Rapid Evolution of Citrate Utilization by Escherichia coli by Direct Selection Requires citT and dctA

Abstract

The isolation of aerobic citrate-utilizing Escherichia coli (Cit(+)) in long-term evolution experiments (LTEE) has been termed a rare, innovative, presumptive speciation event. We hypothesized that direct selection would rapidly yield the same class of E. coli Cit(+) mutants and follow the same genetic trajectory: potentiation, actualization, and refinement. This hypothesis was tested with wild-type E. coli strain B and with K-12 and three K-12 derivatives: an E. coli ΔrpoS::kan mutant (impaired for stationary-phase survival), an E. coli ΔcitT::kan mutant (deleted for the anaerobic citrate/succinate antiporter), and an E. coli ΔdctA::kan mutant (deleted for the aerobic succinate transporter). E. coli underwent adaptation to aerobic citrate metabolism that was readily and repeatedly achieved using minimal medium supplemented with citrate (M9C), M9C with 0.005% glycerol, or M9C with 0.0025% glucose. Forty-six independent E. coli Cit(+) mutants were isolated from all E. coli derivatives except the E. coli ΔcitT::kan mutant. Potentiation/actualization mutations occurred within as few as 12 generations, and refinement mutations occurred within 100 generations. Citrate utilization was confirmed using Simmons, Christensen, and LeMaster Richards citrate media and quantified by mass spectrometry. E. coli Cit(+) mutants grew in clumps and in long incompletely divided chains, a phenotype that was reversible in rich media. Genomic DNA sequencing of four E. coli Cit(+) mutants revealed the required sequence of mutational events leading to a refined Cit(+) mutant. These events showed amplified citT and dctA loci followed by DNA rearrangements consistent with promoter capture events for citT. These mutations were equivalent to the amplification and promoter capture CitT-activating mutations identified in the LTEE.IMPORTANCE E. coli cannot use citrate aerobically. Long-term evolution experiments (LTEE) performed by Blount et al. (Z. D. Blount, J. E. Barrick, C. J. Davidson, and R. E. Lenski, Nature 489:513-518, 2012, http://dx.doi.org/10.1038/nature11514 ) found a single aerobic, citrate-utilizing E. coli strain after 33,000 generations (15 years). This was interpreted as a speciation event. Here we show why it probably was not a speciation event. Using similar media, 46 independent citrate-utilizing mutants were isolated in as few as 12 to 100 generations. Genomic DNA sequencing revealed an amplification of the citT and dctA loci and DNA rearrangements to capture a promoter to express CitT, aerobically. These are members of the same class of mutations identified by the LTEE. We conclude that the rarity of the LTEE mutant was an artifact of the experimental conditions and not a unique evolutionary event. No new genetic information (novel gene function) evolved.

FIG 1

Direct selection of E. coli in minimal M9C yielded Cit⁺ mutants for both E. coli wild-type and E. coli ΔrpoS::kan strains but not for the E. coli ΔcitT::kan mutant strain. (A) Results of two separate representative experiments are shown. The E. coli wild-type (wt), E. coli ΔrpoS::kan, and E. coli ΔcitT::kan strains were inoculated into individual 250-ml flasks containing 50 ml of M9 minimal medium with citrate as the sole carbon source. E. coli wild-type and E. coli ΔcitT::kan levels decreased by 1 log until after day 23, when the wild-type strain (designated DV159) reinitiated exponential growth. (B) Results of a parallel experiment are shown. The level of the E. coli ΔrpoS::kan strain decreased by almost 2 log until after day 37, when it reinitiated exponential growth. This strain was designated DV130.

FIG 2

Modified direct selection of E. coli yielded Cit⁺ mutants independently of strain and citrate concentration. Mutants arose for the E. coli K-12, E. coli K-12 ΔrpoS::kan, E. coli B, and E. coli REL606 strains but not for the E. coli ΔcitT::kan strain or the E. coli ΔdctA::kan strain. Results of four separate representative experiments are shown. Each flask was inoculated with ∼5.0 × 10⁵ CFU/ml, and every 7 days, cultures were diluted 1:100 into fresh medium. OD₆₀₀ levels were measured before transfer. (A and B) The E. coli K-12 wild-type, E. coli ΔrpoS::kan, E. coli ΔdctA::kan, and E. coli ΔcitT::kan strains were inoculated into individual 250-ml flasks containing 50 ml of M9 minimal medium, 6.8 mM citrate, and either 0.0025% glucose (M9C25) or 0.005% glycerol (M9CG50). This level of glucose or glycerol supported six generations of growth. (A) In M9C25, Cit⁺ mutants arose for the E. coli K-12 and E. coli ΔrpoS::kan strains after the forth and ninth transfers, respectively. (B) In M9CG50, Cit⁺ mutants arose for the E. coli K-12 and E. coli ΔrpoS::kan strains after the third transfer. Citrate-supported growth was not recovered for either the E. coli ΔdctA::kan strain or the E. coli ΔcitT::kan strain. (C and D) Similarly, the E. coli K-12 wild-type, E. coli B, and E. coli REL606 strains were inoculated into M9 minimal medium containing 1.7 mM citrate and either 0.0025% glucose (M9LC25) or 0.005% glycerol (M9LCG50). Cit⁺ mutants arose in E. coli K-12 after five or four transfers, in E. coli B after four or five transfers, and in E. coli REL606 after eight and five transfers (panels C and D, respectively).

FIG 3

Extended incubation of actualized E. coli Cit⁺ bacteria on Simmons citrate agar yielded refined E. coli Cit⁺ papillae. Primary direct selection M9C broth that displayed growth was plated on Simmons citrate agar. Small yellow colonies appeared after 3 to 4 days of incubation, and small blue papillae (irregular darkened regions) appeared as shown after 10 to 20 days. Each blue papilla represents an independent mutation to the refinement stage.

FIG 4

Mass spectrometry showed that citrate depletion correlated with E. coli growth. Two E. coli Cit⁺ mutants derived from E. coli wild-type isolates, DV133T (A) and DV159 (B), were inoculated into M9C minimal medium, and the OD₆₀₀ was measured every 6 h. Triplicate medium samples at each time point were subjected to filter sterilization, and a subset of samples corresponding to time zero and the late lag and early-, mid-, and late-exponential and stationary phases were analyzed by mass spectrometry for citric acid. Bars represent ± standard errors (S.E.).

FIG 5

E. coli Cit⁺ mutants grew in clumps in M9C broth but not in LB broth. Results of phase-contrast microscopy comparing levels of growth of E. coli Cit⁺ mutants DV133 (A and B) and DV159 (C and D) in M9 (A and C) C and LB broth (B and D) are shown. Cells in M9C minimal medium were clumped, whereas growth in LB broth was dominated by single and predivisional cells. The pictures are representative of 10 microscopic fields; magnification, ×400.

FIG 6

Genomic DNA sequence analysis of E. coli Cit⁺ phenotypes showed gene amplifications associated with dctA and citT regions and promoter captures. (A) The genomic read depth profiles of E. coli Cit⁺ isolates DV130, DV133, and DV133T. The weakly Cit⁺ phenotype of DV130 shows 4-fold-increased coverage of the citT region and 2-fold-increased coverage of dctA compared to the ∼100× coverage of the rest of the chromosome. (B) Strain DV133 was a strong Cit⁺ phenotype derived from DV130. It showed a 2-fold reduction in citT coverage compared to the parental DV130 strain and the insertion of insl1 5′ into citT. When the E. coli K-12 wild-type strain was transduced with a lysate made using DV133, the same insl1-citT fusion was identified, confirming that this promoter capture was responsible for the transduced strong Cit⁺ phenotype. The genomic sequence of Cit⁺ strain DV159 showed a citT duplication and deletion event that fused the promoter of uspG to citG, another promoter capture event. All read depth profiles show 2-fold-increased coverage for the dctA region compared to the ∼100× coverage for the chromosome. (C) Sequence analysis of DV133, DV133T, and DV159 showed recombination between rhsA and rhsB based on the differences in the positions of flanking genes yibF and yrhC. The genomic organization of the E. coli wild-type chromosome is represented by the top panel. The recombination of sister chromosomes to generate a large 140-kb duplication of this region is shown in the center panel. The gene map determined for both DV133 and DV159 is shown in the bottom panel with the dctA duplication.