(502a) Biomolecular Engineering of Acoustomagnetic Protein Nanostructures for Non-Invasive Imaging of Cellular Function

Genetically encoded optical reporters such as green fluorescent protein
(GFP) have revolutionized biomedical research by enabling observations of
biological processes in engineered cells and transgenic animals. However, such
optical reporters are fundamentally limited by the ~ 1 mm penetration depth of
light in opaque tissues. As cellular therapies advance towards rodent models
and ultimately humans, this limitation becomes increasingly severe. Therefore,
we aim to use biomolecular engineering to develop new classes of genetically
encoded agents capable of communicating with deeply penetrant forms of energy,
such as magnetic field and sound waves.

Here we introduce gas
vesicles (GVs), a class of gas-filled protein nanostructures discovered in
certain photosynthetic microbes as their flotation devices. GVs possess a
hollow gas interior, ~ 200 nm in size, enclosed by a 2 nm protein shell that is
permeable to gas but excludes liquid water (Fig. a-b). We hypothesized that the
gaseous interior of GVs would have a different magnetic susceptibility from the
surrounding aqueous tissue and could thus produce susceptibility-based MRI
contrast (Fig. c). Indeed, we
observed robust T2, T2*, and quantitative susceptibility contrast of GVs at
nanomolar concentrations (Fig. d). A
unique feature of GVs is that they can be irreversibly collapsed under acoustic
pressure above genetically determined thresholds, leading to the rapid
dissolution of their gaseous contents. We demonstrated in several scenarios
including brain tissues of living mice (Fig.
e) that the MRI contrast produced by GVs can be erased remotely in situ by ultrasound, an orthogonal
non-invasive imaging modality. Such acoustically modulated reporters would overcome a major challenge in MRI posed by
background contrast from endogenous sources by allowing reporters to be
identified specifically based on their acoustic responses. Next, we
tested whether GVs can be used as MRI reporter genes. As a proof-of-concept,
the expression of GVs was placed under the control of the chemical IPTG. When E. coli cells expressing this circuit were exposed to IPTG, they showed MRI
contrast, which was absent from controls (Fig.
f). Furthermore,
multiplexed imaging to differentially visualize several species of GVs could be
achieved either by scanning ultrasound pressures from low to high to
selectively collapse certain GVs (Fig. g)
or from distinct MRI fingerprints of GVs that differ in size and shape (Fig. h). Finally, we ask whether GVs
can be engineered as dynamic sensors of biological signals. We found that
clustering of GVs induced potentially by a biological signal could produce a
marked change in T2*-weighted images, as experimentally shown by streptavidin-biotin
cross-linking (Fig. i).
These in vitro experiments
established a set of principles allowing us to use GVs as acoustically
modulated MRI reporters in vivo (Lu
et al, Nature Materials, 2018).

By coupling the
complementary physics of MRI and ultrasound, this genetically encoded and
engineerable nanostructure gives rise to a distinct modality for molecular
imaging with unique advantages and capabilities, especially in basic research
of biological processes in deep-lying tissues and the development and tracking
of therapeutic microbes.

References: Lu
GJ, Farhadi A, Szablowski JO, Lee-Gosselin A, Barnes
SR, Lakshmanan A, Bourdeau RW, Shapiro MG (2018)
Acoustically modulated magnetic resonance imaging of gas-filled protein
nanostructures. Nature Materials. http://doi.org/10.1038/s41563-018-0023-7