Research Article

Microbial single-cell RNA sequencing by split-pool barcoding

See allHide authors and affiliations

Science  19 Feb 2021:
Vol. 371, Issue 6531, eaba5257
DOI: 10.1126/science.aba5257

You are currently viewing the abstract.

View Full Text

Log in to view the full text

Log in through your institution

Log in through your institution

Bacterial cell gene expression

Single-cell genomics in bacteria has lagged relative to in eukaryotes because of their tough bacterial cell walls, low messenger RNA content, and lack of many posttranscriptional modifications. To tackle this challenge, Kuchina et al. developed microbial split-pool ligation transcriptomics, or microSPLiT, a single-cell sequencing method for both Gram-negative and Gram-positive bacteria. Sequencing both Escherichia coli and Bacillus subtilis showed differences in the heat shock response. Examining B. subtilis transcriptional patterns revealed that a small fraction of cells grown in laboratory medium express a myo-inositol catabolism pathway, which the cell could use in nonlaboratory environments, thus highlighting how microSPLiT can identify rare cellular states.

Science, this issue p. eaba5257

Structured Abstract


Bacteria heterogeneously activate gene expression programs in response to environmental changes, stress, and other stimuli. Such behavior may serve as a bet-hedging strategy and is essential for population growth and survival. Because bet-hedging programs are often activated only in a small fraction of cells, their unbiased discovery and characterization remains challenging. High-throughput single-cell RNA sequencing (scRNA-seq) has become ubiquitous for analyzing cell types and states in eukaryotes, but existing methods are not easily adapted to bacteria. Given the need for bacterial scRNA-seq, we developed microSPLiT (microbial split-pool ligation transcriptomics), a low-cost and high-throughput approach tailored to microbes. microSPLiT labels the cellular origin of RNA through combinatorial barcoding and uses only basic laboratory equipment to profile tens of thousands of cells in a single experiment. microSPLiT overcomes challenges specific to bacteria such as their low mRNA content, diversity in cell size, and cell wall architecture.


The lack of single-cell resolution in bacterial RNA sequencing is a severe limitation, considering that even isogenic bacteria frequently exhibit functional subpopulations essential for fitness and survival. Gene expression heterogeneities are likely magnified in natural environments where microbes are exposed to complex chemical, physical, and biological factors, but to date, such functional variation remains obscured. Functional subpopulations may be rare and hard to culture, making their analysis challenging with existing fluorescence-based single-cell methods.


In a species-mixing experiment performed with Gram-negative Escherichia coli and Gram-positive Bacillus subtilis, microSPLiT detected on average 235 unique transcripts per cell in E. coli and 397 in B. subtilis. MicroSPLiT reliably distinguished heat-shocked from non–heat-shocked bacteria by expression of classical heat shock genes, demonstrating its ability to detect differential gene expression in mixed populations. We applied microSPLiT to >25,000 B. subtilis cells grown in rich media and sampled across the growth curve [optical density (OD) 0.5 to 6.0]. MicroSPLiT provided a comprehensive view of carbon utilization across all growth stages, detecting a surprising subpopulation associated with myo-inositol catabolism at intermediate ODs. Single-cell microscopy confirmed heterogeneous expression of the myo-inositol uptake and utilization operons. MicroSPLiT also identified gene expression programs associated with a range of bacterial lifestyles and stress responses. Our data revealed a rare subpopulation of less than 0.2% of all cells associated with prophage activation and another subpopulation of around 2% of cells in a state of competence. Gene expression within these clusters was in close agreement with prior reports on competence and prophage activation.


We developed a high-throughput bacterial scRNA-seq method and applied it to characterize B. subtilis growth in rich media. Our analysis revealed several heterogeneously activated gene expression programs, even though exponential growth is not typically associated with cellular heterogeneity. We were able to detect subpopulations of cells as rare as 0.142%, pointing to microSPLiT’s potential to uncover physiologically relevant rare cell states that are hard to study by bulk or low-throughput methods. The same microSPLiT protocol worked reliably for a mix of Gram-negative and Gram-positive model bacteria; we expect that this versatility together with the scalability inherent to combinatorial barcoding will make microSPLiT a powerful tool for the investigation of complex natural and engineered microbial communities.

MicroSPLiT uncovers transcriptional states in bacterial populations.

MicroSPLiT is a scRNA-seq method for bacteria that uses combinatorial barcoding to label bacterial RNA by cell of origin; the steps in the process are shown. In an experiment performed with >25,000 B. subtilis cells sampled along a growth curve from early exponential to early stationary phase, microSPLiT uncovered a wide range of developmental and metabolic gene expression programs, including rare cell states associated with competence, prophage activation, and myo-inositol metabolism.


Single-cell RNA sequencing (scRNA-seq) has become an essential tool for characterizing gene expression in eukaryotes, but current methods are incompatible with bacteria. Here, we introduce microSPLiT (microbial split-pool ligation transcriptomics), a high-throughput scRNA-seq method for Gram-negative and Gram-positive bacteria that can resolve heterogeneous transcriptional states. We applied microSPLiT to >25,000 Bacillus subtilis cells sampled at different growth stages, creating an atlas of changes in metabolism and lifestyle. We retrieved detailed gene expression profiles associated with known, but rare, states such as competence and prophage induction and also identified unexpected gene expression states, including the heterogeneous activation of a niche metabolic pathway in a subpopulation of cells. MicroSPLiT paves the way to high-throughput analysis of gene expression in bacterial communities that are otherwise not amenable to single-cell analysis, such as natural microbiota.

View Full Text

Stay Connected to Science