Proximal Biotinylation-Based Combinatory Approach for Isolating Integral Plasma Membrane Proteins

Abstract

Comprehensive profiling of the cell-surface proteome has been challenging due to the lack of tools for an effective and reproducible way to isolate plasma membrane proteins from mammalian cells. Here we employ a proximity-dependent biotinylation approach to label and isolate plasma membrane proteins without an extra in vitro labeling step, which we call Plasma Membrane-BioID. The lipid-modified BirA* enzyme (MyrPalm BirA*) was targeted to the inner leaflet of the plasma membrane, where it effectively biotinylated plasma membrane proteins. Biotinylated proteins were then affinity-purified and analyzed by mass spectrometry. Our analysis demonstrates that combining conventional sucrose density gradient centrifugation and Plasma Membrane-BioID is ideal to overcome the inherent limitations of the identification of integral membrane proteins, and it yields highly pure plasma components for downstream proteomic analysis.

Rapid Enzyme-Mediated Biotinylation for Cell Surface Proteome Profiling

“…As seen in Table S7, in terms of the number of identified likely cell surface proteins, which have a surface prediction consensus (SPC) score larger than 0), PECSL technology (713), lysine-targeting methods (202–734), and sucrose density gradient centrifugation method (816–1631) outperform the glycan chain method (180–436) and carboxyl group method (208). As for the highly likely cell surface proteins (SPC score 3, 4), PECSL technology (11.3%) has a similar specificity compared with lysine-targeting methods (10.1–26.7%) and sucrose density gradient centrifugation method (7.8–13.1%) ,,, and a lower specificity than glycan chain-targeting methods (14.2–69.3%) and carboxyl group-targeting method (49.5%), ,,, but a much higher specificity than the whole cell lysate data (3.4%) with SurfaceGenie. It reminded us that as previous studies reported, ,− many noncell surface proteins were also copurified through strong interactions with highly hydrophobic cell surface proteins or avidin-affinity material.…”

“…As for the highly likely cell surface proteins (SPC score 3, 4), PECSL technology (11.3%) has a similar specificity compared with lysine-targeting methods (10.1–26.7%) and sucrose density gradient centrifugation method (7.8–13.1%) ,,, and a lower specificity than glycan chain-targeting methods (14.2–69.3%) and carboxyl group-targeting method (49.5%), ,,, but a much higher specificity than the whole cell lysate data (3.4%) with SurfaceGenie. It reminded us that as previous studies reported, ,− many noncell surface proteins were also copurified through strong interactions with highly hydrophobic cell surface proteins or avidin-affinity material. To determine which proteins are truly present on the cell surface, an LFQ proteomics experiment was employed to compare the enriched proteins between the control and experimental cell samples.…”

“…To further confirm the performance of PECSL technology, GenieScore, 36 where four previous bioinformatics-based constructions of the human cell surface proteome of 5407 proteins were compiled for a consensus-based prediction of cell surface localization, was utilized to analyze the PECSL data and multiple surfaceome data generated by using different 13,23,37,38 and a lower specificity than glycan chain-targeting methods (14.2−69.3%) and carboxyl group-targeting method (49.5%), 13,37,39,40 but a much higher specificity than the whole cell lysate data (3.4%) with SurfaceGenie. It reminded us that as previous studies reported, 38,41−43 many noncell surface proteins were also copurified through strong interactions with highly hydrophobic cell surface proteins or avidin-affinity material.…”

Mass spectrometry and the cellular surfaceome

“…A major drawback of this approach was the long incubation time (quite often overnight) for the labeling, which is reduced to minutes with TurboID and MiniTurbo. As an orthogonal method, Akdag et al developed a BirA* mediated proximity labeling on the inner leaflet of the membrane, termed membrane-BioID (Akdag et al, 2020). By fusing BirA* to a myristoylation and palmitoylation prone sequence, the enzyme is incorporated in the inner leaflet of the membrane through endogenous Lyn kinase activity where it can label integral proteins in a 100 Å radius.…”

Dynamic extracellular proximal interaction profiling reveals Low-Density Lipoprotein Receptor as a new Epidermal Growth Factor signaling pathway component

“…BioID has been used to identify new components of signaling pathways ( 22 , 23 ), uncover protein composition of membraneless organelles ( 24 ) and nuclear structures ( 21 , 25 ), identify proteins involved in lumen formation in 3D culture models ( 26 ), and map out the protein distribution in human cellular organelles ( 18 ). BioID was also applied to define the plasma membrane proteome by directing BirA* to the inner leaflet of the lipid bilayer ( 16 ) or by tagging the cytoplasmic portion of plasma membrane proteins to identify proteins in the vicinity of the membrane-spanning baits ( 15 , 18 , 27 ). However, these strategies generally overlook interactions occurring on the extracellular side of the plasma membrane, such as those involving the extracellular domain of membrane proteins, extracellular matrix proteins or proteins that are extracellularly tethered to the plasma membrane via a glycosylphosphatidylinositol (GPI) anchor.…”