplementation of the ALA enriched diet with B. breve, compared to the ALA enriched group, whereby CD-36 and FABP-1 are involved in transport and uptake of fatty acids to the liver 58,59 . This indicates a strain-specific effect of these two B. breve strains toward reducing hepatic TAG levels while also reducing the hepatic mRNA expression of genes relating to reduced fatty acid uptake to the liver. A potential therapeutic role for commensal bacteria in the treatment of NAFLD have been described 60 . For example, it was reported that dietary supplementation with Lactobacillus plantarum MA2 with a cholesterol enriched diet resulted in reduced liver TAG and increased numbers of fecal Lactobacillus and Bifidobacterium in rats 61 . Furthermore, the potential of VSL#3, a multistrain preparation composed of Streptococcus thermophilus and several strains of Lactobacillus and Bifidobacterium for reducing inflammation (reduced TNFα and cyclooxygenase expression, increased PPARα expression) and oxidative damage (inducible nitric oxide synthase) to the liver while ameliorating the liver lipid profiles has been shown 62 .
The data presented in this study confirmed that dietary ALA positively impacted on tissue fatty acid composition, increased hepatic n-3 PUFA metabolism and reduced hepatic TAG levels, hepatic fatty acid synthesis enzymes and hepatic uptake of fatty acids and demonstrated effects on the ECS. Previously, we demonstrated that B. breve NCIMB 702258 in an ALA enriched diet increased hepatic EPA concentrations, compared with ALA alone 17 . We have elaborated on these previous findings by demonstrating an impact not only of B. breve NCIMB 702258 in an ALA enriched diet, but also of B. breve DPC 6330 toward reducing liver triglyceride levels, altering fatty acid and EC profiles and expression of key enzymes involved in the metabolism of fatty acids. Supplementation with either B. breve strain in combination with an ALA enriched diet reduced liver TAG levels which was associated with reduced hepatic CD-36 and FABP-1 mRNA expression, involved in the uptake of fatty acids to the liver. While it is clear from the data that ALA impacted on desaturase activity, there was little effect of B. breve supplementation on n-3 PUFA metabolism gene expression in the liver. Instead, the increased peroxisomal beta-oxidation caused by the strains, could explain why there was no increase in CLA identified in the liver. Adipose tissue CLA was increased, as a result of B. breve supplementation. Future studies will be devoted to evaluating the possible synergistic effects of CLA enriched products with B. breve: (i) in maintaining/re-equilibrating energy and lipid metabolism, by modulating tissue EC and congener profiles, (ii) acting on energy metabolism and EPA and DHA related molecules, and (iii) to assess if there is an optimal level of dietary EPA and DHA to allow such a synergistic activity. This novel study identifies mechanisms through which these lipid-modulating Bifidobacterium strains may be acting in vivo. Considering that the relative proportions of Bifidobacterium in the gut microbiota of C57BL/6 J have previously been reported to compose 2-3% only of the total gut microbial content 18 (mainly Bifidobacterium animalis and Bifidobacterium pseudolongum 63 ), it is indeed intriguing to identify probiotic manipulations of the gut microbiota which can have a positive impact on health. This nutritional approach may prove helpful in the alleviation of metabolic disorders such as those causing NAFLD, where CLA 45 , n-3 PUFA 64,65 and manipulation of the gut microbiota 66 have been shown to decrease adiposity accumulation in the liver. NCIMB 702258 and B.  Detroit, MI, USA), supplemented with 0.05% (w/v) L-cysteine hydrochloride (Sigma-Aldrich, Wicklow, Ireland) (mMRS), containing 500 μ g/mL rifampicin (Sigma-Aldrich). Following anaerobic incubation (anaerobic jars with Anaerocult A gas packs; Merck, Darmstadt, Germany) at 37 °C for 3 days, colonies were stocked in mMRS broth containing 40% (v/v) glycerol and stored at − 80 °C. To confirm that the rifampicin resistant variants were identical to the parent strain, molecular fingerprinting using pulse-field gel electrophoresis was employed.

Preparation and administration of B. breve
Prior to freeze drying, B. breve NCIMB 702258 and B. breve DPC 6330 were grown in mMRS by incubating for 48 hr at 37 °C under anaerobic conditions. The culture was washed twice in phosphate buffered saline (PBS) and resuspended at a concentration of ~2 × 10 10 cells/mL in 15% (w/v) trehalose (Sigma-Aldrich) in dH 2 O. One millilitre aliquots were freeze-dried using a 24 hr programme (freeze temp. − 40 °C, condenser set point − 60, vacuum set point 600 m Torr). Animals were each administered approximately 1 × 10 9 live microorganisms per day, which was achieved by resuspending appropriate quantities of freeze-dried powder in water which was consumed ad libitum. Mice that did not receive the bacterial strains received placebo freeze-dried powder (15% (w/v) trehalose in dH 2 O). Water containing either the bacterial strains or placebo freeze-dried powder was the only water supplied to the mice throughout the trial. Freeze dried powders with the bacterial strains underwent continuous quality control checks of cell counts for the duration of the trial, by plating serial dilutions on mMRS agar supplemented with 100 μ g/mL of mupirocin (Oxoid) and 100 μ g/mL rifampicin (Sigma-Aldrich) and incubating plates anaerobically for 72 hr at 37 °C. Body weight and food intake were assessed weekly. Following 6 weeks on experimental diets, the animals were sacrificed by cervical dislocation. Liver, brain, fat pads (epididymal, perirenal and mesenteric), gastrointestinal tract from stomach to anus, and cecal contents were removed, blotted dry on filter paper, weighed and flash-frozen immediately in liquid nitrogen. All samples were stored at − 80 °C until processed. Blood samples were collected into serum collection tubes (BD Diagnostics, Oxford, UK) from fasted animals, and allowed to clot for at least 30 min at 4 °C before centrifugation for 20 min at 10,000 g to separate the serum.

Animals and
Culture Dependent Microbial Analysis. Fresh fecal samples were taken from BALB/c mice every second week for microbial analysis. Microbial analysis of the fecal samples involved enumeration of the B. breve strains by plating serial dilutions on MRS agar supplemented with 100 μ g/mL of mupirocin (Oxoid), 100 μ g/mL of rifampicin (Sigma-Aldrich) and 50 units/mL of nystatin (Sigma-Aldrich). Agar plates were incubated anaerobically for 72 h at 37 °C.

Lipid Extraction and Fatty Acid
Analysis. An aliquot of the lipid fraction for each sample was mildly saponified using a procedure in order to obtain free fatty acids (FFA) for high performance liquid chromatography (HPLC) analysis. Lipid extracts were dissolved in 5 ml of ethanol, 100 μ l of desferal (25 mg/ml ddH 2 O), 1 ml of a 25% (v/v) solution of ascorbic acid in water, 0.5 ml of 10 N KOH, and left for 14 h in the dark at room temperature. Following this, 10 ml of n-hexane and 7 ml of ddH 2 O were added. The samples were then acidified with 0.35 ml of 37% (v/v) HCI, to pH 3-4. Samples were centrifuged for 1 h at 900 g. The hexane phase containing FFAs was collected, the solvent evaporated, and the residue was dissolved in 0.5 ml of CH 3 CN/0.14% (v/v) CH 3 COOH.
Separation of FAs, including CLA and its metabolites, was performed on an Agilent 1100 HPLC system (Agilent, Palo Alto, Calif., USA) equipped with a diode array detector. A C-18 Inertsil 5 ODS-2 Chrompack column (Chrompack International BV, Middleburg, The Netherlands), 5 μ m particle size, 150 × 4.6 mm, was used with a mobile phase of CH 3 CN/H 2 O/CH 3 COOH (70/30/0.12, v/v/v) at a flow rate of 1.5 ml/min 67 . Conjugated diene unsaturated FAs were detected at 234 nm. Spectra (195-315 nm) of the eluate were obtained every 1.28 s and were electronically stored. Second-derivate UV spectra of the conjugated diene FAs were generated using Phoenix 3D HP Chemstation software (Agilent, Palo Alto, CA). These spectra were acquired to confirm identification of the HPLC peaks 68 .
Since SFAs are transparent to UV, after derivatization, they were measured as fatty acid methyl esters (FAME) using gas chromatography (Agilent, Model 6890, Palo Alto, CA) equipped with split ratio of 20:1 injection port, a flame ionization detector (FID), an autosampler (Agilent, Model 7673), a 100 m HP-88 fused capillary column (Agilent). Finally, data were analysed by the Agilent ChemStation software system. The injector and detector temperatures were set up at 250 °C and 280 °C, respectively. H 2 served as carrier gas (1 ml/min), and the FID gases were H 2 (30 ml/min), N 2 (30 ml/min), and purified air (300 ml/min). The temperature program was as follows: initial temperature was 120 °C, programmed at 10 °C/min to 210 °C and 5 °C/min to 230 °C, then programmed at 25 °C/min to 250 °C and held for 2 min 64 . Measurement of TAG in the Liver and Serum. The lipids from 50 mg of frozen liver were extracted and purified according to the method of Folch 69 . Liver lipids were extracted using chloroform-methanol (2:1, v:v; Thermo Scientific) and an aliquot of the organic phase was collected, dried and resuspended in duplicate in Infinity-TAG lipid-stable reagent (Thermo Scientific). TAG levels were determined according to the manufacturer's instructions and lipids were quantified using a TAG chemistry calibrator (Pointe Scientific Inc., MI, USA). Serum TAG levels were analysed in duplicate using the commercial L-Type TAG M kit (Wako Diagnostics, Neuss, Germany).

Analysis of Endocannabinoids and their Congeners.
Aliquots of the organic phase containing extracted lipids were evaporated to dryness under vacuum, and reconstituted with 0.4 ml of 100% (v/v) methanol. Quantification of AEA, 2-AG, PEA and OEA, was performed by liquid chromatography-atmospheric pressure chemical ionization-mass spectrometry (LC-APCI-MS), and using selected ion monitoring (SIM) at M + 1 values for the four compounds and their deuterated homologues 70 . A C-18 Zorbax Eclipse Plus column (Agilent, Palo Alto, CA) 5 μ m particle size, 50 × 4.6 mm was used with a mobile phase of CH 3 OH/H 2 O/CH 3 COOH (80/20/0.3, v/v/v) at a flow rate of 0.5 ml/min.

RNA Extractions and Complementary DNA Synthesis.
Total RNA was isolated from liver and ileum using the commercial RNeasy Mini-Kit (Qiagen, West Sussex, UK), according to the manufacturer's instructions. Total isolated RNA was quantified using the Nanodrop (Thermo Scientific). Single stranded complementary DNA (cDNA) was synthesised from 1 μ g of total RNA using 2.5 ng/μ L random primers (Promega, WI, USA), 10 mM PCR nucleotide mix (Promega), 40 units/μ L RNasin Plus RNase inhibitor (Promega) and the Im-Prom II reverse transcriptase (Promega).

Real-Time PCR Analysis.
Amplification of generated cDNA was performed in the Lightcycler 480 system (Roche Diagnostics Ltd., West Sussex, UK) using 0.25 μ M primers (MWG Eurofins, Ebersberg, Germany), 1 μ L cDNA and the Lightcycler 480 SYBR Green I Master kit (Roche Diagnostics Ltd). Real-time PCR conditions were set at: 95 °C for 10 min followed by 50 cycles at 95 °C for 10 sec, 60 °C for 5 sec and 72 °C for 15 sec. Specific forward and reverse primers used to amplify cDNA were newly designed and are listed in Table 4. All samples were analysed in duplicate and normalised to β -actin, as a constitutively expressed control gene. Melting curve analysis allowed the validation of the authenticity of the real-time PCR products. Basic relative quantification of expression was determined using the comparative 2 −ΔΔCt method. Statistical Analysis. All results are presented as means ± SD (per group). To assess whether differences between the treatment groups were significant, data were analysed using one-way ANOVA followed by post hoc Tukey's multiple comparison test using GraphPad Prism version 4.0 for Windows (GraphPad Software). Statistical significance was accepted at P < 0.05. Grubbs method was used to test for outliers 71 .