• 2022-07
  • 2022-06
  • 2022-05
  • 2022-04
  • 2021-03
  • 2020-08
  • 2020-07
  • 2020-03
  • 2019-11
  • 2019-10
  • 2019-09
  • 2019-08
  • 2019-07
  • br Differential metabolites characterization and metabolic


    2.6. Differential metabolites characterization and metabolic pathway analysis
    OPLS-DA, a supervised multivariate statistical model, was applied to visualize the metabolic difference between control group and different fraction-treated groups. In the OPLS-DA model, vari-ables with Variable Importance in the Projection (VIP) >1 and Mann-Whitney U test (p < 0.05) were considered as the differential metabolites related with each fraction. The potential differen-tial metabolites were putatively identified by MS/MS fragments in combination with the METLIN database (http://www.metlin. and HMDB database ( Then, the predictive power of the identified differential metabolites was evaluated according to receiver operating characteristic (ROC) curves (SPSS 19.0). Pathway analysis was based on KEGG database (, MetaboAnalyst 4.0 (http://www., METLIN database and HMDB database. Finally, the visualized distribution of identified differential metabolites in different groups was characterized by a heatmap analysis.
    3. Results and discussion
    3.1. Anti-cancer effects of different fractions of turmeric
    The TcE inhibited A549 Liproxstatin-1 proliferation in a dose-dependent manner and the IC50 value was the concentration at 5-fold dilu-tion, while dilution at 11-fold maintained 70% survival rate of A549 cells, in which the concentrations of Cur, DMCur and BMCur were of 15 M, 8 M, and 15 M, respectively (see Fig. 1A). Consequently, the TcE and CE, BMcur, DMcur and Cur fractions were diluted at
    Fig. 1. The anti-cancer activity of TcE at different concentration (A) and of different curcuminoid-containing fractions (B). Data was represented as means ± SD (n = 6, ***P < 0.001 vs. the control group).
    11-fold for subsequent experiments including cytotoxic activity evaluation on A549 cells and metabolomics.
    As shown in Fig. 1B, all the samples including the TcE, CE, BMcur, DMcur and Cur groups led to significant growth inhibition on A549 cells compared with the control group. Among them, the inhibition activity of CE fraction is comparative with TcE which suggested that the curcuminoids are the major anti-cancer compounds on A549 cells in turmeric. Furthermore, at the same concentration (15 M), Cur and BMcur showed the similar growth inhibition activity, while DMcur showed lower activity at a low concentration (8 M). This result indicated that not only Cur, but also BMcur and DMcur could exert considerably anti-cancer effects on A549 cells. These results have proven that the target cell extraction-chemical pro-filing method we developed before is satisfactory for screening of anti-lung cancer bioactive compounds from turmeric. On the other hand, turmeric showed proficient anti-cancer effects in vitro, and such effects were likely attributed to its curcuminoids.
    However, to study the anti-cancer activity of individual com-pound in turmeric is not the purpose of this paper. What we concern is whether the multi-compounds with natural proportions in turmeric have integration effects on anti-cancer potency because people intake turmeric as a food additive or herbal medicine which contains multiple compounds. With this aim and due to the advan-tage of metabolomics for hinting to potential targets/mechanisms of natural compounds with holistic therapeutic efficacies, we designed the chemical markers’ knockout samples (including CE, BMcur, DMcur and Cur fractions) treated A549 cells metabolomics to reveal the possible mechanism of actions of these bioactive cur-cuminoids on metabolic pathways and their underlying integration effects.
    3.2. Method validation of the UHPLC-LTQ Orbitrap MS-based metabolomics
    Before obtaining metabolic profiles from certain sample admin-istration, the repeatability and stability of the developed method was firstly validated. The overlapping total ion chromatograms 
    3.3. Cell metabolic profiles in response to different fractions treatment
    After validation, the vectors of each samples were extracted as a matrix with 427/517 (ESI+ /ESI− ) variables as columns and 43 runs as rows. With these variables a multivariate statistical anal-ysis was performed, before which the obtained variables needed to be normalized to the same dimension. Currently, there are two types of normalization methods, including sampling quantity nor-malization and statistical normalization. In this study, these two normalization methods were equally essential to guarantee reliable data for further multivariate statistical analysis. Hence, the protein concentration of each group was determined for sampling quantity normalization before UHPLC-LTQ Orbitrap MS analysis. Statistical normalization strategy was routinely applied by the normalization to the IS peak intensity.