Establishment of Proximity-dependent Biotinylation Approaches in Different Plant Model Systems

1 Ghent University, Department of Plant Biotechnology and Bioinformatics, Technologiepark 8 71, 9052 Ghent University, Ghent, Belgium. 9 2 VIB Center for Plant Systems Biology, Technologiepark 71, 9052 Ghent, Belgium. 10 3 Faculty of Biology, Cell Biology, University of Freiburg, Germany. 11 4 Department of Plant Biology, Uppsala BioCenter, Swedish University of Agricultural 12 Sciences and Linnean Center for Plant Biology, Uppsala, Sweden. 13 5 Department of Biochemistry and Microbiology, Ghent University, Ghent, Belgium 14 6 Department of Biochemistry, Ghent University, Ghent, Belgium. 15 7 VIB Center for Medical Biotechnology, Ghent, Belgium. 16 8 VIB Proteomics Core, Ghent, Belgium. 17 9 Department of Applied Genetics and Cell Biology (DAGZ), University of Natural 18 Resources and Life Sciences (BOKU), Vienna, Austria. 19 10 CIBSS – Centre for Integrative Biological Signalling Studies, University of Freiburg, 20 Germany 21 11 Department of Biology, University of Crete, Heraklion, Greece. 22 12 Institute of Molecular Biology and Biotechnology, Foundation for Research and 23 Technology Hellas, Heraklion, Greece. 24 25 * joint first authors 26 # joint senior and corresponding authors 27 28 Running title: Proximity-dependent biotinylation in plants 29

TPLATE complex. We furthermore present a straightforward strategy to identify both non-48 biotinylated as well as biotinylated peptides in a single experimental setup. Finally, we provide 49 initial evidence that our approach has the potential to infer structural information of protein 50 complexes. 51 INTRODUCTION growth conditions used (i.e. cultivation of hairy roots was performed at 22-25°C) (Branon et 161 al., 2018). Noteworthy, only residual trans-biotinylation was observed when no exogenous 162 biotin was added to the liquid grown hairy root cultures. Therefore, the addition of surplus 163 (free) biotin seems also to function as a trigger of PDL in this system. This observation 164 indicates that PDL in plants (to some extent) might also have the capacity to identify the 165 spatiotemporal dynamics of interactome composition. 166 169 We used transient transformation of Nicotiana benthamiana leaf mesophyll cells to test the 170 applicability of PDL in a second model system commonly used for protein expression in planta 171 under various conditions. In this case, biotin was infiltrated directly into leaf tissue 24 h after 172 transformation and harvested 24 h post-biotin infiltration (Supplemental Figure 3A). We 173 confirmed that also in this system, the highest cis-biotinylation level was observed in case of 174 TurboID, and supplementation of biotin was important for the efficient detection of cis-175 biotinylation (Supplemental Figure 3B). Furthermore, the overall biotinylation output signal 176 in tobacco leaves was higher when biotin concentration was increased from 50 μM to 1 mM experimental conditions such as expression levels and exposure time to biotin is greatly 255 advised. 256

PDL-efficiency depends on growth temperatures and PBL can facilitate trans-biotinylation 168 in Nicotiana benthamiana
In summary, these data clearly show that TurboID-mediated PDL can be efficiently 257    Figure 6). 300 To test the effect of the linker and to further evaluate the activity of different PBLs in 301 Arabidopsis cell culture, transgenic cultures were grown for 24h, with and without exogenous 302 biotin at 28C, and expression and biotinylation were assessed via Western blotting 303 (Supplemental Figure 6). Protein abundance of the BioID and BioID2 constructs was 304 comparable to their respective controls in our cell cultures and was not affected by the addition 305 of biotin. Only TPLATE-BioID2 levels were rather lower. At the level of cis-and trans-306 biotinylation, we observed different patterns for each of the fusion proteins used. As several of 307 the detected bands which increased significantly in the presence of biotin, did not correspond 308 to bands in the control or GFP-BioID culture and varied between the different PBLs, they likely 309 represent different trans-biotinylated interactors and suggest that the outcome of a BioID-based 310 interaction assay might partially depend on the PBL used. TPLATE-linker PBL showed the 311 most complex biotinylation pattern when comparing to the other setups expressing BioID and 312 BioID2 fusions (Supplemental Figure 6), suggesting that the addition of a linker may be used 313 to enhance proximity labelling. Consistent with the results described for tobacco, TurboID 314 constructs showed some residual biotinylation without the addition of exogenous biotin, 315 increased biotinylation after 1 h incubation with biotin and gave rise to an extensive 316 biotinylation pattern after 24 h incubation with biotin in both control and bait cultures, 317 suggesting it is highly promiscuous. 318 part) be skewed due to the lower expression levels of the latter. Adding a flexible linker 321 increased cis-biotinylation levels of the bait compared to the constructs without linker 322 (Supplemental Figure 6A  In agreement with the higher stringency of the isolation procedure, the smallest TPC 354 subunit, LOLITA, which was robustly detected using AP-MS (Gadeyne et al., 2014) and, as 355 shown here, without being denatured before binding to streptavidin beads (Figure 3), was no 356 longer detected ( Figure 4B, Supplemental Data Set 2). LFQ revealed that the remaining seven 357 TPC subunits, including the bait TPLATE, were detectable using BioID, linkerBioID, 358 linkerBioID2 and linkerTurboID, although not all subunits were significantly enriched 359 compared to the GFP PBL control using our statistical threshold criteria (FDR 0.05 and S0 of 360 0.5). The TASH3 and TWD40-2 subunits, for example, could not be confidently identified with 361 all PBLs. For BioID2, this might be caused by the reduced expression level of the bait in these 362 cultures (Supplemental Figure 6), yet this does not explain why this low level of detection is 363 not observed for the other subunits as well (Figure 4). We also conclude that adding a long 364 linker increased the robustness of prey identification. For example, using TPLATE-365 linkerBioID, the TASH3 subunit was detected with 15 peptides compared to only 2 peptides 366 when using TPLATE-BioID (Supplemental Table 3). We did not identify TASH3 with 367 TPLATE-BioID2, in contrast to TPLATE-linkerBioID2, where we identified TASH3 with 59 368 peptides (Supplemental Table 3). 369 Noteworthy, increasing the concentration of biotin from 50 M to 2 mM adversely affected 370 TPC subunit detection as only the bait itself could be identified. It is likely that increasing 371 biotin concentrations causes residual free biotin to accumulate in the protein extract, even after 372 protein desalting to deplete free biotin, thereby occupying the streptavidin binding sites on the 373 beads which are saturated at >9 µM of biotin. We tested this "saturation hypothesis" using N.  It should be noted that the fold change by which the other TPC subunits were detected 381 with TurboID was comparable or sometimes even lower (e.g. AtEH2/Pan1) compared to the 382 other BioID forms tested (Figure 4). This was caused by the fact that TPC subunits were 383 identified with higher abundance in the TurboID control samples, resulting in lower relative 13 fold changes. All individual TPC subunits were detected with more than 20 unique peptides 385 using the GFP-linkerTurboID whereas TWD40-2 was the only TPC subunit detected in the 386 other control GFP-PBLs, which explains its overall low fold change (Supplemental Table 3). 387 Nevertheless, TurboID identified most of the TPC subunits more robustly compared to the 388 other PBLs, as evidenced by the overall higher -log10p-values. So, although in our case, 389 TurboID showed to be superior to all others in identifying the other TPC subunits, the lower 390 signal/noise ratio of TurboID, due to its increased activity, might work as a disadvantage to 391 observe differences between bait proteins and control samples, which might even be enhanced 392 if the proteins are targeted to specific subcellular locations. 393 394

PDL and AP-MS 396
To further evaluate PDL, we compared the relative levels compared to the bait by which the 397 different TPC subunits were detected using PDL using our stringent washing protocol with a 398 one-step IgG-based pull-down (PD) protocol using the GS rhino tandem affinity purification 399

Identification of biotinylated peptides enhances the identification power of PDL and allows 433 identifying structural relationships between complex subunits 434
The interaction between biotin-streptavidin is strong enough to be maintained even under harsh 435 conditions (Supplemental Figure 8). Thus, biotinylated peptides are expected to be retained 436 on the streptavidin beads. Following stringent washing under denaturing conditions, on-bead 437 digest will release non-biotinylated proteins, which can subsequently be identified using LC-438 MS/MS. This approach, however, does not provide direct evidence for biotinylation and it 439 relies on the assumption that only biotinylated proteins remain bound to the beads after the 440 washing steps. To acquire direct proof of biotinylation, and to further enhance the power of 441 PDL to identify interactors, release of biotinylated peptides from the Streptavidin beads and 442 their subsequent MS-based identification is required. 443 Thus, we expanded the protocol ( Figure 5B) to also be able to identify biotinylated 444 peptides. For this, we included a second elution step (see Materials and Methods) to release 445 the biotinylated peptides from the beads using an adapted protocol based on previous work 446 (Schiapparelli et al., 2014). This approach enables the detection of both non-biotinylated as 447 well as biotinylated peptides in the same experimental setup. proteins with respect to the bait. We, therefore, tested whether biotinylated peptides could 479 reveal differential proximity between specific domains of TPC subunits using the TPLATE- We provide a comprehensive comparison of various PBL based proximity labelling strategies 498 in plants. We show that TurboID is the most promiscuous PBL, and that this sometimes leads 499 to a lower signal to noise ratio. We also provide guidelines and approaches for interactome 500 capture in various plant systems specifically focusing on proteins that are intrinsic or peripheral 501 to the plasma membrane. Furthermore, we show that for each bait/system conditions might 502 benefit from independent optimization. 503 We observed that in all three plant systems tested, the exogenous application of biotin Furthermore, controls may express at high levels and show increased diffusion due to their 537 smaller hydrodynamic radius, further skewing results. 538 We provide evidence that our methods and conditions apply to plasma-membrane 539 complexes. We showed that the interaction of the symbiotic RLKs NFR5 and SYMRK can be 540 identified by exploiting PDL and particularly the PBL TurboID. Furthermore, the use of proper 541 negative controls is imperative. However, even though the brassinosteroid receptor BRI1 was 542 not co-immunoprecipitated with the symbiotic receptors in a previously published dataset 543 (Antolin-Llovera et al., 2014), we detected weak biotinylation of this RLK and the immune-544 receptor FLS2. While it could be interpreted as unspecificity within the PBL system, it should 545 also be considered, that PBL allows labelling of transient interactions or proximal proteins. As 546 a consequence, continuous unstable interactions accumulate to detectable amounts of proteins 547 and would thus allow their identification. As PDL using TurboID is capable of trans-548 biotinylation in the range of minutes (15 min under our experimental conditions), the 549 enrichment of unstable interactions would thus be more prominent. Therefore, putative 550 interactions identified by PBL still need to be verified using independent experimental systems 551 but comparisons between the different experimental systems should always reflect the technical 552 limitations of each approach. 553 By expanding our protocols and PBLs into Arabidopsis cell cultures, we could not only 554 reproduce the composition of the TPC except for one subunit, but we could also robustly 555 identify and confirm other CME players and novel interactors using the third generation PBL. 556 We show that MS-based identification of interactors is more robust using prolonged biotin 557 exposure of Arabidopsis cell cultures and that the use of linkers can be advantageous when it 558 comes to identifying protein-protein interactions of multi-subunit complexes. Gateway reaction resulted in translational fusions between the eGFP and the proximity labels, miniTurbo and Pro35S::eGFP-BioID construct (in pKm43GW), with a C-terminally triple HA-620 tagged BioID fused to eGFP.
were synthesized and codon-optimized using the codon optimization tool of Integrated DNA 625 Technologies, Inc. The ORFs were synthesized with BsaI overhands and were ligated to the 626 Level1/2 vector pICSL86900 and pICSL86922, as previously described (Patron et al., 2015). 627 The Ratiometric BiFC images were obtained using an Olympus FV1000 inverted confocal 865 microscope equipped with a UPLSAPO 60x water immersion objective (NA 1.2). Images were 866 acquired in line sequential mode, using 515 nm excitation and an emission window between 867 530 nm to 548 nm for the YFP detection and using 559 nm excitation and an emission window 868 between 580 nm to 615 nm for RFP detection. All images were taken using the exact same 869 settings. The experiment was independently repeated twice with similar outcome. 870 For the quantification of the YFP/RFP ratio, only images with less than 1% saturation in the 871 RFP or YFP channel were analysed. For each confocal image, parts of the cortical cytoplasm 872 in the RFP channel were traced in ImageJ using the selection brush tool with a width of 15 873 pixels. Histogram analysis was performed to confirm that less than 1% saturated pixels were 874 present in the ROI. The average intensity from the obtained ROI was calculated and divided 875 by the average intensity of the same region in the YFP channel. Ratios were quantified for 15 876 to 19 individual cells.  Table  1000 6. In figure 6, a subset of proteins is presented.