Article

pubs.acs.org/JAFC

In Vitro Bile Acid Binding Capacities of Red Leaf Lettuce and Cruciferous Isabelle F. Yang, Guddadarangavvanahally K. Jayaprakasha,* and Bhimanagouda S. Patil* and Fruit Improvement Center, Department of Horticultural Sciences, Texas A&M University, 1500 Research Parkway, Suite A120, College Station, Texas 77843, United States

*S Supporting Information

ABSTRACT: In the present study, we tested the bile acid binding capacity of red leaf lettuce, red , red , green kale, and Brussels sprouts through in vitro digestion process by simulating mouth, gastric, and intestinal digestion using six bile acids at physiological pH. Green and red kale exhibited significantly higher (86.5 ± 2.9 and 89.7 ± 0.9%, respectively) bile acid binding capacity compared to the other samples. Further, three different compositions of bile acids were tested to understand the effect on different health conditions. To predict the optimal dose for bile acid binding, we established a logistic relationship between kale dose and bile acid binding capacity. The results indicated that kale showed significantly higher bile acid binding capacity (82.5 ± 2.9% equivalent to 72.06 mg) at 1.5 g sample and remained constant up to 2.5 g. In addition, minimally processed (microwaved 3 min or steamed 8 min) green kale showed significantly enhanced bile acid binding capacity (91.1 ± 0.3 and 90.2 ± 0.7%, respectively) compared to lyophilized kale (85.5 ± 0.24%). Among the six bile acids tested, kale preferentially bound hydrophobic bile acids chenodeoxycholic acid and deoxycholic acid. Therefore, regular consumption of kale, especially minimally processed kale, can help excrete more bile acids and, thus, may lower the risk of hypercholesterolemia. KEYWORDS: oleracea, kale, red leaf lettuce, bile salt, in vitro digestion

1. INTRODUCTION membranes and enhance the defense against oxidative stress, 12 Bile acids are amphipathic molecules that function as thus protecting the liver. Bile acid sequestrants prevent bile emulsifiers to assist in digestion and absorption of lipids in acid absorption and promote synthesis of primary bile acids from cholesterol, which leads to a reduction of serum the gastrointestinal tract. Cholesterol can be biosynthesized 13 into bile acids, including cholic acid (CA), chenodeoxycholic cholesterol. Some foods can function like bile acid acid (CDCA), lithocholic acid (LCA), deoxycholic acid (DCA), sequestrants; for instance, guar gum can reduce the glycochenodeoxycholic acid (GCDCA), glycocholic acid concentration of serum bile acids by binding them in the 1 gastrointestinal tract, which relieves the symptoms associated (GCA), and glycodeoxycholic acid (GDCA). The two primary 14 bile acids, CA and chenodeoxycholic acid, are synthesized by with intrahepatic cholestasis of pregnancy and reduces plasma cholesterol levels.15 the liver, stored in the gallbladder, and then secreted into the fi small intestine, when food is being digested.2 Human intestinal Beyond their functions in digestive emulsi cation and bacteria can convert CA and chenodeoxycholic acid to reduction of cholesterol, bile acids are also associated with secondary bile acids, such as DCA and LCA, though glucose . Because bile acids can activate farnesoid X dehydroxylation. After assistance in lipid digestion, bile acids receptor and G-protein coupled bile acid receptor 1, which are are reabsorbed through the ileum and transferred back to the involved in glucose metabolism, the control of bile acid 1,2 absorption can be used as a potential novel therapy for some liver. Up to 95% of bile acids are reabsorbed, which limits the 2,14 synthesis of new bile acids from cholesterol. Some bile acids metabolic diseases, such as type 2 diabetes. In fact, one of can escape the enterohepatic circulation and reach the colon, the bile acid sequestrants, colesevelam hydrochloride, has been 1 approved by the U.S. Food and Drug Administration (FDA) in where gut bacteria form more DCA and LCA. LCA, DCA, and 16 chenodeoxycholic acid are hydrophobic bile acids, which are 2008 as a medicine for treatment of type 2 diabetes. In considered more toxic to liver and colon cells, and may induce addition, bile acid binding has recently emerged as a supporting 3−5 fi approach to strengthen the standard treatments used to control colon cancer. However, dietary ber increases the excretion 17 of bile acids in the stool,6 which can potentially prevent colon plasma glucose in type 2 diabetic patients. Several studies 7 fi suggested that bile acid sequestrants reduced the fasting plasma cancer and lower cholesterol levels. Besides dietary ber, 17−20 D and calcium supplements also have a protective glucose level in type 2 diabetic patients. Therefore, a function against colon cancer.8 Hence, binding bile acids can vegetable with good bile acid binding capacity may also lower reduce intestinal cancer and reduce plasma cholesterol the plasma glucose level of type 2 diabetic patients. levels.9,10 When hydrophobic bile acids are conjugated with taurine or Received: June 2, 2017 glycine, they become more hydrophilic and much less toxic to Revised: August 7, 2017 hepatic and intestinal cells.11 Some conjugated bile acids, such Accepted: August 16, 2017 as tauroursodeoxycholic acid, can even stabilize hepatocyte Published: August 16, 2017

© 2017 American Chemical Society 8054 DOI: 10.1021/acs.jafc.7b02540 J. Agric. Food Chem. 2017, 65, 8054−8062 Journal of Agricultural and Food Chemistry Article

Figure 1. Schematic representation of in vitro bile acid binding capacity of different vegetables and extraction of bound bile acids.

Previous studies have examined the bile acid binding amylase. The samples were vortexed and then placed in a shaking capacities of various vegetables, but few studies have examined water bath (37 °C and 180 rpm) for 5 min. During the gastric the binding preferences for individual bile acids by specific digestion phase, the mixture from the oral digestion phase was 21 adjusted to pH 2.0 with 1 N HCl, followed by the addition of 600 μL vegetables. According to previous research, greens, ff μ kale, mustard greens, , Brussels sprouts, spinach, green of pepsin bu er (200 g of pepsin in 1 mL of 0.1 M HCl), vortexed and incubated in a shaking water bath (37 °C and 180 rpm) for 1 h. bell peppers, and cabbage have good bile acid binding 21 For the intestinal phase, chyme from the gastric digestion phase was capacity. However, whether hydrophobic or hydrophilic bile first adjusted to pH 6.8 with 1 N NaOH. The sample was then mixed acids are preferentially bound by these vegetables has yet to be with 5 mL of pancreatin (6.25 mg/mL in 50 mM phosphate buffer) tested. Whether the different compositions of bile acid mixtures and 4 mL of bile acid mixture solution prior to incubation in a shaking − can influence the in vitro binding capacity also remains water bath, shaking at 37 °C and 180 rpm, for 3 h.22 25 The bile acid unknown. This study compares the in vitro bile acid binding mixture contained 10 mM cholate, 10 mM deoxycholate, 10 mM capacities of different vegetables, including red cabbage, red leaf glycochenodeoxycholate, 10 mM glycocholate, and 10 mM cheno- lettuce, Brussels sprouts, red kale, and green kale. The most deoxycholate in potassium phosphate buffer (pH 6.8). After ff incubation with the bile acids, the in vitro digestion was terminated active vegetable was used to test di erent bile acid ° compositions and optimal amounts for bile acid binding. by inactivating enzymes in a 78 C water bath for 7 min. The analyses were conducted using triplicate samples of freeze-dried red cabbage, red kale juice, green kale, red leaf lettuce, and Brussels sprouts (2.0 g 2. MATERIALS AND METHODS each). Digestion chymes were centrifuged at 3600 rpm for 30 min, and fi fi 2.1. Materials. Fresh green kale ( var. acephala), supernatants were collected and ltered through Whatman lter paper. red cabbage (Brassica oleracea var. capitata f. rubra), red kale (Brassica The residue was washed with 20 mL of nanopure water by mixing on a shaking water bath for 3 h and centrifuged at 800g for 30 min. Both of oleracea convar. acephala var. sabellica), Brussels sprouts (Brassica fi oleracea var. gemmifera), and red leaf lettuce (Lactuca sativa) were the supernatants were ltered, combined, and concentrated under ° purchased from a local supermarket. All vegetables were cut into small vacuum by rotary evaporation at 40 C to obtain 7 mL and passed μ fi pieces, lyophilized, powdered, and stored at −20 °C until further use. through 0.45 m cellulose lters for high-performance liquid 2.2. Chemicals. Sodium glycocholate (G7132), sodium cholate chromatography (HPLC) analysis. In Vitro ff (C1254), sodium glycochenodeoxycholate (G0759), sodium glyco- 2.4. Digestion of Di erent Bile Acid Compositions. ff deoxycholate (G9910), sodium chenodeoxycholate (C8261), sodium Bile acid composition di ers in various individuals depending upon deoxycholate (D6750), ammonium nitrate, potassium dihydrogen their health conditions and gender. To understand that the binding fl phosphate, potassium chloride, potassium citrate, uric acid sodium salt, capacity is in uenced by the composition of the bile acid mixture, we ff urea, lactic acid sodium salt, porcine gastric mucin, α-amylase, pepsin, used three di erent compositions equivalent to male patients with and pancreatin were obtained from Sigma-Aldrich Co. (St. Louis, MO, gallstones (BAC-1), healthy females (BAC-2), and males with type 2 26,27 U.S.A.). diabetes (BAC-3). Mixtures with three different bile acid 2.3. Screening of Different Vegetables for Bile Acid Binding compositions (BACs) were prepared as shown in Table S2 of the Ability. The in vitro digestion was performed to mimic digestion in Supporting Information. The BAC-1 bile acid mixture contained the mouth, stomach, and small intestine (Figure 1). Simulated saliva 33.3% CA in both conjugated and unconjugated forms, 24.9% DCA, fluid (SSF) was prepared by dissolving the reagents (Table S1 of the and 41.8% chenodeoxycholic acid. The BAC-2 mixture contained Supporting Information) in pH 6.8 phosphate buffer.22,23 In the oral 33.1% CA, 17% DCA, and 49.9% chenodeoxycholic acid. The BAC-3 digestion phase, 2.0 g of lyophilized vegetable was mixed with an equal mixture contained 33.3% CA, 39.3% DCA, and 27.4% chenodeox- weight of water and 10 mL of the SSF containing 0.31 mg of α- ycholic acid. Lyophilized green kale was used for digestion steps

8055 DOI: 10.1021/acs.jafc.7b02540 J. Agric. Food Chem. 2017, 65, 8054−8062 Journal of Agricultural and Food Chemistry Article according to the previously described in vitro digestion protocol, and Table 1. Calibration Curves Used for Quantification of during the intestinal digestion phase, 4 mL of the bile acid mixture Bound Bile Acids with different compositions was used. Except the bile acid mixture, other steps of the in vitro experiment are the same, as shown in Figure bile acid equation amount (mg) 1. After in vitro digestion, the supernatant was concentrated by rotary glycocholate y = 83017x + 52.71 2.42 fi fi evaporation and ltered through a 0.45-micron cellulose lter for cholate y = 9990.8x + 3.15 22.72 HPLC analysis. glycochenodeoxycholate y = 77178x + 31.06 9.68 2.5. Impact of Different Kale Parts on Bile Acid Binding Capacity. Lyophilized kale stem, kale leaf, and whole kale were tested glycodeoxycholate y = 73395x + 25.22 4.84 − for in vitro bile acid binding capacity through oral digestion, gastric chenodeoxycholate y = 12607x 22.76 9.68 digestion, and intestinal digestion phases according to the above deoxycholate y = 11559x + 3.77 22.72 protocol, except that, during the intestinal digestion phase, the bile acid mixture containing taurocholate (4.84 mM), cholate (13.9 mM), Waldbronn, Germany). The separation was carried out at 65 °C with a and deoxycholate (6.16 mM) was used. After the in vitro digestion, the flow rate of 0.2 mL/min using gradient elution with an increasing supernatant was concentrated to achieve the minimal volume under strength of acetonitrile in 0.1% formic acid. Mass spectral analyses vacuum and passed through a cellulose filter for HPLC analysis. were performed using an ESI−Q-TOF mass spectrometer equipped 2.6. Optimization of Different Amounts of Kale for Bile Acid with an electrospray ionization (ESI) source in positive ion mode. The Binding. Kale was found to have the highest bile acid binding capacity capillary voltage was maintained at 2.9 kV; the source temperature was among the tested vegetables, and therefore, to find the minimal kale set at 65 °C; and nitrogen was used as the desolvation gas (12 L/min). amount needed to bind the maximum amount of bile acids, different 2.11. Fiber Analysis. Lyophilized green kale, red kale, red cabbage, weights of kale (0.5, 0.75, 1.0, 1.25, 1.5, 1.75, 2.0, and 2.5 g) were red leaf lettuce, and Brussels sprouts were analyzed for total dietary examined in the in vitro experiment. For this experiment, in the fiber (TDF), soluble fiber (SF), and insoluble fiber (ISF) according to intestinal digestion phase, the bile acid mixture BAC-3 was used. All AOAC protocols at Medallion Laboratories, Minneapolis, MN. other procedures were the same as those described before. 2.12. Statistical Analysis. All results were expressed as means ± 2.7. Minimal Processing of Green Kale for Bile Acid Binding. standard error (SE). The data were evaluated by JMP Pro 12, and To understand the impact of mild processing of kale on bile acid probability of p ≤ 0.05 was considered as statistically significant. All binding ability, two minimal processing techniques were used as experiments were conducted in triplicate with two independent follows: experiments. 2.7.1. Microwave Processing. Fresh-cut kale (18.3 g equivalent to 2.0 g of dried sample) was weighed into three 150 mL beakers, and 20 3. RESULTS AND DISCUSSION mL of water was added to the vegetables and microwaved for 1, 2, and − 3.1. Bile Acid Binding Capacities of Leafy Vegetables. 3 min separately and homogenized to a paste for 2 3 min. All samples Binding bile acids has multiple health benefits.21,23,25 Although were processed in triplicate and used for bioassays. 2.7.2. Steaming. Fresh-cut kale (18.3 g) was taken in a 150 mL certain vegetables have been tested for in vitro bile acid binding ° capacity, individual bile acids were not quantified sepa- beaker, and steaming was performed for 8 and 10 min at 95 C using a 21,28,29 water bath and homogenized to obtain a paste. Analysis of both of the rately. It would be desirable to include human fecal time points was conducted in triplicate and used to test the bile acid bacteria in the in vitro model; however, the intestinal binding capacity. microbiome diversity differs in different parts of the world, 2.8. Extraction of Bound Bile Acids. Digested residue was varies from individual to individual, and can be altered by − extracted with methanol/water/formic acid (80:19:1), followed by 2 diet.30 32 Thus, understanding bile acid binding using human min of homogenization and 2 h of sonication at 60 °C, centrifuged for intestinal bacteria will be challenging for large-scale screening fi fi 30 min, and ltered through Whatman lter paper. The residue was re- studies. The present in vitro study simulated the digestion from extracted with the same solvent, using 10 mL of solvent as mentioned mouth to small intestine and provides segmental results, which above, combined and concentrated by rotary evaporation at 40 °C under a reduced pressure to obtain 5−10 mL, and analyzed by HPLC can be a reference for further studies with human fecal bacteria 2.9. Quantification of Bile Acids by HPLC. The HPLC system involved in the digestion for the most-active samples. The consisted of an Agilent HPLC 1200 series (Foster City, CA) with a present study compared the bile acid binding capacities of degasser, quaternary pump, autosampler, column oven, and photo- different vegetables. Figure 2 shows that green and red kale diode array detector. Elution of bile acids was carried out with gradient exhibited the highest total bile acid binding capacities of 87 and mobile phases of (A) 0.3 M phosphoric acid in water and (B) 90%, respectively. The bile acid binding capacity of kale and acetonitrile with a flow rate of 0.8 mL/min at 30 °C using a Gemini Brussel sprouts was higher than the results previously C18 (Phenomenex, Torrance, CA) column. Bile acids were separated reported.21 This may be due to the low amount (0.1 g) of − using the gradient, from 75 to 45% A in 0 10 min, from 45 to 10% in samples used in the digestion process as well as the sensitivity 10−20 min, and from 10 to 75% in 20−25 min, and maintained an fi fl of the colorimetric assay used for the quanti cation of bile isocratic ow for 5 min. Data were processed using the (Agilent, Foster fi City, CA). All seven bile acids at different concentrations were injected acids. Among the ve tested bile acids, only CA showed the lowest binding capacity by kale. The rest of the vegetables, such to obtain the area. The calibration graphs were prepared by plotting fi the graph of area versus different concentrations (31.25−1000 ppm) of as red leaf lettuce (78%), exhibited signi cantly higher binding bile acids to obtain regression equations Table 1. Bound bile acid than (62%) and red cabbage (58%). DCA and samples were injected, and the presence of bile acids in each sample chenodeoxycholic acid are the more hydrophobic bile acids and was examined by comparing the retention times to those of the may induce colon cancer and liver stress;33,34 however, CA is standard bile acids. This assay was conducted in triplicate with three more hydrophilic, and its liver toxicity is milder than that of the independent experiments, and results were averaged. fi hydrophobic bile acids. The decreasing order of bile acid 2.10. Identi cation of Bile Acids by Liquid Chromatog- binding capacities of five vegetables was found to be red kale < raphy−Mass Spectrometry (LC−MS). Bile acids were identified by ultrahigh-performance liquid chromatography time-of-flight-mass green kale < red leaf lettuce < Brussels sprouts < red cabbage. spectrometry (UHPLC−Q-TOF−MS, Maxis Impact, Bruker Dal- Table S3 of the Supporting Information shows the moisture, tonics, Billerica, MA). Bound bile acid samples were separated on a fiber, and protein content present in different vegetables and Zorbax Eclipse Plus C18 rapid resolution column (1.8 μm partial size, their bile acid binding capacities. The difference in bile acid 100 × 2.1 mm) using an Agilent 1290 UHPLC instrument (Agilent, binding between various green vegetables may be due to their

8056 DOI: 10.1021/acs.jafc.7b02540 J. Agric. Food Chem. 2017, 65, 8054−8062 Journal of Agricultural and Food Chemistry Article

Figure 2. Bile acid binding capacities of red cabbage, green kale, red kale, Brussels sprouts, and red leaf lettuce. Results are expressed as the mean from three replicates ± SD. Different letters in each bile acid indicate significant differences (p ≤ 0.05) between the different vegetable bile acid binding capacities. Cholate (CA), glycochenodeoxycholate (GCDCA), glycodeoxycholate (GDCA), chenodeoxycholate (CDCA), and deoxycholate (DCA) were used in this study.

Figure 3. Green kale bile acid binding capacity in three different models of BAC. For instance, BAC-1, BAC-2, and BAC-3 equivalent to different health conditions, such as male with gallstone, healthy female, and type 2 male diabetic patients, respectively. Results were expressed as the mean from three replicates with SD. Different letters in each treatment indicate significant differences (p ≤ 0.05) between the different compositions of bile acids. different chemical compositions, including the levels of acid binding capacity based on the ISF was the highest in red phenolics, , flavonoids, dietary fiber, and other kale, followed by green kale, red leaf lettuce, Brussels sprouts, primary metabolites. Kale binds significantly higher chenodex- and red cabbage (Table S3 of the Supporting Information). oycholic acid and DCA than red cabbage, Brussels sprouts, and Each gram of ISF in these vegetables can bind 93.11−138.68 red leaf lettuce. Kale also showed the highest total bile acid mg of total bile acids (Table S4 of the Supporting Information). binding capacity, significantly higher than the other tested The results show that soluble fiber can bind significantly more vegetables. Notably, the bile acid binding capacity of green kale bile acids than ISF. Soluble fiber mainly includes pectin, gum, and red kale showed no significant difference from each other. and hemicellulose, whereas ISF includes hemicellulose and Green kale has 81 and 140% more flavonols and glucosinolates cellulose. Gum and pectin have good in vitro bile acid binding than red kale.35 Therefore, we used green kale for further capacity, but cellulose does not bind bile acids very well. The studies. protein contents of the tested vegetables were obtained from 3.2. Correlation between In Vitro Bile Acid Binding the United States Department of Agriculture (USDA) nutri- Capacity and Fiber. There is ample evidence that bile acids tional database, and the concentrations of protein per 100 g of can be bound by fiber and protein in foods.36,37 The total bile fresh weights are displayed in Table S3 of the Supporting

8057 DOI: 10.1021/acs.jafc.7b02540 J. Agric. Food Chem. 2017, 65, 8054−8062 Journal of Agricultural and Food Chemistry Article

Information. The protein content of kale is the highest, followed by Brussels sprouts, red cabbage, and red leaf lettuce. Proteins from various foods, such as soybean, milk, and fish, have demonstrated bile acid binding capacity.36,37 Soybean protein even has promising therapeutic applications against hypercholesterolemia, by binding bile acid and suppressing the absorption of cholesterol.37 Among the examined vegetables, kale has the highest portion of protein and exhibited maximum bile acid binding capacity. This observation may suggest a positive correlation between the protein content and in vitro bile acid binding capacity. 3.3. In Vitro Digestion of Green Kale with Different Bile Acid Compositions. The composition of the bile acids varies among individuals based on different health conditions and is intrinsically linked to their physiology. This occurs via a variety of regulatory processes, variation in toxicity among the − bile acids, and the microbial ecology of the gut.38 40 To ff Figure 4. Bile acid binding capacity of different parts of the kale. understand the positive e ects of kale, we used three bile acid ± ff compositions to understand different health conditions, Results were expressed as the mean of three replicates SD. Di erent letters in each bile acid indicate significant differences (p ≤ 0.05) including male with gallstone (BAC-1), healthy female (BAC- between stem, leaf, and whole kale. 2), and male with type 2 diabetes (BAC-3) (Table S2 of the Supporting Information). The results showed that BAC-1 and BAC-2 were bound in significantly higher amounts by kale than 3.5. Optimization of Different Amounts of Kale for BAC-3 (Figure 3). The binding capacity of DCA remained the Bile Acid Binding Capacity. To understand the optimum same in all three compositions. In the type 2 diabetic amount for green kale to achieve the maximum bile acid composition BAC-3, green kale bound significantly less binding capacity, we tested 0.5−2.5 g of samples. The portion GDCA, GCDCA, GCA, CA, and CDCA. According to Ho et of bound bile acids correlated with the amount of freeze-dried 41 26 al., the human bile pool size is 3055 mg, and Brufau et al. green kale. As shown in Figure 5, the samples had limited reported no significant difference of total bile pool size between binding capacities for each individual bile acid and total bile healthy individuals and type 2 diabetic patients. Thus, the acids. Also, each bile acid needs different amounts of kale for results of kale binding of bile acids in different compositions in maximum binding capacity and then became saturated even if the in vitro experiment are relevant to the amounts in humans. more green kale was added to the in vitro digestion. Table 2 The U.S. FDA has approved colesevelam hydrochloride as a shows the logistic curve explaining the relationship between the bile acid sequestrant for the treatment of type 2 diabetes. kale amount and bile acid binding capacity. The limits for each Clinical trials demonstrated that colesevelam significantly individual and total bile acid binding capacity ranged from 64 to improves the effect of three different drugs (metformin, 90%. To reach the 95% binding limitation of each bile acid, the sulfonylurea, and insulin) for type 2 diabetes. The addition of optimal amount of kale required was 1.56 g for GCA, 1.75 g for colesevelam involved in these treatments helps to reduce CA, 1.49 g for GCDCA, 1.29 g for GDCA, 1.24 g for CDCA, HbA1C from −0.32 to −0.41%.42 The mechanism by which 0.70 g for DCA, and 1.31 g for total bile acids (Table 2). To colesevelam lowers the blood glucose level is not yet compare to the results for kale bile acid binding capacity, established; however, it might be related to the farnesoid X cholestyramine resin (positive control) was tested and bound receptor (FXR) and G-protein-coupled receptor TGR5. The 93.45% total bile acids at 0.33 g (Figure 5). bile acid receptor FXR is involved in both lipid and glucose 3.6. Minimally Processed Green Kale for Bile Acid metabolism, and FXR can inhibit hepatic glucose production. Binding Capacity. Fresh kale (18.3 g, equivalent to 2.0 g of Meanwhile, bile acids can induce the G-protein-coupled lyophilized kale) was tested for the effects of minimal receptor TGR5, which can increase energy expenditure.43,44 It processing by microwaving or steaming. Figure 6 shows that seems that green kale has good potential to explore in an in vivo minimal processing significantly enhanced the fresh kale in vitro model to be used as a natural bile acid sequestrant for type 2 binding capacity for GCA, CA, GCDCA, GDCA, and total bile diabetic patients to reduce blood levels. acids compared to lyophilized kale. However, both microwaving 3.4. In Vitro Bile Acid Binding of Different Parts of and steaming methods did not show significant differences for Kale. Most of the time, kale leaf or whole kale is consumed. To most of the bile acids, except CDCA and for GCDCA and understand the potential health benefits of kale, lyophilized kale DCA in the 2 min microwaved samples. In general, minimally leaf, stem, and whole (total) kale were tested for in vitro bile processed samples bound 90% total bile acids. Microwaved kale acid binding. As seen in Figure 4, kale stem binds less (2 min) bound 93% of DCA, while kale microwaved for 1 and 3 taurocholate than kale leaf and whole kale, while it bound min bound 97% DCA. Steamed kale bound amounts of total similar amounts of cholate, deoxycholate, and total bile acids bile acids similar to microwaved samples. Kale steamed for 8 compared to leaf and whole kale. Additionally, whole kale and 10 min bound significantly less CDCA than 1 and 3 min bound more bile acids than kale stem and kale leaf. The kale microwaved kale. In other words, both steaming and micro- stem bound the lowest amount of total bile acids (63%), waving can significantly improve the bile acid binding capacity including 86.3% deoxycholate, 56.7% cholate, and 54.8% of kale and microwave processing for 1 or 3 min can give the taurocholate, whereas whole kale bound the maximum best in vitro bile acid binding capacity. (67.4%) amount of total bile acids, including 93.8% Moutafis et al.41 measured the fecal excretion of bile acids deoxycholate, 59.4% cholate, and 59.4% taurocholate. (1897.76 mg) in response to an intake of 30 g of

8058 DOI: 10.1021/acs.jafc.7b02540 J. Agric. Food Chem. 2017, 65, 8054−8062 Journal of Agricultural and Food Chemistry Article

Figure 5. Relationship between the amount of kale and cholestyramine resin (positive control) on individual bile acids and total bile acid binding capacity. Different amounts of whole kale were tested for bile acid binding capacity in triplicates.

Table 2. Logistic Curve for Individual Bile Acid Binding Capacity in Response to the Kale Amount

saturation point bile salt equation x valuea y valueb glycocholate y = 67.472958/(1 + exp(−2.2221099(x − 0.2364677))) 1.561 64.1 cholate y = 70.21116/(1 + exp(−2.7243291(x − 0.6714658))) 1.752 66.7 glycochenodeoxycholate y = 81.507683/(1 + exp(−3.0382878(x − 0.5179148))) 1.487 77.4 glycodeoxycholate y = 81.055733 /(1 + exp(−4.011064(x − 0.5510876))) 1.285 77.0 chenodeoxycholate y = 87.215157/(1 + exp(−3.6301377(x − 0.4235821))) 1.235 82.9 deoxycholate y = 94.784765/(1 + exp(−7.5850513(x − 0.3074524))) 0.696 90.0 total bile acids y = 81.883651/(1 + exp(−3.1635124(x − 0.3817649))) 1.312 77.8 ax refers to the amount of green kale (g). by refers to the bile acid binding percentage. cholestyramine. The logistic equation obtained in the present including the microbiome in the large intestine, and, thus, is study for the total bile acid binding is y = 98.36458/(1 + expected to have a difference from the present in vitro exp(−11.570799(x + 0.077039))), where y refers to the experiment. To compute the difference between our experi- percentage of bound bile acids and x refers to the amount ment and the clinical trial, the results of the positive control (g) of cholestyramine (Table 2). The total amount of bile acids (cholestyramine) can be used. For instance, if human intestinal used in the present study was 72.06 mg, while the bile pool of bacteria have no effect on bile acid composition or enter- the human body is 3055 mg, which is 42.4 times higher in this ohepatic circulation, then the in vitro bile acid binding capacity experiment. The present in vitro experiment involves the can be extrapolated to the human body. Therefore, 30 g of digestion from the mouth to the small intestine; the previous cholestyramine taken by the individual in the clinical trial is clinical trial involved the complete human digestive system, equivalent to 0.7075 g (30/42.4 = 0.7075) of cholestyramine in

8059 DOI: 10.1021/acs.jafc.7b02540 J. Agric. Food Chem. 2017, 65, 8054−8062 Journal of Agricultural and Food Chemistry Article

Figure 6. Bile acid binding capacity of minimally processed fresh kale compared to lyophilized kale. Each value is the mean from three replicates with SD. Different letters in each bile acid indicate significant differences at p ≤ 0.05. the in vitro experiment, and 98.35% total bile acids can be hiranonis, Clostridium hylemonae, and Clostridium sordelli can bound when 0.7075 g of cholestyramine is used, according to convert CA to DCA and CDCA to LCA.45,47 the equation. If this in vitro experiment can be extrapolated to Secondary bile acids cannot be absorbed by the liver through the human body, about 3004.5 mg of bile acids should be the portal vein, and therefore, they accumulate in the bile excreted into the human feces. However, Brufau et al.26 pool.45 The DCA content in the bile acid pool can be up to 45,47 reported that 1897.76 mg of bile acids was excreted with 30 g of 75% in some individuals favoring the Western diet. cholestyramine daily intake; hence, the computed difference Moreover, DCA can induce oxidative stress in colon cells, between this in vitro experiment and the clinical trial is 0.6316, can increase the risk of gene mutation, and eventually may 3 (1897.76/3004.53 = 0.6316), which can be used as an index promote the development of tumors or cancer. Berstein et al. reported that DCA induced 94% of colon tumors in mice used factor to link the current in vitro experiment to the human 5 body. The index factor can generally represent the impact of in the experiment, including 56% of colon cancers. In the the human intestinal microbiome on bile acid transformation present study, kale preferentially bound DCA, indicating that and enterohepatic circulation. Without considering the impact kale consumption may have the potential to reduce colon of human fecal bacteria on bile acids, this in vitro experiment cancer risk, which should be explored in future studies. 3.7. Confirmation of Bound Bile Acids. The bound bile determined that no more than 81.81% total bile acids in fi − humans36 can be bound; adjustment by the index factor of acids in the digested residue were con rmed by LC MS using 0.6316 yields the result that 1578.55 mg of bile acids can be accurate experimental mass with the theoretical value of individual bile acids. The digested residue of each vegetable excreted by consuming green kale. With extrapolation to the was extracted with MeOH/water/formic acid (80:19:1) and human body to achieve 95% of the limit of bile acid excretion, a analyzed by high-resolution mass spectrometry. Figure 7A person must consume 55.63 g of dry green kale or shows the base peak chromatogram of five bile acids extracted approximately 425.96 g of fresh green kale. A previous study fi from kale samples. Figure 7B shows the accurate mass spectra reported that additional bile acid sequestrant signi cantly of each bile acid obtained by electrospray negative ionization improved the reduction of HbA1c and low-density lipoprotein 42 mode. All of the molecular weights matched those of their (LDL) cholesterol by insulin-based therapy. Therefore, type 2 fi fi respective molecular ions, which con rm the presence of bile diabetic patients may bene t from the daily consumption of acids without degradation during the digestion process and also kale. authenticate the quantified data as highly accurate. The optimal bile acid binding amounts for kale from Table 2 In conclusion, red kale and green kale have higher in vitro bile can be used as a reference for future in vivo experiments and acid binding capacity than red cabbage, red leaf lettuce, and clinical trials for bile acid binding. When additional factors Brussels sprouts. Green kale equally bound bile acids in the involved in the digestive system, such as intestinal bacteria, the composition equivalent to healthy female and male with bile acid composition may be altered as a result of the gallstone and bound slightly lower amounts of bile acids in type biotransformation effect from microbiomes. Certain bacteria 2 diabetic male. In the bile acid mixture equivalent to type 2 45 could also increase the hydrophobicity of bile salts. For diabetic patients, green kale bound no more than 81.9% of the instance, Lactobacillus acidophilus in the human intestinal tract bile acids, which was accomplished by the amount of 1.3 g of can deconjugate glycocholate and taurocholate, which increases kale. Green kale preferentially bound the hydrophobic bile acids the CA content in the digestive tract.46 Human intestinal DCA and chenodeoxycholic acid. In addition, steaming bacteria can also transform primary bile acids into secondary significantly enhanced the bile acid binding capacity of kale. bile acids. For instance, Clostridium scindens, Clostridium The recommended green kale amount demonstrated the ideal

8060 DOI: 10.1021/acs.jafc.7b02540 J. Agric. Food Chem. 2017, 65, 8054−8062 Journal of Agricultural and Food Chemistry Article

conditions used in the present study (Table S2), levels of moisture, fiber, and protein in fresh vegetables and samples used for bile acid binding capacity (Table S3), and bile acid binding capacity based on the dietary fiber content (Table S4) (PDF)

■ AUTHOR INFORMATION Corresponding Authors *Telephone: +1-979-845-3864. Fax: +1-979-862-4522. E-mail: [email protected] and/or [email protected]. *Telephone: +1-979-862-4521. Fax: +1-979-862-4522. E-mail: [email protected]. ORCID Guddadarangavvanahally K. Jayaprakasha: 0000-0003-1749- 9699 Bhimanagouda S. Patil: 0000-0001-7189-0432 Funding This study was supported by Texas State Funding 2013-121277 VFIC-TX state appropriation. Notes The authors declare no competing financial interest. ■ REFERENCES (1) Chiang, J. Y. L. Bile acids: Regulation of synthesis. J. Lipid Res. 2009, 50, 1955−1966. (2) Kuipers, F.; Bloks, V. W.; Groen, A. K. Beyond intestinal soap Bile acids in metabolic control. Nat. Rev. Endocrinol. 2014, 10, 488− 498. (3) Ajouz, H.; Mukherji, D.; Shamseddine, A. Secondary bile acids: An underrecognized cause of colon cancer. World J. Surg. Oncol. 2014, 12, 164−164. (4) Pakarinen, M.; Miettinen, T. A.; Lauronen, J.; Kuusanmaki, P.; Raivio, P.; Kivisto, T.; Halttunen, J. Adaptation of cholesterol absorption after proximal resection of porcine small intestine. J. Lipid Res. 1996, 37, 1766−1775. (5) Bernstein, C.; Holubec, H.; Bhattacharyya, A. K.; Nguyen, H.; Payne,C.M.;Zaitlin,B.;Bernstein,H.Carcinogenicityof deoxycholate, a secondary bile acid. Arch. Toxicol. 2011, 85, 863−871. (6) Jenkins, D. J.; Wolever, T. M.; Rao, A. V.; Hegele, R. A.; Mitchell, S. J.; Ransom, T. P.; Boctor, D. L.; Spadafora, P. J.; Jenkins, A. L.; Mehling, C.; Relle, L. K.; Connelly, P. W.; Story, J. S.; Furumoto, E. J.; Corey, P.; Wursch, P. Effect on blood lipids of very high intakes of fiber in diets low in saturated and cholesterol. N. Engl. J. Med. 1993, 329,21−26. (7) Wakai, K.; Hirose, K.; Matsuo, K.; Ito, H.; Kuriki, K.; Suzuki, T.; Kato, T.; Hirai, T.; Kanemitsu, Y.; Tajima, K. Dietary Risk: Factors for Colon and Rectal Cancers A Comparative Case-Control Study. J. Epidemiol. 2006, 16, 125−135. Figure 7. (A) Confirmation of bound bile acids in kale samples by (8) Newmark, H. L.; Yang, K.; Kurihara, N.; Fan, K.; Augenlicht, L. LC−MS. (B) Base peak of bound bile acids separated on eclipse plus H.; Lipkin, M. Western-style diet-induced colonic tumors and their − C18 column. Peaks 1 5 indicate individual bile acids. High-resolution modulation by calcium and vitamin D in C57Bl/6 mice: A preclinical quadrupole time-of-flight (HR-Q-TOF) mass spectra of each bound model for human sporadic colon cancer. Carcinogenesis 2009, 30,88− bile acids were obtained by electrospray negative ionization. 92. (9) Newmark, H. L.; Wargovich, M. J.; Bruce, W. R. Colon cancer and dietary fat, phosphate, and calcium: A hypothesis. J. Natl. Cancer bile acid binding in the in vitro experiment; this amount can be Inst. 1984, 72, 1323−1325. used as a reference for future in vivo experiments. (10) Fedirko, V.; Bostick, R. M.; Flanders, W. D.; Long, Q.; Sidelnikov, E.; Shaukat, A.; Daniel, C. R.; Rutherford, R. E.; Woodard, ■ ASSOCIATED CONTENT J. J. Effects of vitamin D and calcium on proliferation and *S Supporting Information differentiation in normal colon mucosa: A randomized clinical trial. − The Supporting Information is available free of charge on the Cancer Epidemiol., Biomarkers Prev. 2009, 18, 2933 2941. ACS Publications website at DOI: 10.1021/acs.jafc.7b02540. (11) Hofmann, A. F. The continuing importance of bile acids in liver and intestinal disease. Arch. Intern. Med. 1999, 159, 2647−2658. Chemical composition of simulated saliva fluid prepared (12) Attili, A. F.; Angelico, M.; Cantafora, A.; Alvaro, D.; Capocaccia, for the bile acid binding assay (Table S1), different bile L. Bile acid-induced liver toxicity: Relation to the hydrophobic− acid compositions equivalent to different human health hydrophilic balance of bile acids. Med. Hypotheses 1986, 19,57−69.

8061 DOI: 10.1021/acs.jafc.7b02540 J. Agric. Food Chem. 2017, 65, 8054−8062 Journal of Agricultural and Food Chemistry Article

(13) Insull, W., Jr. Clinical utility of bile acid sequestrants in the (32) Blum, H. E. The human microbiome. Adv. Med. Sci. 2017, 62, treatment of dyslipidemia: A scientific review. South. Med. J. 2006, 99, 414−420. 257−273. (33) Perez, M. J.; Briz, O. Bile-acid-induced cell injury and (14) Riikonen, S.; Savonius, H.; Gylling, H.; Nikkila, K.; Tuomi, A. protection. World J. Gastroenterol. 2009, 15, 1677−1689. M.; Miettinen, T. A. Oral guar gum, a gel-forming relieves (34) Degirolamo, C.; Modica, S.; Palasciano, G.; Moschetta, A. Bile pruritus in intrahepatic cholestasis of pregnancy. Acta Obstet. Gynecol. acids and colon cancer: Solving the puzzle with nuclear receptors. Scand. 2000, 79, 260−264. Trends Mol. Med. 2011, 17, 564−572. (15) Butt, M. S.; Shahzadi, N.; Sharif, M. K.; Nasir, M. Guar gum: A (35) Olsen, H.; Grimmer, S.; Aaby, K.; Saha, S.; Borge, G. I. miracle therapy for hypercholesterolemia, hyperglycemia and obesity. Antiproliferative effects of fresh and thermal processed green and red Crit. Rev. Food Sci. Nutr. 2007, 47, 389−396. of curly kale (Brassica oleracea L. convar. acephala var. − (16) Sonnett, T. E.; Levien, T. L.; Neumiller, J. J.; Gates, B. J.; Setter, sabellica). J. Agric. Food Chem. 2012, 60, 7375 7383. S. M. Colesevelam hydrochloride for the treatment of type 2 diabetes (36) Perez-Galvez, R.; Garcia-Moreno, P. J.; Morales-Medina, R.; mellitus. Clin. Ther. 2009, 31, 245−259. Guadix, A.; Guadix, E. M. Bile acid binding capacity of fish protein (17) Hansen, M.; Sonne, D. P.; Knop, F. K. Bile Acid Sequestrants: hydrolysates from discard species of the West Mediterranean Sea. Food − Glucose-Lowering Mechanisms and Efficacy in Type 2 Diabetes. Curr. Funct. 2015, 6, 1261 1267. Diabetes Rep. 2014, 14, 482. (37) Nagaoka, S.; Nakamura, A.; Shibata, H.; Kanamaru, Y. Soystatin (18)Garg,A.;Grundy,S.M.Cholestyraminetherapyfor (VAWWMY), a novel bile acid-binding peptide, decreased micellar solubility and inhibited cholesterol absorption in rats. Biosci., dyslipidemia in non-insulin-dependent diabetes mellitus. A short- − term, double-blind, crossover trial. Ann. Intern. Med. 1994, 121, 416− Biotechnol., Biochem. 2010, 74, 1738 1741. (38) Hirota, M.; Shimosegawa, T. Bile Acids and and Pancreatic 422. Disease. In Bile Acids in Gastroenterology: Basic and Clinical; Tazuma, (19) Bays, H. E.; Goldberg, R. B.; Truitt, K. E.; Jones, M. R. S., Takikawa, H., Eds.; Springer: Tokyo, Japan, 2017; pp 169−176, Colesevelam hydrochloride therapy in patients with type 2 diabetes DOI: 10.1007/978-4-431-56062-3_12. mellitus treated with metformin: Glucose and lipid effects. Arch. Intern. (39) Mudaliar, S.; Henry, R. R.; Sanyal, A. J.; Morrow, L.; Marschall, 2008 − Med. , 168, 1975 1983. H. U.; Kipnes, M.; Adorini, L.; Sciacca, C. I.; Clopton, P.; Castelloe, E.; (20) Fonseca, V. A.; Rosenstock, J.; Wang, A. C.; Truitt, K. E.; Jones, Dillon, P.; Pruzanski, M.; Shapiro, D. Efficacy and safety of the M. R. Colesevelam HCl improves glycemic control and reduces LDL farnesoid X receptor agonist obeticholic acid in patients with type 2 cholesterol in patients with inadequately controlled type 2 diabetes on − diabetes and nonalcoholic fatty liver disease. Gastroenterology 2013, sulfonylurea-based therapy. Diabetes Care 2008, 31, 1479 1484. 145, 574−582.e1. (21) Kahlon, T. S.; Chapman, M. H.; Smith, G. E. In vitro binding of (40) Carey, M. C.; Small, D. M. The physical chemistry of cholesterol bile acids by spinach, kale, brussels sprouts, broccoli, mustard greens, solubility in bile. Relationship to gallstone formation and dissolution in − green bell pepper, cabbage and collards. Food Chem. 2007, 100, 1531 man. J. Clin. Invest. 1978, 61, 998. 1536. (41) Ho, K.-J.; Ho, L.-h. C.; Hsu, S.-C.; Chen, J.-S. Bile acid pool size ́ (22) Espinal-Ruiz, M.; Parada-Alfonso, F.; Restrepo-Sanchez, L.-P.; in relation to functional status of gallbladder and biliary lipid ́ Narvaez-Cuenca, C.-E.; McClements, D. J. Impact of dietary fibers composition in Chinese. Am. J. Clin. Nutr. 1980, 33, 1026−1032. [methyl cellulose, chitosan, and pectin] on digestion of lipids under (42) Fonseca, V. A.; Handelsman, Y.; Staels, B. Colesevelam lowers simulated gastrointestinal conditions. Food Funct. 2014, 5, 3083−3095. glucose and lipid levels in type 2 diabetes: The clinical evidence. (23) Hur, S. J.; Lim, B. O.; Decker, E. A.; McClements, D. J. In vitro Diabetes, Obes. Metab. 2010, 12, 384−392. human digestion models for food applications. Food Chem. 2011, 125, (43) Thomas, C.; Gioiello, A.; Noriega, L.; Strehle, A.; Oury, J.; 1−12. Rizzo, G.; Macchiarulo, A.; Yamamoto, H.; Mataki, C.; Pruzanski, M.; (24) Ulleberg, E. K.; Comi, I.; Holm, H.; Herud, E. B.; Jacobsen, M.; Pellicciari, R.; Auwerx, J.; Schoonjans, K. TGR5-mediated bile acid Vegarud, G. E. Human Gastrointestinal Juices Intended for Use in In sensing controls glucose homeostasis. Cell Metab. 2009, 10, 167−177. Vitro Digestion Models. Food Dig. 2011, 2,52−61. (44) Katsuma, S.; Hirasawa, A.; Tsujimoto, G. Bile acids promote (25) Zacherl, C.; Eisner, P.; Engel, K.-H. In vitro model to correlate glucagon-like peptide-1 secretion through TGR5 in a murine viscosity and bile acid-binding capacity of digested water-soluble and enteroendocrine cell line STC-1. Biochem. Biophys. Res. Commun. insoluble dietary fibres. Food Chem. 2011, 126, 423−428. 2005, 329, 386−390. (26) Brufau, G.; Stellaard, F.; Prado, K.; Bloks, V. W.; Jonkers, E.; (45)Ridlon,J.M.;Kang,D.J.;Hylemon,P.B.Bilesalt Boverhof, R.; Kuipers, F.; Murphy, E. J. Improved glycemic control biotransformations by human intestinal bacteria. J. Lipid Res. 2006, with colesevelam treatment in patients with type 2 diabetes is not 47, 241−259. directly associated with changes in bile acid metabolism. Hepatology (46) Gilliland, S. E.; Speck, M. L. Deconjugation of bile acids by (Hoboken, NJ, U. S.) 2010, 52, 1455−1464. intestinal lactobacilli. Appl. Environ. Microbiol. 1977, 33,15−18. (27) Fisher, M. M.; Yousef, I. M. Sex differences in the bile acid (47) Ridlon, J. M.; Kang, D. J.; Hylemon, P. B.; Bajaj, J. S. Bile Acids composition of human bile: Studies in patients with and without and the Gut Microbiome. Curr. Opin. Gastroenterol. 2014, 30, 332− gallstones. Can. Med. Assoc. J. 1973, 109, 191−193. 338. (28) Ngoh, Y.-Y.; Choi, S. B.; Gan, C.-Y. The potential roles of Pinto bean (Phaseolus vulgaris cv. Pinto) bioactive peptides in regulating physiological functions: Protease activating, lipase inhibiting and bile acid binding activities. J. Funct. Foods 2017, 33,67−75. (29) Kahlon, T. S.; Chiu, M. C.; Chapman, M. H. Steam cooking significantly improves in vitro bile acid binding of collard greens, kale, mustard greens, broccoli, green bell pepper, and cabbage. Nutr. Res. (N. Y., NY, U. S.) 2008, 28, 351−357. (30) Othman, M.; Agüero, R.; Lin, H. C. Alterations in intestinal microbial flora and human disease. Curr. Opin. Gastroenterol. 2008, 24, 11−16. (31) Antonio, M. A.; Hawes, S. E.; Hillier, S. L. The identification of vaginal Lactobacillus species and the demographic and microbiologic characteristics of women colonized by these species. J. Infect. Dis. 1999, 180, 1950−1956.

8062 DOI: 10.1021/acs.jafc.7b02540 J. Agric. Food Chem. 2017, 65, 8054−8062