Background: Gut microbiota has demonstrated tremendous metabolic capability in metabolizing endogenous compounds as well as xenobiotics. Comparing to the well-characterized host drug-metabolizing enzyme systems, gut microbiome drug metabolism is relatively unexplored. β-Glucuronidase and β-glucosidase, the dominant metabolic enzymes among most of the gut microbial enzymes, could highly affect the oral bioavailability and systemic exposure of drugs. So far, most in vitro gut bacterial metabolism carried out with fecal slurries or intestinal bacteria suspensions, which has inherited disadvantages in dynamically altered gut microbiota compositions during in vitro anaerobic incubation. Ginsenosides, the major active components of the traditional Chinese medicine ginseng, have shown multifaceted bioactivity. Deglycosylation by microbial β-glucosidases and the hydrolysis in gastric fluid were revealed to be important factors for poor bioavailability of ginsenosides. In present study, we aimed to explore a stable in vitro microbial drug metabolism system with gut bacterial protein. With this in vitro microbial metabolism reaction system, the regioselectivity of the microbial deglycosylation of ginsenosides was characterized in pooled human gut bacterial proteins. This result, together with the data obtaining from the degradation of ginsenosides in simulated gastric fluid, offers an explaination of the low bioavaiablity of the ginsenosides. Methods: One-step active bacterial protein extraction, sonication in PBS, and sonication with lysozyme, lysozyme extraction were carried out to compare the efficiency of these methods in extracting gut bacterial protein by measuring the β-glucuronidase and β-glucosidases activities. The β-glucuronidase and β-glucosidases activities were determined by measuring the formation of 4-nitrophenol from 4-nitrophenyl-β-D-glucuronide and 4-nitrophenyl-β-D-glucopyranoside spectrophotometrically, respectively. The optimal gut bacterial protein solution preparation method was chosen by comparing the enzyme activities by different extraction, or the addition of metal ion or protease inhibitor. Stability of the gut bacterial suspension and protein solution were studied by determining the enzyme activities at different cryopreserved time. Then, the method was applied to study β-glucuronidase and β-glucosidases activities in fecal samples of different origins. Six PPD-type or PPT-type ginsenosides monosides and biosides were incubated with human gut bacterial proteins and the parent compounds and metabolites were determined on a 4000 QTRAP® LC/MS/MS System. The effects of time and substrate concentration on deglycosylation of these ginsenosides were examined. On the other hand, the stability of 12 ginsenosides, as mixed standards or in a Ginseng extract, in simulated gastric fluid was quantitatively determined in both positive and negative modes by HPLC-MS/MS. Results: The gut bacterial protein was better extracted by 10-15min sonication which showed high efficiency in extracting β-glucuronidase and β-glucosidase with the yield of the bacterial protein proportional to the enzyme activity. The enzyme activities were not affected by adding metal ions or protease inhibitor. Cryopreservation of gut bacterial suspensions in PBS system keeps the enzyme activity stable within 9 months tested. In gut bacterial proteins, the enzyme activities were stable within 2-month cryopreservation and after 10 freeze-thawing rounds. The β-glucuronidase activity per unit protein in gut bacterial proteins from different species was in the following descending order: mouse > rat > human > monkey, and that of the β-glucosidase activity was human > rat > mouse > monkey. The deglycosylation of the monosides of both PPD and PPT-type ginsenosides followed typical kinetics, while that of the bioside PPD-type and PPT-type ginsenoside followed sigmoidal and biphasic kinetics, respectively. The kinetics of enzyme catalyzed reaction showed the preference of deglycosylation of the 3 PPD-type ginsenosides: C-3 in bioside > C-3 in monoside > C-20 in monoside > C-20 in bioside. The deglycosylation activity of the monosides was in the following order: C-3 > C-20 of PPD-type > C-20 of PPT-type > C-6. While in the biosides, the deglycosylation activity follows the following order: C-3 > C-6 > C-20 of PPT-type > C-20 of PPD-type. In simulated gastric fluid, the more polar PPD-type multi-glycosides and the PPT-type ginsenosides containing C-20 glycosyl were unstable at low pH of the gastric acid, while the PPT-type ginsenosides containing C-6 glycosyl without C-20 glycosyl was relatively stable in gastric acid. Conclusion: The present study characterized, for the first time, the gut bacterial proteins as an in vitro system for microbial drug metabolism. Using this in vitro metabolic reaction system, we found the regioselectivity of deglycosylation of ginsenosides in gut bacterial proteins. This regioselective deglycosylation of ginsenosides, together with that observed in simulated gastric fluid offer a better understanding of the poor oral bioavailability of ginsenoside and provided an important basis for clinical applications of ginsenosides.

背景：腸道菌是生活在宿主腸道內的所有微生物，能夠代謝多種內源性和外 源性化合物。腸道菌中含有多種藥物代謝酶，β-葡萄糖醛酸苷酶和β-葡萄糖苷 酶是其中重要的代謝酶，能影響多種藥物的口服生物利用度以及系統暴露水平。 然而，與宿主自身藥物代謝系統的研究相比較，目前對於腸道菌藥物代謝系統的 研究較為粗放，多採用糞便或腸道菌總菌的體外培養體系，存在嚴重雜質干擾、 腸道菌組成隨培養時間延長而發生動態改變，不能準確反映體內腸道菌代謝藥物 的真實情況。因此，亟需建立穩定可控的體外反應體系用於研究腸道菌群對藥物 的代謝。此外，作為中國傳統著名中藥人參，三七等人參屬藥物中的活性成分， 人參皂苷具有廣泛的生物活性，但較低的生物利用度成為進一步研發人參皂苷類 化合物的瓶頸。研究表明，口服人參皂苷後其在胃液中以及腸道菌的水解可能是 影響其生物利用度的重要因素，腸道菌代謝酶對底物的結構選擇性，可能是導致 不同人參皂苷的生物利用度不同的原因之一。本研究旨在建立腸道菌蛋白體外代 謝反應體系，為研究藥物的腸道菌代謝提供快速穩定的研究平臺；在此基礎上， 研究比較人參皂苷脫糖基反應在人腸道菌蛋白以及人工胃液作用下的結構選擇 性，為人參皂苷合理利用提供重要的數據。 方法：採用一步法蛋白提取試劑，超聲提取法以及溶菌酶提取法對腸道菌蛋 白進行提取，並用探針底物測定腸道菌的β-葡萄糖醛酸苷酶和β-葡萄糖苷酶的 酶活性，比較不同提取方法、加入不同金屬離子或蛋白酶抑制劑對酶活性的影 響；同時研究了腸道菌蛋白直接低溫凍存，或以 PBS 或甘油 PBS 體系凍存腸道 菌等不同條件下酶活性的穩定性；在此基礎上，比較了不同種屬來源、正常/病 理狀態、有/無服用中藥等條件下的腸道菌的酶活性差異。通過比較人腸道菌蛋 白對單糖或二糖連接的 PPD 或 PPT 型人參皂苷在不同孵育時間以及底物濃度下 的脫糖基速率和酶動力學參數，分析人腸道菌蛋白對人參皂苷的脫糖基的位點選 擇性。人工胃液水解人參皂苷定量分析分別採用 HPLC-MS/MS 負離子以及正離 子檢測。 結果：比較人腸道菌蛋白提取方法（一步法菌蛋白提取、超聲法、加溶菌酶 超聲法、溶菌酶提取法），發現 10-15 min 內超聲提取法提取的腸道菌蛋白的β -葡萄糖醛酸苷酶以及β-葡萄糖苷酶的酶活性較高，且提取液的蛋白濃度與酶活性呈現很好的線性，不需要加入金屬離子或蛋白酶抑制劑，因此確定超聲提取法 為提取腸道菌活性蛋白的較優方法。選擇 PBS 溶液作為腸道菌混懸液的凍存體 系能較好的維持酶活性，直接凍存人腸道菌蛋白，酶活性在 2 個月維持穩定，反 復快速凍融 10 次以內，對β-葡萄糖醛酸苷酶以及β-葡萄糖苷酶的活性基本沒 有影響。比較不同動物種屬正常菌群的酶活性，發現種屬差異較為顯著，單位蛋 白的β-葡萄糖醛酸苷酶：小鼠>大鼠>人>獼猴，單位蛋白的β-葡萄糖苷酶：人> 大鼠>小鼠>獼猴。DSS 誘導的急性潰瘍結腸炎大鼠腸道菌的β-葡萄糖醛酸苷酶 的活性增高；口服給予受試中藥會降低健康獼猴的腸道菌β-葡萄糖醛酸苷酶活 性。不同的人參皂苷的代謝在人腸道菌蛋白中的代謝呈現不同的類型。PPD 以及 PPT 型人參皂苷單糖苷符合經典的米氏方程，而 PPD 型和 PPT 型人參皂苷二糖 苷分別符合自我活化類型和雙相代謝類型。人腸道菌蛋白對人參皂苷脫糖基呈現 一定的結構選擇性，總體而言，腸道菌蛋白對 PPD 型人參皂苷的脫糖基活性遠 高於 PPT 型人參皂苷。PPD 型人參皂苷脫糖基的優先順序是二糖苷 C-3 位>單糖 苷C-3位>單糖苷C-20位>二糖苷C-20位。人參皂苷單糖苷的脫糖基是C-3 > PPD 型C-20 > PPT型C-20 > C-6，而二糖苷脫糖基活性順序是C-3 > C-6 > PPT型C-20 > PPD 型 C-20。人參皂苷在人工胃液中的水解研究數據顯示，大極性多糖苷連 接的 PPD 型人參皂苷極不穩定，人工胃液中含有 C-20 位糖基的多糖苷連接的 PPT 型人參皂苷極不穩定，而含 C-6 位糖基的 PPT 型人參皂苷在人工胃液中較 穩定，不容易被水解。 結論：本研究第一次報導了以腸道菌蛋白作為腸道菌體外代謝藥物的研究體 系，觀察了提取、凍存、金屬離子、蛋白酶抑制劑等對代謝酶活性的影響，並初 步比較了不同來源的腸道菌蛋白的β-葡萄糖醛酸苷酶以及β-葡萄糖苷酶的活 性，為腸道菌代謝藥物研究提供了重要研究平臺。利用該體外腸道菌代謝研究體 系，較系統地研究了人腸道菌蛋白對人參皂苷的脫糖基作用，並與人參皂苷在人 工胃液中的非酶催化的脫糖基反應相結合，更好地解釋了人參皂苷（尤其是 PPD 型人參皂苷）口服生物利用度低的現象，為進一步合理利用人參皂苷提供了重要 數據。關鍵词： 腸道菌群，β-葡萄糖醛酸苷酶，β-葡萄糖苷酶，人參皂苷，代谢，水解， 結構選擇性，人工胃液，液相-串聯質譜