Sulforaphane Bioavailability from Glucoraphanin-Rich Broccoli: Control by Active Endogenous Myrosinase

Abstract

Glucoraphanin from broccoli and its sprouts and seeds is a water soluble and relatively inert precursor of sulforaphane, the reactive isothiocyanate that potently inhibits neoplastic cellular processes and prevents a number of disease states. Sulforaphane is difficult to deliver in an enriched and stable form for purposes of direct human consumption. We have focused upon evaluating the bioavailability of sulforaphane, either by direct administration of glucoraphanin (a glucosinolate, or β-thioglucoside-N-hydroxysulfate), or by co-administering glucoraphanin and the enzyme myrosinase to catalyze its conversion to sulforaphane at economic, reproducible and sustainable yields. We show that following administration of glucoraphanin in a commercially prepared dietary supplement to a small number of human volunteers, the volunteers had equivalent output of sulforaphane metabolites in their urine to that which they produced when given an equimolar dose of glucoraphanin in a simple boiled and lyophilized extract of broccoli sprouts. Furthermore, when either broccoli sprouts or seeds are administered directly to subjects without prior extraction and consequent inactivation of endogenous myrosinase, regardless of the delivery matrix or dose, the sulforaphane in those preparations is 3- to 4-fold more bioavailable than sulforaphane from glucoraphanin delivered without active plant myrosinase. These data expand upon earlier reports of inter- and intra-individual variability, when glucoraphanin was delivered in either teas, juices, or gelatin capsules, and they confirm that a variety of delivery matrices may be equally suitable for glucoraphanin supplementation (e.g. fruit juices, water, or various types of capsules and tablets).

Fig 1. Mean daily excretion / conversion of SF- or GR-rich broccoli preparations to SF metabolites.

Preparations are: SF-, GR-, or GR plus Myrosinase-rich; they are delivered either in water, juice, or gel-caps; and they are derived from either sprouts or seeds, as indicated at top of graph. Boxes around data points delineate the 25^th to 75^th percentile of values, and whiskers delineate the 5^th and 95^th percentiles. Solid line through middle of box is the mean. “Cohort numbers” in the following are from Table 1: (A) Hospitalized subjects given SF-rich BSE (mean 68.9%, data from Ref. [19]); (B) Hospitalized subjects given GR-rich BSE (mean 18.0%, data from Ref. [19]); (C) Free-living subjects given GR-rich BSE (mean 11.8%, data from Ref. [7]); (D) Cohort 1 (BSE in water; mean 9.4%); (E) Cohort 3 (Xymogen BSdE in gel-caps; mean 10.4%); (F) Cohort 2 (JHU BSE in gel-caps; mean 10.3%); (G) Cohort 6 (FDBS in dilute pineapple-lime juice; mean 44%); (H) Cohort 5 (FDBS in acid-resistant gel-caps; mean 32.7%); (I) Cohort 5 (FDBS in regular gel-caps; mean 35.1%); (J) Cohort 4 (BSdP; 35.9%). There is a highly significant difference between the means of D through F^a (10.3%), and G through J^b (39.6%), (p = 0.0000, by oneway anova).

Fig 2. Mean daily excretion / conversion of GR-rich broccoli preparations to SF metabolites.

Preparations are: Commercial broccoli seed extract (BSdE) formulated into gel-caps at 69 and 230 μmol/dose (10.4% conversion overall; Table 1, Cohort 3), and JHU BSE formulated (by Xymogen) into molar equivalent doses (10.3% conversion overall; Table 1, Cohort 2). There is no difference among the treatment groups overall (p = 0.0651 by repeated measures anova)

Fig 3. Mean daily excretion / conversion of freeze-dried broccoli sprouts (FDBS) containing active myrosinase, to SF metabolites.

(see also Table 1, Cohort 4). Preparations were either pre-mixed in dilute pineapple-lime juice for 15 min at room temperature to allow autolysis in-vitro, and then consumed directly by volunteers (40.5% conversion), or swallowed in either acid-resistant or standard gel-caps (33.8% conversion). The same 5 subjects performed each test. There was no effect of subject on conversion (p = 0.2740) by repeated measures analysis of variance. There was a significant difference (by oneway anova with Bonferroni adjustment and Bartlett’s test for equal variances), between replicates at the 50 μmol dose level (p = 0.0107), but no significant difference between dose levels (pooled replicates; p = 0.3329) or dose matrices (p = 0.3735 between capsule types and p = 0.0255 between capsules and juice).

Fig 4. Mean daily excretion / conversion of broccoli seed powder (BSdP) containing active myrosinase, to SF metabolites.

Preparations were delivered in standard gel-caps. The same 5 subjects performed each test (Table 1, Cohort 4). There was no significant difference between replications (p = 0.3747 using a repeated measures analysis of variance). The mean conversion of all 3 replicates was 36% of dose.