Ultrasensitive ultrasound imaging of gene expression with signal unmixing
Acoustic reporter genes (ARGs) that encode
air-filled gas vesicles enable ultrasound-based imaging of gene expression in
genetically modified bacteria and mammalian cells, facilitating the study of
cellular function in deep tissues. Despite the promise of this technology for
biological research and potential clinical applications, the sensitivity with
which ARG-expressing cells can be visualized is currently limited. They present
BURST, a method that improves the cellular detection limit by more than 1,000-fold
compared to conventional methods. BURST takes advantage of the unique temporal
signal pattern produced by gas vesicles as they collapse under acoustic
pressure above a threshold defined by the ARG. BURST can detect ultrasound
signals from individual bacteria and mammalian cells, enabling quantitative
single-cell imaging.
https://www.nature.com/articles/s41592-021-01229-w
Dynamic modulation of enzyme
activity by synthetic CRISPR-Cas6 endonucleases
RNA-scaffolds
can increase flux through a given metabolic pathway by keeping the pathway’s
enzymes in close proximity to increase the local concentration of pathway
intermediates. Here, RNA scaffolds are built using the crRNA-Cas6 system:
enzymes of interest are fused to Cas6, which in turn binds a specific loop of
RNA. By engineering complementarity into the RNA sequence upstream of the
Cas6-bound loop, the RNA scaffold assembles by RNA:RNA hybridisation. The
authors can then use other input RNA strands to trigger the assembly and
disassembly of the scaffold through TSMD reactions. With this method, the
authors demonstrate controllable scaffold assembly, break-down and cycles of
assembly/disassembly in vivo.
A
domain-level DNA strand displacement reaction enumerator allowing arbitrary
non-pseudoknotted secondary structures
Domain-level
DNA strand displacement crn simulators. I've read up on visualDSD and
peppercorn enumerator. The aim of these computational models is to 1) enumerate
WC bonded domain level complexes which could from when designing DNA strand
displacement circuits, 2) to use approximate the kinetics of the systems and 3)
simulate them so that mechanisms can be inspected and designs can be updated to
yield desired results, perhaps by inspection or by algorithmic approaches.
Classic visual DSD (https://ph1ll1ps.github.io/visualdsd/index.html)
could describe only a very limited set of toehold mediated strand displacement
and other binding mechanisms. More recently visualDSD has been converted to use
"logic dsd" semantics (https://pubs.acs.org/doi/pdf/10.1021/acssynbio.8b00229)
which can describe a very wide range of customizable reactions, including DNA
strand displacements and enzymatic reactions such as ligation and cleaving. It
is unclear where to find and how to use the logicDSD version, as the link
provided in the paper no longer works. Classic visualDSD is incapable of
simulating the remote toehold used in handhold mediated strand displacement.
Peppercorn
enumeration does biomolecular binding interactions and a multitude of
intramolecular reactions for non pseudoknotted secondary structured DNA
complexes (where kinetics and thermodynamics are well characterised).
Approximate
rates for each type of reaction based upon domain lengths can be generated
within the software. Once the complexes have been enumerated and the rates of
each reaction have been estimated, a CRN can be constructed. The combinatorics
of enumerating all possible domain-level complexes can be challenging as the
number of complexes may explode with increasing number of domains. Therefore
the software has built-in coarse-graining methods based upon timescale
separation. Simulation can be separated into fast and slow reactions.
Unimolecular reactions may be fast, slow or negligible and bimolecular
reactions are slow (which is valid for low concentrations < 10nM, uni-rates
go as conc^1 and bi-rates go as conc^2). A "condensed" CRN which
features fewer species can be constructed on the basis of timescale separation
that has the same slow dynamics as the original CRN but can be simulated more
easily. Care must be taken in cases in which sequential bimolecular reactions
are required, as if all unimolecular reactions are assumed to be fast compared
to biomolecular reactions, the subsequent bimolecular reaction may never occur
in the condensed CRN. No net production or degradation reactions allowed in
peppercorn, and therefore all strands are conserved. Toeholds less than 7 nt by
default are reversible. Branch migrations are irreversible. Zero-toehold branch
migrations aren't included. peppercorn would be capable of simulating hmsd, but
this would require the implementation of custom reactions and custom rates.
https://doi.org/10.1098/RSIF.2019.0866
Molecular filters for noise
reduction
In classical
signal processing a filter takes an input signal and produces an output signal
with reduced noise. This paper investigates three classes of bimolecular
chemical reaction networks (CRNs) that act as filters by producing a new output
chemical that tracks the input chemical but with reduced noise. One critical
difference between classical filters and CRN filters is that CRNs have
intrinsic noise associated with the stochastic firing of reactions as well as
noise in the input signal, where as classical filters only have noise in the
input signal.
The first
class of CRNs that are analysed are the linear filters, the paper shows through
classical frequency domain analysis that linear filters have the same transfer
function as the low-pass filter. Low-pass filters attenuate high-frequency
oscillations while preserving the low-frequency ones. They show that the output
signal of linear filters are limited by the Poisson’s level, a lower bound on
the variance of your output signal that is at minimum the mean of the output
signal.
They then
go onto to investigate the annihilation module which includes complex formation
reactions. They show that this can reduce the noise of the output to below the
Poisson’s level. They then introduce the annihilation filter which is similar
to the annihilation module except that the mean of the output signal is
proportional to the mean of the input signal.
Finally
they suggest that the translation and transcription of mRNA to produce proteins
can be seen as two cascading linear filters and that certain microRNA pathways
might be annihilation filters.
