Effect of Iron and Zinc Supplementation on B-carotene Bioavailability in Healthy Males

The study followed a double-blind crossover design with three study arms separated by one week washout periods. In short, the participants were asked to follow a diet low in carotenoids by avoiding all orange, yellow, red and green fruits and vegetables for four days. This was followed by three days of a strictly carotenoid-free diet, which only allowed foods from a specified list (Supplemental Table A2). On each study day, β-carotene was administered in the morning after a >10 hour overnight fast (Supplemental Figure A1). All participants orally ingested, in random order, a single dose of 15 mg β-carotene (BIOVEA) with either a placebo (empty capsule), 25 mg iron (FeSO4; Woerwag Pharma GmbH \& Co. KG, Boeblingen, Germany) or 30 mg zinc (ZnSO4; Woerwag Pharma) capsule. A standardised dinner was provided on the evening before the trial and standardised meals were provided during the entire intervention day (Supplemental Table A3). On the first study day, the amount of food (weight or volume) consumed by each participant for each meal was recorded and the same amounts provided during the following two arms to ensure similar food consumption, especially of fat. Water was available unrestricted for consumption throughout the day. Blood samples were drawn from an indwelling venous cannula and collected at 0 hours directly before β-carotene supplementation and then every hour for 10 hours.

For the determination of plasma concentrations of β-carotene, LDL- and HDL-cholesterol, and triacylglycerols (TAG), blood was collected in tubes containing EDTA (Sarstedt AG \& Co, Nuebrecht, Germany) and immediately centrifuged (3000 × g, 10 min, 4 °C). From the obtained plasma samples, three aliquots were stored at -80 °C until further analysis and the rest ultracentrifuged to obtain the triacylglycerol-rich fraction (TRF). For the analyses of liver and kidney function markers, plasma and serum were obtained from blood sampled at the 0- and 4-hour time points.

The TRF was prepared according to [10]. Briefly, plasma (3.5 mL) was transferred to an ultracentrifuge tube and carefully overlaid with 8 mL 1.3% sodium chloride and then ultracentrifuged (Beckman Coulter, OptimaTM L-80 XP Ultracentrifuge) using a swinging bucket rotor (SW41Ti) at 150 000 x g for 1 hour at 4 °C. Afterwards, the TRF was isolated by transferring the upper ~6 mL, which was then overlaid with nitrogen gas to minimize oxidation and stored at -80 °C until extraction.

The plasma samples were randomly extracted and analysed by HPLC according to [15]. Briefly, 40 µL plasma was extracted with an ethanol/n-butanol mixture (50:50) containing apo-80-carotenal-methyloxime (12µL/100 mL; Fluka Analytical (Merck Group KGaA), Darmstadt, Germany) as internal standard. After centrifugation, the clear supernatant was analysed by HPLC.

The TRF was extracted and analysed by HPLC [15]. For the extraction, 100 µl apo-80-carotenal-methyloxime (12 µL/100 mL) and 2 mL ethanol (for deproteination) were added to 3 mL of the TRF and vortexed for 30 sec. The solution was extracted twice with 2 mL hexane. The hexane layers were removed, combined and evaporated in a centrifugal vacuum concentrator (Christ, RVC 2-25 CD plus) and the dried sample re-dissolved in 100 µL acetonitrile and immediately analysed by HPLC.

Both the plasma and TRF samples were analysed using a Shimadzu HPLC (LC-10AD) equipped with a UV-Vis detector (SPD 20A, set at 450 nm). Carotenoids were separated using a ReproSil 80 ODS-2 column (3 µm, 250 x 4.6 mm; Dr. Maisch GmbH, Ammerbuch, Germany) and an eluent in recirculation mode (82% acetonitrile, 15% 1,4-dioxin, and 3% methanol (vol/vol) containing 100 mM ammonium acetate and 10 mM triethylamine) at a flow rate of 1.5 mL/min [15]. A β-carotene standard (≥97.0% purity, Sigma-Aldrich) was used to construct a standard curve.

Plasma TAG, HDL- and LDL-cholesterol were analysed by a clinical laboratory (Laborärzte Sindelfingen, Sindelfingen, Germany).