Informatics framework of traditional Sino-Japanese medicine (Kampo) unveiled by factor analysis
Kampo, an empirically validated system of traditional Sino-Japanese medicine, aims to treat patients holistically. This is in contrast to modern medicine, which focuses in principle on treating the affected parts of the body of the patient. Kampo medicines formulated as combinations of crude drugs are prescribed based on a Kampo-specific diagnosis called Sho (in Japanese), defined as the holistic condition of each patient. Therefore, the medication system is very complex and is not well understood from a modern scientific perspective. Here, we show the informatics framework of Kampo medication by multivariate factor analysis of the elements constituting Kampo medication. First, the variation of Kampo formulas projected by principal component analysis (PCA) indicated that the combination patterns of crude drugs were highly correlated with Sho diagnoses of Deficiency and Excess. In an opposite way, partial least squares projection to latent structures (PLS) regression analysis could also predict Deficiency/Excess only from the composed crude drugs. Secondly, to chemically verify the correlation between Deficiency/Excess and crude drugs, we performed mass spectrometry (MS)-based metabolome analysis of Kampo prescriptions. PCA and PLS regression analysis of the metabolome data also suggested that Deficiency/Excess could be theoretically explained based on the variation in chemical fingerprints of Kampo medicines. Our results show that factor analysis of Kampo concepts and of the metabolomes of Kampo medicines enables interpretation of the complex system of Kampo. This study will theoretically form the basis for establishing traditionally and empirically based medications worldwide, leading to systematically personalized medicine.
KeywordsKampo Sho Factor analysis Metabolome analysis
Abbreviations of Kampo formulas
With a history dating back 1,600 years, Kampo is a traditional Japanese medicine of therapeutic strategies and diagnostics adapted from traditional Chinese medicine (TCM). Kampo medicines are prescribed and generally prepared with combinations of crude drugs derived from herbs, animals, and minerals. Curative effects are based on synergism between pharmacologically and biologically active constituents producing minimum side-effects. Currently, 254 regulated crude drugs, as well as their processed forms, and 22 Kampo medicines are listed in the Japanese Pharmacopoeia, 16th edition , and 294 over-the-counter prescriptions were approved in 2012. A revival of Kampo has been accompanied by scientific re-evaluation of its relevance to modern healthcare [2, 3].
Recently we proposed that informatics based on the comprehensive and simultaneous analysis of numerous factors may be clarified to classify the complex relationships among crude drugs and formulas, chemical constituents, and pharmacological and biological effects in traditional and modern medicine (reviewed in Refs. [7, 8, 9] ). In the present study, we investigated the relationship between the diagnostic criteria Sho and the prescriptions of Kampo medicines based on statistical factorial analysis. Specifically, this study focused on the complex correlation between the combination of patterns of crude drugs in formulating Kampo medicines and a Sho diagnosis of Deficiency/Excess. To systematically understand and interpret the theory of empirical medication in Kampo, multivariate statistics including principal component analysis (PCA) and partial least squares projection to latent structures (PLS) modeling were applied to the correlation between crude drug patterns and Deficiency/Excess. Metabolome analysis, a comprehensive and global chemical analysis of metabolites contained in samples of decoctions of Kampo prescriptions actually prepared from mixtures of crude drugs was also incorporated using mass spectrometry (MS), to substantiate the relationships between Kampo formulas and Deficiency/Excess using the chemical fingerprints of Kampo prescriptions based on the similarities and differences among their chemically complex features. In this study, we begin to unveil the complex system of Kampo medication.
Materials and methods
Kampo formulas analyzed in this study are listed in the KAMPO section of the KNApSAcK family database [8, 10] and in several Kampo reference texts [11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25]. Medicinal resources of crude drugs used in Kampo formulas are listed in two references texts for crude drugs [1, 26] and are described along with their scientific names and medicinally active region in Table S1. Kampo formulas are introduced using the structured Romanized notation recommended by The Japan Society of Oriental Medicine (Tokyo, Japan), and are abbreviated in subsequent appearances. For metabolome analysis, 25 Kampo prescriptions containing Cinnamon bark (Cinnamomi Cortex), as well as 9 other prescriptions, were selected from Refs [11, 13].
Preparation of decoctions of Kampo prescriptions for metabolome analysis
Crude drugs for the preparation of 34 Kampo prescriptions (Table S2) were purchased from Uchida Wakanyaku (Tokyo, Japan) and Tsumura (Tokyo, Japan). The decoctions for metabolome analysis were prepared in accordance with the standard method clinically used at the Diagnosis and Treatment Department of Kampo Medicine, Chiba University Hospital (Chiba, Japan) as follows. Kampo prescriptions tested were packed in an L-size filter bag (Uchida Wakanyaku). The packed prescription was boiled with 600 mL of water for 60 min by using a decocting pot with an electric heater (HMJ3-1000 W; Uchida Wakanyaku).
Acquisition of chemical fingerprints by MS-based metabolome analysis of Kampo prescriptions
High-accuracy quadrupole time-of-flight (Q-TOF)–MS analysis by direct infusion was performed for acquisition of the chemical fingerprints of Kampo prescriptions using a Q-TOF mass spectrometer (Q-TOF micro™ Mass Spectrometer; JASCO International, Tokyo, Japan) and the resolution was set at 5,000. Ionization of the analyzed samples was performed by positive electrospray ionization (ESI) at m/z range of 85–1,200. The flow rates of cone and desolvation N2 gases were set at 50 and 500 L h−1, respectively. In the positive ESI source, capillary and sample cone voltages were set at 2,800 and 30 V, respectively, with the desolvation and source temperatures set to 150 and 100 °C, respectively. The ion energy was set at 2.0 V in the Q setting. TOF flight tube and tube lens voltages were set at 5,630 and 90 V, respectively, and a microchannel plate detector was set at 2,400 V.
The spectral intensity in MS analysis was acquired under the unsaturated condition of peak detection. Test samples were diluted 100-fold with water and injected by syringe pump at a constant rate. The basal conditions of Q-TOF–MS analysis were calibrated by measured values of sodium formate solution (0.1 % formic acid and 5 mM NaOH/90 % acetonitrile). MS data were collected at a scanning rate of 1.0 s per one data scan with 0.1 s of internal delay. In addition, leucine-enkephalin, of which the m/z value in positive ESI analysis was 556.277 ([M + H]+), was analyzed as a reference compound for calibration of the measured values using a syringe pump with a lock spray module, permitting one data scan every 5.0 s.
Processing of metabolome data for multivariate analysis
A metabolomics data matrix for multivariate analysis was constructed by m/z values and peak intensities of mass spectra (Table S3). Data points of m/z values were reduced by integration of the peak intensities per m/z range of 0.5 (m/z: \(85.0,85.5,86.0 \ldots 1199.0,1199.5\)). Accordingly, a data matrix consisting of 2,230 m/z values and the corresponding peak intensities was generated. The acquired metabolome data were applied to PCA and PLS regression analysis.
PLS regression analysis
Results and discussion
Factorial analysis of Kampo formulas
The scores of PCA assigned to Kampo formulas prescribed for Deficiency (382 formulas), Middle (218 formulas), and Excess (226 formulas), were projected onto a two-dimensional space (Fig. 2). In this unsupervised analysis, a discriminant boundary between the formulas prescribed for Deficiency and Excess was broadly observed in both axes (dotted line in Fig. 2). The formulas prescribed for Middle also formed a cluster at a nearby discriminant boundary between the formulas for Deficiency and Excess. The contribution of crude drugs for variation of formulas was indicated by the factor loading in PCA. For example, Pueraria root (Puerariae Radix) (0.673) and apricot kernel (Armeniacae Semen) (−0.293) in PC 24, and fennel (Foeniculi Fructus) (0.041) and ass-hide glue (Asini Gelatinum) (−0.252) in PC 8 contributed to the variation of formulas in each degree. These results suggested that the prescribing theory of Kampo medicines may be considered from the point of view of the formulations and quantities of the crude drugs composing the medicines.
Linear regression coefficients b of PLS regression analysis assigned to crude drugs composing Kampo formulas prescribed for Deficiency and Excess
b Value (<−0.1)
No. of formulas
Japanese angelica root
Rehmannia root (steamed)
b Value (>0.1)
No. of formulas
Saposhnikovia root and rhizome
Japanese cherry bark
34 Kampo prescriptions analyzed by direct infusion Q-TOF–MS for metabolomic analysis
Patient constitution according to Sho
Thus, PLS regression analysis distinguished between Kampo formulas for Deficiency and Excess as well as did PCA (Fig. 2), and enabled an assessment of the contributions of crude drugs to those groups.
Metabolome analysis of Kampo prescriptions
In conclusion, systematic and comprehensive interpretation of Kampo medication was achieved by integration of the accumulated informatics. The factor analyses simultaneously processed a number of Kampo medicines, enabling comprehensive classification and correlation analysis of multiple factors; namely, Sho diagnosis of Deficiency/Excess, Kampo formulas, and the crude drugs. This finding and suggestion could be chemically substantiated by the metabolome analysis of Kampo prescriptions. The approaches in this study could lead to systems medicine, a comprehensive analysis of correlations between ingredients and practices in traditional (empirically verified) and modern medicines . Factor and correlation analyses may lead to generalization of the diagnostic criteria and prescribing methods of Kampo, and be applied to other traditional medicines such as TCM and Jamu (an Indonesian traditional medicine). Moreover, Kampo medicines more suitable for the health of people in the modern age may be proposed from the new information assembled by systems medicine. Practically, TCM tends to create new combinations of crude drugs for therapeutic use, while Kampo medicine is often dependent on traditional prescriptions . We believe that factor analysis of a complex medical system by informatics will have implications for systems medicine.
This study was partially funded by a research grant from the Japanese Ministry of Health, Labour and Welfare (Grant Number 24202001) and a Grant-in-Aid for Young Scientists (B) from the Japan Society for the Promotion of Science (Grant Number 25860085).
Conflict of interest
The authors declare that they have no conflict of interest.
- 1.Ministry of Health, Labour and Welfare of Japan (2011) The Japanese pharmacopoeia, 16th edn. Ministry of Health, Labour and Welfare of Japan, JapanGoogle Scholar
- 3.Watanabe K, Matsuura K, Gao P, Hottenbacher L, Tokunaga H, Nishimura K, Imazu Y, Reissenweber H, Witt CM (2011) Traditional Japanese Kampo medicine: clinical research between modernity and traditional medicine—the state of research and methodological suggestions for the future. Evid Based Complement Altern Med 2011:513842Google Scholar
- 4.Otsuka K (2010) Kampo: a clinical guide to theory and practice. Churchill Livingstone/Elsevier, PhiladelphiaGoogle Scholar
- 5.Terasawa K (1993) Kampo, Japanese-oriental medicine: insights from clinical cases. Standard McIntyre, TokyoGoogle Scholar
- 8.Okada T, Afendi FM, Katoh A, Hirai A, Kanaya S (2013) Multivariate analysis of analytical chemistry data and utility of the KNApSAcK family database to understand metabolic diversity in medicinal plants, chap 18. In: Chandra S, Lata H, Varma A (eds) Biotechnology for medicinal plants: micropropagation and improvement. Springer, Berlin, pp 413–438CrossRefGoogle Scholar
- 9.Okada T, Katoh A (2011) Metabolomics: data collection and analysis, chap 27. In: Cseke LJ, Kirakosyan A, Kaufman PB, Westfall MV (eds) Handbook of molecular and cellular methods in biology and medicine, 3rd edn. Taylor & Francis Group (CRC Press), London, pp 471–484Google Scholar
- 10.Afendi FM, Okada T, Yamazaki M, Hirai-Morita A, Nakamura Y, Nakamura K, Ikeda S, Takahashi H, Altaf-Ul-Amin M, Darusman LK, Saito K, Kanaya S (2012) KNApSAcK family databases: integrated metabolite-plant species databases for multifaceted plant research. Plant Cell Physiol 53:e1CrossRefPubMedGoogle Scholar
- 11.Chiba Koho Kampo Kenkyukai (1993) Kampo houi note. Maruzen, Tokyo (in Japanese) Google Scholar
- 12.Fujihira K (1972) Jitsuyou Kampo ryouhou–gendai igaku no mouten wo tsuku. Hokendohjin-Sha, Tokyo (in Japanese) Google Scholar
- 13.Nakamura K (2004) Wakanyaku houi jiten. Midori-Shobou, Tokyo (in Japanese) Google Scholar
- 14.Nishiyama H (1969) Kampo igaku no kiso to shinryou. Sogen-Sha, Osaka (in Japanese) Google Scholar
- 15.Okuda K (1973) Kampo koho yoho kaisetsu. Ido-no-Nippon-Sha, Yokosuka (in Japanese) Google Scholar
- 16.Otsuka K (1963) Shokou ni yoru Kampo chiryo no jissai. Nanzan-Do, Tokyo (in Japanese) Google Scholar
- 17.Otsuka K, Yakazu D, Kiga R (1966) Keiken Kampo shoho bunryo shu. Ido-no-Nippon-Sha, Yokosuka (in Japanese) Google Scholar
- 18.Otsuka K, Yakazu D, Shimizu T (1969) Kampo shinryo iten. Nanzan-Do, Tokyo (in Japanese) Google Scholar
- 19.Takahashi S, Nishioka K (1966) Meikai Kampo shoho. Naniwa-Sha, Osaka (in Japanese) Google Scholar
- 20.Takeda T, Yoshikawa T, Takahashi K, Saito K (2006) Tennen iyaku shigen-gaku, 2nd edn. Hirokawa-Shoten, Tokyo (in Japanese) Google Scholar
- 21.Tatsuno K (1973) Kampo shoho-shu. Chugoku Kampo Igakusho Kankokai, Japan (in Japanese) Google Scholar
- 22.Tatsuno K (1974) Kampo nyumon kouza. Chugoku Kampo Igakusho Kankokai, Japan (in Japanese) Google Scholar
- 23.Yakazu D (1959) Kampo goseiyouho kaisetsu. Ido-no-Nippon-Sha, Yokosuka (in Japanese) Google Scholar
- 24.Yakazu D (1966) Rinsho ouyo Kampo shoho kaisetsu. Sogen-Sha, Osaka (in Japanese) Google Scholar
- 25.Yamada K (1972) Kampo shoho ohyo no jissai, 3rd edn. Nanzan-Do, Tokyo (in Japanese) Google Scholar
- 26.Research Institute for Wakan-Yaku of Toyama Medical and Pharmaceutical University, Nanba T (2007) Wakan-yaku no jiten, the new format edn. Asakura-Shoten, Tokyo (in Japanese)Google Scholar
- 27.Johnson RA, Wichern DW (1982) Applied multivariate statistical analysis. Prentice-Hall, NJGoogle Scholar
- 28.Afendi FM, Ono N, Nakamura Y, Nakamura K, Darsman LK, Kibinge N, Morita AH, Tanaka K, Horai H, Altaf-Ul-Amin M, Kanaya S (2013) Data mining methods for omics and knowledge of crude medicinal plants toward big data biology. Comput Struct Biotechnol J 4:e201301010PubMedCentralCrossRefPubMedGoogle Scholar
- 29.Ishibashi A, Kosoto H, Ohno S, Sakaguchi H, Yamada T, Matsuda K (2005) Genaral introduction to Kampo. In: Sato Y, Hanawa T, Arai M, Cyong J, Fukuzawa M, Mitani K, Ogihara Y, Sakiyama T, Shimada Y, Toriizuka K et al, The Japan Society for Oriental Medicine (eds) Introduction to KAMPO—Japanese traditional medicine. Elsevier, Tokyo, pp 2–13Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.