64 4 Rational Efforts to Streamline the Escherichia coli Genome
Recently, rapid streamlining and genome-wide inactivation of IS elements
were accomplished by genome shuffling between different E. coli strains, fol-
lowed by multiplex genome modifications by CRISPR/Cas-assisted MAGE [95].
First, prophages were deleted by shuffling prophage-free segments of multiple
deletion series (MDS) genomes into E. coli BL21 by P1 transduction. This was
followed by subsequent rounds of CRISPR/Cas-assisted MAGE on multiplex IS
targets, disrupting the transposases of the IS elements. With the growing num-
ber of reduced-genome strains, such recycling of streamlined genomes might
accelerate strain construction.4.7 Genome-Reducing Efforts and the Impact of Streamlining
of Streamlining
4.7.1 Comparative Genomics-Based Genome Stabilization
and Improvement
The first systematic, large-scale genome reduction project was aimed at remov-
ing the largest K12-specific genomic islands from the MG1655 genome [85]
(Figure 4.4). Identification of the K-islands was based on the sequence com-
parison of three E. coli genomes available at that time (MG1655, enterohemor-
rhagic O157:H7, and uropathogenic CFT073). Via a series of linear
DNA-mediated recombineering steps, including a novel way of I-SceI-
stimulated scarless resolution of the recombination intermediate, 12 precise
deletions were created and combined in a single strain. Compared with the
parental MG1655, the resulting MDS12 (multiple deletion series strain with 12
deletions) had a genome reduced by 8.1%, with 9.3% of the genes deleted. All
prophages and 24 of the 44 transposable elements present in the MG1655
genome were deleted. Growth rates of MDS12 in minimal and rich medium
were similar to those of MG1655. Doubling times were nearly identical, but
MDS12 reached 10% higher density in stationary phase. Electroporation and
transformation efficiencies of the parental and the MDS12 strain were identi-
cal. This first attempt of drastic genome streamlining proved that by applying
a rational design strategy, a large fraction of the genes can be removed from an
organism that has been shaped by billions of years of evolution. Moreover, this
could be done without losing robustness and rapid growth, at least under the
laboratory conditions tested.
The next milestones of this project were the IS-free MDS41, MDS42, and
MDS43 strains with 14.28, 14.30, and 15.27% of the genome deleted, respec-
tively [29]. Deletion targets were primarily selected by comparative genomics of
several sequenced strains (RS218, CFT073, Shigella flexneri 2457T, O157:H7
EDL933, and DH10B) and by assessment of literature data on the particular
gene functions. Major K-islands were targeted, but deletions were in several
cases extended to include neighboring nonessential genes with no impact on
growth in either rich or minimal media. Deletions were tested for growth prop-
erties both individually and when combined in a single strain. Growth rates of
the MDS cells were similar to that of the parental MG1655. Elimination of