Modulation of inflammatory gene transcription after long-term coffee consumption

Modulation of inflammatory gene transcription after long-term coffee consumption

    Modulation of inflammatory gene transcription after long-term coffee consumption Swantje Winkler, Natalie Dieminger, Volker Blust, An...

955KB Sizes 0 Downloads 21 Views

    Modulation of inflammatory gene transcription after long-term coffee consumption Swantje Winkler, Natalie Dieminger, Volker Blust, Annett Riedel, Tamara Bakuradze, Gina Montoya, Ute Hassmann, Roman Lang, Thomas Hofmann, Veronika Somoza, Elke Richling, Gerhard Bytof, Herbert Stiebitz, Ingo Lantz, Dorothea Schipp, Jochen Raedle, Doris Marko PII: DOI: Reference:

S0963-9969(14)00388-3 doi: 10.1016/j.foodres.2014.05.073 FRIN 5315

To appear in:

Food Research International

Received date: Revised date: Accepted date:

7 January 2014 22 April 2014 26 May 2014

Please cite this article as: Winkler, S., Dieminger, N., Blust, V., Riedel, A., Bakuradze, T., Montoya, G., Hassmann, U., Lang, R., Hofmann, T., Somoza, V., Richling, E., Bytof, G., Stiebitz, H., Lantz, I., Schipp, D., Raedle, J. & Marko, D., Modulation of inflammatory gene transcription after long-term coffee consumption, Food Research International (2014), doi: 10.1016/j.foodres.2014.05.073

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Modulation of inflammatory gene transcription after long-term coffee consumption Swantje Winkler1, Natalie Dieminger2, Volker Blust1, Annett Riedel3, Tamara Bakuradze4,

T

Gina Montoya4, Ute Hassmann1, Roman Lang2, Thomas Hofmann2, Veronika Somoza3, Elke

RI P

Richling4, Gerhard Bytof5, Herbert Stiebitz5, Ingo Lantz5, Dorothea Schipp6, Jochen Raedle7

1

MA NU

SC

and Doris Marko1*

University of Vienna, Department of Food Chemistry and Toxicology, Währingerstr.38, A-

1090 Vienna, Austria 2

Chair of Food Chemistry and Molecular Sensory Science, Technical University Munich, Lise-

University of Vienna, Department of Nutritional and Physiological Chemistry, Althanstrasse

PT

3

ED

Meitnerstrasse 34, 85354 Freising

Department of Chemistry, Division of Food Chemistry and Toxicology, University of

AC

4

CE

14 (UZA II), A-1090 Vienna, Austria

Kaiserslautern, Kaiserslautern, Germany 5

Tchibo GmbH, Überseering 18, D-22297 Hamburg, Germany

6

www.ds-statistik.de, Rosenthal-Bielatal, Germany

7

Innere Medizin III, Westpfalzklinikum, Kaiserslautern, Germany

Corresponding author: *Doris Marko, University of Vienna, Department of Food Chemistry and Toxicology, Währingerstr. 38, 1090 Vienna, Austria, email: [email protected], phone: +43 (0)1-4277-70800

ACCEPTED MANUSCRIPT Abbreviations: ARE, antioxidative response element; BC, blood collection; BMI, body mass index; CGA, chlorogenic acids; CRP, c-reactive protein; DHFAglcd, dihydroferulic acid

T

glucuronide; FA, ferulic acid; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; HO1,

RI P

heme oxygenase; IL6, Interleukin 6; LPS, lipopolysaccharide; Nf-κB, nuclear factor 'kappalight-chain-enhancer' of activated B-cells; NMP, N-methylpyridinium; Nrf2, nuclear factor

SC

(erythroid-derived 2)-like 2; PPAR, peroxisome proliferator-activated receptor; RPL13,

MA NU

ribosomale protein L13a; SD, standard deviation; TNFα, tumor necrosis factor α

AC

CE

PT

ED

Keywords: Nrf2, inflammation, obesity, coffee, chlorogenic acid, polymorphism

ACCEPTED MANUSCRIPT Abstract Scope: Obesity has been found to be associated with low grade inflammation accompanied

T

by chronic oxidative stress. The transcription factor Nrf2 is likely involved in lipid metabolism

RI P

and inflammation processes, possibly mediated by an antioxidant response element (ARE)similar region located in the promoter of lipogenic genes like peroxisome-proliferator

SC

activated receptor γ (PPARγ) and pro-inflammatory interleukin 6 (IL6). The present study

MA NU

investigates the influence of coffee consumption on the transcription of obesity-associated genes in human peripheral blood lymphocytes (PBL). Two different coffee blends with comparable caffeine concentrations were provided, either rich in chlorogenic acids and trigonelline (market blend, MB) or in N-methylpyridinium (NMP, study blend, SB).

ED

Methods and results: In a cross-over randomized double blind intervention study 84

PT

volunteers (male and female, 25.6 ± 5.8 years, BMI 22.9 ± 1.9 kg/m2, healthy, nonsmokers, regular coffee drinker) daily consumed 750 mL of the respective coffee over a period of 4

CE

weeks, respectively. Transcription of IL6 in PBL was found to be positively associated with

AC

body fat. After consumption of MB transcription of Nrf2, PPARγ and IL6 decreased significantly while concomitantly an enhanced level of PPARα mRNA was found. In contrast, effects of SB were not quite pronounced. The changes in gene transcription appear to correlate with the level of different CGA metabolites in the plasma of the volunteers. Initial results further indicate a potential contribution of genetic polymorphisms in the nrf2 promoter and the pparγ-gene to the influence of coffee consumption on PPARγ transcription. Conclusion: Regular coffee consumption affects the transcription of genes associated with obesity and/or inflammation. Metabolites of chlorogenic acids as well as genetic polymorphisms may be relevant influencing factors.

ACCEPTED MANUSCRIPT 1 Introduction In recent years the prevalence of adults being overweight has alarmingly increased to 1.4

T

billion people worldwide. Overweight is ranked as the fifth leading risk of global deaths

RI P

through sequels such as cardiovascular diseases, diabetes mellitus, musculoskeletal

SC

disorders and some types of cancer (Fujioka, 2002; J. G. Kang & Park, 2012; Visscher & Seidell, 2001).

MA NU

Hence, it is aimed to identify substances of our daily nutrition that bear weight losing properties. After water coffee is the most consumed beverage worldwide with USA, Brazil and Germany as leading countries (ICO, 2011). Obesity is accompanied by low-grade

ED

inflammation increasing oxidative stress, since adipocytes attract macrophages that infiltrate into adipose tissue (Medzhitov, 2008). The key regulator tumor necrosis factor α (TNFα)

PT

subsequently enhances nuclear factor 'kappa-light-chain-enhancer' of activated B-cells (Nf-

CE

κB) action that in turn initiates an immune reaction by binding to the DNA and starting transcription of various cytokines. Thus, during weight loss decreased levels of the

AC

proinflammatory factors like TNFα, IL6 and hepatic C-reactive protein (CRP) can be observed (Higdon & Frei, 2003; Stienstra, Duval, Muller, & Kersten, 2007). Furthermore, oxidative stress activates the transcription factor Nrf2 by detaching it from its cytosolic anchor Keap1. After release from Keap1, Nrf2 translocates into the nucleus where it binds jointly with co-factors to the cis-acting antioxidant response elements (ARE). Consequently, ARE-regulated transcription of phase II biotransformation enzymes such as glutathione S-transferases (GST), UDP-glucuronyltransferases (UGT) or heme oxygenase 1 (HO1) is activated (Ute Boettler et al., 2011; Cho et al., 2010). Besides, Nrf2 is likely involved in lipid metabolism and inflammation processes, since adipose tissue mass is decreased in

ACCEPTED MANUSCRIPT Nrf2-knockout models and these animals are protected against weight gain and obesity from a high fat diet (Hou et al., 2012; Huang, Tabbi-Anneni, Gunda, & Wang, 2010; Pi et al., 2010).

T

In Nrf2-deficient 3T3-L1 cells differentiation is repressed (Hou et al., 2012; Pi et al., 2010). In

RI P

contrast, Keap1-knockdown and –deficiency animal models lead to accelerated differentiation of preadipocytes, higher body mass, fat mass and hepatic triacylglycerides (Pi

SC

et al., 2010; Xu, Kulkarni, Donepudi, More, & Slitt, 2012). This process is possibly mediated

MA NU

by an ARE-similar region located in the promoter of lipogenic genes like peroxisomeproliferator activated receptor γ (PPARγ) and C/EBPα that are therefore up regulated after Nrf2-activation (Pi et al., 2010). PPARs are ligand-activated transcription factors belonging to a superfamily of nuclear receptors controlling mainly lipid metabolism via regulation of

ED

related enzymes (Huang et al., 2010; Pi et al., 2010).

PT

Nrf2 may also be responsible for down regulation of PPARα, a lipolytic gene that is involved in the oxidation of fatty acids due to enhancing the transcription of enzymes (Huang et al.,

CE

2010) like carnitine palmitoyltransferase-I and medium-chain acyl-CoA dehydrogenase

AC

(Sugden, Caton, & Holness, 2010). Besides enhancing enzymes required for fatty acid oxidation this latter receptor moreover reduces inflammation possibly via accelerated expression of IκB that controls Nf-κB and therefore down regulates the expression of proinflammatory cytokines (Stienstra et al., 2007). Furthermore Nrf2 is directly able to influence inflammatory processes by binding to ARE in the promoter of IL6 thereby up regulating its expression (Wruck et al., 2011). Coffee consumption is prevalently associated with lower risk of type II diabetes (T2D) in men and women (Fujioka, 2002; Greenberg, Boozer, & Geliebter, 2006; Higdon & Frei, 2003; Esther Lopez-Garcia et al., 2006; Olthof, van Dijk, Deacon, Heine, & van Dam, 2011; van Dam & Feskens, 2002). This may be due to decreased obesity and inflammation markers (Kempf &

ACCEPTED MANUSCRIPT Martin, 2010; Lopez-Garcia, van Dam, Qi, & Hu, 2006) for which caffeine is mostly taken into account (Fujioka, 2002; Esther Lopez-Garcia et al., 2006). Caffeine is known to stimulate

T

thermogenesis by up regulating the expression of uncoupling proteins (UCP) in mitochondria

RI P

and to enhance lipolysis, fat oxidation and insulin secretion while decreasing blood glucose levels. Therefore, it may promote weight loss and improve diabetes (Butt & Sultan, 2011;

SC

Rudelle et al., 2007; Tunnicliffe & Shearer, 2008). However, decaffeinated coffee has been

MA NU

reported as well to have anti-diabetic or anti-obesity properties (Greenberg et al., 2006; Olthof et al., 2011), suggesting that other constituents than caffeine, e.g. chlorogenic acids (CGA) (Cho et al., 2010; Johnston, Clifford, & Morgan, 2003; Li, Chang, Ma, & Yu, 2009; Olthof et al., 2011), trigonelline (Greenberg et al., 2006; Johnston et al., 2003; Olthof et al.,

ED

2011) or other xanthines may contribute to the reported effects on lipid and glucose

PT

metabolism. CGA and their metabolites are already hypothesized to contribute to weight loss after consumption due to a decrease of body fat (Jin Son, W. Rico, Hyun Nam, & Young

CE

Kang, 2010; Tunnicliffe & Shearer, 2008) and to improve glucose and insulin metabolism

AC

(Butt & Sultan, 2011) mediated by reduced uptake of glucose from the gut (Greenberg et al., 2006) and extenuated activity of glucose-6-phosphatase in the liver (Thom, 2007). In a recent human intervention study we observed a significant weight loss in lean subjects (BMI < 25 kg/m2) after a 4 week consumption of a dark roasted coffee brew that contained high concentrations of NMP and CGA (Bakuradze et al., 2011). In the present study we addressed the question whether long-term consumption of coffee affects the transcription of genes involved in lipogenesis and inflammation, known to be associated with weight regulation and whether differences between two different coffee blends, reflected by the levels of CGA, trigonelline and NMP, result in differences in the transcriptional response pattern in PBL of the probands.

ACCEPTED MANUSCRIPT

2 Materials and Methods Coffee brews

T

2.1

RI P

Two different coffee blends were used in this study. The CGA- and trigonelline-rich market

SC

blend (MB) was manufactured from five commercially available regular ground coffee. Quantified compounds were: 0.39 (± 0.01) mg/g NMP, 6.27 (± 0.12) mg/g trigonelline, 12.4

MA NU

(± 0.13) mg/g caffeine and 19.3 (± 0.28) mg/g CGA. The NMP-rich study blend (SB) was obtained as a mixture from a dark roast coffee and a light roast coffee and contained 1.20 (± 0.02) mg/g NMP, 3.42 (± 0.20) mg/g trigonelline, 12.8 (± 0.24) mg/g caffeine and 10.0 (±

2.2

ED

0.28) mg/g CGA. Study design

PT

84 healthy, male and female (25.6 ± 5.8 years, BMI 22.9 ± 1.9 kg/m2), nonsmoking, regularly

CE

coffee consuming volunteers participated in this study and were randomly divided into two groups (group A, n= 43, 24 male + 19 female; group B, n=41, 22 male + 19 female) Overall

AC

the study took 20 weeks and was subdivided into 4 week-periods of alternating wash-outs and interventions Error! Reference source not found.as follows: wash-out 1, intervention 1, wash-out 2, intervention 2, wash-out 3. Prior to and after every period blood of each subject was collected (BC) after an overnight fasting and analysis of body composition was performed. During the first intervention period group A daily consumed 750 mL of market blend, whereas group B obtained the same amount of study blend. This was reversed during the second intervention period. Compliance to coffee consumption was controlled by analysis of urinary NMP, a compound unique in its abundance to coffee and therefore a suitable biomarker for coffee consumption (Lang, Wahl, Stark, & Hofmann, 2012). (For details see Bakuradze et al., submitted as accompanying paper).

ACCEPTED MANUSCRIPT 2.3

Ethics

The study was approved by the Ethics Committee of the medical chamber Rhineland-

T

Palatinate, Mainz, Germany (no. 837.414.10 (7423)) in December 2010. Signed letters of

Quantitation of coffee constituents and metabolites in urine and plasma

SC

2.4

RI P

agreement were obtained from all subjects.

Coffee constituents were quantified by HPLC-DAD and HPLC-SIDA-MS/MS techniques using

MA NU

literature methods (Lang et al., 2012; Weiss et al., 2010). Compliance control was done by analysis of urinary NMP by HPLC-SIDA-MS/MS (Lang et al., 2012). Analysis of chlorogenic acid metabolites in human plasma was done as reported recently (Lang, Dieminger, et al.,

2.5

ED

2013).

Isolation of human peripheral blood lymphocytes and RNA extraction

Determination of RNA-yield, -purity and –integrity

CE

2.6

PT

Performed as described previously (Volz et al., 2012).

AC

Yield and purity of RNA-samples were determined photometrically using Nanodrop2000 (Peqlab). RNA-integrity was electrophoretically analysed via Agilent RNA 6000 Pico Kit by comparing the ratio of 18S and 28S to an intact RNA ladder to gain a RNA-Integrity Number. 2.7

Realtime (Q)-PCR

In order to obtain cDNA out of 1 µg RNA the Quantitect Reverse Transcription Kit (QIAGEN, Hilden, Germany) was used and quantitative realtime PCR was performed according to the manuals of Quantitect SYBR Green PCR and Quantitect Primer Assays (both QIAGEN, Hilden, Germany). All samples were determined in at least duplicates and controls without templates were included. PCR reaction protocol: 15 min at 95°C, 40 cycles of 15 sec at 94°C, 30 sec at 55°C, 30 sec at 72°C. Relative transcription was calculated using the 2-ΔΔCt method

ACCEPTED MANUSCRIPT with normalised Ct-values on endogenous control genes GAPDH (glyceraldehyde 3phosphate dehydrogenase) and RPL13 (ribosomale protein L13a). Sequencing

T

2.8

RI P

Saliva samples of a subset of 59 probands of the intervention study, consenting to

SC

polymorphism analysis, were collected using Oragene Tubes (DNA genotek). DNA was extracted using prepIT (DNA genotek). 80 ng DNA/4 µL H2O were amplified with 4 µL of

MA NU

GoTaq® Flexi buffer (Promega), 0.2 mM dNTP, 2.5 mM MgCl2, 0.13 µL GoTaq® Polymerase, 7.87 µL H2O and each with 0.8 µL NFE2L2 primer (forward primer 5’– GACCACTCTCCGACCTAAAGG–3’ and reverse primer 5’–CGAGATAAAGAGTTGTTTGCGAA-3’) in a Dyad DiscipleTM (BioRad) using the following program: 10 min at 94°C, 30 cycles of: 45

ED

sec. at 94°C, 45 sec. at 58°C and 45 sec. at 72°C, final step: 7 min at 72°C. 4 µL of this product

PT

were electrophoresed in a 2% agarose gel for 75 min at 70-90 Volt, 1000 mA and 150 Watt. 5 µL of PCR product were purified by mixing 1 µL SAP, 1 µL exonuclease and 3 µL H 2O and

CE

incubating as follows: 15 min at 37°C and 15 min at 80°C (ExoSAP-IT® PCR cleanup kit,

AC

Affymetrix). 30 ng DNA / 1.5 µL H2O of the purified product was used for the sequence reaction. For this 1.3 µL of each primer was added respectively, 5 µL of BigDye Terminator v3.1 (Applied Biosystems), 3 µL 5x BDT v3.1 sequencing buffer (Applied Biosystems) and 9.2 µL H2O. Thermal cycling as follows: 1 min at 96°C, 30 cycles of: 10 sec at 96°C, 5 sec at 50°C, 4 min at 60°C, last steps: 5 min at 4°C and 5 min at 10°C. An ethanol precipitation was then performed by adding 2 µL 3 M sodium acetate (pH = 5.2, icecold) and 2 µL 125 mM EDTA (pH = 8.9, icecold) to 20 µL BDT-product. Vortexing and centrifugation at 10.000 g for 5 min. 50 µL 100% ethanol were added and vortexing and centrifugation repeated. Allow to stand for 15 min at RT and centrifuge at 10.000 g and 4°C for 20 min. Supernatant was discarded and 70 µL of 70% ethanol were added. Repeated centrifugation for 5 min and carefully removing

ACCEPTED MANUSCRIPT of supernatant. The pellet was then dried completely using a speed vac and was then resuspended into 15 µL dH2O. These products were stored at -20 °C overnight or directly

T

added into a Micro AmpTM optical 96 well reaction plate (Applied Biosystems) to be analyzed

Pro12Ala PPARγ polymorphism

SC

2.9

RI P

on a 3100 genetic analyzer (Applied Biosystems).

Analysis of the Pro12Ala mutation was carried out according to the method of Ek et al.

MA NU

(1999) with some modifications. Briefly, the 241 bp polymorphic region of the PPAR-2 gene was amplified using 100 ng of genomic DNA with 5.0 µL of 10 x GoTaq Flexi buffer (Promega), 3.0 mM MgCl2, 0.2 mM dNTP, 0.2 mM of each primer (forward primer 5’CAAGCCCAGTCCTTTCTGTG- 3’ and reverse primer 5’- AGTGAAGGAATCGCTTTCCG-3’) and

ED

0.0625 U GoTaq polymerase to a total reaction volume of 25 µL. Amplification was

PT

performed on a Dyad Disciple (BioRad) with an initial step of 94 °C for 10 min followed by 40 cycles of denaturation for 30 s, annealing at 53 °C for 30 s, extension at 72 °C and a final

CE

extension at 72 °C for 9 min. Prior to restriction fragment length polymorphism (RFLP)

AC

analysis PCR products were purified using the GF-1 PCR Clean-up Kit (Vivantis) according to the manufacturer’s instruction following elution in ddH2O (30 µL). The purified products were then either stored at-20 °C or directly used for enzymatic digestion. RFLP was detected after digestion of cleaned-up DNA (15 µL, approximately 500 ng/µL) for 4 h with 1 U of HpaII (New England Biolabs), 2 µL 10 x NEBuffer 1 in a final volume of 20 µL. HpaII cuts the mutant allele at a site introduced by the reverse primer (mismatched base indicated by bold letter). Products were subsequently separated by polyacrylamid gel electrophoresis (12%) for 2 h at 140 V, stained with ethidium bromide and visualized by ultraviolet transillumination. 2.10

Statistics:

ACCEPTED MANUSCRIPT The cross-over design is represented by a statistical model that includes effects of group, treatment and time, as well as interaction of treatments and time and carry-over effect.

T

Details of the model and the appropriate statistical analysis methods are described in (Jones

RI P

& Kenward, 1988). The assumption of normality, underlying the t-Tests was verified by the Shapiro-Wilk test. If normality was rejected, data were logarithmically transformed. If this

SC

transformation was not effective, data were analyzed by Wilcoxon tests.

MA NU

If significant interactions or carry over effects were detected, the data of the second intervention period were excluded from the calculation of treatment effects. Effectiveness of both coffees was compared by two sample t-Test (or Wilcoxon´s rank sum test). To compare data before and after consumption of either market blend (MB) or study blend (SB) coffee

ED

one-sample t-Tests (or Wilcoxon´s signed rank test) on the basis of intra-individual

PT

differences between measurements after and before coffee consumption were performed. The statistical model allows estimation of differences between product effects. This enables

CE

a comparison of the two coffee blends, although the effect of a single coffee blend cannot

AC

be distinguished from time effects in the cross-over design. In order to investigate the impact of Nrf2-polymorphisms volunteers were divided into five subgroups according to their genotype: Wildtype (group 1, N=21), only -617 (group 2, N=7), only -651 (group 3, N=5), only -653 (group 4, N=21) and two polymorphisms simultaneously (group 5, N=5). Groups of Nrf2- and PPARγ polymorphism types were compared with respect to distinctions in Nrf2, PPARγ, IL6 and body weight concerning absolute height of measurements at each BC as well as differences to the preceding wash-out phase. When differences between groups could be detected by Kruskal-Wallis-Test, Wilcoxon rank sum tests followed to compare groups pairwise.

ACCEPTED MANUSCRIPT Statistical hypotheses were tested at the 5% level of significance against a two-sided alternative. Calculations were done by the statistical software SPSS.

T

Spearman correlation coefficients were calculated based on the changes of the first wash-

RI P

out period to the first intervention period. Significant correlations were determined by

MA NU

SC

means of t-Tests.

3 Results 3.1

Metabolites of chlorogenic acid in human blood plasma after coffee consumption

Coffee is a rich source for CGAs, compounds known to exert antioxidant effects in vitro

ED

(Somoza et al., 2003). During coffee making, these water-soluble compounds are nearly

PT

quantitatively extracted into the brew (Lang, Yagar, et al., 2013). After consumption of the coffee brew the chlorogenic acids are heavily metabolized leading to sulfated or

CE

glucuronated hydroxycinnamic acid derivatives with plasma appearance/cmax around 1 h,

AC

and to hydroxyphenylpropanoic acid derivatives appearing roughly 4-6 h after coffee ingestion thanks to the involvement of the intestinal microbiota (Lang, Dieminger, et al., 2013; Stalmach et al., 2009). In our studies blood samples were collected from the probands after an overnight fasting period. Despite their rather short half lifes in human plasma, some of the known metabolites of chlorogenic acids were detected. Particularly the quantitatively important phenols catechol and guaiacol formed from CGAs during roasting, were found as sulfate conjugates in significantly higher plasma concentration during the intervention periods compared to the preceding wash outs, reaching concentrations around 1 µM and 0.1 µM, respectively (table 2). The glucuronides of these compounds were similarly increased, however, at a

ACCEPTED MANUSCRIPT concentration roughly 200 and 50 lower than the respective sulfates, suggesting sulfation was favored. The concentrations of sulfated ferulic acid and i-ferulic acid as well as

T

glucuronated ferulic acid were significantly increased during the intervention periods

RI P

compared to the wash out periods (p < 0.05). Dihydroferulic acid, a metabolite known to originate from microbial processes, appears roughly 4 – 6h after coffee consumption in the

SC

plasma (Lang, Dieminger, et al., 2013). The plasma concentration of this compound was

MA NU

significantly increased during the intervention periods (p < 0.05). The data show that these quantitatively important metabolites of CGA circulate in the vascular system for quite some time before excretion via the urine occurs. It was astonishing to detect short-lived metabolites like catechol sulfate or ferulic acid sulfate present in blood

ED

samples collected 12 h after coffee ingestion. We assume that the 4-week intervention

PT

period with 750 mL of coffee brew distributed over the day led to a continuous supply of coffee constituents and thus a continuous generation of metabolites, resulting in a changed

CE

pharmacokinetic profile different from a that typically obtained after a single intervention

3.2

AC

(Lang, Dieminger, et al., 2013). Impact of coffee consumption on gene transcription

3.2.1. Nrf2 transcription Subjects of group A exhibited slightly higher incoming (BC1) Nrf2-levels as compared to group B with medians 1.36 versus 1.01 (figure 1, table 2). No significant modulation of Nrf2transcription in both groups was observed during the first wash-out period. Female subjects of group A tended to lower Nrf2-transcript levels compared to males (table 2). In group B differences between females and males are not pronounced. After 4 weeks daily coffee consumption merely the market blend (MB) was associated with a significant decrease in Nrf2-levels in subjects of group A (median = 0.69, p < 0.05, figure 1, A, table 2), whereas the

ACCEPTED MANUSCRIPT study blend (SB) showed no significant modulating effect on Nrf2-transcription in group B (median = 1.05, figure 1, B, table 2). Nevertheless different impacts of MB and SB could not

T

be substantiated. In contrast to males, Nrf2-transcript levels in PBLs of females were

RI P

significantly diminished. Neither group A nor group B reached levels of Nrf2 transcription after the second wash-out phase comparable to the first one. This was confirmed statistically

SC

by significant interaction (p < 0.01) and carry over effects (p < 0.01). Therefore, the second

MA NU

intervention period had to be excluded from discussion. 3.2.2. PPARα transcription

At the first visit (BC1) PPARα transcripts were significantly increased in group A (median = 1.51, figure 2, A, table 2), but not in group B (median = 1.03, figure 2, B, table 2) compared to

ED

BC 2 (first wash-out). While female volunteers showed higher transcript levels than male

PT

subjects in group A at BC 1, this was not the case for those of group B where both were on a similar level. The gene response to coffee consumption was more pronounced in group A

CE

after 4 weeks consumption of MB (median = 1.19, table 2) compared to subjects of group B

AC

after consumption of SB (median = 1.11, table 2). Only after consumption of MB a significant increase of PPARα compared to the preceding wash-out phase could be detected (p < 0.01). Regarded in detail this increase was caused mainly by the female subjects which showed a significantly enhanced PPARα-level at BC 3 whereas no such reaction could be seen for male subjects. Due to carry-over effects the second intervention period had to be excluded from discussion.

3.2.3. PPARγ-transcription

ACCEPTED MANUSCRIPT The transcription of lipogenic PPARγ was initially lower in BC 1 compared to BC 2 on average in PBLs of subjects of the study, in group A even on a significant level to median = 0.66 (p <

T

0.01, figure 3, A, table 2). Herein, female test persons showed lower PPARγ-transcript levels

RI P

than males. Coffee consumption of both coffee brews used in this study was associated with a significant reduction of PPARγ-transcripts at BC 3 (group A: median = 0.43, p < 0.001, group

SC

B: median = 0.71, p < 0.01, table 2). In group A a significant reduction of PPARγ transcripts in

MA NU

both female and male subjects was observed, whereas in group B only in male probands a significant decrease was monitored. During the following wash-out phase transcript-levels further declined significantly in both groups. Although the level of the first wash-out (BC 2) was not re-achieved during the second wash-out period (BC 4), statistical interaction and

ED

carry-over effects fall short of the level of significance. The second intervention period was

PT

characterized by significant increases in both groups (figure 3, table 2). The last wash-out phase had a different influence on both groups. While a significant increase of PPARγ

CE

transcripts was observed in group A (median = 1.25, p < 0,001), probands of group B showed

AC

slightly lower levels at BC 6 (median = 0.58). However, since no statistical interaction and carry-over effects were detected, both treatment phases can be used to calculate treatment effects. When data for the respective coffees were pooled, no significant effects of MB or SB on the transcription of PPARγ were apparent when compared to precedent wash-out period. 3.3

IL6 transcription

The results of IL6 transcription during the first BC were characterized by a comparably high incoming level of IL6 in both groups compared to the second BC. Results of BC 1 in group A and B differed significantly from data in BC 2 (median group A = 3.80, figure 4A; median group B = 5.0, figure 4B, table 2, p < 0.001). Differences between BC1 and BC 2 in group A were additionally significant in female subjects, but not in males. Although no significant

ACCEPTED MANUSCRIPT difference can be observed in group B over all, significant results were obtained for each sex if analyzed separately. IL6 transcription of all subjects averagely decreased during the first

T

intervention period, but only in group A after MB consumption on a significant level (median

RI P

= 0.71, p < 0.05, figure 4, A, table 2) although transcripts in group B were also reduced after SB consumption (median = 0.70, figure 4, B, table 2). The following coffee-free period

SC

resulted in a significant increase of IL6 transcripts in both groups (BC 4). During the second

MA NU

coffee period (BC 5), in which group A received SB and group B MB, transcripts declined, merely in group A significantly. Likewise the following increase during the last wash-out phase was present in both groups but again only in group A at a significant level. Since no interaction and carry-over effects were calculated, the data of both intervention periods of

ED

group A and B can be pooled for the respective coffees. With the pooled data, a significant

3.4

CE

PT

decrease of IL6 transcripts was calculated only for SB consumption.

Correlation analysis

AC

Spearman correlation coefficients between genes determined during this human intervention study are listed in table 3. The transcription of Nrf2 showed only a significant negative correlation with β3 adrenergic receptor (r = -0.26), which is involved in a lipolytic pathway that leads to hormone-sensitive lipase (HSL) activation, and a positive one with PPARγ (r = 0.42). Also PPARα, PPARγ and IL6 were correlating significantly with this receptor, PPARα (r = 0.58) and IL6 (r = 0.24) with a positive prefix, PPARγ in a negative manner (r = 0.41). PPARα displayed an additional significant correlation with the satiety regulating leptin receptor (LEPR, data not shown) (r = 0.35). PPARγ transcription revealed except for Nrf2 (r = 0.42) negative significant correlation coefficients in particular with the transcription of β3 adrenergic receptor (r = -0.41), the downstream HSL (r = -0.39), LEPR (r = -0.30), and IL6 (r = -

ACCEPTED MANUSCRIPT 0.23). IL6 itself was significantly correlating with the transcription of β3 adrenergic receptor (r = 0.24), LEPR (r = 0.26) and PPARγ (r = -0.23).

T

In order to get a first hint on possible interactions between single coffee ingredients and the

RI P

analyzed gene transcription Spearman’s Rank calculations were performed with plasma

SC

concentrations of trigonelline, NMP, CGA and its metabolites as well as other selected parameters of the study (table 4). The data indicate a monotone relation of dihydro ferulic

MA NU

acid glucuronide and Nrf2 transcription (r = -0.28). Moreover, the transcription of PPARα showed a significant correlation with this metabolite (r = 0.22), but as well with feruloylglycine (r = 0.22), iso-ferulic acid sulfate (r = 0.26) and trigonelline (r = 0.30). PPARγtranscription revealed no significant correlations with coffee constituents or plasma

ED

metabolites. Transcription of IL6 was correlating significantly with the plasma levels of ferulic

PT

acid glucuronide (r = 0.24) and was associated significantly with the results of body fat

Impact of Nrf2 and PPARγ polymorphisms on inflammatory gene transcription

AC

3.5

CE

measurements (r = 0.25, data see Bakuradze et al., accompanying paper).

Previous studies indicated an influence of genetic polymorphisms in the promoter of the Nrf2 genes on the transcriptional response to coffee (Boettler et al., 2012). Also for the PPARγ gene polymorphisms have been reported which might affect gene transcription and weight regulation (Berhouma et al., 2013). Therefore selected Nrf2 (-C617A, -G651A, A653G) and PPARγ (Pro12Ala) genotypes were analyzed for volunteers who consented to genetic analysis of their saliva DNA (table 5) and compared with respect to the impact on Nrf2, PPARγ, IL6 and body weight.

ACCEPTED MANUSCRIPT Among subjects, 19% were found to carry adenosine instead of cytosine in position -617 of the Nrf2 gene of which 15% were heterozygous and 3% homozygous (table 5). 14% of all

T

analyzed probands were carriers of a polymorphism at position -651 in the same gene. All of

RI P

them were heterozygous. The largest appearance of the analyzed Nrf2 polymorphisms was A653G with 41%. Most of the carriers were heterozygous (34% vs. homozygous 7%). The

MA NU

those were heterozygous, 28% homozygous.

SC

determined Pro12Ala-PPARγ polymorphism had a frequency of 59% in this collective. 31% of

Statistical analysis revealed no influence of the three analyzed single nucleotide polymorphisms (SNPs) of Nrf2 -C617A, -G651A and -A653G on relative levels of Nrf2 transcription. However, due to significant interaction and carry over effects (see 3.1.2.) only

ED

the first coffee intervention period can be used to calculate treatment effects. If the

PT

transcriptional response in both groups (A + B) is summarized irrespective of the consumed coffee, and analyzed with respect to the differences between BC3 (end coffee phase 1) and

CE

BC2 (end wash-out 1), probands carrying the –A653G polymorphism showed less effects on

AC

Nrf2 transcription compared to WT carriers (p<0.05). Furthermore, the Pro12Ala polymorphism of PPARγ had no impact on Nrf2 transcription. IL6 transcription as well as body weight (data see Bakuradze et al., accompanying paper) were also not significantly influenced by all analyzed polymorphisms. However, influences of both polymorphisms (Nrf2 promoter; PPARγ Pro12Ala) could be observed regarding PPARγ transcription (Figures 5 and 6). In group A carriers of -C617A Nrf2 polymorphism had increased PPARγ transcription levels after consumption of MB compared to WT carriers who showed decreased levels (figure 5A). Statistical analysis revealed a significant difference between those groups (p < 0.05). Furthermore, in group A carriers of the polymorphisms either -C617A, -G651A or A653G enhanced significantly (p< 0.05) the level of PPARγ transcription during the first

ACCEPTED MANUSCRIPT coffee phase (BC3 minus BC2) compared to WT carriers. Furthermore significant differences were observed in group A between carriers of the –C617A and the –A653G polymorphisms.

T

Interestingly no significant differences were observed in group B (figure 5B). If the

RI P

transcriptional response in both groups (A + B) is summarized irrespective of the consumed coffee, and analyzed with respect to the differences between BC3 (end coffee phase 1) and

SC

BC2 (end wash-out 1), the transcriptional response of –C617A and –A653G carriers

MA NU

significantly exceeded the effect on WT carriers (p<0.01).

Carriers of Pro12Ala polymorphism in group A had significant different levels of PPARγ than WT carriers at BC1and BC5 (figure 6A, p < 0.05). Also in group B significant differences of PPARγ transcription during both intervention periods (BC3, BC5) were observed (figure 6B, p

ED

< 0.05). If according to the cross-over design the data for the transcription of PPARγ of both

PT

intervention groups were merged with respect to the consumed coffee, statistic significant differences between WT and polymorphism carriers were observed for both coffee blends

AC

CE

(p<0.05).

4 Discussion

In this study the influence of the consumption of two coffee brews on weight-associated inflammatory gene transcription was analyzed in human peripheral blood lymphocytes. The coffees administered in this study contained equal amounts of caffeine but differed in concentrations of CGA, trigonelline and NMP. Plasma concentrations of CGA and trigonelline were therefore higher after MB consumption, containing higher CGA and trigonelline concentrations compared to SB consumption. NMP plasma concentrations reached a maximum after NMP-rich SB consumption.

ACCEPTED MANUSCRIPT Nrf2 is a transcription factor that is mostly implicated with defense of electrophilic and oxidative stress, but is also assumed to be involved into weight-associated inflammation

T

(Hou et al., 2012; Wruck et al., 2011). Transcription of PPARγ and IL6 are discussed to be

RI P

regulated by Nrf2 (Hou et al., 2012; Huang et al., 2010; Wruck et al., 2011). In this study Nrf2 transcription was significantly down-regulated in human peripheral blood lymphocytes after

SC

4 weeks of daily consumption of the CGA- and trigonelline-rich coffee brew (MB) but not

MA NU

after consumption of the NMP-rich SB. Comparable amounts of caffeine in both blends indicate that it seems not to be crucial for Nrf2 modulation. 5-CGA and NMP are already known to enhance Nrf2 activity in vitro (Ute Boettler et al., 2011) but metabolites may mediate different effects. As plasma concentrations of dihydroferulic acid glucuronide

ED

correlate significantly with Nrf2-transcription (R = -0.28) this could possibly be a counterpart

PT

to the parent compound leading to reduction of Nrf2 transcription. This is in agreement with a higher daily intake of CGA after consumption of MB compared to SB (915 vs. 474 mg) and a

CE

greater decrease of Nrf2 transcription by MB. In contrast, trigonelline was already observed

AC

to decrease Nrf2-activity in vitro (Ute Boettler et al., 2011) and may be a further compound of coffee suppressing Nrf2 transcription in this study. This is supported by recent human intervention studies (U. Boettler et al., 2011; Volz et al., 2012) showing that a higher content of trigonelline in a consumed coffee leads to a stronger reduction of Nrf2 transcription and vice versa. In the present study subjects had a daily intake of 330 mg trigonelline after consumption of 750 mL MB, compared to 180 mg trigonelline after SB. However, no significant correlation of trigonelline concentrations in plasma and Nrf2 transcription was observed. Considering the role of Nrf2 in the cellular defense against oxidative stress and electrophiles, these results raise the question on the consequences of decreased Nrf2 transcription levels.

ACCEPTED MANUSCRIPT This down regulated Nrf2 transcription could possibly lead to a downstream diminished translocation of Nrf2 into the nucleus and reduced expression of PPARγ and IL6 (Huang et

T

al., 2010; Wruck et al., 2011). PPARγ and IL6 transcription were indeed observed to be

RI P

significantly reduced in this study after MB consumption. Furthermore, PPARγ transcription was significantly correlated to Nrf2 transcription (R = 0.42). As the transcription factor PPARγ

SC

is mainly involved in the differentiation of preadipocytes into adipocytes via enhancing the

MA NU

expression of lipogenic enzymes (López-Alarcón et al., 2012; Poulsen, Siersbæk, & Mandrup, 2012) it may contribute to weight gain. Besides, it is known to promote the differentiation of monocytes (Stienstra et al., 2007) and to regulate inflammatory genes such as TNFα and IL6 (Prabhakar & Doble, 2011). This is supported by a significant correlation with IL6

ED

transcription (R = -0.23). Chlorogenic acid metabolites were already revealed to decrease

PT

PPARγ transcription in adipocytes (Prabhakar & Doble, 2011). But plasma concentrations of CGA and its metabolites in this study exhibited no significant correlation with PPARγ

CE

transcription, neither have other compounds in this study. In contrast PPARα expression has

AC

already been reported to be enhanced by coffee (Vitaglione et al., 2010), CGA (Li et al., 2009) and different CGA metabolites (Cho et al., 2010). In this study an increase of PPARα transcription was also observed after MB consumption which may be mediated by CGA metabolites or trigonelline. Significant correlations between PPARα and the respective plasma concentrations of these coffee-derived compounds support this hypothesis. PPARα may be activated by the lipolytic generated fatty acids that were observed to be elevated after coffee consumption as well (Riedel et al., accompanying paper). As a consequence activated PPARα may be translocated into the nucleus initiating the transcription of antiinflammatory cytokines or inhibiting the production of pro-inflammatory cytokines like IL6 (Stienstra et al., 2007). The latter indeed was found down regulated in this study after

ACCEPTED MANUSCRIPT consumption of MB. Likewise, PPARγ is known to bind Nf-κB, thereby leading to a suppression of the transcription of genes as TNFα, IL1β and IL6 (Poulsen et al., 2012).

T

Corresponding to another human intervention study caffeine may mediate an increase in

RI P

inflammatory response, as plasma concentrations of IL6 were reported to be enhanced after

SC

4 weeks consumption of a caffeine-containing coffee, but not after a caffeine-free coffee (Wedick et al., 2011). In the present study, IL6 transcription was only significantly decreased

MA NU

in human PBLs by MB consumption although amounts of caffeine were similar in both blends. Consequently other coffee-derived compounds must be taken into account to contribute to the regulation of IL6 transcription as well. CGA may be one of those. It is already known, that they reduce lipopolysaccharide-mediated reactive oxygen species (ROS)

ED

production and decrease phosphorylated I-κB, concentration of nuclear Nf-κB and

PT

transcription of IL6 in different cell lines (Feng et al., 2005; Shi et al., 2013) and rats (di Paola et al., 2010; Wang et al., 2012). Even the CGA metabolites ferulic acid and caffeic acid

CE

possess anti-inflammatory properties, reflected by reduced Nf-κB, TNFα, IL6, activator

AC

protein 1 and cyclooxygenase 2 in vitro and in vivo (N. J. Kang et al., 2009; Kesh et al., 2013). However, a direct link of CGA and their metabolites to IL6 transcription could not be observed in the current study as there were no significant correlations between plasma concentrations and transcription results. Chronically enhanced pro-inflammatory markers as TNFα may provoke T2D via binding to the insulin receptor or phosphorylating insulin receptor substrate 1 (Barbarroja et al., 2010; Stienstra et al., 2007). As IL6 transcription was identified to be the only parameter in this study to correlate with body fat, down regulation of pro-inflammatory IL6 might contribute to the improvement of weight-associated diabetes symptoms by coffee consumption.

ACCEPTED MANUSCRIPT The analysis of the polymorphisms –C617A, -G651A and –A653G in the Nrf2 gene revealed frequencies typically for a European population. Marzec et al. (2007) found out that the

T

frequencies of the SNPs –C617A, -G651A and –A653G are 20%, 10% and 25%. With the

RI P

results of 19% -C617A and 14% -G651A in our study population this is comparable to the results of Marzec et al. (2007). In contrast, the frequency of –A653G with 41% in our study

SC

population is slightly higher compared to their results with 25%. Concerning the frequency of

MA NU

the PPARγ polymorphism Pro12Ala Gouda et al. (2010) have conducted a meta analysis and found frequencies between 6 and 22% in European cohorts. In this work we found a quite higher frequency of Pro12Ala-carriers with 59% which could be due to differences in characteristics as for example BMI. In most of the studies in Gouda et al. (2010) only

ED

overweight people took part. Statistical analysis indicated influences of the analyzed Nrf2

PT

SNP -C617A, -G651A and -A653G as well as the PPARγ polymorphism Pro12Ala on PPARγ transcription. Existence of -C617A Nrf2 or Pro12Ala PPARγ in the genomic DNA of persons

CE

may lead to higher PPARγ transcription after coffee consumption. However, results were not

AC

consistent between groups and time points. Further studies with larger study collective and control group should be conducted to verify or falsify this hypothesis. We hypothesize that MB contained higher amounts of Nrf2-lowering substances than SB entailing a reduced transcription of PPARγ and IL6 as well as enhanced PPARα mRNA. We could demonstrate a simultaneous regulation of these four genes after coffee consumption in human PBL. Possibly, CGA metabolites that are generated within the human organism after coffee consumption may play an essential role in this process. Also individual genotypes as the promoter Nrf2 SNPs -C617A and -A653G and PPARγ Pro12Alapolymorphism may influence this. Transcription of IL6 was positively associated with body fat and may be a link to risk of T2D.

ACCEPTED MANUSCRIPT

Acknowledgement

RI P

T

The project was funded by the German Federal Ministry of Education and Research (BMBF, grant no. 01EA1365) and supported by Tchibo GmbH. The authors G. Bytof, H. Stiebitz and I.

SC

Lantz are employees of Tchibo GmbH. All other authors have declared no conflict of interest. We gratefully acknowledge the most valuable contribution of G. Eisenbrand to the successful

AC

CE

PT

ED

MA NU

realization of the project.

ACCEPTED MANUSCRIPT 5 References

AC

CE

PT

ED

MA NU

SC

RI P

T

Bakuradze, T., Boehm, N., Janzowski, C., Lang, R., Hofmann, T., Stockis, J.-P., Albert, F. W., Stiebitz, H., Bytof, G., Lantz, I., Baum, M., & Eisenbrand, G. (2011). Antioxidant-rich coffee reduces DNA damage, elevates glutathione status and contributes to weight control: Results from an intervention study. Molecular Nutrition & Food Research, 55 (5), 793-797. Barbarroja, N., López‑Pedrera, R., Mayas, M. D., García‑Fuentes, E., Garrido‑Sánchez, L., Macías‑González, M., El Bekay, R., Vidal‑Puig, A., & Tinahones, F. J. (2010). The obese healthy paradox: is inflammation the answer? Biochemical Journal, 430 (1), 141-149. Berhouma, R., Kouidhi, S., Ammar, M., Abid, H., Ennafaa, H., & Benammar-Elgaaied, A. (2013). Correlation of Peroxisome Proliferator-Activated Receptor (PPAR-γ) mRNA Expression with Pro12Ala Polymorphism in Obesity. Biochemical Genetics, 51 (3-4), 256-263. Boettler, U., Sommerfeld, K., Volz, N., Pahlke, G., Teller, N., Somoza, V., Lang, R., Hofmann, T., & Marko, D. (2011). Coffee constituents as modulators of Nrf2 nuclear translocation and ARE (EpRE)-dependent gene expression. The Journal of nutritional biochemistry, 22 (5), 426-440. Boettler, U., Volz, N., Pahlke, G., Teller, N., Kotyczka, C., Somoza, V., Stiebitz, H., Bytof, G., Lantz, I., Lang, R., Hofmann, T., & Marko, D. (2011). Coffees rich in chlorogenic acid or Nmethylpyridinium induce chemopreventive phase II-enzymes via the Nrf2/ARE pathway in vitro and in vivo. Mol Nutr Food Res, 55 (5), 798-802. Boettler, U., Volz, N., Teller, N., Haupt, L., Bakuradze, T., Eisenbrand, G., Bytof, G., Lantz, I., Griffiths, L., & Marko, D. (2012). Induction of antioxidative Nrf2 gene transcription by coffee in humans: depending on genotype? Molecular Biology Reports, 39 (6), 7155-7162. Butt, M. S., & Sultan, M. T. (2011). Coffee and its Consumption: Benefits and Risks. Critical Reviews in Food Science and Nutrition, 51 (4), 363-373. Cho, A.-S., Jeon, S.-M., Kim, M.-J., Yeo, J., Seo, K.-I., Choi, M.-S., & Lee, M.-K. (2010). Chlorogenic acid exhibits anti-obesity property and improves lipid metabolism in high-fat diet-induced-obese mice. Food and Chemical Toxicology, 48 (3), 937-943. di Paola, R., Esposito, E., Mazzon, E., Caminiti, R., Dal Toso, R., Pressi, G., & Cozzocrea, S. (2010). 3,5Dicaffeoyl-4-malonylquinic acid reduced oxidative stress and inflammation in a experimental model of inflammatory bowel disease. Free Radical Research, 44 (1), 74-89. Ek, J., Urhammer, S., Sørensen, T., Andersen, T., Auwerx, J., & Pedersen, O. (1999). Homozygosity of the Pro12Ala variant of the peroxisome proliferation-activated receptor-γ2 (PPAR-γ2): divergent modulating effects on body mass index in obese and lean Caucasian men. Diabetologia, 42 (7), 892-895. Feng, R., Lu, Y., Bowman, L. L., Qian, Y., Castranova, V., & Ding, M. (2005). Inhibition of activator protein-1, NF-kappaB, and MAPKs and induction of phase 2 detoxifying enzyme activity by chlorogenic acid. J Biol Chem, 280 (30), 27888-27895. Fujioka, K. (2002). Management of Obesity as a Chronic Disease: Nonpharmacologic, Pharmacologic, and Surgical Options. Obesity, 10 (S12), 116S-123S. Gouda, H. N., Sagoo, G. S., Harding, A.-H., Yates, J., Sandhu, M. S., & Higgins, J. P. T. (2010). The Association Between the Peroxisome Proliferator-Activated Receptor-γ2 (PPARG2) Pro12Ala Gene Variant and Type 2 Diabetes Mellitus: A HuGE Review and Meta-Analysis. American Journal of Epidemiology, 171 (6), 645-655. Greenberg, J. A., Boozer, C. N., & Geliebter, A. (2006). Coffee, diabetes, and weight control. The American Journal of Clinical Nutrition, 84 (4), 682-693. Higdon, J. V., & Frei, B. (2003). Obesity and Oxidative Stress. Arterioscler Thromb Vasc Biol, 23 (3), 365-367. Hou, Y., Xue, P., Bai, Y., Liu, D., Woods, C. G., Yarborough, K., Fu, J., Zhang, Q., Sun, G., Collins, S., Chan, J. Y., Yamamoto, M., Andersen, M. E., & Pi, J. (2012). Nuclear factor erythroid-derived factor 2-related factor 2 regulates transcription of CCAAT/enhancer-binding protein β during adipogenesis. Free Radical Biology and Medicine, 52 (2), 462-472.

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

MA NU

SC

RI P

T

Huang, J., Tabbi-Anneni, I., Gunda, V., & Wang, L. (2010). Transcription factor Nrf2 regulates SHP and lipogenic gene expression in hepatic lipid metabolism. American Journal of Physiology Gastrointestinal and Liver Physiology, 299 (6), G1211-G1221. ICO. (2011). Monthly Coffee Market Report. In. Jin Son, M., W. Rico, C., Hyun Nam, S., & Young Kang, M. (2010). Influence of Oryzanol and Ferulic Acid on the Lipid Metabolism and Antioxidative Status in High Fat-Fed Mice. Journal of Clinical Biochemistry and Nutrition, 46 (2), 150-156. Johnston, K. L., Clifford, M. N., & Morgan, L. M. (2003). Coffee acutely modifies gastrointestinal hormone secretion and glucose tolerance in humans: glycemic effects of chlorogenic acid and caffeine. The American Journal of Clinical Nutrition, 78 (4), 728-733. Jones, B., & Kenward, M. G. (1988). Design and Analysis of Cross-Over Trial. London, New York: Chapman and Hall. Kang, J. G., & Park, C.-Y. (2012). Anti-Obesity Drugs: A Review about Their Effects and Safety. Diabetes Metab J, 36 (1), 13-25. Kang, N. J., Lee, K. W., Shin, B. J., Jung, S. K., Hwang, M. K., Bode, A. M., Heo, Y.-S., Lee, H. J., & Dong, Z. (2009). Caffeic acid, a phenolic phytochemical in coffee, directly inhibits Fyn kinase activity and UVB-induced COX-2 expression. Carcinogenesis, 30 (2), 321-330. Kempf, K., & Martin, S. (2010). Kaffee und Diabetes. Medizinische Klinik, 105 (12), 910-915. Kesh, S. B., Sikder, K., Manna, K., Das, D. K., Khan, A., Das, N., & Dey, S. (2013). Promising role of ferulic acid, atorvastatin and their combination in ameliorating high fat diet-induced stress in mice. Life Sciences, 92 (17–19), 938-949. Lang, R., Dieminger, N., Beusch, A., Lee, Y.-M., Dunkel, A., Suess, B., Skurk, T., Wahl, A., Hauner, H., & Hofmann, T. (2013). Bioappearance and pharmacokinetics of bioactives upon coffee consumption. Analytical and Bioanalytical Chemistry, 405 (26), 8487-8503. Lang, R., Wahl, A., Stark, T., & Hofmann, T. (2012). Identification of Urinary and Salivary Biomarkers for Coffee Consumption. In Recent Advances in the Analysis of Food and Flavors (Vol. 1098, pp. 13-25): American Chemical Society. Lang, R., Yagar, E. F., Wahl, A., Beusch, A., Dunkel, A., Dieminger, N., Eggers, R., Bytof, G., Stiebitz, H., Lantz, I., & Hofmann, T. (2013). Quantitative Studies on Roast Kinetics for Bioactives in Coffee. J Agric Food Chem, 61 (49), 12123-12128. Li, S.-Y., Chang, C.-Q., Ma, F.-Y., & Yu, C.-L. (2009). Modulating Effects of Chlorogenic Acid on Lipids and Glucose Metabolism and Expression of Hepatic Peroxisome Proliferator-activated Receptor-α in Golden Hamsters Fed on High Fat Diet. Biomedical and Environmental Sciences, 22 (2), 122-129. López-Alarcón, M., Rodríguez-Cruz, M., Vital-Reyes, V. S., Zavala-Ortega, M. I., Hinojosa-Cruz, J. C., Canizales-Quinteros, S., & Fernández, J. (2012). PPARγ2 Pro12Ala polymorphism is associated with improved lipoprotein lipase functioning in adipose tissue of insulin resistant obese women. Gene, 511 (2), 404-410. Lopez-Garcia, E., van Dam, R. M., Qi, L., & Hu, F. B. (2006). Coffee consumption and markers of inflammation and endothelial dysfunction in healthy and diabetic women. Am J Clin Nutr, 84 (4), 888-893. Lopez-Garcia, E., van Dam, R. M., Rajpathak, S., Willett, W. C., Manson, J. E., & Hu, F. B. (2006). Changes in caffeine intake and long-term weight change in men and women. The American Journal of Clinical Nutrition, 83 (3), 674-680. Marzec, J. M., Christie, J. D., Reddy, S. P., Jedlicka, A. E., Vuong, H., Lanken, P. N., Aplenc, R., Yamamoto, T., Yamamoto, M., Cho, H.-Y., & Kleeberger, S. R. (2007). Functional polymorphisms in the transcription factor NRF2 in humans increase the risk of acute lung injury. The FASEB Journal, 21 (9), 2237-2246. Medzhitov, R. (2008). Origin and physiological roles of inflammation. Nature, 454, 428-435. Olthof, M. R., van Dijk, A. E., Deacon, C. F., Heine, R. J., & van Dam, R. M. (2011). Acute effects of decaffeinated coffee and the major coffee components chlorogenic acid and trigonelline on incretin hormones. Nutr Metab (Lond), 8, 10.

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

MA NU

SC

RI P

T

Pi, J., Leung, L., Xue, P., Wang, W., Hou, Y., Liu, D., Yehuda-Shnaidman, E., Lee, C., Lau, J., Kurtz, T. W., & Chan, J. Y. (2010). Deficiency in the Nuclear Factor E2-related Factor-2 Transcription Factor Results in Impaired Adipogenesis and Protects against Diet-induced Obesity. Journal of Biological Chemistry, 285 (12), 9292-9300. Poulsen, L. l. C., Siersbæk, M., & Mandrup, S. (2012). PPARs: Fatty acid sensors controlling metabolism. Seminars in Cell & Developmental Biology, 23 (6), 631-639. Prabhakar, P. K., & Doble, M. (2011). Interaction of Cinnamic Acid Derivatives with Commercial Hypoglycemic Drugs on 2-Deoxyglucose Uptake in 3T3-L1 Adipocytes. J Agric Food Chem, 59 (18), 9835-9844. Rudelle, S., Ferruzzi, M. G., Cristiani, I., Moulin, J., Mace, K., Acheson, K. J., & Tappy, L. (2007). Effect of a Thermogenic Beverage on 24-Hour Energy Metabolism in Humans[ast]. Obesity, 15 (2), 349-355. Shi, H., Dong, L., Dang, X., Liu, Y., Jiang, J., Wang, Y., Lu, X., & Guo, X. (2013). Effect of chlorogenic acid on LPS-induced proinflammatory signaling in hepatic stellate cells. Inflammation Research, 62 (6), 581-587. Somoza, V., Lindenmeier, M., Wenzel, E., Frank, O., Erbersdobler, H. F., & Hofmann, T. (2003). Activity-Guided Identification of a Chemopreventive Compound in Coffee Beverage Using in Vitro and in Vivo Techniques. J Agric Food Chem, 51 (23), 6861-6869. Stalmach, A., Mullen, W., Barron, D., Uchida, K., Yokota, T., Cavin, C., Steiling, H., Williamson, G., & Crozier, A. (2009). Metabolite Profiling of Hydroxycinnamate Derivatives in Plasma and Urine after the Ingestion of Coffee by Humans: Identification of Biomarkers of Coffee Consumption. Drug Metabolism and Disposition, 37 (8), 1749-1758. Stienstra, R., Duval, C., Muller, M., & Kersten, S. (2007). PPARs, Obesity, and Inflammation. PPAR Res, 2007, 95974. Sugden, M. C., Caton, P. W., & Holness, M. J. (2010). PPAR control: it's SIRTainly as easy as PGC. Journal of Endocrinology, 204 (2), 93-104. Thom, E. (2007). The Effect of Chlorogenic Acid Enriched Coffee on Glucose Absorption in Healthy Volunteers and Its Effect on Body Mass When Used Long-term in Overweight and Obese People. The Journal of International Medical Research, 35 (6), 900-908. Tunnicliffe, J. M., & Shearer, J. (2008). Coffee, glucose homeostasis, and insulin resistance: physiological mechanisms and mediators. Applied Physiology, Nutrition, and Metabolism, 33 (6), 1290-1300. van Dam, R. M., & Feskens, E. J. M. (2002). Coffee consumption and risk of type 2 diabetes mellitus. The Lancet, 360 (9344), 1477-1478. Visscher, T. L., & Seidell, J. C. (2001). The Public Health Impact of Obesity. Annual Review of Public Health, 22 (1), 355-375. Vitaglione, P., Morisco, F., Mazzone, G., Amoruso, D. C., Ribecco, M. T., Romano, A., Fogliano, V., Caporaso, N., & D'Argenio, G. (2010). Coffee reduces liver damage in a rat model of steatohepatitis: the underlying mechanisms and the role of polyphenols and melanoidins. Hepatology, 52 (5), 1652-1661. Volz, N., Boettler, U., Winkler, S., Teller, N., Schwarz, C., Bakuradze, T., Eisenbrand, G., Haupt, L., Griffiths, L. R., Stiebitz, H., Bytof, G., Lantz, I., Lang, R., Hofmann, T., Somoza, V., & Marko, D. (2012). Effect of Coffee Combining Green Coffee Bean Constituents with Typical Roasting Products on the Nrf2/ARE Pathway in Vitro and in Vivo. Journal of Agricultural and Food Chemistry, 60 (38), 9631-9641. Wang, S.-x., Guo, H., Hu, L.-m., Liu, Y.-n., Wang, Y.-f., Kang, L.-y., & Gao, X.-m. (2012). Caffeic acid ester fraction from Erigeron breviscapus inhibits microglial activation and provides neuroprotection. Chinese Journal of Integrative Medicine, 18 (6), 437-444. Wedick, N., Brennan, A., Sun, Q., Hu, F., Mantzoros, C., & van Dam, R. (2011). Effects of caffeinated and decaffeinated coffee on biological risk factors for type 2 diabetes: a randomized controlled trial. Nutrition Journal, 10 (1), 93. Weiss, C., Rubach, M., Lang, R., Seebach, E., Blumberg, S., Frank, O., Hofmann, T., & Somoza, V. (2010). Measurement of the Intracellular pH in Human Stomach Cells: A Novel Approach To

ACCEPTED MANUSCRIPT

SC

RI P

T

Evaluate the Gastric Acid Secretory Potential of Coffee Beverages. J Agric Food Chem, 58 (3), 1976-1985. Wruck, C. J., Streetz, K., Pavic, G., Götz, M. E., Tohidnezhad, M., Brandenburg, L.-O., Varoga, D., Eickelberg, O., Herdegen, T., Trautwein, C., Cha, K., Kan, Y. W., & Pufe, T. (2011). Nrf2 Induces Interleukin-6 (IL-6) Expression via an Antioxidant Response Element within the IL-6 Promoter. Journal of Biological Chemistry, 286 (6), 4493-4499. Xu, J., Kulkarni, S. R., Donepudi, A. C., More, V. R., & Slitt, A. L. (2012). Enhanced Nrf2 Activity Worsens Insulin Resistance, Impairs Lipid Accumulation in Adipose Tissue, and Increases Hepatic Steatosis in Leptin-Deficient Mice. Diabetes.

Figure legends

MA NU

Figure 1

Modulation of Nrf2 transcription in human peripheral blood lymphocytes (PBL) during human intervention study of alternated 4 weeks wash-out (between BC 1/2, BC 3/4 and BC

ED

5/6) and 4 weeks of daily coffee consumption (between BC 2/3 and BC 4/5) of two different coffee brews. A: subjects of group A (n = 43) primarily consumed market blend (striped box)

PT

before study blend (double striped box), B: subjects of group B (n = 41) vice versa. Depicted

CE

data are relative transcripts normalized on two endogenous controls and in relation to BC 2. (group A: Wilcoxon signed rank test of ΔCt-values, group B: student’s t-test of logarithmized

Figure 2

AC

ΔCt-values in each case of two neighboring BCs # = p < 0.05; ## = p < 0.01; ### = p < 0.001)

Modulation of PPARα transcripts in human PBL during human intervention study of alternated 4 weeks wash-out (between BC 1/2, BC 3/4 and BC 5/6) and 4 weeks of daily coffee consumption (between BC 2/3 and BC 4/5) of two different coffee brews in a cross over design. A: subjects of group A (n = 43) primarily consumed Market Blend (striped box) before Study Blend (double striped box), B: subjects of group B (n = 41) vice versa. Depicted data are relative transcripts normalized on two endogenous controls and in relation to BC 2. (group A and group B: Wilcoxon signed rank test of ΔCt-values in each case of two neighboring BCs # = p < 0.05; ## = p < 0.01; ### = p < 0.001)

ACCEPTED MANUSCRIPT Figure 3 Modulation of PPARγ transcripts in human PBL during human intervention study of

T

alternated 4 weeks wash-out (between BC 1/2, BC 3/4 and BC 5/6) and 4 weeks of daily

RI P

coffee consumption (between BC 2/3 and BC 4/5) of two different coffee brews in a cross over design. A: subjects of group A (n = 43) primarily consumed Market Blend (striped box)

SC

before Study Blend (double striped box), B: subjects of group B (n = 41) vice versa. Depicted

MA NU

data are relative transcripts normalized on two endogenous controls and in relation to BC 2. (group A: student’s t-test, group B: Wilcoxon signed rank test of ΔCt-values in each case of two neighboring BCs # = p < 0.05; ## = p < 0.01; ### = p < 0.001) Figure 4

ED

Modulation of IL6-transcription in human peripheral blood lymphocytes (PBL) during human

PT

intervention study of alternated 4 weeks wash-out (between BC 1/2, BC 3/4 and BC 5/6) and 4 weeks of daily coffee consumption (between BC 2/3 and BC 4/5) of two different coffee

CE

brews. A: subjects of group A (n = 43) primarily consumed Market Blend (striped box) before

AC

Study Blend (double striped box), B: subjects of group B (n = 41) vice versa. Depicted data are relative transcripts normalized on two endogenous controls and in relation to BC 2. (group A: Wilcoxon signed rank test, group B: student’s t-test of ΔCt-values in each case of two neighboring BCs # = p < 0.05; ## = p < 0.01; ### = p < 0.001) Figure 5 PPARγ transcription plotted against the Nrf2 polymorphisms -C617A, -G651A, -A653G and the combination of two of these SNPs during a human intervention study of alternated 4 weeks wash out (between BC 1/2, BC 3/4 and BC 5/6) and 4 weeks of daily coffee consumption (between BC 2/3 and BC 4/5) of two different coffee brews. A: subjects of group A (n = 43) primarily consumed Market Blend (BC3) before Study Blend (BC5), B:

ACCEPTED MANUSCRIPT subjects of group B (n = 41) vice versa. Depicted data are relative transcripts normalized on two endogenous controls and in relation to BC 2

RI P

T

Figure 6

PPARγ transcription plotted against PPARγ polymorphism Pro12Ala during a human

SC

intervention study of alternated 4 weeks wash out (between BC 1/2, BC 3/4 and BC 5/6) and 4 weeks of daily coffee consumption (between BC 2/3 and BC 4/5) of two different coffee

MA NU

brews. A: subjects of group A (n = 43) primarily consumed Market Blend (BC3) before Study Blend (BC5), B: subjects of group B (n = 41) vice versa. Depicted data are relative transcripts normalized on two endogenous controls and in relation to BC 2. WT: wild type (Wilcoxon

AC

CE

PT

ED

rank sum test *p < 0.05)

ACCEPTED MANUSCRIPT

PT

6 Tables

CR I

Table 1 Concentrations of CGA-metabolites in blood plasma in subjects of the human intervention study after consumption of market blend and

BC 2 (wash-out)

BC 3 (intervention)

BC 4 (wash-out)

group B

MA

group A group A

NU S

study blend

group B

group A

BC 5 (intervention)

group A

group B

(study blend)

( market blend)

group B

D

(market blend) (study blend)

0.498 ± 0.288

1.141 ± 0.643

1.031 ± 0.509

0.435 ± 0.281

0.542 ± 0.335

0.813 ± 0.360

0.863 ± 0.468

Guajacol sulfate

0.034 ± 0.016

0.045 ± 0.034

0.130 ± 0.085

0.111 ± 0.054

0.052 ± 0.038

0.049 ± 0.028

0.119 ± 0.086

0.093 ± 0.062

Isoferulic acid sulfate

0.001 ± 0.001

0.001 ± 0.001

0.004 ± 0.003

0.003 ± 0.002

0.002 ± 0.001

0.002 ± 0.001

0.003 ± 0.003

0.002 ± 0.003

Ferulic acid sulfate

0.005 ± 0.006

0.009 ± 0.011

0.019 ± 0.019

0.014 ± 0.016

0.007 ± 0.007

0.005 ± 0.005

0.016 ± 0.024

0.011 ± 0.017

N-Cinnamoylglycin

0.032 ± 0.020

0.033 ± 0.027

0.069 ± 0.038

0.055 ± 0.040

0.043 ± 0.030

0.038 ± 0.033

0.061 ± 0.039

0.056 ± 0.042

Feruloylglycin

0.020 ± 0.007

0.024 ± 0.022

0.061 ± 0.031

0.035 ± 0.015

0.027 ± 0.010

0.026 ± 0.018

0.035 ± 0.023

0.036 ± 0.027

CE P

0.344 ± 0.152

AC

Catechol sulfate

TE

Concentration [µM]

ACCEPTED MANUSCRIPT

0.003 ± 0.002

0.003 ± 0.003

0.006 ± 0.006

0.010 ± 0.009

0.002 ± 0.002

0.002 ± 0.001

0.004 ± 0.004

0.005 ± 0.006

Guajacol glucuronide

0.002 ± 0.001

0.001 ± 0.000

0.003 ± 0.004

0.003 ± 0.002

0.003 ± 0.001

0.002 ± 0.000

0.003 ± 0.001

0.002 ± 0.002

Ferulic acid glucuronide

0.016 ± 0.018

0.029 ± 0.031

0.037 ± 0.030

0.033 ± 0.037

0.038 ± 0.045

0.045 ± 0.064

0.044 ± 0.058

0.033 ± 0.031

Dihydroferulic acid

0.013 ± 0.005

0.027 ± 0.031

0.053 ± 0.052

0.068 ± 0.036

0.053 ± 0.019

0.022 ± 0.008

0.053 ± 0.026

0.039 ± 0.023

m-Cumaric acid glucuronide

LLOQ

LLOQ

LLOQ

LLOQ

0.031 ± 0.011

0.018 ± 0.008

LLOQ

Ferulic acid

0.231 ± 0.016

0.225 ± 0.015

0.197 ± 0.030

0.212 ± 0.011

0.214 ± 0.017

0.205 ± 0.014

0.207 ± 0.022

0.249 ± 0.021

Dihydroferulic acid

0.041 ± 0.031

0.062 ± 0.079

0.270 ± 0.308

0.162 ± 0.191

0.055 ± 0.063

0.069 ± 0.098

0.144 ± 0.168

0.116 ± 0.181

Chlorogenic acid

LLOQ

LLOQ

LLOQ

LLOQ

LLOQ

LLOQ

LLOQ

LLOQ

CR I

NU S

MA

glucuronide

TE

D

0.041 ± 0.030

CE P

AC

PT

Catechol glucuronide

ACCEPTED MANUSCRIPT Table 2 Transcription results of the human intervention study. Average, median and standard deviation of relative transcription Nrf2, PPARα, PPARγ and IL6 during BC 1, 3 -6 in

1.04 ± 0.87 0.69 0.82 ± 0.7 1.20 ± 0.9 0.6 / 0.8

0.97 ± 0.63 0.66 0.80 ± 0.5 1.09 ± 0.7 0.6 / 0.7

0.86 ± 0.56 0.69 0.79 ± 0.6 0.91 ± 0.5 0.6 / 0.8

1.15 ± 0.49 1.05 1.13 ± 0.4 1.17 ± 0.5 1.2 / 1.0

0.63 ± 0.41 0.49 0.65 ± 0.3 0.61 ± 0.5 0.5 / 0.5

0.80 ± 0.36 0.68 0.85 ± 0.4 0.76 ± 0.3 0.7 / 0.7

1.39 ± 0.60 1.19 1.45 ± 0.5 1.33 ± 0.7 1.3 / 1.1

1.03 ± 0.39 0.97 1.00 ± 0.3 1.05 ± 0.5 0.9 / 1.0

1.30 ± 1.00 1.04 1.16 ± 0.7 1.40 ± 1.2 1.0 / 1.0

1.15 ± 0.43 1.11 1.20 ± 0.5 1.10 ± 0.4 1.1 / 1.1

0.93 ± 0.41 0.86 0.89 ± 0.3 0.96 ± 0.5 0.9 / 0.8

0.95 ± 0.44 0.85 1.00 ± 0.5 0.91 ± 0.4 0.8 / 0.9

0.84 ± 1.10 0.43 0.89 ± 1.2 0.80 ± 1.0 0.3 / 0.6

0.49 ± 0.49 0.29 0.43 ± 0.5 0.53 ± 0.5 0.2 / 0.4

1.63 ± 4.17 0.54 2.83 ± 6.2 0.73 ± 0.7 0.7 / 0.5

1.18 ± 1.79 0.71

0.56 ± 0.55 0.41

2.12 ± 6.60 0.63

ED

PT

CE

SC

1.45 ± 0.94 median both 1.36 f 1.11 ± 0.8 average m 1.61 ± 1.0 median (f/m) 0.8 / 1.6 group B 1.09 ± average both 0.71 median both 1.01 f 1.10 ± 0.6 average m 1.29 ± 0.7 median (f/m) 1.0 / 1.1 group A 1.75 ± average both 0.91 median both 1.51 f 2.02 ± 1.9 average m 1.63 ± 0.9 median (f/m) 1.9 / 1.5 group B 1.12 ± average both 0.39 median both 1.03 f 1.11 ± 0.5 average m 1.12 ± 0.3 median (f/m) 1.0 / 1.0 group A 0.79 ± average both 0.77 median both 0.66 f 1.01 ± 1.1 average m 0.70 ± 0.6 Median (f/m) 0.6 / 0.7 group B 2.60 ± average both 7.14 median both 0.99

AC

PPARγ

BC 4

group A average both

PPARα

BC 3

MA NU

Nrf2

BC 1

BC 5

RI P

sex

T

relation to BC 2. BC 6 1.17 ± 0.93 0.86 0.85 ± 0.5 1.42 ± 1.1 0.7 / 1.1 0.95 ± 0.49 0.95 1.02 ± 0.5 0.89 ± 0.5 1.0 / 0.8 1.56 ± 0.97 1.39 1.30 ± 0.6 1.76 ± 1.2 1.2 / 1.5 1.15 ± 0.57 1.02 1.05 ± 0.4 1.24 ± 0.7 1.0 / 1.0 3.45 ± 6.26 1.25 3.66 ± 4.9 3.29 ± 7.2 1.3 / 1.2 0.83 ± 0.95 0.58

ACCEPTED MANUSCRIPT

0.2 / 0.6

0.6 / 0.7

0.3 / 0.9

1.22 ± 1.4 0.71 1.27 ± 1.6 1.21 ± 1.2 0.5 / 1.1

2.34 ± 2.4 1.35 3.00 ± 3.1 1.70 ± 1.6 1.3 / 1.4

1.98 ± 2.1 0.95 1.46 ± 1.7 2.44 ± 2.5 0.8 / 1.6

2.2 ± 2.0 1.41 1.97 ± 2.2 2.28 ± 1.9 1.0 / 2.5

1.28 ± 1.50 0.70

1.87 ± 1.85 1.18

SC

RI P

T

0.7 / 0.7

2.25 ± 3.61 1.15

3.29 ± 8.27 1.56

1.72 ± 2.0 2.40 ±2.4 3.44 ± 5.0 4.85 ± 11.7 0.89 ± 0.8 1.32 ± 1.1 1.22 ± 1.1 1.72 ± 1.5 0.9 / 0.7

ED PT CE AC

IL6

1.60 ± 2.5 0.48 ± 0.6 1.53 ± 3.2 0.63 ± 0.7 0.82 ± 0.8 0.62 ± 0.5 2.63 ± 8.6 1.01 ± 1.1

MA NU

4.47 ± 10.7 1.17 ± 1.0 median (f/m) 0.8 / 1.0 group A average both 6.41 ± 6.0 median both 3.80 f 7.20 ± 6.7 average m 6.21 ± 6.1 median (f/m) 3.8 / 5.0 group B 10.04 ± average both 13.89 median both 5.00 12.09 ± f average 18.8 8.48 m ± 8.7 median (f/m) 4.2 / 5.0 f average m

1.4 / 1.0

1.2 / 1.0

1.3 / 1.6

ACCEPTED MANUSCRIPT Table 3 Correlation coefficients of determined transcripts. Grey shaded cells indicate significant correlations (*p < 0.05) ADR: adrenergic receptor, GHR: ghrelin, HSL: hormone-

T

sensitive lipase, IL6: interleukin 6, LEPR: leptin receptor, Nrf2: Nuclear factor (erythroid-

RI P

derived 2)-like 2, PPAR: peroxisome proliferator-activated receptor 1Bakuradze et al.,

IL6

LEPR2 Nrf2

-0.11 0.14

0.24* 0.26* -0.26*

PPARα PPARγ 0.58*

-0.41*

MA NU

ADR3

GHR1 HSL3

SC

accompanying paper, 2Riedel et al., accompanying paper, 3data not shown

-0.08 0.03

0.21

0.12

-0.01

0.09

HSL3

0.03

0.02

-0.07

0.01

-0.39*

0.14

-0.23*

0.35*

-0.30*

-0.15

0.42*

ED

GHR1

0.26* 0.17

PT

IL6

PPARα

AC

Nrf2

CE

LEPR2

0.06

-0.10

ACCEPTED MANUSCRIPT Table 4 Correlation coefficients between inflammatory parameters and plasma concentrations of trigonelline, NMP and CGA-metabolites as well as body parameters1, *p <

Trigonelline

-0.18

0.30*

NMP

-0.04

-0.04

Catechol glucuronide

-0.06

0.12

Catechol sulfate

-0.04

Guajacol glucuronide

PPARγ

IL6

RI P

PPARα

0.02

0.13

0.00

0.20

-0.03

0.18

0.03

-0.11

0.05

0.02

0.08

0.06

Guajacol sulfate

-0.05

0.16

0.07

-0.03

m-Cumaric acid glucuronide

-0.01

0.02

0.19

-0.04

-0.02

-0.05

-0.05

0.13

-0.15

0.2

-0.14

0.15

0.22*

-0.09

0.14

-0.07

0.07

0.19

-0.04

Ferulic acid glucuronide

-0.09

0.21

0.03

0.24*

Ferulic acid sulfate

-0.17

0.12

0.04

0.18

Feruloylglycin

-0.18

0.22*

-0.11

0.06

Isoferulic acid sulfate

-0.19

0.26*

0.12

0.06

body weight

-0.02

0.07

-0.16

0.19

body fat

0.07

0.16

0.02

0.25*

Dihydroferulic acid

MA NU

PT

N-Cinnamoylglycin

AC

CE

Dihydroferulic acid glucuronide -0.28* Ferulic acid

SC

-0.03

ED

Nrf2

T

0.05. 1Bakuradze et al., accompanying paper

ACCEPTED MANUSCRIPT Table 5 Frequencies of determined Nrf2- and PPARγ-polymorphisms in subjects of the human intervention study. Detected in DNA of saliva samples by sequencing and restriction

-G651A

-A653G

81% (48)

86% (51)

59% (35)

14% (8)

heterozygous 15% (9)

14% (8)

homozygous

0% (0)

AC

CE

PT

3% (2)

Pro12Ala 41% (24)

41% (24)

59% (34)

34% (20)

31% (18)

7% (4)

28% (16)

ED

Polymorphism 19% (11)

RI P

-C617A

MA NU

WT

PPARγ

SC

Nrf2

T

enzyme/PCR (n=58-59). Numbers in brackets are absolute.

ACCEPTED MANUSCRIPT Figure 1, A

###

T

#

RI P

4

#

SC MA NU

2

1

1. BC

2. BC

3. BC

CE

PT

ED

0

AC

relative transcription

3

4. BC

5. BC

6. BC

ACCEPTED MANUSCRIPT Figure 1, B

4

T

### ###

RI P #

1

1. BC

2. BC

3. BC

CE

PT

ED

0

MA NU

SC

2

AC

relative transcription

3

4. BC

5. BC

6. BC

ACCEPTED MANUSCRIPT Figure 2, A

###

##

5

###

###

RI P

T

4

2,0

SC

1,5

0,5

2. BC

3. BC

4. BC

PT

ED

1. BC

CE

0,0

MA NU

1,0

AC

relative transcription

2,5

5. BC

6. BE

ACCEPTED MANUSCRIPT Figure 2, B

##

##

T

3,0

RI P

2,0

SC

1,5

0,5

1. BC

2. BC

3. BC

4. BC

CE

PT

ED

0,0

MA NU

1,0

AC

relative transcription

2,5

5. BC

6. BE

ACCEPTED MANUSCRIPT Figure 3, A

###

##

###

T

###

10

RI P

##

SC

3,0 2,5

1,5 1,0 0,5 1. BC

2. BC

3. BC

CE

PT

ED

0,0

MA NU

2,0

AC

relative transcription

3,5

4. BC

5. BC

6. BC

ACCEPTED MANUSCRIPT Figure 3, B

###

##

T

##

RI P

10

SC

2,0

1,0 0,5

1. BC

2. BC

3. BC

4. BC

CE

PT

ED

0,0

MA NU

1,5

AC

relative transcription

2,5

5. BC

6. BC

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

MA NU

SC

RI P

T

Figure 4, A

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

MA NU

SC

RI P

T

Figure 4, B

ACCEPTED MANUSCRIPT Figure 5, A

25

SC

6

MA NU

relative transcription

15

RI P

20

T

WT -617 -651 -653 Combination

4

2

0 BC2

AC

CE

PT

ED

BC1

BC3

BC4

BC5

BC6

ACCEPTED MANUSCRIPT Figure 5, B

25

T

WT -617 -651 -653 Combination

RI P

20

SC

6

MA NU

relative transcription

15

4

2

ED

0

BC2

AC

CE

PT

BC1

BC3

BC4

BC5

BC6

ACCEPTED MANUSCRIPT

WT Pro12Ala

RI P

*

8

*

2

0 BC2

BC3

BC4

CE

PT

ED

BC1

MA NU

4

SC

6

AC

relative transcription

20

T

Figure 6, A

BC5

BC6

ACCEPTED MANUSCRIPT

12

WT Pro12Ala

3

*

1

0 BC2

BC3

BC4

CE

PT

ED

BC1

MA NU

2

SC

RI P

*

AC

relative transcription

10

T

Figure 6, B

BC5

BC6

ACCEPTED MANUSCRIPT Highlights

CE

PT

ED

MA NU

SC

RI P

T

Coffee intake affects gene transcription associated with obesity and inflammation Transcription of IL6 in peripheral lymphocytes positively associated with body fat Changes in gene transcription associated with plasma level of CGA metabolites Genetic polymorphisms in the nrf2 promoter and the pparγ-gene Impact of genetic polymorphisms on the transcriptional response to coffee

AC

    