10.3 Managing Transcript Stability 201Improving the ability to integrate biochemical models and refined RNA
transcript folding design algorithms should therefore lead to better tools for
engineering genetic control systems that employ RNA sequence and structure
design to quantitatively program expression.
10.3 Managing Transcript Stability
10.3.1 Transcript Stability as a Confounding Factor
Perhaps the greatest obstacle on the road to predictable biological engineering is
the joint confounding effect of cellular subsystems that interact with synthetic
biological components in unanticipated ways. In this regard, it is important to
realize that any – and all – synthetic RNA in the cell is affected by the degradation
systems regulating transcript stability. Except in cases where transcript stability
is the explicit genetic control mechanism, efforts aimed at engineering gene
expression tend to neglect dimensions of transcript stability. However, as new
tools are developed, it should become much easier to circumvent limitations
imposed by variations in RNA stability and instead fine-tune transcript stabil-
ity in concert with other engineering strategies to rapidly implement genetic
controls to meet performance requirements.
10.3.2 Anticipating Transcript Stability Issues
Because so many factors can affect RNA stability, it is important to consider the
ways that experimental results may be impacted by unexpected changes in tran-
script stability. Moreover, it may be prudent to routinely determine whether
transcript stability is a parameter requiring attention, either through computa-
tional design variable sensitivity analysis or through wet-lab experimentation.
The roles and binding behaviors of all ribonucleases and associated proteins have
yet to be elucidated [97], but study of major players such as RNase E, PNPase,
RppH, Hfq, and RNase III has unveiled structures and many key roles of these
enzymes. Though the binding interactions of these enzymes with RNA and each
other are not completely understood, it is possible to analyze sequences and
attempt to avoid unwanted degradation, or increase degradation, by changing
codons within the open reading frame or UTR sequences to eliminate, or insert,
putative binding sites.
Computationally, the potential impact of transcript stability on a given syn-
thetic biological device output can be assessed with GSA using coarse-grained
mechanistic model simulations and Monte Carlo sampling [98]. With this
method, the global space of potential designs is mapped by simulating genetic
device outputs with Monte Carlo sampled values for the model parameters taken
randomly from biochemically reasonable ranges. By computing quantitative
GSA measures to relate the potential genetic device outputs to transcript stabil-
ity parameter inputs (e.g., partial correlation coefficients) [98, 99], the impact of
variations in RNA degradation rate, relative to other tunable design variables,
can be readily discerned [7].