Genome evolution of Acinetobacter baylyi ADP1 during laboratory domestication: acquired mutations impact competence and metabolism

Abstract

The bacterium Acinetobacter baylyi is a model organism known for its extreme natural competence and metabolic versatility. It is capable of taking up environmental DNA at a high rate across all growth phases. The type strain ADP1 was created by random mutagenesis of a precursor strain, BD4, to prevent it from forming cell chains in culture. ADP1 has since been distributed between research groups over several decades and acquired subsequent mutations during this time. In this study, we compare the genome sequences of A. baylyi BD4 and its modern descendants to identify and understand the effects of mutations acquired and engineered during its domestication. We demonstrate that the ADP1 variants in use today differ in their competence, growth on different carbon sources, and autoaggregation. In addition, we link the global carbon storage regulator CsrA and a transposon insertion that removes its C-terminal domain specifically to changes in both overall competence and an almost complete loss of competence during the stationary phase. Reconstructing the history of ADP1 and the diversity that has evolved in the variants currently in use improves our understanding of the desirable properties of this experimentally and industrially important bacterium and suggests ways that its reliability can be improved through further genome engineering.IMPORTANCEAcinetobacter baylyi ADP1 is a bacterial chassis of interest to microbiologists in academia and industry due to its extreme natural competence and wide metabolic range. Its ability to take up DNA from its environment makes it straightforward to efficiently edit its chromosome. We identify and characterize mutations that have been passed down to modern strains of ADP1 from the initial work in the 1960s, as well as subsequent mutations and genome edits separating strains in use by different research groups today. These mutations, including one in a global regulator (CsrA), have significant phenotypic consequences that have affected the reproducibility and consistency of experiments reported in the literature. We link a mutation in this global regulator to unexpected changes in natural competence. We also show that domesticated A. baylyi strains have impaired growth on a variety of carbon sources.

Keywords: Acinetobacter baylyi; domestication; laboratory evolution; natural competence; post-transcriptional regulation.

Fig 1

Maximum parsimony phylogeny of A. baylyi strains inferred from mutations that evolved during domestication. Strains were grouped by the presence or absence of mutations shown in Table S1. Branch labels correspond to sets of mutations that evolved along each branch. Targeted deletions of IS elements during the creation of strain ISx are described in (21). Circles indicate strains used in assays in this study.

Fig 2

Mutations leading to aggregation in A. baylyi strains. (A) Microscope images of strains ATCC 33304 and 33305 grown in LB medium. (B) Metabolic pathways showing conversion of glucose to capsular polysaccharides. Genes mutated in ATCC 33304 and 33305 are shown. Pathways adapted from KEGG (29). (C) Reverting pgi or galU mutations in ATCC 33304 and 33305, respectively, produced single cells in LB. Scale bars represent 5 µM.

Fig 3

A. baylyi strains form cell chains in minimal media. (A) Domestication in BD413_rev and ADP1(D) prevented the formation of cell chains in S2 medium with glucose. (B) Cell chains formed for all strains when succinate was used as a carbon source in S2 medium. (C) All strains, including BD4_rev, grew unicellularly in LB medium. The image of BD413_rev in LB is taken from the same field of view as in Fig. 2C. Scale bars are 5 µM.

Fig 4

Transformation efficiency varies between domesticated A. baylyi strains. (A) Transformation frequencies of BD4_rev (black), BD413_rev (blue), ADP1(D) (orange), and ISx (green) strains incubated overnight with genomic DNA carrying a specR resistance gene. (B) Transformation frequencies of the same strains during log (squares) and stationary (triangles) phases. At each time point, each strain was incubated with transforming DNA for 30 min, then the remaining free DNA was digested by DNase I before plating to count transformants. (C) Transformation frequencies during log (squares) and stationary (triangles) phases for ADP1 and ISx compared to an ISx mutant with a truncated csrA gene (yellow). *P < 0.05 for Welch’s t-test.

Fig 5

Domestication impacted growth and catabolite repression in A. baylyi. Growth curves of A. baylyi strains in (A) LB medium and in S2 media with (B) 25 mM succinate, (C) 3 mM glucose, (D) 3 mM glucose and 2.5 mM succinate, (E) 2 mM benzoate, or (F) 20 mM vanillate. Lines represent the average OD₆₀₀ of six replicates each and error bars depict standard error. The strains tested were BD4_rev (black), BD413_rev (blue), ADP1(D) (orange), and ISx (green). E and F include an ISx mutant with the csrA truncation from ADP1(D) (yellow).

Fig 6

(A) Mutations evolved in BD4_rev (MRV001) during 1 month of evolution in ABMS medium. Full details are provided in Table S1. (B) Potential post-transcriptional regulatory network inferred in A. baylyi from interactions between CsrA, Hfq, and RNase D. (1) Hfq protects csrA mRNA from degradation by RNase E (30). (2) CsrA proteins are bound by a CsrB-like small RNA (31). (3) CsrA binds to Shine-Dalgarno sites, preventing ribosome assembly on mRNAs (32). (4) CsrA protects mRNAs from degradation by RNase D (rnd). (5) CsrA protects small regulatory RNAs from degradation by RNases (33). Panel B created with Biorender.