fed to the cells, whereas the latter uses an abortive variant of a
protein biotin ligase (BPL) that activates native biotin. In both
the cases, the activated substrate diffuses around the enzyme and
leads to the biotinylation of proximal proteins in an estimated
range of approximately 10 nm for BioID [4]. When fused to a
POI, both enzymes thus mediate the biotinylation of nearby
factors including interacting proteins. These biotinylated proteins
can then be efficiently isolated by streptavidin pulldown and
identified by MS. As opposed to AP-MS, APEX2- and BioIDMS do not aim at purifying assembled protein complexes by
pulling one of its components but rather identify proteins that
were marked within cells because they were in close proximity to
the POI. In proximity-dependent labeling approaches, it thus
does not matter whether interacting proteins are still associated
during the pulldown procedure, making these techniques more
powerful than AP-MS for detecting transient interactions [5] or
PPI that depend on intact or poorly soluble cellular structures
[6]. Yet, a limitation to both approaches is that they cannot easily
resolve the remodeling and/or maturation of protein complexes.
In the common situation in which a POI is part of several distinct
complexes depending on the cellular context, both AP- and
APEX2/BioID-MS will identify all possible PPI but will not
assign individual interactions to specific context-dependent complexes. We and others have recently introduced split-BioID assays
[7, 8] in which BirA*, the BPL of BioID, is split into two inactive
and poorly interacting fragments that can reassemble into an
active enzyme when fused with two interacting proteins. Provided
a POI is known to interact with a given factor within a specific
protein complex, applying split-BioID to these two proteins
allows the labeling of additional components of that specific
assembly, ignoring any additional PPI the POI may have within
other protein complexes (Fig. 1). The general strategy of a splitBioID experiment is outlined in Fig. 2. Briefly, the coding
sequences for two putative interacting POI are cloned in frame
with the two BirA* fragments, termed NBirA* and CBirA*, and
transfected in mammalian cells. Biotinylation of proximal proteins
is induced by the addition of excess biotin in the growth medium.
Biotinylated proteins are then isolated by streptavidin pulldown
and can be analyzed either by immunoblotting or MS. Of note
the protocol described here is virtually identical to a classical
BioID procedure, the only difference being that in BioID the
POI is fused to the full BirA* enzyme.
304
Cinthia Amaya Ramirez et al.
protein biotin ligase (BPL) that activates native biotin. In both
the cases, the activated substrate diffuses around the enzyme and
leads to the biotinylation of proximal proteins in an estimated
range of approximately 10 nm for BioID [4]. When fused to a
POI, both enzymes thus mediate the biotinylation of nearby
factors including interacting proteins. These biotinylated proteins
can then be efficiently isolated by streptavidin pulldown and
identified by MS. As opposed to AP-MS, APEX2- and BioIDMS do not aim at purifying assembled protein complexes by
pulling one of its components but rather identify proteins that
were marked within cells because they were in close proximity to
the POI. In proximity-dependent labeling approaches, it thus
does not matter whether interacting proteins are still associated
during the pulldown procedure, making these techniques more
powerful than AP-MS for detecting transient interactions [5] or
PPI that depend on intact or poorly soluble cellular structures
[6]. Yet, a limitation to both approaches is that they cannot easily
resolve the remodeling and/or maturation of protein complexes.
In the common situation in which a POI is part of several distinct
complexes depending on the cellular context, both AP- and
APEX2/BioID-MS will identify all possible PPI but will not
assign individual interactions to specific context-dependent complexes. We and others have recently introduced split-BioID assays
[7, 8] in which BirA*, the BPL of BioID, is split into two inactive
and poorly interacting fragments that can reassemble into an
active enzyme when fused with two interacting proteins. Provided
a POI is known to interact with a given factor within a specific
protein complex, applying split-BioID to these two proteins
allows the labeling of additional components of that specific
assembly, ignoring any additional PPI the POI may have within
other protein complexes (Fig. 1). The general strategy of a splitBioID experiment is outlined in Fig. 2. Briefly, the coding
sequences for two putative interacting POI are cloned in frame
with the two BirA* fragments, termed NBirA* and CBirA*, and
transfected in mammalian cells. Biotinylation of proximal proteins
is induced by the addition of excess biotin in the growth medium.
Biotinylated proteins are then isolated by streptavidin pulldown
and can be analyzed either by immunoblotting or MS. Of note
the protocol described here is virtually identical to a classical
BioID procedure, the only difference being that in BioID the
POI is fused to the full BirA* enzyme.
304
Cinthia Amaya Ramirez et al.