PUB-MS - a mass-spectrometry-based method to monitor protein-protein proximity in vivo

The common techniques to study protein-protein proximity in vivo are not well-adapted to the capabilities and the expertise of a standard proteomics laboratory, typically based on the use of mass spectrometry. With the aim of closing this gap, we have developed PUB-MS (for Proximity Utilizing Biotinylation and Mass Spectrometry), an approach to monitor protein-protein proximity, based on biotinylation of a protein fused to a biotin-acceptor peptide (BAP) by a biotin-ligase, BirA, fused to its interaction partner. The biotinylation status of the BAP can be further detected by either Western analysis or mass spectrometry. The BAP sequence was redesigned for easy monitoring of the biotinylation status by LC-MS/MS. In several experimental models, we demonstrate that the biotinylation in vivo is specifically enhanced when the BAP- and BirA- fused proteins are in proximity to each other. The advantage of mass spectrometry is demonstrated by using BAPs with different sequences in a single experiment (allowing multiplex analysis) and by the use of stable isotopes. Finally, we show that our methodology can be also used to study a specific subfraction of a protein of interest that was in proximity with another protein at a predefined time before the analysis.

💡 Research Summary

The paper introduces PUB‑MS (Proximity Utilizing Biotinylation and Mass Spectrometry), a novel method that translates protein‑protein proximity into a covalent biotin tag detectable by standard proteomics workflows. Two proteins of interest are expressed as fusions: one with the bacterial biotin ligase BirA, the other with a short Biotin Acceptor Peptide (BAP). When the two proteins come into close proximity inside a living cell, BirA catalyzes the transfer of biotin onto the lysine within the BAP. The extent of biotinylation serves as a quantitative proxy for proximity.

Key technical innovations include redesigning the BAP sequence to contain flanking arginine residues, producing a tryptic peptide of optimal size for LC‑MS/MS, and adding a 7×His tag for independent protein quantification. Two expression vectors were built: a strong CMV‑driven BAP‑fusion vector and a weaker MoMuLV‑driven BirA‑fusion vector, ensuring an excess of the BAP substrate over the enzyme, which is crucial for quantitative interpretation.

The authors first validated the system by co‑expressing BirA‑GFP and BAP‑GFP in HEK293T cells. Biotinylation increased linearly with the duration of biotin exposure, confirming that the engineered BAP is efficiently labeled yet slower than the original BAP, providing a broader dynamic range for detecting proximity‑dependent labeling.

To demonstrate biological relevance, several model systems were examined: (1) oligomerization of TAP54α and HP1γ, (2) DNA‑damage‑induced interaction between Rad18 and histone variant H2AZ, and (3) multiplexed analysis using BAP variants of different primary sequences within a single LC‑MS/MS run. In each case, biotinylation of the BAP‑fusion was markedly higher when the partner proteins were known to interact or share the same subcellular compartment, compared with appropriate negative controls.

Multiplexing was further extended by incorporating SILAC labeling. Cells grown in light (¹²C₆‑Lys) or heavy (¹³C₆‑Lys) media expressed distinct BAP‑fusions; mass‑spectrometric quantification of the biotinylated peptides allowed direct comparison of proximity levels across conditions, demonstrating the method’s capacity for accurate, high‑throughput, quantitative interaction profiling.

A major advantage of PUB‑MS is its ability to capture a temporal “snapshot” of proximity. By performing a short biotin pulse followed by a chase, the covalent biotin mark remains on the BAP‑fusion even after the interacting partner dissociates or the cellular context changes. This irreversible labeling enables downstream analysis of proteins that were in proximity at a defined time point, a capability not readily achievable with reversible fluorescence‑based methods such as FRET, BRET, or protein complementation assays.

Potential limitations are acknowledged. Overexpression of BirA can lead to background biotinylation due to random collisions, necessitating careful titration of BirA levels and optimization of labeling times. Moreover, because biotinylation can occur for proteins residing in the same compartment without direct contact, rigorous negative controls and spatially resolved validation are essential for accurate interpretation.

In summary, PUB‑MS bridges a methodological gap by providing a mass‑spectrometry‑centric, multiplexable, and temporally resolved approach to study protein proximity in vivo. It leverages the high resolution and quantitative power of modern LC‑MS/MS, expands the repertoire of proteomics tools for interaction mapping, and opens new avenues for dissecting dynamic cellular processes that involve transient or compartment‑based protein associations.