Weijun
Kong
ab,
Jiabo
Wang
b,
Xiaohe
Xiao
*b,
Shilin
Chen
a and
Meihua
Yang
*a
aInstitute of Medicinal Plant Development, Chinese Academy of Medical Sciences, Peking Union Medical College, Beijing, 100193, China. E-mail: kongwj302@126.com; Tel: +86 10 66933322 Tel: +86 10 57833277
bChina Military Institute of Chinese Materia Medica, 302 Military Hospital of China, Beijing, 100039, P.R. China
First published on 7th November 2011
In this study, the antibacterial effect and mode of Coptidis rhizoma on Escherichia coli was evaluated by microcalorimetry coupled with chemometric techniques. Using an isothermal microcalorimeter, the metabolic profiles of E. coligrowth at 37 °C affected by 15 batches of C. rhizoma were measured. Through principal component analysis (PCA) on nine quantitative thermo-kinetic parameters obtained from the metabolic power–time profiles of E. coli, the antibacterial effects of C. rhizoma from various sources could be easily evaluated by analyzing the change of the two main thermo-kinetic parameters, growth rate constant k2 and maximum heat-output power P2m, in the second exponential phase of E. coligrowth. Then, hierarchical clustering analysis (HCA) was carried out on the two parameters to distinguish those C. rhizoma samples in respect to their antibacterial effects. Clear results were obtained to show that all 15 C. rhizoma samples with different antibacterial effects could be successfully grouped in accordance with their origins. Ranked in decreasing order, the antibacterial mode of C. rhizoma samples that were from Sichuan province had the strongest antibacterial effects, followed by samples from Chongqing city and Hubei province. Our results revealed that the developed microcalorimetry with chemometric techniques had the potential perspective for evaluating the effect and mode of Coptidis rhizoma and other Chinese materia medicas.
As we know, the quality evaluation of CMM should consist of two aspects. One is the chemical analysis including qualitative and quantitative determination of one or several high-content components and fingerprinting analysis of different samples,8 which has been introduced and accepted by the WHO,9 and State Food and Drug Administration (SFDA) of China.10 The other is the evaluation of pharmacodynamic action. Some reports have demonstrated the quality differences of C. rhizoma by chemical analysis.11 So the evaluation of pharmacodynamic action including antimicrobial effects of this CMM should be studied deeply and widely by useful and sensitive methods.12,13 To our knowledge, no previous studies have been published on evaluating the antibacterial mode of C. rhizoma.
Microcalorimetry, with its high sensitivity, high accuracy and ability of automation, has been used widely for evaluating the antimicrobial effects of many drugs and compounds.14–16 From the metabolic profile of microbial growth affected by various substances, some important qualitative and quantitative information could be obtained to detect and analyze the changes of the total metabolic state of microbes due to patho-physiological stimuli, further to present the antimicrobial effects of these substances. Nowadays, metabolic profile is being used to evaluate the effects of these substances and has been an attractive candidate for mode-of-action studies.17
In this study, the antibacterial effect and mode of 15 batches of C. rhizoma on Escherichia coli (E. coli) growth were evaluated by microcalorimetry. And the metabolic profile of E. coli affected by C. rhizoma from various sources was acquired. Then, chemometric techniques including principal component analysis (PCA) and hierarchical clustering analysis (HCA) were carried out on the quantitative parameters obtained from the metabolic profile of E. coli to distinguish samples according to their antibacterial effects; further, the possible antibacterial mode of C. rhizoma was explored.
Sample no. | Sources | Origins | Harvesting time | Description |
---|---|---|---|---|
A | Shizhu, Sichuan | Coptis chinensis Franch. | February 2007 | Dried |
B | Shizhu, Sichuan | Coptis chinensis Franch. | August 2007 | Dried |
C | Shizhu, Sichuan | Coptis chinensis Franch. | December 2007 | Dried |
D | Emei, Sichuan | Coptis chinensis Franch. | August 2007 | Dried |
E | Dayi, Sichuan | Coptis chinensis Franch. | November 2006 | Dried |
F | Dayi, Sichuan | Coptis chinensis Franch. | August 2007 | Dried |
G | Pengxian, Sichuan | Coptis chinensis Franch. | July 2006 | Dried |
H | Chengkou, Sichuan | Coptis chinensis Franch. | February 2007 | Dried |
I | Wuxi, Sichuan | Coptis chinensis Franch. | August 2007 | Dried |
J | Huangshui, chongqing | Coptis chinensis Franch. | February 2007 | Dried |
K | Huangshui, chongqing | Coptis chinensis Franch. | August 2007 | Dried |
L | Enshi, Hubei | Coptis chinensis Franch. | August 2007 | Dried |
M | Lichuan, Hubei | Coptis chinensis Franch. | August 2007 | Dried |
N | Fangxian, Hubei | Coptis chinensis Franch. | July 2007 | Dried |
O | Laifeng, Hubei | Coptis chinensis Franch. | July 2007 | Dried |
Strain E. coli (CCTCC AB91112) was provided by China Center for Type Culture Collection, Wuhan University, Wuhan, China. It was inoculated in Luria–Bertani (LB) culture medium which contained per 1000 mL (pH 7.2–7.40): 10 g peptone, 5 g yeast extract and 5 g NaCl and was sterilized in high pressure (0.1 MPa) steam at 121 °C for 30 min.
Fig. 1 Metabolic power–time profiles of E. coligrowth (a) without drug, and affected by C. rhizoma samples of (b) 9 from Sichuan province, (c) 2 samples from Chongqing city and 4 from Hubei province. The final concentration of all C. rhizoma samples were 9.0 mg mL−1. |
The metabolic power–time profiles of E. coligrowth in Fig. 1(a) showed that the bacterial growth was an exponential model in the growth processes giving the following equation:20
Pt = P0 exp(kt) or ln Pt = ln P0 + kt | (1) |
For the equivalency of the eight channels, an E. coli suspension without the drug was added into each channel in triplicate experiments. The reproducibility of the method was evaluated by running sample B of C. rhizoma in triplicate. As shown in Table 2, the R values of k1 and k2 for the equivalency of the eight channels were all more than 0.9975 with RSDs of 0.73% and 0.65%, respectively, and the R values of k1 and k2 for the reproducibility of the microcalorimetric method were in the range of 0.9974–0.9988 and 0.9983–0.9993 with RSDs of 0.71% and 1.37%, respectively. The above results illustrated that the eight channels were equivalent, and that the microcalorimetric method was precise, stable and sensitive enough for evaluation of the antibacterial effects of C. rhizoma.
Channel | Equivalency (n = 3) | Reproducibility (n = 3) | ||||||
---|---|---|---|---|---|---|---|---|
k 1/min−1 | R a | k 2/min−1 | R | k 1/min−1 | R | k 2/min−1 | R | |
a Correlation coefficient. b Relative standard deviation. | ||||||||
1 | 0.01206 | 0.9983 | 0.00647 | 0.9991 | 0.00903 | 0.9985 | 0.00394 | 0.9991 |
2 | 0.01211 | 0.9981 | 0.00652 | 0.9993 | 0.00908 | 0.9980 | 0.00399 | 0.9993 |
3 | 0.01183 | 0.9978 | 0.00649 | 0.9991 | 0.00913 | 0.9974 | 0.00395 | 0.9988 |
4 | 0.01192 | 0.9976 | 0.00650 | 0.9989 | 0.00916 | 0.9975 | 0.00401 | 0.9984 |
5 | 0.01201 | 0.9977 | 0.00651 | 0.9990 | 0.00911 | 0.9983 | 0.00404 | 0.9990 |
6 | 0.01205 | 0.9988 | 0.00655 | 0.9988 | 0.00902 | 0.9988 | 0.00389 | 0.9983 |
7 | 0.01198 | 0.9989 | 0.00643 | 0.9992 | 0.00904 | 0.9981 | 0.00397 | 09989 |
8 | 0.01199 | 0.9991 | 0.00644 | 0.9991 | 0.00900 | 0.9979 | 0.00400 | 0.9991 |
RSD (%)b | 0.73 | 0.65 | 0.71 | 1.37 |
Sample | k 1/min−1 | k 2/min−1 | P 1 m/mW | P 2 m/mW | t 1 m/min | t 2 m/min | Q sta1/J | Q sta2/J | Q t/J |
---|---|---|---|---|---|---|---|---|---|
Control | 0.01355 | 0.00677 | 1.257 | 2.444 | 121.5 | 522.8 | 6.73 | 22.70 | 29.43 |
A | 0.01181 | 0.00435 | 1.231 | 1.928 | 141.3 | 592.3 | 5.73 | 16.59 | 22.32 |
B | 0.00911 | 0.00405 | 1.079 | 1.753 | 143.7 | 646.7 | 4.98 | 16.07 | 21.05 |
C | 0.01108 | 0.00411 | 1.217 | 1.873 | 141.3 | 639.5 | 5.23 | 16.17 | 21.40 |
D | 0.01243 | 0.00492 | 1.142 | 1.970 | 138.8 | 602.2 | 5.21 | 17.27 | 22.48 |
E | 0.00989 | 0.00597 | 1.173 | 2.153 | 136.2 | 602.2 | 5.24 | 17.74 | 22.98 |
F | 0.01224 | 0.00556 | 1.179 | 2.058 | 136.2 | 614.5 | 5.94 | 17.28 | 23.22 |
G | 0.01148 | 0.00609 | 1.108 | 2.179 | 136.2 | 594.7 | 6.13 | 18.96 | 25.09 |
H | 0.00979 | 0.00605 | 1.103 | 2.160 | 129.0 | 560.2 | 6.42 | 21.58 | 28.00 |
I | 0.01037 | 0.00574 | 1.118 | 2.148 | 133.7 | 599.8 | 6.02 | 18.71 | 24.73 |
J | 0.01016 | 0.00627 | 1.165 | 2.231 | 136.5 | 518.0 | 6.65 | 21.88 | 28.53 |
K | 0.01208 | 0.00615 | 1.084 | 2.193 | 129.2 | 574.8 | 5.58 | 17.27 | 22.85 |
L | 0.01313 | 0.00624 | 1.135 | 2.342 | 133.7 | 530.3 | 5.16 | 17.87 | 23.03 |
M | 0.01168 | 0.00634 | 1.215 | 2.373 | 131.3 | 545.0 | 6.06 | 19.64 | 25.70 |
N | 0.01161 | 0.00640 | 1.187 | 2.404 | 123.8 | 518.0 | 6.22 | 21.90 | 28.18 |
O | 0.01148 | 0.00638 | 1.235 | 2.385 | 151.2 | 555.0 | 6.33 | 22.40 | 28.73 |
The data in Table 3 showed that the values of the nine metabolic parameters had different changing trends with various C. rhizoma samples. Compared with the control, the values of k1, k2, P1m, P2m, Qsta1, Qsta2 and Qt decreased and the values of t1m, t2m increased, demonstrating that all C. rhizoma samples had the ability to inhibit E. coligrowth to different extents and that the inhibitory extent varied with various sources of C. rhizoma. The increasing values of t1m and t2m indicated that E. coligrowth was inhibited and the bacterial culture might take a longer time to produce a sufficient number of cells for a detectable signal. This probably resulted from the fact that the C. rhizoma samples combined with the cell to inhibit the duplication of DNA, resulting in damage of the membrane structure and functions of cells. The thermo-kinetic parameters of E. coligrowth in Table 3 were different for all C. rhizoma samples. The smaller the values of k1, k2, P1m, P2m, Qsta1, Qsta2, Qt and the bigger the values of t1m, t2m are, the stronger the antibacterial effects are that the drug possesses.21 For explaining intuitively the inhibitory effects of different C. rhizoma samples on E. coli, the column maps of the nine quantitative metabolic parameters in Fig. 2 showed the changes of the nine thermo-kinetic parameters. The general results from Fig. 1(b) and (c), Fig. 2 and Table 3 showed that the sequence of antibacterial effect of C. rhizoma from various sources on E. coligrowth was: B > C > A > D > F > I > E > H > G > K > J > L > M > O > N. Sample B sample in Table 1 with the smallest k1, k2, P1m, P2m, Qsta1, Qsta2, and Qt values had the strongest antibacterial effect, but sample N had adverse results. As we know, sample B was collected from Shizhu city, Sichuan province, in the southwest area of China, while sample N was produced in Fangxian, Hubei province, in the central area of China. These indicated that the antibacterial effect of C. rhizoma was related to the production place and the latitude and longitude of the place of this CMM.
Fig. 2 Column maps of the nine parameters metabolic parameters. The nine quantitative metabolic parameters were: (1) the growth rate constant k1 and k2; (2) the maximum heat-output power P1m and P2m; (3) the appearance time t1m and t2m of P1m and P2m; (4) the heat output Qsta1, Qsta2 and Qt. |
It could be seen from the results of HCA for C. rhizoma samples from various sources in Fig. 3 that all the samples could be distinguished into three clusters. But the samples in each group were not necessarily close. Cluster I consisted of four samples from Sichuan province, one sample from Chongqing city and two samples from Hubei province. Cluster II consisted of four samples from Sichuan province, while cluster III consisted of two samples from Hubei province, one from Sichuan province and one from Chongqing city. So, it was difficult to find the antibacterial mode of all C. rhizoma samples. This phenomenon might result from the multivariate problems. There were nine metabolic parameters for HCA and the internal information of these parameters was different and the contribution of each parameter for HCA was different. Then, it was necessary to find out the main parameter(s) which played a more important role in representing the antibacterial effect of C. rhizoma and HCA of these samples. By analyzing the change tendency of the main parameter(s), the antibacterial effect could be evaluated better and faster and the antibacterial mode could be explored well by HCA.
Fig. 3 HCA results of C. rhizoma samples from various sources. This dendrogram was acquired based on the nine metabolic parameters obtained from the metabolic profiles of E. coligrowth affected by C. rhizoma. |
The loadings plot of PCA on the nine metabolic parameters was shown in Fig. 4(a). For the log centered data set with 99.62% of explained variance by the first two principal components (parameters), the scatter plot of the loading showed a good distribution of these parameters. This plot indicated that parameters 2 (k2) and 4 (P2m) may be the two main parameters playing a more important role in evaluating the antibacterial effect and mode of C. rhizoma samples from various sources. From the values of k2 and P2m in Table 3, it could be quickly and clearly found that the sequence of antibacterial effect of C. rhizoma on E. coligrowth was: B > C > A > D > F > I > E > H > G > K > J > L > M > O > N, which was same as the above-mentioned results.
On the basis of eigenvalues >1 from the first two principal components PC1 and PC2, the scores plot of PCA for C. rhizoma samples from various sources was shown in Fig. 4(b). The results showed three site-related groups, which were marked as groups I, II and III according to different provinces and cities, respectively. Although producing exactly the same pattern of loadings, the map gave a clear classification of all C. rhizoma samples. From these plots, one could see clearly that the samples of the same or similar source (province and city) would cluster together in the PCA projections. Group I consisted of C. rhizoma samples from Shizhu city Sichuan province, group II consisted of the samples from Chongqing city and other cities in Sichuan province, and group III consisted of the samples from Hubei province. Combined with the above results, it could be found that the samples in group I from Shizhu city Sichuan province had the strongest antibacterial effects and the samples in group III from Hubei province had the adverse results. These 15 samples with different antibacterial effects could be successfully grouped in accordance with the province and city of origin. Ranked in decreasing order, the antibacterial mode of C. rhizoma was that samples from Sichuan province had the strongest antibacterial effects, followed by samples from Chongqing city and then Hubei province.
Fig. 4 (a) Loadings plot of PCA on the nine metabolic parameters, and (b) scores plot of PCA for C. rhizoma samples from various sources on the first two PCs. These plots were obtained by PCA on the nine metabolic parameters from the metabolic power–time profiles of E. coligrowth using software of Unscrambler 9.7 from Camo AS (Trondheim, Norway). |
Fig. 5 New HCA results of C. rhizoma samples from various sources. This dendrogram was acquired based on the two main metabolic parameters got from PCA on the nine parameters. The C. rhizoma samples could be distinguished into three clusters: (I) samples from Sichuan province, (II) samples from Chongqing city, and (III) samples Hubei province. |
This study also illustrated that it was possible for microcalorimetry coupled with chemometric techniques to evaluate the antibacterial effect and mode of C. rhizoma samples from various sources with the help of PCA and HCA, which further proved that the developed microcalorimetry with chemometric techniques had potential for evaluating the effect and mode of Coptidis rhizoma and other Chinese materia medicas.
This journal is © The Royal Society of Chemistry 2012 |