Differences in Metabolism of Ellagitannins by ... - ACS Publications

Nov 23, 2016 - This is also the case for other ellagitannin-containing fruits and nuts including strawberry, raspberry, blackberry, walnuts, and musca...
1 downloads 8 Views 3MB Size
Article pubs.acs.org/jnp

Differences in Metabolism of Ellagitannins by Human Gut Microbiota ex Vivo Cultures Jakub P. Piwowarski,*,† Sebastian Granica,† Joanna Stefańska,‡ and Anna K. Kiss† †

Department of Pharmacognosy and Molecular Basis of Phytotherapy, Faculty of Pharmacy, Medical University of Warsaw, Banacha 1, 02-097 Warsaw, Poland ‡ Department of Pharmaceutical Microbiology, CePT, Medical University of Warsaw, Banacha 1B, 02-097 Warsaw, Poland S Supporting Information *

ABSTRACT: Ellagitannin-rich plant materials are used as popular remedies in the treatment of various inflammatory diseases. Urolithins are gut microbiota metabolites of ellagitannins and are considered responsible for in vivo health effects. Various natural products have been studied that are known sources of urolithins. However, few studies have focused on the metabolism of ellagitannin molecules. The aim of the study was to examine the metabolic fate of select ellagitannins using ex vivo cultures of human gut microbiota. Fifteen monomeric and dimeric ellagitannins, 1-O-galloyl-4,6(S)-HHDP-β-D-glucose (2), pedunculagin (3), potentillin (4), casuarictin (5), coriariin B (6), vescalagin (7), castalagin (8), stachyurin (9), casuarinin (10), stenophyllinin A (11), stenophyllanin A (12), salicarinin A (13), gemin A (14), agrimoniin (15), and oenothein B (16), and ellagic acid (1) were studied. The formation of the metabolites in ex vivo human microbiota cultures was monitored using UHPLC-DAD-MS/MS. Ellagitannins possessing hexahydroxydiphenoyl moieties were metabolized to 6Hdibenzo[b,d]pyran-6-one derivatives, i.e., urolithins. The observed differences in amounts of produced urolithins indicated that the individual microbiota composition and type of ingested ellagitannins could determine the rate of urolithin production. When the oral ingestion of natural products containing ellagitannins with hexahydroxydiphenoyl groups is considered, the formation of urolithins and their bioactivity should be addressed.

T

in ingested products and extracts, the bioavailability of urolithins has been well established in many animal and human studies. Urolithins have been shown to reach micromolar concentrations in feces, serum, tissues, and urine. On the basis of the results of in vivo and in vitro bioactivity studies, urolithins are known to inhibit inflammatory response and were recently shown to improve mitochondrial and muscle function;4 as such, these molecules are potentially considered as factors responsible for observed beneficial health effects of orally administered, ET-rich, natural products.5 The majority of gut microbiota metabolism studies have been conducted either using ET mixtures, e.g., food products and plant extracts, or using pure ellagic acid. Among pure ETs, only punicalagins and vescalagin have been directly shown to be substrates in human gut microbiota metabolism that result in urolithin production.3,6,7 Despite the structural variety of ETs and their presence in different natural products with postulated anti-inflammatory therapeutic properties, studies on the metabolism of single molecules have not yet been conducted. For the evaluation of the metabolism of pure ETs, representatives of different structural groups were selected

he gut microbiota is an important metabolic and immunological organ that plays a key role in whole organism homeostasis. A number of studies have investigated host/gut microbiota interactions. The mechanisms by which the microbiota affect gut health include immune signaling, toxin release, nutrient and xenobiotic metabolism, and modulation of mucosal barrier function and integrity.1 Since its recognition as an important player in human metabolism, the impact of gut microbiota on ingested natural products has become a subject of extensive studies. As a result, beneficial health effects of microbiota in clinical and interventional trials have been discovered. Among the most thoroughly studied natural products are those that contain ellagitannins (ETs). In vivo and ex vivo studies have shown that popular food products (e.g., pomegranates, walnuts, strawberries, raspberries, blackberries, cloudberries, oak-aged red wine, tea, and acorns) and medicinal plants (e.g., Filipendula ulmaria (L.) Maxim. herb, Geranium pratense L. herb, Geranium robertianum L. herb, Geum urbanum L. root and rhizome, Lythrum salicaria L. herb (Ph. Eur.), Potentilla anserina L. herb, Potentilla erecta (L.) Raeusch rhizome, Quercus robur L. bark, Rubus idaeus L. leaf, and Rubus f ruticosus L. leaf) rich in ETs provide compounds that are metabolized by gut microbiota to 6H-dibenzo[b,d]pyran-6-one derivatives, i.e., urolithins.2,3 In contrast to compounds present © XXXX American Chemical Society and American Society of Pharmacognosy

Received: June 30, 2016

A

DOI: 10.1021/acs.jnatprod.6b00602 J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products

Article

Chart 1. Chemical Structures of Tested ETs

widely distributed in various plant species of the Casuarinaceae, Stachyuraceae, Myrtaceae, Betulaceae, Fagaceae, Hamamelidaceae, Lythraceae, Melastomataceae, Rosaceae, Elaeagnaceae, Theaceae, and Juglandaceae.8 The complex ETs, represented by stenophyllinin A (11) and stenophyllanin A (12), have a flavan3-ol moiety linked to the C-1 of the acyclic glucosyl core through a C−C bond and occur in some species of the Fagaceae, Combretaceae, Myrtaceae, Theaceae, and Melastomataceae as well as in oak-aged wines.8 Salicarinin A (13) is a dimeric, C-glucosidic ET. Two dimeric ETs, gemin A (14) and

(Chart 1). 1-O-Galloyl-4,6-(S)-HHDP-β-D-glucose (2), pedunculagin (3), potentillin (4), casuarictin (5), and coriariin B (6) represent monomeric ETs with galloyl, hexahydroxydiphenoyl (HHDP), and/or dehydrodigalloyl (for 6) moieties attached to the glucopyranosyl moiety. This type of tannins is found in plants belonging to the families Betulaceae, Coriariaceae, Cornaceae, Fagaceae, Hamamelidaceae, Lecythidaceae, Lythraceae, Melastomataceae, Myrtaceae, Cornaceae, Onagraceae, Rosaceae, and Theaceae.8 The monomeric C-glucosidic ETs vescalagin 7, castalagin 8, stachyurin 9, and casuarinin 10 are B

DOI: 10.1021/acs.jnatprod.6b00602 J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products

Article

agrimoniin (15), contain cyclic glucose cores. Dimeric, macrocyclic ET oenothein B (16) consists of only valoneoyl and galloyl moieties. Dimeric ellagitannins have been found in many plant families including Fagaceae, Betulaceae, Rosaceae, Coriariaceae, Melastomataceae, Onagraceae, Lythraceae, Theaceae, Myrtaceae, Cornaceae, Combretaceae, and Euphorbiaceae.8 The aim of the present study is to provide direct evidence that human gut microbiota is capable of metabolizing pure ETs to urolithins and to evaluate whether the structural features of ETs influence their metabolic fate. Because pure urolithins are required to conduct in vitro studies on their biological activities, simple methods for their production and isolation are also presented.

Table 1. Presence of Urolithins in Human Gut Microbiota ex Vivo Cultures Incubated with Selected Pure ETs after 24/48 ha



RESULTS AND DISCUSSION Production and Isolation of Urolithins. To obtain an iso-urolithin A (iUA) standard, a biosynthesis using a gut microbiota culture of D3 (iUA producer) was performed. A simple two-step protocol of isolation using liquid−liquid extraction and preparative HPLC purification was developed. From 24 and 48 h cultures, 7.0 mg of iUA and 12.5 mg of urolithin B (UB) were obtained from 1600 mg of Lythrum salicaria aqueous extract. The UV, MS (Table 1), and NMR (Table 3) spectra confirmed the identity of iUA. The identity of UB was confirmed by comparing its retention time and UV and MS data with those of a standard compound obtained from chemical synthesis. This is the first time an alternative method of producing iUA has been developed as compared to earlier synthesis.9 iUA is not commercially available but is required to study the health effects of ETs. A significant proportion of the population (10−50%) has been identified as iUA producers.10 Additionally, no methods of production and isolation of urolithins using human gut microbiota cultures have been previously reported. UB, urolithin A (UA), iUA, and urolithin M7 (M7) were previously isolated from the feces of Trogopterus xanthipes.11 UA and UB were isolated from the urine of volunteers who had ingested pomegranate juice.12 Isolation of urolithins from urine of rats fed with geraniin and rat microbiota cultures was previously reported by Ito and coworkers.13 Later, Zhao and co-workers developed countercurrent chromatography methods for the isolation of UA and UB from rat microbiota cultures of pomegranate husk extracts.14 Our developed method is capable of producing significant quantities of pure urolithins. This allows for their bioactivity studies in vitro. Moreover, the production of urolithins can be scaled by increasing the number of cultures or by prolonging the incubation time to produce higher yields of less hydroxylated compounds like UB. The introduction of easy and inexpensive methods, i.e., using cultures of human microbiota, as alternatives to chemical syntheses of urolithins is of particular importance. Commercially available compounds are expensive, and traces of chemical catalysts, such as copper, may be present in the final product, which may potentially influence the bioassay results.15 Determination of Gut Microbiota Metabolism of ETs. The presence of urolithins M6 (M6), UC, M6 isomer (iM6), C isomer (iUC), M7 (M7), iso-urolithin A (iUA), UA, and UB in the gut microbiota cultures is summarized in Table 1. Detailed chromatograms are provided in the Supporting Information (Figures S1−S3). Quantification results of the main metabolites are provided in Figure 2. The identification of major gut microbiota metabolites, i.e., UA, iUA, UB, and UC, was based

a

The most abundant metabolite is in bold font. Detailed chromatograms are provided in Figures S1−S3, Supporting Information.

on their retention times and UV and MS data, which were compared with those of respective standard substances (Table 2 and Figure 1). The identities of other metabolites were determined by UV and MS data and characteristic retention pattern comparisons with available literature data.5,16 Their quantification (Table S4, Supporting Information) was done based on relative response factors (RRFs) toward UA used in the quantification method for the urolithins developed by ́ Garcia-Villalba and co-workers.17 Parent ETs were not detected in culture media at any time. No urolithins were detected in samples incubated without ETs. C

DOI: 10.1021/acs.jnatprod.6b00602 J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products

Article

Table 2. Retention Times, UV Spectra, and MS Data for Metabolites

Tomas-Barberan and co-workers,10 volunteers D1 and D3 were

The microbiotas of all tested volunteers after 24 h were capable of producing urolithins from selected pure ETs. On the basis of the metabotype classification system introduced by

classified as metabotype B (producing iUA, UA, and UB), D

DOI: 10.1021/acs.jnatprod.6b00602 J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products

Article

exceed 4 μM after 48 h. For the monomeric ETs with glucopyranosyl cores, the amount of UA ranged from 7.0 to 11.5 μM after 24 h. Only in the dimeric ETs 14 and 15 were traces of M6 detected. D3 and D1 were assigned to metabotype B. However, unlike D1, D3 produced iUA as a major compound, and small amounts of UA were observed. The gut microbiota of D3 was capable of producing two urolithins possessing four hydroxy groups, i.e., M6 and its isomer iM6, and two urolithins with three hydroxy groups, i.e., UC and iUC, after 24 h. M6 and UC were no longer detected in the culture media after 48 h, whereas their isomers iM6 and iUC were still detectable. D3, classified as metabotype B, did not produce M7 like D1. Instead, D3 produced iM6 and iUC, which were not observed in the samples originating from D1 and D2. Compounds with these characteristic UV spectra (Table 2) and molecular weights (244 and 260 amu), which indicated three and four hydroxy groups, respectively, have not been previously identified in any studies on microbiota metabolism of ETs. The hypsochromic shift of band I and increased absorption in band II (approximately 258 nm) suggest the presence of a C-9 hydroxy group.16,18 Nevertheless, to accurately determine the hydroxylation pattern, the isolation of these compounds was required. However, this is not achievable using current methods wherein L. salicaria extract is the ET source. Here, ETs were not present in the culture media or in the Et2O or EtOAc fractions. The high complexity of ET metabolites for D1 and D3 (metabotype B) in comparison with D2 (metabotype A) agreed with previous in vivo observations and metabotype characterizations introduced by Tomas-Barberan and coworkers.10 Among the tested ETs, those possessing an HHDP moiety were metabolized by human gut microbiota to urolithins. The ability of gut microbiota to directly metabolize the native ET molecules indicated that the initial steps of their hydrolyses in the upper compartments of the gastrointestinal tract were not necessary for the ETs to become substrates 1 for urolithin production by the gut microbiota. All ETs and EA were tested at an equimolar concentration (250 μM), which was set based on the limits of urolithin detection and their quantification in culture media. However, no correlation between the number of HHDP groups in the ET molecule and the amounts of produced urolithins was observed. The incubation of dimeric ETs 14 and 15, despite providing larger amounts of ellagic acid equivalents, did not result in higher urolithin production rates (Figure 1). Surprisingly, urolithin production dynamics from 1, being the final product of ET hydrolysis, was relatively low when compared with those of the ETs. In all cultures of pure compound 1 traces remained in the culture media after 24 h. The production rates of urolithins from monomeric ETs with glucopyranosyl cores were generally higher than those for dimeric compounds for all three donors. The lowest urolithin production rate was observed for 13, a dimeric C-glucosidic ET possessing two HHDP groups. The parallel examination of pairs of isomeric compounds with different orientations at the anomeric carbon of the glucose core (4 and 5; 7 and 8; and 9 and 10) showed that the patterns of produced urolithins were very similar in each pair. This indicated that this structural feature did not significantly influence ET metabolism. Furthermore, compounds with closely related structures, i.e., complex C-glucosidic ETs with an attached flavan-3-ol moiety as in 11 and 12 and dimeric ETs 14 and 15, were metabolized in comparable ways by the microbiota of each volunteer.

Table 3. NMR Spectroscopic Data (300 MHz, DMSO-d6) for Iso-urolithin A (iUA) position 1 2 3 4 4a 6 6a 7 8 9 10 10a 10b

δC, type 124.6, 113.0, 159.8, 102.9, 152.4, 160.3, 110.5, 132.4, 116.5, 163.9, 106.1, 137.4, 109.3,

CH CH C CH C C C CH CH C CH C C

δH (J in Hz) 7.97, d (8.8) 6.81, dd (8.8, 2.1) 6.70, d (2.2)

8.03, d (8.7) 6.97, dd (8.7, 1.8) 7.43, d (1.8)

Figure 1. Chemical structures of identified gut microbiota metabolites.

whereas volunteer D2 was identified as a typical metabotype A (producing main metabolite UA but not iUA or UB). The main metabolite present in the D1 cultures was UA. Minor amounts of iUA were also detected for some of the tested ETs. After 24 h, the D1 cultures produced M6, UC, M7, iUA, UA, and UB. After 48 h of incubation, due to further deoxygenation, M6 and UC were no longer present in the culture media, and the amounts of UA and UB significantly increased. Interestingly, the levels of trihydroxylated M7, i.e., a UC isomer, did not significantly change during the incubation time and were stable after 24 and 48 h. iUA, which was initially produced in minor amounts after 24 h, was no longer present after 48 h, with a couple of exceptions. The biotransformations of two isomeric compounds, 9 and 10, resulted in increased iUA concentrations over time. The D2 microbiota belonged to metabotype A, with the main metabolites being UC and UA. No UB production was observed. For non-C-glucosidic ETs, UC production dominated all other urolithins after 24 h of incubation with a small accompanying UA peak. An additional 24 h incubation resulted in UC deoxygenation and increased amounts of UA. Significantly lower yields of urolithins were produced from Cglucosidic ETs. For this group of ETs, traces of 1 were still detected in cultures with small amounts of UA and UC after 24 h. For the C-glucosidic ETs the concentration of UA did not E

DOI: 10.1021/acs.jnatprod.6b00602 J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products

Article

Figure 2. Results of quantitative analysis of the major metabolites in gut microbiota cultures. The exact urolithin amounts determined in cultures are provided in Table S1, Supporting Information. The estimated amounts of remaining metabolites are provided in Table S4, Supporting Information.

No urolithin production was observed for the dimeric macrocyclic ET (16), which possesses only valoneoyl and galloyl moieties. This outcome suggested that gut microbiota is unable to cleave the ether bond connecting HHDP with the galloyl group in the valoneoyl moiety or produces metabolites that were not detectable by the applied methods. Alternatively, 16 could have simply inhibited the growth of the urolithinproducing microbiota. However, in additional studies using a 5fold lower concentration of 16 in gut microbiota cultures followed by Et2O extraction and concentration, no urolithin production was observed (data not shown). The previous studies conducted with Epilobium hirsutum aqueous extract, in which 16 is a dominating constituent, suggested that urolithins could be produced in the presence of 16, but were more likely sourced from other minor ETs and ellagic acid present in the extract.19 Nevertheless, the metabolism of 16 still requires elucidation, as neither 16 nor valoneic acid dilactone (the product of cleaving the valoneoyl moiety from the glucose core) were found in the culture media. Observed differences in the amounts of produced urolithins, independent of the amounts of ellagic acid equivalents provided to the culture, indicate that the individual microbiota composition and type of ingested ETs could have determined the rate of urolithin production. These observations were

consistent with the results obtained by Cerda and co-workers, who noted that the amount of excreted urolithins was not proportional to the amount of consumed ETs while testing the in vivo metabolism of different ET-rich food products, e.g., strawberries, raspberries, walnuts, and oak-aged red wine.20 Gonzalez-Sarrias and co-workers observed that significantly more urolithins were detected in prostate samples from patients who consumed walnuts than in patients who consumed pomegranate juice, despite the latter possessing more ETrelated compounds.21 It is well established that the HHDP group detaches from ET glucose cores and is further metabolized to urolithins. However, not much is known about the metabolic fate of other moieties present in ET structures. To detect a wider range of potential metabolites, a UHPLC gradient applied in qualitative examinations was used to separate and detect compounds within a wide range of polarities. For this reason, sample preparations without exclusionary purification steps were used, e.g., SPE or liquid−liquid extractions. Two unknown compounds, N1 and N2 at m/z 309 and 367, respectively, were detected from 6 in the cultures of D1 and D2 (Figures S1 and S2, Supporting Information). However, whether these were products derived from 6 or produced by the microbiota under the influence of 6 remains unclear. UV and MS (total MS and F

DOI: 10.1021/acs.jnatprod.6b00602 J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products

Article

among the study volunteers agreed with previously conducted in vivo and ex vivo studies.6,10,12,20,29,30 The characteristic and independence of the metabolic fate of the HHDP groups in ET molecules of each group of ETs suggests that the ET molecules not only provided HHDP groups as substrates for microbiota metabolism but also, depending on their structure and microbiota metabotype, influenced the growth and/or metabolism among the microbiota ecosystem in the urolithin metabolic pathways. The complexity of these interactions could increase in the case of ET mixtures, which are found in medicinal plant extracts and food products. The resultant production of ET metabolites was difficult to predict accurately. However, if the ET compositions of a certain natural product were known, the formation of gut microbiota metabolites can be assessed. When natural products containing ETs are ingested, the formation of bioavailable gut microbiota metabolites has to be taken into consideration, and their bioactivity needs to be addressed.

extracted ion chromatograms of hypothetical degradation products) chromatograms of all other ET cultures displayed peaks representing compounds other than those associated with the urolithin metabolic pathway. The expected characteristic products of ET degradation were, for the C-glucosidic 9, 10, and 13, ellagic acid C-glycosides formed from the hexahydroxybiphenoyl moiety; for 7, 8, and 13, vescalin, castalin, or related compounds formed from the nonahydroxyterphenoyl moiety; for 11 and 12, catechin or its metabolites; for 13 and 16, the valoneic acid dilactone; and for 14 and 15, metabolites of the linking dehydrodigalloyl moiety. The lack of metabolites other than urolithins could be caused by different factors. The urolithins were readily detected in cultures due to their characteristic UV spectra and high absorbance/concentration ratios. Other groups of metabolites, even if produced, were likely not observed due to low concentrations below their limits of detection or the applied methods were unable to even detect molecules such as short-chain fatty acids. Furthermore, the protein binding activity of hypothetical polyphenolic metabolites could have bound the undetected metabolites to the protein-rich BHI medium, and thus, highly hydroxylated phenolic compounds could have precipitated and been filtered away during sample filtration. This could also be a reason that the parent ETs were not detectable at any time, as was noted during studies conducted on punicalagin.6 In the bile extracts of Iberian pigs fed with acorns, proposed metabolites of tergallic acid, a valoneic acid dilactone isomer, were detected. This trihydroxytergallagic degradation metabolite with aglycone at m/z 361 was conjugated with glucuronic acid.22 In the present study, no compound with tergalloyl moiety was tested. However, in cultures of ETs with an isomeric valoneoyl moiety, no corresponding metabolite was found. The issue of undetectable metabolite formation (i.e., CO2) was previously considered by Cerda and co-workers, who reported that only 3−6% of punicalagin ingested by rats was detected as such or as its microbiota metabolites in urine and feces.23 The majority of previous in vivo and ex vivo studies regarding microbiota metabolism of ETs were conducted for mixtures of ETs present in either food products2 or ET-rich medicinal plant extracts.3 Direct evidence of gut microbiota metabolism of single ETs has primarily been evaluated using animal models. Geraniin, corilagin, punicalagins, and ellagic acid were shown to be urolithin sources using in vivo rat models,23−25 while geraniin metabolism was confirmed using ex vivo cultures of rat microbiota.24 No in vivo human trials have been conducted for pure ETs. Direct proof of single ET metabolism to urolithins by human gut microbiota was conducted ex vivo only for 1, 7, and punicalagins.3,6,7,26 Previous studies conducted on 1 indicated two species of bacteria, Gordonibacter urolithinfaciens and G. pamelaeae DSM 19378, being able to sequentially metabolize 1 to urolithins M5, M6, and UC.27,28 The species of microbiota metabolizing UA and UB and those responsible for alternative modes of deoxygenation remain unidentified. For the first time, a complex study regarding human gut microbiota metabolism of different ETs was conducted, providing a direct proof of their biotransformation to urolithins. Therefore, the peroral use of ET-rich product-containing compounds included in the present study is likely to result in the production of a series of bioavailable gut microbiota metabolites, i.e., urolithins. The observed strong interindividual differences in urolithin concentrations and production dynamics due to the variability in the gut microbiota composition



EXPERIMENTAL SECTION

General Experimental Procedures. Human fecal samples were donated by three healthy volunteers (D1−D3) aged 25 to 40 without a history of gastrointestinal disease. Donors had not used antibiotics in the six months before sample collection. The study complied with the Helsinki Declaration. The intake of ellagitannin-containing products was strictly forbidden for 1 week before sample collection. Samples were processed within 30 min from defecation. The growth medium, brain heart infusion (BHI) (DIFCO, Detroit, MI, USA), was prepared according to the manufacturer’s instructions. To acquire anaerobic conditions, BHI was boiled for 20 min and immediately cooled before experiment. Fecal slurries (FS) were prepared by suspending human feces in BHI (1:10 w/v). Solutions of each ellagitannin (10 mM) were prepared in deionized water and sterilized by filtration through 0.2 mm Ophtalsart hydrophilic syringe filters (Sartorius Stedim Biotech GmbH, Göttingen, Germany). Then, 1 mL of FS and 0.25 mL of compound solution were added to 8.75 mL of BHI. Subsequently, 0.25 mL of deionized water was added to the control blank sample. The batch cultures were incubated in sealed containers under anaerobic conditions using GENbox Anaer sachets (bioMerieux, France) at 37 °C. Then, 0.75 mL of batch culture was collected at t = 0, 24, and 48 h, mixed with an equal volume of MeOH subjected to an ultrasonic bath for 5 min, and filtered through a 0.45 μm Chromafil polyester membrane (Macherey-Nagel, Duren, Germany). The samples were stored in the dark at −20 °C until injection to the UHPLC system. Isolation and Structure Determination of Urolithins. Briefly, 10 mL of FS from D3 and 5 mL of Lythrum salicaria L. aqueous extract, prepared as previously described31 (40 mg/mL), were added to 85 mL of BHI. Four cultures were prepared and incubated in a sealed container under anaerobic conditions in 37 °C. After 24 h, cultures were combined and subsequently extracted with Et2O (4 × 0.5 L) and ethyl EtOAc (4 × 0.5 L) and evaporated to dryness at 40 °C. The Et2O fraction (250 mg) was dissolved in 1.5 mL of DMSO and subjected to preparative HPLC, 16 runs, 100 μL injection each time, using a Shimadzu LC-20AP instrument and a Zorbax SB-C18 (21.2 mm × 150 mm × 5 μm) column, eluted with mobile phase A (0.1% HCOOH in H2O) and mobile phase B (0.1% HCOOH in MeCN), flow rate 9.00 mL/min using the gradient 0−40 min, 20− 45% B, oven temperature 40 °C, and detection wavelengths 254 and 350 nm, to yield pure compounds iUA (6.1 mg, tr = 16.8 min) and UB (0.8 mg, tr = 29.6 min). 1H and 13C NMR spectra were recorded at 25 °C on a Varian NMR instrument (300 MHz for 1H and 75 MHz for 13 C NMR). DMSO-d6 was used as a solvent. Chemical shifts (δ) are reported in ppm, and coupling constants (J) are reported in Hz. The assignments were based on HSQC and HMBC experiments. The same method was used to obtain pure UB, but the incubation time was extended to 48 h. The obtained Et2O fraction (600 mg) was subjected G

DOI: 10.1021/acs.jnatprod.6b00602 J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products

Article

to preparative HPLC, yielding pure compounds iUA (0.9 mg, tr = 16.8 min) and UB (11.7 mg, tr = 29.6 min). Standard Solution Preparation. Accurately weighed UA, iUA, and UB were dissolved in DMSO to obtain 20 μM stock solutions. Serial dilutions of the stock solution with H2O/MeOH (7:3, v/v) yielded calibration standards at 0.133, 0.334, 0.667, 1.000, 1.334, 2.667, 5.334, and 8.004 μM. After preparation, each solution was stored in the dark at 5 °C for a maximum of 72 h. Then, 5 μL of each solution was injected into the UHPLC system, 0.667−48.024 pmol per injection of standard. UHPLC-DAD Analysis. Quantitative analyses were performed using a Kinetex XB-C18 column (150 mm × 3 mm × 2.6 μm) (Phenomenex, Torrance, CA, USA). The mobile phase consisted of 0.1% HCOOH in H2O (A) and 0.1% HCOOH in MeCN (99.9:0.1; v/v) (B). The flow rate was set at 0.400 mL/min. The column was eluted as follows: 0−20 min 20−40% B, 20−22 min 100% B. The column was equilibrated for 7 min between injections. Column temperature was maintained at 40 °C. The UV spectra were recorded over a 200−400 nm range, and chromatograms were acquired at 305 nm. Method Validation. Standard solutions of UA, iUA, and UB at eight concentration levels were analyzed in triplicate. Calibration curves were generated using linear regression on the plots of peak areas for each standard versus the amount injected to the UHPLC column. The F-test was applied to determine the significance of the regression equation for the linear model at the 99% confidence level. The calculations were performed using the REGLINP function in MS Excel 2016. Limits of detection (LOD) and quantification (LOQ) for the used standards were calculated using the equations LOD = 3.3SD/ a and LOQ = 10SD/a, respectively, where SD is the standard deviation of the response (y-intercept) and a is the slope of the calibration curve according to ICH guidelines.32 Repeatability and intermediate precision were performed using standard mixtures at eight levels from 0.667 to 48.024 pmol per injection. Both parameters were calculated as relative standard deviations of multiple determinations. For repeatability, one injection from six independent samples was performed. Intermediate precision was investigated using three samples prepared according to the same protocol and analyzed within 24 h, interday assay, or over different days, intraday assay. Validation results are provided in the Supporting Information (Tables S2 and S3). Qualitative Analysis. UHPLC-DAD-MSn analyses were performed using a UHPLC-3000 RS system (Dionex, Sunnyvale, CA, USA) with DAD detection and an AmaZon SL mass spectrometer with ESI interface (Bruker Daltonik GmbH, Bremen, Germany). The column was Zorbax SB-C18 (150 mm × 2.1 mm × 1.9 μm) (Agilent, Santa Clara, CA, USA). The mobile phase consisted of 0.1% HCOOH in H2O (A) and 0.1% HCOOH in MeCN (B). The gradient was 0−5 min 0% B, 5−15 min 0−10% B, 15−25 min 10−20% B, 25−35 min 20−30% B, 35−45 min 30−50% B, 45−50 min 50−100% B, and 50− 60 min 100% B. The column temperature was maintained at 25 °C, and the flow rate was 0.200 mL/min. The LC eluate was introduced into the ESI interface without splitting, and compounds were analyzed in the negative and positive ion modes with the following settings: nebulizer pressure of 40 psi; drying gas flow rate of 9 L/min; nitrogen gas temperature of 300 °C; and a capillary voltage of 4.5 kV. The mass scan ranged from 100 to 2200 m/z. UV spectra were recorded in the range of 200−400 nm. The presence of UA, iUA, UB, and UC was confirmed by comparing retention times and UV and MS data with authentic samples. Other compounds were tentatively assigned based on their chromatographic properties, UV, and MS data. Reagents and Materials. Chromatographic grade MeCN was purchased from Merck (Darmstadt, Germany). Water was purified with a Millipore Simplicity System (Bedford, MA, USA). Formic acid and MeOH were purchased from POCh (Gliwice, Poland). UA, UB, and UC were synthesized according to Bialonska and co-workers.33 Ellagitannins used in the study were isolated in the Department of Pharmacognosy and Molecular Basis of Phytotherapy, Warsaw, Poland. 1-O-Galloyl-4,6-(S)-HHDP-β-D-glucose (2), pedunculagin (3), potentillin (4), casuarictin (5), coriariin B (6), stachyurin (9),

casuarinin (10), stenophyllinin A (11), stenophyllanin A (12), and gemin A (14) were isolated from the roots of Geum urbanum L.34,35 Vescalagin (7), castalagin (8), and salicarinin A (13) were isolated from the aerial parts of Lythrum salicaria L.31 Agrimoniin (15) was obtained from the aerial parts of Agrimonia eupatoria L.36 Oenothein B (16) was isolated from aerial parts of Oenothera hoelsheri Renner ex Rostanski.37 The purities of the examined compounds (≥95%) were confirmed using a UHPLC-DAD-MSn method; their identities were confirmed by NMR and MS spectra. Ellagic acid (1) was purchased from Carl Roth (Karlsruhe, Germany).



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jnatprod.6b00602. Additional information (PDF)



AUTHOR INFORMATION

Corresponding Author

*Tel/fax: +48 22 572 09 85. E-mail: jakub.piwowarski@wum. edu.pl. ORCID

Jakub P. Piwowarski: 0000-0002-5011-0983 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The project was financially supported by a Polish Ministry of Science and Higher Education research grant, Iuventus Plus (IP2015 062274). J.P.P. was financially supported by a Foundation for Polish Science START scholarship (START 84.2016). The project was carried out with the use of CePT infrastructure financed by the European Regional Development Fund within the Operational Programme “Innovative Economy” for 2007−2013.



REFERENCES

(1) Tuohy, K.; Del Rio, D. Diet-Microbe Interactions in the Gut. Effects on Human Health and Disease; Academic Press: London, 2015. (2) Espin, J. C.; Larrosa, M.; Garcia-Conesa, M. T.; Tomas-Barberan, F. Evid. Based Complement. Alternat. Med. 2013, 2013, 270418. (3) Piwowarski, J. P.; Granica, S.; Zwierzynska, M.; Stefanska, J.; Schopohl, P.; Melzig, M. F.; Kiss, A. K. J. Ethnopharmacol. 2014, 155, 801−809. (4) Ryu, D.; Mouchiroud, L.; Andreux, P. A.; Katsyuba, E.; Moullan, N.; Nicolet-Dit-Felix, A. A.; Williams, E. G.; Jha, P.; Lo Sasso, G.; Huzard, D.; Aebischer, P.; Sandi, C.; Rinsch, C.; Auwerx, J. Nat. Med. 2016, 22, 879−888. (5) Tomas-Barberan, F. A.; Gonzalez-Sarrias, A.; Garcia-Villalba, R.; Nunez-Sanchez, M. A.; Selma, M. V.; Garcia-Conesa, M. T.; Espin, J. C. Mol. Nutr. Food Res. 2016, DOI: 10.1002/mnfr.201500901. (6) Cerda, B.; Periago, P.; Espin, J. C.; Tomas-Barberan, F. A. J. Agric. Food Chem. 2005, 53, 5571−5576. (7) Bialonska, D.; Ramnani, P.; Kasimsetty, S. G.; Muntha, K. R.; Gibson, G. R.; Ferreira, D. Int. J. Food Microbiol. 2010, 140, 175−182. (8) Quideau, S. Chemistry and Biology of Ellagitannins. An Underestimated Class of Bioactive Plant Polyphenols; World Scientific Publishing Co. Pte. Ltd.: Singapore, 2009. (9) Pandey, J.; Jha, A. K.; Hajela, K. Bioorg. Med. Chem. 2004, 12, 2239−2249. (10) Tomas-Barberan, F. A.; Garcia-Villalba, R.; Gonzalez-Sarrias, A.; Selma, M. V.; Espin, J. C. J. Agric. Food Chem. 2014, 62, 6535−6538. (11) Jeong, S. J.; Kim, N. Y.; Kim, D. H.; Kang, T. H.; Ahn, N. H.; Miyamoto, T.; Higuchi, R.; Kim, Y. C. Planta Med. 2000, 66, 76−77. H

DOI: 10.1021/acs.jnatprod.6b00602 J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products

Article

(12) Cerda, B.; Espin, J. C.; Parra, S.; Martinez, P.; Tomas-Barberan, F. A. Eur. J. Nutr. 2004, 43, 205−220. (13) Ito, H. Planta Med. 2011, 77, 1110−1115. (14) Zhao, W.; Wang, Y.; Hao, W.; Yang, H.; Song, X.; Zhao, M.; Peng, S. J. Chromatogr. B: Anal. Technol. Biomed. Life Sci. 2015, 990, 111−117. (15) Tomás-Barberan, F. A.; Garcia-Conesa, M. T.; Larrosa, M.; González-Barrio, R.; Bermúdez-Soto, M. J.; González-Sarrias, A.; Espin, J. C. In Recent Advances in Polyphenol Research; Daayf, F.; Lattanzio, V., Eds.; Blackwell Publishing Ltd: Chichester, UK, 2008; Vol. 1, Chapter 11, pp 263−277. (16) Gonzalez-Barrio, R.; Truchado, P.; Ito, H.; Espin, J. C.; TomasBarberan, F. A. J. Agric. Food Chem. 2011, 59, 1152−1162. (17) Garcia-Villalba, R.; Espin, J. C.; Tomas-Barberan, F. A. J. Chromatogr. A 2016, 1428, 162−175. (18) Garcia-Villalba, R.; Beltran, D.; Espin, J. C.; Selma, M. V.; Tomas-Barberan, F. A. J. Agric. Food Chem. 2013, 61, 8797−8806. (19) Stolarczyk, M.; Piwowarski, J. P.; Granica, S.; Stefanska, J.; Naruszewicz, M.; Kiss, A. K. Phytother. Res. 2013, 27, 1842−1848. (20) Cerda, B.; Tomas-Barberan, F. A.; Espin, J. C. J. Agric. Food Chem. 2005, 53, 227−235. (21) Gonzalez-Sarrias, A.; Gimenez-Bastida, J. A.; Garcia-Conesa, M. T.; Gomez-Sanchez, M. B.; Garcia-Talavera, N. V.; Gil-Izquierdo, A.; Sanchez-Alvarez, C.; Fontana-Compiano, L. O.; Morga-Egea, J. P.; Pastor-Quirante, F. A.; Martinez-Diaz, F.; Tomas-Barberan, F. A.; Espin, J. C. Mol. Nutr. Food Res. 2010, 54, 311−322. (22) Espin, J. C.; Gonzalez-Barrio, R.; Cerda, B.; Lopez-Bote, C.; Rey, A. I.; Tomas-Barberan, F. A. J. Agric. Food Chem. 2007, 55, 10476−10485. (23) Cerda, B.; Llorach, R.; Ceron, J. J.; Espin, J. C.; TomasBarberan, F. A. Eur. J. Nutr. 2003, 42, 18−28. (24) Ito, H.; Iguchi, A.; Hatano, T. J. Agric. Food Chem. 2008, 56, 393−400. (25) Doyle, B.; Griffiths, L. A. Xenobiotica 1980, 10, 247−56. (26) Gonzalez-Barrio, R.; Edwards, C. A.; Crozier, A. Drug Metab. Dispos. 2011, 39, 1680−1688. (27) Selma, M. V.; Beltran, D.; Garcia-Villalba, R.; Espin, J. C.; Tomas-Barberan, F. A. Food Funct. 2014, 5, 1779−1784. (28) Selma, M. V.; Tomas-Barberan, F. A.; Beltran, D.; GarciaVillalba, R.; Espin, J. C. Int. J. Syst. Evol. Microbiol. 2014, 64, 2346− 2352. (29) Gonzalez-Barrio, R.; Borges, G.; Mullen, W.; Crozier, A. J. Agric. Food Chem. 2010, 58, 3933−3939. (30) Nunez-Sanchez, M. A.; Garcia-Villalba, R.; Monedero-Saiz, T.; Garcia-Talavera, N. V.; Gomez-Sanchez, M. B.; Sanchez-Alvarez, C.; Garcia-Albert, A. M.; Rodriguez-Gil, F. J.; Ruiz-Marin, M.; PastorQuirante, F. A.; Martinez-Diaz, F.; Yanez-Gascon, M. J.; GonzalezSarrias, A.; Tomas-Barberan, F. A.; Espin, J. C. Mol. Nutr. Food Res. 2014, 58, 1199−1211. (31) Piwowarski, J. P.; Kiss, A. K. Phytochem. Anal. 2013, 24, 336− 348. (32) ICH. 2005. http://www.ich.org/fileadmin/Public_Web_Site/ ICH_Products/Guidelines/Quality/Q2_R1/Step4/Q2_R1__ Guideline.pdf, accessed on June 23, 2016. (33) Bialonska, D.; Kasimsetty, S. G.; Khan, S. I.; Ferreira, D. J. Agric. Food Chem. 2009, 57, 10181−10186. (34) Piwowarski, J. P.; Granica, S.; Kosinski, M.; Kiss, A. K. Biochem. Syst. Ecol. 2014, 53, 46−50. (35) Granica, S.; Klebowska, A.; Kosinski, M.; Piwowarski, J. P.; Dudek, M. K.; Kazmierski, S.; Kiss, A. K. J. Ethnopharmacol. 2016, 188, 1−12. (36) Granica, S.; Krupa, K.; Klebowska, A.; Kiss, A. K. J. Pharm. Biomed. Anal. 2013, 86, 112−122. (37) Granica, S.; Kiss, A. K. Biochem. Syst. Ecol. 2012, 44, 44−47.

I

DOI: 10.1021/acs.jnatprod.6b00602 J. Nat. Prod. XXXX, XXX, XXX−XXX