Carotenoids from gac fruit aril (Momordica cochinchinensis [Lour.] Spreng.) are more bioaccessible than those from carrot root and tomato fruit

FRIN-06486; No of Pages 8 Food Research International xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Food Research International journal homepage: www.elsevier.com/locate/foodres

Carotenoids from gac fruit aril (Momordica cochinchinensis [Lour.] Spreng.) are more bioaccessible than those from carrot root and tomato fruit Judith Müller-Maatsch a, Jasmin Sprenger a, Judith Hempel a, Florence Kreiser a, Reinhold Carle a,b, Ralf M. Schweiggert a,⁎ a b

University of Hohenheim, Institute of Food Science and Biotechnology, Chair of Plant Foodstuff Technology and Analysis, Garbenstrasse 25, 70599 Stuttgart, Germany King Abdulaziz University, Faculty of Science, Biological Science Department, P. O. Box 80257, Jeddah 21589, Saudi Arabia

a r t i c l e

i n f o

Article history: Received 30 September 2016 Received in revised form 28 October 2016 Accepted 31 October 2016 Available online xxxx Keywords: Lycopene β-Carotene In vitro digestion Vitamin A Vitamin A deficiency Liberation Chromoplasts Aggregation

a b s t r a c t Using a simulated digestion procedure in vitro, liberation and bioaccessibility of β-carotene (29.5 ± 1.7% and 22.6 ± 0.9%, respectively) and lycopene (51.3 ± 2.6% and 33.2 ± 3.1%, respectively) from gac fruit aril were found to be significantly higher than from carrot root (β-carotene, 5.2 ± 0.5% and 0.5 ± 0.2%, respectively) and tomato fruit (lycopene, 15.9 ± 2.8% and 1.8 ± 0.5%, respectively). Gac fruit aril naturally contained significantly more lipids (11% on fresh weight base) than carrot root and tomato fruit (b 1%). However, when test meals were supplemented with an O/W emulsion to match the content of gac fruit aril, carotenoid bioaccessibility was still considerably lower than that from genuine gac fruit aril. Carotenoids in gac fruit aril were found to be stored in small, round-shaped chromoplasts. Despite the high lipid content, these carotenoids are unlikely to occur in a lipid-dissolved state according to simple solubility estimations, instead being possibly deposited as submicroscopic crystallites. In contrast, carotenoids of carrot root and tomato fruit were stored in large, needle-like crystallous chromoplasts. Consequently, we hypothesized the natural deposition form to be majorly responsible for the observed differences in bioaccessibility. A favorable surface-to-volume ratio of the deposition form in gac fruit aril might have allowed a more rapid micellization during digestion, and thus, an enhanced bioaccessibility. Irrespective of the ultimate reason, gac fruit aril provided a highly bioaccessible form of both lycopene and provitamin A (β-carotene), thus offering a most valuable dietary source of both carotenoids. Currently, gac is majorly grown in Southeast Asia, where its consumption might help to diminish the ‘hidden hunger’ namely the insufficient supply with vitamin A. Ultimately, gac fruit might thus contribute to alleviating most severe health implications of vitamin A deficiency, such as anaemia and xerophthalmia, the prevailing cause of preventable childhood blindness, as well as mortality from infectious diseases. © 2016 Elsevier Ltd. All rights reserved.

1. Introduction Carotenoids, constituting essential functional components of the photosynthetic apparatus, are ubiquitously synthesized in plants. Furthermore, they impart yellow, orange, and red colors to numerous fruits, flowers, leaves, seeds, and roots of plants. Although humans and animals are incapable of carotenoid biosynthesis (Britton & Khachik, 2009), they play important bio-functional roles such as their contribution to embryonic development and reproduction, immune modulation, ⁎ Corresponding author at: University of Hohenheim, Institute of Food Science and Biotechnology, Chair of Plant Foodstuff Technology and Analysis, Garbenstrasse 25, D70599 Stuttgart, Germany. E-mail addresses: [email protected] (J. Müller-Maatsch), [email protected] (J. Sprenger), [email protected] (J. Hempel), fl[email protected] (F. Kreiser), [email protected] (R. Carle), [email protected] (R.M. Schweiggert).

and the maintenance of the visual system. Furthermore, carotenoids exert anti-inflammatory and anti-oxidant effects (Saini, Nile, & Park, 2015). β-Carotene (Fig. 1) is the most abundant carotenoid in many plant foods, representing a C40 tetraterpenoid with two unmodified βionone rings, which may be symmetrically cleaved to form vitamin A by humans (Tang & Russel, 2009). This vitamin is essential for the mammalian visual system as well as for sound growth and development (Saini et al., 2015). Although not being a provitamin A carotenoid, the acyclic tetraterpenoid lycopene (Fig. 1) represents one of the most potent natural antioxidants, and its frequent dietary intake has been associated with a lower risk of cardiovascular diseases and prostate cancer (Müller, Caris-Veyrat, Lowe, & Böhm, 2016; Rock, 2009). In a Western diet, β-carotene is mainly ingested from apricot, carrot, mango, sweet potato, and green leafy vegetables like kale and spinach, whereas lycopene is mostly ingested via tomatoes and watermelons. Although the mentioned plant foods contain high levels of both β-carotene and

http://dx.doi.org/10.1016/j.foodres.2016.10.053 0963-9969/© 2016 Elsevier Ltd. All rights reserved.

Please cite this article as: Müller-Maatsch, J., et al., Carotenoids from gac fruit aril (Momordica cochinchinensis [Lour.] Spreng.) are more bioaccessible than those from carrot root and tomato fruit..., Food Research International (2016), http://dx.doi.org/10.1016/j.foodres.2016.10.053

2

J. Müller-Maatsch et al. / Food Research International xxx (2016) xxx–xxx

Fig. 1. Chemical structures of β-carotene and lycopene.

lycopene, their bioavailability is often poor, particularly when consumed raw and unprocessed. Most rich sources, e.g. green leafy vegetables and fruits need to be processed, i.e. mechanically and thermally treated and co-consumed with lipids, to enhance their liberation, bioaccessibility, and bioavailability during human digestion (Lemmens et al., 2014). In particular, in South-East-Asian diets, the seed membrane (aril) of gac fruit (Fig. 2), belonging to the Cucurbitaceae family, represents an extremely rich source of both β-carotene and lycopene, at levels that easily surpass those of other main dietary sources. For instance, recent publications reported highly variable total carotenoid contents of 48.1 mg/100 g fresh weight (FW) (Aoki, Kieu, Kuze, Tomisaka, & Chuyen, 2002), 49.7 mg/100 g FW (Vuong, Franke, Custer, & Murphy, 2006), 89.2 mg/100 g FW (West & Poortvliet, 1993), 97.7 mg/100 g FW (Vuong, Dueker, & Murphy, 2002), 287.4 mg/100 g FW (Ishida, Turner, Chapman, & McKeon, 2004), and 410.7 mg/100 g FW (Nhung, Bung, Ha, & Phong, 2010). The enormous range of the carotenoid contents so far reported was suspected to be caused by genetic and environmental factors, by post-harvest degradation during transportation and storage, and also by differences in the applied analytical extraction methods (Kubola & Siriamornpun, 2011). Traditionally, gac fruit aril is used as a coloring component in rice dishes, served at important

celebrations in Vietnam (Rodriguez-Amaya, 2016). The repeated consumption of gac-enriched rice dishes, containing 3.5–5.0 mg β-carotene per dish, has been shown to induce significantly greater increases of βcarotene and retinol concentrations in human plasma than the consumption of rice dishes supplemented with an equivalent amount of synthetic β-carotene (Vuong, 2000; Vuong et al., 2002). Furthermore, levels of lycopene were found to be increased in human plasma after the 30 day-long consumption of rice dishes with added gac fruit aril (Vuong et al., 2002). Providing deep insights into the liberation, micellization, and uptake of gac fruit carotenoids into Caco-2 cells, Failla, Chitchumroonchokchai, and Ishida (2008) reported that 30–45% of βcarotene and 2–28% of lycopene had been micellized, and thus, had become “bioaccessible” during digestion of a rice dish enriched with aril or oil. Although these studies indicated that gac fruit aril provides a readily bioaccessible and bioavailable form of β-carotene and lycopene, quantitative studies to compare bioaccessibility and bioavailability of both carotenoids from gac fruit aril to that of other β-carotene- and lycopenerich fruits and vegetables are lacking. Therefore, our study aimed at comparing the liberation and bioaccessibility of carotenoids from the different food matrices namely from gac fruit aril, tomato fruit, and carrot root. Due to its potential importance for explaining differences in carotenoid bioaccessibility and bioavailability, we further investigated the yet unknown genuine deposition form of gac fruit aril carotenoids, and compared it to that found in carrot root and tomato fruit. The natural deposition form in plant foods was earlier shown to be highly different, i.e. being either solid-crystalline, liquid-crystalline, protein-complexed, or lipid-dissolved. A detailed overview on different carotenoid deposition forms in plants has been currently compiled by Schweiggert and Carle (2015). 2. Materials and methods 2.1. Reagents, emulsifiers and enzymes Methanol, light petroleum (b. p. 40–60 °C), methyl tert-butyl ether (MTBE), ammonium acetate, sodium chloride, magnesium chloride hexahydrate, D-(+)-maltose, and di-sodium hydrogen phosphate dihydrate were obtained from VWR International (Leuven, Belgium). Ethyl acetate, n-hexane, 3-tert-butyl-4-hydroxy-anisole (BHA), 3,5-di-tert-butyl-4-hydroxy-toluene (BHT), potassium chloride, potassium dihydrogen phosphate, sodium hydrogen carbonate, hydrochloric acid (37%), sodium chloride, sodium potassium tartrate, trichloroacetic acid, tris-(hydroxymethyl)-aminomethane, and soluble potato starch were from Merck (Darmstadt, Germany). α-Amylase from porcine pancreas (EC 3.2.1.1), pepsin from porcine gastric mucosa (EC 3.4.23.1), cholesterol esterase from porcine pancreas (EC 3.1.1.13), pancreatin from porcine pancreas (EC 232–468-9), lipase from Candida rugosa (EC 3.1.1.3), porcine bile extract, ammonium carbonate, bovine blood hemoglobin, tributyrin, p-toluene-sulfonyl-L-arginine methyl ester, and 3,5-dinitrosalicylic acid were purchased from Sigma Aldrich (Taufkirchen, Germany). Calcium chloride dihydrate was from Fluka (Buchs, Switzerland). All the above-mentioned chemicals were of analytic purity. Food grade soy lecithin (Ultralec P) was supplied by Archer Daniels Midland Company (Chicago, IL, USA) and Miglyol 812N (medium chained triglycerides) was purchased from Cremer Olio (Hamburg, Germany). (all-E)-Lycopene (purity 97.5%) was prepared from tomato paste as described by Kopec et al. (2010), while (all-E)-β-carotene (purity N 99%) was purchased from Sigma Aldrich (Taufkirchen, Germany). Ultrapure water was used throughout the experiments. 2.2. Plant materials

Fig. 2. Morphology of a freshly-cut, halved gac fruit.

Gac fruit (Momordica cochinchinensis Lour. Spreng., Fig. 1), tomato fruit (Solanum lycopersicum L.), and carrot root (Daucus carota L. ssp. sativus) were obtained from a local market in Stuttgart (Germany). Peel, mesocarp, placenta, aril, and seeds from gac fruit were manually

Please cite this article as: Müller-Maatsch, J., et al., Carotenoids from gac fruit aril (Momordica cochinchinensis [Lour.] Spreng.) are more bioaccessible than those from carrot root and tomato fruit..., Food Research International (2016), http://dx.doi.org/10.1016/j.foodres.2016.10.053

J. Müller-Maatsch et al. / Food Research International xxx (2016) xxx–xxx

separated. After removing the stem ends and green parts, carrot roots were washed and peeled with a knife. Similarly, tomato fruit were washed and green parts were removed. Subsequently, the edible plant parts were cut into small cubes (ca. 1 × 1 × 1 cm3) and homogenized with a kitchen blender (Retsch Grindomix GM 200, Retsch, Haan, Germany) at 10,000 rpm for 30 s. The homogenized samples were vacuum-packed into aluminum pouches and stored at − 80 °C until starting the simulated digestion. In addition, aliquots of the same material were freeze-dried, prior to the determination of dry matter, lipid, and carotenoid content. Furthermore, fresh aliquots of all plant materials were used for light microscopic inspection.

3

methanol/MTBE/water (8:90:2, v/v/v with 0.04% w/v ammonium acetate) as eluent B. The solvent gradient was as follows: 15% B isocratic (6 min), 15% to 25% B (1 min), 25% B isocratic (3 min), 25% to 55% B (17 min), 55% B isocratic (7 min), 55% to 100% B (15 min), 100% B isocratic (1 min), 100% B to 15% B (1 min), and 15% B isocratic (2 min). Total run time was 53 min at a flow rate of 1.0 mL/min. Carotenoids were monitored at 446 nm. Additional UV/Vis spectra were recorded in the range of 200–600 nm. Compounds were identified by comparing retention times and UV/Vis spectra to those of authentic standards. For quantitation of the carotenoids lycopene, α-carotene, and β-carotene, linear calibrations of (all-E)-lycopene and (all-E)-β-carotene (α-carotene and β-carotene) were used. Quantification was carried out at 446 nm for all carotenoids.

2.3. Light microscopy 2.7. Preparation of O/W emulsion and particle size distribution analyses Bright field light microscopy and differential interference contrast (DIC) were carried out with an Axioplan microscope (Zeiss, Oberkochen, Germany) coupled to a digital camera (Canon Powershot A640, Canon, Krefeld, Germany). Freehand razorblade sections and squash specimens of fresh gac fruit aril, carrot root, and tomato fruit were used to investigate form, size, and color of the chromoplasts. 2.4. Dry matter and lipid content Dry matter was determined gravimetrically by freeze-drying (VaCo 10-II-E, Zirbus technology, Bad Grund, Germany). Lipid content was analyzed using a Soxhlet extraction apparatus (Matissek, Steiner, & Fischer, 2013). In brief, aliquots of 5, 10, and 20 g of freeze-dried gac fruit aril, tomato fruit, and carrot root were weighed into an extraction thimble, respectively. Subsequently, total lipids were extracted with 200 mL n-hexane in a Soxhlet extractor for 5 h. n-Hexane was removed by evaporation under reduced pressure at 40 °C and the lipid content was determined gravimetrically. 2.5. Ultrasound-assisted extraction of carotenoids of samples prior to and after digestion After grinding the freeze-dried samples with mortar and pestle, aliquots of 30 mg gac fruit aril, tomato fruit, and carrot root were soaked for 1 h in 1 mL water. Regarding the samples derived from the in vitro digestion, aliquots of 3 mL of the supernatant or micellar phase (filtrate) were used for extraction. The respective samples were combined with 3 mL of a ternary mixture (1:1:1, v/v/v) of methanol/ethyl acetate/ light petroleum containing each 0.1 g/L BHA and 0.1 g/L BHT to prevent oxidation. The sample-solvent mixture was then probe-sonicated (Sonopuls HD 3100 equipped with a MS 72 microtip, Bandelin, Berlin, Germany) at 70% amplitude for 15 s and subsequently centrifuged at 5000 U/min (~2600 ×g) for 10 min. The organic phase was separated and the remaining lower phase or pellet was re-extracted three times likewise. The combined organic phases were washed with 3 mL of aqueous 10% (w/v) sodium chloride by vigorously shaking for 30 s. Again, the organic phase was separated, evaporated to dryness using a gentle stream of nitrogen, and stored at −80 °C until HPLC analyses. 2.6. HPLC-DAD analyses and quantitation of major carotenoids Prior to HPLC analyses, the dried extracts were made up in MTBE and methanol (90:10, v/v) and membrane-filtered (0.45 μm, PTFE, Chromafil, O-45/15 MS, Macherey-Nagel, Düren, Germany). Chromatographic separation was conducted with a Hewlett-Packard 1100 series HPLC (Palo Alto, CA, USA) equipped with a G1315A photodiode array detector. Carotenoids were separated at 25 °C on a YMC (YMC Europe, Dinslaken, Germany) Carotenoid Column (C30, 250 × 4.6 mm i.d., 3 μm particle size) protected by a guard column (10 × 4.0 mm i.d.) of the same material. The mobile phase consisted of methanol/MTBE/water (80:18:2, v/v/v with 0.04% w/v ammonium acetate) as eluent A and

To prepare the aqueous emulsifier solution, 1% (w/w) soy lecithin was dispersed in ultrapure water and stirred overnight at 7 °C. Subsequently, the emulsifier solution was slowly combined with a total of 20% (w/w) Miglyol 812N under continuous shearing at 15,000 rpm for 2 min (Homogenizer SilentCrusher, Heidolph Instruments, Schwabach, Germany). The formed pre-emulsion was homogenized at 100,000 kPa in four passes using a Microfluidizer M-110 EH (Microfluidics, Newton, MA, USA). Particle size distribution of the final O/W emulsion was analyzed using a static light-scattering instrument (Horiba LA-950, Retsch Technology, Germany). Samples were diluted in water to a droplet concentration of approximately 0.005% (w/w), thus preventing multiple scattering effects. A relative refractive index of 1.44 was used. Particle size measurements were recorded as mean volume-weighted diameter d43 (de Brouckere mean). 2.8. In vitro digestion model 2.8.1. Experimental design and test food assembly The design of the four different experiments, i.e. two carotenoid dose-normalized (experiments 1 and 2) and two test meal size-normalized digestion (experiments 3 and 4) experiments, is summarized in Table 1. In experiment 1, equal carotenoid doses, i.e. 0.19 ± 0.03 mg of lycopene and 0.02 ± 0.01 mg of β-carotene per digestion system were fed to the digestion system irrespective of the food source. Experiment 2 was identical to experiment 1 except that lipid contents of the tomato fruit and carrot root test meals were increased by the addition of the above-mentioned O/W emulsion to match that of the gac fruit aril test meal. Furthermore, the lipid content of the gac fruit test meal was adjusted by adding the same amount of lipids. In experiment 3, equal portions of plant material homogenates (0.5 g each) of the respective samples were subjected to the digestion model. Experiment 4 was identical to experiment 3; however, lipid contents of the tomato fruit and carrot root test meals were adjusted to that of the gac fruit aril test meal in order to allow a digestion at equal fat contents. Furthermore, the lipid content of the gac fruit test meal was increased by the same amount of lipids (see Table 1). 2.8.2. In vitro digestion The in vitro digestion followed the INFOGEST protocol of Minekus et al. (2014) and that of Schweiggert, Mezger, Schimpf, Steingass, and Carle (2012) with modifications to allow the digestion of the comparably lipid-rich gac fruit aril. The activities of α-amylase, pepsin, trypsin, and lipases were determined in the respective reagents according to Minekus et al. (2014) in order to adjust the activities in the respective digestive fluids as recommended. The in vitro digestion protocol will be briefly described in the following. First, the respective amount of test meal (gac fruit aril, tomato fruit, and carrot root homogenate) and that of O/W emulsion (see Table 1 for exact amounts) were combined with phosphate buffer saline (Xia, McClements, & Xiao, 2015) to reach

Please cite this article as: Müller-Maatsch, J., et al., Carotenoids from gac fruit aril (Momordica cochinchinensis [Lour.] Spreng.) are more bioaccessible than those from carrot root and tomato fruit..., Food Research International (2016), http://dx.doi.org/10.1016/j.foodres.2016.10.053

4

J. Müller-Maatsch et al. / Food Research International xxx (2016) xxx–xxx

Table 1 Experimental design of carotenoid dose-normalized and test meal size-normalized in vitro digestions with and without added lipids. The amounts of sample and added lipids are given in grams [g]. Carotenoid dose-normalized Experiment

1

b,c

Gac fruit aril Tomato fruitb Carrot rootc a b c

Meal size-normalized

2

Sample

Sample

+

Added lipids

0.1 4 0.2

0.1 4 0.2

+ + +

0.05 0.05 0.05

a

3

4

Sample

Sample

+

Added lipidsa

0.5 0.5 0.5

0.5 0.5 0.5

+ + +

0.3 0.3 0.3

Lipids were added in form of a 20% (w/w) oil-in-water emulsion. Lycopene contents of gac fruit aril and tomato fruit were 164.4 and 4.5 mg/100 g FW, respectively. β-Carotene contents of gac fruit aril and carrot root were 20.5 and 8.3 mg/100 g FW, respectively.

a total bolus weight of 5 g. Subsequently, the test meal was combined with 3 mL simulated salivary fluid (SSF, Minekus et al. (2014)), 1 mL of α-amylase solution (750 U/mL SSF), 25 μL of aqueous 0.3 M CaCl2, and 975 μL of water. The mixture was carefully shaken (95 rpm) at 37 °C for 2 min (shaking incubator, OLS, Grant Instruments, Shepreth, UK) in order to model the oral phase. Simulating the gastric phase, the oral bolus was combined with 6 mL of simulated gastric fluid (SGF, 37 °C, Minekus et al. (2014)) and the pH was adjusted to 4 ± 0.1 using 1 M HCl. Subsequently, 2 mL of porcine pepsin solution (20,000 U/mL SGF), 10 μL of aqueous 0.15 M CaCl2, and 1.99 mL of water were added. Then, the pH was adjusted to pH 3 ± 0.1 using 1 M HCl, the mixture was flushed with nitrogen, and incubated for 2 h while shaken as mentioned above. In the following intestinal phase, 3.515 mL of SIF (37 °C) was added to the gastric chyme and the pH was set to pH 6 ± 0.1 using 1 M NaOH. Subsequently, 6 mL porcine pancreatin solution (2400 U lipase/mL SIF), 472 μL cholesterol esterase solution (31.8 U/mL SIF), 13.4 μL Candida rugosa lipase solution (1117 U/mL SIF), 6 mL bile extract solution (12 mg/mL SIF), 40 μL aqueous 0.3 M CaCl2, and 3.96 mL water were added. Consequently, the pH of the mixture was adjusted to 7 ± 0.1 using 1 M HCl. This mixture was flushed with nitrogen and incubated for 2 h in the aforementioned shaking incubator (95 rpm, 37 °C). After completion of the intestinal phase, samples were made up to 50 mL with water and centrifuged at 75,000 ×g and 10 °C for 60 min (Avanti J-26 XPI, Beckman-Coulter, Krefeld, Germany) to separate solids from the aqueous phase. An aliquot of the supernatant was membrane-filtrated (0.2 μm, cellulose acetate, Klaus Ziemer, Langerwehe, Germany) to obtain the micellar phase. An aliquot of the micellar phase and that of the unfiltered supernatant were stored at −80 °C until further analyses. The supernatant was considered to contain the liberated carotenoids. Thus, the term “liberation” as used below refers to the percentage of the incorporated carotenoid

doses being released from the food matrix into the digestive fluids. The filtrate of the supernatant (micellar phase) was considered to solely contain micellized carotenoids, i.e. those that were incorporated in mixed micelles, and thus, were considered to be bioaccessible and potentially ready for absorption by the enterocyte (Table 2). Hence, the term “bioaccessibility” as used below refers to the percentage of the fed carotenoid dose that has been micellized during the simulated digestion. 2.9. Statistical analyses All in vitro digestions were carried out in triplicate and randomized order. Analytical determinations described above were conducted in duplicate. The results of the in vitro digestions were statistically examined using the general linear model (GLM) procedure and the GLM one-way ANOVA procedures of SPSS 22 (IBM, Armonk, NY, USA). Differences as identified by the Tukey's test were considered significant at p b 0.05 and results are given as mean ± standard deviation. 3. Results and discussion 3.1. Identification and quantitation of carotenoids Major carotenoids identified in the gac fruit aril samples were (allE)-β-carotene (peak 3 in Fig. 3) and (all-E)-lycopene (peak 10 in Fig. 3), amounting to 20.5 ± 0.5 and 164.4 ± 5.4 mg/100 g FW, respectively. Furthermore, α-carotene (peak 2 in Fig. 3) and lycopene (Z)-isomers (peaks 4–9, and 11 in Fig. 3) were detected. As compared to gac fruit aril, tomato fruit contained substantially less (all-E)-lycopene (4.5 ± 0.8 mg/100 g FW) and (all-E)-β-carotene (0.5 ± 0.0 mg/100 g FW). Noteworthy, the proportion of (Z)-lycopene isomers present in the

Table 2 Actual carotenoid doses subjected to the in vitro digestion (in μg per sample) and respective absolute amount of carotenoid (in μg) present in the micellar fraction after their respective digestion. Carotenoid dose-normalized Experiment

1

Meal size-normalized 2

With added lipids β-Carotene

Gac fruit aril

b

b

Carrot root Lycopene

Gac fruit arilc Tomato fruitc

a b c

Subjected dose Amount micellized Subjected dose Amount micellized Subjected dose Amount micellized Subjected dose Amount micellized

23.2 ± 1.2 5.2 ± 0.0 17.5 ± 0.3 0.1 ± 0.0 185.5 ± 9.7 61.4 ± 0.0 182.0 ± 1.5 3.2 ± 0.9

28.6 ± 7.2 7.7 ± 0.0 17.6 ± 0.6 0.5 ± 0.2 229.0 ± 57.9 66.3 ± 0.0 181.6 ± 0.6 4.1 ± 1.3

3 a

4 With added lipidsa

110.4 ± 5.8 32.5 ± 0.0 43.7 ± 0.9 0.3 ± 0.1 883.6 ± 46.8 203.5 ± 0.0 25.6 ± 0.6 0.3 ± 0.0

111.1 ± 5.8 31.0 ± 0.0 43.8 ± 1.5 3.0 ± 1.1 889.7 ± 46.3 160.0 ± 0.0 25.5 ± 1.2 0.5 ± 0.2

Lipids were added in form of a 20% (w/w) oil-in-water emulsion. β-Carotene contents of gac fruit aril and carrot root were 20.5 and 8.3 mg/100 g FW, respectively. Lycopene contents of gac fruit aril and tomato fruit were 164.4 and 4.5 mg/100 g FW, respectively.

Please cite this article as: Müller-Maatsch, J., et al., Carotenoids from gac fruit aril (Momordica cochinchinensis [Lour.] Spreng.) are more bioaccessible than those from carrot root and tomato fruit..., Food Research International (2016), http://dx.doi.org/10.1016/j.foodres.2016.10.053

J. Müller-Maatsch et al. / Food Research International xxx (2016) xxx–xxx

Fig. 3. HPLC chromatogram of carotenoids from gac fruit aril, carrot root, and tomato fruit monitored at 446 nm. Vertical axes of the different panes are not the same scale. Peak assignment was as follows: 1: (all-E)-lutein; 2: (all-E)-α-carotene; 3: (all-E)-β-carotene; 4–9: lycopene (Z)-isomers; 10: (all-E)-lycopene; 11: lycopene (Z)-isomer.

carotenoid extract from tomato fruit (12.5% of total lycopene content) was similar to that obtained from gac fruit aril (7.8% of total lycopene content). In addition, the major carotenoid present in carrot root was (all-E)-β-carotene (8.3 ± 0.9 mg/100 g FW), being accompanied by lower amounts of (all-E)-α-carotene (4.1 ± 0.3 mg/100 g FW) and (all-E)-lutein (peak 1 in Fig. 3). The obtained results are in agreement with previous reports (Ishida et al., 2004; Marx, Schieber, & Carle, 2000; USDA, 2015). 3.2. Characterization of plant tissue and chromoplastidal carotenoid deposition form Dry matter of the homogenated test meal samples was 29.8 ± 0.01 g/100 g FW, 13.2 ± 0.3 g/100 g FW, and 5.4 ± 0.1 g/100 g FW in gac fruit aril, carrot root, and tomato fruit, respectively. Furthermore, gac fruit aril contained 11.17 ± 0.01 g/100 g FW of lipids, by far exceeding the lipid contents of carrot roots (0.16 ± 0.00 g/100 g FW) and tomato fruit (0.05 ± 0.00 g/100 g FW). Similar average values for the lipid contents have been reported previously (Kha, Nguyen, Roach, Parks, & Stathopoulos, 2013; USDA, 2015). For instance, Kha et al. (2013) reported gac fruit aril to contain 10.2 g lipids/100 g FW and 20 g dry matter/100 g FW. Furthermore, we aimed at characterizing the gac fruit chromoplasts, as they represent the cellular organelle containing lipid-dissolved, protein-bound, solid-crystalline, or liquid-crystalline carotenoid aggregates in all non-green plant tissues rich in carotenoids. As depicted in Fig. 4a and b, the chromoplasts of the gac fruit aril were intensely red colored, round to oval shaped mostly having an approximate diameter between

5

3 and 5 μm. Despite exhibiting poor cell-to-cell adhesion, cell walls and vacuoles were still intact. These findings were un-expected due to the pasty, liquid appearance of the aril. In contrast to gac fruit aril, carrot root (Fig. 4c) and tomato fruit (Fig. 4d) cells contained large needleshaped carotenoid crystals. The crystalloid nature of carrot and tomato chromoplasts has been reported previously (Schweiggert, Steingass, Heller, Esquivel, & Carle, 2011; Zhou, Gugger, & Erdman, 1994). The high lipid content of gac fruit aril gave rise to the hypothesis that carotenoids contained therein might be deposited in a lipid-dissolved form. In order to support or disprove this hypothesis, we conducted a solubility estimation by analogy to Schweiggert et al. (2011) and Hempel et al. (2014). Our estimation was based on the aforementioned total lipid content (11.2 g/100 g FW), total carotenoid content (20.5 and 164.4 mg/100 g FW for β-carotene and lycopene, respectively), and previously reported maximum solubilities of the respective carotenoids in dietary lipids (0.141% w/w for β-carotene and 0.075% w/w for lycopene, Hempel et al. (2014)). According to the calculation, the lipid content in gac fruit aril would be by far insufficient for the dissolution of all gac fruit aril carotenoids, even if all lipids were available in the chromoplasts. The hypothetical amount of lipids required to dissolve total carotenoids in 100 g of fresh gac fruit aril would be 233.7 g, i.e. 219.2 g for 164.4 mg lycopene and 14.5 g for 20.5 mg β-carotene. Moreover, the low proportion of carotenoid (Z)-isomers in the gac fruit aril (7.8%), similar to that in tomato fruit (12.5%) with a crystalloid deposition form of carotenoids, supports the hypothesis of a non-lipid-dissolved deposition form. In plant tissues with lipid-dissolved carotenoid forms, higher ratios of (Z)-isomers were observed previously (Cooperstone et al., 2015; Hempel et al., 2014; Pott, Marx, Neidhart, Mühlbauer, & Carle, 2003). For instance, Hempel et al. (2014) found the fruit of peach palm (Bactris gasipaes Kunth) to contain 29–57% of their total carotenoids as (Z)-isomers. These fruits contained typical globular chromoplasts and sufficient lipids to dissolve their total carotenoids, thus supporting the assumption of completely lipid-dissolved carotenoids in peach palm fruit (Hempel et al., 2014). Furthermore, Cooperstone et al. (2015) reported tangerine tomatoes to similarly contain globular chromoplasts with apparently lipid-dissolved carotenoids. In such tomatoes, (Z)-isomers were present at a high ratio of up to 94% of total carotenoids. Since tangerine tomatoes do not contain appreciable amounts of lipids, Cooperstone et al. (2015) suggested that the (Z)isomers themselves might have been responsible for the observed oily physical form. In brief conclusion, the carotenoids of gac fruit aril are unlikely to be deposited in a lipid-dissolved form. At present, the exact deposition form of gac fruit aril carotenoids remains uncertain, since our preparations of the pasty gac fruit aril for transmission electron microscopy have been yet unsuccessful. Further study is on-going. Nevertheless, our results suggest that the main proportion of the gac fruit aril carotenoids might be deposited as small, submicroscopic crystallites within the chromoplasts. Such rare submicroscopic aggregate forms have earlier been reported in Physalis sp. berries (Schweiggert & Carle, 2015; Sitte, Falk, & Liedvogel, 1980). 3.3. Carotenoid liberation and bioaccessibility after in vitro digestion 3.3.1. In vitro digestion of test foods without added lipids Comparing the carotenoid dose-normalized digestion experiments (Table 2, experiment 1), test food boli either containing 0.1 g of gac fruit aril, 4 g of tomato fruit, and 0.2 g of carrot root homogenate, respectively, were subjected to the in vitro digestion model. As depicted in Fig. 5, β-carotene liberation from the test foods into the intestinal fluids was 29.5 ± 1.7% from gac fruit aril, being 5.7-fold higher than that from carrot root (5.2 ± 0.5%). By analogy, bioaccessibility of β-carotene from gac fruit aril (22.6 ± 0.9%) was 45-fold higher than from carrot root (0.5 ± 0.2%, Fig. 5). Furthermore, both the liberation and bioaccessibility of lycopene from gac fruit aril (51.3 ± 2.6% and 33.2 ± 3.1%, respectively) clearly exceeded those of lycopene from tomato fruit (15.9 ± 2.8% and 1.8 ± 0.5%, Fig. 5). Since the highly different test meal sizes (see Table

Please cite this article as: Müller-Maatsch, J., et al., Carotenoids from gac fruit aril (Momordica cochinchinensis [Lour.] Spreng.) are more bioaccessible than those from carrot root and tomato fruit..., Food Research International (2016), http://dx.doi.org/10.1016/j.foodres.2016.10.053

6

J. Müller-Maatsch et al. / Food Research International xxx (2016) xxx–xxx

Fig. 4. Light micrographs of chromoplasts of gac fruit aril (a (DIC) and b), carrot root (c), and tomato fruit (d).

1) might have influenced the digestion, we conducted further experiments with test foods normalized by their sample weight instead of carotenoid doses. Thus, equal amounts of 0.5 g fresh sample of gac

fruit aril (with 822 μg lycopene and 102 μg β-carotene per test meal), tomato fruit (with 23 μg lycopene per test meal) or carrot root (42 μg β-carotene per test meal) were subjected to the digestion model

Fig. 5. Liberation and bioaccessibility of β-carotene and lycopene from gac fruit aril, carrot root, and tomato fruit with and without added lipids. The term “equal carotenoid dosage” refers to experiments 1 and 2, and the term “equal fresh weight” refers to experiments 3 and 4 as shown in more detail in Table 1. Data is presented as mean of three digestion experiments with error bars as standard deviation. Carotenoids from gac fruit aril were consistently more bioaccessible than those from carrot and tomato (p b 0.05), irrespective if lipids had been added to the test meal or not.

Please cite this article as: Müller-Maatsch, J., et al., Carotenoids from gac fruit aril (Momordica cochinchinensis [Lour.] Spreng.) are more bioaccessible than those from carrot root and tomato fruit..., Food Research International (2016), http://dx.doi.org/10.1016/j.foodres.2016.10.053

J. Müller-Maatsch et al. / Food Research International xxx (2016) xxx–xxx

(Table 2). As a result, likewise, liberation and bioaccessibility of both lycopene and β-carotene remained to be significantly higher from gac fruit aril than from carrot root and tomato fruit (Fig. 5). Increasing the gac fruit aril test meal size from 0.1 g to 0.5 g (experiments 1 and 3 in Table 1) consequently resulted in a 5-fold increase of the subjected carotenoid dose (Table 2), i.e., from 23.2 to 110.4 μg β-carotene per experiment. After digestion, the absolute amounts of micellized β-carotene increased approximately 6-fold from 5.2 to 32.5 μg per digestion (cf. experiment 1 and 3 in Table 2). Briefly, the 5-fold increased dose led to a 6-fold higher response, resulting in an increase of β-carotene bioaccessibility from 22.6 ± 0.9% to 29.4 ± 1.7% (experiment 1 and 3 in Fig. 5). In contrast, a reverse trend was observed for lycopene, which was slightly less bioaccessible from the greater (23.0 ± 2.6%) than from the smaller test meal (33.2 ± 3.1%). In contrast to the deviations found for the digestion of gac fruit aril, increasing the carrot test meal size from 0.2 to 0.5 g (experiments 1 and 3, Table 1) did not significantly alter liberation and bioaccessibility of β-carotene from the root matrix (Fig. 5). Regarding tomato fruit, lowering the test meal size from 4 g to 0.5 g (experiments 1 and 3, Table 1), lycopene liberation decreased from 15.9 ± 2.8% to 7.1 ± 1.7%, while the bioaccessibility from tomato fruit homogenate remained unchanged (p b 0.05, Fig. 5). In brief, our bioaccessibility results were widely independent of the digested amount of test food, except for some slight deviations regarding the digestion of gac fruit aril. However, even when considering these deviations, carotenoids from gac fruit aril were consistently and significantly more bioaccessible than those of tomato fruits and carrot roots. The poor bioaccessibility of carotenoids from raw carrot root and tomato fruit has been exhaustively discussed in literature, being most likely related to their large, solid-crystalline aggregation forms in the chromoplasts (Fig. 4c and d), their negligible lipid content, and their potentially more rigid polymeric carbohydrates in the cell walls (Lemmens et al., 2014; Schweiggert & Carle, 2015; Xia et al., 2015; Zhang et al., 2016b). We believe that superior carotenoid bioaccessibility from gac fruit aril is related to their genuine deposition form in the arils, despite the fact that the notably exceptional, presumably submicroscopic aggregates deposited in gac fruit arils have not yet been fully elucidated. Since the comparably high lipid content of the gac fruit aril might have contributed to the above described findings, we conducted further experiments with added lipids in order to equalize the lipid content of all test foods. 3.3.2. In vitro digestion of test foods with added lipids In order to add lipids in a well-defined, reproducible way, an O/W emulsion with 20% (w/w) lipid content was prepared, consisting of droplets with a mean diameter d43 of 0.29 ± 0.05 μm, a D10 of 0.23 μm, a D50 of 0.29 μm, and D90 of 0.35 μm. Accordingly, O/W emulsions with droplet sizes between 0.14 μm and 0.46 μm were previously used as successful excipients in digestion experiments (Zhang et al., 2016a). The O/W emulsion was used to adjust the lipid content of the carrot root and tomato fruit test meals to that of genuine gac fruit aril, i.e. 11.17 ± 0.01 g per 100 g FW. In consent with previous studies (Lemmens et al., 2014), the amount of liberated carotenoids as depicted in Table 2 from test meals was improved by lipid enrichment. Bioaccessibility in the carotenoid dose-normalized experiments increased from 0.5 to 2.9% for β-carotene from carrot root test meals and from 1.8 to 2.2% for lycopene from tomato fruit test meals (Fig. 5); however, falling far short of the superior striking liberation and bioaccessibility of β-carotene and lycopene from genuine gac fruit aril without added lipids (Fig. 5), irrespective of the test meal size. Unexpectedly, carotenoid bioaccessibility from gac fruit aril without added lipids was not significantly improved by the addition of lipids (Fig. 5, Table 2). Possibly, genuine lipid contents of gac fruit aril sufficed to liberate and micellize a certain amount of carotenoids within the limited times of oral, gastric, and intestinal phase (in total: 6–7 h), possibly being too short to reach saturation of the genuine lipids with carotenoids. Thus, adding lipids might have been ineffective to increase carotenoid bioaccessibility from gac fruit aril. Noteworthy, Palmero, Panozzo, Simatupang, Hendrickx, and

7

van Loey (2014) have previously observed a non-digestible lipid phase in the simulated intestinal fluids when exceeding a certain amount of lipids by the addition of an O/W emulsion to the test foods, evidently surpassing the capacity of the used in vitro digestion model. However, in our experiments, a non-digestible oil phase was not observed. 4. Conclusion The present study supports previous findings providing further evidence of the influence of food matrix and chromoplastidal deposition form of carotenoids on their liberation and bioaccessibility during human digestion. The gac fruit aril contained 11 g lipids per 100 g FW, approximately 30 g dry matter per 100 g FW, and outstandingly high levels of two nutritionally important carotenoids, β-carotene and lycopene. According to our in vitro digestion experiments, liberation and bioaccessibility of both carotenoids were up to eight times higher from gac fruit aril than from carrot root and tomato fruit. Although bioaccessibility from the aforementioned sources was enhanced by increasing lipid contents in carrot root and tomato fruit test meals to the level of gac fruit aril meals, the superior carotenoid bioaccessibility of non-supplemented gac fruit arils was not reached. Microscopic observation of chromoplasts, solubility estimations, and the low content in carotenoid (Z)-isomers supported our hypothesis that gac fruit aril carotenoids might be deposited in a non-lipid-dissolved form, possibly as submicroscopic crystallites. Such small carotenoid aggregates would possess a much higher surface-to-volume ratio than the large crystalloid aggregates of carrot and tomato, thus dissolving more rapidly in dietary lipids during digestion, and ultimately boosting carotenoid bioaccessibility. Irrespective of the reason for its high bioaccessibility of lycopene and β-carotene, gac fruit aril may provide a valuable source of highly bioavailable, nutritionally relevant carotenoids for humans. Noteworthy, our findings need to be corroborated by a human intervention trial. Nevertheless, food supplements based on gac fruit aril as well as increasing the daily intake of gac fruit may support the fight against vitamin A deficiency, especially in Asian countries, where gac fruit is a traditional, but underestimated component of the diet. Conflict of interest All authors declare no conflict of interest. Acknowledgments We acknowledge the helpful support of Claudia Gras, Lutz Großmann, Christiane Schweiggert (all from University of Hohenheim), and Tuyen Chan Kha (Nong Lam University, Vietnam) in fruit photography sessions, emulsion preparation and gac fruit research, respectively. References Aoki, H., Kieu, N. T. M., Kuze, N., Tomisaka, K., & Chuyen, N. V. (2002). Carotenoid pigments in gac fruit (Momordica cochinchinensis Spreng.). Bioscience, Biotechnology, and Biochemistry, 66, 2479–2482. Britton, G., & Khachik, F. (2009). Carotenoids in food. In G. Britton, S. Liaaen-Jensen, & H. Pfander (Eds.), Carotenoids (pp. 45–66). Basel: Birkhäuser Verlag. Cooperstone, J. L., Ralston, R. A., Riedl, K. M., Haufe, T. C., Schweiggert, R. M., King, S. A., ... Schwartz, S. J. (2015). Enhanced bioavailability of lycopene when consumed as cisisomers from tangerine compared to red tomato juice, a randomized, cross-over clinical trial. Molecular Nutrition & Food Research, 59, 658–669. Failla, M. L., Chitchumroonchokchai, C., & Ishida, B. K. (2008). In vitro micellarization and intestinal cell uptake of cis isomers of lycopene exceed those of all-trans lycopene. Journal of Nutrition, 138, 482–486. Hempel, J., Amrehn, E., Quesada, S., Esquivel, P., Jimenez, V. M., Heller, A., ... Schweiggert, R. M. (2014). Lipid-dissolved gamma-carotene, beta-carotene, and lycopene in globular chromoplasts of peach palm (Bactris gasipaes Kunth) fruits. Planta, 240, 1037–1050. Ishida, B. K., Turner, C., Chapman, M. H., & McKeon, T. A. (2004). Fatty acid and carotenoid composition of gac (Momordica cochinchinensis Spreng) fruit. Journal of Agricultural and Food Chemistry, 52, 274–279.

Please cite this article as: Müller-Maatsch, J., et al., Carotenoids from gac fruit aril (Momordica cochinchinensis [Lour.] Spreng.) are more bioaccessible than those from carrot root and tomato fruit..., Food Research International (2016), http://dx.doi.org/10.1016/j.foodres.2016.10.053

8

J. Müller-Maatsch et al. / Food Research International xxx (2016) xxx–xxx

Kha, T. C., Nguyen, M. H., Roach, P. D., Parks, S. E., & Stathopoulos, C. (2013). Gac fruit: Nutrient and phytochemical composition, and options for processing. Food Reviews International, 29, 92–106. Kopec, R. E., Riedl, K. M., Harrison, E. H., Curley, R. W., Jr., Hruszkewycz, D. P., Clinton, S. K., & Schwartz, S. J. (2010). Identification and quantification of apo-lycopenals in fruits, vegetables, and human plasma. Journal of Agricultural and Food Chemistry, 58, 3290–3296. Kubola, J., & Siriamornpun, S. (2011). Phytochemicals and antioxidant activity of different fruit fractions (peel, pulp, aril and seed) of Thai gac (Momordica cochinchinensis Spreng). Food Chemistry, 127, 1138–1145. Lemmens, L., Colle, I., van Buggenhout, S., Palmero, P., van Loey, A., & Hendrickx, M. (2014). Carotenoid bioaccessibility in fruit- and vegetable-based food products as affected by product (micro) structural characteristics and the presence of lipids: A review. Trends in Food Science & Technology, 38, 125–135. Marx, M., Schieber, A., & Carle, R. (2000). Quantitative determination of carotene stereoisomers in carrot juices and vitamin supplemented (ATBC) drinks. Food Chemistry, 70, 403–408. Matissek, R., Steiner, G., & Fischer, M. (2013). Lebensmittelanalytik. (5., vollst. überarb. u. akt. Aufl.). Springer-Lehrbuch. Berlin: Springer. Minekus, M., Alminger, M., Alvito, P., Ballance, S., Bohn, T., Bourlieu, C., ... Brodkorb, A. (2014). A standardised static in vitro digestion method suitable for food - An international consensus. Food & Function, 5, 1113–1124. Müller, L., Caris-Veyrat, C., Lowe, G., & Böhm, V. (2016). Lycopene and its antioxidant role in the prevention of cardiovascular diseases-A critical review. Critical Reviews in Food Science and Nutrition, 56, 1868–1879. Nhung, D. T. T., Bung, P. N., Ha, N. T., & Phong, T. K. (2010). Changes in lycopene and beta carotene contents in aril and oil of gac fruit during storage. Food Chemistry, 121, 326–331. Palmero, P., Panozzo, A., Simatupang, D., Hendrickx, M., & van Loey, A. (2014). Lycopene and β-carotene transfer to oil and micellar phases during in vitro digestion of tomato and red carrot based-fractions. Food Research International, 64, 831–838. Pott, I., Marx, M., Neidhart, S., Mühlbauer, W., & Carle, R. (2003). Quantitative determination of beta-carotene stereoisomers in fresh, dried, and solar-dried mangoes (Mangifera indica L.). Journal of Agricultural and Food Chemistry, 51, 4527–4531. Rock, C. L. (2009). Carotenoids and cancer. In G. Britton, S. Liaaen-Jensen, & H. Pfander (Eds.), Carotenoids (pp. 269–286). Basel: Birkhäuser Verlag. Rodriguez-Amaya, D. B. (2016). Natural food pigments and colorants. Current Opinion in Food Science, 7, 20–26. Saini, R. K., Nile, S. H., & Park, S. W. (2015). Carotenoids from fruits and vegetables: Chemistry, analysis, occurrence, bioavailability and biological activities. Food Research International, 76, 735–750.

Schweiggert, R. M., & Carle, R. (2015). Carotenoid deposition in plant and animal foods and its impact on bioavailability. Critical Reviews in Food Science and Nutrition. http://dx.doi.org/10.1080/10408398.2015.1012756. Schweiggert, R. M., Mezger, D., Schimpf, F., Steingass, C. B., & Carle, R. (2012). Influence of chromoplast morphology on carotenoid bioaccessibility of carrot, mango, papaya, and tomato. Food Chemistry, 135, 2736–2742. Schweiggert, R. M., Steingass, C. B., Heller, A., Esquivel, P., & Carle, R. (2011). Characterization of chromoplasts and carotenoids of red- and yellow-fleshed papaya (Carica papaya L.). Planta, 234, 1031–1044. Sitte, P., Falk, H., & Liedvogel, B. (1980). Chromoplasts. In F. -C. Czygan (Ed.), Pigments in plants (pp. 117–148). Stuttgart New York: Fischer. Tang, G., & Russel, M. R. (2009). Carotenoids as provitamin A. In G. Britton, S. LiaaenJensen, & H. Pfander (Eds.), Carotenoids (pp. 149–172). Basel: Birkhäuser Verlag. USDA (2015). US Department of Agriculture, Agricultural Research Service, Nutrient Data Laboratory: National Nutrient Database for standard reference, release 28.: Version current: September 2015. http://www.ars.usda.gov/nea/bhnrc/ndl (Accessed 28.07. 2016) Vuong, T. L. (2000). Underutilized β-carotene–rich crops of Vietnam. Food and Nutrition Bulletin, 21, 173–181. Vuong, L. T., Dueker, S. R., & Murphy, S. P. (2002). Plasma β-carotene and retinol concentrations of children increase after a 30-d supplementation with the fruit Momordica cochinchinensis. American Journal of Clinical Nutrition, 75, 872–879. Vuong, L. T., Franke, A. A., Custer, L. J., & Murphy, S. P. (2006). Momordica cochinchinensis Spreng. (gac) fruit carotenoids reevaluated. Journal of Food Composition and Analysis, 19, 664–668. West, C. E., & Poortvliet, E. J. (1993). The carotenoid content of foods with special reference to developing countries. Arlington, VA, USA: VITAL, International Science and Technology Institute. Xia, Z., McClements, D. J., & Xiao, H. (2015). Influence of physical state of β-carotene (crystallized versus solubilized) on bioaccessibility. Journal of Agricultural and Food Chemistry, 63, 990–997. Zhang, R., Zhang, Z., Zou, L., Xiao, H., Zhang, G., Decker, E. A., & McClements, D. J. (2016a). Enhancement of carotenoid bioaccessibility from carrots using excipient emulsions: Influence of particle size of digestible lipid droplets. Food & Function, 7, 93–103. Zhang, R., Zhang, Z., Zou, L., Xiao, H., Zhang, G., Decker, E. A., & McClements, D. J. (2016b). Impact of lipid content on the ability of excipient emulsions to increase carotenoid bioaccessibility from natural sources (raw and cooked carrots). Food Biophysics, 11, 71–80. Zhou, J. R., Gugger, E. T., & Erdman, J. W. (1994). Isolation and partial characterization of an 18 kDa carotenoid-protein complex from carrot roots. Journal of Agricultural and Food Chemistry, 42, 2386–2390.

Please cite this article as: Müller-Maatsch, J., et al., Carotenoids from gac fruit aril (Momordica cochinchinensis [Lour.] Spreng.) are more bioaccessible than those from carrot root and tomato fruit..., Food Research International (2016), http://dx.doi.org/10.1016/j.foodres.2016.10.053