The idea of using a soluble APEX2 enzyme to introduce an enrichment handle onto cell surface proteins¹ emerged after a brief brainstorming session while one of us was trying to explore alternatives to Cell Surface Capture² (CSC, pioneered by the Wollscheid lab) where the protein glycosylations, rather than proteins themselves, are labeled, and the other was using APEX2 for proximity labeling³ and interactome studies (after attending a seminar by Prof. Alice Ting).

Our biochemistry-oriented minds immediately tallied a list of attractive features of this approach over other surface protein mapping tools: First, the fast kinetics compared to the much slower NHS-biotin (the classical surface biotinylation method) and even slower CSC method could open avenues to capture plasma membrane remodeling events rather than just to catalog their proteomes.  Second, we like APEX2 labeling sites (at least Tyr, Trp, His, Met, Cys as per various reports in the literature, but check our publication, we settled the issue!), that do not clash with Lys (unlike NHS-biotin), a site for trypsin digestion, as well as TMT-labeling and potentially other useful chemistries.  Third, as we both are trained glycobiologists, we are well aware that data from the CSC labeling requires cautionary interpretation since the biotin is added to the glycan portion (whose abundances can change in different cellular states) of the proteins, and such concern becomes a moot point in APEX2-mediated biotinylation.  Thus, we decided to invest a few of the following weekends collaborating on a side project building a labeling tool.

After we subcloned APEX2 onto a bacterial expression plasmid, overexpressed and purified it, the next question in our minds was, can we immediately assay its activity?  Scrounging both of our labs’ chemical shelves, we found an old bottle of catechol that allowed us to move forward.  Results from the first experiment: very low APEX2 activity!  Further reading of literature indicated that catechol is transformed by the peroxidase activity into very (amine-) reactive ortho-quinone that may immediately inactivate our enzyme.  We then modified the reaction buffer to include 1 M glycine as a sink for this product and repeated the activity assay.  Result: a beautiful slowly plateauing absorbance curve! We were very lucky as things were going super smoothly from the very beginning!  After scaling up the purification, we systematically went through a panel of optimization in a short time, obtaining results from several streptavidin blots per day.

As the project grew legs, things started to get a bit challenging.  Our first set of proteomic data using U2OS cells indicated a large number of soluble intracellular proteins labeled by APEX2.  After several control experiments to rule out labeling-induced artifacts, several more repeats of the proteomic preparations, and hours of PubMed scrolling and bouncing ideas back and forth, we were convinced that what we saw was not an artifact, but rather a manifestation of normal cellular workings that have never been captured before.  We were then ready to test the method’s main selling point, its capability to label cells quickly enough to capture differentially exposed proteins after some cellular signaling event.

We had a conversation with Prof. Art Weiss and his post-doc Wen Lu, who suggested we look into plasma membrane protein remodeling of cytotoxic T cells, a long-time interest of Art’s.  After analyzing data collected from freshly isolated and FACS-sorted cytotoxic T cells kindly prepared by Wen, we found that as low as 10% of non-viable cells from freshly isolated cells would throw off the data due to cytosolic materials being labeled through their leaky plasma membrane (that said, our labeling approach should work for primary cells too, after nurturing them to health and high viability for a while, something that we didn’t have patience for at the time).  After evaluating the investment and rewards, we decided to switch course.  Hence, back to the drawing board we went.

At that time, in one of the Jan lab group meetings, Lily mentioned a curious TMEM16F- calcium-dependent membrane expansion phenomenon that the Hilgemann group reported in Scientific Reports⁴.  Thus, we meandered into CRISPR-Cas9 engineering with an aliquot of Jurkat cells the Weiss lab provided (sidenote:  we considered ourselves very lucky to get an aliquot of Jurkat cells from the person who first isolated them.  Thank you, Art!).  Given the dearth of biological information on the intersection of membrane proteome and membrane expansion, we were stuck making sense of the massive proteomic data we collected.  It wasn’t until a follow-up Nature Communications⁵ paper from the Hilgemann group reported the crosstalk between TMEM16F and PIEZO1 that prompted us to examine our data further, and found the absence of extracellularly exposed PIEZO1, which gave us more confidence in our data.  The rest is history (and described in our Communications Biology paper¹).

As the saying goes, a great method deserves an equally great name.  Over time, our lab members knew this method under various names.  First iterations included SLURP for Surfaceome Labeling Using Recombinant Peroxidase (short, funny, onomatopoeia, and epicurean) and then BURP for Biotinylation Using Recombinant Peroxidase (ditto).  Good sense (and a veto from the hockey fan) eventually prevailed, and our method was eventually dubbed SLAPSHOT for Surface-exposed protein Labeling using PeroxidaSe, H₂O_2, and Tyramide-derivative.  SLAPSHOT is, in reality, a backronym, since the acronym was conceived first after pondering words such as “powerful, accurate, and fast” that are descriptive of the new method, and the constituent words just fell in place after that effortlessly.

References:

1. Tuomivaara, S. T., Teo, C. F., Jan, Y. N., Wiita, A. P. & Jan, L. Y. SLAPSHOT reveals rapid dynamics of extracellularly exposed proteome in response to calcium-activated plasma membrane phospholipid scrambling. Commun Biol 7, 1060 (2024).
2. Wollscheid, B. et al. Mass-spectrometric identification and relative quantification of N-linked cell surface glycoproteins. Nat Biotechnol 27, 378–386 (2009).
3. Rhee, H.-W. et al. Proteomic mapping of mitochondria in living cells via spatially restricted enzymatic tagging. Science 339, 1328–1331 (2013).
4. Bricogne, C. et al. TMEM16F activation by Ca(2+) triggers plasma membrane expansion and directs PD-1 trafficking. Sci Rep 9, 619 (2019).
5. Deisl, C., Hilgemann, D. W., Syeda, R. & Fine, M. TMEM16F and dynamins control expansive plasma membrane reservoirs. Nat Commun 12, 4990 (2021).