Mechanisms for Differential Protein Production in Toxin-Antitoxin Systems

Abstract

Toxin-antitoxin (TA) systems are key regulators of bacterial persistence, a multidrug-tolerant state found in bacterial species that is a major contributing factor to the growing human health crisis of antibiotic resistance. Type II TA systems consist of two proteins, a toxin and an antitoxin; the toxin is neutralized when they form a complex. The ratio of antitoxin to toxin is significantly greater than 1.0 in the susceptible population (non-persister state), but this ratio is expected to become smaller during persistence. Analysis of multiple datasets (RNA-seq, ribosome profiling) and results from translation initiation rate calculators reveal multiple mechanisms that ensure a high antitoxin-to-toxin ratio in the non-persister state. The regulation mechanisms include both translational and transcriptional regulation. We classified E. coli type II TA systems into four distinct classes based on the mechanism of differential protein production between toxin and antitoxin. We find that the most common regulation mechanism is translational regulation. This classification scheme further refines our understanding of one of the fundamental mechanisms underlying bacterial persistence, especially regarding maintenance of the antitoxin-to-toxin ratio.

Keywords: RNA-seq; Ribo-Seq; persister; ribosome profiling; toxin–antitoxin systems.

Figure 1

Schematic of a typical Type II toxin–antitoxin (TA) system. A TA system operon is transcribed to produce a corresponding mRNA, which is then translated to produce toxin and antitoxin proteins. With sufficient concentration of anti-toxin protein, toxin protein can be primarily neutralized in a complex, which allows the cell to maintain a non-persister state, or else the toxin can exist as a free and active protein in the cell, which leads to persistence [15]. The antitoxin-to-toxin protein ratio, which is expected to be sufficiently greater than 1.0 in the non-persister state, controls these scenarios. Antitoxin is actively degraded by proteases (at a greater rate than the toxin), which requires the excess production of antitoxin to ensure the non-persister state. T: toxin. A: antitoxin. Not to scale.

Figure 2

TA (toxin–antitoxin) systems were classified into four different classes based on DNA sequence and measured mRNA products. (A–D) Representative wire diagrams of the operon organization and mRNA products of each class (not to scale). (A) Class 1 TA systems (FicAT, MazEF, MqsAR, PrlF-YhaV, RelBE, YefM-YoeB), the most abundant class in our study, have a single transcript for the operon and should rely on translational regulation to ensure higher antitoxin production relative to toxin production; (B) Class 2 is characterized by a second promoter that produces a slightly shorter transcript, which is predicted by our work to have a lower toxin translation rate than the transcript of the first promoter. The one example (HicAB) available has non-overlapping coding regions with the toxin upstream from the antitoxin; (C) Class 3 TA systems (DinJ-YafQ, YafNO) have a truncated transcript in addition to the whole transcript for the operon, due to some unknown mechanism (perhaps a terminator or post-transcriptional degradation); (D) Class 4 TA systems (RnlAB) produce two transcripts: a transcript of the whole operon (both toxin and antitoxin mRNA), and a transcript that can only be translated to antitoxin; (E) A summary of the classification of TA systems in this study. For the promoter column, the number indicates the number of promoters that are located upstream the coding regions (external, Ex), or within the coding regions (internal, In). Each class has a different A-to-T RNA ratio (see Table 2) based on analysis of 13 different RNA-seq datasets from a variety of conditions, such as growth in rich and minimal media, and cell densities (Table S1). * Some genes may have additional promoters, but they did not affect the ratio of mRNA expression. P: Promoter. RBS: Ribosome Binding Site. A: Antitoxin. T: Toxin.

Figure 3

RNA-seq coverage of antitoxin vs. toxin open reading frames. Left: dataset GSE48829 [26] triplicate biological replicates sampled during exponential growth in minimal media. Right: dataset GSE74809 [27] duplicate biological replicates sampled from five different stages of growth in M9 (glucose) media. Both plot the quantitative analysis of sequence coverage of antitoxin and toxin (see Methods) on a common axis for a variety of TA systems. Most TA systems in the conditions considered have less than a two-fold difference in coverage (1:1 coverage is indicated by a dashed line) between antitoxin-to-toxin mRNA, suggesting expression of the mRNA as a single transcript. TA systems that fall within the dotted lines had a 1:2 to 2:1 ratio of antitoxin to toxin coverage. TA systems were not included if either toxin or antitoxin had an average of less than one read per base for more than half of the datasets (the minimum read count for a gene is 168 reads). The error was calculated in two different directions (ratio and magnitude, see Methods), and error bars are aligned to these primary directions to illustrate the low error of the ratio. The individual replicates had similar groupings (Figure S2). Our results are also supported by a global error analysis (Figure S3), which shows a typically small error for the replicates. Units of coverage are Reads Per Kilobase Million (RPKM).

Figure 4

Protein synthesis rates plotted for each TA system based on Ribo-Seq data. A systematic bias towards higher antitoxin protein synthesis rate relative to toxin is evident in all but one case, and we explain the one outlier RnlAB in Section 2.5. The dashed line represents a 1:1 ratio of antitoxin-to-toxin, and the dotted lines mark a two-fold difference. TA systems with values of low confidence (less than 128 reads) in protein synthesis data [39] were not included in the figure. Error bars assume a 30% error, as estimated in Li et al. 2014 [39]. Arbitrary units: AU.