Article pubs.acs.org/JAFC
Flavonoid Metabolites in Human Urine during Blueberry Anthocyanin Intake Wilhelmina Kalt,*,† Jane E. McDonald,† Yan Liu,§ and Sherry A. E. Fillmore† †
Kentville Research and Development Centre, Agriculture and Agri-Food Canada, 32 Main Street, Kentville, Nova Scotia B4N 1J5, Canada § Institute of Special Economic Animal and Plant Sciences, Chinese Academy of Agricultural Sciences, No. 4899 Juye Street, Changchun 130112, China ABSTRACT: The human health benefits of anthocyanins (Anc) and other flavonoids are widely recognized. However, the flavonoid-based urinary metabolites arising in vivo after Anc intake are not well described. Human (n = 17) urine was collected while blueberry juice (BJ) was consumed daily for 28 days and once after a 7 day washout. MS/MS scanning of 664 urine samples for 18 parent Anc (PAnc) and 42 predicted Anc metabolites (AncM) yielded 371 products (i.e., MS/MS × retention time (RT)). Flavonoid-based AncM, which were likely underestimated, were almost 20 times more abundant than PAnc. Together, PAnc and AncM accounted for about 1% of the daily Anc dose. Aglycone forms were >94% of the total. Cluster analysis of the 371 Anc identified about 55 major Anc that contributed about 80% to the total Anc. The abundance of flavonoidbased Anc-derived products in the gastrointestinal tract could contribute to the health benefits of Anc-rich berries. KEYWORDS: Vaccinium angustifolium, bioavailability, LC-MS, enterohepatic, metabolite, profile
■
INTRODUCTION Although anthocyanins (Anc) are arguably the most important dietary flavonoid contributing to the health benefits of berry crops, understanding the in vivo mechanisms underlying Anc health effects is made difficult by evidence of their low bioavailability. Topics related to Anc bioactivity and bioavailability are reviewed in recent comprehensive works (see refs 1−4). Collectively animal and human studies demonstrate that 300 Anc were identified as minor contributors. The mean of the minor Anc was included with means of more than 50 major Anc in a second iteration of cluster analysis. Cluster analysis was done for results obtained for 24−0 and 0 h.
Figure 2. Schematic description of study design, sample collection, and results for total anthocyanins (Anc) and major and minor Anc identified by cluster analysis. Results are described for parent Anc (PAnc) and for Anc metabolites (AncM).
a smaller group of Anc that were notable for their large contribution to the total Anc (major Anc). These groups are described in relation to their 24−0 h profile (n = 582 samples) and 0 h profile (n = 82 samples) and with respect to the type of anthocyanidin and metabolite. The kinds of major and minor Anc for 24−0 h are also described. Total Nanomoles of Anc Excreted. Mean 24−0 h Anc excretion was 3301 ± 324.7 nmol, of which 167.3 ± 42.11 nmol (5.1%) was PAnc and which was contained in an average of seven voids (Table 2). The 0 h void, just before BJ intake, contained 1042 ± 330.8 nmol Anc, of which 51.89 ± 13.00 nmol (5.0%) was PAnc. The single void at 0 h, which was typically the first void of the morning (data not shown), accounted for 24% of the total 24 h Anc excretion. The amount of PAnc excreted in 24 h accounted for 0.049% of the Anc dose, and AncM accounted for another 0.92%. Therefore, the total 24 h urinary Anc accounted for 0.97% of the daily Anc dose administered (Table 2). On the basis of the total nanomoles of Anc and milliliters of urine excreted, the mean 24 h urinary total Anc concentration was about 1.65 μM, of which about 0.083 μM was due to PAnc. This estimate compares well with several studies (for a review, see ref 4). An average of 154 of the 371 Anc were detected in each of the 582 urine samples that made up the 24−0 h samples. This included 15 of the 18 PAnc and 142 AncM. In the 0 h urine samples (n = 82) an average of 129 Anc were detected, including 14 PAnc and 115 AncM (Table 2). The only PAnc that was not detected in any urine sample was Cyn 3,5-diglucoside. Major Anc Identified by Cluster Analysis. In the cluster analysis for 24−0 h Anc, 316 of the 371 Anc had a >99% similarity, whereas the remaining 55 Anc were separated by greater differences (Table 3). The mean (2.214 nmol) for the 316 similar, minor Anc was analyzed with the 55 major Anc. The 55 major Anc ranged from a minimum of 13.17 ± 7.135 nmol to a maximum of 130.9 ± 11.447 nmol with a mean of 48.81 nmol (Table 4). The 55 major Anc contributed 79.3% to the total 24−0 h excretion (Table 3). Forty-four of the 55 major Anc were observed in ≥75% of 664 urine samples (Table 4). In the 0 h cluster analysis of the 371 Anc, 311 shared a >99% similarity, whereas the remaining 60 Anc were separated by greater differences. The mean (0.766 nmol) of the 311 minor Anc was analyzed with the 60 major Anc. The 60 major 0 h Anc ranged from a minimum of 2.519 ± 1.160 nmol to a maximum of 66.0 ± 26.70 nmol with an average 15.19 nmol (Table 3). The 60 major Anc contributed 79.6%, whereas the 311 minor
■
RESULTS Results are reported (Figure 2) for 371 Anc with respect to relative amounts of PAnc and AncM. Cluster analysis identified 1584
DOI: 10.1021/acs.jafc.6b05455 J. Agric. Food Chem. 2017, 65, 1582−1591
Article
Journal of Agricultural and Food Chemistry
Table 2. Cumulative Output of Anthocyanins (Anc) by 17 Human Volunteers during Five 24 h Urine Collections Conducted over 5 Weeks with Daily Blueberry Juice Intakea
a
24−0 h
0h
total Anc
cumulative excretion mean signals/371
nmol no.
3301 ± 324.7 154
1042 ± 330.8 129
parent Anc
cumulative excretion percent of total mean signals/18
nmol % no.
167.3 ± 42.11 5.1 15
51.89 ± 13.00 5.0 14
Anc metabolites
cumulative excretion percent of total mean signals/353
nmol % no.
3133 ± 196.3 94.5 139
990.1 ± 110.5 95.0 115
urine samples
samples in mean mean volume
no. mL
580 2282
82 358
Anc of the urinary void just prior to ingesting blueberry juice (0 h) is compared to output for the following 24 h period (i.e. 24−0 h).
Table 3. Summary of Data Reduction by Cluster Analysis for the Nanomoles of Anthocyanins (Anc) Excreteda 24−0 h no. of Anc total Anc (nmol) mean Anc (nmol) % of total Anc with max nmol Anc with min nmol a
0h
major Anc
minor Anc
major Anc
minor Anc
55 2684 48.81 79.3 130.9 ± 11.45 13.17 ± 7.135
316 699.8 2.526 20.7 20.11 ± 3.263 0.0006
61 927.0 15.20 79.6 66.00 ± 26.70 2.519 ± 1.160
310 237.5 0.7661 20.4 6.145 ± 1.687 0.001
The 371 Anc are arranged as major and minor contributors to the total 24-0 h cumulative excretion and the 0 h void.
Anc contributed 20.4% to the total Anc in the 0 h void. Fortynine of the 60 major 0 h Anc were also among the 55 major Anc at 24−0 h listed in Table 4. Anthocyanidin Classes among Major Anc at 24−0 and 0 h. For 24−0 h excretion, the relative nmol in the pool of 55 major Anc included Cyn with 9.57%, Del with 24.5%, Mal with 7.71%, Pel with 23.4%, Peo with 9.88%, and Pet with 24.91% (Figure 3). By also including the minor Anc, these contributions were as follows: Cyn, 7.58%; Del, 19.4%; Mal, 6.11%; Pel, 18.6%; Peo, 7.83%; Pet, 19.7% (Figure 3 inset). Among the major Anc were 5 Cyn-based Anc, 11 based on Del, 7 based on Mal, 10 based on Pel, 7 based on Peo, and 15 based on Pet (Table 4). For 0 h excretion, the relative nanomoles in the pool of 60 major Anc included Cyn with 10.5%, Del with 21.4%, Mal with 7.41%, Pel with 33.0%, Peo with 7.51%, and Pet with 20.3% (Figure 2 inset). Included were 5 Cyn-based Anc, 11 Anc based on Del, 7 Anc based on Mal, 10 Anc based on Pel, 7 Anc based on Peo, and 15 Anc based on Pet (Table 5). Types of Anc Metabolites at 24−0 and 0 h. 24−0 h PAnc and Anthocyanidin Glycosides. Four PAnc made up 4.25% of the total excretion of major Anc and included all three parent Mal glycosides (glucose, galactose, and arabinose) and Peo 3-glucoside (Figure 4; Table 4). There was one major nonPAnc anthocyanidin glycoside (methylated Pel 3-glucoside). 24−0 h Aglycone Glucuronides. Anthocyanidin glucuronide conjugates were the most abundant class of major urinary Anc with 27 forms contributing 54.8% to the total nmol Anc excreted. The number ranged between 4 Mal glucuronides and 10 Del glucuronides (Figure 4; Table 4). 24−0 h Simple Aglycones. Aglycones with no phase II conjugation were the second most abundant urinary Anc,
making up 28.1% of the 55 major Anc (Figure 4). Only aglycones based on Del (1), Pel (3), and Pet (9) were among the major Anc (Figure 4; Table 4). 24−0 h Aglycone Methyl Glucuronides. Eight anthocyanidin methyl glucuronide conjugates made up 12.3% of the major Anc. Multiple major forms were seen for Mal (2), Pel (3), and Pet (3) (Figure 4; Table 4). 0 h PAnc and Anthocyanidin Glycosides. Among the 60 major Anc were 3 PAnc, specifically Mal glucoside, galactoside, and arabinoside. These three PAnc made up 2.97% of the total 24 h excretion (Figure 4). There were no other anthocyanidin glycosides among the major Anc at 0 h. 0 h Aglycone Glucuronides. Anthocyanidin glucuronide conjugates were the most abundant metabolite class, contributing 59.9% of the nmol in the 0 h void (Figure 4). Their number (MRM × RT) of major forms ranged between three forms based on Mal and eight forms for Del. 0 h Simple Aglycones. Aglycones with no phase II conjugation were the second most abundant urinary Anc making up 22.8% of the major Anc in the 0 h void (Figure 3 inset). As with the 24−0 h, only aglycones based on Del (1), Pel (4), and Pet (9) were among the major 0 h Anc. 0 h Aglycone Methyl Glucuronides. Nine aglycone methylated glucuronides contributed 14.4% to the nanomoles of major Anc excreted in the 0 h void (Figure 4). This included three forms based on Pet, two forms each for Pel and Mal, and one each for Peo and Del. Comparison of 55 Major Anc with 316 Minor Anc at 24−0 h. Types of Major and Minor Anthocyanidins. The 24−0 h nanomole excretion of major anthocyanin forms exceeded that of minor anthocyanin forms for all six anthocyanidins (Figure 3). Anc forms based on Del, Pel, and 1585
DOI: 10.1021/acs.jafc.6b05455 J. Agric. Food Chem. 2017, 65, 1582−1591
Article
Journal of Agricultural and Food Chemistry Table 4. For 24−0 h, the 55 Major Urinary Anthocyanins (Anc) Identified by Cluster Analysis among 371 Anca IDb
anthocyanidin
conjugate 1
conjugate 2
mean (nmol)
SEM
rank (nmol), 1 = max, 49 = min
observations (% of 82)
51 44 45 55 46 134 115 114 113 133 120 112 116 121 119 117 167 174 175 182 466 220 212 224 258 266 267 274 269 272 271 230 231 318 286 347 344 343 331 348 446 436 444 431 432 440 449 447 448 421 427 420 378 381 379
Cyn Cyn Cyn Cyn Cyn Del Del Del Del Del Del Del Del Del Del Del Mal Mal Mal Mal Mal Mal Mal Pel Pel Pel Pel Pel Pel Pel Pel Pel Pel Peo Peo Peo Peo Peo Peo Peo Pet Pet Pet Pet Pet Pet Pet Pet Pet Pet Pet Pet Pet Pet Pet
glucuronide glucuronide glucuronide glucuronide glucuronide 0 glucuronide glucuronide glucuronide glucuronide glucuronide glucuronide glucuronide glucuronide glucuronide glucuronide arabinoside galactoside glucoside glucuronide glucuronide glucuronide glucuronide methyl 0 0 0 glucuronide glucuronide glucuronide glucuronide glucuronide glucuronide glucoside glucuronide glucuronide glucuronide glucuronide glucuronide glucuronide 0 0 0 0 0 0 0 0 0 glucuronide glucuronide glucuronide glucuronide glucuronide glucuronide
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 methyl methyl glucoside 0 0 0 0 0 0 0 methyl methyl 0 methyl 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 methyl methyl methyl
21.79 23.66 67.40 73.07 86.44 21.91 45.67 48.24 50.31 51.10 55.64 71.63 74.35 96.70 18.19 120.78 25.85 29.91 38.47 32.43 45.53 16.99 16.85 16.02 126.51 79.48 47.14 25.23 31.72 76.14 77.21 65.04 80.33 19.67 19.94 35.31 35.99 37.34 69.64 46.83 13.17 19.30 17.36 20.05 25.03 30.12 57.14 109.31 130.85 18.909 29.86 61.17 27.33 43.74 58.76
6.205 3.917 13.394 13.014 11.064 5.560 7.403 12.850 10.838 6.486 9.304 24.852 20.692 19.056 4.677 25.173 2.578 3.555 3.639 3.906 5.450 3.209 1.116 2.782 10.903 11.616 8.919 3.666 5.675 20.390 26.719 15.867 12.365 1.868 3.247 4.207 5.797 6.482 8.616 4.687 7.135 3.909 3.549 4.015 4.023 6.097 3.980 12.743 11.447 3.750 4.581 12.734 2.349 8.328 6.695
44 42 14 17 6 43 26 23 21 22 20 12 11 5 50 3 39 37 29 33 27 52 53 54 2 8 24 40 34 10 9 15 7 47 46 32 31 30 13 25 55 48 51 45 41 35 19 4 1 49 36 16 38 28 18
54 84 95 100 100 94 63 85 90 100 94 70 76 84 74 92 99 99 100 94 94 65 99 95 100 100 99 79 99 81 67 68 98 100 98 99 96 98 95 100 38 74 72 81 94 85 99 100 100 85 79 95 99 85 100
total mean
2684.5 48.81
Frequency (%) of occurrence for major Anc for 24−0 h based on means among 17 participants and five collection dates. Parent Anc are shown in bold font. bID, identification number; where ID is underlined, compound was not among top 60 in 0 h cluster analysis. a
1586
DOI: 10.1021/acs.jafc.6b05455 J. Agric. Food Chem. 2017, 65, 1582−1591
Article
Journal of Agricultural and Food Chemistry
Figure 3. Distribution of urinary anthocyanins (Anc) based on six common anthocyanidins. Shown is the contribution (nanomole) of the 55 major Anc at 24−0 h and the 60 major Anc at 0 h, identified by cluster analysis. The contributions of >300 minor Anc for 24−0 and 0 h are included. (Inset) Relative contribution (%) of six common anthocyanidins to the total Anc, with minor Anc not included.
Pet each contributed about 25% to the pool 55 major Anc and between about 14 and 25% to the pool of 316 minor Anc. Cyn and Mal each contributed about 5−16% to the pool of major and minor Anc. Peo was distinctive in contributing 26% to the pool of major Anc and only 10% to the pool of minor Anc. Types of Major and Minor Anc Metabolites. PAnc for 24−0 h contributed about twice as many nanomoles (8.34%) to the pool of minor Anc than major Anc (4.25%). Notably, among the minor Anc were 138 non-PAnc glycosides (Table 5) that contributed 12.73% to the minor Anc. Also among the minor Anc forms were 12 glycosidic sulfate conjugates, which contributed 0.085% of total 24−0 h Anc. Major aglycone glucuronides contributed about 6.2 times more nanomoles than their minor forms. Twenty-nine major aglycone glucuronides contributed almost 55.5% to the pool of major Anc (Table 4). Fifty-eight minor aglycone glucuronide (Table 5) contributed about 33.6% to the minor Anc pool. Eight methyl aglycone glucuronides contributed about 12% to the major Anc pool (Table 4), whereas 54 minor methyl
Table 5. Distribution of 55 Major and 316 Minor Urinary Anthocyanins (Anc) at 24−0 h Based on Six Anthocyanidins and Five Types of Anc Metabolitesa,b MRM × RT (no.)
MRM × RT (no.)
major Anc
minor Anc
major Anc
minor Anc
cyanidin delphinidin malvidin pelargonidin peonidin petunidin
5 11 7 10 7 15
55 53 38 49 71 50
4 1 13 29 8
13 138 53 58 54
total
55
316
55
316
parent Anc other glycosides simple aglycone aglycone glucuronide methyl aglycone glucuronide
Major and minor Anc were identified by cluster analysis. bMRM × RT refers to MS transition at a specific retention time.
a
Figure 4. Distribution of urinary anthocyanins (Anc) based on five kinds of metabolites. Shown is the contribution (nmol) of the 55 major Anc at 24−0 h and the 60 major Anc at 0 h identified by cluster analysis. The contributions of >300 minor Anc for 24−0 and 0 h are included. (Inset) Relative contribution (%) of five kinds of anthocyanin metabolites, with minor Anc not included. 1587
DOI: 10.1021/acs.jafc.6b05455 J. Agric. Food Chem. 2017, 65, 1582−1591
Article
Journal of Agricultural and Food Chemistry
chromatographic behavior. (Five of these Cyn aglycone glucuronides were among the 55 major Anc (Table 4).) Anc behavior during chromatography can be affected by the Anc chemical milieu, such as when associated with bile components. During SPE and HPLC, urinary AncM copurified with amphipathic phospholipids,21 which are rich in bile and membranes. The rapid uptake of Anc into bile is wellknown.1,31,35,36 Positional isomers of Anc share the MRM but will chromatograph differently due to relative differences in structure and polarity. Chalcones formed from anthocyanidins in vivo may also contribute to apparent multiple forms of AncM when detected by MS/MS. Each anthocyanidin can form a chalcone intermediate by the hydrolytic opening of the Cring, leading to a mass increase of +17 amu by OH addition. Like Anc, chalcones undergo phase II conjugation and can form positional isomers.37 The chalcone backbone can also occur as cis- and trans-isomers.38 It should be noted that neither the urine SPE cleanup nor the LC-MS/MS analysis was optimized for Anc-derived chalcone analysis. Acidification of chalcones can lead to isomeric flavanone formation; however, it is not known whether chalcones derived from Anc can undergo the same reaction under the low-pH conditions employed here. The six major anthocyanidins are distinguished from each other by their B-ring hydroxylation and methylation patterns. Therefore, when detected with MS/MS, Anc are being effectively interconverted during phase II Anc methylation and hydroxylation and enteric demethoxylation and dehydroxylation.21 Novel anthocyanin forms with unknown MRMs were therefore likely present in urine and, if so, would contribute to an underestimation of urinary C6−C3−C6 content. Profiles of AncM. Although profiles of major and minor Anc at 24−0 and 0 h were complex, they were relatively similar to each other (Figures 3 and 4; Tables 4 and 5). Indeed, 49 of the 55 major Anc at 24−0 h were also among the 60 major Anc at 0 h (Table 4). The profile of major Anc, which were about 80% of the total nanomoles of Anc excreted in 24 h, shed light on the routes of PAnc decline. PAnc will decline due to the impact of chemical and physical factors38 and biochemically by human and bacterial enzyme-catalyzed processes.39,40 PAnc appeared to be rapidly deglycosylated by human and bacterial intra- and extracellular glycosidases41 because >94% of the total urinary Anc nmol were deglycosylated forms (Figure 4), including 51 of the 55 major Anc for 24−0 h (Table 4). Despite the preponderance of deglycosylated forms, 138 non-PAnc glycosides and glycoside conjugates were found among the minor 24−0h Anc (Table 5), indicating that anthocyanin glycosides can persist in vivo. Glucuronidation of Anc during first-pass metabolism is extensive42 and was apparent because glucuronide conjugates (including methylated forms) were greatest in number (149) (Table 5) and in total nanomoles (67% of total Anc) in the 24−0h urine, and in all other Anc pools examined (Figure 4; Tables 4 and 5). Anc glucuronidation is catalyzed by various uridine diphosphate glucuronate transferase (UGT) isoforms in the GIT and in the liver and kidneys during EHC. The substrate specificities of UGT isoforms vary,39 which can contribute diversity to the pool of glucuronides during enteric metabolism and EHC. The relative abundance of aglycone glucuronides suggests that they are more stable than simple aglycone forms, just as anthocyanidin glycosides are more stable than their aglycones. Removal of glucuronide groups
aglycone glucuronides (Table 5) contributed about 23.6% to the pool of minor Anc. Thirteen simple aglycones contributed 28% to the pool of major Anc (Table 4), whereas 53 simple aglycones (Table 5) contributed almost 20% to the nanomoles of minor Anc.
■
DISCUSSION Complexity of Anc Digestive Absorption. No Anc bioavailability study can detect all of the myriad products arising in vivo after Anc ingestion. When humans were fed 13Clabeled Cyn 3-glucoside, a multitude of labeled phenolic moieties and their phase II metabolites were detected in blood, urine, and especially feces, along with a small amount of 13C Cyn-based compounds.7 About 7% of the 13C dose was even recovered in breath as 13CO2. At 48 h after ingestion, the rate of cumulative fecal excretion of 13C was still unchanged since intake. At 48 h, only 44% of the 13C label had been accounted for. Results with 13C-labeled Anc feeding in humans illustrate the complex and extended elimination kinetics of Anc.7 Whereas Anc bioavailability research to date has often focused on the simple phenolics (C6−Cn) arising from Anc breakdown,1,5−8,28 the present study focused on the flavonoidlike (C6−C3−C6) metabolites of Anc. Our focus on flavonoidlike compounds was due in part to the recognized healthful bioactivities of flavonoids25,26 and was necessitated by the diverse Anc profile of wild blueberry23,24 (Figure 1). A large number of C6−C3−C6 anthocyanin metabolites were detected ex vivo when only Cyn 3-glucoside was administered to mice.19 The present study further informs because it examines Anc metabolism based on all six anthocyanidins. Considerations with MS/MS Analysis. Tracking AncM with a single parent/daughter mass transition (Table 1) has been used widely in Anc bioavailability research (for example, see refs 28−33). In the present study scanning up to 70 MRM in a single run provided sufficient data points for signal integration (data not shown). More predicted xenobiotic conjugates could have been scanned (e.g., aglycone sulfates and methyl aglycones were not scanned) (Table 1); thus, quantities of AncM are probably underestimated. MS/MS with ±0.5 amu resolution cannot provide definite compound identification. Urinary analytes that had the same unit mass MRM could have led to an overestimation of Anc per se. This issue was mitigated by scanning in positive ion mode, with which detection is best for Anc but not for other flavonoids. The flavonol quercetin has the same MS/MS as Del; however, it is not detected well in positive ion mode. Notably, wild blueberries have a very low flavonol content, which is limited mainly to quercetin products in nonedible plant parts.34 Pel and Pet aglycone were scanned only on the basis of their parent/parent mass (Table 1); this less specific detection could have led to overestimation of these two aglycones. It is notable that among the major 24−0 h Anc, aglycones of Pel (n = 3) and Pet (n = 9) were relatively abundant in numbers and in their nanomole contribution (Table 4). Multiple AncM Forms. Multiple forms of Anc have previously been reported ex vivo20,22 and in another paper arising from the present study.21 In the present study each of the 42 AncM MRM was detected, on average, at nine different RTs. Anc forms that shared the same MRM, but chromatographed differently, must have been chemically distinct, although the differences in their nature are not clear. Nine RTs for the Cyn aglycone glucuronide (463.0/287.0) ranged between about 3 and 25 min, reflecting a broad range in 1588
DOI: 10.1021/acs.jafc.6b05455 J. Agric. Food Chem. 2017, 65, 1582−1591
Article
Journal of Agricultural and Food Chemistry occurs by enteric β-glucuronidases and their isoforms in the liver and kidneys.39 Whereas almost 30% of wild blueberry PAnc are Mal glycosides23,24 Mal-based urinary Anc were in the lowest concentration (Figure 3) and in the least number of forms (Table 5) among the six anthocyanidins. Mal PAnc made up six of seven occurrences of major PAnc in 24−0 and 0 h (Table 4). These results support the notion that the B-ring dimethoxy moiety of Mal may confer steric hindrance to UGT isoforms during glucuronidation39 and possibly to other enzymes that transform Anc (e.g., glycosidases, glucuronidases, methyl transferases, sulfatases). Deglucuronidation and deglycosylation yielding anthocyanin aglycones will have implications in Anc bioavailability, in that flavonoid aglycones are more membrane soluble than their glucuronides or glycosides.39,43,44 Association of the abundant Anc aglycones with biological membranes could increase the Anc concentration at membrane surfaces and regions that provide a stable Anc matrix in situ.44 Anc are known to be widely distributed in animal tissues,45,46 which supports the notion of Anc association with membranes. Phase II methylation will reduce Anc polarity and enhance their membrane solubility.47 These aspects should be considered in relation to the purported low systemic Anc bioavailability in humans. Methylation of Anc, which is due at least in part to catecholO-methyl transferase,39,40 gives rise to methylated AncM. Phase II Anc methylation can produce ambiguous Anc identification in relation to native methylated PAnc forms (e.g., Peo and methyl Cyn). Similarly, catabolic removal of anthocyanin functional groups during GIT transit can effectively interconvert Anc (e.g., dehydroxylation of Cyn gives Pel) when detected by MS/MS. One significant example of Anc interconversion in vivo relates to Pel metabolism in vivo. Although wild blueberries do not contain Pel glycosides23,24 Pel-based urinary AncM were abundant in both their nanomole and number of forms (Figure 3; Tables 4 and 5). It is notable that Pel has the least number of B-ring functional groups so that catabolic demethoxylation and dehydroxylation by colonic bacteria48 will eventually give rise to Pel from other Anc detected by MS/MS. The catabolism of Mal, which is the most substituted Anc, into Pel would require two demethoxylation and two dehydroxylation steps. Therefore, as seen here, Pel-like products appear to occur in vivo even when Pel Anc are not in the food source. The high apparent bioavailability reported for Pel29,49,50 could be at least partially explained in this way. A recent study48 employed timeof-flight MS to identify major breakdown steps for isorhamnetin 3-O-glucoside by human intestinal bacteria in vitro. Major breakdown steps were deglycosylation, dehydroxylation, and demethoxylation, which supports the current results with respect to events during Anc catabolism. Estimate of Anc Concentration. The recovery in urine of 0.047% of the administered PAnc aligns well with other results.4 Recovery of an additional 0.92% of the dose as AncM raised the total recovery 20-fold to 0.97% (Table 2). Also, this recovery was likely an underestimate because some predicted AncM categories were not scanned (Table 1), and novel AncM whose MRM are unknown are predicted due to EHC.21 The recovery of 1% of the ingested Anc dose, which is likely a minimum, reflects an excretion of about 5 μmol in 24 h and with significant Anc retention apparent by the high 0 h Anc content (Table 2).
Persistence of Anc in vivo was previously documented when urinary AncM were very abundant even after a 5 day Anc-free run-in.21 In the present study persistence was also observed by the high estimated Anc concentration (2.91 μM) in the 0 h void when BJ had not been consumed for nearly 24 h, compared to the 24−0 h mean concentration of 1.47 μM (Table 2). The persistence and fluctuation in urinary Anc probably reflects Anc dissolution with bile phospholipids. Anc circulation with bile during EHC will involve Anc reconcentration with bile acids recovered from the GIT, its storage in the gall bladder, and release into the GIT in relation to food intake.51 Uptake and association of Anc with bile is relatively wellknown.1,31,35 It is a feature common to flavonoids and is related to the approximate MW of flavonoids.51 In contrast,
