Discussion
BioID has been used to profile the proteome of different cellular
compartments and molecular complex systems
(Varnaitė & MacNeill,
2016). However, this is the first time that BioID has been used to
identify proteome of the secretory pathway during the recombinant
protein expression. Numerous PPIs guiding the folding, modifications,
and trafficking of the secreted and membrane proteins through the
secretory pathway are transient
(Nyfeler, Michnick, &
Hauri, 2005; Schreiber, Haran, & Zhou, 2009) and therefore cannot be
captured by conventional methods such as co-immunoprecipitation.
Consistent with previous studies
(Sears, May, & Roux, 2019),
these results showed the BioID can detect weak and transient
interactions in situ, and therefore it is a powerful approach to
study luminal processes involved in protein secretion. We found that
disulfide bridge formation enzymes showed the most significant
association with bait proteins enriched in disulfide bonds, supporting
their critical roles in protein secretion and maintaining ER stability.
A previous study on difficult to express (DTE) monoclonal antibodies
showed less recognition by PDI impairs disulfide bridge formation within
the antibody light chain (LC) which can initiates the intracellular
degradation by the ubiquitin proteasome system via ERAD
(Mathias et al., 2020).
Thus, insufficient interaction between the secreted proteins with
enriched disulfide bond and PDIs can limit secretion efficiency and
serve as a rate-limiting step for protein production. In another study,
the tissue specific analysis of SecMs expression showed a positive
correlation between the expression of P4HB and PDIA4 and liver tissue
where numerous disulfide bond enriched proteins are secreted by
hepatocytes (Feizi et al.,
2017). These observations are clear evidence that suggests the
tissue-specific fine-tuning of the PDI family expression level in
response to the enrichment of the disulfide sites. Together, these
results showed PDIs are actively involved in adaptive responses and
secretion of proteins with dominant disulfide bonds which are crucial
for restoring ER stability, and therefore yielding the recombinant
proteins. Given the associations between the SecMs and the features of
the model proteins, we also hypothesized that SecMs preferentially
interacting with bait-BirA proteins that carry certain structural
features may be essential for the secretion of those proteins. While
evidence linking SecMs to the structural motifs is lacking, many
molecular chaperones selectively interact with certain sequence and
structural elements to favor the particular folding pathways
(Gidalevitz, Stevens, &
Argon, 2013). For example, chaperones of the HSP70 family evolved to
bind extended β strand peptides; interestingly, the associations
identified between chaperones and asx motif and ST turn represent a
novel association for further study.
While we show BioID works well for studying the synthesis of secreted
proteins, we acknowledge that biotin-based methods have some limitations
as well. Biotin is actively imported into the cytoplasm of cells and can
freely diffuse to the nucleus, but it may not be as accessible in the
secretory pathway, thus reducing labeling efficacy in that compartment
(Kim & Roux, 2016). Here we
showed this challenge is not an insurmountable issue, in that the BioID2
construct successfully identified many expected luminal interactions.
BioID2 requires less biotin supplementation, and exhibits enhanced
labeling of proximate protein
(Kim et al., 2016) allowing
for BioID to be introduced to new systems where biotinylation
supplementation may not be easily accomplished
(Sears et al., 2019). More
recently, two promiscuous mutants of biotin ligase, TurboID and
miniTurbo, have been developed to catalyze proximity labeling even with
much greater efficiency
(Branon et al., 2020) and
therefore can be considered as an effective method when proximal
labeling of the endomembrane organelles is desired.
In summary, we demonstrate here an approach to identify the protein
interactions that synthesize and support secreted proteins, and thus
define the product-specific secretory pathway. The identification of
such machinery opens avenues for mammalian synthetic biology, wherein
biotherapeutic production hosts can be rationally engineered to improve
the titer and quality of diverse proteins in a client specific manner.