JOURNAL OF CAFFEINE RESEARCH Volume 3, Number 4, 2013 ª Mary Ann Liebert, Inc. DOI: 10.1089/jcr.2013.0025
Association Between Caffeine Intake and the Plasma Proteome in Humans Ouxi Tian,1 Andrea R. Josse,1 Christoph Borchers,2 and Ahmed El-Sohemy1
Background: Caffeine intake has been associated with both an increased and decreased risk of various health conditions. However, many of the physiological pathways affected remain unclear. CYP1A2 is the major enzyme that metabolizes caffeine, and a single nucleotide polymorphism (rs762551) affects the rate of caffeine metabolism. Objective: The aim of this study was to determine the association between caffeine intake and the plasma proteome, and whether CYP1A2 genotype modifies any association. Methods: Subjects (n = 1,095) aged 20–29 years from the Toronto Nutrigenomics and Health Study completed a 196-item semi-quantitative food frequency questionnaire and provided a fasting blood sample from which DNA and plasma were obtained for genotyping and proteomics analysis. Fifty-four proteins were measured by mass spectrometry multiple reaction monitoring (MS-MRM). Subjects were categorized into three groups according to habitual caffeine intake (< 100 mg/d, 100–200 mg/d, and >200 mg/d) and stratified by CYP1A2 genotype. Results: Among carriers of the C allele (slow metabolizers), plasma concentrations of gelsolin isoform 1 were significantly ( p < 0.005) lower among those in the highest category of caffeine intake compared to those with the lowest level of intake. No differences in protein concentration were observed for AA homozygotes (fast metabolizers). Conclusion: These findings show that caffeine intake is associated with lower gelsolin levels only among slow caffeine metabolizers, and suggest that gelsolin might mediate some of the biological effects of caffeine.
Introduction
C
affeine (1,3,7-trimethylxanthine) is the most widely consumed stimulant in the world and has been associated with several diseases, including those of the cardiovascular, neuropsychiatric, and endocrine systems.1–4 Although numerous studies have examined the effects of caffeine on the risk of various diseases, the results have been inconsistent.5–8 Genetic variation that influences caffeine metabolism may modify the association between caffeine intake and disease risk.9,10 Cytochrome P450 1A2 (CYP1A2) is responsible for the majority of caffeine metabolism in the liver.11 A common A/C single nucleotide polymorphism (rs762551) in the CYP1A2 gene is associated with decreased enzyme inducibility and activity.12 Carriers of the C allele can be considered ‘‘slow’’ caffeine metabolizers and tend to break down caffeine at a slower rate.13 Individuals who are homozygous for the A allele are ‘‘fast’’ caffeine metabolizers,13 and may be less susceptible to any effects of caffeine compared to 1 2
slow metabolizers. Cornelis et al. showed that slow metabolizers are at an increased risk of myocardial infarction with increased coffee consumption, while fast metabolizers experienced a protective effect with moderate coffee consumption.9 CYP1A2 genotype was also shown to modify the association between coffee consumption and risk of hypertension.10 Although CYP1A2 variants may modify the association between caffeine and cardiovascular disease risk, the underlying mechanisms and physiological pathways affected by caffeine are not well understood. Recent advances in proteomics, such as the development of mass spectrometry-based multiple reaction monitoring (MRM), allow for the targeted quantification of numerous high-abundance plasma proteins across a wide range of concentrations.14 Many of these highabundance plasma proteins belong to disease-associated pathways, which are useful for predicting disease risk and evaluating the effects of drug and dietary exposures.15,16 For example, a cross-sectional study on dietary patterns and the plasma proteome reported that different dietary habits
Department of Nutritional Sciences, University of Toronto, Toronto, Ontario, Canada. Genome British Columbia Proteomics Centre, University of Victoria, Victoria, British Columbia, Canada.
175
176 and ethnocultural groups have distinct plasma proteomic profiles, which may be predictive of future cardiometabolic disease risk.16 Since the rs762551 variant in CYP1A2 has been shown to modify the association between caffeine and disease risk, it is possible that it also modifies the association between caffeine intake and plasma proteomic profiles. Accounting for genetic variation in caffeine metabolism also minimizes the likelihood of any association being due to residual confounding.17 The objectives of this study were to determine whether different levels of caffeine intake are associated with different concentrations of 54 high-abundance plasma proteins in an ethnically diverse population of healthy young adults, and whether CYP1A2 genotype modifies any associations observed. Materials and Methods
TIAN ET AL. standard procedures.19 Genotyping for the candidate CYP1A2 rs762551 polymorphism was completed by real-time polymerase chain reaction (PCR) using TaqMan SNP Genotyping Assays from Applied Biosystems (Foster City, CA), as described previously.18 Plasma proteomics measurements We measured 54 high-abundance plasma proteins with coefficients of variation < 15%, as described previously.16 In brief, a mass spectrometry-based MRM assay was designed to measure high-abundance plasma proteins belonging to major biological pathways, including hemostasis, innate immunity, lipid metabolism, amyloid disease, transmembrane transport, and signal transduction.14,21–23 Protein measurements were completed at the Genome British Columbia Proteomics Centre at the University of Victoria (Victoria, BC, Canada).
Subjects and data collection Subjects were participants of the Toronto Nutrigenomics and Health Study (TNH), which is a cross-sectional examination of an ethnically diverse group of men and women aged 20–29 years (n = 1,639). Subjects were recruited from the University of Toronto between the fall of 2004 and the fall of 2010. All subjects completed a one-month, 196-item semi-quantitative Toronto-modified Willet food frequency questionnaire (FFQ) and a general health and lifestyle questionnaire (GHLQ), and provided a venous blood sample after a 12-hour fast. Women who were pregnant or breastfeeding and individuals unable to provide a blood sample were excluded from the study. Based on self-reported ethnicity, subjects were categorized into four ethnocultural groups: Caucasians (n = 778), East Asians (n = 565), South Asians (n = 175), and other (n = 119), as described previously.16 Written consent was obtained from all subjects, and the study was approved by the Research Ethics Board at the University of Toronto. Physical activity and anthropometric measurements Physical activity, as measured by questionnaire, was expressed as metabolic equivalent task (MET) hours per week.16 Anthropometric measures, including height and weight, were performed with subjects dressed in light clothing and without shoes, as described previously.16 Resting systolic blood pressure, diastolic blood pressure, and heart rate were measured, as described elsewhere.18 Biochemical measurements and genotyping Blood samples were collected from subjects after a 12-hour overnight fast and analyzed at LifeLabs medical laboratory services (Toronto, ON, Canada). Biomarkers of glycemic control, lipid metabolism, and the systemic inflammatory marker C-reactive protein (CRP) were measured on-site following standard procedures, as described elsewhere.19 At the time of participation, subjects were screened for potential acute inflammatory conditions, and if such a condition was found, a 2-week recovery period was allowed before drawing blood. The homeostasis model of insulin resistance (HOMA-IR) and beta-cell function (HOMA-b) were calculated from previously established formulas.20 DNA was extracted from fasting whole blood samples using
Statistical analysis All statistical analyses were completed using SAS v9.2 (SAS Institute Inc., Cary, NC). During the time of study, proteomics analysis was completed for 1,116 out of 1,639 subjects from the TNH study cohort. Current smokers were excluded from the study, as chemicals in cigarette smoke induce CYP1A2 activity and increase caffeine clearance.24 Furthermore, we excluded seven subjects because of possible over(> 3,500 kcal/day for women, > 4,000 kcal/day for men) or under-reporting of energy intake (< 800 kcal/day), nine subjects because of missing genotype data, and five subjects because of missing anthropometric or biochemical data, which left 1,095 subjects in the final study sample. These subjects were categorized into three levels of caffeine intake (< 100 mg/d, 100–200 mg/d, and > 200 mg/d). Subject characteristics were compared across the three levels of caffeine intake using chi-square tests for categorical variables and analysis of covariance (ANCOVA) for continuous variables with adjustments for age, sex, ethnocultural group, and body mass index (BMI). The distribution of continuous variables was assessed prior to analysis and, when necessary, loge or square-root transformations were used to improve normality. The Tukey–Kramer procedure was used when examining the differences in means across the three intake groups. All reported p-values are two-sided and considered significant when p < 0.05. Biomarkers of glycemic control and lipid metabolism were also compared amongst the three intake groups using ANCOVA adjusted for age, sex, ethnocultural group, and BMI. ANCOVA was used to examine associations between 54 plasma proteins from the proteomics panel and the three categories of caffeine consumption. p-Values from models using transformed variables are reported, and untransformed mean plasma protein concentrations and standard errors are reported for ease of interpretation. Covariates included in the adjusted model (Model 2) were age, sex, ethnocultural group, oral contraceptive use among women, and BMI. p-Values from the unadjusted model (Model 1) were also reported. Significant comparisons of means across the three caffeine intake categories were determined using the Tukey– Kramer procedure. Interactions were tested between caffeine intake and CYP1A2 genotype on plasma protein concentrations. For
CAFFEINE INTAKE AND THE PLASMA PROTEOME IN HUMANS proteins that showed a significant p-value for the interaction term, the associations between caffeine and those proteins were examined further after stratifying by CYP1A2 genotype (A/A vs. A/C + C/C). Results The study population was divided into three groups according to levels of caffeine consumption measured by the FFQ. A total of 660 subjects were in the lowest caffeine consumption category (< 100 mg/d), while 243 subjects and 192 subjects were grouped in the 100–200 mg/d and >200 mg/d intake categories respectively (Table 1). Overall, fasting serum insulin was significantly higher in subjects consuming 200 mg/d of caffeine compared to the two lower levels of caffeine consumption. HOMA-IR was significantly higher among subjects consuming 200 mg/d. HOMA-b was significantly higher in subjects within the lowest intake category when compared to those consuming caffeine at the 100–200 mg/d level.
177
Initially, while a few of the 54 proteins analyzed (specifically, apolipoprotein CIII, coagulatin factor XIIa, and gelsolin) were significantly different between caffeine intake categories in unadjusted models, none of the 54 plasma proteins were significantly different between caffeine intake categories in adjusted models (Table 2). After testing for interactions between caffeine intake and CYP1A2 genotype on plasma protein concentrations, there was a significant interaction only with gelsolin ( p = 0.04). Therefore, gelsolin concentrations were compared between caffeine consumption categories after subjects were stratified by CYP1A2 (rs762551) genotype. After stratification, there were 502 fast metabolizers (A/A) and 593 slow metabolizers (A/C or C/C). Among slow metabolizers, mean plasma concentrations of gelsolin were significantly lower for those consuming >200 mg/d caffeine compared to those consuming 200 mg/d (n = 192)
pValue2
22.7 – 2.5
22.4 – 2.3a3
22.9 – 2.5b
23.4 – 2.7b
< 0.0001
324 (29) 771 (71)
221 (68) 439 (57)
60 (18) 183 (24)
43 (14) 149 (19)
0.002
514 (47) 385 (35) 116 (11) 80 (7) 22.7 – 3.5 7.6 – 3.1 73.8 – 8.9 113.7 – 11.3 68.9 – 8.0
273 (53) 263 (68) 68 (58) 56 (70) 22.6 – 3.4a 7.5 – 3.2 73.5 – 8.8 113.8 – 11.4 68.8 – 7.9
123 (24) 75 (19) 36 (31) 9 (11) 22.6 – 3.6a 7.9 – 3.1 73.5 – 9.0 112.8 – 11.5 68.9 – 8.1
118 (23) 47 (12) 12 (10) 15 (19) 23.5 – 3.6b 7.7 – 2.8 75.3 – 8.8 114.4 – 10.8 69.5 – 8.4
< 0.0001
537(70) 234 (30) 1.4 – 0.9 108.9 – 71.3 4.2 – 0.8 1.6 – 0.4 2.2 – 0.6 2.8 – 0.8 1.0 – 0.5 489.7 – 244.3 4.8 – 0.4 47.7 – 29.3 1.3 – 2.7 1952.4 – 634 5.0 – 8.3
304 (69) 135 (31) 1.5 – 1.0a 110.8 – 66.6a 4.2 – 0.7 1.5 – 0.4 2.2 – 0.6 2.8 – 0.7 1.0 – 0.5 491.5 – 244.4 4.8 – 0.4 49.1 – 30.1a 1.3 – 2.9 1880.1 – 604a 3.6 – 5.5a
130 (71) 53 (29) 1.4 – 0.9ab 105.9 – 78.0b 4.3 – 0.8 1.6 – 0.4 2.3 – 0.7 2.8 – 0.8 0.9 – 0.4 491.8 – 239.7 4.8 – 0.4 46.6 – 29.2b 1.1 – 1.9 2060.4 – 682b 6.6 – 9.2b
103 (69) 46 (31) 1.3 – 0.9b 107.7 – 79.3ab 4.3 – 0.8 1.6 – 0.4 2.3 – 0.6 2.8 – 0.7 1.0 – 0.5 481.5 – 250.5 4.7 – 0.4 44. 9 – 27.0b 1.5 – 2.8 2074.0 – 672b 7.8 – 12.9b
0.90
0.004 0.23 0.09 0.45 0.92
0.01 0.02 0.10 0.17 0.17 0.98 0.83 0.91 0.31 0.01 0.75 < 0.0001 < 0.0001
All values are mean – standard deviation or n (%). p-Value calculated from chi-square analysis for categorical variables and analysis of covariance (ANCOVA) for normalized continuous variables. Except for age, sex, ethnocultural group, and body mass index, p-values for all other variables are adjusted for age, sex, ethnocultural group, and body mass index. 3 Means with different superscript letters are significantly different from one another ( p < 0.05) after Tukey–Kramer adjustment. MET, metabolic task. 2
178
TIAN ET AL. Table 2. Mean Plasma Protein Concentration by Categories of Caffeine Consumption1 p-Value2
Caffeine consumption categories Plasma proteins, lmol/L Alpha-1B-glycoprotein Alpha-2-antiplasmin Alpha-2-macroglobulin Alpha-1-Anti-trypsin Alpha-1-antichymotrypsin Adiponectin Afamin Alpha-1-acid glycoprotein 1 Angiotensinogen Alpha-2-HS-glycoprotein Albumin Apolipoprotein A-I Apolipoprotein A-II precursor Apolipoprotein A-IV Apolipoprotein B-100 Apolipoprotein C-I Apolipoprotein C-III Apolipoprotein D Apolipoprotein E Apolipoprotein L1 Antithrombin-III Zinc-alpha-2-glycoprotein Beta-2-glycoprotein I Complement C1 inactivator Complement C3 Complement C4 beta chain Complement C4 gamma chain Complement C9 Complement factor B Complement factor H Clusterin Ceruloplasmin Vitamin D-binding protein Coagulation factor XIIa HC Fibrinogen alpha chain Fibrinogen beta chain Fibrinogen gamma chain Fibronectin Fibrinopeptide A Gelsolin, isoform 1 Heparin cofactor II Haptoglobin beta chain Hemopexin Histidine-rich glycoprotein Inter-alpha-trypsin inhibitor HC Kininogen-1 L-selectin Plasminogen Prothrombin Plasma retinol-binding protein Serum amyloid P-component Transferrin Transthyretin Vitronectin 1
< 100 mg/d
100–200 mg/d
> 200 mg/d
Model 1
Model 2
1.65 – 0.02 1.92 – 0.02 5.91 – 0.07 11.14 – 0.12 3.37 – 0.03 0.06 – 0.00 0.26 – 0.00 1.76 – 0.03 0.95 – 0.03 8.87 – 0.08 966.5 – 5.8 43.20 – 0.39 24.95 – 0.22 1.42 – 0.02 0.79 – 0.01 3.17 – 0.03 2.33 – 0.03 0.34 – 0.00 0.50 – 0.01 0.42 – 0.01 3.57 – 0.02 1.05 – 0.02 2.77 – 0.02 4.74 – 0.05 19.77 – 0.19 1.46 – 0.02 1.59 – 0.02 2.70 – 0.03 1.46 – 0.02 0.60 – 0.01 1.52 – 0.01 2.31 – 0.04 2.83 – 0.03 0.26 – 0.00 12.16 – 0.26 9.62 – 0.17 9.64 – 0.19 0.63 – 0.05 7.10 – 0.13 1.23 – 0.01 0.69 – 0.01 10.79 – 0.22 10.10 – 0.09 1.33 – 0.02 0.63 – 0.01 2.14 – 0.02 0.07 – 0.00 1.23 – 0.01 0.58 – 0.00 0.91 – 0.01 0.45 – 0.01 12.57 – 0.12 5.74 – 0.05 3.71 – 0.04
1.71 – 0.03 1.95 – 0.03 5.91 – 0.11 11.11 – 0.19 3.33 – 0.05 0.07 – 0.00 0.26 – 0.00 1.75 – 0.04 1.00 – 0.05 8.84 – 0.14 951.4 – 10.7 44.76 – 0.69 25.58 – 0.38 1.46 – 0.03 0.82 – 0.02 3.26 – 0.05 2.49 – 0.05 0.35 – 0.01 0.50 – 0.01 0.42 – 0.01 3.57 – 0.05 1.07 – 0.03 2.84 – 0.05 4.70 – 0.09 19.64 – 0.32 1.43 – 0.03 1.58 – 0.04 2.70 – 0.05 1.44 – 0.03 0.60 – 0.01 1.52 – 0.02 2.35 – 0.06 2.86 – 0.05 0.27 – 0.01 11.86 – 0.34 9.51 – 0.24 9.45 – 0.27 0.59 – 0.06 7.07 – 0.17 1.21 – 0.02 0.71 – 0.01 10.44 – 0.34 10.23 – 0.14 1.32 – 0.03 0.62 – 0.01 2.19 – 0.04 0.07 – 0.00 1.24 – 0.02 0.58 – 0.01 0.94 – 0.02 0.44 – 0.01 12.57 – 0.20 5.77 – 0.09 3.77 – 0.06
1.69 – 0.04 1.94 – 0.03 5.77 – 0.13 11.08 – 0.20 3.44 – 0.06 0.07 – 0.00 0.26 – 0.01 1.79 – 0.05 1.02 – 0.05 8.73 – 0.16 952.8 – 11.7 44.41 – 0.73 25.73 – 0.44 1.42 – 0.03 0.80 – 0.02 3.33 – 0.07 2.49 – 0.07 0.34 – 0.01 0.50 – 0.01 0.42 – 0.01 3.57 – 0.05 1.03 – 0.03 2.82 – 0.05 4.60 – 0.09 19.60 – 0.35 1.45 – 0.04 1.61 – 0.04 2.74 – 0.06 1.46 – 0.03 0.61 – 0.01 1.54 – 0.02 2.41 – 0.07 2.87 – 0.05 0.29 – 0.01 12.30 – 0.34 9.78 – 0.24 9.53 – 0.23 0.63 – 0.05 7.25 – 0.17 1.17 – 0.02 0.72 – 0.02 10.68 – 0.38 10.23 – 0.17 1.32 – 0.03 0.62 – 0.01 2.21 – 0.04 0.07 – 0.00 1.25 – 0.02 0.58 – 0.01 0.95 – 0.02 0.45 – 0.01 12.63 – 0.24 5.75 – 0.09 3.74 – 0.06
0.23 0.68 0.45 0.97 0.35 0.08 0.78 0.78 0.29 0.63 0.23 0.09 0.16 0.68 0.29 0.09 0.01 0.60 0.99 0.93 1.00 0.78 0.47 0.34 0.75 0.90 0.89 0.77 0.70 0.73 0.77 0.20 0.76 0.01 0.41 0.41 0.75 0.29 0.34 0.03 0.18 0.84 0.64 0.93 0.52 0.24 0.26 0.61 0.84 0.15 0.53 0.98 0.96 0.63
0.60 0.76 0.84 0.20 0.26 0.76 0.58 0.48 0.60 0.39 0.55 0.31 0.35 0.82 0.25 0.56 0.06 0.41 0.99 0.32 0.89 0.57 0.57 0.69 0.09 0.83 0.98 0.59 0.29 0.40 0.86 0.91 0.85 0.65 0.59 0.72 0.70 0.38 0.86 0.09 0.93 0.28 0.45 0.90 0.48 0.89 0.15 0.86 0.91 0.28 0.79 0.62 0.37 0.55
Data presented are unadjusted plasma protein means – standard errors. p-Values from ANCOVA with log or square-root transformed plasma protein concentrations where necessary. Model 1 p-values are unadjusted, while Model 2 p-values are adjusted for age, sex, ethnocultural group, oral contraceptive use among women, and body mass index. HC, heavy chain; HS, Heremans–Schmid. 2
CAFFEINE INTAKE AND THE PLASMA PROTEOME IN HUMANS
FIG. 1. Mean plasma concentrations of gelsolin, isoform 1 by caffeine consumption levels in fast (A/A) and slow caffeine metabolizers (A/C and C/C). Unadjusted means and standard errors are shown. p-Values were calculated with analysis of covariance (ANCOVA) using the transformed gelsolin variable and adjusted for age, sex, ethnocultural group, body mass index, and oral contraceptive use among women. In carriers of the C allele, mean plasma gelsolin concentration is significantly lower in subjects consuming >200 mg/d of caffeine compared to those consuming 200 mg/d of caffeine compared to East Asians, South Asians, and others. This consumption pattern is in agreement with the global trend of caffeine intake. Data from Europe and North America show that 90% of adults consume caffeine on a daily basis, with an average intake of 227 mg/d.25,26 Fur-
179
thermore, 75% of dietary sources of caffeine come from drinking coffee in the North American diet.27 In contrast, the preferred caffeinated beverage in East Asian countries such as China and Japan is tea, which has a lower caffeine content compared to coffee on a per cup basis.28 We found no significant associations between caffeine intake and any of the plasma proteins when the entire study population was analyzed. In order to account for the biologically effective ‘‘dose’’ of caffeine and any residual confounding, we tested for the caffeine–CYP1A2 interaction, and further stratified our analysis by the rs762551 SNP in CYP1A2. After stratifying by CYP1A2 genotype, no significant differences in mean plasma protein concentrations were observed between different levels of caffeine consumption among the fast metabolizers. However, gelsolin isoform 1 (gelsolin) concentrations decreased significantly with increasing caffeine intake in the slow metabolizers. In slow metabolizers, caffeine is metabolized less efficiently and remains in the body for a longer period of time.29 Therefore, the sustained effects of caffeine may have contributed to the observed differences in plasma gelsolin concentration in subjects with the slow metabolizer genotype. This observation in the slow caffeine metabolizers supports findings from previous studies that some of the cardiovascular effects of caffeine are modified by CYP1A2 genotype.9,10 The association between caffeine intake and gelsolin concentrations may offer insight into the biological pathways affected by caffeine intake. Gelsolin is a multifunctional actinbinding protein expressed both extracellularly in the plasma (isoform 1) and intracellularly in the cytoplasm (isoform 2).30 Plasma gelsolin is secreted by most cell types in the body with a higher amount being produced by smooth, skeletal, and cardiac muscles.31 This protein is known to cleave and scavenge circulating actin filaments, which helps to counteract the potentially harmful effects of actin release during cellular injury.32 Therefore, plasma gelsolin has been described as a ‘‘debris cleaner’’ that may limit inflammation and reduce blood clotting by preventing the polymerization of actin filaments.33 In addition to its established role as an actin remodelling protein, gelsolin is also involved in numerous physiological processes.32 Altered expression of gelsolin is linked to diseases such as amyloidosis,34 cancer,35 and rheumatoid arthritis.36 Furthermore, low plasma gelsolin levels are associated with increased inflammation and poor prognosis of acute illnesses.37 Therefore, plasma gelsolin is suggested to be a prognostic biomarker for various health outcomes.38 Interestingly, plasma gelsolin levels have also been shown to decrease shortly after acute myocardial infarction.39 Suhler et al. observed that plasma gelsolin decreased by almost half of its normal level in several disease states associated with acute cellular injuries such as myocardial infarction and hepatic failure.39 A study on gelsolin and various types of heart disease also showed that gelsolin levels decreased significantly in patients with acute coronary syndrome.40 Therefore, the observed decrease in plasma gelsolin concentrations in subjects consuming the highest amount of caffeine suggests that gelsolin may be involved in pathways linking caffeine consumption to cardiovascular disease. There are several limitations to the present study. The study population consisted of young adults that are generally healthy and are at low risk of chronic disease. Therefore, plasma proteomic profiles of these subjects may have fewer
180 deviations from normal, and potential biomarkers of disease risk may not be as evident. Because the majority of our study population were not heavy caffeine consumers, a smaller range was used when defining the three caffeine consumption categories. Thus, the effect of heavier caffeine consumption (e.g., > 400 mg/d) on plasma protein concentrations may not have been captured by the current population. Conclusions To our knowledge, this is the first study to examine the association between caffeine intake and the plasma proteome. Plasma gelsolin concentrations were inversely associated with caffeine consumption in carriers of the C allele (slow metabolizer genotype) for rs762551 in CYP1A2. Our study suggests that plasma gelsolin levels are affected by varying amounts of caffeine and this association was unmasked after stratifying by CYP1A2 genotype. These findings also suggest that caffeine may have physiological effects that impact disease-related pathways involving gelsolin such as cardiovascular diseases, in a genetic subset of the population with slow caffeine metabolism. Acknowledgment This work was supported by the Canadian Institutes of Health Research (CIHR; MOP-89829). Author Disclosure Statement A.E.-S. holds shares in Nutrigenomix Inc., a genetic testing company for personalized nutrition. No competing financial interests exist for the remaining authors. References 1. Wang Y, Ho CT. Polyphenolic chemistry of tea and coffee: a century of progress. J Agric Food Chem 2009;57:8109–8114. 2. Bohn SK, Ward NC, Hodgson JM, Croft KD. Effects of tea and coffee on cardiovascular disease risk. Food Funct 2012;3:575–591. 3. Liu R, Guo X, Park Y, et al. Caffeine intake, smoking, and risk of Parkinson disease in men and women. Am J Epidemiol 2012;175:1200–1207. 4. Huxley R, Lee CM, Barzi F, et al. Coffee, decaffeinated coffee, and tea consumption in relation to incident type 2 diabetes mellitus: a systematic review with meta-analysis. Arch Intern Med 2009;169:2053–2063. 5. Panagiotakos DB, Pitsavos C, Chrysohoou C, Kokkinos P, Toutouzas P, Stefanadis C. The J-shaped effect of coffee consumption on the risk of developing acute coronary syndromes: the CARDIO2000 case-control study. J Nutr 2003; 133:3228–3232. 6. Mostofsky E, Schlaug G, Mukamal KJ, Rosamond WD, Mittleman MA. Coffee and acute ischemic stroke onset: the Stroke Onset Study. Neurology 2010;75:1583–1588. 7. Li XJ, Ren ZJ, Qin JW, et al. Coffee consumption and risk of breast cancer: an up-to-date meta-analysis. PLoS One 2013;8: e52681. 8. Braem MG, Onland-Moret NC, Schouten LJ, et al. Coffee and tea consumption and the risk of ovarian cancer: a prospective cohort study and updated meta-analysis. Am J Clin Nutr 2012;95:1172–1181.
TIAN ET AL. 9. Cornelis MC, El-Sohemy A, Kabagambe EK, Campos H. Coffee, CYP1A2 genotype, and risk of myocardial infarction. JAMA 2006;295:1135–1141. 10. Palatini P, Ceolotto G, Ragazzo F, et al. CYP1A2 genotype modifies the association between coffee intake and the risk of hypertension. J Hypertens 2009;27:1594–1601. 11. Kot M, Daniel WA. Caffeine as a marker substrate for testing cytochrome P450 activity in human and rat. Pharmacol Rep 2008;60:789–797. 12. Gunes A, Dahl ML. Variation in CYP1A2 activity and its clinical implications: influence of environmental factors and genetic polymorphisms. Pharmacogenomics 2008;9:625–637. 13. Sachse C, Brockmoller J, Bauer S, Roots I. Functional significance of a C/A polymorphism in intron 1 of the cytochrome P450 CYP1A2 gene tested with caffeine. Br J Clin Pharmacol 1999;47:445–449. 14. Kuzyk MA, Smith D, Yang J, et al. Multiple reaction monitoringbased, multiplexed, absolute quantitation of 45 proteins in human plasma. Mol Cell Proteomics 2009;8:1860–1877. 15. Josse AR, Garcia-Bailo B, Fischer K, El-Sohemy A. Novel effects of hormonal contraceptive use on the plasma proteome. PLoS One 2012;7:e45162. 16. Garcia-Bailo B, Brenner DR, Nielsen D, et al. Dietary patterns and ethnicity are associated with distinct plasma proteomic groups. Am J Clin Nutr 2012;95:352–361. 17. Cahill L, El-Sohemy A. Nutrigenomics: a possible road to personalized nutrition. In: Comprehensive Biotechnology, Vol. 4, second ed. Moo-Young M (ed). Amsterdam: Elsevier, 2011:703–712. 18. Brathwaite JM, Da Costa LA, El-Sohemy A. Catechol-Omethyltransferase genotype is associated with self-reported increased heart rate following caffeine consumption. J Caffeine Res 2011;1:123–130. 19. Cahill L, Corey PN, El-Sohemy A. Vitamin C deficiency in a population of young Canadian adults. Am J Epidemiol 2009;170:464–471. 20. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985;28:412–419. 21. Anderson L. Candidate-based proteomics in the search for biomarkers of cardiovascular disease. J Physiol 2005;563:23–60. 22. DeKosky ST, Ikonomovic MD, Wang X, et al. Plasma and cerebrospinal fluid alpha1-antichymotrypsin levels in Alzheimer’s disease: correlation with cognitive impairment. Ann Neurol 2003;53:81–90. 23. Polanski M, Anderson NL. A list of candidate cancer biomarkers for targeted proteomics. Biomark Insights 2007;1: 1–48. 24. Schreiber GB, Robins M, Maffeo CE, Masters MN, Bond AP, Morganstein D. Confounders contributing to the reported associations of coffee or caffeine with disease. Prev Med 1988; 17:295–309. 25. Reissig CJ, Strain EC, Griffiths RR. Caffeinated energy drinks—a growing problem. Drug Alcohol Depend 2009;99: 1–10. 26. Temple JL. Caffeine use in children: what we know, what we have left to learn, and why we should worry. Neurosci Biobehav Rev 2009;33:793–806. 27. Nawrot P, Jordan S, Eastwood J, Rotstein J, Hugenholtz A, Feeley M. Effects of caffeine on human health. Food Addit Contam 2003;20:1–30. 28. Heckman MA, Weil J, Gonzalez de Mejia E. Caffeine (1,3,7trimethylxanthine) in foods: a comprehensive review on
CAFFEINE INTAKE AND THE PLASMA PROTEOME IN HUMANS
29. 30.
31.
32.
33.
34.
35. 36.
consumption, functionality, safety, and regulatory matters. J Food Sci 2010;75:R77–87. Cornelis MC, El-Sohemy A. Coffee, caffeine, and coronary heart disease. Curr Opin Lipidol 2007;18:13–19. Silacci P, Mazzolai L, Gauci C, Stergiopulos N, Yin HL, Hayoz D. Gelsolin superfamily proteins: key regulators of cellular functions. Cell Mol Life Sci 2004;61:2614–2623. Kwiatkowski DJ, Mehl R, Izumo S, Nadal-Ginard B, Yin HL. Muscle is the major source of plasma gelsolin. J Biol Chem 1988;263:8239–8243. Li GH, Arora PD, Chen Y, McCulloch CA, Liu P. Multifunctional roles of gelsolin in health and diseases. Med Res Rev 2012;32:999–1025. Le HT, Hirko AC, Thinschmidt JS, et al. The protective effects of plasma gelsolin on stroke outcome in rats. Exp Transl Stroke Med 2011;3:1–9. Solomon JP, Page LJ, Balch WE, Kelly JW. Gelsolin amyloidosis: genetics, biochemistry, pathology and possible strategies for therapeutic intervention. Crit Rev Biochem Mol Biol 2012;47:282–296. Kwiatkowski DJ. Functions of gelsolin: motility, signaling, apoptosis, cancer. Curr Opin Cell Biol 1999;11:103–108. Osborn TM, Verdrengh M, Stossel TP, Tarkowski A, Bokarewa M. Decreased levels of the gelsolin plasma isoform in
37. 38.
39.
40.
181
patients with rheumatoid arthritis. Arthritis Res Ther 2008; 10:R117. DiNubile MJ. Plasma gelsolin as a biomarker of inflammation. Arthritis Res Ther 2008;10:124. Peddada N, Sagar A, Ashish, Garg R. Plasma gelsolin: a general prognostic marker of health. Med Hypotheses 2012;78: 203–210. Suhler E, Lin W, Yin HL, Lee WM. Decreased plasma gelsolin concentrations in acute liver failure, myocardial infarction, septic shock, and myonecrosis. Crit Care Med 1997;25:594–598. Yue L, Huijun Y, Yuerong J, Mei X, Keji C. Correlation between platelet gelsolin level and different types of coronary heart disease. Heart. 2011;97:A137–A137.
Address correspondence to: Ahmed El-Sohemy Department of Nutritional Sciences Room 350 University of Toronto 150 College Street Toronto, Ontario Canada, M5S 3E2 E-mail:
[email protected]
