Research

 

1) トランスポーターの機能解析と薬物動態への影響
  Effects of transporters on pharmacokinetics

 消化管に発現する排出トランスポーターは薬物の吸収性に大きく関与します.薬物の消化管吸収機構を明らかにするために考案された既存の方法には,消化管内の生理的状態を保ったままin vitro系で評価する手法や,小腸管腔側に発現する排出トランスポーターの機能を簡便に評価する実験系が少ないことが問題点でした.小腸から作製されるエンテロイドは,陰窩に含まれる幹細胞が種々の腸上皮細胞へ分化・増殖し,それら細胞同士が接着して自己組織化したオルガノイドです.エンテロイドは吸収上皮細胞以外の細胞や組織が混在する生理的状態をよく維持しており,実際の小腸を模倣した物質輸送が行われていると考えられるため,in vivo系に近いin vitro系として有用な物質輸送解析のツールになる可能性があります.そこで,エンテロイドを用いる新規の物質輸送実験系を確立して排出トランスポーターの機能評価へ応用することを目的として研究を進めています.本実験系は,蛍光顕微鏡で視覚的に変化を観察できるところも利点です.
 これまでの検討から,代表的排出タンパク質であるP-糖タンパク質(P-gp)の蛍光基質であるローダミン123がエンテロイド内腔へ蓄積する様子を観察でき,P-gp阻害薬を共存させることによりエンテロイド内腔への基質の移行率が大きく減少したことから,小腸管腔側に発現するP-gpの活性が阻害されことによりエンテロイドを形成する上皮細胞に蓄積した基質の細胞外への流出抑制が再現されたものと考えられました.これらの結果より,エンテロイドを用いてP-gpによる基質の輸送を測定することができ,本実験系が排出トランスポーターの機能を評価するのに有用であることが示唆されています.現在,3D解析を導入しさらなる解析精度の向上を目指しています.

  Efflux transporters expressed on the intestinal membrane contribute to drug absorption. A number of methods for clarifying the mechanism of drug absorption in the intestine in vitro have been developed. However, it is difficult to assess the function of efflux transporters in the intestine under physiological conditions in vitro. Therefore, we are conducting research to establish a novel experimental method using “enteroids” by which the function of efflux transporters in the intestine can be assessed. Enteroids are organoids prepared from stem cells in the small intestinal crypts and are known to be comprised of intestinal epithelial cell lineages. In addition to absorptive epithelial cells, enteroids maintain a mixed population of various cells and tissues, and the transport system in enteroids is therefore considered to be close to that in physiological conditions. Accordingly, we have been using enteroids to investigate the transport system in the small intestine in in vitro conditions that are close to in vivo conditions. This method has the advantage of enabling detection of the accumulation of a substrate for efflux transporters in the luminal space of enteroids by using a fluorescence microscope.
  P-glycoprotein (P-gp) is a representative efflux transporter expressed on the apical membrane of epithelial cells in the small intestine. We confirmed the accumulation of rhodamine 123 (Rh123), a fluorescent substrate for P-gp, in the luminal space when Rh123 was exposed to enteroids, suggesting that Rh123 was transported from the basolateral side to the apical side by P-gp. In contrast, this accumulation was suppressed by the addition of a P-gp inhibitor. Instead, we found that Rh123 accumulated in the epithelial cells in the presence of the P-gp inhibitor. These results indicate that enteroids can mimic the inhibition of substrate efflux from epithelial cells to the intestinal lumen by a P-gp inhibitor and that enteroids are a useful tool for assessment of the function of efflux transporters in the intestine. Recently, we have also been using 3D analysis and are currently addressing improvement of this method using enteroids by 3D analysis.

 

2) 吸収改善のための製剤学的検討
  Pharmaceutical studies for improvement of absorption

 近年,生活習慣病の増加や高度高齢社会の進行などにより,高騰する国民医療費の軽減を期待してセルフメディケーションが推奨されています.それに伴って,機能性食品の成分に関する研究が広く行われており,その成果を反映した多様な製品が開発されています.しかしながら,食品成分の体内動態、特に吸収動態の解析は詳細に行われていないものが多いのが現状です.私たちの研究室では,経口吸収率の良くない機能性食品成分の消化管吸収機構を解明し,それをもとに最適な剤形を設計することを目的として研究を進めています.また,従来法での難水溶性物質の吸収性評価は難しかったことから,吸収性の新たな評価方法としてリンパ移行量の測定方法を確立しました.さらに,併用すると吸収過程でトランスポーターなどの特殊な輸送系が関与する相互作用を生じる組み合わせを明らかにし,その回避策を提案しています.これらの結果は,医薬品と食品成分との相互作用の科学的な根拠を示すとともに,健康保持・増進のためのより効率的な摂取法の提案や日常的な健康食品成分の使用への有用な情報となると考えています..

  Preventive medicine and anti-aging medicine have received much attention recently due to increases in the proportion of elderly people in the population and patients with lifestyle diseases such as hypertension, dislipidemia and diabetes. Oxidative stress is involved in the onset of lifestyle diseases, and various antioxidant supplements and antioxidant-fortified functional foods have recently become available. Many epidemiological studies have shown relationships between consumption of polyphenol or carotenoid-rich foods and prevention of these lifestyle diseases. However, the mechanisms by which these components are absorbed have not been studied as much detail as the effects of these components. Absorption of food components and drugs from the gastrointestinal tract after oral administration is one of the important determinants of bioavailability, and the solubility of these components is thought to be a critical issue. Generally, poorly water-soluble components with poor membrane permeability classified into BCS (biopharmaceutics classification system) Class 4 show very low bioavailability. We have been investigating the absorption mechanisms of these components and we have prepared some pharmaceutical formulations for improvement of their absorption and have evaluated their absorption by various methods.
  Our research has shown possibility of undesirable food-food, drug-food or drug-drug interactions and how to avoid these interactions. Based on the evidence, our studies could also contribute to the establishment of an efficient intake method for these components and health promotion.

 

3) 加齢による薬物動態変動の定量的評価と投与設計
  Quantitative evaluation of the variation of pharmacokinetics with aging and optimization of dosing for elderly people

 ⼀般に加齢に伴い⽣理的機能が低下するため,⾼齢者への薬物療法はその機能を考慮した投与量の調節が求められます.しかしながら腎機能の指標であるクレアチニン・クリアランス(CCR)の推算式であるCockcroft-Gault式(C-G式)は,特に⾼齢者において推定値との乖離が指摘されています.そこで,当研究室では、C-G式を含む腎機能推定式により算出した推測値と実測CCRとの乖離状況の検討および,これら既存の腎機能推定式を,⾼齢者(特に後期⾼齢者)を多く含む集団を対象として補正することで腎機能の定量的評 価の精度向上を試みました.
 まず共同研究施設である砂川市⽴病院の患者を対象として,蓄尿によるCCR測定値(実測値)のデータのある患者の診療情報から,解析に必要な臨床検査値を収集しました.その結果,40-64歳の患者88名,64-74歳122名,75歳以上103名と⾼齢者のデータを⼗分確保できました.男⼥ともに40-64歳においては加齢に伴う実測CCRの変化はほぼ認められず,65歳以上を境にCCRの低下がはじまり,75歳以上の群で最も顕著に低下していることが明らかとなりました.また既存の腎機能推定式によるCCR推定値と実測値の乖離状況を検討した結果,C-G式および折⽥・堀尾の式ともに腎機能を過⼩評価しており,その乖離は加齢に伴い⼤きくなることが⽰されました.次にC-G式および折⽥・堀尾の式を本研究の対象集団を⽤いて補正した結果,乖離が特に⼤きかった75歳以上の群においても予測精度の⼤幅な向上がみられました.

  In general, physiological functions decrease with aging, and pharmacotherapy for people therefore needs adjustments of the drug dosages with consideration of their physiological function. The Cockcroft‐Gault equation (C-G equation) has been widely used for estimating renal function (creatinine clearance: CCR). However, it has been reported that estimated CCR values calculated by the C-G equation tend to deviate from the values measured in elderly people. Therefore, we compared the differences between measured CCR values and CCR values calculated by conventional renal function estimation equations such as the C-G equation. To try to improve the accuracy of these estimation equations, we modified the estimation equations by fitting parameters using data from a population that included many elderly patients.
  First, we collected clinical laboratory data for patients in whom CCR values were measured by the 24-hour urine collection method at Sunagawa City Hospital, a joint research facility. As a result, 88 patients aged 40 to 64 years, 122 patients aged 65 to 74 years and 103 patients aged 75 years or over were included in our study. Almost no change in measured CCR with aging was found in either males or females between the ages of 40 and 64 years, and CCR began to decline after the age of 65 years, and the decrease rate with aging was the sharpest in the group of 75 years or older. Examination of the deviation of estimated CCR values from measured CCR values showed that renal function was underestimated in both the C-G equation and the Orita/Horio equation and that the deviation increased with aging. We therefore modified the C-G equation and the Orita/Horio equation using patient data obtained in this study. As a result, a significant improvement in prediction accuracy was achieved even for the group aged 75 years or over, a group for which the deviation was particularly large.

 

4) 血小板の活性化による薬物動態変動を解析
  Effects of platelets on transporter activity

 血小板は、直接的および間接的に止血機構を制御する一方、病的な血栓の形成においても中心的な役割を担っています。また、血小板は活性化に伴い凝集すると共に、自身の顆粒中の物質を放出することが知られています。血小板内の顆粒には成長因子をはじめとした様々な生理活性物質が含有されており、これらを放出することにより血小板は周辺の細胞や組織に多様な生理作用を及ぼします。従来、炎症の誘導やがん細胞の増殖・転移を促進といった血小板のネガティブな作用が注目されてきましたが、肝再生や病原微生物の捕捉といったポジティブな作用についても報告されるようになってきました。さらに、近年では血小板が自己免疫疾患や非アルコール性脂肪肝炎の発症にも関与することが明らかとなり、止血や血栓の形成以外の血小板の多彩な作用が多方面で脚光を浴びています。しかしながら、血小板と薬物動態との関連に着目した検討はほとんど行われていません。我々は、薬物動態の中でも吸収への寄与が大きいトランスポーターに着目し、血小板によりトランスポーターの輸送活性が変動するかを解析しています。

  Platelets play a central role not only in physiological hemostasis but also in pathological thrombosis. Activated platelets are known to achieve their roles by aggregating and releasing various bioactive substances such as growth factors, lysophospholipids and chemokines. Accordingly, platelets exert various physiological actions on surrounding cells and tissues. Several studies have demonstrated negative effects of platelets such as promotion of inflammation, cancer cell proliferation and metastasis. However, positive effects of platelets on liver regeneration and on trapping of pathogens have also been reported. Furthermore, recent studies have revealed that platelets are also involved in the development of autoimmune diseases and nonalcoholic steatohepatitis. Thus, numerous actions of platelets other than hemostasis and thrombosis are in the spotlight in various fields. However, there have been few studies in which the relationships between platelets and pharmacokinetics were investigated. Among pharmacokinetics, drug absorption is strongly associated with the activity of transporters. We have been focusing on transporters and have been examining whether activity of transporters is potentiated by platelets.

 

5) データマイニングを用いた副作用発現リスク推定モデルの開発
  Construction of a risk prediction model for adverse drug reactions using the data mining method

 データマイニングはデータの中から隠れた規則性を発見し,重要な知識を見出すための手法です.その中でもDecision tree(DT)モデルは予測・判別モデルの一つであり,マーケティング等の分野において,顧客の背景因子から行動を予測するなどの目的で用いられています.本モデルはフローチャート状の構造を有し,従属変数と関連の強い順に独立変数が配置されます.利用者はそれぞれの変数に関わる条件に対してYes/Noで順次回答していくことで,複数の要因の組み合わせによる相互関係を考慮したイベントの発現割合を定量的に評価可能となります.
 我々はこのDT分析の副作用発現リスク推定への応用可能性に着目し,各種検討を実施しています.まず,抗MRSA薬であるバンコマイシン(Vancomycin;VCM)による腎機能障害の発現割合推定モデルを構築し,その精度が従来の手法であるロジスティック回帰モデルと同等であることを明らかにしました(Imai S, et al. j eval clin pract, 2017).次に,モデル使用タイミング(初回TDM[Therapeutic Drug Monitoring]時)と対象患者(非複雑性感染症患者)を具体的に設定することで,より臨床において活用可能なVCMによる腎機能障害発現割合推定モデルの構築に成功しました(Imai S, et al. j eval clin pract, 2019).さらに,抗ウイルス薬であるガンシクロビル誘発性好中球減少症ならびに血小板減少症の発現割合推定モデルを構築し,構築されたモデルが臨床的に妥当であることを証明しました(今井俊吾ら,医薬品情報学,2019;今井俊吾ら,医療薬学,2017).
 これまでの研究成果に基づき,現在はより一般化可能な副作用発現リスク推定モデル構築を目指し,多施設共同研究を実施しています.

  Data mining is a method of predictive analysis in which a large amount of data is explored it to discover underlying patterns and relationships in complex datasets and it enables the construction of predictive models. Decision tree (DT) analysis, which is a flowchart-like tree framework, is a typical technique for data mining. A DT model is commonly used to establish classification systems based on multiple covariates or to develop prediction algorithms for a target variable. It classifies a population into segments, i.e., “branches”, that form an inverted tree.
  We have been conducting various studies with focus on the applicability of DT analysis to prediction of the risk of adverse reactions. First, we have constructed a risk prediction model for vancomycin (VCM)-induced nephrotoxicity and have shown that the accuracy of this model is equivalent to the accuracy of conventional statistical analysis using a logistic regression model (Imai S, et al. j eval clin pract, 2017). We then successfully constructed a useful DT model for predicting nephrotoxicity in clinical practice by setting the “optimal timing and subjects” to be "the time when initial therapeutic drug monitoring (TDM)" and "patients with a target trough level within 10.0–15.0 mg/L" (Imai S, et al. j eval clin pract, 2019). We also constructed risk prediction models for ganciclovir-induced neutropenia and thrombocytopenia and showed these clinical validity of the models (Imai S, et al. Jpn. J. Pharm. Health Care Sci, 2017 and Imai S, et al. Jpn. J. Drug Inform, 2017).
  Based on the results of research obtained so far, we are currently conducting a multi-center study to construct a more generalizable model of risk prediction for adverse drug reactions.