far form complexes of suboptimal brightness. The brightness of a
fluorogen/aptamer complex results from the number of photons
the fluorogen can absorb (quantified by the absorption coefficient
ε) and from the number of photons re-emitted per absorbed
photons (quantified by the quantum yield, QY). Therefore, to be
as bright as possible, a complex should have both the highest
possible ε value and a QY as close as possible to 1 (i.e., every
absorbed photon is re-emitted). Consequently, even though
DFHBI-1T/Spinach2 displays an excellent QY of 0.98, the low ε
of the complex makes it just as bright as eGFP. On the contrary,
complexes with much higher ε such as TO1-Biotin/Mango, DIR-Pro/DIR2s, and Cbl-Cy5/Riboglow are limited by their low QY
(Table 1). Therefore, to be efficient, the aptamer selection pipeline
should ideally include a screening step during which the capacity of
each molecule to light up the fluorogen is also taken into account.
Following the strategy originally used to identify improved
variants of GFP [36], the aptamer variants contained in a SELEXenriched pool can be further selected for light-up aptamers upon
expression in bacteria grown on agar plates supplemented with
fluorogen and selecting colonies that are fluorescent [26]. However, this approach allows only analyzing a few hundreds of aptamers. Significant gain in throughput can be reached by analyzing
bacteria with a fluorescence-activated cell sorter (FACS, Fig. 3b).
Indeed, combining SELEX pre-enrichment with FACS screening
of aptamer-expressing bacteria allowed to identify Broccoli, a
DFHBI-binding aptamer optimized for imaging RNA in living
cells [35]. While live-cell FACS screening allows to select aptamers
with lighting-up capacity directly in the cellular context, it may also
face significant drawbacks given by the cell-based system such as
(1) limited bacteria transformation efficiency, (2) requirement of
the fluorogen to be cell membrane permeable (dispensable for the
development of extracellular probes), or (3) difficulty to apply
harsh and controlled selection pressures. To overcome these limitations, aptamer libraries should ideally be screened in vitro. Such
in vitro screening can be performed using gene-linked RNA aptamer particle (GRAP) display [37], a technology in which aptamercoding genes contained in a library are first individualized at the
surface of beads onto which they are then clonally amplified by
emulsion PCR. Upon washing, beads are emulsified in a second set
of droplets into which DNA is transcribed into RNA aptamers that
are then captured on the bead surface. Finally, incubating beads
with a fluorogen allows to fluorescently label the beads all the better
the aptamer is efficient and makes it possible to FACS-sort them
(Fig. 3b). Advantageously, this method enables to select for both
brightness and affinity and was successfully used to isolate new
MG-binding aptamers forming higher affinity or brighter complexes with the fluorogen. Such a screening would have been
Light-Up RNA Aptamers
81
fluorogen/aptamer complex results from the number of photons
the fluorogen can absorb (quantified by the absorption coefficient
ε) and from the number of photons re-emitted per absorbed
photons (quantified by the quantum yield, QY). Therefore, to be
as bright as possible, a complex should have both the highest
possible ε value and a QY as close as possible to 1 (i.e., every
absorbed photon is re-emitted). Consequently, even though
DFHBI-1T/Spinach2 displays an excellent QY of 0.98, the low ε
of the complex makes it just as bright as eGFP. On the contrary,
complexes with much higher ε such as TO1-Biotin/Mango, DIR-Pro/DIR2s, and Cbl-Cy5/Riboglow are limited by their low QY
(Table 1). Therefore, to be efficient, the aptamer selection pipeline
should ideally include a screening step during which the capacity of
each molecule to light up the fluorogen is also taken into account.
Following the strategy originally used to identify improved
variants of GFP [36], the aptamer variants contained in a SELEXenriched pool can be further selected for light-up aptamers upon
expression in bacteria grown on agar plates supplemented with
fluorogen and selecting colonies that are fluorescent [26]. However, this approach allows only analyzing a few hundreds of aptamers. Significant gain in throughput can be reached by analyzing
bacteria with a fluorescence-activated cell sorter (FACS, Fig. 3b).
Indeed, combining SELEX pre-enrichment with FACS screening
of aptamer-expressing bacteria allowed to identify Broccoli, a
DFHBI-binding aptamer optimized for imaging RNA in living
cells [35]. While live-cell FACS screening allows to select aptamers
with lighting-up capacity directly in the cellular context, it may also
face significant drawbacks given by the cell-based system such as
(1) limited bacteria transformation efficiency, (2) requirement of
the fluorogen to be cell membrane permeable (dispensable for the
development of extracellular probes), or (3) difficulty to apply
harsh and controlled selection pressures. To overcome these limitations, aptamer libraries should ideally be screened in vitro. Such
in vitro screening can be performed using gene-linked RNA aptamer particle (GRAP) display [37], a technology in which aptamercoding genes contained in a library are first individualized at the
surface of beads onto which they are then clonally amplified by
emulsion PCR. Upon washing, beads are emulsified in a second set
of droplets into which DNA is transcribed into RNA aptamers that
are then captured on the bead surface. Finally, incubating beads
with a fluorogen allows to fluorescently label the beads all the better
the aptamer is efficient and makes it possible to FACS-sort them
(Fig. 3b). Advantageously, this method enables to select for both
brightness and affinity and was successfully used to isolate new
MG-binding aptamers forming higher affinity or brighter complexes with the fluorogen. Such a screening would have been
Light-Up RNA Aptamers
81