Human Nutrition and Metabolism Coffee Oil Consumption Increases Plasma Levels of 7␣-Hydroxy-4- cholesten-3-one in Humans1 Mark V. Boekschoten,*†2 Maaike K. Hofman,*,** Rien Buytenhek,** Evert G. Schouten,* Hans M. G. Princen,** and Martijn B. Katan*† *Division of Human Nutrition, Wageningen University, Wageningen, The Netherlands; †Wageningen Centre for Food Sciences, Wageningen, The Netherlands; and **TNO Prevention & Health, Gaubius Laboratory, Leiden, The Netherlands ABSTRACT Unfiltered coffee brews such as French press and espresso contain a lipid from coffee beans named cafestol that raises serum cholesterol in humans. Cafestol decreases the expression and activity of cholesterol 7␣-hydroxylase, the rate-limiting enzyme in the classical pathway of bile acid synthesis, in cultured rat hepatocytes and livers of APOE3Leiden mice. Inhibition of bile acid synthesis has been suggested to be responsible for the cholesterol-raising effect of cafestol. Therefore, we assessed whether cafestol decreases the activity of cholesterol ␣ ␣ 7 -hydroxylase in humans. Because liver biopsies were not feasible, we measured plasma levels of 7 -hydroxy- Downloaded from 4-cholesten-3-one, a marker for the activity of cholesterol 7␣-hydroxylase in the liver. Plasma 7␣-hydroxy-4- cholesten-3-one was measured in 2 separate periods in which healthy volunteers consumed coffee oil containing cafestol (69 mg/d) for 5 wk. Plasma levels of 7␣-hydroxy-4-cholesten-3-one increased by 47 Ϯ 13% (mean Ϯ SEM, n ϭ 38, P ϭ 0.001) in the first period and by 23 Ϯ 10% (n ϭ 31, P ϭ 0.03) in the second treatment period. Serum cholesterol was raised by 23 Ϯ 2% (P Ͻ 0.001) in the first period and by 18 Ϯ 2% (P Ͻ 0.001) in the second period. We corrected individual 7␣-hydroxy-4-cholesten-3-one levels for serum cholesterol levels, because coffee oil jn.nutrition.org increases serum cholesterol and 7␣-hydroxy-4-cholesten-3-one is probably present in the lipoprotein fraction of serum. After correction, the increase in 7␣-hydroxy-4-cholesten-3-one was 24 Ϯ 11% (P ϭ 0.04) in the first period and there was no effect in period 2. Our study showed that coffee oil did not decrease, and actually increased, plasma levels of 7␣-hydroxy-4-cholesten-3-one in humans in 2 separate treatment periods. Therefore, this study does not support the hypothesis that cafestol decreases bile acid synthesis in humans. J. Nutr. 135: 785–789, by on March 26, 2008 2005. KEY WORDS: ● coffee oil ● cafestol ● cholesterol ● bile acid synthesis Humans and animals regulate their metabolism in response cholesterol by 0.13 mmol/L (4). Therefore, elucidation of the to dietary compounds. Elucidation of underlying molecular mechanism by which cafestol achieves this large effect on mechanisms using in vitro and animal studies allows us to serum cholesterol would be helpful in understanding how identify novel dietary components that can affect metabolism certain food components can affect lipid metabolism. A num- and potentially health. However, before one can apply knowl- ber of potential mechanisms have been proposed, such as edge of molecular mechanisms to identify such compounds in downregulation of the LDL receptor (5) and regulation of lipid the human diet, it must be established that the potential transfer proteins such as cholesteryl ester transfer protein, mechanism indeed is important in humans. phospholipids transfer protein, and lecithin:cholesterol acyl- In this study we used coffee oil to explore a new potential transferase (6,7). However, it remains unclear by which mech- pathway for the mode of action of food components that affect anism cafestol affects these proteins. serum lipids. Oil from coffee beans contains the diterpenes Cafestol feeding causes a rise in serum cholesterol in cafestol and kahweol, which are responsible for the cholester- APOE3Leiden mice after 3 wk. This rise is accompanied by a ol-raising effect of unfiltered coffee types (1–4). Cafestol raises decrease in cholesterol 7␣-hydroxylase (Cyp7a1)3 expression serum cholesterol more potently than the related diterpene of 58% (8) and fecal bile acid content is reduced by 41% after kahweol, which is also found in coffee beans (3). The com- cafestol feeding (8). Cyp7a1 is the rate-limiting enzyme in the bined results of 11 intervention trials show that each 10 mg of classical pathway of bile acid synthesis from cholesterol in the cafestol after 4 wk of daily consumption raises serum total liver. When rats or mice are fed cholesterol, Cyp7a1 is up- regulated; this enzyme converts the extra cholesterol into bile 1 Supported by the Netherlands Organisation for Scientific Research (Grant 98–10-001) and the Wageningen Centre for Food Sciences (Project A-201-C). 3 Abbreviations used: ALAT, alanine aminotransferase; ASAT, aspartate ami- 2 To whom correspondence should be addressed. notransferase; Cyp7a1, cholesterol 7␣-hydroxylase; LXR, liver X receptor; PXR, E-mail: [email protected]. pregnane X receptor. 0022-3166/05 $8.00 © 2005 American Society for Nutritional Sciences. Manuscript received 14 October 2004. Initial review completed 14 November 2004. Revision accepted 14 January 2005. 785 786 BOEKSCHOTEN ET AL. acids, which are excreted with the feces. In this way, mice do Subjects were asked to maintain their lifestyles and dietary habits not exhibit the rise in serum cholesterol induced by dietary for the duration of the study. They reported in diaries changes in diet, cholesterol that occurs in other species. smoking, physical activity, use of medication, illness, and the number Suppression of bile acid synthesis increases the levels of of capsules taken daily. Laboratory measurements. Blood samples were taken after sub- hepatic cholesterol and this could lead to downregulation of jects had fasted overnight on 4 separate days in the last 2 wk of the the LDL receptor. This could explain the increase in serum run-in and coffee oil stages. Both serum and plasma samples were LDL-cholesterol in APOE3Leiden mice fed cafestol. In hu- prepared. Levels of total cholesterol, HDL-cholesterol, triglycerides, mans, cafestol intake also increases serum LDL-cholesterol. ALAT, and ASAT were measured in the 16 serum samples of each Whether this is caused by a decrease in Cyp7a1 activity is not person as previously described (13). LDL-cholesterol levels were known. However, Cyp7a1 deficiency in humans does increase calculated (15). serum lipid levels (9). Due to the differences between mice In addition, a number of liver function markers were measured in and humans in regulation of bile acid synthesis, it is essential the serum samples: alkaline phosphatase, bilirubin, ␥-glutamyl to establish whether cafestol also regulates activity of Cyp7a1 transpeptidase, and lactate dehydrogenase. We also measured levels in humans. of amylase, a marker for pancreatitis. All tests were performed using Flex reagents cartridges with the Dimension system (Dade Behring). Because it is ethically unacceptable to take liver biopsies to Levels of bile acids were also determined in serum using an Enzabile study the effect of cafestol on expression and activity of kit (Bio-Stat Diagnostic Systems). Cyp7a1 in human livers, we used an indirect method. In this For the analysis of the bile acid precursor 7␣-hydroxy-4-cholesten- method, the level of 7␣-hydroxy-3-cholesten-4-one, a metab- 3-one we pooled plasma samples per subject. This yielded 4 plasma olite of cholesterol, is measured in plasma. 7␣-Hydroxy-3- samples, 1 for each run-in and 1 for each coffee oil stage. We cholesten-4-one is an intermediate in bile acid synthesis and determined plasma levels of 7␣-hydroxy-4-cholesten-3-one using the its level in human plasma is considered to reflect Cyp7a1 HPLC method described by Ga¨lman et al. (12). We were not able to activity in the liver (10–12). measure 7␣-hydroxy-4-cholesten-3-one levels in samples of 2 subjects in period 1 and of 1 subject in period 2 due to the presence of an Downloaded from unidentified compound in the plasma that interfered with our inter- METHODS nal standard. In total we measured 7␣-hydroxy-4-cholesten-3-one levels in samples of 38 subjects in period 1 and of 31 subjects in period Subjects. Subjects were recruited and screened as previously 2. The 4 samples were analyzed within 1 run. The within-run CV described (13). Fifty subjects, including 13 men and 37 women, were was 9.7%. enrolled. During the study 5 subjects withdrew: 3 subjects suffered Statistics. A subject’s response to coffee oil in period 1 was from stomach complaints, 1 went abroad, and 1 had a gastrointestinal defined as the mean level at the end of coffee oil stage 1 minus the jn.nutrition.org infection. Another 13 subjects were excluded during the study be- mean at the end of run-in 1. For the samples in which liver function cause their serum activities of alanine aminotransferase (ALAT) and variables and bile acids were determined, the mean for a period was aspartate aminotransferase (ASAT) exceeded previously determined calculated as the mean of the 4 repeated measurements in each stage. boundaries of 2.7 times the upper limit of normal for ALAT and 1.5 The response in period 2 was calculated in the same way. We times the upper limit of normal for ASAT. The Medical Ethical compared means using Student’s t test for paired samples. Values are Committee of Wageningen University and Research Centre ap- means Ϯ SEM. Differences were considered significant at P Ͻ 0.05. by on March 26, 2008 proved the study. Each volunteer gave written informed consent. Study design. The study was originally designed to assess the within-subject reproducibility of the serum lipid response to coffee RESULTS oil. Therefore, the response to coffee oil was measured in 2 separate treatment periods (13).