| Single-cell RNA sequencing(sc RNA-seq),which allows scientists to parse intracellular gene expression profiles at the single-cell scale,has dramatically changed transcriptome research strategies.sc RNA-seq has become a highly useful tool for interrogating biological systems,and permitting a finer understanding of the transitory cellular states following stimulation in both homogenous and heterogeneous populations.A number of methods have been developed to generate sc RNAseq libraries,each with distinctive advantages.Dropletbased approaches such as in Drop and Microwell-seq have emphasized the capture of large amounts of cells at the expense of individual cell detail.In contrast,commonly used plate-based approaches,such as SMARTseq2 and CEL-seq,can generate highly detailed profiles of small numbers of cells,including more information on splicing events and non-coding sequences.Despite their divergent protocols,all of these methods share a requirement for initial amplification of RNA content through whole transcriptome amplification(WTA),under the presumption that the initial c DNA quantity of any given cell is too minute to work with.However,biases in the transcriptome pool may result from the initial amplification,as some c DNA evade amplification/m RNA capture,and the unique molecular identifier(UMI)tagged date with low protein-coding gene coverage rate.These biases contribute to technical dropout,wherein uneven and pseudo-random detection of medium-and low-expressed genes significantly occlude sc RNA-seq results.Dropout events in sc RNA-seq are not exactly missing values.Those events have zero expression that mixed with real zeros,and has been recognized as a key concern for single-cell experiments.While several computational approaches have been designed to help overcome dropout through dimensionality reduction and imputation,such methods have difficulty in recovering true gene-gene co-expression relationships and consequent co-expression networks.In order to lower technical dropout,we streamline sc RNA-seq library construction at the experimental level.As such,in the sc STAT-seq(Streamlined Transcription And Tagmentation)workflow presented here,sorted cells by microfluidic chip are rapidly lysed,and subsequently reverse transcribed into c DNA with the use of conventional oligo-d T,together with not-sorandom hexamers(NSR)designed to help capture 5’ information.A rapid second-strand synthesis step is used to generate paired c DNA strands,which are then immediately tagmented via homebrew Tn5,such that c DNA amplification only occurs after fragmentation is completed.Samples are only pooled together following the completion of sequencing index ligation,and sample transfer steps were minimized,in order to prevent potential cross-contamination.The entire workflow can be completed by hand in 7 hours,with favorable per-cell cost compared to other common plate-based workflows(SMART-seq2).From our initial applications of sc STATseq to the mouse RAW264.7 macrophage cell line,we could recover a median of just under 8,000 protein-coding genes per cell with a detailed analysis of gene-gene co-expression relationships and mapping of cell physiological state.By comparing the sequencing information obtained from SMART-seq2,sc STAT-seq libraries are insulated against technical dropout,allowing for simple steps and more obvious cost advantages.Investigation of osteoclastogenesis using sc STAT-seq allowed us to detect the transcriptome changes at critical time nodes.Rich vesicle movements were found in osteoclastogenesis by using single-cell transcriptome profiling analysis with low technical dropout,in which Rab32 participated in the intracellular vesicle transport process.We observe the osteoporosis in Rab32flox/flox;Lys MCre mouse with the increasing number and activating function of osteoclasts.Rab32 knockout(Rab32KO)cell line(RAW264.7)was constructed by using the CRISPR-Cas9 system.Rab32 KO cells were observed to promote osteoclast differentiation in vitro,which repeated the result in vivo.The overexpression cell lines of Rab32 and its activated(Q83L)or inhibitory(T37N)mutants were constructed by the lentiviral system.It was found that overexpression of Rab32 could inhibit the differentiation process and the inhibition process was related to the form of Rab32 activation type.The transcriptome changes of Rab32 cells at the key differentiation nodes were analyzed by Bulk RNA-seq,and it was verified that Rab32 regulated the transport of DMT1 and affected the distribution of iron ions.Abnormal iron transport inhibited osteoclastogenesis.Overall,sc STAT-seq is the first solution aiming to solve the dropout problem in single cell sequencing.Using sc STAT-seq to detect the key nodes of osteoclastogenesis,we can explore the process of cell transcriptome changes at accurate level,and lay a foundation for the treatment of osteoporosis and other bone-related diseases. |
PDF Full Text Request