posted on 2022-08-17, 18:20authored byRan Zheng, Rui Su, Fan Xing, Qing Li, Botong Liu, Daguang Wang, Yechao Du, Keke Huang, Fei Yan, Jianfeng Wang, Huanwen Chen, Shouhua Feng
The application of rapid and accurate diagnostic methods
can improve
colorectal cancer (CRC) survival rates dramatically. Here, we used
a non-targeted metabolic analysis strategy based on internal extractive
electrospray ionization mass spectrometry (iEESI-MS) to detect metabolite
ions associated with the progression of CRC from 172 tissues (45 stage
I/II CRC, 41 stage III/IV CRC, and 86 well-matched normal tissues).
A support vector machine (SVM) model based on 10 differential metabolite
ions for differentiating early-stage CRC from normal tissues was built
with a good prediction accuracy of 92.6%. The biomarker panel consisting
of lysophosphatidylcholine (LPC) (18:0) has good diagnostic potential
in differentiating early-stage CRC from advanced-stage CRC. We showed
that the down-regulation of LPC (18:0) in tumor tissues is associated
with CRC progression and related to the regulation of the epidermal
growth factor receptor. Pathway analysis showed that metabolic pathways
in CRC are related to glycerophospholipid metabolism and purine metabolism.
In conclusion, we built an SVM model with good performance to distinguish
between early-stage CRC and normal groups based on iEESI-MS and found
that LPC (18:0) is associated with the progression of CRC.