Identification Of NPM1 Mutant Leukemic Cells With Raman Spectroscopy

As a label-free,nondestructive and real-time detection method,Raman spectroscopy of single cell can detect the changes of intracellular molecular components before the clinical symptoms and medical imaging abnormalities caused by diseases.It can be regarded as an essential alternative method for early classification of diseases and tumor grading,and has potential application value in medical fields such as cancer diagnosis.Based on the review and analysis of the theory and application of Raman spectroscopy,this paper studied the feasibility of Raman spectroscopy as a method for the detection and diagnosis of leukemia samples.The specific research work was as follows:(1)Raman spectroscopy acquisition and average spectral analysis of acute myeloid leukemia(AML).The Raman spectra of AML cell lines without NPM1 mutation(THP-1 and HL-60)and OCI-AML3 cell line with NPM1 mutation gene were cultured and collected by the spontaneous Raman spectrometer built in the laboratory.After preprocessing the spectral data,the average spectra of NPM1 mutant and non-mutated cells were analyzed and compared.It was found that there were obvious differences in the spectral peaks between the two.The peaks of the characteristic protein and chondroitin sulfate in the NPM1 mutant leukemia cells were stronger than those in the non-mutated leukemia cells,while the peak intensity of the characteristic nucleic acid was less than that in the non-mutated leukemia cells.In addition,the significance P value was calculated by independent sample T test or Mann-Whitney U test,and the intensity of the above differential peaks was statistically analyzed.The results showed that the spectral differences between leukemia cells with NPM1 mutation and those without NPM1 mutation were statistically significant(P < 0.05).(2)Raman spectroscopy combined with multivariate statistical technique was used to identify different types of human leukemia cells.A principal component analysis(PCA)model was constructed to accurately distinguish NPM1 mutant and non-NPM1 mutant leukemia cells.On this basis,the two types of cells were further identified by partial least squares discriminatory analysis(PLS-DA).The accuracy rate reached98.92%,the sensitivity was 98.08%,and the specificity was 99.25%.In addition,the PLS-DA model was constructed to classify THP-1,HL-60 and OCI-AML3 leukemia cell lines,and the recognition accuracy was 99.46%.The classification accuracy of the model after ten-fold cross-validation was 87.1%.The above analysis results showed that Raman spectroscopy was suitable for distinguishing different types of leukemia cells.(3)Raman spectroscopy combined with transcriptome analysis was used for AML typing.The transcriptome data of 140 patients with NPM1 mutation and 317 patients without NPM1 mutation were obtained from the GEO public database.The gene expression matrix of the two types of cells was quantitatively analyzed by R language,and the differentially expressed genes were screened out,and their roles in regulating chondroitin sulfate proteoglycan and protein synthesis were analyzed.Studies have shown that the differences between the two types of cells expressed by single-cell Raman spectroscopy were consistent with the differences in transcription profiles.The combination of Raman spectroscopy and transcriptomics can explain the differences in Raman spectra of different types of cells from the perspective of biological mechanisms.