Deep mutational scanning reveals the molecular determinants of RNA polymerase-mediated adaptation and tradeoffs

Abstract

RNA polymerase (RNAP) is emblematic of complex biological systems that control multiple traits involving trade-offs such as growth versus maintenance. Laboratory evolution has revealed that mutations in RNAP subunits, including RpoB, are frequently selected. However, we lack a systems view of how mutations alter the RNAP molecular functions to promote adaptation. We, therefore, measured the fitness of thousands of mutations within a region of rpoB under multiple conditions and genetic backgrounds, to find that adaptive mutations cluster in two modules. Mutations in one module favor growth over maintenance through a partial loss of an interaction associated with faster elongation. Mutations in the other favor maintenance over growth through a destabilized RNAP-DNA complex. The two molecular handles capture the versatile RNAP-mediated adaptations. Combining both interaction losses simultaneously improved maintenance and growth, challenging the idea that growth-maintenance tradeoff resorts only from limited resources, and revealing how compensatory evolution operates within RNAP.

Fig. 1. Mutations in the “target” sequence of the RNAP β subunit (RpoB) improve fitness in diverse conditions.

a CREPE-based experimental setup to estimate the fitness of RNAP mutations in multiple environments. b An adaptive evolution mutation map of RpoB depicting mutations identified in different environments (colored peaks), with the target region highlighted. Conditions corresponding to the peaks have been highlighted below the figure. c The target region (red cartoon) in the RNAP is proximal to the catalytic core (pink MG²⁺ sphere), DNA (blue), and the RpoC subunit (the Bridge-Helix) (Gray). d The rolling average Consurf score (blue line, averaged over a 15 amino acid window size) for RpoB. The target region occurs between the dashed red lines. Source data are provided as a Source Data file.

Fig. 2. Correlation of fitness between environments to identify cross-stress adaptation and tradeoffs for target mutations.

a Expected correlation of fitness of variants between environments in the case of cross-environment adaptation (left) or tradeoffs (right). b Correlation of the fitness for target mutations measured in M9 minimal media with glucose to the fitness measured in M9 media (blue spheres, left to right) with glycerol, galactose, glucose, and 0.3 M NaCl, glucose, and 0.6% (v/v) butanol respectively. The Spearman correlation coefficient and the associated p-value are mentioned at the top of each curve. c Correlation of fitness for different variants within the target region measured in M9 minimal media (blue spheres) with glucose with the enrichment of the same variants in the ΔrelAΔspoT strain of Escherichia coli upon plating on M9 minimal media agar with Glucose (see Supplementary Fig. 3 for detailed information). Source data are provided as a Source Data file.

Fig. 3. Modular residues clusters are important for growth and stringent response.

a Residue-wise mean growth-associated fitness for all mutations (green triangles, y-axis on the left) and mean stringent enrichment (blue triangles, y-axis on the right). b Residues with high mean growth-associated fitness (green spheres) and high mean stringent enrichment (blue spheres) were mapped on the structure of the target region. c Correlation between growth-associated fitness and the minimum distance of the residue from the β’ RNA polymerase subunit (RpoC) (left), and growth-associated fitness and the minimum distance of the residue from the DNA (RpoC) (right). d Correlation between stringent enrichment and the minimum distance of the residue from the β’ RNA polymerase subunit (RpoC) (left), and growth-associated fitness and the minimum distance of the residue from the DNA (RpoC) (right). Source data are provided as a Source Data file.

Fig. 4. Stringent mutations within the target.

a Residue-wise distribution of stringent enrichment for each variant (dots). Variants (blue dots) above the cut-off (grey dashed line, Supplementary Fig. 3C) are stringent mutations of the RNAP. The yellow triangles represent the mean enrichment score. b Internal (dark blue) and DNA-proximal (light blue) target residues with greater than five unique stringent mutations mapped on the target structure. c (Left) DNA-proximal residues with high mean stringent enrichment (blue sticks). (Right) A bar plot (blue) of the stringent enrichment score of non-synonymous (blue) and synonymous (red) mutations in the residue G534 relative to the cutoff stringent-enrichment score (grey dashed line, Supplementary Fig. 3). Within each box, the horizontal black lines represent median values, lower 25^th percentile and upper 75th percentile bounds, and the whiskers represent extreme values in the 1.5× interquartile range. There were 28, 10, 64, 122, 4, 2, 40, 6, and 2 independent observations for variants top-bottom in the plot. An observation represents an independent fitness measurement of a synonymous variant of the focal mutation in two biological replicates (Supplementary Note 1). d (Left) Residues with high mean stringent enrichment (blue sticks) with internal polar interactions (red lines). (Right) A bar plot (blue) of the stringent enrichment score of non-synonymous (blue) and synonymous (red) mutations for some highlighted residues relative to the cutoff stringent-enrichment score (grey dashed line, Supplementary Fig. 3). Within each box, the horizontal black lines represent median values, lower 25th percentile and upper 75th percentile bounds, and the whiskers represent extreme values in the 1.5× interquartile range. There were 70, 2, 2, and 6 independent observations for T533 variants and 44, 42, 12, and 5 independent observations for R557 variants left to right in the plot. An observation represents an independent fitness measurement of a synonymous variant of the focal mutation in two biological replicates (Supplementary Note 1). e Residues with high mean stringent enrichment (blue) with red lines connecting two residues with significant positive epistatic interaction (Supplementary Fig. 4 and Supplementary Note 2). Source data are provided as a Source Data file.

Fig. 5. Growth-improving mutations within the target.

a Distribution of fitness effects for the growth-associated fitness measures in M9 minimal media + Glucose for all (blue) and synonymous mutations (orange). b Residue-wise distribution of growth-associated fitness for beneficial variant (blue dots), synonymous variants (gray dots), and deleterious variants (red dots) with the cut-off for beneficial fitness score (top gray line) and deleterious fitness score (bottom gray line). c Residue-level mean growth-associated fitness heatmap on the target structure, with increasing fitness represented as a gradient from red to blue. d Residues with high mean growth-associated fitness (green sticks), and highly conserved and functionally important Bridge Helix (BH) residues (orange sticks). (Left) BH (bridge helix) residues with charged side chains (blue sticks) and proximal target residues with high growth-associated fitness means (red sticks). e (Right) A bar plot (blue) of mean growth-associated fitness for substitutions to amino acids with positively charged (blue) and negatively charged (red) sidechains respectively. Within each box, the horizontal black lines represent median values, lower and upper bounds correspond to the 25th and 75th percentile, and the whiskers extend to the extreme values within the 1.5× interquartile range. The top and bottom dashed lines represent the cutoff for beneficial and deleterious fitness scores respectively. There were 6, 2, 2, 28, 34, 10, and 4 independent observations for each variant left to right in the plot. An observation represents an independent fitness measurement of a synonymous variant of the focal mutation in two biological replicates (Supplementary Note 1). f Correlation of growth-associated fitness with the enrichment for CR703 resistance for target mutations. The dashed vertical and horizontal lines represent the cut-off for beneficial growth-associated fitness and CBR703 resistance respectively. g Residues with greater than five CBR703 resistant mutations (all spheres) and ones with high growth-associated mean fitness (green spheres). h Residues with high growth-associated fitness (green sticks) with red lines connecting two residues with significant positive epistatic interaction (Supplementary Fig. 4 and Supplementary Note 2). Source data are provided as a Source Data file.

Fig. 6. Some mutations are both growth-improving and stringent.

a The residue N573 (cyan sticks) with internal polar interaction (red line) with residue C559 is close to BH positively charged residue R780. b A bar plot (blue) of mean growth-associated fitness for substitutions to amino acids with positively charged (blue), negatively charged (red), and synonymous substitutions (red). On the left, the top and bottom gray lines represent the cut-off for beneficial fitness score and deleterious fitness score respectively. On the right, the dashed line represents cut-off for stringent mutations. Within each box, the horizontal black lines represent median values, lower and upper bounds correspond to the 25th and 75th percentile, and the whiskers extend to the extreme values within the 1.5× interquartile range. There were 34, 10, 4, and 2 independent observations for Glucose-associated fitness (left) and 72, 2, 4, and 2 independent observations for stringent enrichment (right) for each variant left to right in the plot. An observation represents an independent fitness measurement of a synonymous variant of the focal mutation in two biological replicates (Supplementary Note 1). c Correlation of growth-associated fitness and stringent enrichment of all double mutations (grey), and ones with combined synonymous mutations (green), combined growth-improving and stringent mutations (64, blue), and combined mutations on positions important for growth and stringent phenotypes (41, pink. 105 total). d Box plots comparing stringent enrichment (left) and growth-associated fitness (right) of double synonymous mutations (136 variants, blue) and double growth-stringent mutations (91 variants, orange). The red line represents cut-off for stringent mutations (left) and the cut-off for beneficial growth-associated mutations (right). Source data are provided as a Source Data file.

Fig. 7. Conservation and compensatory evolution.

a Bar plot of percentage variability in residues with high growth-associated fitness means (green), high mean stringent enrichment (blue), and growth-stringent residues (pink). (Bottom) An alignment of the target sequence from E. coli with the RNAP IV sequence from plant chloroplasts from Sansevieria trifsciate (ST), Arabidopsis thaliana (AT), RPB2 subunit of RNA polymerase II from Homo sapiens (HS), Saccharomyces cerevisiae (SC) and RPO1N RNAP subunit from Methanocaldococcus jannaschii (MJ) and Saccharolobus solfataricus (SS). Source data are provided as a Source Data file. b Distribution of growth-associated fitness for synonymous mutations and compensatory mutations for Rifampicin resistance. The statistics were derived using n-synonymous = 186 and n-Compensatory = 356 measurements of Glucose fitness. The red line represents the cutoff for beneficial growth-associated mutations. c Residues with high growth-associated fitness (green sticks) with significant positive epistasis (red lines) with rifampicin-resistant stringent mutations (orange spheres). d Summary: two modular clusters in the target region associated with unique interactions and possibly functions control the growth and stringent phenotypes.