Evaluation And Forensic Application Of Protective Capabilities Of Nucleosome Against Degraded DNA

Objective:The analysis of highly degraded DNA is always a major challenge for forensic genetics. It is encountered in the case of natural and mass disasters, homicide cases, and terrorist attacks. Many factors can destroy the DNA molecular, such as high temperature, moisture, ultraviolet, soil microorganism, strong acid and alkali, resulting in molecular fracture, molecular reduction, large fragments loss, base deletion and base modification. DNA profiles from these samples always produce amplification failure or genotype failure. The common problem-solving methods are: ① to increase the sensitivity of the PCR reaction, such as increased PCR cycles and template amount ② to reduce the size of amplicons, such as mini-STR designs and SNP strategy ③ to use mtDNA as a marker which has a high copy number. However, the intrinsic structural properties of DNA always are ignored. Nucleosome is the subunits of all chromatins, containing 200 bp of DNA-the core DNA and linker DNA. The association of DNA with the histone octamer forms a core particle containing 146 bp of DNA.By analyzing and summarizing the reference literatures, some circumstantial evidence suggested that the protective capabilities of nucleosomes probably exist. In apoptotic cells, Endogenous endonucleases target the linker DNA between portions of the nucleosome, resulting in a ladder of multiple fragments in 200bp steps. Dixon et al. also suggested that a nucleosome could offer protection of the 147bp of DNA that are bound to it from the attacks of endonucleases. Another study showed the majority of extracted ancient DNA fragments were 100-200bp, which is consistent with the length of core DNA or the mononulceosome DNA. Thanakiatkrai et al. and Freire-Aradas et al. focused on nucleosome potential protection for forensic STRs and SNPs, respectively. These two studies using software predictions produced controversial results. The selection of nuclesome )iomarkers based on software prediction have virtues and limitations, although it is convenient and cost-effective, it is different from the actual ituation due to the predictions only based on several characteristics such as mcleosome positioning signals and DNA bendability. Our studies strived for i breakthrough of degraded DNA analysis by taking advantage of the DNA ntrinsic structural properties. We used the Illumina Hiseq 2000 Genome nalyzer platform to characterize nucleosome positions in a multicellular opulation of human leucocytes. We evaluated the protective capabilities of mcleosome STRs and developed a nucleosome STR multiplex which rovided a new strategy for degraded DNA analysis.Methods:1 The leukocytes were isolated from the peripheral blood of healthy ionors.Mono-nucleosome-sized DNA from MNase-digested chromatin were solated and sequenced by Solexa sequencing technology. The general nformation of sequencing data were analyzed. The distribution of SNVs, ndels and STRs in obtained reads were analyzed. Biomarkers in nucleosome egions and non-nucleosome regions were selected.2 Here we obtained nucleosome STR information by sequencing the core DNA of leucocVtes. Five nucleosome STRs and five non-nucleosome STRs vere selected. After redesigning the primer, all amplicon sizes were reduced to )elow 147 bp. We then compared nucleosome protected biomarkers with other ive non-nucleosome STRs using artificially degraded DNA and real casework samples.3 We performed a population genetic investigation and forensic pplication parameter analysis of unpublished nuclesome loci. Two qualified npublished nuclesome loci were selected to perform further evaluation. Five oci were selected to construct the nuclesome STR multiplex. We estimated he protective capabilities of the nucleosome STR multiplex and MiniFilerTM oci using artificial degraded DNA and case samples. We also analyzed the different performances between sequencing results and software predicted results.Results:1 The results of part one:①we isolated mono-nucleosome-sized DNA rom MNase-digested chromatin and sequenced the DNA ends using Solexa sequencing technology. We obtained 17,752,5592×100 pair-ended reads, and 75.34% of them align to unique genomic loci. The average GC content of the eads from the nucleosome core fragment experiment is 52%. All regions with ead coverage account for about 34% of the entire genome ,042,463,676/3,095,693,983), ranging from 27% （chr 22） to 52% （chr X）, nd the average sequencing depth reaches approximately 3× in these regions. The short reads obtained from sequencing were mapped to the human genome hg19) and the distribution in the genome was analyzed.The majority of ell-positioned nucleosomes were detected in intergenic and intronic regions 52.5% and 38.8%, respectively).②We identified 2,462 SNVs and 128 indels, n the nucleosome-binding regions, with at least 10 supporting reads, most of vhich occurred in intergenic regions. Among the detected SNVs,2,220 of hem had records in dbSNP 138, and 242 were excluded.89 of the indels had ecords in dbSNP₁38, and 39 were excluded. We compared our direct equence reads from nulceosome core regions with 44 commonly used rensic STR loci.. Five of these STRs （TH01, TPOX, D18S51, DYS391, and ）10S1248) were discovered in our reads. After run through an STR screening rogram,10,167 potential STRs in our nucleosome core region reads met all ix criteria of program 1; 295,126 potential STRs satisfied the first 4 criteria of program 2,468 type of motif satisfied all the 5 criteria of program 2. The ibove-mentioned 5 commonly used forensic STR loci were found both in hese two programs.2 The results of part 2:①We compared these "nucleosome protected TRs" （D10S1248, D18S51, TH01, TPOX and DYS391） with five other on-NPSTRs （CSF1PO, D5S818, D8S1179, D16S539 and DYS392） about heir protective capabilities against degradation. ㎝ann-Whitney U test showed no differences between these two groups for amplicons length. ③Degraded samples exhibited a downward trend in DNA concentration with ncreasing incubation time in all five samples. We used repeated measures and nultivariate analysis of variance （MANOVA） process from the general linear odel in SPSS and giving comparison among different groups and different aeasure time pairwise. There were significant differences between the NPSTR roup and the non-NPSTR group in all samples at three time points （5,10, and 0 min）.④In 20 blood samples the average locus detection rates of the PSTR group is significantly higher than the non-NPSTR group （P= 0.001 .05）. the average locus detection rates of the NPSTR group is 64.75%, the iverage locus detection rates of the non-NPSTR group is 26.75%. Among the n miniSTRs, D10S1248 of the NPSTR group was most likely to be detected )y capillary gel electrophoresis （100% detected in 20 artificially degraded amples）. D18S51 is the only locus from the NPSTR group whose detection ate was less than 60%. The single locus detection rates of the non-NPSTR roup was low, varying from 10% to 45%. For all 12 samples the average ocus detection rates of the NPSTR group was 50%; however, the values for he non-NPSTR group were 27%. The value of the NPSTR group was significantly higher than that of the non-NPSTR group （P= 0.008< 0.05）. We urther analyzed the effect of different fixation times on the detection rates. There was a downward trend of detection rates with increasing fixation times n NPSTR group, but regardless of fixation time length, the loci detection rates or the NPSTR group is always higher than that for the non-NPSTR group. We conclude that the four NPSTR loci （TPOX, TH01, D10S1248, and DYS391） ;ould be well suited for nucleosome multiplex constructions in the third chapter.3 The results of part three: ①Two qualified unpublished nucleosome loci AC012568.7 and AC007160.3) were selected to construct the nucleosome TR multiplex. The observed heterozygosity （Ho） for AC012568.7 and C007160.3 were 0.648 and 0.722, respectively. The polymorphism nformation component （PIC） for AC012568.7 and AC007160.3 were 0.52 and 0.59, respectively. The power of discrimination （PD） for AC012568.7 andAC007160.3 were 0.755 and 0.787, respectively, and the power of exclusion PE) for AC012568.7 and AC007160.3 were 0.353 and 0.463, respectively. ② The nucleosome STR multiplex were successfully developed which included ive nucleosome STRs（TPOX, TH01, D10S1248, AC012568.7 and C007160.3）.③We used a capillary gel electrophoresis （CGE） separation chnique to type performance 20 artificially degraded DNA segments between nucleosome STR multiplex and MiniFilerTM. In 20 artificially iegraded DNA samples, the average loci detection ratio of nucleosome STR nultiplex was 54%; however, the value in MiniFilerTM （including all the loci） vas 19%, the value in MiniFilerTM （< 147bp） group was 17%. The Vlann-Whitney U Test results showed that there was a significant difference in he loci detection ratio between the nucleosome STR multiplex and VliniFilerTM （Pminifiler-all=0.000< 0.05;Pminifiler-147=0.000< 0.05）.12 formalin ixed samples were genotyped by nucleosome STR multiplex and MiniFilerTM. n 12 formalin fixed samples, the average loci detection ratio of the mcleosome STR multiplex was 40%; however, the value in MiniFilerTM including all the loci) was 11%, the value in MiniFilerTM （< 147bp） group vas 18%. The Mann-Whitnev U Test results showed that there was a significant difference in the loci detection ratio between the nucleosome STR nultiplex and MiniFilerTM （Pminifiler-all=0.005< 0.05; Pminifiler-147=0.043< 0.05）. or the single locus detection ratio, all loci of the nucleosome STR multiplex lad a detection ratio more than or equal to that of MiniFilerTM （< 147bp） roup. These results were consistent with the nucleosome positioning results )btained by large scale sequencing. However the results of software prediction vere opposite. The potential nucleosome dyad of MiniFilerTM （< 147bp） roup was higher than that of the nucleosome STR multiplex. Comparing with software prediction, experimentally mapped nucleosome positions conducted n our study should provide more accurate and direct results.Conclusions:1 We isolated mono-nucleosome-sized DNA from human blood leucocytes sequenced the DNA ends using Solexa sequencing technology. In the nucleosome-binding regions,2,462 SNVs,128 indels,10,167 potential TRs（program 1） and 295,126 potential STRs （program 2）, including five orensic common STRs（TH01, TPOX, D18S51, DYS391, and D10S1248） were identified.2 The loci detection rates for the five forensic common nuclesome STRs higher than that for the non-NPSTR group in artifically degraded samples nd formalin fixed samples. The results showed that the protective capabilities )f nucleosome STRs did exist which would beneficial for the analysis of ighly degraded DNA.3 At last, the nucleosome STR multiplex were successfully developed hich included five nucleosome STRs （TPOX, TH01, D1,D10S 1248, C012568.7 and AC007160.3）.The mlutiples had better performance than MiniFilerTM Kit in artifically degraded samples and formalin fixed samples. The results revealed the existence of protective capabilities of nucleosome. The nucleosome loci and the nucleosome STR multiplex could provided a new trategy for degraded DNA analysis.4 Comparing with software prediction, experimentally mapped ucleosome positions conducted in our study should provide more accurate nd direct results. The results revealed that large-scale sequencing technology important for obtaining accurate information on nucleosome positioning of forensic common profile-human leucocytes.