extremely challenging to perform with living cells given the significant toxicity of the fluorogen [38]. Recently, the methodology was
further improved by making the second emulsification step (in vitro
transcription step) dispensable. Indeed, in the so-called R-CAMPS
technology the RNA produced from DNA coupled on the bead
surface is co-transcriptionally captured at the surface of the same
bead via a specific sequence [39]. The functionality of this approach
was demonstrated by screening small libraries of Baby Spinach
aptamer. However, the significant polydispersity of the emulsions
generated during PCR and in vitro transcription steps may affect
the overall efficiency of the bead-based methods. Thus, a significant
gain in accuracy could be reached by performing the amplification
and transcription reactions in more homogenous emulsions.
An attractive way of producing highly identical water-in-oil
droplets is to use microfluidics as it allows to generate and manipulate highly monodisperse emulsions and to gain control over selection conditions. In this view, we recently introduced microfluidicassisted in vitro compartmentalization (μIVC) as an alternative to
the in vitro ultrahigh-throughput screening approach [40]. In
μIVC, genes contained in a library are individualized into small
water-in-oil droplets (Fig. 3c) in which they are PCR amplified,
expressed into RNA (or even protein), and analyzed for their
properties (e.g., enzyme activity, capacity to light up a fluorogen).
The droplets with the genes of interest are then sorted at rates of
several millions per day. Such ultrahigh-throughput is primarily
possible due to the extreme miniaturization of the reaction vessels
(down to a few picoliters) and the use of microfluidic devices that
allow to produce highly monodisperse emulsions and to manipulate
(fusion, injection, and sorting) individual droplets with electricity
(avoiding the use of moving parts that would limit the throughput)
[41]. We first demonstrated μIVC efficiency in light-up aptamer
development by improving the folding and fluorescence properties
of the light-up aptamer Spinach, which led to the isolation of
iSpinach, an aptamer optimized for in vitro applications [42]. Yet,
so far, the best illustration of μIVC efficiency was obtained with the
isolation of new Mango aptamers. Indeed, the original Mango
aptamer was found to form a high-affinity complex with the fluorogen TO1-Biotin, but the complex also suffered from a low QY of
0.14, making it dimmer than half an eGFP (Table 1). Rescreening
the SELEX-enriched library from which Mango was identified
using a μIVC procedure enabled us to identify three new Mango
variants (Mango-II, -III, and -IV) [43]. Of these news aptamers,
Mango-II displayed an improved affinity for TO1-Biotin while
having unchanged lighting-up capacity, whereas Mango-III (the
major sequence found at the end of process) had preserved affinity
for the fluorogen while forming a complex with a nearly fourfold
higher QY, making it significantly brighter than eGFP (Table 1).
82
Michael Ryckelynck
further improved by making the second emulsification step (in vitro
transcription step) dispensable. Indeed, in the so-called R-CAMPS
technology the RNA produced from DNA coupled on the bead
surface is co-transcriptionally captured at the surface of the same
bead via a specific sequence [39]. The functionality of this approach
was demonstrated by screening small libraries of Baby Spinach
aptamer. However, the significant polydispersity of the emulsions
generated during PCR and in vitro transcription steps may affect
the overall efficiency of the bead-based methods. Thus, a significant
gain in accuracy could be reached by performing the amplification
and transcription reactions in more homogenous emulsions.
An attractive way of producing highly identical water-in-oil
droplets is to use microfluidics as it allows to generate and manipulate highly monodisperse emulsions and to gain control over selection conditions. In this view, we recently introduced microfluidicassisted in vitro compartmentalization (μIVC) as an alternative to
the in vitro ultrahigh-throughput screening approach [40]. In
μIVC, genes contained in a library are individualized into small
water-in-oil droplets (Fig. 3c) in which they are PCR amplified,
expressed into RNA (or even protein), and analyzed for their
properties (e.g., enzyme activity, capacity to light up a fluorogen).
The droplets with the genes of interest are then sorted at rates of
several millions per day. Such ultrahigh-throughput is primarily
possible due to the extreme miniaturization of the reaction vessels
(down to a few picoliters) and the use of microfluidic devices that
allow to produce highly monodisperse emulsions and to manipulate
(fusion, injection, and sorting) individual droplets with electricity
(avoiding the use of moving parts that would limit the throughput)
[41]. We first demonstrated μIVC efficiency in light-up aptamer
development by improving the folding and fluorescence properties
of the light-up aptamer Spinach, which led to the isolation of
iSpinach, an aptamer optimized for in vitro applications [42]. Yet,
so far, the best illustration of μIVC efficiency was obtained with the
isolation of new Mango aptamers. Indeed, the original Mango
aptamer was found to form a high-affinity complex with the fluorogen TO1-Biotin, but the complex also suffered from a low QY of
0.14, making it dimmer than half an eGFP (Table 1). Rescreening
the SELEX-enriched library from which Mango was identified
using a μIVC procedure enabled us to identify three new Mango
variants (Mango-II, -III, and -IV) [43]. Of these news aptamers,
Mango-II displayed an improved affinity for TO1-Biotin while
having unchanged lighting-up capacity, whereas Mango-III (the
major sequence found at the end of process) had preserved affinity
for the fluorogen while forming a complex with a nearly fourfold
higher QY, making it significantly brighter than eGFP (Table 1).
82
Michael Ryckelynck