Cytometry meets next-generation sequencing - RNA-Seq of sorted subpopulations reveals regional replication and iron-triggered prophage induction in Corynebacterium glutamicum

Abstract

Phenotypic diversification is key to microbial adaptation. Currently, advanced technological approaches offer insights into cell-to-cell variation of bacterial populations at a spatiotemporal resolution. However, the underlying molecular causes or consequences often remain obscure. In this study, we developed a workflow combining fluorescence-activated cell sorting and RNA-sequencing, thereby allowing transcriptomic analysis of 10⁶ bacterial cells. As a proof of concept, the workflow was applied to study prophage induction in a subpopulation of Corynebacterium glutamicum. Remarkably, both the phage genes and flanking genomic regions of the CGP3 prophage revealed significantly increased coverage upon prophage induction - a phenomenon that to date has been obscured by bulk approaches. Genome sequencing of prophage-induced populations suggested regional replication at the CGP3 locus in C. glutamicum. Finally, the workflow was applied to unravel iron-triggered prophage induction in early exponential cultures. Here, an up-shift in iron levels resulted in a heterogeneous response of an SOS (P_divS) reporter. RNA-sequencing of the induced subpopulation confirmed induction of the SOS response triggering also activation of the CGP3 prophage. The fraction of CGP3-induced cells was enhanced in a mutant lacking the iron regulator DtxR suffering from enhanced iron uptake. Altogether, these findings demonstrate the potential of the established workflow to gain insights into the phenotypic dynamics of bacterial populations.

Figure 1

Heterogeneous prophage induction in C. glutamicum populations. For proof-of-concept studies, prophage induction was triggered by counter silencing as described previously. Phage induction was visualized by the means of fluorescent protein production using a fusion of the P_lys phage promoter to eyfp. (A) Growth of the strain C. glutamicum::Plys-eyfp/pAN6_N-cgpS treated with different IPTG concentrations. (B) Flow cytometry analysis after six hours revealed a significant fraction of prophage induced cells upon addition of >100 µM IPTG. (C) Contour plots of an induced (blue) and an uninduced sample (red). Gating strategy applied for the isolation of 10⁶ cells from phage positive and negative populations are shown in the right plot.

Figure 2

Experimental workflow for the transcriptome analysis of bacterial populations after cell sorting. Cultures were analyzed by flow cytometry and sorted by the means of the fluorescent reporter signal. One million cells were sorted and immediately treated with an RNA stabilization agent (RNAlater or RNAprotect). Subsequently, cells were concentrated on a filter plate, flash frozen in liquid nitrogen and stored at −80 °C. Prior to RNA extraction, the cells were treated with lysozyme and mutanolysine. The quality of RNA was determined as RIN value (>7 for samples used for sequencing). Extracted RNA of appropriate quality was then used for cDNA library preparation and sequencing.

Figure 3

Differential gene expression analysis of prophage induction in C. glutamicum. C. glutamicum::Plys-eyfp/pAN6_N-cgpS cells were cultivated in CGXII medium with 2% (w/v) glucose and either with or without 150 µM IPTG. (A) Flow cytometry data show the activation of the phage reporter (blue), compared to the uninduced control (red). (B,C) From both samples RNA was extracted (without cell sorting) and analysed by RNA-Sequencing. The Log2(foldchange) was plotted against the coding regions (orfs) of the C. glutamicum genome. (D) Fluorescence-activated cell sorting to separate induced from uninduced cells from a heterogeneous population. (E,F) Differential gene expression analysis of the induced versus the uninduced subpopulation (see also Table S2). (G) Comparison of expression values of the sorted samples versus the unsorted reference data set (R² = Pearson correlation coefficient).

Figure 4

Genome re-sequencing reveals regional replication at the CGP3 locus. C. glutamicum::Plys-eyfp/pAN6_N-cgpS cells were cultivated in the presence of 150 µM IPTG to induce N-cgpS expression triggering CGP3 prophage induction. Samples were taken at different time points (0, 3 and 6 hours), the DNA was extracted and sequenced. Plots show the respective mean normalized genomic coverage on a logarithmic scale against the respective genomic position. The CGP3 locus is highlighted in orange; flanking regions affected by regional replication are shaded in light orange.

Figure 5

Activation of the SOS response in C. glutamicum lag phase cultures triggered by iron fluctuations. C. glutamicum/pJC1_PdivS-venus cells were cultivated in CGXII medium with 2% (w/v) glucose under iron limitation (1 µM FeSO₄) and under conditions of sufficient iron supply (36 µM FeSO₄). (A) Stationary phase cells from the preculture (either 1 or 36 µM FeSO₄) were transferred into fresh CGXII medium with 36 µM FeSO₄ and analysed by flow cytometry. (B) Growth of the strains after the transfer into fresh medium. (C) Microscopy images of cells upon iron upshift confirmed a heterogeneous reporter output.

Figure 6

Iron triggered prophage induction in C. glutamicum. Shown is a differential gene expression analysis of subpopulations isolated from an iron upshift experiment. C. glutamicum/pJC1_PdviS-venus cells were first cultivated in CGXII medium with 1 µM iron and then shifted to medium with 36 µM iron. (A) Based on the output of the divS promoter fusion, 10⁶ cells of each subpopulation were sorted and treated according to the established RNA-Seq workflow. (B) Differential gene expression analysis: As expected, several SOS genes showed an upregulation in the divS⁺ subpopulation. Remarkably, also the expression of genes located within the CGP3 prophage element (red dots) were significantly increased indicating iron-triggered prophage induction.

Figure 7

Iron-triggered prophage activation is dependent on the cellular SOS response. (A) Time lapse studies of a ΔdtxR strain containing the genomically integrated prophage reporter P_lys-eyfp (strain: C. glutamicum ΔdtxR::P_lys-eyfp). Cells were grown in a microfluidic chip device (see material and methods) in CGXII minimal medium with 2% (w/v) glucose and under a constant flow rate of 300 nl min⁻¹. (B) C. glutamicum wild type, ΔdtxR and ΔrecA strains, containing the integrated P_lys-eyfp reporter, were cultivated under iron up-shift conditions (1 → 36 µM FeSO₄). After 4 h samples were taken and analyzed by flow cytometry. Whereas a prophage-induced subpopulation is clearly visible in Wt and ΔdtxR cells (1.28% and 4.61%, respectively), almost no activity of the phage reporter was observed in ΔrecA cells. Deletion of dtxR and recA was complemented by transforming the respective strains with the plasmids pAN6-dtxR or pAN6-recA, respectively. Shown are representative scatter plots from three biological replicates; absolute values varied between replicates whereas the overall trend was consistent.

Figure 8

Model of iron-triggered prophage induction. Under conditions of sufficient iron supply, the global iron regulator DtxR is in its (Fe²⁺-bound) active state and represses genes encoding high-affinity iron uptake systems while activating expression of ferritin. Under iron limiting conditions, DtxR dissociates from its target promoters and iron uptake is highly upregulated. A fast transition to medium containing high iron concentrations (e.g. 36 μM) will now lead to fast uptake of iron. Increased intracellular Fe²⁺ levels may cause oxidative stress and lead to SOS-triggered prophage induction in a fraction of cells (Figs 6 and 7).