Study On Establishment And Validation Of Multiple Gene Expressions-based Biodosimetric Assays For Rapid Radiation Dose Estimation

ObjectiveIn the event of a nuclear accident or terrorist attack,a larger number of individuals may be exposed to ionizing radiation.Rapid and reliable radiation dose estimation of possible exposed individuals is the key to emergency preparedness and clinical treatment in a large-scale nuclear event.Conventional unstable chromosome aberration(dicentric chromosome and centric ring,dic+r)analysis is reliable and sensitive,but it is labor-intensive and time-consuming for triaging numerous victims.Analysis of ionizing radiation-induced gene expression provides solutions to many of these difficulties.Previously,we have conducted a systematic study of radiation-induced changes in the expression levels of multiple individual genes and found that the use of a single gene to estimate the dose of irradiation had great uncertainty in practical applications.The present study is to screen and validate radiation-sensitive genes with established dose-effect relationship and time response based on current studies.The expression levels of selected radiation-responsive genes will be examined and identified in human peripheral blood samples irradiated ex vivo.Optimal multiple gene expressions-based biodosimetric model at each time-point after irradiation will be obtained by stepwise regression analysis.The feasibility of multiple gene expressions-based biodosimetric assays for rapid dose estimation will be validated in both ex vivo and in vivo irradiated human peripheral blood samples.Method1.Bibliometric methods were applied to analyze the global research trend of publications related to differential expressed radiation-responsive genes from 2000 to 2017.The highly reproducible radiation-responsive genes must possess consistent dose-effect relationships in more than three independent studies.Bioinformatics analysis was used to identify the functional classification,signaling pathway and protein interactions of these genes.2.Human B lymphoblastoid cell line(AHH-1)was irradiated with 60Co ?-rays,at doses of 0,1,3and 5 Gy with dose rate of 1 Gy/min.Following exposure,AHH-1 cells were incubated at 37 ? for 4 and 24 h.3.AHH-1 cells were irradiated with 60Co ?-rays,at doses of 0,1,3,5,8 and 10 Gy with dose rate of 1 Gy/min.Irradiated AHH-1 cells were incubated at 37 ? for 8-168 h;Neutron irradiation was performed at the General Hospital of Armed Police Forces(Beijing,China),using a 252Cf neutron source.AHH-1 cells were irradiated with 0,0.4,0.8,1.2 and 1.6 Gy,at dose-rate of 0.073 Gy/min.Irradiated AHH-1 cells were incubated at 37 ? for 48 h.GDF-15 mRNA and protein expression levels following exposure to y-rays and neutron irradiation were assessed by reverse transcription-quantitative polymerase chain reaction and Western blot analysis in AHH-1 cells.4.Human peripheral blood samples from 30 adult donors was acutely exposed to 60Co ?-rays at doses of 0(control),0.5,1,2,3,4,6 and 8 Gy with dose rate of 1 Gy/min.Irradiated peripheral blood samples were incubated at 37 ? for 6,12,24 and 48 h;the expression levels of highly reproducible radiation-responsive genes will be examined and identified in human peripheral blood samples irradiated ex vivo.Optimal multiple gene expressions-based biodosimetric model at each time-point after irradiation was obtained by stepwise regression analysis.5.For the double blind test,human peripheral blood from 10 adult donors was acutely exposed to 60Co ?-rays at doses of 0.8,1.5,2.5,3.5,4.5 and 5.0 Gy with dose rate of 1 Gy/min.Following exposure,the peripheral blood samples were incubated at 37 ? for 24 h.Dose estimations were obtained by using single genes(GDF-15,TNFSF4,PCNA,CDKN1A,ASTN2 and FDXR),chromosome aberration analysis and multiple gene expressions-based biodosimetric models at 24 h after irradiation,respectively.6.Human peripheral blood samples were collected from four acute leukemia patients who reveiving total-body irradiation(TBI)before irradiation and at 24 after exposure to 60Co ?-rays at dose of 5 Gy with dose rate of 1 Gy/min.Dose estimations were obtained by using multiple gene expressions-based biodosimetric models at 24 h.Results1.There were 129 publications on radiation-responsive genes from 2000 to 2017.Some fluctuations were observed with regard to the total number of publications over the past 17 years;a total of 35 highly reproducible radiation-responsive genes were identified;most genes are involved in response to DNA damage,cell proliferation,cell cycle regulation,and DNA repair;the p53 signal pathway was the top enriched pathway.2.The expression levels of 18 radiation-responsive genes(CDKN1A,DDB2,GADD45A,BAX,MDM2,FDXR,XPC,PIG3,GDF-15,SESN1,PHPT1,PHPLDA3,POLH,ASTN2,RPS27L,TRIM22,PLK3 and BBC3)in human B lymphoblastoid cell line(AHH-1)were significantly up-regulated in a dose-dependent manner at 4 or 24 h after exposure to 0-5 Gy 60Co ?-rays irradiation.Ionizing radiation-induced GDF-15 gene expression changes are most significant among all radiation-responsive genes.3.GDF-15 mRNA and protein expression levels in AHH-1 cells were significantly up-regulated from 1 to 10 Gy of y-rays at 12 to 120 h or 12 to 48 h after exposure,respectively.The enhancement was also observed in AHH-1 cells from 0.4 to 1.6 Gy of neutrons 48 h post-irradiation.4.The expression levels of 18 radiation-responsive genes(CDKN1A,BAX,MDM2,XPC,PCNA,FDXR,GDF-15,DDB2,TNFRSF10B,PHPT1,ASTN2,RPS27L,BBC3,TNFSF4,POLH,CCNG1,PPM1D and GADD45A)in human peripheral blood samples were significantly up-regulated in a dose-dependent manner at 6,12,24 and 48 h after exposure to 0-8 Gy 60Co ?-rays irradiation;The expression levels of radiation-responsive genes may be affected by individual variations,gender and ages.Of the 18 radiation-responsive genes,only the expression levels of GDF-15 and PPM ID genes were not affected by gender and ages at each time-point after irradiation.Optimal stepwise regression models based on gender and age groups at each time-point after irradiation were established,R2 values were 0.802-0.854 and 0.776-0.935,respectively.In particular,the statistic models of different genders were composed of 8 and 10 radiation-responsive genes at 24 h after exposure.PCNA,CCNG1,CDKN1A,XPC,and TNFSF4 were included in two models,the R2 values of both models were approximately 0.850.5.In the double blind tests,there was a large gap between the actual doses and the predicted doses estimated by single genes.Only the average estimated doses in the specific dose range of PCNA,TNFSF4 and GDF-15 genes were close to the actual doses.The average estimated dose in the dose range of 0.80?3.50 Gy for unstable chromosome aberration analysis were very close to the actual doses.Using stepwise regression models based on gender at 24 h after irradiation,the average predicted doses in an independent test were close to actual doses with the dose range of 0 to 5 Gy.6.Using stepwise regression models based on gender at 24 h after irradiation,the estimated dose of four radiotherapy patients at 24 h after receiving total-body irradiation of 5 Gy were 4.80,5.02,5.50 and 5.04 Gy,respectively.The average estimated dose was 5.09 ± 0.29 Gy,and the 95%confidence interval was 4.62 to 5.56 Gy.Conclusion1.Thirty-five radiation-responsive genes were screened and identified based on bibliometrics and bioinformatics methods from 129 publications in the past 17 years.2.The alterations in expression levels of radiation-induced 13 genes(CDKNIA,DDB2,GADD45A,BAX,MDM2,FDXR,XPC,GDF-15,PHPT1,POLH,ASTN2,RPS27L,and BBC3)are not affected by cell types.There was a more consistent dose-response relationship between human lymphoblastoid cells and peripheral blood samples,of which GDF-15 has the potential to become a radiation biomarker and can be combined with other radiation-sensitive genes for biological dose estimation.3.Establishing multiple gene expressions-based biodosimetric assays for rapid radiation dose estimation has certain feasibility by using 3 to 10 radiation-responsive genes in human peripheral blood samples.Optimal stepwise regression models were obtained based on age and gender at each time-point after irradiation.4.Multiple gene expressions models based on gender at 24 h after irradiation can be utilized for accurate dose estimation in human peripheral blood samples irradiated ex vivo and in radiotherapy patients.The dose estimation range is 0 to 5 Gy,which is better than single genes and unstable chromosomal aberration analysis.Multiple gene expressions-based biodosimetric assays for rapid radiation dose estimation is simple,rapid,accurate and high-throughput.Innovations1.The present study was closely focused on the feasibility of ionizing radiation-induced gene expression as a radiation biodosimetry.This study mainly included the screening and verification of radiation-sensitive genes,the construction,optimization and validation of multiple gene expressions-based biodosimetric assays.Stepwise regression models composed of 3 to 10 radiation-responsive genes were obtained.The study has a certain degree of innovation in research strategy.2.The multiple gene expressions-based biodosimetric models established in this study can detect the radiation-sensitive genes in human peripheral blood samples by real-time fluorescence quantitative PCR;and improve the accuracy of biological dose estimation by comprehensively analyzing the expression changes of multiple genes.Compared with the analysis of unstable chromosomal aberrations,this method has the advantages of simplicity,rapidity,accuracy,and high throughput.Meanwhile,the results of the study suggested that the multiple gene expressions-based biodosimetric assays can be used not only for biological dose estimation of exposed individuals,but also for the monitoring and evaluation of radiotherapy in cancer patients.