| Background DNA methylation is an important mechanism of epigenetics and plays an important role in regulating gene duplication and protein expression.Cp G islands are cytosine guanine dinucleotide-rich regions in the mammalian genome.Most of the gene promoters are located within the Cp G island,and in particular,the promoter of the housekeeping gene is usually embedded in Cp G islands.Therefore,understanding the distribution of Cp G islands is particularly important for further study of HBV DNA methylation and HBV biological characteristics.Due to the lack of proofreading function of HBV DNA polymerase,HBV DNA has a high mutation rate.There were studies which had divided HBV DNA into 10 genotypes of A-J,each genotype was divided into several subtypes.Three traditional Cp G islands have been recognized: ?,? and ?.Recently,some scholars have found three new Cp G islands named ?,? and ?,and analyzed the distribution of Cp G islands ? ~ ? in some genotypes.However,due to the large variation rate of DNA in each strain of virus,the distribution of Cp G islands of the virus strains may not be able to more accurately reflect the general situation of the Cp G island distribution of the genotype,and moreover,the characteristics of Cp G islands of each subgenotype can not be reflected.The purpose of this study was to establish a reference sequences of HBV DNA genotypes and subtypes using large samples and to analyze the distribution of Cp G islands in the reference sequences and their virus strains.Methods The whole genome sequence of A-H genotype of HBV DNA is searched from Gen Bank,and the virus strains with the whole gene sequence of 3100 to 3300 bases in length are selected.Reference sequences of HBV DNA A-H genotypes and some subtypes in the published papers were used as reference.MEGA7 software was used to carry out multiple genome sequence alignment of virus strains of all genotypes to establish phylogenetic tree and phylogenetic analysis,verify and reclassify the genotype or subtype of each strain.For a large number of virus strains in a genotype,the branch in the evolutionary tree is more obvious,we further classified its subtypes.Then,all the virus strains of each genotype or subtype were subjected to multiple sequence multiple alignment using Align X of Vector NTI Advance11.5 software,respectively,to establish a consensus sequence and to finally obtain the reference sequence.Cp G islands of each reference sequence and each representative virus strain`s whole gene sequence of each subgenotype were calculated by Meth Primer and Cp G Plot two online methods.The distribution differences of Cp G islands between the reference sequences of each genotype and different subgenotypes in the same genotype,and between the representative virus strains in the same subgenotype were analyzed.We conducted a Chi-Square test of SPSS 16.0 on the composition of Cp G islands among the various genotypes,and performed non-parametric tests of SPSS 16.0 on the length and position of each Cp G island of each genotype to determine whether Cp G islands exist significant diversity between different subgenotypes of the same HBV genotype.Statistical significance of the test was defined as P<0.05.Results We selected 3037 complete genomes of HBV A-H genotypes from the downloaded strains,accounting for A 433,B 512,C 924,D 785,E 198,F 135,G 28,H 22.The reference sequences of 28 subgenotypes of A-H genotypes were established by using these strains.The length of the reference sequences of B,C,F and H genotypes were 3215 bp,and A,D,E and G were 3182-3248 bp.We calculated Cp G islands from 28 subgenotype reference sequences and their 939 representative strains.The 28 reference sequences each had 2-3 Cp G islands,and 17 reference sequences of genotypes B,D and E and subgenotypes A1,A2 and C6 contained three conventional Cp G islands of I,II and ?,11 reference sequences of subgenotypes A5,C1,C2 and C5 and genotypes F,G and H lacked Cp G island ?.Only the reference sequence of subgenotype F4 contained a novel Cp G island: Cp G island ?.Among the 939 representative strains,there were 1 to 5 Cp G islands in each strain,and 515,939,938,65,47 and 8 strains containing Cp G island ? ~ ?,respectively.The median length of each island was 102,439,157,112,104 and 105.Among them,454 strains contained only three conventional Cp G islands: ? ~ ?,which were A 81,B 124,C 24,D 183,E 40,F 0,G 2 and H 0 in each genotype.423 strains lacked Cp G island ?,mainly concentrated in genotypes C,F,G and H;117 strains contained 120 novel Cp G islands: ?,? and ?,Cp G islands ? and ? accounting for the majority and mainly focusing in genotypes B,C,D and F.One strain contained only one Cp G island: ?.366 strains contained only Cp G islands ? and ? and they had neither Cp G islands nor novel Cp G islands.The Cp G islands I,II and ? were all cut off by non-Cp G-rich regions of varying length,which was more common with Cp G island ?.The composition of strains with or without Cp G island ? or new island between different subtypes of the HBV each genotype were all significant differences(P<0.05).The composition of strains containing Cp G islands ?,?,? and new islands between different subtypes of the HBV some genotype were significant differences(P<0.05 of genotypes A,C and F,B and D are opposite).There were significant differences in the positions of Cp G islands ?,? and ?between the strains from different subtypes of HBV genotypes(P<0.05),while the Cp G island ? of genotype C was the exception(P>0.05).Conclusion This study established 28 HBV genotype reference sequences,which provided reliable materials and basis for further study on the biological characteristics of each(sub)genotype.The distribution of Cp G islands of different subgenotype reference sequences has obvious differences,while the distribution of Cp G islands among the same subgenotype strains has some commonalities. |
PDF Full Text Request