Biology Reference
In-Depth Information
FROM mRNA TO PROTEIN: TUNING TRANSLATION OF PATHWAY PROTEINS
The overall process of how an mRNA transcript is translated into protein is well-understood.
However, making accurate predictions of how much protein will be generated from a given
transcript are very difficult. Because fluxes depend strongly on pathway enzyme abundances,
especially for rate-limiting enzymes, modulating protein translation from mRNA can help to
fine-tune enzyme expression. Translation is also indirectly controlled by mRNA transcript
stability. We will focus on efforts to engineer two of the phases of translation (initiation and
elongation) in the context of biosynthesis pathways.
It is thought that the rate-limiting step in protein production from a transcript is
transcription initiation, a complex process largely determined by binding of the ribosome
RNA to the ribosome binding site (RBS), a sequence upstream of the initiation codon. 58
This binding step is controlled by several factors: the presence of secondary structures within
the mRNA that occludes the RBS and reduces the availability of RBS to the ribosome; and
the equilibrium binding strength of the RBS to the ribosomal RNA. Because these factors are
highly dependent on the mRNA sequence upstream and downstream of the RBS, a single
RBS will likely result in widely varying translation initiation rates for different surrounding
sequences. This may complicate efforts to precisely tune protein levels across a biofuel
production pathway.
An elegant method for predicting and controlling RBS-determined translation initiation in
E. coli has been developed. 59 It estimates protein expression levels using a thermodynamic
model for translation initiation. The model uses an mRNA folding algorithm to calculate
the free energies involved in disrupting mRNA secondary structure occluding the RBS, as
well as the binding energy of the RBS with ribosomal RNA. Using the RBS calculator, the
authors were able to predict protein levels that result from a variety of different RBS
sequences reasonably well. Furthermore, their model enables forward engineering of desired
expression levels by generating RBS sequences with predicted expression levels for specific
proteins. The authors confirmed that RBS sequences that result in high expression activity
for a given CDS do not universally apply to other proteins. Using a simple genetic switch,
the authors demonstrate the utility of their model using RBS sequences generated from their
calculator.
216
The rate of protein elongation can also be engineered. The most common approach to
increasing protein yield from a transcript is to optimize the codon usage within each gene
for its new host by replacing codons that occur rarely in the host coding sequences with
more commonly used codons. This is thought to speed translation by using codons that
correspond to tRNA species that are more abundant in the host organism. Codon
optimization of pathway genes has been shown to lead directly to increased protein
expression, which in turn results in higher yields. Redding-Johanson and coauthors used
transcriptomics and proteomics to understand the rate-limiting steps of an amorphadiene
production pathway in E. coli . 60 They found that two proteins in the pathway, mevalonate
kinase (MK) and phosphomevalonate kinase (PMK), were expressed at very low levels. By
codon-optimizing the genes encoding MK and PMK, the authors were able to increase their
protein abundances by about two-fold each, resulting in a two-fold increase in titer.
However, the presence of the codon-optimized MK and PMK sequences resulted in a
dramatic decrease in the levels of the other pathway enzymes encoded within the operon.
The addition of a transcriptional terminator and a second promoter sequence restored the
levels of the proteins, and led to a 2.5-fold further increase in amorphodiene production, a
total production increase of five-fold over the original operon design.
The previous study also demonstrates the dangers of treating individual genes transcribed
within a common operon as noninteracting entities. Because of poorly-understood and
context-dependent phenomena such as folding and degradation of the mRNA transcripts,
Search WWH ::




Custom Search