Black Tea Theaflavin Detoxifies Metabolic Toxins in the Intestinal Tract of Mice

Abstract

Scope:

The goal of this study was to determine the in vivo efficacy of theaflavin (TF), one of the major bioactive components in black tea, to detoxify two major metabolic toxins, ammonia and methylglyoxal (MGO), in mice.

Methods and results:

Under in vitro conditions, TF was able to react with ammonia, methylglyoxal (MGO), and hydrogen peroxide (H₂O₂) to produce its aminated, MGO conjugated, and oxidized products, respectively. In TF-treated mice, the aminated TF, the MGO conjugates of TF and aminated TF, and the oxidized TF were searched using LC-MS/MS. Our results provided the first in vivo evidence that the unabsorbed TF was able to trap ammonia to form the aminated TF; furthermore, both TF and the aminated TF had the capacity to trap MGO to generate the corresponding mono-MGO conjugates. Moreover, TF was oxidized to dehydrotheaflavin, which underwent further amination in the gut. By exposing TF to germ-free (GF) mice and conventionalized mice (GF mice colonized with specific-pathogen-free microbiota), the gut microbiota was demonstrated to facilitate the amination and MGO conjugation of TF.

Conclusion:

TF has the capacity to remove the endogenous metabolic toxins through oxidation, amination, and MGO conjugation in the intestinal tract, which could potentially explain why TF still generates in vivo efficacy while showing a poor systematic bioavailability.

Graphical Abstract

This study reported that theaflavin, one of the major bioactive components in black tea, has the capacity to remove the endogenous metabolic toxins, such as ammonia and methylglyoxal, through oxidation, amination, and methylglyoxal conjugation in the intestinal tract. By exposing theaflavin to germ-free mice and conventionalized mice, the gut microbiota was demonstrated to facilitate the amination and methylglyoxal conjugation of theaflavin.

1. Introduction

Black tea, a type of fermented tea, is one of the most popular type of teas in western countries. During fermentation, two major enzymes, polyphenol oxidase (PPO) and peroxidase (POD), oxidize and polymerize the major components in tea leaves, catechins, into two types of the major color components in black tea, theaflavins (TFs) and thearubigins (TRs).^[2] Studies, especially those from in vitro assays, have shown that TFs exhibit anti-microbial, hypolipidemic, anti-inflammatory, anti-mutagenicity, and anti-cancer activities.^[1] However, TFs have a poor systematic bioavailability due to their large structural size, larger than catechins, after undergoing oxidation and polymerization.^[2] Chen et al. ^{[4, 5]} reported that the gallate group in theaflavin-3-mono gallate (TF3G), theaflavin-3’-mono gallate (TF3’G), and theaflavin-3,3’-di gallate (TFDG) was released by gut microbiota to form TF. It is still unknown how TF is metabolized and especially how the large amount of unmetabolized TF functions in the intestinal tract.

Many small molecular metabolites have been reported as endogenous toxins, which are the potential causal factors in certain diseases such as neurological disorders and diabetic complications.^{[6, 7]} Among them, ammonia and methylglyoxal (MGO) are the two major endogenous toxins. Ammonia is mainly produced in the intestine by phosphate-activated glutaminase and bacteria-related urease, and is primarily a waste product and is deleterious in high concentrations.^[2] MGO is a toxic metabolite produced through enzymatic and nonenzymatic reactions in mammalian cells.[3] Furthermore, it is the most important precursor of the advanced glycation end products (AGEs), and is associated with several pathologies, including diabetes, aging, and neurodegenerative diseases.^{[5, 4, 6]} Therefore, being able to scavenge these deleterious metabolites is believed to provide beneficial health effects.^[2] Previous studies have demonstrated the ability of TF to efficiently trap MGO under in vitro conditions.^[12] Recently, we reported that dietary flavonoids were able to scavenge ammonia in mice and humans. Since TF is poorly absorbed and the majority of TF remains in the intestinal tract, it is important to study how TF interacts with ammonia and MGO, especially in the intestinal tract, and how gut microbiota impact its effects. The aim of this study is to investigate the formation of the aminated and/or MGO conjugated TF and its metabolites in mice and the impact of gut microbiota on the formation of these novel TF metabolites.

2. Experimental Section

2.1. Materials

TF was synthesized in house from the reaction between epicatechin (EC) and epigallocatechin (EGC) in the presence of POD and hydrogen peroxide (H₂O₂) according to our previous procedure.^[2] Ammonium hydroxide solution, H₂O₂ (35%), MGO solution (40% in water), and type II POD from horseradish were purchased from Sigma (St. Louis, MO). Deuterated methanol (CD₃OD) solvent was purchased from Acros Organics (now Thermo Fisher Scientific, Waltham, MA). Other common chemicals and solvents were purchased from VWR International (Radnor, PA). LC-MS-grade solvents were obtained from Thermo Fisher Scientific (Waltham, MA).

2.2. Animal experiments and dosage information/dosage regimen

2.2.1. Metabolism of TF in CD-1 mice

The experiment with CD-1 mice was conducted according to a protocol approved by the Institutional Animal Care and Use Committee of the North Carolina Research Campus (IACUC No.16–016). Five-week old female CD-1 mice were purchased from Charles River Laboratories (Wilmington, MA) and acclimated for at least 1 week before being randomly assigned to different experimental groups in air-conditioned quarters with a room temperature of 20 ± 2 °C, relative humidity of 50 ± 10%, and a light:dark cycle of 12:12 h (7 am to 7 pm). The mice were allowed free access to water and normal chow diet.

Two groups of mice (5 mice/group) were washed out with AIN-93G diet for three days. After fasting overnight, one group of mice (n=5) were administered TF by oral gavage at the dose of 200 mg/kg body weight.^{[4, 5]} The other group of mice (n=5) were treated with the same volume vehicle as the control. After administration, TF-treated mice and control mice were transferred to the metabolic cages (5 mice/cage) to collect their urine and stool for 24 hours, respectively. The urine and stool samples were stored at −80 °C before analysis.

2.2.2. Metabolism of TF in germ-free (GF) and conventionalized mice

The experiment with GF mice and conventionalized mice was carried out according to the protocol approved by the Institutional Animal Care and Use Committee at the University of Florida (IACUC# 201609606). GF wild type (WT) 129/SvEv mice (10–14-week old), were gavaged with SPF microbiota and transferred to regular housing for 2 weeks (conventionalized mice). For metabolic experiments, both GF mice (n=7 with 3 female and 4 male mice) and conventionalized mice (n=8 with 5 male and 3 female mice) were individually housed in metabolic cages, fasted overnight, and then orally gavaged with 200 mg/kg TF before collecting their urine and stool for 24 hours. The urine and stool samples were stored at −80 °C before analysis.

2.2.3. Preparation of mouse fecal samples

After drying by compressed nitrogen, the feces of each group were mashed and mixed completely, then 100 mg of each mouse fecal samples were extracted with 1 mL of 80% aqueous methanol with 0.1% formic acid under ultrasonication for 10 min. After centrifugation at 16100g for 15 min at 4 °C, the supernatants were collected and analyzed by LC-MS directly.

2.3. In vitro chemical reactions

2.3.1. Amination of TF

The amination reaction was carried out according to our previous procedure with slight modification due to TF having poor stability in high concentration of ammonia. Briefly, 5 mg of TF was stirred with 500 μL of 3M ammonium hydroxide for 30 minutes with air at room temperature. The reaction was stopped by neutralization with 3M hydrochloric acid (final pH 4~5). The resulting solution was partitioned with ethyl acetate, and the ethyl acetate extract was used for LC-MS/MS analysis of the aminated products and for further MGO conjugation reaction.

To purify the aminated TF, 300 mg of TF was aminated according to the method described above. The aminated product was purified by a flash C18 column (5×15 cm) that was eluted with gradient methanol/water with 0.1% formic acid from 20% to 90%. The separation was monitored by LC-MS. The fractions containing the target aminated product was combined and dried (2, 5 mg) to record its NMR spectra in CD₃OD.

2.3.2. MGO conjugation with TF and the aminated TF

TF (1.5 mg) and the aminated products of TF (1 mg) were incubated with 100 μM of MGO in 100 mM potassium phosphate buffer (PBS, pH 7.4) by shaking at 37 °C under dark, respectively. After 3 h, the reactions were stopped by adding 2 μL of acetic acid. The reaction solutions were diluted with the same volume of methanol. After centrifugation, the MGO conjugation products were checked by LC-MS/MS.

2.3.3. Oxidation of TF

TF (5 mg) and POD (1 mg) were mixed in water (5 mL). During stirring at 700 rpm at room temperature, H₂O₂ (35%, 200 μL) was added. After 1 h of reaction, the oxidation products were extracted with 5 mL of ethyl acetate for three times. The ethyl acetate extract was dried by nitrogen and the residue was dissolved in methanol for LC-MS/MS analysis.

2.4. LC-MS/MS analysis

The analysis of TF related samples was performed with a Thermo Scientific Vanquish UHPLC system consisting of a Vanquish pump, Vanquish autosampler, and column temperature compartment, which was coupled to a Q-Exactive Plus Orbitrap tandem mass spectrometer via electrospray ionization interface (Thermo Scientific, Waltham, MA). The separation of TF and its metabolites was performed on a Gemini C18 110Å column (50 mm × 2.0 mm, 3μm; Phenomenex, Torrance, CA) at a flow rate of 0.2 mL/min. The mobile phases were 100% water with 0.1% formic acid as solvent A and 100% acetonitrile with 0.1% formic acid as solvent B. The gradient program was: 5% B from 0 to 1 min, 5%−100% B from 1 to 8 min, 100% B from 8 to 9 min, then the column was balanced to 5% B. The injection volume was 5 μL.

Mass spectrometric detection was carried out in negative ion mode. Mass spectrometer was calibrated before detection and the source was tuned with TF. The source parameters were as follows: spray voltage, 2.5 kV; capillary temperature, 300 °C; sheath gas, 40 units; auxiliary gas, 10 units; auxiliary gas heater temperature, 300 °C; sweep gas, 3 units. Data were acquired from 100–1000 Da with parallel reaction monitoring (PRM) mode at a resolution of 17500, 1e5 of automatic gain control (AGC), and 15, 35, and 65 of stepped collision energy.

2.5. Nuclear Magnetic Resonance (NMR) analysis

¹H- and ¹³C-NMR of TF and b-NH₂-TF, Heteronuclear Single Quantum Correlation (HSQC), and Heteronuclear Multiple Bond Correlation (HMBC) of b-NH₂-TF NMR spectra were recorded on a Bruker AVANCE 600 MHz spectrometer (Bruker Inc., Silberstreifen, Rheinstetten, Germany). Multiplicities were indicated by s (singlet), d (doublet), dd (double doublet) and m (multiplet). The ¹³C-NMR spectra were proton decoupled. CD₃OD was used as solvent for NMR data acquisition.

2.6. Statistical Analysis

All results are presented as means ± standard deviation. A Student’s t test was performed using GraphPad Prism version 5.04. A p-value < 0.05 was considered significant.

3. Results

3.1. Chemical synthesis and structure elucidation of the aminated TF

The aminated TF was synthesized by stirring TF in 3M ammonia at room temperature (Figure 1). Since TF and its products were not stable in base solution, only 5 mg of the aminated product (2) was obtained from 300 mg of TF. The structure of this product was characterized after analyzing its HR-MS and NMR data.

Figure 1.

The amination reaction and its underlying mechanism of TF and the key HMBC correlations of b-NH₂-TF.

Its molecular formula was elucidated as C₂₉H₂₅NO₁₁ based on its HR-ESI-MS at m/z 562.1345 [M-H]⁻ (Calcd. 562.1349 [M-H]⁻) (Figure 2B), which was one hydrogen and one nitrogen more, and one oxygen less than TF, suggesting one hydroxyl group on TF was replaced with an amino group on this new product. Compared to the ¹H- and ¹³C-NMR spectra of TF, the proton and carbon signals from the A- and C-ring on compound 2 were identical to those of TF (Table 1). However, the signals from the benzotropolone moiety on compound 2 were slightly shifted, which indicated that the amino group was connected to this moiety. The analysis of its HSQC and HMBC spectra revealed that H-2 at δ_H 4.87 on the C-ring showed HMBC correlations to the carbons at δ_C 130.6 and δ_C 133.0, and these two carbon signals were associated to the protons at δ_H 7.71 and 7.94 on the HSQC spectrum, respectively, indicating these two carbons and protons belonged to CH-c and CH-e. The HMBC correlation from H-2 to the carbon at δ_C 134.1 revealed that this carbon was C-d. The proton at δ_H 7.71 (H-c) showed HMBC correlations to the carbons at δ_C 136.7, δ_C 134.1 (C-d), δ_C 133.0 (C-e), and δ_C 188.3 (C-a) suggesting that the carbon at δ_C 136.7 was C-b. The high-field shift of C-b was due to the lower electronegativity of the amino group compared to the hydroxyl group. Therefore, the amine group was detemined to be bonded to C-b and the aminated product of TF was characterized as b-NH₂-TF (2, Figure 1). The ¹H- and ¹³C-NMR data of b-NH₂-TF were assigned in Table 1. The reaction mechanism was proposed in Figure 1.

Figure 2.

The formation of the aminated metabolites of TF in mice. (A) The LC chromatograms under parallel reaction monitoring (PRM) mode of the aminated TF in mouse fecal samples collected from TF (1) treated mice (Feces), and the synthetic standard of b-NH₂-TF (Std). (B) The ESI-MS² (negative ion) spectra of the ions at m/z 562.1349 [M-H]⁻ from the TF-treated mouse sample and the synthetic standard of b-NH₂-TF. (C) The key ESI-MS fragment pathways of b-NH₂-TF.

Table 1.

¹H- and ¹³C-NMR data of TF (1) and its aminated product (2)^a

Position	1		2
	1H	13C	1H	13C
	δ (ppm), J (Hz)	δ (ppm)	δ (ppm), J (Hz)	δ (ppm)
2	4.88b	81.6	4.87b	79.4
3	4.31–4.33 (m)	66.9	4.25–4.28 (m)	65.2
4	2.98 (dd, 4.4, 16.8) 2.82 (d, 17.0)	30.3	2.97–3.01 (m) 2.80–2.85 (m)	28.9
4a	-	100.5	-	101.3
5	-	158.4c	-	154.7c
7	-	158.3c	-	155.3c
5’	-	158.2c	-	155.8c
7’	-	158.1c	-	156.0c
8a	-	157.9c	-	156.3c
8a’	-	157.6c	-	156.4c
6	6.02 (d, 2.3)c	97.1c	6.02 (s)c	98.4c
6’	6.00 (d, 2.3)c	97.1c	6.00 (s)c	98.4c
8	5.98 (d, 2.3)c	96.4c	5.98 (s)c	97.8c
8’	5.95 (d, 2.3)c	96.0c	5.93 (s)c	97.8c
a	-	186.1	-	188.3
b	-	156.9	-	136.7
c	7.34 (s)	118.6	7.71 (s)	130.6
d	-	134.9	-	134.1
e	7.83 (s)	124.2	7.94 (s)	133.0
f	-	131.9	-	134.8
g	7.96 (s)	126.7	7.96 (s)	124.9
h	-	146.7	-	149.4
i	-	151.3	-	152.8
j	-	122.5	-	124.3
k	-	129.3	-	128.5
2’	5.62 (s)	77.5	5.63 (s)	75.4
3’	4.44–4.46 (m)	65.9	4.35–4.38 (m)	64.3
4’	2.94 (dd, 4.4, 16.8) 2.84 (d, 17.0)	29.7	2.89–2.93 (m) 2.80–2.85 (m)	28.3
4a’	-	100.1	-	100.3

^arecorded in CD₃OD, 600 MHz (¹H), 150 MHz (¹³C).

^boverlapped in water, assigned based on HMQC and HMBC data.

^cnot assign to exact position because the shifts were too close.

3.2. Identification of the aminated metabolite of TF in mice

To verify that TF can be aminated in vivo, TF was administrated to CD-1 mice at 200 mg/kg by oral gavage. After scanning for aminated TF at m/z 562.1345 [M-H]⁻, there was a major peak at 6.38 min that showed an identical retention time and mass spectrum to those of the synthetic standard in the TF-treated mouse fecal samples but not in the control samples (Figure 2A and 2B). The fragment at m/z 424.1038 [M-138-H]⁻ was from the typical loss of the A-ring, and the fragments at m/z 125.0242 [M-H]⁻ and 137.0241 [M-H]⁻ were from the A-ring (Figure 2B, 2C). Altogether, this metabolite from the TF-treated mouse fecal sample was characterized as b-NH₂-TF. Besides the major peak at 6.38 min, one minor peak at 5.86 min with m/z 562.1343 [M-H]⁻ in TF-treated mouse fecal sample shared identical fragments with b-NH₂-TF, suggesting it is possible that the amination is at position C-h or C-i on the benzotropolone core structure on this minor metabolite. Taken together, TF has the capacity to trap endogenous ammonia to produce aminated metabolites in mice.

3.3. Conjugation of TF with MGO in mice

Many studies have reported that tea polyphenols, including theaflavins, showed strong trapping capacity towards reactive dicarbonyl species.^{[12, 10, 3, 3, 12]} Theaflavins were found to be more reactive to MGO than epigallocatechin 3-gallate (EGCG), the most abundant catechin in green tea, and decreased MGO by more than 60% in a molar ratio of 3 (MGO/TFs) within 1 h incubation.^[12] However, there is no report on the in vivo trapping of MGO by TFs. To this end, TF-treated mouse fecal samples were analyzed and compared with products from the in vitro incubation of TF with MGO using LC-MS/MS.

As shown in Figure 3, three major peaks were found in response to the search of mono-MGO conjugated TF at m/z 635.1401 [M-H]⁻ (C₃₂H₂₈O₁₄) at 4.76, 4.94, and 5.06 min in TF-treated mouse fecal samples, in which they had almost identical retention times and tandem mass spectra with those of the three major mono-MGO conjugated TF products from the incubation of TF with MGO in vitro (Figure 3). All three major peaks had fragment ions at m/z 563.1190 [M-72-H]⁻, which was from the loss of one MGO unit (Figure 3B and 3C). Furthermore, other daughter ions including m/z 425.0880, 241.0503, 137.02442, and 125.0242 were similar to the typical fragment pattern of TF. It has been confirmed that C-6 and C-8 on the A-ring of catechins were the two active positions to conjugate with MGO. TF retains the same two A-rings from catechins. Therefore, we excepted that MGO was conjugated with the C-6 and C-8 positions on the A-rings of TF (Figure 3). These results showed that TF conjugated with MGO in vivo to produce the mono-MGO conjugates.

Figure 3.

Mono-MGO conjugation of TF. (A) The LC chromatograms under parallel reaction monitoring (PRM) mode of mono-MGO conjugates with TF (mono-MGO-TF) in mouse fecal samples collected from TF (1) treated mice (Feces), and the synthetic standard (Std) from the incubation of TF with MGO under the physiological conditions and the key ESI-MS fragment pathways of mono-MGO-TF. The ESI-MS² (negative ion) spectra of the ion at m/z 635.1401 [M-H]⁻ from TF-treated mouse fecal sample (B), and synthetic standard (C) at different retention times.

3.4. Conjugation of aminated TF with MGO in mice

Due to the amination occurring on the benzotropolone core structure of TF, the aminated TF retains the same C-6 and C-8 positions on the A-rings as those of TF. Therefore, we hypothesized that the aminated TF had the capacity to trap MGO. To test this hypothesis, we searched the mono-MGO conjugated NH₂-TF from TF-treated mouse fecal samples. As shown in Figure 4, NH₂-TF further trapped MGO to form three clear mono-MGO conjugates with the target m/z 634.1560 [M-H]⁻, which had almost identical retention times and tandem mass spectra with those of the three major products formed from the in vitro incubation of NH₂-TF with MGO. These metabolites showed the strongest daughter ions at m/z 562.1350 [M-72-H]⁻ from the typical loss of MGO (Figure 4B and 4C), which confirmed they are the mono-MGO conjugates of the NH₂-TF. Furthermore, these metabolites also had the typical fragment ions of NH₂-TF, which include m/z 424.1070, 137.0243, and 125.0241. Therefore, the aminated metabolite of TF is capable of traping MGO in vivo to generate the corresponding mono-MGO conjugates.

Figure 4.

Mono-MGO conjugation of the aminated TF. (A) The LC chromatograms under parallel reaction monitoring (PRM) mode of mono-MGO conjugates of the aminated TF (mono-MGO-NH₂-TF) in mouse fecal samples collected from TF (1) treated mice (TF-fecal), and of the synthetic standard (Std) from the incubation of the aminated TF with MGO under the physiological conditions and the key ESI-MS fragment pathways of mono-MGO-NH₂-TF. The ESI-MS² (negative ion) spectra of the ion at m/z 634.1560 [M-H]⁻ from TF-treated mouse fecal sample (B), and the synthetic standard (C) at different retention times.

3.5. Oxidation and further amination of TF in mice

Previous studies have reported that TF can be oxidized to dehydrotheaflavin during model tea fermentation or under the presence of metal.^[24–27] However, whether dehydrotheaflavin is the oxidized metabolite of TF in vivo and has the capacity to trap ammonia in mice are still unknown. Under the search of dehydrotheaflavin at m/z 577.0971 [M-H]⁻, one peak appeared at 4.47 min in TF-treated mouse fecal sample (Figure 5A). Its molecular formula was deduced as C₂₉H₂₂O₁₃ based on its HR-ESI-MS at m/z 577.0971 [M-H]⁻ (Calcd. 577.0982 [M-H]⁻), which was one oxygen more and two hydrogens less than TF. To confirm this peak was dehydrotheaflavin, we synthesized dehydrotheaflavin as a reference by stirring TF with POD and H₂O₂.^{[4, 43]} As shown in Figure 5A, the in vivo oxidized metabolite showed identical retention time and tandem mass spectrum with those of the standard generated during the enzymatic oxidation of TF. The fragment at m/z 125.0242 was from the A-ring, and the fragments at m/z 409.0556 [M-168-H]⁻, 451.0662 [M-126-H]⁻, and 439.0659 [M-138-H]⁻ were the results of the characteristic loss of the A-ring (Figure 5A). Therefore, this oxidized metabolite was tentatively characterized as dehydrotheaflavin. This is the first time to report dehydrotheaflavin as the oxidized metabolite of TF in vivo.

Figure 5.

Oxidation of TF (dehydrotheaflavin) and amination of oxidized TF (NH₂-dehydrotheaflavin) in mice. (A) The LC chromatograms under parallel reaction monitoring (PRM) mode and the ESI-MS² (negative ion) spectra of the oxidized metabolite of TF (dehydrotheaflavin) in TF treated mouse fecal sample (Feces) and the synthetic standard by stirring TF with peroxidase and H₂O₂. (B) The LC chromatograms under parallel reaction monitoring (PRM) mode and the ESI-MS² (negative ion) spectra of the aminated metabolites of oxidized TF (NH₂-dehydrotheaflavin) in TF treated mouse fecal sample (Feces).

Under the search of the aminated metabolite of dehydrotheaflavin at m/z 576.1142, two major peaks appeared at 4.18 and 4.29 min (Figure 5B). Their molecular formulas were deduced as C₂₉H₂₃NO₁₂ based on their HR-ESI-MS at m/z 576.1140 [M-H]⁻ (Calcd. 576.1142 [M-H]⁻), indicating that they were the aminated metabolites of dehydrotheaflavin. Similar to dehydrotheaflavin, the fragment at m/z 125.0242 was from the A-ring, and the fragments at m/z 408.0715 [M-168-H]⁻, 450.0831 [M-126-H]⁻, and 438.0821 [M-138-H]⁻ were the characteristic loss of the A-ring. These data indicate that the amination occurred on the oxidized benzotropolone core instead of the A-ring or C-ring of dehydrotheaflavin. Therefore, these metabolites were tentatively identified as NH₂-dehydrotheaflavin, and their structures are shown in Figure 5B.

3.1.6. The impact of gut microbiota on the capacity of TF to trap deleterious reactive endogenous metabolites

The impact of gut microbiota on the capacity of TF to trap ammonia and MGO was investigated by administrating TF to GF and conventionalized mice. The production of two representative metabolites, NH₂-TF and mono-MGO-TF were compared between TF-treated GF and SPF fecal samples. As shown in Figure 6, the levels of NH₂-TF and mono-MGO-TF were significantly higher in ex-GF mice than GF mice (p<0.05 and p<0.01, respectively). These results demonstrated that gut microbiota played an important role in the formation of NH₂-TF and mono-MGO-TF in mice.

Figure 6.

The effect of gut microbiota on trapping deleterious reactive endogenous metabolites by TF. The averages of peak area of the metabolites of TF (NH₂-TF and mono-MGO-TF) from TF treated germ-free (GF) (n=7) and conventionalized (n=8) mouse fecal samples were used to detect the impact of gut microbiota on the formation of these metabolites. Data are presented as the means ± standard deviation. A 2-tailed Student’s t test distribution with paired groups was evaluated for statistical significance. * p <0.05 and ** p < 0.01 were considered statistically significant.

4. Discussion

An increasing amount of evidence has demonstrated that TFs have diverse biological functions, which are believed to arise from their antioxidant activity.^{[1, 6]} The benzotropolone moiety of TFs has been observed to be able to undergo multiple oxidation pathways forming various oxidation products due to a quinone’s large resonance system. ^{[4, 43, 15, 15]} This study demonstrated for the first time that TF can be oxidized in vivo to form the quinone structure on the benzotropolone moiety, which can further react with ammonia in the intestinal tract to produce the aminated TF or generate the oxidized dehydrotheaflavin and aminated dehydrotheaflavin metabolites, respectively.

We have demonstrated that the A-ring in flavonoids is the active site to trap MGO. ^[12] As the oxidized product of epicatechin (EC) and epigallocatechin (EGC), TF retains the A-ring structures of EC and EGC, suggesting TF has the capacity to trap MGO. This study provides the first in vivo evidence that TF can trap MGO in mice to generate the mono-MGO conjugated TF metabolites. As the major metabolite of TF, the aminated TF has the same A-ring structures as TF, allowing this metabolite to further trap MGO. This was confirmed from the observation of the mono-MGO conjugated metabolites of the aminated TF in mice.

The gut microbiota consists of trillions of microbial cells and thousands of diverse bacterial species with an extensive metabolic function that contribute not only to the formation of ammonia ^{[3, 2, 1]} and MGO, ^[3] but also to the metabolism of dietary compounds. ^[1] In this study, we observed that gut microbiota facilitates the formation of the aminated TF and the mono-MGO conjugated TF, suggesting that the gut microbiota plays an important role on the detoxification effects that TF exhibits. It is possible that certain gut microbes have the enzyme to oxidize TF to its quinone, which further reacts with ammonia to generate the aminated metabolite or to produce the oxidized metabolite. Interestingly, both the aminated TF and the MGO conjugated TF were detected in GF mice, suggesting that a host-derived gut mucosa enzyme also contributes to the amination and MGO conjugation of TF.

Due to its poor systematic bioavailability, large amounts of the unabsorbed TF remain in the intestinal tract. Our observation that TF has the capacity to remove metabolic toxins through oxidation, amination, and MGO conjugation in the intestinal tract suggests a beneficial impact of TF in host homeostasis. In this scenario, the unabsorbed compound could mediate its protective effects via removing of metabolic toxins in the intestinal tract, preventing dissemination of these metabolic toxins throughout the body. Whether TF or black tea consumption could prevent ammonia and MGO associated chronic diseases would require future investigation.

Abbreviations:

TF

theaflavin

TF3G

theaflavin 3-mono gallate

TF3’G

theaflavin 3’-mono gallate

TFDG

theaflavin 3,3’-digallate

TFs

theaflavins

TRs

thearubigins

EC

epicatechin

ECG

epicatechin gallate

EGC

epigallocatechin

EGCG

epigallocatechin-3-gallate

PPO

polyphenol oxidase

POD

peroxidase

HSQC

Heteronuclear Single Quantum Correlation

HMBC

Heteronuclear Multiple Bond Correlation

GF

germ free

SPF

specific pathogen free

MGO

methylglyoxal

H₂O₂

hydrogen peroxide

RCSs

reactive carbonyl species