Screening And Identification Of Serum Potential Biomarkers For Multidrug-Resistant Tuberculosis Based On Data-Independent Acquisition And Targeted Proteomics

Background: Multidrug-resistant tuberculosis(MDR-TB)is defined as resistance to at least isoniazid and rifampicin,posing a serious threat to global public health.China ranks second in the world with a high burden of MDR-TB.The incidence rate of MDR-TB in China is 8.32%.There are estimated 120,000 new cases of MDR-TB in China per year,and the situation is grim.There is currently no rapid diagnostic method for detecting MDR-TB.The process of diagnosing MDR-TB is relatively complicated.It is necessary to conduct bacteriological confirmation of tuberculosis and use rapid molecular testing or culture methods for drug resistance detection.Culture-based methods such as drug susceptibility testing(DST)of solid medium are the main criteria for diagnosing MDR-TB,which usually takes long culture time(6-12 weeks).The XPERT MTB/RIF assay,recommended by the World Health Organization(WHO),can detect resistance to rifampicin.However,the detection rate of the XPERT MTB/RIFassay in smear-negative/culture-positive samples is low.Another problem is false positivity.The current diagnostic methods for MDR-TB are directly related to the bacterial load of the specimen,and are suitable for patients with positive sputum smear,but not for patients with negative sputum smear or people with low sputum.The dynamic changes of the proteome can directly elucidate the underlying mechanism of the disease pathological state,and show promise in the diagnosis or drug targets of tuberculosis.Therefore,it’s valuable to investigate biomarkers of proteomics and may help to improve the diagnosis of MDR-TB.Methods: In order to explore potential diagnostic biomarkers for MDR-TB,we identified differential serum proteins using data-independent acquisition(DIA)and parallel reaction monitoring(PRM)by high performance liquid chromatography(HPLC)system Easy n LC1200 and Q-Exactive HF mass spectrometer in the MDR-TB group,the drug sensitive tuberculosis(DS-TB)group,and the healthy control(HC)group.The DIA data between each sample was calibrated using i RT(a normalized retention time for peptides),and then a comparative analysis of the original data of each sample was performed by LC-MS/MS.In the characterization of proteins,the false discovery rate(FDR)method was used to screen the score threshold.Differential proteins were set as:the fold change > 1.2 times or < 0.7 times with the p value < 0.05.According to target qualitative analysis of candidate proteins,the identified target peptides were screened,and credible peptides were retained.The peptides separated by HPLC were analyzed by PRM to quantify the target protein and target peptides.Bioinformatics,GO,KEGG and correlation analyses were used to establish a MDR-TB diagnostic model.Results: A total of 1,000 proteins were identified in the MDR-TB,DS-TB and HC groups by DIA.A total of 157 differential proteins were identified between the HC group and the MDR-TB group,of which 143 differential proteins were up-regulated and14 differential proteins were down-regulated.The results of KEGG analysis showed that the differential proteins were mainly concentrated in the complement coagulation cascade.A total of 33 differential proteins were identified between the MDR-TB group and the DS-TB group,of which 28 differential proteins were up-regulated and 5differential proteins were down-regulated.The results of KEGG analysis showed that differential proteins were related to the cell surface adhesion and extracellular matrix(ECM)receptor interaction.A total of 170 differential proteins were identified between the DS-TB group and the HC group,of which 135 differential proteins were up-regulated and 35 differential proteins were down-regulated.The results showed that the differential proteins were mainly concentrated in the complement coagulation cascade reaction and other immune reactions(such as,leukocyte-mediated production and regulation).The identified differential candidate peptides were retained with a total of 41 peptides,and then a targeted analysis methodology was established.Compared with the DS-TB group,the abundance of peptidoglycan recognition protein 2(PGLYRP2,Swissprot: Q96PD5),fibrinogen alpha chain(FGA,Swissprot: P02671)and monocyte differentiation antigen CD14(s CD14,Swissprot: B2R888)were significantly up-regulated(p < 0.001),up-regulated(p < 0.001),and down-regulated(p< 0.01)in the MDR-TB group,respectively.We combined the three candidate proteins to establish a diagnostic model.The area under the ROC curve value was 0.934(95%CI,0.799 to 0.990)with the sensitivity of 81.2% and specificity of 90.0%.The leave-one-out cross-validation showed that the predictive accuracy of the candidate biomarker fitting model was 80.0% with the sensitivity of 75.0% and specificity of85.0%,respectively,which can distinguish the MDR-TB group and the DS-TB group.In addition,we detected differential proteins derived from Mycobacterium tuberculosis(Mtb)in the serum of tuberculosis patients,and established a correlation based network between the host candidate proteins and the differential Mtb proteins.The results suggested changes between immune-and inflammation-related proteins in the progression of tuberculosis.Conclusion: We used DIA and PRM to identify three potential protein biomarkers(s CD14,FGA and PGLYRP2),and established a combinatorial model for MDR-TB.Our results may help establish a biological basis for the diagnosis of potential biomarkers for MDR-TB,and may help to elucidate the underlying mechanisms of MDR-TB.Our study may also illustrate the application of proteomics in the field of disease biomarkers,and may contribute to future research on tuberculosis pathology-related functions.