Noninvasive Prenatal Diagnosis For ?- Thalassemia Based On Next-generation Sequencing And Pseudo-tetraploid Genotyping

Objective:Thalassemia is serious hereditary hemolytic disease.Patients with?-thalassemia major present with severe anemia in infancy and become transfusion dependent for life,and need iron chelation therapy.?-thalassemia cause a huge burden on patients,families and society.Prenatal diagnosis for the couples at high-risk to have?-thalassemia major fetuses is the only effective way to avoid the birth of?-thalassemia major patients.Prenatal diagnosis for thalassemia are mainly through invasive techniques,with inevitably damage to the mothers and fetuses.Therefore,safe and effective non-invasive methods for prenatal diagnostic have always been the direction and goal of prenatal diagnosis research for hereditary diseases.We try to set up a new techniques for non-invasive prenatal diagnosis by using cell free fetal DNA in maternal plasma through high-throughput target region capture sequencing and pseudo-tetraploid genotype bioinformatics analysis.Methods:The selected subjects were the couples at high risk having?-thalassemia major fetuses and need to do prenatal diagnosis.Only normal singletons were selected.The gestation age was between 12 to 26 weeks.The chorionic villus sampling,amniocentesis or fetal blood sampling were performed and fetal DNA was extracted.The fetal genotypes were detected by the method of reverse dot blot hybridization and fluorescence PCR melting curve.Meanwhile,peripheral blood of pregnant women were collected and cell free DNA was extracted.Gene panels for targeted sequencing were designed to detect 12common?-thalassemia mutations in Chinese.The DNA library were built and then hybridizations with the gene panel were performed to capture the targeted allele for sequencing.A pseudo-tetraploid genotype bioinformatics analysis method was established based on the relationship among fetal genotypes,free fetal DNA concentration,and pseudo-tetraploid genotypes.Through the maximum likelihood estimation,the optimal fetal fraction is continuously iterated.Using Bayesian model to obtain the optimal genotype of each locus,the most accurate fetal concentration was estimated by the correspondence between genotype and fetal concentration,and the iterative until the difference between the two concentrations is less than 0.001.The fetal fraction,which in turn,yielded the corresponding fetal genotypes.To verify the efficacy of this new method,the fetal genotypes obtained by this non-invasive method were compared with results of invasive one.The predicted fetal DNA fraction was also compared with the measured value.Results:1.General information:Of the 146 groups of cases,133 of the cases met the requirements of this study.The age of pregnant women was 28.47�6.05years old(16 years to 47 years old).The gestational age was 15.24�3.31 weeks(12.0 weeks to 26.4 weeks).The height of the pregnant women was 156.44cm�5.02cm(145cm to 171cm),and the body weight was 54.78kg�7.84kg(40.5kg to79.0kg).The pregnant women's BMI was 20.3�3.2 m~2/kg(18.4 kg/m~2 to 23.98kg/m~2).2.The concentration of free DNA and fetal fraction in maternal plasma:The concentration of free DNA library in maternal plasma was 20.3 ng/?l�12.8ng/?l.The DNA concentration was positively linear with gestational age.The concentration of free DNA library in maternal plasma at gestation age of 12,13and 18 weeks had no correlation with age of the pregnant women.This indicated the age of pregnancy had no effect on the concentration of cell free DNA.The predicted fetal fraction was 10.26%�5.82%.The measured fetal fraction was11.37%�6.86%.There was a positive linear relationship between fetal fraction and gestational age.The fetal fraction at gestation age of 12,13 and 18 weeks had no correlation with the age of the pregnant women.This indicated the age of pregnancy had no effect on fetal fraction.3.The genotypes of the fetal DNA obtained from villus,amniotic fluid and cord blood:The genotypes detected by reverse dot hybridization were consistent with that of detected by fluorescence PCR melting curve method.Among 133 fetuses,there were 15 double heterozygotes,7 cases of homozygotes,79 cases of heterozygotes and 32 cases without mutations.Their genotypes included mutations at:CD 41/42,CD 17,CD71/72,CD 43,-28,-29,IVS-I-I,IVS-II-654 and Hb E.4.Results of plasma free DNA library construction:For 133 pregnant women peripheral blood DNA library,quality identification showed that the main peak absorption spectrum was between 200bp and 400bp.The peak map was high and sharp with smooth line.The second peak range was 400bp to 1000bp and peak map was short and blunt.There was no obvious peak map after the Lower marker.The absorption peak of DNA library in high-throughput sequencing was 167bp�8bp.5.High-throughput sequencing results in the target region:The average lateral depth of 133 eligible DNA documents was 379.475ױ172.928.The lateral depth in 120 of 133 cases(90.22%)was more than 200�.The pool sequencing depth was 600�.Raw reads was 7825310.57�1254716.368.The Q30 rate was 81.8317%�2.09838%.GC content was 41.0253%�0.8747%.The total number of libraries in the mutation site was 181.The number of effective libraries captured was 161,and the mutual efficiency was 88.95%.The capture efficiency of the CD 41/42 was 85.14%,and the capture efficiency in other regions was 91.59%.6.Analysis results of pseudo-tetraploid genotype bioinformatics:Compared with the results of invasive prenatal diagnosis,113 cases were consistent.20 cases were not consistent,including 11 cases of CD 41/42,3 cases of CD 17,2 cases of-28,1 case of CD43,1 case of CD 71/72 and 2 cases of IVS-II-654.The overall compliance rate is84.96%.7.The results of the efficacy evaluation of the pseudo-tetraploid genotype analysis method:A paired sample t test was performed on the two sets of data of predicted and measured fetal fraction,t=0.436,P=0.762.The genotype evaluation results showed a Kappa coefficient of 0.71,a total sensitivity of 83.5%(95%CI:61.6%-93.2%),and a specificity of 92.8%(95%CI:76.7%-97.9%).Conclusion:1.The gene panel designed for the 12 genotypes of?-thalassemia mutations which are common in Chinese,can well capture the corresponding regions.2.The concentration of free DNA and fetal fraction in maternal plasma increased with the gestational age,but they are not related to the age of pregnant women.3.The method of pseudo-tetraploid genotype bioinformatics analysis can directly infer the proportion of fetal genotypes and free fetal DNA in the free DNA in maternal plasma.It is feasible for non-invasive prenatal detection on?-thalassemia.4.The method based on high-throughput sequencing and pseudo-tetraploid bioinformatics analysis has the advantages.It does not need to know the genetic information of probands and the couples,and not need to construct RHDO and RMD models.5.The method combined the high-throughput sequencing with pseudo-tetraploid bioinformatics analysis are theoretically applicable to the single-gene genetic diseases which conform to Mendel's law of inheritance.