A Photo‐Caged Cross‐Linker for Identifying Protein‐Protein Interactions Authors: Xingyu Chen, Cristina Jiménez López, André Nadler, Florian Stengel Published: 2024 ABSTRACT AbstractCross‐linking mass spectrometry (XL–MS) has seen significant improvements which have enhanced its utility for studying protein‐protein interactions (PPIs), primarily due to the emergence of novel crosslinkers and the development of streamlined analysis workflows. Nevertheless, poor membrane permeability and side reactions with water limit the extent of productive intracellular crosslinking events that can be achieved with current crosslinkers. To address these problems, we have synthesized a novel crosslinker with o‐nitrobenzyl‐based photoresponsive groups. These o‐nitrobenzyl ester (o‐NBE) groups enhance the stability and hydrophobic properties of the crosslinker and add the potential for temporal resolution, i. e. the ability to control the initiation of the crosslinking reaction. Upon exposure to UV light the resulting aldehyde product reacts with adjacent amino groups and subsequent reductive amination of the formed Schiff‐bases yields stable secondary amine linkages. This controlled activation mechanism enables precise UV‐triggered protein crosslinking. We demonstrate proof‐of principle of our o‐NBE cross‐linker to reliably detect PPIs by XL–MS using a recombinant model protein. We also demonstrate its ability to enter intact Hela cells, thereby indicating its future potential as a useful tool to study PPIs within the cellular environment.Keywords: Uncaging, Cross-linker, Mass spectrometry, Proteomics, Protein-protein interaction Introduction The ability of mass spectrometry (MS)‐based proteomics to identify and quantify thousands of proteins has fundamentally altered the way in which biological systems are studied today. [1] Whether individual proteins, protein‐protein interactions (PPIs) or system‐wide studies of whole organisms, all levels of biological complexity can be addressed using different MS‐based approaches. PPIs as key constituents of protein networks are increasingly explored by the rapidly advancing technology of chemical cross‐linking coupled to mass spectrometry (XL–MS).[ 2 , 3 , 4 , 5 ] Over the past few years significant advancements in our understanding of PPIs have been made using XL–MS in various systems including cell lysates, cultured cells, [6] cellular organelles,[ 7 , 8 , 9 ] and tissues.[ 10 , 11 ] However, to effectively capture a PPI by XL–MS within the cell remains a significant challenge. [12] Conventional crosslinkers, such as disuccinimidyl suberate (DSS) or bis(sulfosuccinimidyl)suberate (BS3), rely on N‐hydroxysuccinimide (NHS) esters that rapidly react with primary amines, and particularly the amine groups of lysines. While these NHS cross‐linkers are easy to handle and lead to reliable and reproducible results, they have limitations. NHS cross‐linkers react with water and have limited abilities to cross the polar environment of the cellular membrane, which constrains their ability to capture PPIs in the aqueous environment of the cell. [13] Excellent recent attempts to address these problems involve designated handles for biochemical enrichment and improved membrane permeability such as PhoX and tBu‐PhoX (t‐butyl featured phosphoric acid based crosslinkers).[ 14 , 15 ] Alternatively, the use of photoresponsive cross‐linkers capable of interacting with various amino acids via highly reactive carbene intermediates has been shown to allow for a controlled initiation of the cross‐linking reaction and to reduce side reactions with water[ 16 , 17 , 18 ] even though diazirine based crosslinkers present other drawbacks such as poor efficiency and dependence on pH. In this work, we introduce a novel crosslinker with o‐nitrobenzyl‐based photoresponsive groups. Its o‐nitrobenzyl ester (o‐NBE) groups enhance the hydrophobic properties of the crosslinker. Upon exposure to UV light (365 nm), the resulting di‐aldehyde product reacts with adjacent amino groups and subsequent treatment with reducing agent results in stable secondary amine linkages. The UV‐light‐triggered ortho‐nitrobenzyl ester reaction has been investigated for many years and has been effectively utilized in numerous applications, where o‐NBEs can serve as light‐activated probes that target the amine group of lysines.[ 19 , 20 , 21 , 22 ] However, in the context of cross‐linkers used for the detection of PPIs by MS, this specific reaction represents a new strategy. o‐NBE potentially offers multiple downstream advantages for monitoring protein complexes in the cell, as the hydrophobicity of o‐NBE should facilitate the crosslinker's penetration of the lipid membranes of the cell and as the initiation of the reaction can simultaneously be controlled by UV‐light. Here, we show the synthesis of a novel o‐NBE cross‐linker and demonstrate its basic functionality as a protein‐protein linker for XL‐MS using the recombinant protein BSA as a test case. We also show the ability of o‐NBE to enter intact Hela cells, thereby indicating the future use of o‐NBE as a cellular cross‐linker. Experimental Section Synthesis of the o‐Nitrobenzylester (o‐NBE) Crosslinker All chemicals were purchased from commercial sources and were used without further purification. Solvents for flash chromatography were obtained from VWR and dry solvents were obtained from Sigma. Deuterated solvents were obtained from Deutero GmbH, Karlsruhe, Germany. TLC was performed on precoated silica plates (Merck, 60 9 F254) using UV light (254 or 366 nm) or a solution of phosphomolybdic acid in EtOH (3 % phosphomolybdic acid in 100 mL EtOH) for analysis. Preparative column chromatography was performed with 40–63 μm silica gel from VWR Chemicals with a pressure of 1–1.5 bar. 1H and 13 C‐NMR‐spectra were obtained on a 400 MHz Bruker Ascend TM Nanobay spectrometer. J values are given in Hz, and chemical shifts are in ppm. Splitting patterns are designated as follows: s, singlet; d, doublet; t, triplet; q, quartet; m, multiplet; b, broad. Mass spectra (ESI) were recorded using a QExactive instrument (Thermo Fisher Scientific) equipped with a robotic nanoflow ion source. FTMS spectra were acquired within the range of m/z 200–1200 with a target mass resolution of m/z 200=140000. The spectra were evaluated using the Xcalibur Qual Browser software. tert‐butyl (2‐(4‐(hydroxymethyl)‐2‐methoxy‐5‐nitrophenoxy)ethyl)carbamate (1) 0.39 g of 4‐(hydroxymethyl)‐2‐methoxy‐5‐nitrophenol (1 equiv., 1.95 mmol), 0.61 g of tert‐butyl (2‐bromoethyl) carbamate (1.4 equiv., 2.73 mmol), and 0.49 g of potassium carbonate (K2CO3, 1.8 equiv., 3.51 mmol) were dissolved in acetonitrile (MeCN, 10 mL). This mixture was subsequently heated to reflux and stirred overnight. After 18 hours, the solvent was removed under reduced pressure and the reaction mixture was purified by flash column chromatography (eluents 30 % cyclohexane in ethyl acetate), yielding 220 mg (46 %) of tert‐butyl (2‐(4‐(hydroxymethyl)‐2‐methoxy‐5‐nitrophenoxy)ethyl)carbamate (1) as a yellow oil. The 1H and 13C NMR figures are shown in Supplementary Figure 6 and 7. 4‐(2‐((tert‐butoxycarbonyl)amino)ethoxy)‐5‐methoxy‐2‐nitrobenzyl acetate (2) 135 mg (394 μmol, 1 equiv.) of (1) and 10 mg of 4‐dimethylaminopyridine (DMAP, 0.2 equiv., 93.5 μmol) were dissolved in dry dichloromethane (DCM, 8 mL) under an argon atmosphere., Trimethylamine (Et₃N, 0.133 mL) was added to this solution. The reaction mixture was stirred at room temperature for 5 minutes and subsequently treated with acetic anhydride (1.35 mL, 14.28 mmol, 36 equiv.). After 2 hours, the solvents were removed under reduced pressure. The crude product was purified by flash column chromatography (eluent 30 % cyclohexane in ethyl acetate), which yielded 141.8 mg (369 μmol, 94 %) of 4‐(2‐((tert‐butoxycarbonyl)amino)ethoxy)‐5‐methoxy‐2‐nitrobenzyl acetate (2). The 1H and 13C NMR figures are shown in Supplementary Figure 8 and 9. 4‐(2‐aminoethoxy)‐5‐methoxy‐2‐nitrobenzyl acetate (3) 140 mg (364 μmol) of (2) were dissolved in dichloromethane (DCM, 5 mL), to which 250 μL of triisopropylsilane (TIPS) were added. After 2 minutes, trifluoroacetic acid (TFA, 5 mL) was added to the reaction mixture which was stirred at room temperature for 10 min. Solvents were removed under reduced pressure and the crude product purified by flash column chromatography (eluent 25 % MeOH in chloroform), yielding 96.1 mg (337 μmol, 93 %) of 4‐(2‐aminoethoxy)‐5‐methoxy‐2‐nitrobenzyl acetate (3) as a colorless solid. The 1H and 13C NMR figures are shown in Supplementary Figure 10 and 11. HR‐MS (ESI positive) m/z calculated for C12H16N2O6: 284.1; identified: 285.1 [M+H]+. ((((octanedioylbis(azanediyl))bis(ethane‐2,1‐diyl))bis(oxy))bis(5‐methoxy‐2‐nitro‐4,1‐phenylene))bis(methylene) diacetate (4) 96 mg (337 μmol) of (3) and 62.9 mg of disuccinimidyl glutarate (193 μmol) were dissolved in dimethylformamide (DMF, 3 mL) with a 0.1 M sodium phosphate buffer at pH 8 (2 mL). The reaction was allowed to proceed at room temperature. After 1 hour, the reaction was quenched by evaporating all volatiles under reduced pressure. The crude mixture was purified by flash column chromatography (eluent 10 % MeOH in chloroform), yielding 63.5 mg (89.9 μmol, 27 %) of a white solid. The 1H and 13C NMR figures are shown in Supplementary Figure 12 and 13. HR‐MS (ESI positive) m/z calculated for C32H42N4O14: 706.7; identified: 724.3 [M+NH4]+. Cross‐Linking Mass Spectrometry (XL–MS) 100 μg of Bovine Serum Albumin (BSA) at a concentration of 1 μg/μl was mixed with BS3, DSS or the crosslinker o‐Nitrosobenzaldehyde (o‐NBE) at a final concentration of 1.5 mM. Cross‐linking with DSS was carried out essentially as described.[ 23 , 24 ] With o‐NBE the reaction was allowed to proceed over time intervals of 0 minutes, 10 minutes, 20 minutes and 30 minutes, using 365 nm UV light (Thorlabs M365 L2, 190 mW (min). Following UV‐light activation, the crosslinked samples were incubated at 37 °C under continuous shaking at 650 rpm for 3 hours in a Thermomixer (Eppendorf) before sodium triacetoxyhydroborate (STAB) at a concentration of 10 mM was added for 20 minutes to convert generated Schiff bases into stable secondary amine linkages by reductive amination. In‐gel digest was performed as previously described. [25] The crosslinked samples were separated by SDS‐PAGE and stained overnight. Protein bands were excised into small segments of around 1 cm2 and destained with a 50/50 (v/v) solution of acetonitrile (ACN) and 50 mM ammonium bicarbonate (ABi) and incubated for 15 minutes at 37 °C at 650 rpm. Subsequently, protein bands were reduced in 100 mM NH₄HCO3, 10 mM DTT for 30 min at 56 °C, alkylated in 100 mM NH₄HCO3, 55 mM iodoacetamide for 20 min at 23 °C in the dark, and digested with 13 ng/μl trypsin (Promega, sequencing grade) in 10 mM 100 mM NH₄HCO3, 10 % ACN for 2 h on ice, followed by 18 h at 37 °C. Samples were extracted and dried using a SpeedVac. Dried peptides were re‐suspended in 0.1 % TFA for 15 min at 37 °C and adjusted to pH <3 with TFA. The peptides were loaded onto Pierce C18 SpinTip columns (Thermo, 10 μg) and desalted according to protocol. Crosslinked peptides were separated on an EASY‐nLC 1200 system (Thermo Scientific) at a flow rate of 300 nl/min over a 60 min gradient (5 % acetonitrile in 0.1 % formic acid for 4 min, 5 % – 35 % acetonitrile in 0.1 % formic acid in 55 min, 35 % ‐ 80 % acetonitrile in 1 min) and measured on a Orbitrap Fusion Tribrid mass spectrometer (Thermo Electron, San Jose, CA). Full scan mass spectra were acquired in the Orbitrap at a resolution of 120,000, a scan range of 400–1600 m/z, and a maximum injection time of 50 ms. Precursor ions with charge states 3–8 were fragmented by CID at 35 % normalized collision energy in a data dependent mode. The duration for dynamic exclusion was set to 60 seconds. Analysis of MS data was conducted using Proteome Discoverer 2.4 (Thermo Scientific). The XlinkX mode was applied for the identification of cross‐linked peptides, while the SEQUEST HT search engine was employed for the detection of unmodified and dead‐end‐modified peptides using the following settings: acquisition strategy “non‐cleavable” and a minimum S/N value of 1.5. The calculated mass of the crosslinker was 552.22 Da (product 1), 550.21 Da (product 2), and 554.24 Da (product 3), respectively and lysines set as the primary reaction targets [20] (Figure 2). The maximum number of missed cleavage sites was set to 2 per cross‐linked peptide (not counting cross‐linked lysine residues) and minimum peptide length to 5. Dead‐end‐modified peptides (i. e. where o‐NBE reacted only one side with an amino acid) were allowed to contain multiple functional groups at the non‐reactive side of the cross‐linker including a reduced hydroxyl group (+572.248 Da), aldehyde group (+570.233 Da) or carboxylic acid group (+586.227 Da). These modifications have been collectively introduced as dynamic modifications, alongside methionine oxidation (+15.996 Da) and acetylation of the N‐terminus (+42.011 Da). Carbamidomethylation (+57.021 Da) on cysteines was used as a static modification. The Uniprot P02769 fasta file of BSA was used and a 0.05 FDR cut off in the percolator mode chosen. Cross‐linking with DSS was carried out essentially as described.[ 23 , 24 ] Hela Cells Cross‐Linking with o‐NBE 100 μL of Hela cells containing roughly 106 cells (2 mg/mL) were mixed with o‐NBE at a final concentration of 1.5 mM and BS3 at and a final concentration of 1.5 mM and 3 mM, respectively. o‐NBE was added to Hela cells at 1.5 mM using different UV‐ activation (365 nm, 190 mW) times from 0 minute to 20 minutes. Results and Discussion Synthesis of the o‐Nitrobenzylester (o‐NBE) Cross‐Linker We aimed to generate a crosslinker with o‐nitrobenzyl‐based photoresponsive groups, i. e. o‐nitrobenzyl esters as reactive groups (Figure 1). Therefore, of 4‐(hydroxymethyl)‐2‐methoxy‐5‐nitrophenol was coupled with tert‐butyl (2‐bromoethyl) carbamate to yield intermediate (1). Acetylation of compound 1 was performed using acetic anhydride with triethylamine and 4‐dimethylaminopyridine as catalyst, resulting in the formation of 4‐(2‐((tert‐butoxycarbonyl)amino)ethoxy)‐5‐methoxy‐2‐nitrobenzyl acetate (2). Subsequent acidic deprotection and amide coupling of compound (3) with disuccinimidyl glutarate yielded the desired o‐Nitrobenzylester crosslinker (4). Synthesis and structure of the synthesized o‐Nitrobenzylester (o‐NBE) cross‐linker. Reaction Mechanism of o‐NBE The o‐Nitrobenzylester serves as a photoresponsive moiety, undergoing cleavage upon UV activation. This cleavage releases an acetate group, generating an o‐Nitrosobenzaldehyde intermediate.[ 19 , 21 , 27 , 28 , 29 ] The o‐Nitrosobenzaldehyde intermediate then reacts with neighboring amino groups, facilitating crosslinking reactions in the synthesis of nitrogen heterocycles. [30] There are two possible reaction steps for ortho‐nitrobenzyl and amine derivates. The potential products are labeled as product 1, 2 and 3 and we tested all these various reaction products in our XL–MS workflow (Figure 2) and find evidence for all of them (Supplementary Table 1 and Supplementary Data 1). Proposed mechanism of the cross‐linking reaction between the functional groups of the o‐NBE cross‐linker and proteins with primary amino groups (Lys, N‐terminus).[ 19 , 27 , 28 , 29 ] Validation and Benchmarking of o‐NBE using BSA After successful synthesis of the o‐NBE linker, we tested and validated the cross‐linking reaction in a next step. We used Bovine Serum Albumin (BSA) [31] to evaluate the o‐NBE crosslinker and benchmarked it using the established NHS‐based cross‐linker as a positive control (Figure 3).[ 23 , 24 ] Figure 3A shows the SDS‐PAGE gel of BSA crosslinked with both DSS and o‐NBE cross‐linker, after UV‐light exposure of varying time length. The appearance of a band of cross‐linked BSA dimers together with a reduction of the monomeric BSA band observed after 10 minutes of UV light activation indicated the efficacy of the o‐NBE crosslinking reaction. A SDS‐PAGE gel of BSA without cross‐linker (control) and after crosslinking with DSS (1.5 mM) or o‐NBE (1.5 mM) using the indicated UV‐light (365 nm) activation times (0 minutes, 10 minutes and 20 minutes). “I” indicates area of cross‐linked BSA. B SDS Page gel of o‐NBE after different exposure times of UV light in Hela cells. BS3 or o‐NBE crosslinker was added to 100 μl Hela cells containing roughly 106 cells (2 mg/mL) at the indicated concentrations and UV‐activation time (365 nm, 190 mW) times. * Band with cross‐linked proteins and higher oligomers formed during cross‐linking. A SDS page gel of Hela cells plus/minus addition of o‐NBE after 20 minutes exposure to UV light (365 nm) as a control is shown in Supplementary Figure 1B. Additional optimization of UV‐light exposure times, ranging from 0 to 30 minutes, confirmed that a significant part of the o‐NBE linker had reacted after around 5 to 10 minutes of UV activation time and indicated 10 to 20 minutes of activation as optimal to obtain a maximal cross‐linking yield (Supplementary Figure 1A). However, when comparing the overall efficacy of the cross‐link reaction to DSS, it is evident that the cross‐linking reaction with o‐NBE is less effective, as indicated by the complete disappearance of the monomeric BSA band and the appearance of more crosslinked BSA species in the case of DSS (Figure 3A). As the relative hydrophobicity of o‐NBE should facilitate its transversal through the lipid membrane of the cell, we used intact HeLa cells to test the membrane permeability of our o‐NBE cross‐linker in a next step. We therefore added o‐NBE or BS3 to intact Hela cells before the cells were lysed and probed for the appearance of bands containing cross‐linked proteins, which typically serve as evidence of a successful crosslinking reaction within the cellular environment. [15] We find for o‐NBE the appearance of a UV‐dose dependent band of cross‐linked proteins, which is stronger than when BS3 is applied (Figure 3B). We therefore conclude from the data, that o‐NBE exhibits promising membrane permeability. Identified Protein‐Protein Cross‐Links Using o‐NBE Next, we extracted cross‐linked peptides from the SDS‐PAGE gel via in gel digestion and identified potential cross‐links using XlinkX. [32] To do so, we implemented o‐NBE within XlinkX as a novel, non‐cleavable cross‐linker with the properties as described above (see Experimental Section and Figure 2). The three possible cross‐linking products were implemented in XlinkX separately, containing the respective mass shifts of the possible products of the o‐NBE reaction (see product 1, 2 and 3 in Figure 2). We find that after sufficient UV‐activation time (i. e. 20 minutes) all the expected products are readily detected across replicates (Supplementary Table 1). Products 2 and 3 appear to be the primary products, likely because product 1 is depleted by the reductive amination that generates product 3. We find that after 20 minutes of UV light exposure more crosslinked peptides can be detected in the BSA sample cross‐linked with oNBE than after no activation or UV‐activation for only 10 minutes. Together with our SDS‐PAGE qualitative data (Figure 3), these results indicate that the cross‐linking reaction works in a reproducible, UV‐dose dependent manner (Figure 4A). A Number of identified cross‐link‐spectral matches (CSMs) within BSA after cross‐linking with o‐NBE after different UV‐light activation times (n=2). B Overlap of identified unique cross‐linking sites on BSA using BS3 or o‐NBE as a cross‐linker. C Mapping of identified unique cross‐links onto BSA (PDB identifier: 6qs9) using BS3 or o‐NBE (20‐minutes activation time). Cross‐links with a Ca−Ca distance <35 Å for BS3 and <50 Å for o‐NBE are shown and colored in blue (for Ca−Ca distances of all identified crosslinks see Supplementary Figure 3). However, a comparison with the widely used NHS cross‐linker BS3 showed that more cross‐links were identified using BS3 compared to o‐NBE. A closer look reveals that there are several cross‐links that were identified by both BS3 and o‐NBE (Figure 4B). As both cross‐linkers are quite distinct in their reaction mechanism ‐ even though a lysine‐lysine cross‐link is in both cases the most likely protein‐protein cross‐link product (with an estimated yield of over 90 % according to literature [20] ) ‐ one would not expect a complete or even large overlap in terms of the exact amino acid composition of the cross‐linked peptides. Moreover, when mapping the identified cross‐links onto BSA (PDB identifier: 6qs9), the cross‐links show a significant overlap in terms of connected peptide pairs that are identified for the two linkers but at the same time show that complementary information can be gained by using NHS linkers together with o‐NBE (Figure 4C, Supplementary Figures 2, 3, 4 and 5, Supplementary Table 2). In summary, our data shows that o‐NBE can reliably form covalent cross‐links using the test protein BSA. Comparison and benchmarking o‐NBE against BS3 and DSS, two commonly used NHS‐based cross‐linkers, confirms that similar but distinct peptide pairs are linked by these crosslinkers, even though the overall cross‐linking yield is not as high for o‐NBE as for NHS‐based cross‐linkers. Conclusions We report the design and synthesis of a cross‐linker with o‐nitrobenzyl‐based photoresponsive groups. The o‐NBE group enhances the hydrophobic properties of the cross‐linker and enables light‐controlled activation, thereby providing control over the protein crosslinking reaction. We demonstrate the basic function of our o‐NBE cross‐linker using a recombinant model protein and benchmark it again two commonly used NHS‐ester based cross‐linkers, BS3 and DSS, and show that o‐NBE can link comparable but distinct protein sites, even though the overall cross‐linking yield is not as high. It is likely that only the most reactive lysines react with the generated aromatic aldehyde, which might be exploited to increase selectivity in future applications. We likewise reason that the lower cross‐linking yield could be at least partly explained by a suboptimal fragmentation behavior of our current o‐NBE linker. Also, the efficiency of the reduction reaction to convert the Schiff bases into amines to form a stable linkage is another likely reason, as we deliberately chose a mild reductive reagent (i. e. STAB [33] ) and moderate reaction temperature of 37 °C in order to be amenable to future in vivo applications at the price of optimal cross‐link yields. Moreover, we find that a sufficient number of crosslinked peptides is observed only when using collision‐induced dissociation (CID), but not under a regime where HCD is used to fragment the linker. A cross‐link search operating the MS in HCD‐only fragmentation mode did yield zero potential cross‐linked peptide pairs (data not shown). This hampers interpretation of the data by some workflows that are optimized for HCD collision of non‐cleavable linkers, as for example pLink. [34] The currently limited solubility of o‐NBE remains likewise a challenge, as solvent‐based stock solutions of o‐NBE tend to easily precipitate in the aqueous environment of a protein complex or, eventually, the cell. However, our preliminary data using intact Hela cells indicates a potential future application of our o‐NBE crosslinker for in‐vivo crosslinking. We also anticipate that alternative designs, potentially including a MS‐cleavable version of the linker, will alleviate some of its current problems. Despite these challenges, the demonstrated capabilities of o‐NBE as a protein‐protein cross‐linker and its hydrophobic basic chemical structure illustrate its future potential as a useful complementary tool to study PPIs within the cellular environment. Author Contributions Supplementary Data containing Figures 1 to 13 and Supplementary Table 1 to 2 (PDF file). Supplementary Data 1 (referring to Figure 4) (Excelfile) Author Contributions X.C., C.J.L., A. N. and F.S. designed research; X.C. and C.J.L. performed research; X.C., C.J.L., A. N. and F.S. analyzed data; X.C. C.J.L., A.N. and F.S. wrote the paper. All authors provided critical feedback on the paper. Conflict of Interests The authors declare no competing interests. Supporting information As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors. Supporting Information Supporting Information Acknowledgments This work was supported by the German Research Foundation, project number 496470458, SFB969 and TRR 353/1–471011418 (Project A07). Open Access funding enabled and organized by Projekt DEAL. Chen X., Jiménez López C., Nadler A., Stengel F., ChemBioChem 2024, 25, e202400620. 10.1002/cbic.202400620 Contributor Information André Nadler, Email: nadler@mpi-cbg.de. Florian Stengel, Email: florian.stengel@uni-konstanz.de. Data Availability Statement The MS raw files, crosslink databases and original XlinkX result files have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository[26] with the project accession number (PXD052552). References Associated Data Supplementary Materials As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors. Supporting Information Supporting Information Data Availability Statement The MS raw files, crosslink databases and original XlinkX result files have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository[26] with the project accession number (PXD052552).