Tracking deposition of a 14 C-radiolabeled kudzu hairy root-derived isoflavone-rich fraction into bone

Induction and maintenance of kudzu hairy root cultures

Seeds from four kudzu genotypes, P. montana var. lobata (United States Department of Agriculture [USDA] no. PI9227, Origin: Japan), P. montana var. lobata (USDA no. PI 434246, Origin: USA), P. lobata (Binkley Rd population, Williamson County, IL, USA) ²⁶ and P. phaseoloides (Origin: Guatemala), were scarified, soaked overnight in tap water containing polyoxyethylene sorbitan monolaurate (0.1% v/v, Tween 20, Sigma-Aldrich, St Louis, MO, USA), surface disinfested in 0.9% sodium hypochlorite for 10 min and rinsed with sterilized distilled water. Seeds were then explanted onto media, ²⁹ containing 1/2× (2.165 g/L) Murashige and Skoog (MS) basal salts (Phytotechnology Laboratories, Shawnee Mission, KS, USA), ³⁰ Gamborg B5 vitamins, ³¹ 0.1 g/L myoinositol (Phytotechnology Laboratories), 30 g/L sucrose, 0.1% FeEDTA and 2 g/L gellan gum (PhytoTechnology Laboratories). Seeds were allowed to germinate and grow for seven days under fluorescent lights. Sterile Luria-Bertani broth (LB) (Phytotechnology Laboratories) was prepared, inoculated with the cucumopine-producing A. rhizogenes strain, K599, and cultured overnight. Seedlings were then harvested, cotyledon explants were wounded on the abaxial side with a scalpel, placed in LB media containing A. rhizogenes and 250 μL acetosyringone for 5–10 min and then set onto sterile Petri dishes lined with sterile filter paper to co-cultivate for two days under fluorescent lights. The cotyledon explants were then transferred to Petri dishes containing a growth-regulator-free MS media, containing 4.330 g/L MS basal salts, Gamborg B5 vitamins, 0.1 g/L myoinositol, 30 g/L sucrose, 0.1% FeEDTA, 2 g/L gellan gum and 500 mg/L carbenicillin to inhibit bacterial growth. After two weeks, explants were transferred to MS media with 250 mg/L carbenicillin. When hairy roots were sufficiently large, they were severed from the cotyledon and transferred to dark growing conditions at 25°C and subcultured every two weeks onto MS media containing 250 mg/L carbenicillin until no signs of bacterial contamination were evident, at which point roots were subcultured to carbenicillin-free media. Kudzu hairy roots on growth-regulator-free media were identified by their rapid growth rates and extensively branching phenotype, and further authenticated by high-voltage paper electrophoresis (HVPE) of aqueous kudzu extracts for the synthesis of cucumopine, which is unique to hairy roots of this strain. ^32,33 An authentic cucumopine standard, and aqueous extracts from the hairy root genotypes and four non-transgenic genotypes were blotted onto Whatman #3 filter paper and allowed to dry. Filter paper was placed into a formic-acetic acid buffer of pH 1.8 and subjected to electrophoresis at 290 V for 70 min. Electrophoretograms were dried thoroughly in a stream of warm air and then visualized by spraying lightly with Pauly reagent. ³² Samples of kudzu hairy roots were dried on a freeze-dryer (Labconco Freezone 4.5, Kansas City, MO, USA) and stored at −80°C until ready for extraction and isoflavone analysis.

Extraction, identification and quantification of isoflavones from hairy roots

Extraction was performed as previously described ^26,27 with some modifications. Briefly, roots were extracted in methanol, freeze-dried and analyzed after filtration through a 0.45 μm nylon syringe filter (Fisher Scientific, Pittsburgh, PA, USA). Isoflavones were identified by high-performance liquid chromatography equipped with electrospray ionization–mass spectrometry (HPLC–ESI–MS) using a LCQ Deca XP mass spectrometer under positive ESI mode (m/z 100–1000) attached to a photodiode array (PDA) detector (Thermo Finnigan Corp., San Jose, CA, USA) at ultraviolet (UV) wavelength 262 nm. Separations were performed on a 2.1 mm × 150 mm, 3.5 μm, Waters XBridge C18-reversed phase column (Waters Corp., Milford, MA, USA) using a mobile phase consisting of 5% acetonitrile (ACN) in double-distilled H₂O with 0.1% formic acid (Fisher Scientific) in solvent A and 95% ACN in double-distilled H₂O with 0.1% formic acid in solvent B. ^26,27 Isoflavones were quantified by high-performance liquid chromatography with photodiode array detection (HPLC-PDA), using an Agilent 1100 HPLC system (Agilent Technologies, Inc, Santa Clara, CA, USA) with an autosampler (G1313A), a degasser (G1322A), a quadratic pump (G1311A) and a temperature-controlled column compartment (G1316A). Separations were performed using a 250 mm × 4 mm, 5 μm Supelcosil LC-18 reversed-phase column (25°C) (Supelco, Bellefonte, PA, USA) with a mobile phase consisting of 10% ACN in double-distilled H₂O with 0.1% trifluoroacetic acid (TFA) (Sigma-Aldrich) in solvent A and 90% ACN in double-distilled H₂O with 0.1% TFA in solvent B. The elution gradient was composed of 0%, 70%, 100% and 0% of solvent B at 0, 30, 35 and 40 min, respectively, and UV detection was at a wavelength of 262 nm. The injection volume was 15 μL/min and the flow rate was 1.5 mL/min. Data were processed using ChemStation Software for LC 3D systems (Rev. A.10.02, Agilent Technologies, Inc). Authentic standards for puerarin, daidzein and genistein from Sigma-Aldrich and daidzin, genistin and malonyl-genistin from LC Laboratories (Woburn, MA, USA) were used to verify UV absorption, retention time and molecular weight of isoflavones, and those that could not be authenticated with standards due to commercial unavailability were identified by comparison to literature data and quantified by molecular weight adjustment of related standards. ³⁴

Radiolabeling of kudzu hairy roots in vivo

Hairy roots were cultured on Petri dishes containing 20 mL solid MS media with 2 g/L gellan gum for two weeks. Fresh hairy root mass was recorded from 30 uniformly growing cultures of each genotype after the two-week culture period. Measurements were taken after four separate culture periods (n = 4). Average fresh mass per hairy root culture was calculated for each culture interval. Isoflavones were extracted from approx. 10 g fresh kudzu hairy roots (n = 4) of each genotype, identified by HPLC–ESI–MS, and quantified by HPLC-PDA. Total isoflavone content was compared by one-way analysis of variance, followed by post hoc analysis with Tukey's least significant difference multiple comparison test at a level of significance of 0.05 using SPSS Statistics software (Release 17.0, SPSS Inc, Chicago, IL, USA)

¹⁴C-sucrose (MP Biomedicals, Inc, Irvine, CA, USA), 120 mg, with a specific activity of 370.0 MBq/mmol (10.00 mCi/mmol) was dissolved in 125 mL distilled water to create a concentrated stock solution, then filter sterilized (0.2 μm filter, Nalgene, Rochester, NY, USA).

Tip segments of approx. 15 cm from actively growing kudzu (P. montana var. lobata, USDA no. PI 434246) hairy roots were transferred from solid MS media to 40 mL concentrated liquid MS media (prepared by dissolving all constituents in distilled water at 80% of the final concentration prior to autoclaving) in 250 mL flasks, on a gyrotory shaker (New Brunswick Scientific Co, Edison, NJ, USA) at 150 rpm. A volume of 10 mL ¹⁴C-sucrose stock solution 10.79 MBq (291.7 μCi) was added to each flask to bring the final media volume to 50 mL. Hairy roots grew for three weeks in a dark, enclosed Plexiglas labeling chamber, designed to capture respired ¹⁴CO₂, ^27,35 and were then harvested by emptying cultures into a Buchner funnel containing a Whatman #4 filter, connected to a suction filtration flask. Fresh weight was recorded, roots were frozen at −20°C, and frozen samples were loaded onto a freeze-dryer until all roots were completely dry, then extracted as previously described.

Fractionation of radiolabeled isoflavone-containing extracts

Radiolabeled kudzu dry crude extract was successively extracted with petroleum ether, ethyl acetate, butanol and water fractions by liquid–liquid partitioning as follows below. The crude extract was mixed with distilled water in a separatory funnel, and partitioned with petroleum ether. This process was repeated four more times, adding fresh petroleum ether each time. Petroleum ether extracts were combined and solvent was removed using a rotary evaporator to afford the petroleum ether extract, then the aqueous phase was partitioned with ethyl acetate 10 times. The ethyl acetate was removed by a rotary evaporator to afford the ethyl acetate extract. The aqueous layer was partitioned with butanol five times, and solvent evaporated to afford the butanol extract. The aqueous layer was then freeze-dried. Dried ethyl acetate and butanol fractions were then combined and partitioned again by the above procedure.

The combined ethyl acetate and butanol fraction from the second partitioning procedure was fractionated on a Sephadex LH-20 gel filtration column (Sigma-Aldrich), using ethanol as the eluting solvent, followed by methanol to wash the column of residual extract. Sephadex fractions were tested by thin-layer chromatography (TLC) on Silica Gel 60 F₂₅₄ Plates (EMD Chemicals Inc, Gibbstown, NJ, USA), with ethyl acetate, methanol and distilled water (77:13:10) solvent. TLC plates were visualized using vanillin/sulfuric acid reagent and heating to 105°C for 5 min. According to TLC profiles, similar fractions were combined, solvent was removed with a rotoevaporator and residual water was removed with a freeze-dryer. Isoflavone content of each fraction was also evaluated by HPLC-PDA and the isoflavone-rich fractions were combined.

Metabolic tracking of ¹⁴C-labeled, isoflavone-rich fraction in rats

Male Sprague–Dawley rats (n = 7, body weight 250–300 g), acclimatized to a 12 h photoperiod and a polyphenol-free diet (AIN93G, Dyets Inc, Bethlehem, PA, USA) for seven days were anesthetized by intraperitoneal injection (ip) with ketamine–xylazine (0.1 mL/100 g body weight ip, 100 mg ketamine–10 mg xylazine/mL; Henry Schein Inc, Melville, NY, USA). A locking intracerebral guide and stylet (MD-2251, Bioanalytical Systems, Inc, West Lafayette, IN, USA) was inserted stereotaxically into the hippocampal fissure of the brain, using the three-dimensional stereotaxic coordinates from a rat brain atlas. ³⁶ After surgery, the anesthesia was changed to 3–5% isofluorane, delivered with oxygen at 2–4 L/min, and a femoral catheter (CX-2020S, Bioanalytical Systems, Inc) was inserted into the right femoral vein and an ultrafiltrate probe (MF-7023/UF-3-12, Bioanalytical Systems, Inc) was implanted subcutaneously along the dorsal midline. Following surgery, the rats were administered buprenex (0.1 mL/100 g body weight ip, 0.03 mg buprenex/mL; Henry Schein Inc) in saline (0.9% NaCl w/v) for pain relief and connected to a Culex^® Automated Blood Sampling System (Bioanalytical Systems, Inc) for blood, interstitial fluid (ISF), brain microdialysate, urine and feces collection. Rats were given 48 h for recovery. Food (AIN93G) and water were provided ad libitum and the catheter was flushed with 10 μL of heparinized saline (10 units heparin/mL saline [0.9% NaCl w/v]) every 15 min. The ultrafiltrate probe was attached to a peristaltic pump to collect ISF at a rate of 1 μL/min. All procedures were approved by the Purdue Animal Care and Use Committee, which adheres to policies set forth by the USDA and the United States Public Health Service.

Rats were food-deprived for eight hours prior to gavage. One hour before dosing, the brain microdialysate stylet was removed from the intracerebral guide and a brain microdialysate probe with a 4 mm membrane (MD-2204/BR-4, Bioanalytical Systems, Inc) was inserted into the hippocampal fissure through the intracerebral guide. The probe inlet was attached to a syringe pump and perfused with artificial cerebrospinal fluid at a rate of 1 μL/min. Microdialysate was collected from the probe outlet, ISF was collected from the ultrafiltrate probe and 200 μL blood was drawn prior to gavage as the baseline.

Approximately 60 mg/kg body weight of ¹⁴C-labeled, isoflavone-rich kudzu extract 8.614 MBq/g (232.8 μCi/g) in 1 mL of distilled water was administered to the rats by gavage, followed by an immediate rinse of the gavage needle with 0.5 mL of distilled water. The Culex system automatically sampled 200 μL blood at 5, 15, 30, 45, 60, 90, 120, 180, 240, 300, 360, 480, 600, 720 and 1440 min after gavage and collected ISF and brain microdialysate at 60-min intervals for a total of 24 h. Urine and feces were collected for 24 h. Food was returned 240 min after gavage.

Rats were sacrificed 24 h after gavage. They were anesthetized with ketamine–xylazine (0.1 mL/100 g body weight ip, 100 mg ketamine–10 mg xylazine/mL; Henry Schein Inc) and exsanguinated. Blood was transferred to heparinized microcentrifuge tubes, centrifuged at 5000 g for 10 min, and plasma was collected for analysis. The vascular system was perfused with cold saline to remove blood from tissues. An incision was made through the left jugular vein and the rat was perfused through the right femoral catheter with 240 mL chilled saline, which was sufficient to produce clear perfusate that exited the incision in the left jugular vein. Perfusate was also collected for analysis. The stomach, small intestines, large intestines, brain, heart, lungs, liver, kidneys and testes were placed into scintillation vials. The contents of the stomach, small intestines and large intestines were collected by flushing the organs through with 3, 5 and 10 mL saline (0.9% NaCl), respectively. Femurs, tibias and vertebrae were carefully removed and placed into scintillation vials. All samples were stored at −80°C until ready for analysis.

¹⁴C activity from serum, ISF, gut contents, feces, urine, bone and tissue samples was measured with a Beckman LS 6500 scintillation counter (Beckman Coulter, Inc, Fullerton, CA, USA). Bio-Safe II scintillation cocktail (Research Products International Corp., Mt. Prospect, IL, USA), 20 mL, was added to all sample vials and allowed to sit overnight in the dark, prior to analysis in the scintillation counter. Tissue samples, gut contents and feces were prepared by freeze-drying, grinding dried samples to a fine powder and extracting ¹⁴C-labeled compounds overnight with 80% methanol. Bones were cleaned, freeze-dried and ground to a fine powder. Samples of ground bone of a known mass were dissolved in concentrated nitric acid overnight to achieve uniformity for scintillation counting. Residual radiolabel in each animal carcass, which included all bones, muscle, adipose and connective tissue not previously removed, was analyzed by liquefying in 10% potassium hydroxide for one week, neutralizing with 12 N hydrochloric acid and homogenizing for even consistency. Brain microdialysate was analyzed by an accelerator mass spectrometer (Purdue University, West Lafayette, IN, USA).

ISF was collected at an average rate of 60 μL/h. The ultrafiltrate probe tubing was 48″ in length, with a total internal volume of approximately 12 μL. This internal volume presents a delay between the time ISF is collected and the time ISF reaches the sample collection vial. To account for this delay and to more accurately reflect the amount of ¹⁴C label present in the ISF at any given time, the ISF collection time points illustrated in Figure 6 were shifted 12 min earlier than actually collected. Total serum and ISF volumes were approximated based on the body weight of each rat ³⁷ and were used to estimate the percent of ¹⁴C-labeled compounds remaining in each respective fluid 24 h after gavage.

Approximately 1 g of dried, finely ground femurs was added to 50 mL of 80% methanol for extraction of metabolites. Methanolic extract was centrifuged at 5000 rpm for 10 min and filtered with a Whatman #4 filter, a Whatman #1 filter and then a Millipore 0.20 μm nylon filter (Millipore, Billerica, MA, USA). Filtered bone extract was dried with a rotoevaporator and a freeze-dryer. Dried extract was weighed, redissolved in 500 μL of 80% methanol, and filtered with a 0.45 μm nylon syringe filter. A 200 μL sample of the femur extract was spiked with commercial isoflavone standards (puerarin, daidzin, genistin, daidzein and genistein) for comparison. The remaining 300 μL and the spiked extract were analyzed by HPLC–ESI–MS using the same methodology described previously.

Induction of hairy roots in kudzu genotypes

During the co-cultivation of A. rhizogenes with kudzu cotyledons, the bacterium transferred a segment of DNA from its root-inducing (Ri) plasmid to the plant genome, which allows roots to grow rapidly in the absence of exogenous growth regulators and to produce opines (not produced by non-transformed roots), which are modified amino acids used by the bacterium as an energy source. ³⁸ Transformation of kudzu hairy roots was positively confirmed by observation of the typical highly branched hairy root phenotype (Figure 1), ³³ and rapid root proliferation in growth-regulator-free media also served to indicate the transfer of root loci (rol) genes from the Ri plasmid. ³⁸ DNA transfer was further validated by analysis of hairy root extracts by HVPE for the presence of cucumopine, an opine produced specifically by the strain of A. rhizogenes. The presence of cucumopine in all transformed extracts was confirmed by comparison to an authentic standard, as shown in Figure 2. Cucumopine was absent in all non-transgenic kudzu root extracts, which served as negative controls. The synthesis of cucumopine in hairy root samples verified the transfer of opine synthase genes from the A. rhizogenes Ri plasmid into each kudzu genotype. ³³ Endogenous histidine was present in all root samples and was stained in the electrophoretogram when sprayed with Pauly reagent, while no histidine was present in the cucumopine standard. By exploiting the rapid growth rate of kudzu hairy root cultures, isoflavone production was improved over non-transformed root cultures.

OPEN IN VIEWER

Figure 1

Hairy root proliferation induced by Agrobacterium rhizogenes (strain K599) in kudzu (Pueraria montana var. lobata, USDA no. PI 434246, Origin: USA)

OPEN IN VIEWER

Figure 2

Paper electrophoretogram of cucumopine extracts from kudzu hairy roots

In order to determine the most promising secondary metabolite profiles and concentrations within these four genotypes, the major isoflavones were identified using HPLC–ESI–MS, in the positive ion mode, in each kudzu hairy root genotype as hydroxy-puerarin, puerarin, daidzin, apiosyl/xylosyl-glucosyl-genistein, genistin, malonyl-daidzin, malonyl-genistin, daidzein and genistein at retention times 7.3, 8.7, 10.2, 10.8, 11.8, 12.4, 14.2, 15.7 and 18.7 min, respectively. Quantification using HPLC-PDA showed that P. montana var. lobata (USDA no. PI 434246, Origin: USA) yielded the highest puerarin content and total isoflavone content at 53.4 and 129.9 mg/g dry extract (Table 1 and Figure 3), respectively. P. lobata exhibited lower total isoflavone content at only 58.9 mg/g dry extract. P. phaseoloides yielded no detectable quantities of hydroxy-puerarin, puerarin, apiosyl/xylosyl-glucosyl-genistein or genistin, but instead produced the highest quantities of daidzein and genistein among the genotypes evaluated at 21.9 and 18.8 mg/g dry extract, respectively. The isoflavone profile of P. montana var. lobata from Japan was similar to that of P. montana var. lobata from the USA; however, the Japanese genotype contained lower puerarin and total isoflavone contents at 50.5 and 108.2 mg/g dry extract, respectively. P. montana var. lobata from the USA yielded a significantly greater total isoflavone content than the other three genotypes tested (P < 0.001), and also yielded the greatest amount of new fresh mass per two-week subculture cycle among the four genotypes tested at approx. 3.57 g per culture (Table 2). This genotype also yielded the greatest amount of isoflavones per quantity of fresh roots at 58.5 mg isoflavones per 10 g fresh root mass. Based on the isoflavone yields and isoflavone profiles of the four genotypes tested, hairy roots from P. montana var. lobata (USDA no. PI 434246, Origin: USA) were selected for subsequent radiolabeling trials.

OPEN IN VIEWER

Figure 3

Isoflavone profiles of hairy root-induced kudzu genotypes

Extraction of radiolabeled isoflavones

From each 15 cm root segment (fresh mass approx. 0.1 g at time of subculture) of P. montana var. lobata (USDA no. PI 434246, Origin: USA), final root cultures averaging 13.1 g each were harvested after 21 days. This indicates that on average, 99.2% of the fresh root mass from each of the 12 harvested root cultures was comprised of new growth. New growth incorporates and utilizes both ¹⁴C-radiolabeled and unlabeled sucrose without discrimination. From this root growth data, we conclude that all new root tissue and its synthesized phytochemicals were homogeneously radiolabeled with ¹⁴C. Radiolabeled kudzu hairy roots (156.7 g fresh weight) were lyophilized to a final dry weight of 8.8 g. Dry hairy roots were extracted with 80% methanol, rotoevaporated and then freeze-dried. The crude extract (3.8 g) contained homogeneously labeled isoflavones at a concentration of 52.4 mg/g extract. In the current study, approx. 17.0% of the total ¹⁴C dose was incorporated into the crude extract, which had an average radioactivity of 5.839 MBq/g (157.8 μCi/g) extract, which was comparable to previous research with non-transgenic kudzu (P. lobata) roots that reported radiolabel incorporation of 18.7% of the ¹⁴C dose. ²⁶ The predominant isoflavones in the crude, radiolabeled kudzu extract were puerarin, malonyl-daidzin and daidzin at 27.2, 11.4 and 4.1 mg/g extract, respectively, as identified by HPLC–ESI–MS, quantified by HPLC-PDA, and shown in Table 3 and Figure 4.

OPEN IN VIEWER

Figure 4

HPLC–ESI–MS spectra showing total ion current (a), HPLC-UV chromatogram (262 nm) (b) and MS spectrum (c) of ¹⁴C-labeled, crude kudzu extract. HPLC–ESI–MS, high-performance liquid chromatography with electrospray ionization–mass spectrometry; UV, ultraviolet

Enrichment of radiolabeled isoflavones

To metabolically track radiolabeled isoflavones, removal of extraneous phytochemicals was required. Partitioning of the crude extract produced isoflavone-enriched ethyl acetate and butanol fractions, while removing unwanted compounds in petroleum ether and water fractions. After the first solvent extraction, 65.8% of the crude extract was isolated in the water-soluble fraction and 1.8% in the petroleum ether fraction, both of which contained negligible quantities of isoflavones. The ethyl acetate fraction contained primarily puerarin, apiosyl/xylosyl-glucosyl-genistein, daidzin and malonyl-genistin at 109.0, 24.5, 12.5 and 9.4 mg/g, respectively, while the butanol fraction contained predominantly puerarin, malonyl-daidzin, malonyl-genistin and daidzin at 61.1, 28.0, 7.5 and 6.9 mg/g, respectively. Enrichment of major isoflavones increased from 5.2% in the crude extract to 17.8% in the ethyl acetate fraction and 11.9% in the butanol fraction. Ethyl acetate and butanol fractions were combined, repartitioned by the same method, dried down and reanalyzed. Additional undesirable material was separated out through the second partitioning, in the water and petroleum ether fractions, which contained very low concentrations of isoflavones. The percentage of total isoflavones in ethyl acetate and butanol fractions improved in the second partitioning from 17.8% to 31.5% and from 11.9% to 20.7%, respectively. While liquid–liquid partitioning greatly improved the concentration of the major isoflavones in these two fractions over the crude extract, these two fractions were combined again and further fractionated to boost enrichment.

Sephadex LH-20 column fractionation of the combined ethyl acetate and butanol fractions yielded 83 new fractions, each of which were rotoevaporated, freeze-dried and evaluated by TLC and HPLC-PDA for isoflavone content. Isoflavone-rich fractions were combined, dried down and reassessed by HPLC–ESI–MS (Figure 5) and HPLC-PDA (Table 3). The resulting fraction weighed 322.7 mg, contained 588.1 mg isoflavones/g fraction (58.8% enrichment), and had an average radioactivity of 8.614 MBq/g (232.8 μCi/g) fraction. The predominant isoflavones in the isoflavone-enriched fraction were puerarin, daidzin and malonyl-daidzin at 395.4, 65.2 and 57.0 mg/ g fraction, respectively, as shown in Table 3. The level of enrichment of total isoflavones in the final fraction was over 10-fold higher than in the crude extract and the concentration of individual isoflavones was increased in the enriched fraction by as much as 15 times. Loss of malonyl-genistin in the isoflavone-rich fraction despite an increase in concentration of every other major isoflavone may be attributed to its strong susceptibility to degradation. ³⁹ The presence of genistein in the isoflavone-rich fraction despite its absence in the crude extract was due to its rise in concentration into detectable levels.

OPEN IN VIEWER

Figure 5

HPLC–ESI–MS spectra showing total ion current (a), HPLC-UV chromatogram (262 nm) (b) and MS spectrum (c) of ¹⁴C-labeled, isoflavone-rich kudzu extract. HPLC–ESI–MS, high-performance liquid chromatography with electrospray ionization–mass spectrometry; UV, ultraviolet

Metabolic tracking of radiolabeled, isoflavone-rich fraction in rats

In a previous study, we demonstrated that a dose of 0.1480 MBq (4 μCi) of ¹⁴C-labeled crude extract containing isoflavones could be traced effectively in vivo using a scintillation counter ²⁷ and these results were used as a basis for calculating the dose size for the current study. With an average radioactivity of 8.614 MBq/g (232.8 μCi/g) fraction, an oral dose of 60 mg of ¹⁴C-labeled, kudzu isoflavone-rich fraction per kg body weight was needed for each rat. Naturally occurring levels of ¹⁴C in each rat were established by analysis of baseline serum, ISF, urine and feces by a scintillation counter.

Serum pharmacokinetics showed rapid absorption of the isoflavone-rich fraction, with serum reaching peak ¹⁴C-label concentrations approx 30 min after gavage, at an average of 0.053% of the administered ¹⁴C dose per mL serum (Figure 6). As seen previously in the metabolic tracking of radiolabeled kudzu crude extract, ²⁷ a second peak in serum concentration was observed one hour after gavage of the isoflavone-rich fraction and may be attributed to varied rates of isoflavone absorption. Isoflavone aglycones, for instance, are absorbed faster and in higher amounts than their glucosides. ⁴⁰ A plateau in serum concentration at four hours after gavage, which followed the trend seen in the metabolic tracking of ¹⁴C-kudzu crude extract, ²⁷ is consistent with the enterohepatic circulation of kudzu isoflavones. ⁴¹ Plasma collected 24 h after gavage showed no difference in ¹⁴C activity compared with serum collected 24 h after gavage. Approximately 0.459% of the ¹⁴C-labeled dose remained in the perfusate 24 h after gavage.

OPEN IN VIEWER

Figure 6

Distribution of ¹⁴C-label for isoflavone-rich kudzu fraction in serum and interstitial fluid of rats during 24 h after administration

ISF concentrations reached peak levels of 0.029% of the administered dose 1–2 h after gavage. Analysis of serum and ISF over the course of 24 h showed that kudzu phytochemicals were rapidly metabolized and eliminated. Excrement contained the greatest amount of the administered dose at 8.384% and 26.157% in the urine and feces, respectively (Table 4). The amount of ¹⁴C in the urine was likely underestimated, due to losses in the urine-collection funnel.

The radiolabeled, isoflavone-rich fraction accumulated an average of 0.016% of the ¹⁴C-dose in the brain, which is consistent with puerarin (and/or its metabolites) crossing the blood–brain barrier. ⁴² The liver accumulated 0.335% of the isoflavone-rich dose and accumulated the greatest quantity of radiolabel among tissues analyzed (Table 4), which is consistent with the distribution of ¹⁴C-labeled kudzu crude extract. ²⁷ Kidneys, lungs, heart, testes and spleen accumulated 0.029%, 0.012%, 0.004%, 0.042% and 0.004% of the administered dose, respectively.

Stomach, small intestine and large intestinal tissues contained 0.030%, 0.335% and 0.883% of the total dose, respectively, and the gastrointestinal contents contained a significant portion of the isoflavone-rich dose, at 0.042%, 1.306% and 11.821% in the stomach contents, small intestinal contents and large intestinal contents, respectively. In this study, a much greater percentage of the ¹⁴C-labeled dose was accounted for in the large intestinal contents and feces than was previously observed for the metabolic tracking of kudzu crude extract. ²⁷ This may be attributed to the freeze-drying of each sample and scintillation counting of a sample of dry mass rather than by scintillation counting of fecal slurry in the previous study, resulting in a more accurate measure of radiolabel in each sample.

The radiolabeled fraction also reached muscle and adipose tissues, accumulating an average of 0.011% and 0.019% of the total ¹⁴C dose in each gastrocnemius muscle and in the total abdominal adipose of each animal, respectively. All residual ¹⁴C in the carcass of each rat was successfully analyzed by scintillation counting of chemically liquefied bones and tissues. Each rat retained an average of 5.208% of the administered dose in the tissues and bones that were not removed for individual analysis.

All tissues accumulated less ¹⁴C-label when rats were gavaged with the radiolabeled isoflavone-rich fraction in the current study, than was observed when rats were gavaged with the radiolabeled crude kudzu extract in an earlier study. ²⁷ This may be attributed to the removal of non-target phytochemicals from the crude extract, including sugars, which would also have acquired radiolabel, and would have been readily absorbed and distributed throughout all animal tissues. Rats gavaged with the isoflavone-rich kudzu fraction accumulated 0.011%, 0.09% and 0.003% of the administered dose in each femur, tibia and vertebrae, respectively (Table 4). While these values were lower than those collected from the metabolic tracking of ¹⁴C-labeled kudzu crude extract, ²⁷ it was with greater certainty that kudzu isoflavones reached each target tissue and were available to exert an effect on bone health. This conclusion was further supported by HPLC–ESI–MS analysis of the methanolic extract from femur bones. A peak that eluted at retention time (t _R) 8.7 min in the UV chromatogram (262 nm) contained peaks in the MS spectrum with m/z (mass-to-charge ratio) 593 and 417 (Figure 7b). These peaks are indicative of the presence of puerarin (m/z 417) and puerarin glucuronide (m/z 593, 417) as observed by Prasain et al. ⁴¹ in organs of spontaneously hypertensive rats that were gavaged with puerarin. The presence of puerarin glucuronide in bone tissue is also consistent with the identification of glucuronidated metabolites of puerarin by Luo et al. ⁴³ in serum, liver and intestines of rats administered with intravenous puerarin. Two peaks at t _R 8.9 and at t _R 15.7 in the UV chromatogram (262 nm) corresponded to puerarin (Figure 7c, m/z 417) and daidzein (Figure 7d, m/z 255), respectively, as compared with the femur extract that was spiked with several isoflavone standards. Peaks that corresponded with isoflavone-metabolites in isoflavone-treated rat femur extracts were not present in sham-treated rat femur extracts.

OPEN IN VIEWER

Figure 7

HPLC–ESI–MS spectra showing HPLC-UV chromatogram (UV 262 nm) (a), and MS spectrum at t _R 8.7 (puerarin glucuronide; m/z 593, 417), t _R 8.9 (puerarin; m/z 417) and t _R 15.7 (daidzein; m/z 255) (b, c and d, respectively) of methanolic extract of rat femurs. HPLC–ESI–MS, high-performance liquid chromatography with electrospray ionization–mass spectrometry; UV, ultraviolet