论文部分内容阅读
Single-cell genomics has become a new strategy for the exploitation of uncultivated microorganisms from environment,as well as the interpretation of genetic heterogeneity in microbial populations or consortia.Current methods for isolating single cells,such as fluorescence activated cell sorting (FACS) and micromanipulation,were limited by either need for external labeling or the cell size trackable.Our Raman-activated Cell Sorter provides three approaches for isolation of single-cells with targeted Raman phenotypes (the chemical ‘fingerprint’ of a cell): (1) laser-trapping (for cells in aqueous phase),(2) laseractivated ejection (for cells in non-aqueous phase) and (3) microfluidic device (for cells in flowing phase).We employed these approaches to isolate microbial cells including yeast and bacteria.The single-cells obtained then were used for whole genomic amplification (WGA) to get sufficient DNA template for genomic sequencing.By using multiple displacement amplification (MDA),typical yield of approximately 10 μg of genomic DNA was achieved from 1-6 E.coli cells.We then applied pyrosequencing to investigate the singlecell genome from two E.coli DH5α cells.Totally,309,507 reads of high quality were obtained with an average length of 410bp.Among these reads,164,017 which accounted for 53.0% of the total amount were fully mapped to the reference genome of E coli K12 substrain MG1655,whose genome size was 4.64Mbp and contains approximately 4,466 predicted genes.Afterwards 239,503 reads were assembled to generate 1,520 contigs by Newbler Metrics (Roche 454 Technology),with an average length of 1,704bp.The total length of contigs reached 2.59Mbp,N50 contig size was 5,409bp and the largest contig size was 38,761bp.Among these contigs,1,367 were mapped to the reference genome with BLASTN evalue<1e-20.The contigs covered 52.1% of the whole genome,with read coverage reaching 36.8.The contig GC-content was 51.3%,consistent with the GC-content of 50.8% of the reference genome.SNPs were further identified and compared using three different softwares including SOAP,Bowtie and BWA.The coverage of different SNPs varied from 36 to 2094,with ratio ranging between 4.35e-7 and 8.82e-7.Our pilot study has demonstrated the feasibility of phenotype-genotype based screening of single cells.