Discussion
The occurrence of random mutations can improve genetic diversity and
play an important role in many fields.[1-4, 6, 9]To date, researchers have developed a handful of mutagenesis techniques
to increase the mutation rate and mutation
diversity[6]. These tools have produced achievable
effects in various chassis cells such as E. coli , S.
cerevisiae , and mammalian cells. In this work, we developed a targetedin vivo mutagenesis tool that can significantly improve the
mutation rate and broaden the mutation spectrum by fusing DNA-modifying
enzymes, cytidine deaminase and T7 RNAP.
We constructed different mutagenesis proteins that could accelerate the
evolution of the target gene. We first constructed the plasmid
containing the cytidine deaminase PmCDA1-T7 RNAP expression cassette and
inserted T7 promoter sequence upstream of the target gene. PmCDA1-T7
RNAP fusions could raise the mutation frequency to about
1.0~1.2x10-3, and exhibited a strong
bias toward C→T mutations, which is consistent with the mutation
characteristic of PmCDA1.[21, 22] We hypothesized
that enhancing the removal of mismatched U as well as the subsequent DNA
repair processes would lead to diverse mutation
types.[34, 35, 46] Therefore, we fused different
DNA-modifying enzymes with PmCDA1 to improve the mutation effect. The
results indicated that the introduction of DNA-modifying enzymes could
indeed improve the mutation effect. Among these candidates, MAG1 and
REV3 could significantly increase the diversity of mutations. In the
mutations generated by MAG1-PmCDA1-T7 RNAP, C→T mutations accounted for
64.5%, followed by G→A mutations (19.2%), C→G mutations (12.5%), and
G→C mutations (3.8%). The ratio of non-C→T mutations was 36.5%, which
was 11-fold higher than that of PmCDA1-T7 RNAP. REV3-PmCDA1-T7 RNAP
could also generate more diverse mutations, of which C→T mutations
comprised about 29.6%, nearly 10-fold higher than that of PmCDA1-T7
RNAP. The mutation diversity is crucial to the evolutionary process and
the wider mutation spectrum would help us to obtain desired strains more
efficiently. Smolke et al. developed TRIDENT system and increased the
ratio of non-C→T mutations to about 20%,[23]while Shoulders et al. fused evolved adenosine deaminase to T7 RNAP and
developed MutaT7A→G and eMuataT7A→G,
which could generate all transition mutations when being employed with
cytidine-bearing mutators.[25] As our mutagenesis
fusions could generate higher proportion of C→G and G→A mutations, they
could play a complementary role with the above-mentioned tools. Except
for the change on the mutation spectra, the addition of DNA-modifying
enzymes could also raise the mutation frequency. EXO1-PmCDA1-T7 RNAP
could increase the mutation rate up to 2.2x10-3, which
is twice as high as PmCDA1-T7 RNAP.
We then added a reverse T7 promoter at the end of the target gene and
compared the mutation effects with that of the single T7 promoter
system. The introduction of the second T7 promoter could significantly
increase the mutation frequency of the target gene. The mutation
frequencies of strains with dual T7 promoters were about
1.5~2.5-fold higher than that in the single T7 promoter
system. We hypothesized that the dual T7 promoters might increase the
possibility of T7 RNAP binding to the T7 promoter, thus increasing the
mutation rate. The proportions of different mutations generated in the
dual T7 promoter system were basically the same as the single promoter
system. Both Kim et al. and Shoulders et al. have found that the
introduction of the second T7 promoter increased the ratios of G→A
mutations.[22, 24] However, we did not observe
apparent change of the frequency of G→A mutations in the dual T7
promoter system. Instead, we found that the ratios of some transversion
mutations (G→T, G→C etc.) increased than that in the single promoter
system. We suspect that this difference might be due to the different
DNA repair mechanisms in the hosts (S. cerevisiae vs. E.
coli ). Our mutators still showed bias toward C→T mutations, and we
hypothesized that the introduction of other DNA-modifying enzymes and
their combinations might further broaden the mutation spectra.
Finally, we applied our mutagenesis tools to the evolution of the key
enzymes in the β-carotene biosynthetic pathway. After induction, we
observed apparent color changes compared with the original strain. Most
of the mutations were C and G mutations, which was consistent with the
mutation characteristic of our mutators. These results demonstrated that
our mutagenesis tools could evolve the non-growth-limiting genes and
generate diverse genotypes even without the growth pressure or
selection.
Our mutagenesis tools are inducible, so the mutation rate could be
flexibly tuned by changing the inducer concentration and induction time.
The expression level of our mutagenesis fusion proteins may also
influence the mutagenic activity and needs further investigation. In
addition, using more DNA-modifying enzymes or their combinations may
further alter the mutation effect of the mutagenesis proteins, thus
developing evolutionary tools with diverse mutation characteristics to
meet different needs of the evolutionary process. Our mutagenesis tools
are compatible with continuous evolution, with the help of biosensors or
other screening techniques, researchers can efficiently obtain desired
mutants. In addition, our mutagenesis tools can also work complementary
to the single deaminase-bearing mutators that are biased toward
generating transition mutations, effectively creating both transition
and transversion mutations, and can be applied to many aspects, such as
industrial strain breeding, protein engineering, and so on.