Mechanistic insights into adhesion GPCR autoproteolysis by a multiscale computational approach

Mechanistic insights into adhesion GPCR autoproteolysis by a multiscale computational approach | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Mechanistic insights into adhesion GPCR autoproteolysis by a multiscale computational approach Florian Seufert, Yin Kwan Chung, Robin Schick, Fabian Pohl, Norbert Sträter, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7121662/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract In cis -autoproteolysis, proteins self-catalyze cleavage of their peptide chains. In the GAIN domain of adhesion G protein-coupled receptors, cis -autoproteolysis yields a tethered agonist critical for receptor activation. Contrary to current reaction models, the GAIN HL|T catalytic triad is insufficient for cleavage. Here, we investigated the cleavage mechanism in the rat ADGRL1 GAIN domain using multiscale modeling combining molecular dynamics and QM/MM simulations to investigate cis -autoproteolysis as a once-in-a-lifetime event. We present an updated and unique GAIN domain cis -autoproteolysis mechanism. The initial N-O acyl shift proceeds via a hydroxy-oxazolidine intermediate, sterically shifting subsequent reaction steps away from the triad base H GPS.-2 . Water on its opposing side, coordinated by acidic residue E656 H6.50 , is essential for completing the N-O acyl shift facilitating ester hydrolysis. Our study provides detailed mechanistic insights into cis -autoproteolysis, highlighting how conserved residues and structured water networks adjacent to the triad enable a chemically precise reaction within an inherently flexible protein not evolutionarily optimized for high catalytic rate. These findings deepen our understanding of aGPCR processing and may apply broadly to other autoproteolytic proteins, offering a framework to explore similarly elusive one-time cleavage events in biology. Computational Chemistry Molecular Biology Biophysics GAIN Domain adhesion GPCR autoproteolysis QM/MM Molecular Dynamics ADGRL1 Multiscale Modeling Stachel peptide Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Adhesion G-protein coupled receptors (aGPCR) comprise a family of GPCRs involved in cell adhesion, migration, paracrine signaling and various disease pathologies 1 – 3 . aGPCRs are characterized by a heptahelical transmembrane (7TM) domain and their large extracellular regions containing various protein domains. Their hallmark domain is the GPCR autoproteolysis-inducing (GAIN) domain containing the GPCR proteolysis site (GPS) 4 , 5 . The GAIN domain is composed of two subdomains, a variable helical subdomain A and a more structurally conserved β-sandwich subdomain B containing the GPS with the conserved catalytic triad of H GPS.−2 , L GPS.−1 and S/T GPS.+1 (superscript indicates GAIN domain generic residue numbers 6 ). The GPS is flanked by two flexible regions enveloping the catalytic triad termed Flap 1 and 2, mediating triad solvent-accessibility 7 . The GPS mediates cis -autoproteolysis, forming an intact complex between the N-terminal fragment (NTF) composed of all extracellular domains, and the C-terminal fragment (CTF) with the most C-terminal one of 14 β-strands of the GAIN domain followed by the 7TM region 1 , 8 , 9 . The cleaved β-strand comprises the N-terminus of the CTF in the NTF-CTF complex. It has been shown to mediate receptor activation as an intramolecular agonist in many aGPCRs like ADGRL1, ADGRD1, ADGRG1/G5/G6, and is termed the “ Stachel ” element 10 – 13 . The NTF-CTF complex can be disrupted by e.g. mechanostimulation, resulting in NTF shedding off the CTF 4 , 14 – 18 . Shed NTF of some receptors moves away from the cell surface 19 , 20 , to stimulate aGPCR signaling in the extracellular space 16 , 21 and leave the exposed Stachel acting as an intramolecular receptor agonist on the 7TM domain. Despite recent structural insights into Stachel mediated G protein activation 1 , 22 , the cleavability of aGPCRs and the role of cleavage for their physiological functions remains unclear. The population of some receptors exist cleaved, others in an uncleaved state or even partially in both states like ADGRG1 9,23 .⁠ Moreover, the impact of cleavage on receptor function seems to be receptor-specific, as some uncleaved receptors also exhibit activity 24 . It has therefore been postulated that separate pathways may exist for aGPCR signaling depending on their cleavage competence 4 , 25 . Finally, the GPS turned out to be necessary but not sufficient for cleavage, pointing towards additional structural determinants of cleavage competence 8 , 26 . A detailed structural and mechanistic understanding of cis -autoproteolysis is thus a key issue of current aGPCR research. In this work we used a multiscale computational approach combining quantum mechanics (QM), molecular mechanics (MM) and molecular dynamics (MD) simulations to investigate the mechanism of cis -autoproteolysis. QM allows for simulating chemical reactions by resolving the electronic configurations on a set of atoms. For simple systems, reactions are modeled in vacuum, however for reactions in proteins, the more complex environment of the reaction site is critical for catalysis. For that purpose, hybrid methods such as QM/MM expand the explicit QM system by a flexible, however inert environment treated on the MM level 27 , 28 . QM/MM hybrid approaches are established methods to simulate cleavage of peptide bonds by proteases. The local environments of the active sites of these enzymes have evolutionarily been optimized for high catalytic rate, with a k cat of e.g. trypsin in the per second range 29 . For maximum catalytic effect, the amino acids involved in the reaction and water required for hydrolysis are usually well-coordinated. In this reaction, a base abstracts a proton, creating a nucleophile for acylation of the peptide bond, resulting in an oxyanionic tetrahedral intermediate. In the next step, the N-O acyl shift is completed through protonation of the backbone amide nitrogen and coordination of water for subsequent ester hydrolysis, which finally yields the cleaved product. In contrast to proteolytic enzymes, the cis -autoproteolytic reaction is orders of magnitude slower and has not been evolutionary optimized, likely because it only occurs once in protein lifetime. For such systems, reaction pathways may be less well-defined, with chemical reactions proceeding through multiple conformationally and chemically distinct intermediates. This mechanistic ambiguity limits the applicability of QM/MM approaches that were developed based on studies of classical enzymes. The QM/MM pathway analysis was thus supplemented with MD simulations to sample conformational changes of putative intermediates on time scales exceeding the orders of magnitude faster chemical reactions. Because force field-based MD simulations employ fixed protonation states and cannot account for changes in electronic configurations, the GAIN domain structure was simulated at different protonation states to identify the most thermodynamically favorable protonation configuration. Energetic evaluations of these configurations provided the appropriate starting geometries for subsequent QM/MM reaction pathway analysis. Our multiscale MD and QM/MM approach suggests an updated sequence of events during GAIN domain autoproteolysis that involves, in addition to the HL|T catalytic triad residues, conserved residues in the direct environment of the GPS: In the first reaction step, the H GPS.−2 acts as a base, abstracting a proton from the S/T GPS.+1 hydroxyl group via a pre-formed hydrogen bond (Fig. 1 a). The nascent nucleophilic alkoxide initiates a nucleophilic attack on the carbonyl carbon of L GPS.−1 , forming the five-membered heteroatomic ring intermediate T1, with its oxyanion solely coordinated by water (Fig. 1 b). Subsequently, the intermediate collapses through protonation of the ring amide by a water molecule located behind the catalytic triad in a pocket between T GPS.+1 and the conserved E H6.50 , completing the N-O acyl shift, resulting in the ester intermediate E (Fig. 1 c). The nascent hydroxide anion is coordinated by E H6.50 , which forms part of a larger polar network between the catalytic triad and the C-terminal helix of GAIN subdomain A. Our analysis suggests a one-step mechanism with ester hydrolysis occurring directly after the completed N-O acyl shift and provides explanations why the catalytic triad is necessary but not sufficient for cleavage 8 , 30 , 31 . Our computational findings are in agreement with previous experimental data 26 and further supplemented by mutational and functional studies on E H6.50 , a residue that coordinates a water network enveloping the catalytic triad and which protonation state impacts GAIN domain cleavage. Results Preparation of the reactant R for the QM/MM analysis workflow We selected the GAIN domain of rat ADGRL1 (L1), a neuronal aGPCR 32 involved in glucose homeostasis and energy regulation 33 , 34 as model to investigate the detailed autoproteolysis mechanism. For this particular system, there is a wealth of experimental data that can inform further interpretation 7 , 11 , 30 , 33 – 36 . To obtain an uncleaved GAIN domain, we re-formed the cleaved peptide bond at the GPS from the X-ray structure of the L1 (Ext. Data Fig. 1 a-b). After several equilibration runs interposed by steered MD simulations, the backbone of the triad residues shows excellent agreement to the available uncleaved GAIN domain structure of ADGRB3, which lacks the canonical H-L-S/T motif (Ext. Data Fig. 1 d) 8 . The reactant model of L1 selected from the MD simulations (Supp. Figures 4 –9) was further validated and cross-referenced to the structure of the uncleaved ADGRB2 GAIN domain, with a canonical H-L-S/T motif 26 . Finally, the product of ester hydrolysis at the end of the reaction trajectory reveals the configuration of the cleaved peptide bond at the GPS from the X-ray structure of L1, further corroborating the uncleaved reactant model. To simulate bond formation and proton transfer reactions we selected the residues from the GPS triad and proximal regions for QM calculations and defined their protonation states (Fig. 2 ). The QM selection contains two titratable residues, H GPS.−2 and E H6.50 . From the catalytic triad, H GPS.−2 was protonated at N ε (HSE). Only in this protonation state, the system adopts the reactant-characterizing hydrogen bond between T GPS.+1 and H GPS.−2 in L1 and B2 26 . Our mechanistic approach furthermore reveals that in L1 the probability of this bond to be formed is influenced by the protonation state of E H6.50 (Fig. 2 b). This residue appears to coordinate a water network enveloping the catalytic triad (see below). Empirical pKa estimation by PropKa3 initially indicated E H6.50 as charged (referred to as E-) 37 . Regardless, MD simulations on both E656 protonation states revealed that protonated E H6.50 (referred to as EH) enhances formation of the hydrogen bond between T GPS.+1 and H GPS.−2 (Ext. Data Fig. 2 , Supp. Table 1). From the proximal GPS triad region, the highly conserved F S10.49 was included in the QM selection. In the reactant, F S10.49 forms a T-shaped π-π interaction with H GPS.−2 . Functional analysis of L1 and B2 variants show that F S10.49 is essential for cleavage competence in Pohl et al. , 2023 26 . The QM selection, visualized in Fig. 2 , further includes N839 S14.45 (GPS.+2), Q818 s11s12 , E656 H6.50 , S770 S8.54 and additional atoms of adjacent residues to allow only carbon-carbon bonds to form the QM system boundary (see Supp. Data 1). The QM/MM system was finally created with the explicit QM atoms surrounded by a MM-treated sphere consisting of all residues with at least one atom within a 5.0 Å sphere of the QM selection (Fig. 2 c). The N-O acyl shift is favored by protonated E H6.50 and proceeds on both sides of the peptide plane The N-O acyl shift is the first chemical reaction step in autoproteolysis (Fig. 1 a, b, 3 a). Before we studied the reaction mechanism in detail by QM/MM, we performed a potential energy surface (PES) scan of the reaction to obtain a rough estimation of the putative reaction path. We used the self-consistent charge density functional tight binding (SCC-DFTB) employed via DFTB3, a promising approach in treating large biomolecular systems with reasonable accuracy/performance trade-off 38 – 40 . To increase the level of resolution of the resulting energy landscapes, intermediate state geometries were generated to be optimized with the DFT workflow. The PES scans comparing both reactant systems EH (protonated E H6.50 ) and E- (deprotonated). (Fig. 3 k-l) show that the initial reaction proceeds in concert to reach the hydroxy-oxazolidine intermediate T1 (Fig. 1 b). Moreover, our comparison indicates that the N-O acyl shift is energetically favored with protonated E H6.50 . Finally, it suggests that the ester E (Fig. 1 c) is reached in the next step via two subsequent reactions, amide protonation and removal of the amine leaving group (Fig. 3 m-n). To decipher which of the two protonation states of E H6.50 provides the energetically more favorable pathway, we started our QM/MM pathway analysis from the two optimized respective reactant conformations R EH and R E− . In both setups, we observe that H GPS.−2 abstracts a proton from the hydroxyl moiety of T GPS.+1 , resulting in an alkoxide, attacking the peptide carbon between L GPS.−1 and T GPS.+2 in concert with the proton transfer reaction (Fig. 1 a,b, 3 a,f). The resulting tetrahedral intermediate T1 EH /T1 E− is characterized by a five-membered heteroatomic ring and an oxyanion (Fig. 3 b,g) 41 . This hydroxy-oxazolidine is transient due to ring strain and three electronegative binding partners at the former peptide carbon 42 . From both T1 EH and T1 E− , the reaction immediately proceeds via protonation of the ring amide 43 . In serine proteases, this step is mediated by the catalytic histidine imidazole group and was suggested to be rate-limiting 44 . In L1, the histidine imidazole is sterically hindered by the five-membered oxazolidine ring to form a direct interaction to the amidic nitrogen. A water molecule located on the opposite ( trans ) side of the peptide from H GPS.−2 ( cis side) forms a hydrogen bond to the unprotonated amide nitrogen instead. Strikingly, in both T1 EH /T1 E− this water is additionally coordinated by the carboxyl group of E H6.50 assigning it a key-role for these first steps (Ext. Data Figs. 4 b, 5 b). From the two QM/MM systems, the EH system provides the energetically more favorable mechanism, in agreement with the energy landscape from the surface energy scan (Fig. 3 m-n). A water-mediated proton hop occurs from neutral E H6.50 to the amide, leading to charged E H6.50 in TH EH (Fig. 3 b-c). In the E- system, E H6.50 is already charged. As a result, the proton is abstracted from the nearby water, forming a hydroxide. The resulting electrostatic repulsion of multiple negatively charged moieties in proximity significantly increases the energy barrier towards TH E− (Fig. 3 g-h), making this protonation state/reaction path unlikely. The EH system also represents the energetically more favorable configuration during the final step that completes the N-O acyl shift. The ester intermediate E is formed as a result of expulsion of the amidic nitrogen atom in both TH EH /TH E− , (Fig. 1 d-e). At + 37.4 kcal/mol, however, the increase in energy of the ester compared to the reactant is significantly higher in the E- system than in the EH system with + 26.0 kcal/mol (Fig. 3 e,j). E EH energy is consistent with ester intermediates observed in related studies on proteases 45 – 48 . In addition, we see proton hopping in E E− with the hydroxide anion located at the edge of the QM system (Ext. Data Fig. 5 d-e), whereas the water molecule remains coordinated close to the ester bond and the free amine in E EH (Ext. Data Fig. 4 d). GAIN domain cleavage is hampered by E656 H6.50 mutations Our analysis attributes a catalytic role to E H6.50 as a proton donor for the amide protonation step and coordinating water adjacent to the catalytic triad. Interestingly, water accessibility of L GPS.−1 depends on the protonation state of this residue (Ext. Data Fig. 2 d-e). Examining the radial density function around the leucine backbone oxygen (forming the oxyanion in the mechanism, Fig. 1 b), we find significantly more water in one- and two-layer distances in MD simulations of the reactant R with charged E H6.50 (Ext. Data Fig. 2 f-g, Supp. Figure 1 ). Charged E H6.50 reduces solvent-accessibility of the GPS through formation of a salt-bridge with R777 s8s9 , a residue located in the flexible Flap 1 (Supp. Figure 2 , Supp. Table 2) 7 . Notably, R777 does not coordinate to the putative oxyanion of L GPS.−1 , leaving only water to interact with the anionic oxygen in T1 (Supp. Table 3, Ext. Data Fig. 3 ). Of note, the determined configuration leaves water molecules coordinated in the proteolytic site, still sufficient for the reaction to take place. To investigate the influence of E H6.50 on self-cleavage, we mutated it to Ala, Leu, Gln and Lys, and quantified the effect using two sets of experiments. Firstly, using surface expressed full-length rat L1 receptor (Fig. 4 a-b), and secondly, using secreted and purified HormR-GAIN domains of rat L1 (Fig. 4 c-d). To assess the extent of cleavage, determined 48 hours after transfection, the reaction was slowed down by the additional mutation F803Y S10.49 in the secreted ectodomains. The first mutation E656Q substitutes the carboxyl with an amide group. Replacing the proton donor with an H-bond donor preserves the capability of this residue to coordinate water, but prevents the proton hop during the last step of the N-O acyl shift. This mutation retains the highest cleavage activity of the four variants, with minor reduction compared to native E656 in the full-length constructs, but around 40% reduction of cleavage in the secreted ectodomains. The observation that cleavage is not completely abrogated by the loss of E656 as proton donor suggests that coordination of the water molecule for protonation of the leaving amide nitrogen is the most important role of E H6.50 for cleavage. To test this hypothesis, we substituted E H6.50 by Ala and Leu. E656A should abolish the coordination effect of the side chain while still preserving solvent-accessibility of the amide. E656L should have the strongest impact by introducing a hydrophobic bulky sidechain, altering solvation and polarity of the cleavage site. In agreement, the experiments show minor diminishing of cleavage for E656A in the full-length receptor, while cleavage is reduced by > 80% for the secreted ectodomain. For E656L, as expected, we see significant cleavage reduction in both experiments, with 40% and more than 90% reduction in the full-length receptors and secreted GAIN domain, respectively. Finally, we replaced E656 with a lysine to alter the water coordination but allow the nascent hydroxide after amide protonation to be coordinated by the protonated lysine. E656K also causes a significant reduction in cleavage activity, with ~ 30% reduction in the full-length L1 variant and 75% reduction in the full-length ectodomains. Multiple avenues for ester hydrolysis exist for the ester intermediate In the previously proposed cis -autoproteolytic mechanism, H GPS.−2 acts as a base to generate a hydroxide anion for ester hydrolysis 41 . However, H GPS.−2 is protonated in the obtained ester conformation E (Fig. 1 c), suggesting a different mechanism. The E EH path described above suggests that the ester hydrolysis is mediated by a water molecule, coordinated between the ester carbon, the neutral amino group and charged E H6.50 (Ext. Data Fig. 4 e). This water could be deprotonated by either the nearby neutral amine or the carboxylate moiety of E H6.50 , resulting in a hydroxide for nucleophilic attack on the ester carbon (Fig. 1 f-g). With the given low efficiency of GAIN domain autoproteolysis, the ester intermediate, however, may have a sufficiently long lifetime to allow conformational re-arrangements and changes in protonation states of H GPS.−2 . The motility of the ester bond compared to the more rigid planar peptide bond between L GPS.−1 and T GPS.+1 , facilitates rotation and changes of dihedral angles. Together with the accessibility of the titratable residues H GPS.−2 and E H6.50 to bulk solvent, it seems reasonable to assume that several conformations and avenues for hydrolysis may exist. With the goal of identifying the most favorable conformation for hydrolysis, we performed MD simulations of different ester intermediate states. In the various setups, we incorporated a parameterized leucine-threonine ester (LTE) with H GPS.−2 in HSD, HSE and HSP states as well as charged and neutral E H6.50 (denoted as E- and EH, respectively; Ext. Data Fig. 7), yielding six different setups. In the obtained trajectories, we observe distinct orientations of the ester dihedral relative to the histidine labelled cis , neutral and trans (Ext. Data Fig. 8, Supp. Figures 10–14). In uncleaved L1 MD simulations, the peptide bond would always correspond to the neutral state. In LTE MD simulations, for HSE and HSP states, the carbonyl oxygen of the ester forms a hydrogen bond with H GPS.−2 . In this conformation, the ester plane is exposed to water for ( trans -side) hydroxide attack. In the trans conformation, the carbonyl group of the ester points away from the histidine and is coordinated by water, so that the ester plane faces to H GPS.−2 for possible ( cis -side) hydroxide attack. With the goal of finding suitable geometries for the water attack in the MD trajectories, we examined solvent-accessible surface area (SASA) and water geometries around the electrophilic ester carbon in the six setups. For LTE SASA, we observe a drop to < 0.1 nm² in the first 250 ns in charged E- systems, while this value fluctuates between 0-0.5 nm 2 in the neutral EH systems (Fig. 5 . a-b, Supp. Figure 14). When we analyzed waters within 4.0 Å and within a 15° cone of the ideal Bürgi-Dunitz angle of 107° relative to the ester plane (Fig. 5 c-e) 49 , we find a clear preference for the trans side in both. From three independent trajectories with 10,000 frames each (1,000 ns with 100 ps trajectory time step), we find 473 frames (1.58%) in trans to 52 frames in cis (0.17%) in the EH system and with 353 frames in trans (1.17%) to 1 frame in cis (0.003%) in the E- system. To examine the best attack geometry, we extracted the four best frames for each recombination of high and low SASA states and cis / trans attack geometry and created a separate QM/MM system for each of them. All systems were minimized with the water restrained to 3.00 Å to the ester carbon before we calculated the natural bond orbital (NBO) charges as partial atomic charges to estimate the electrostatic forces for nucleophilic attack 50 , 51 . Stronger electrostatic attraction between the carbon and the water oxygen would indicate a more favorable geometry for hydrolysis. Notably, the two cis- frames originate from trajectories with the HSD state, where the histidine would act as a base for generating the hydroxide for hydrolysis as suggested previously and observed for hydrolases 41 , 52 . Both trans -frames originate from HSP/HSE trajectories, where either E H6.50 or the R-NH 2 amine could activate the water via base catalysis, corresponding to the E EH path. Trans- attack proceeds differently from the proposed mechanism, with the initial base H GPS.−2 remaining protonated and coordinating the ester oxygen and the oxyanion after water attack. To further assess the reactivity potential of the hydrolytic configurations, we examined charge separation between the reacting water molecule and the electrophilic ester carbon. Charge separation is a key indicator of polarization and potential nucleophilic attack, with a larger difference between partial charges reflecting a more reactive state. To quantify this, we computed Natural Bond Orbital (NBO) charges 51 , which provide chemically intuitive estimates of atomic partial charges based on the electron density distribution. Notably, the two cis -frames originate from trajectories using the HSD protonation state, where histidine acts as a general base, abstracting a proton from water and generating a hydroxide capable of nucleophilic attack, as previously suggested 41 , 53 . In contrast, trans -frames stem from HSP/HSE trajectories, where either E H6.50 or the R-NH₂ amine group can abstract the proton from water, corresponding to the E EH pathway. Interestingly, trans -frames exhibited slightly higher charge separation: +0.892/+0.904 on the ester carbon and − 0.985/-1.003 on the water oxygen in the open and closed GPS states, respectively (Fig. 5 g–h). In comparison, cis conformations showed + 0.878/+0.871 on the carbon and − 0.976/-0.967 on the water oxygen for the open/closed configurations (Fig. 5 i–j), respectively. These values suggest a slightly enhanced polarization in the trans -pathway, which may facilitate nucleophilic attack during hydrolysis. Despite harmonic restraints of 1000 kJ mol − 1 nm − 2 applied to keep the oxygen-carbon distance at 3.00 Å, we observe strong attraction between both atoms in both trans systems, resulting in 2.85 Å and 2.84 Å for trans -open and trans -closed conformations, while in cis -open and cis -closed conformations, distances remained at 3.00 Å and 3.02 Å. Thus, the attractive electrostatic force is higher in trans systems (–33.4 pN for trans -closed, -32.1 pN for trans -open; -27.4 pN for cis -closed, -27.9 pN for cis -open). This higher electrostatic attraction indicates that the trans attack, corresponding to the reaction path taken in the E EH QM/MM system, is the preferable path for hydrolysis (Ext. Data Table 1). Ester hydrolysis can proceed in two pathways with different product energies To examine the ester hydrolysis on the quantum-mechanical level, we employed the EH QM/MM system, that corresponds to the favored trans -attack hydrolysis. Since this system is also the natural product of the N-O acyl shift, the entire mechanism can be described as a single trajectory starting from a protonated E H6.50 reactant geometry. By optimizing the geometry, a hydrogen bond was formed between the ester oxygen and the now cationic H GPS.−2 , resulting in the flipped ester E f with a relative energy of + 27.0 kcal/mol to the reactant R, 1.0 kcal/mol higher than the original E EH (Fig. 1 d). With this configuration, a water molecule is coordinated between the ester carbon, the neutral amino group and charged E H6.50 (Ext. Data Fig. 4 e). During the chemical reaction, the water is first deprotonated before the hydroxide carries out a nucleophilic attack on the ester carbon (Fig. 1 f-g). Both the amine and the acid moiety of E H6.50 are potential bases for the hydrolysis step, resulting in potentially two different pathways. In the amine pathway (denoted with N), the amine acts as a base and the nucleophilic attack proceeds in concert with the proton transfer from water, resulting in the tetrahedral intermediate T2 N (Fig. 6 b). Hydrolysis takes place via the alkoxide, which acts as a leaving group and receives a proton from the positively charged amino group, resulting in a product with neutral termini P Nn (Fig. 6 c,e). This product in turn receives a proton from the nascent C-terminal carboxylic acid group, resulting in a product P N with a relative energy of -3.0 kcal/mol, lower than the reactant (Fig. 6 d-e). In the acid (A) pathway, the water transfers a proton (back) to the negatively charged E H6.50 . The resulting hydroxide attacks the ester carbon, yielding the tetrahedral intermediate T2 A (Fig. 6 f). Hydrolysis proceeds analogously to the amino route, however now with an uncharged amino group. In restrained minimization, both unsubstituted oxygens of the tetrahedral intermediate T2 A coordinate a proton through a hydrogen bond; a hydrogen from the original carbonyl-oxygen with H GPS.−2 and another hydrogen bond from the original hydroxide to the amino group, respectively (Ext. Data Fig. 4 g). During the chemical reaction, the neutral amine first completely abstracts the hydrogen bonded proton (Fig. 5 g), before it transfers a proton to the leaving alkoxide group. With a neutral N-terminal amine and a neutral E H6.50 , the resulting product P A had with − 16.3 kcal/mol the lowest energy of all states (Fig. 6 h,j, Ext. Data Fig. 6 ). Notably, inducing charge separation by simulating water-mediated proton transfers clearly increased the energy to -6 kcal/mol in P Ap , highlighting the impact of the solvated and polar GPS triad environment (Fig. 6 i-j). The obtained product conformations were checked for agreement with the conformational space of the cleaved L1 crystal structure (PDB: 4DLQ) 8 employing MD simulations. Principal component analyses of the whole GAIN and the GPS environment showed good agreement of the non-hydrogen coordinates, indicating a continuum between the obtained product states and available structural data (Supp. Figures 17–18). Discussion Cis -autoproteolytic proteins perform a unique self-cleavage reaction only once in their lifetime, distinguishing them from classical enzymes like serine proteases, which are evolutionarily optimized for rapid and repeated catalysis 46 , 48 , 54 . This one-time reaction found e. g. in aGPCRs, inteins or glycosylasparaginase 55 , 56 (Supp. Table 4), combined with presumably high structural flexibility of the cleavage site creates challenges for detailed mechanistic studies. Contrary to enzymatic studies, structures with a bound transition state analogue are not readily available for autoproteolytic proteins. These differences make it challenging to study the mechanism of cis -autoproteolysis using conventional workflows developed for enzymes with high catalytic rates and optimized reaction geometries. To address these challenges, we employed a multiscale computational approach that integrates molecular dynamics (MD) simulations with hybrid quantum mechanics/molecular mechanics (QM/MM) calculations. MD simulations allowed us to explore the flexible structural intermediates of the GAIN domain, the hallmark domain of aGPCRs 4 , 5 . QM/MM methods provided the detailed electronic and energetic description of the chemical steps involved in the cleavage 27 , 28 , 57 , 58 . With this approach, we propose a more detailed perspective on the cleavage mechanism by aGPCRs influenced by conserved residues and water adjacent to the HL|T catalytic triad of the GAIN domain that coordinate solvent accessibility and co-determine the chemical reaction. This can be extended to assess cleavability via evaluating all human GAIN domain models for presence of cleavage determinants (Supp. Table 5). While the mechanism of cis -autoproteolysis has not so far been structurally elucidated in detail, a basic residue has been proposed to initiate the cleavage reaction not only in the GAIN domain, but also in Nucleoporin98, the panthetine hydrolase ThnT and the zona occludens-1 (ZO-1) protein family 56 , 59 – 61 (Supp. Table 4). By contrast, in glycosylasparaginase or cephalosporin acylase 55 , 59 , 62 , an acidic residue serves as the initial base. Generally, mechanistic insights on cis -autoproteolytic proteins are lacking, with the exception of the SEA domain, where an acid-catalysis of the N-O acyl shift is implied 42 , 43 , 63 . Our results reveal that the cleavage reaction in the GAIN domain of L1 starts with a proton transfer from the threonine residue (T GPS.+1 ) at the cleavage site to a nearby histidine residue (H GPS.−2 ). This proton transfer is stabilized by a T-shaped π–π interaction of H GPS.−2 with a phenylalanine residue (F S10.49 ), which reduces the configurational entropy of the cleavage site and increases proton affinity 26 , 64 , thus preparing the system for catalysis. In a concerted mechanism with proton transfer, the nucleophilic threonine alkoxide attacks the adjacent peptide bond, leading to the formation of a five-membered oxazolidine intermediate T1 (Fig. 3 a,b). This mechanism differs from that of serine proteases, where the histidine protonates the amide nitrogen after nucleophilic attack by serine. In the GAIN domain, the five-membered ring sterically blocks this proton transfer. Instead, a water molecule on the opposite ( trans- ) side of the histidine donates the proton to the amide nitrogen. This water molecule is coordinated by a conserved glutamate residue (E656 H6.50 ), which facilitates a proton hop by quickly regaining a proton, maintaining the catalytic cycle (Ext. Data Fig. 10). This is only possible with neutral E656 H6.50 , while electrostatic repulsion makes a pathway with charged E656 H6.50 less likely. The shift towards the ester is eventually completed, when the ring opens with the ring nitrogen acting as leaving group. A similar catalytic mechanism for cis -autoproteolysis has been proposed in a structural study on the pantetheine hydrolase ThnT 59 , where a threonine forms the alkoxide in the first step of the reaction. On the trans- side opposite the initial nucleophilic attack, two aspartates were suggested to coordinate a water molecule with one in a neutral state facilitating a proton hop for amide protonation. This suggests a conserved chemical reaction theme across different cis -autoproteolytic proteins involving nucleophilic attack by a hydroxyl group and amide protonation via proton hop mediated by acidic residues and water. Mutation of the acidic residue E656 H6.50 indeed shows that this conserved residue is functionally important for cleavage in aGPCRs (Fig. 4 ). The observation that E656Q and E656A do only have a minor effect on cleavage competence compared to E656L, assign a superior role of cleavage competence to water coordination, which is presumably still possible with the former variants. This residue together with the flexible flaps surrounding the cleavage site modulates solvent accessibility, influencing cleavage efficiency. Swapping flap regions between different aGPCRs alters cleavage rates, support the concept of these dynamic protein regions helping to regulate catalysis by controlling water access 7 , 26 . Water in the trans- pocket, coordinated by E H6.50 during the N–O acyl shift, also plays a critical role in the subsequent ester hydrolysis. Unlike serine proteases, where the initiating histidine base activates water for nucleophilic attack, the GAIN domain uses E H6.50 to abstract a proton from water, enabling it to attack the ester bond. This supports a more general catalytic function of E H6.50 in cis -autoproteolysis. An alternative mechanism cannot be excluded, where the amine group serves as the proton abstracting base— a mechanism observed in a study on spontaneous hydrolysis of a glycylserine dipeptide, where water is coordinated by two amino groups and activated through simultaneous proton transfer 45 . Remarkably, in both proposed reaction routes, the leaving alkoxide group is protonated by the same amine group, matching the mechanism proposed for GPS cleavage. Notably, the energy profile of glycylserine hydrolysis shows a maximum transition energy state of 29.4 kcal/mol, with water addition to the ester and serine R–OH attack as rate-limiting steps - values similar to the GAIN domain system 45 . Regardless of mechanistic alternative, the negative partial charges surrounding the water promote proton abstraction, generating the hydroxyl for ester hydrolysis. The two mechanistic routes for ester hydrolysis can directly proceed from the ester geometry obtained by quantum-mechanical modeling of the N–O acyl shift. This supports the idea that cis -autoproteolysis may proceed as a single, continuous reaction pathway. However, the conformational flexibility of the ester intermediate could allow for alternate hydrolysis paths (Fig. 5 ) or simply reduce the reaction rate by increasing conformational entropy and obscuring reactive conformations. Analogously, the uncleaved GAIN domain exhibits a similar effect, with the reactive conformation R EH /R E− representing a small subset of observed conformational space. Thus, assessing reaction speed solely from reaction energies falls short of including the entropic effect of the dynamic GAIN domain on the cleavage rate. In conclusion, our study provides detailed mechanistic insights into cis -autoproteolysis, highlighting how conserved residues and structured water networks enable a chemically precise reaction within an inherently flexible system. These findings deepen our understanding of aGPCR processing and may apply broadly to other autoproteolytic proteins, offering a framework to explore similarly elusive one-time cleavage events in biology. Methods Computational Methods Creation and MD of the uncleaved ADGRL1 Model Based on the existing cleaved GAIN + HormR domain structure of rat ADGRL1 (PDB ID: 4DLQ), the geometry of the backbone of the uncleaved ADGRB3 GAIN structure was aimed to be achieved in the uncleaved ADGRL1 model 8 .First, the T838 GPS.+1 residue was rotated along the C-Cα bond to better face the C-terminal carboxyl group of L837 GPS.−1 . The OT2 atom of L GPS.−1 was deleted and a bond introduced between T838 GPS.+1 :N and L837 GPS.‑1 :C using the Builder utility in PyMol. The OT1 atom was re-named to O and the topology fed into CHARMM-GUI to automatically normalize bond lengths and generate minimization and equilibration inputs using the CHARMM36 forcefield and GROMACS version 2020.2 65–68 . After minimization using the steepest-descent method for 5,000 steps, a 125,000 step equilibration with 1 fs timestep was performed to yield the equilibrated uncleaved model. After comparing the backbone φ and ψ angles to ADGRB3 and the measuring the distance of H836 GPS.−2 side chain nitrogens to the target T838 GPS.+1 O γ of 4.4 Å, in two separate systems for the HSE and HSD state, the respective basic nitrogen atom and the oxygen were pulled together by using biased MD after 5 ns of equilibration with a time step of 1 fs, applying an harmonic umbrella potential on both atoms with a force constant of 1000 kJ/mol*nm² and a pull rate of -0.001 nm/ns until the run terminated due to a low-distance warning (HSD: 510 ps, HSE: 650 ps). With the resulting configurations, an equilibration cascade with decreasing harmonic potential holding the N δ/ε ‑O γ distance constant for 100 ns with a force constant of 1000, 500, 200 and 100 kJ/mol*nm², respectively, in a decreasing order, while simultaneously applying backbone, sidechain and dihedral restraints of 400, 40 and 4 kJ/mol*nm² or kJ/mol*deg, respectively, to the protein. After equilibration, we performed triplicate unbiased MD simulations for 1,500 ns using the CHARMM36 force field in GROMACS. Reactant pose selection The resulting trajectories were concatenated for HSD and HSE, respectively, and clustered with Principal Component Analysis (PCA) using the scikit-learn package in Python 3.9 69 . The distances in the unbiased MD between the H GPS.−2 nitrogen atoms and the T GPS.+1 O γ were analyzed to find a potential occupation of the hydrogen-bond distance between the respective unprotonated nitrogen and the oxygen. Based on the distance data, the HSE state was selected and the centroid of the cluster with the lowest N δ ‑O γ distance extracted (Ext. Data Fig. 9a-d). The H GPS.−2 and F S10.49 π-π-interaction was stable for both HSE runs (Ext. Data Fig. 2 j-k), in contrast, three MD systems (HSD-E, HSD-EH, HSP-EH) showed fluctuations in this interaction (Supp. Figures 8–9). While we initially hypothesized R777 to coordinate the oxyanion of the tetrahedral intermediate during the mechanism, the simultaneous low-distance between the guanidyl moiety of R777 to the L GPS.−1 peptide oxygen and H GPS.−2 N δ to T GPS.+1 Oγ did not coincide (Supp Fig. 2 ). We selected the centroid of cluster 2 of the PCA-clustered trajectory as a reasonable reactant with a favorable hydrogen bond configuration between the triad histidine and threonine and ran geometry optimization with the B3-LYP 70 – 72 functional and the Weigend and Ahlrichs def2-SVP basis set followed by geometry optimization with the def2-TZVP basis set (for detailed QM/MM methods, see below; Ext. Data Fig. 9d-e) 73 . We duplicated the system, adding an additional proton to the E656 H6.50 residue with identical conformation. From there, for the protonated E656 H6.50 system, we traced the pathway along R-E-P and in reverse P-E-R to generate a reverse reactant for validation of the initial reactant R. The validation pathway was generated via DFTB3 scans and intermediates were optimized using the B3-LYP functional and the def2-SVP basis set. (Ext. Data Fig. 9f). While the QM/MM poses of the reactant and the reversed reactant did not significantly differ (Ext. Data Fig. 9g), we selected the reversed reactant as R for the pathway tracing due to more favorable water coordination. QM/MM selection The explicit QM selection comprised 146 or 147 (proton at E656) atoms including 8 link atoms with a total of 16 molecules. Link atoms were used for treating the QM/MM boundary, breaking C-C bonds, therefore some backbone atoms of G819, N835 and A769 were included in the QM selection (full list of QM selection in Supp. Data 1). Furthermore, our reactant frame was bulk-solvent accessible, so we included 11 water molecules covering the GPS triad towards Flap 1 and bulk solvent as well as forming a polar network between L GPS.−1 , T GPS.+1 E H6.50 and S S8.54 (Fig. 2 d). Reaction coordinate scans with semi-empirical DFTB3 Reactions were simulated by employing multidimensional coordinate scans with DFTB3 using CHARMM46. We used the previously defined QM/MM system and for each reaction step (R-T1, T1-E, E-T2, T2-P), the changing interatomic distances were scanned to reach reasonable estimated distances for each intermediate, respectively. For each point in the one- to three-dimensional conformational space, a geometry optimization was performed with DFTB3. Upon meeting energy gradient thresholds, the energy and conformation was stored along with the current interatomic distances. For example, the R EH -T1 EH scan was performed on the distances between H GPS.−2 :N δ – T GPS.+1 :Hγ (2.10 Å − 1.0 Å) and T GPS.+1 :Oγ – L GPS.−1 :C (2.9 Å − 1.5 Å) with 0.1 Å spacing and distances Å, respectively. Low-energy intermediates were extracted for geometry optimization with DFT QM/MM. DFT QM/MM for Generating Intermediate, Product States and coordinate scans QM/MM DFT was performed with TURBOMOLE 7.4 74–77 and CHARMM46b2 78 in conjunction with the CHARMM/TURBOMOLE interface 79 . During setup, initial molecular orbitals were guessed by the Extended Hückel method 80 and the corresponding system charge of 0 (neutral glutamate, EH) or -1 (charged glutamate, E-). Geometry minimizations were performed by first applying the Weigend and Ahlrichs basis set def2-SVP 81 with the B3-LYP functional 70 – 72 followed by geometry optimization/minimization with def2-TZVP 73 and the B3-LYP functional; all employing the subtractive QM/MM scheme and link atoms placed on peptide backbone C-C bonds. Coordinate scans followed a similar system, however the semi-empirical DFTB3 was used for QM treatment 39 . The MM system was treated with the CHARMM36 force field with a flexible sphere of all residues with at least one atom within 5.0 Å of the QM selections and all atoms outside the sphere fixed. Self-consistent field (SCF) threshold was set to 10 − 7 and we used the spherical grid of size m3. Due to the ambiguity of the E656 H6.50 protonation state, two separate QM/MM systems were generated based on both states. By following the initial proposed reaction path, one-, two- or multidimensional pathways were generated by incrementally moving the atoms involved in the respective step by intervals of 0.1–0.2 Å per iteration, applying distance restraints with force constants of 1000 kJ/mol/nm² and minimizing the structure iteratively until reaching the next reaction state. First, from the reactant conformation R the hydroxide moiety H γ of T838 GPS.+1 was moved towards H836 GPS.−2 , while simultaneously the distance between T838 GPS.+1 O γ and L837 GPS.−1 peptide carbon was decreased to facilitate the tetrahedral hydroxyl-oxazolidine intermediate T1 following nucleophilic attack on the peptide bond. The next step consisted of proton addition to the T838 GPS.+1 nitrogen from a nearby water molecule within the five-membered ring, followed by distance increase between T838 GPS.+1 nitrogen and L837 GPS.−1 peptide carbon to complete of the N-O acyl shift, resulting in the ester intermediate E. The resulting hydroxyl anion deprotonated the E656 sidechain in the systems containing protonated E656 H6.50 . The water attack on the ester was facilitated by either moving the free hydroxide ion directly towards the L837 GPS.−1 ester carbon or moving the intact water towards the attack, resulting in a second tetrahedral intermediate T2, with a proton transferred to the free neighboring amino moiety if water was attacking. During T2 minimization – first with restraint of the attacking water oxygen to the carbon atom and subsequently without restraints, various proton transfers occurred and the final products P were generated. Ester intermediate parameterization The leucine-threonine ester (LTE) molecule was extracted and minimized. Link atoms (peptide bond) were simply capped by uncharged hydrogens. Charges and parameters were taken from the respective amino acids in accordance with the CHARMM36 force field parameters where appropriate, leaving the ester bond and its immediate neighbors to parameterize. Point charges, bonds and angles were parameterized using the force field Toolkit (ffTk) 82 plugin in VMD 1.9.4alpha 83 coupled with the individual QM runs performed by ORCA 5.0.3 84 . Upon occurring steric clashes, the torsion intervals were reduced for the generation of dihedral parameters for the dihedrals involved in the ester bond. The resulting parameter files were modified for CHARMM to detect the engineered residue as an artificial amino acid with the additional caveat of not being an alpha-amino acid within the backbone. Ester intermediate MD With the LTE topology and parameter files, ADGRL1-GAIN-LTE systems were set-up by taking the generated ester intermediate frames minimized via DFT on the B3-LYP/def2-TZVP level and renaming the atoms involved in the new LTE residue according to the topology file via python script. MD systems were then generated using CHARMM-GUI 65 , 85 with identical parameters to the uncleaved ADGRL1-GAIN model systems. Simulations were run in triplicate for 1,000 ns with H836 − 2 as HSD, HSE and HSP, with E656 protonated and unprotonated, respectively, using the NAMD2 package on GPU 86 . Biochemical Methods Plasmids An N-terminally HA-tagged and C-terminally FLAG-tagged rat LPHN1/ADGRL1 cDNA (gift from Simone Prömel) was subcloned into pcDNA3.1 by sequence-ligation-independent cloning with the following primers: 5’-ACTCACTATAGGGAGACCCAAGCTTGCCACCATGGCCCGCTTG and 5’-AGTGTGATGGATATCTGCAGAATTCTTACTTATCGTCGTCATCCTTGTAATCACC. GAIN domain cleavage-deficient H838S mutant, and all E656 mutants were generated by Genscript. Expression analyses of receptors HEK293T cells were ­maintained, transfected and lysed as previously described. Lysates were mixed with Laemmli buffer, and directly subjected to SDS-PAGE and Western blotting. Receptor expressions were detected by first incubating the nitrocellulose membrane overnight at 4 o C with rabbit anti-HA primary antibody (1:1000, Cell Signaling, Cat# 3724) or anti-TwinStrepII® primary antibody (1:500, iba Life Sciences, Cat# 2-1507-001), then a 1-hour incubation in room temperature with IRDye® 680RD goat anti-rabbit secondary antibody (1:15000, Licor, Cat# 926-68071). Detection of β-tubulin was done by similar incubation approach, with mouse anti-β-tubulin primary antibody (1:5000, DSHB, Clone E7) and IRDye® 800CW goat anti-mouse secondary antibody (1:15000, Licor, Cat# 926-32210). Visualizations of the immunosignals were done by the Licor Odyssey® XF Dual-Mode Imaging System. Molecular cloning and protein expression of ADGRL1 GAIN domain constructs A DNA fragment encoding residues 460–849 of the rat ADGRL1 ECR was synthesized by Thermo Fisher GeneArt. Point mutation of the r L1 gene was performed on the pMA vector (ThermoFisher Scientific) by Quik Change mutagenesis . The HormR and GAIN domains of the wild-type and rat L1 point mutants, comprising residues 460–849, were subcloned into the AgeI–KpnI site of the pHLsec vector via PCR amplification. A C-terminal Fc tag followed by a hexahistidine sequence was added to increase the mass of the C-terminal fragment generated by GPS cleavage. HEK293T cells were transfected with polyethyleneimine (PEI) at a mass concentration 1.5 times higher than that of the pHLsec expression plasmid. Expression was performed in 10 ml volume for two days at 37°C in a humid atmosphere containing 5% CO 2 . Protein purification and analysis Elution buffer (50 mM Tris-HCl, pH 8; 300 mM NaCl; 500 mM imidazole) was added to the harvested expression medium yielding a final imidazole concentration of 30 mM. Then, 100 µl of washed His-bead slurry (Cat. No. 30410, Qiagen) was added after which the samples were incubated for one hour at 4°C with shaking. The beads were collected by centrifugation at 10000 ×g. The beads were then washed four times with 1 ml of wash buffer (50 mM Tris-HCl, pH 8; 300 mM NaCl; 30 mM imidazole) and the proteins were eluted by adding 250 µl of elution buffer using an incubation time of two minutes at room temperature. After centrifugation the supernatant was collected. 20 µl of the samples were denatured by adding 5 µl of SDS sample buffer (250 mM Tris-HCl, pH 6.8; 50% (v/v) glycerol; 4% (w/v) SDS; 25% (v/v) β-mercaptoethanol; 0.25% (w/v) bromophenol blue) and boiling at 95°C for 600 seconds. The SDS-PAGE gels were analyzed by semi-dry western blotting on methanol-activated PVDF membranes (Cat. No. 10600023, Cytiva Amersham). After skimmed milk blocking and incubation with anti-His-HRP antibodies (Cat.No. 11965085001, Roche), the proteins were detected using a luminol-peroxidase mixture (Cat.No. 10308449, Cytiva Amersham). Chemiluminescence was measured using a Biostep Celvin® S 420 chemiluminescense imager and intensities were scaled using the program ImageJ to a maximum pixel saturation of 65535. The resulting band intensities were quantified using GelAnalyzer software (version 23.1.1, available at www.gelanalyzer.com from Istvan Lazar Jr., PhD and Istvan Lazar Sr., PhD, CSc) using the valley-to-valley analysis method. The plot and the statistical depiction were done by using GraphPad Prism (version 10.4.1 for Windows, GraphPad Software, Boston, Massachusetts USA). Declarations Acknowledgements This work was funded by the Deutsche Forschungsgemeinschaft (DFG) through CRC 1423, project number 421152132 (project A06 to T.L. and N.S., projects C01 and Z04 to P.W.H.). The authors gratefully acknowledge the scientific support and HPC resources provided by the Erlangen National High Performance Computing Center (NHR@FAU) of the Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU) under the NHR project p101ae. NHR funding is provided by federal and Bavarian state authorities. NHR@FAU hardware is partially funded by the German Research Foundation (DFG) – 440719683. We thank Paolo Carloni and Emiliano Ippoliti from Forschungszentrum Jülich GmbH for their support and access to their HPC and QM/MM software infrastructure. We like to acknowledge Ville Kaila and group for their QM/MM tutorial and the interface python script, which we modified to be compatible with Python 3. Author contributions F.S.: Conceptualization; Methodology; Validation; Investigation(computational); Software; Formal Analysis; Visualization; Writing – original draft; Writing – review and editing; Data Curation Y.K.C.: Investigation(biochemical); Visualization R.S.: Investigation(biochemical); Visualization F.P.: Investigation(biochemical) N.S.: Conceptualization; Writing – review and editing P.W.H.: Conceptualization; Validation; Writing – review and editing; Supervision; Project administration; Funding acquisition T.L.: Conceptualization; Writing – review and editing H.B.: Investigation(computational); Writing – review and editing; Data Availability The generated GAIN domain models generated in this study, QM/MM input and output file, MD models and trajectories, LTE MD parameters and additional data have been deposited in the online repository zenodo under accession code DOI: 10.5281/zenodo.14445825. The PDB entries used in this study are available under the following accession codes: 4DLQ, 4DLO, 1MCT. The original template Python script for TURBOMOLE execution is available under https://villekaila.com/news/. The structural models reviewed are available in the alphafold database under: P15941, Q5T601, Q8IZF2. The entries of proteins reviewed are available in the UniProt knowledge base under: O88917, P15941, Q5T601, Q8IZF2, O60242. References Seufert, F., Chung, Y. K., Hildebrand, P. W. & Langenhan, T. 7TM domain structures of adhesion GPCRs: what’s new and what’s missing? Trends in Biochemical Sciences 48 , 726–739 (2023). Rosa, M., Noel, T., Harris, M. & Ladds, G. Emerging roles of adhesion G protein-coupled receptors. Biochem. Soc. Trans. 49 , 1695–1709 (2021). Bondarev, A. D. et al. Opportunities and challenges for drug discovery in modulating Adhesion G protein-coupled receptor (GPCR) functions. Expert Opin. Drug Discov. 15 , 1291–1307 (2020). Morgan, R. K. et al. The expanding functional roles and signaling mechanisms of adhesion g protein–coupled receptors. Annals of the New York Academy of Sciences 1456 , (2019). Langenhan, T., Piao, X. & Monk, K. R. Adhesion G protein-coupled receptors in nervous system development and disease. Nature Reviews Neuroscience 17 , 550–561 (2016). Seufert, F. et al. Generic residue numbering of the GAIN domain of adhesion GPCRs. Nat. Commun. 16 , 246 (2025). Beliu, G. et al. Tethered agonist exposure in intact adhesion/class B2 GPCRs through intrinsic structural flexibility of the GAIN domain. Molecular Cell 81 , 905-921.e5 (2021). Araç, D. et al. A novel evolutionarily conserved domain of cell-adhesion GPCRs mediates autoproteolysis. The EMBO Journal 31 , 1364–1378 (2012). Nieberler, M., Kittel, R. J., Petrenko, A. G., Lin, H.-H. & Langenhan, T. Control of Adhesion GPCR Function Through Proteolytic Processing. Handbook of Experimental Pharmacology 234 , 83–109 (2016). Liebscher, I. et al. A Tethered Agonist within the Ectodomain Activates the Adhesion G Protein-Coupled Receptors GPR126 and GPR133. Cell Reports 9 , 2018–2026 (2014). Nazarko, O. et al. A Comprehensive Mutagenesis Screen of the Adhesion GPCR Latrophilin-1/ADGRL1. iScience 3 , 264–278 (2018). Stoveken, H. M., Hajduczok, A. G., Xu, L. & Tall, G. G. Adhesion G protein-coupled receptors are activated by exposure of a cryptic tethered agonist. Proc. Natl. Acad. Sci. 112 , 6194–6199 (2015). Wilde, C. et al. The constitutive activity of the adhesion GPCR GPR114/ADGRG5 is mediated by its tethered agonist. FASEB J. 30 , 666–673 (2016). Frenster, J. D. et al. Functional impact of intramolecular cleavage and dissociation of adhesion G protein–coupled receptor GPR133 (ADGRD1) on canonical signaling. J. Biol. Chem. 296 , 100798 (2021). Araç, D. et al. Dissecting signaling and functions of adhesion G protein-coupled receptors. Annals of the New York Academy of Sciences 1276 , 1–25 (2012). Scholz, N. et al. Molecular sensing of mechano- and ligand-dependent adhesion GPCR dissociation. Nature 615 , 945–953 (2023). Vizurraga, A., Adhikari, R., Yeung, J., Yu, M. & Tall, G. G. Mechanisms of adhesion G protein–coupled receptor activation. Barros-Álvarez, X. et al. The tethered peptide activation mechanism of adhesion GPCRs. Nature 604 , 757–762 (2022). Koh, J. T. et al. Extracellular fragment of brain-specific angiogenesis inhibitor 1 suppresses endothelial cell proliferation by blocking α vβ5 integrin. Experimental Cell Research 294 , 172–184 (2004). Vallon, M. & Essler, M. Proteolytically Processed Soluble Tumor Endothelial Marker (TEM) 5 Mediates Endothelial Cell Survival during Angiogenesis by Linking Integrin αvβ3 to Glycosaminoglycans *. Journal of Biological Chemistry 281 , 34179–34188 (2006). Lehmann, L., Groß, V. E., Behlendorf, R. & Prömel, S. The N terminus-only function of adhesion GPCRs: emerging concepts. Trends Pharmacol. Sci. (2025) doi:10.1016/j.tips.2025.01.004. Kleinau, G. et al. Intramolecular activity regulation of adhesion GPCRs in light of recent structural and evolutionary information. Pharmacol. Res. 197 , 106971 (2023). Scholz, N. et al. Mechano-dependent signaling by latrophilin/CIRL quenches cAMP in proprioceptive neurons. eLife 6 , (2017). Kishore, A. & Hall, R. A. Versatile Signaling Activity of Adhesion GPCRs. Handbook of experimental pharmacology 234 , 127–146 (2016). Liebscher, I., Schöneberg, T. & Thor, D. Stachel-mediated activation of adhesion G protein-coupled receptors: insights from cryo-EM studies. Signal transduction and targeted therapy 7 , 227 (2022). Pohl, F. et al. Structural basis of GAIN domain autoproteolysis and cleavage-resistance in the adhesion G-protein coupled receptors. bioRxiv (2023) doi:10.1101/2023.03.12.532270. Senn, H. M. & Thiel, W. QM/MM Methods for Biomolecular Systems. Angew. Chem. Int. Ed. 48 , 1198–1229 (2009). Senn, H. M. & Thiel, W. QM/MM studies of enzymes. Curr. Opin. Chem. Biol. 11 , 182–187 (2007). Ásgeirsson, B. & Cekan, P. Microscopic rate‐constants for substrate binding and acylation in cold‐adaptation of trypsin I from Atlantic cod. FEBS Lett. 580 , 4639–4644 (2006). Fu, C. et al. Unveiling Mechanical Activation: GAIN Domain Unfolding and Dissociation in Adhesion GPCRs. Nano Lett. 23 , 9179–9186 (2023). Kordon, S. P. et al. Conformational coupling between extracellular and transmembrane domains modulates holo-adhesion GPCR function. Nat. Commun. 15 , 10545 (2024). Matúš, D., Lopez, J. M., Sando, R. C. & Südhof, T. C. Essential Role of Latrophilin-1 Adhesion GPCR Nanoclusters in Inhibitory Synapses. J. Neurosci. 44 , e1978232024 (2024). Chhabra, K. H. et al. ADGRL1 is a glucose receptor involved in mediating energy and glucose homeostasis. Diabetologia 67 , 170–189 (2024). Dietzsch, A. N. et al. Dysfunction of the adhesion G protein-coupled receptor latrophilin 1 (ADGRL1/LPHN1) increases the risk of obesity. Signal Transduct. Target. Ther. 9 , 103 (2024). Ichtchenko, K. et al. A novel ubiquitously expressed α-latrotoxin receptor is a member of the CIRL family of G-protein-coupled receptors. Journal of Biological Chemistry 274 , 5491–5498 (1999). Krasnoperov, V. et al. Dissociation of the Subunits of the calcium-independent receptor of α-Latrotoxin as a result of two-step proteolysis. Biochemistry 48 , 3230–3238 (2009). Olsson, M. H. M., Søndergaard, C. R., Rostkowski, M. & Jensen, J. H. PROPKA3: Consistent Treatment of Internal and Surface Residues in Empirical p K a Predictions. Journal of Chemical Theory and Computation 7 , 525–537 (2011). König, P. H., Hoffmann, M., Frauenheim, Th. & Cui, Q. A Critical Evaluation of Different QM/MM Frontier Treatments with SCC-DFTB as the QM Method. J. Phys. Chem. B 109 , 9082–9095 (2005). Gaus, M., Cui, Q. & Elstner, M. DFTB3: Extension of the Self-Consistent-Charge Density-Functional Tight-Binding Method (SCC-DFTB). Journal of Chemical Theory and Computation 7 , 931–948 (2011). Spiegelman, F. et al. Density-functional tight-binding: basic concepts and applications to molecules and clusters. Adv. Phys.: X 5 , 1710252 (2020). Lin, H. H. et al. Autocatalytic cleavage of the EMR2 receptor occurs at a conserved G protein-coupled receptor proteolytic site motif. Journal of Biological Chemistry 279 , 31823–31832 (2004). Johansson, D. G. A. et al. Protein autoproteolysis: Conformational strain linked to the rate of peptide cleavage by the pH dependence of the N → O acyl shift reaction. Journal of the American Chemical Society 131 , 9475–9477 (2009). Johansson, D. G. A., Macao, B., Sandberg, A. & Härd, T. SEA Domain Autoproteolysis Accelerated by Conformational Strain: Mechanistic Aspects. Journal of Molecular Biology 377 , 1130–1143 (2008). Díaz-Cervantes, E., Robles, J., Solà, M. & Swart, M. The peptide bond rupture mechanism in the serine proteases: an in silico study based on sequential scale models. Phys. Chem. Chem. Phys. 25 , 8043–8049 (2023). Mihaylov, T. T., Parac‐Vogt, T. N. & Pierloot, K. A Mechanistic Study of the Spontaneous Hydrolysis of Glycylserine as the Simplest Model for Protein Self‐Cleavage. Chem. A Eur. J. 20 , 456–466 (2014). Elsässer, B. & Goettig, P. Molecular Sciences Mechanisms of Proteolytic Enzymes and Their Inhibition in QM/MM Studies. (2021) doi:10.3390/ijms22063232. Viegas, M. F., Neves, R. P. P., Ramos, M. J. & Fernandes, P. A. QM/MM Study of the Reaction Mechanism of Thermophilic Glucuronoyl Esterase for Biomass Treatment. ChemPhysChem 23 , e202200269 (2022). Du, S. et al. Conformational Ensembles Reveal the Origins of Serine Protease Catalysis. bioRxiv 2024.02.28.582624 (2024) doi:10.1101/2024.02.28.582624. Bürgi, H. B., Dunitz, J. D., Lehn, J. M. & Wipff, G. Stereochemistry of reaction paths at carbonyl centres. Tetrahedron 30 , 1563–1572 (1974). Reed, A. E., Weinstock, R. B. & Weinhold, F. Natural population analysis. J. Chem. Phys. 83 , 735–746 (1985). Weinhold, F. & Labdus, C. R. Natural Bond Orbitals and Extensions of Localized Bonding Concepts. Chem Educ Res Pr. 2 , 91–104 (2001). Merz, T. et al. Autoproteolytic and Catalytic Mechanisms for the β-Aminopeptidase BapA—A Member of the Ntn Hydrolase Family. Structure 20 , 1850–1860 (2012). Elsässer, B. et al. Distinct Roles of Catalytic Cysteine and Histidine in the Protease and Ligase Mechanisms of Human Legumain As Revealed by DFT-Based QM/MM Simulations. ACS Catal. 7 , 5585–5593 (2017). Radisky, E. S., Lee, J. M., Lu, C.-J. K. & Koshland, D. E. Insights into the serine protease mechanism from atomic resolution structures of trypsin reaction intermediates. (2006). Xu, Q., Buckley, D., Guan, C. & Guo, H. C. Structural insights into the mechanism of intramolecular proteolysis. Cell 98 , 651–661 (1999). Rosenblum, J. S. & Blobel, G. Autoproteolysis in nucleoporin biogenesis. Proceedings of the National Academy of Sciences of the United States of America 96 , 11370–5 (1999). Karplus, M. Development of Multiscale Models for Complex Chemical Systems: From H+H2 to Biomolecules (Nobel Lecture). Angew. Chem. Int. Ed. 53 , 9992–10005 (2014). Kubař, T., Elstner, M. & Cui, Q. Hybrid Quantum Mechanical/Molecular Mechanical Methods For Studying Energy Transduction in Biomolecular Machines. Annu. Rev. Biophys. 52 , 525–551 (2023). Buller, A. R., Freeman, M. F., Wright, N. T., Schildbach, J. F. & Townsend, C. A. Insights into cis-autoproteolysis reveal a reactive state formed through conformational rearrangement. Proc. Natl. Acad. Sci. 109 , 2308–2313 (2012). D’Osualdo, A. et al. CARD8 and NLRP1 Undergo Autoproteolytic Processing through a ZU5-Like Domain. PLOS ONE 6 , e27396 (2011). Liao, Y., Pei, J., Cheng, H. & Grishin, N. V. An ancient autoproteolytic domain found in GAIN, ZU5 and Nucleoporin98. Journal of Molecular Biology 426 , 3935–3945 (2014). Kim, J. K. et al. Insight into autoproteolytic activation from the structure of cephalosporin acylase: A protein with two proteolytic chemistries. Proc. Natl. Acad. Sci. 103 , 1732–1737 (2006). Sandberg, A., Johansson, D. G. A., Macao, B. & Härd, T. SEA Domain Autoproteolysis Accelerated by Conformational Strain: Energetic Aspects. Journal of Molecular Biology 377 , 1117–1129 (2008). Calinsky, R. & Levy, Y. Histidine in Proteins: pH-Dependent Interplay between π–π, Cation–π, and CH–π Interactions. J. Chem. Theory Comput. 20 , 6930–6945 (2024). Lee, J. et al. CHARMM-GUI Input Generator for NAMD, GROMACS, AMBER, OpenMM, and CHARMM/OpenMM Simulations Using the CHARMM36 Additive Force Field. Journal of Chemical Theory and Computation 12 , 405–413 (2016). Spoel, D. V. D. et al. GROMACS: Fast, flexible, and free. Journal of Computational Chemistry 26 , 1701–1718 (2005). Abraham, M. J. et al. Gromacs: High performance molecular simulations through multi-level parallelism from laptops to supercomputers. SoftwareX 1–2 , 19–25 (2015). Huang, J. & MacKerell, A. D. CHARMM36 all-atom additive protein force field: Validation based on comparison to NMR data. Journal of Computational Chemistry 34 , 2135–2145 (2013). Pedregosa, F. et al. Scikit-Learn: Machine Learning in Python. J. Mach. Learn. Res. 12 , 2825–2830 (2011). Becke, A. D. Density‐functional thermochemistry. III. The role of exact exchange. J. Chem. Phys. 98 , 5648–5652 (1993). Stephens, P. J., Devlin, F. J., Chabalowski, C. F. & Frisch, M. J. Ab Initio Calculation of Vibrational Absorption and Circular Dichroism Spectra Using Density Functional Force Fields. The Journal of Physical Chemistry 98 , 11623–11627 (1994). Lee, C., Yang, W. & Parr, R. G. Development of the Colle-Salvetti correlation-energy formula into a functional of the electron density. Physical Review B 37 , 785–789 (1988). Weigend, F., Häser, M., Patzelt, H. & Ahlrichs, R. RI-MP2: optimized auxiliary basis sets and demonstration of efficiency. Chem. Phys. Lett. 294 , 143–152 (1998). GmbH, T., Karlsruhe, U. of & GmbH, F. K. TURBOMOLE V7.4 . (2019). Ahlrichs, R., Bär, M., Häser, M., Horn, H. & Kölmel, C. Electronic structure calculations on workstation computers: The program system turbomole. Chem. Phys. Lett. 162 , 165–169 (1989). Treutler, O. & Ahlrichs, R. Efficient molecular numerical integration schemes. J. Chem. Phys. 102 , 346–354 (1995). Häser, M. & Ahlrichs, R. Improvements on the direct SCF method. J. Comput. Chem. 10 , 104–111 (1989). Brooks, B. R. et al. CHARMM: The biomolecular simulation program. Journal of Computational Chemistry 30 , 1545–1614 (2009). Riahi, S. & Rowley, C. N. The CHARMM–TURBOMOLE interface for efficient and accurate QM/MM molecular dynamics, free energies, and excited state properties. J. Comput. Chem. 35 , 2076–2086 (2014). Hoffmann, R. An Extended Hückel Theory. I. Hydrocarbons. J. Chem. Phys. 39 , 1397–1412 (1963). Schäfer, A., Horn, H. & Ahlrichs, R. Fully optimized contracted Gaussian basis sets for atoms Li to Kr. J. Chem. Phys. 97 , 2571–2577 (1992). Mayne, C. G., Saam, J., Schulten, K., Tajkhorshid, E. & Gumbart, J. C. Rapid parameterization of small molecules using the force field toolkit. Journal of Computational Chemistry 34 , 2757–2770 (2013). Humphrey, W., Dalke, A. & Schulten, K. VMD: Visual molecular dynamics. Journal of Molecular Graphics 14 , 33–38 (1996). Neese, F. The ORCA program system. Wiley Interdisciplinary Reviews: Computational Molecular Science 2 , 73–78 (2012). Jo, S., Kim, T., Iyer, V. G. & Im, W. CHARMM-GUI: A web-based graphical user interface for CHARMM. Journal of Computational Chemistry 29 , 1859–1865 (2008). Phillips, J. C. et al. Scalable molecular dynamics on CPU and GPU architectures with NAMD. J. Chem. Phys. 153 , 044130 (2020). Weinhold, F. Chemical Bonding as a Superposition Phenomenon. J. Chem. Educ. 76 , 1141 (1999). Additional Declarations The authors declare no competing interests. Supplementary Files 3SupplementaryFile.docx Supplementary File 2ExtDataFig.docx Extended Data Figures Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7121662","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":485218823,"identity":"158692a7-9017-44ca-b12d-14d537438997","order_by":0,"name":"Florian Seufert","email":"","orcid":"https://orcid.org/0000-0002-0664-7169","institution":"Institute of Medical Physics and Biophysics, Medical Faculty, Leipzig University, Germany","correspondingAuthor":false,"prefix":"","firstName":"Florian","middleName":"","lastName":"Seufert","suffix":""},{"id":485218824,"identity":"bb2551f6-33bf-4bdd-91d3-870a29f78368","order_by":1,"name":"Yin Kwan Chung","email":"","orcid":"https://orcid.org/0000-0003-2868-8508","institution":"Rudolf Schönheimer Institute of Biochemistry, Division of General Biochemistry, Medical Faculty, Leipzig University, Germany","correspondingAuthor":false,"prefix":"","firstName":"Yin","middleName":"Kwan","lastName":"Chung","suffix":""},{"id":485218825,"identity":"fd84261e-6f63-4032-85a5-dc59052a8762","order_by":2,"name":"Robin Schick","email":"","orcid":"https://orcid.org/0009-0003-3491-5997","institution":"Institute of Bioanalytical Chemistry, Leipzig University, Germany","correspondingAuthor":false,"prefix":"","firstName":"Robin","middleName":"","lastName":"Schick","suffix":""},{"id":485218826,"identity":"134dfecd-33df-4f6c-9cef-c5421fe025ce","order_by":3,"name":"Fabian Pohl","email":"","orcid":"https://orcid.org/0009-0009-8912-3472","institution":"Rudolf Schönheimer Institute of Biochemistry, Division of General Biochemistry, Medical Faculty, Leipzig University, Germany","correspondingAuthor":false,"prefix":"","firstName":"Fabian","middleName":"","lastName":"Pohl","suffix":""},{"id":485218827,"identity":"ab83af4f-6898-4a6d-aacb-a9398e3260ab","order_by":4,"name":"Norbert Sträter","email":"","orcid":"https://orcid.org/0000-0002-2001-0500","institution":"Institute of Bioanalytical Chemistry, Leipzig University, Germany","correspondingAuthor":false,"prefix":"","firstName":"Norbert","middleName":"","lastName":"Sträter","suffix":""},{"id":485218828,"identity":"f5882be1-bb9d-4057-b76a-adeeba19e62f","order_by":5,"name":"Tobias Langenhan","email":"","orcid":"https://orcid.org/0000-0002-9061-3809","institution":"Rudolf Schönheimer Institute of Biochemistry, Division of General Biochemistry, Medical Faculty, Leipzig University, Germany","correspondingAuthor":false,"prefix":"","firstName":"Tobias","middleName":"","lastName":"Langenhan","suffix":""},{"id":485218829,"identity":"f2ab296f-89b1-42aa-b1f8-a1f3827e6d0b","order_by":6,"name":"Hossein Batebi","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA90lEQVRIiWNgGAWjYBACAwkg8YGBgbEBQh9gYGAHsdnwamFsnAHSwsbAOAOshZkILc08qFpA4ni0mEs3P39s88dGtn9+A0jvHTnzZua2BwxlNji1WM45Zticw5NmPOMY2LpnxjKHGdsNGM6l4XbYjQSgFonDiQ3HGNgf8/47nDiDmbFNgrHtMB4t6R+bLQz+J86H2HK4HqrlPx4tOYbNDAkHEjdAtSRIQLQcwK3lzpnCmT0Hko03HktsbJzD8MwQbEvCuWTcWm63b/jw44+d7LzDhw82vGG4Iy/B3v5M4kOZHU4tSACcAKAggRgNo2AUjIJRMApwAgCRXFciVd0vHwAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0002-0714-9700","institution":"Freie Universität Berlin, Germany","correspondingAuthor":true,"prefix":"","firstName":"Hossein","middleName":"","lastName":"Batebi","suffix":""},{"id":485218830,"identity":"6b18dc73-b15f-44cd-bc97-f531e1b126d7","order_by":7,"name":"Peter Werner Hildebrand","email":"data:image/png;base64,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","orcid":"https://orcid.org/0000-0003-0063-1104","institution":"Institute of Medical Physics and Biophysics, Medical Faculty, Leipzig University, Germany","correspondingAuthor":true,"prefix":"","firstName":"Peter","middleName":"Werner","lastName":"Hildebrand","suffix":""}],"badges":[],"createdAt":"2025-07-14 13:26:58","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-7121662/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7121662/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":86756590,"identity":"d76d6ba8-e41e-41dc-b7cd-fb9f8f3196d9","added_by":"auto","created_at":"2025-07-15 09:29:52","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":133162,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMechanistic detail of GAIN autoproteolysis.\u003c/strong\u003e\u0026nbsp; \u003cstrong\u003ea-c\u003c/strong\u003e, mechanism of the N-O acyl shift: From the the T\u003csup\u003eGPS.+1\u003c/sup\u003e hydroxyl moiety of the reactant R a proton is abstracted\u0026nbsp; by H\u003csup\u003eGPS.-2\u003c/sup\u003e in concert with a nucleophilic attack onto the L\u003csup\u003eGPS.-1\u003c/sup\u003e peptide carbon atom (\u003cstrong\u003ea\u003c/strong\u003e), yielding a hydroxy-oxazolidine anionic tetrahedral intermediate T1 (\u003cstrong\u003eb\u003c/strong\u003e)\u003cstrong\u003e.\u003c/strong\u003e The ring nitrogen abstracts a proton from a nearby water, leaving the ring and completing the N-O acyl shift to the ester intermediate E, while the nascent hydroxide anion abstracts a proton from nearby conserved E\u003csup\u003eH6.50\u003c/sup\u003e (\u003cstrong\u003ec\u003c/strong\u003e). \u003cstrong\u003ed-f\u003c/strong\u003e, mechanism of ester hydrolysis: The flipped ester E\u003csub\u003ef\u003c/sub\u003e forms a hydrogen bond to the now protonated H\u003csup\u003eGPS.-2\u003c/sup\u003e side chain, leaving its carboxyl exposed (Figs. 3,5, \u003cstrong\u003ed\u003c/strong\u003e). A water molecule attacks the ester bond, yielding the tetrahedral intermediate T2. The glutamate acts as a proton acceptor. (\u003cstrong\u003ee\u003c/strong\u003e). The leaving alcoholate group abstracts a proton from the charged amino group, resulting in concerted proton transfers yielding the cleaved product P (\u003cstrong\u003ef\u003c/strong\u003e), with an uncharged N-terminus, protonated E\u003csup\u003eH6.50\u003c/sup\u003e and a charged H\u003csup\u003eGPS.+2\u003c/sup\u003e side chain. \u003cstrong\u003eg\u003c/strong\u003e, scheme of computational scope applied to interrogate the mechanism via molecular mechanics (MM) in molecular dynamics (MD) simulations and quantum mechanics (QM) via hybrid quantum mechanics / molecular mechanics simulations (QM/MM) via means of coordinate scanning and geometry optimizations.\u0026nbsp; From the uncleaved model MD, the reactant R is selected and fed into the reaction scan to the ester intermediate E via tetrahedral intermediate T1. The mechanism is evaluated in one-step fashion to feed into QM/MM for the ester hydrolysis via intermediate T2 to the cleaved product P.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-7121662/v1/a87c5721edd3eaf9f8995494.png"},{"id":86756591,"identity":"9fbf5dd1-3cf2-49ac-b48a-8b2491398a39","added_by":"auto","created_at":"2025-07-15 09:29:52","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":267694,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe QM selection defined in a frame from reactant MD trajectories. a, \u003c/strong\u003eoverview of the rat ADGRL1 GAIN domain uncleaved model based on PDB: 4DLQ\u003csup\u003e8\u003c/sup\u003e\u0026nbsp; with subdomain A (blue), subdomain B (orange) and the \u003cem\u003eStachel\u003c/em\u003e (yellow). \u003cstrong\u003eb\u003c/strong\u003e, MD frame geometry showing the GPCR proteolysis site (GPS) triad where HLT defines the cleavage site; F\u003csup\u003eS10.49\u003c/sup\u003e and E\u003csup\u003eH6.50\u003c/sup\u003e form mechanistically important interactions to the GPS and are included in the QM selection (Supp. Table. 1). \u003cstrong\u003ec\u003c/strong\u003e, the QM/MM system consists of the MD GAIN domain system (cartoon; water and ions not shown), with the explicit QM atom selection (sticks) surrounded by a sphere of residues with atoms within 5.0 Å of the QM selection treated flexible with MM, where atoms outside the sphere remain fixed with MM treatment. \u003cstrong\u003ed\u003c/strong\u003e, explicit QM atom selection consisting of 146/147 atoms including the GPS triad, N839\u003csup\u003eS14.45\u003c/sup\u003e, E656\u003csup\u003eH6.50\u003c/sup\u003e, S770\u003csup\u003eS8.54\u003c/sup\u003e, the backbone (bb) oxygen of Q818 and atoms of adjacent residues (grey) to enable boundary treatment at C-C bonds only. Boundary treatment consists of introducing QM link atoms (pink) at the cut C-C bonds. \u003cstrong\u003ee\u003c/strong\u003e, side-view of the QM selection shows full coverage of the GPS triad by water on one side (Ext. Data Fig. 3), with 11 QM waters included in the selection, connecting to bulk solvent and defining a polar network between all polar QM selection residues.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-7121662/v1/cc397e055cbd1e87305b6258.png"},{"id":86757845,"identity":"cdc91238-9774-41e4-8715-ad1de1e0b924","added_by":"auto","created_at":"2025-07-15 09:45:52","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":273976,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe N-O acyl shift is energetically favored by protonated E\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003eH6.50\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e. \u003c/strong\u003e\u0026nbsp;\u003cstrong\u003ea-e, \u003c/strong\u003eN-O acyl shift with protonated E\u003csup\u003eH6.50\u003c/sup\u003e. a), From the reactant R\u003csub\u003eEH\u003c/sub\u003e, the reaction proceeds in a concerted mechanism of proton transfer (magenta) and the nucleophilic attack on the peptide carbon (blue) towards the tetrahedral intermediate T1\u003csub\u003eEH \u003c/sub\u003e(\u003cstrong\u003eb\u003c/strong\u003e). For completion of the shift, the ring amide is protonated by water (seagreen), with another proton transferred from E\u003csup\u003eH6.50 \u003c/sup\u003e(orange), forming TH\u003csub\u003eEH\u003c/sub\u003e (\u003cstrong\u003ec\u003c/strong\u003e). The ring nitrogen acts as leaving group (yellow), completing the shift towards the ester E\u003csub\u003eEH\u003c/sub\u003e (\u003cstrong\u003ed\u003c/strong\u003e). \u003cstrong\u003ee\u003c/strong\u003e, Energies of optimized QM/MM geometries on the def2-TZVP/B3-LYP levels with R\u003csub\u003eEH\u003c/sub\u003e energy as zero reference show reasonable energy of +26.0 kcal/mol for the ester intermediate. For roughly comparing energies in the transient intermediates, we employed restrained geometry optimization on T1 and TH (indicated in grey; see Ext. Data. Figs. 4-5). From T1\u003csub\u003eEHr\u003c/sub\u003e to TH\u003csub\u003eEHr\u003c/sub\u003e the energy decreased from +29.5 to +25.8 kcal/mol. \u003cstrong\u003ef-j\u003c/strong\u003e, the reaction mechanism of the E- system (deprotonated E\u003csup\u003eH6.50\u003c/sup\u003e) proceeds analogously from R\u003csub\u003eE-\u003c/sub\u003e (\u003cstrong\u003ef\u003c/strong\u003e) to T1\u003csub\u003eE-\u003c/sub\u003e (\u003cstrong\u003eg\u003c/strong\u003e). Amide protonation leads to a hydroxide anion in TH\u003csub\u003eE- \u003c/sub\u003e(\u003cstrong\u003eh\u003c/strong\u003e), which remains in proximity while the nitrogen leaves the ring towards the ester E\u003csub\u003eE-\u003c/sub\u003e (\u003cstrong\u003ei\u003c/strong\u003e). Here, we observe charge repulsion between the amine, hydroxide and carboxyl group (pink). \u003cstrong\u003ej,\u003c/strong\u003e optimized energies reflect repulsion. The amide protonation increases the energy from +27.2 kcal/mol (T1\u003csub\u003eE-\u003c/sub\u003e) to +45.2 kcal/mol (TH\u003csub\u003eE-\u003c/sub\u003e), indicating that the electrostatic repulsion of the hydroxide and E\u003csup\u003eH6.50\u003c/sup\u003e impacts the energy at this point in the mechanism. In the ester E\u003csub\u003eE-\u003c/sub\u003e, the hydroxide is subject to multiple proton transfers towards the edge of the QM system with an energy of +37.4 kcal/mol (see Ext. Data Fig. 5). \u003cstrong\u003ek-n\u003c/strong\u003e, Coordinate scans obtained on the DFTB3 level approximate the potential energy surface. \u003cstrong\u003ek-l, \u003c/strong\u003ePES traces for R-T1 showed a modest energy increase from R\u003csub\u003eEH\u003c/sub\u003e to T1\u003csub\u003eEH\u003c/sub\u003e in the EH (k) system, where in the E- system, energy increases more substantially towards T1\u003csub\u003eE-\u003c/sub\u003e. \u003cstrong\u003em-n\u003c/strong\u003e, In the two-step reaction T1\u003csub\u003eEH\u003c/sub\u003e-E\u003csub\u003eEH\u003c/sub\u003e, the proton from E\u003csup\u003eH6.50\u003c/sup\u003e to the amide lowers energies considerably relative to T1\u003csub\u003eEH\u003c/sub\u003e, again favoring the protonated pathway R\u003csub\u003eEH\u003c/sub\u003e-E\u003csub\u003eEH\u003c/sub\u003e. For the E- system energy between T1\u003csub\u003eE-\u003c/sub\u003e and TH\u003csub\u003eE-\u003c/sub\u003e remains almost constant and contrastingly increases towards E\u003csub\u003eE-\u003c/sub\u003e. For full energies, see Ext. Data Fig. 6.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-7121662/v1/deac4376fce222d3609d1599.png"},{"id":86756944,"identity":"0933b483-5a8e-4a1c-b705-4efbddee2b58","added_by":"auto","created_at":"2025-07-15 09:37:52","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":200293,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMutations of E656\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003eH6.50\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e in the rat ADGRL1 GAIN domain influence fraction of cleaved protein. a\u003c/strong\u003e, Representative blot of the rL1 E656 mutants. Expressions of the receptors were detected by WB against the N-terminal HA tag. β-tubulin was also detected as loading control. Black triangles represent full-length (FL), while red triangles indicate the cleaved N-terminal fragment (NTF). Filled and open triangle represent full and immature glycosylation, respectively. Unspecific immunosignal was indicated by open circle. \u003cstrong\u003eb\u003c/strong\u003e, the extent of cleavage of the receptors was calculated as the percentage of the background-subtracted intensity representing NTF over the sum of the intensity representing NTF and FL (I\u003csub\u003eNTF\u003c/sub\u003e / (I\u003csub\u003eNTF\u003c/sub\u003e + I\u003csub\u003eFL\u003c/sub\u003e) x 100%), regardless of the glycosylation states. Data were represented mean ± standard error of mean (n=3). Data were first confirmed of normality by Shapiro-Wilk test, before one-way ANOVA followed by Tukey’s multiple comparison test. P-values from the multiple comparisons were indicated, and those lower than 0.05 were highlighted in red. \u003cstrong\u003ec\u003c/strong\u003e, In vitro analysis of the E656X/F803Y double mutation in rat ADGRL1 HormR-GAIN domain constructs. After purification via Ni-NTA beads the proteins were detected by western blot analysis against the C-terminal His-tag, following reducing SDS-PAGE. GPS-cleaved fragments including the Fc-tag are marked by ‘cl.’ while ‘uncl.’ indicates uncleaved constructs. The ECL Rainbow Marker Full Range was used as size standard. \u003cstrong\u003ed\u003c/strong\u003e, The fraction of GPS-cleaved GAIN constructs as detected in the western blots was calculated as the percentage of the background-subtracted intensity representing the cleaved construct (cl.) over the sum of the cleaved and uncleaved constructs (Icl. / (Icl. + Iuncl.) ×100 %). Data are presented as mean ± standard error of the mean (n = 2). Data sets were analyzed using a Student’s t‑test and statistical significance is marked. One star (*) is assigned for p \u0026lt; 0.05 and three stars are assigned for p \u0026lt; 0.001 between the corresponding columns. The difference observed between F803Y and all four double mutants E656X/F803Y has a significance of p \u0026lt; 0.001. See also Supp. Fig. 3.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-7121662/v1/16a44fcb88e7ad22c2d110f2.png"},{"id":86756600,"identity":"bae0752f-c635-461d-8cff-0aa04e1fd65b","added_by":"auto","created_at":"2025-07-15 09:29:52","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":190632,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eWater attack for ester hydrolysis favors a \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003etrans-closed \u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003econfiguration. a-b\u003c/strong\u003e,\u003cstrong\u003e \u003c/strong\u003esolvent-accessible surface area (SASA) of the leucine-threonine ester (LTE) in three independent 1000 ns Molecular Dynamics (MD) trajectories shows quick and persistent desolvation of LTE in unprotonated E656 trajectories (\u003cstrong\u003ea\u003c/strong\u003e) while SASA fluctuates in the protonated trajectories (\u003cstrong\u003eb\u003c/strong\u003e). For attack, trajectories were filtered, searching for waters within 4.0 Å of the electrophilic carbon and close to the ideal Bürgi-Dunitz attack angle of 107°\u003csup\u003e49\u003c/sup\u003e, with waters on the same side as H836\u003csup\u003eGPS.-2\u003c/sup\u003e defined as \u003cem\u003ecis \u003c/em\u003eand opposite defined as \u003cem\u003etrans\u003c/em\u003e respectively (\u003cstrong\u003ec\u003c/strong\u003e). Even in closed configurations, water is still present with waters favoring the \u003cem\u003etrans\u003c/em\u003e side (\u003cstrong\u003ed\u003c/strong\u003e), while in the fluctuating SASA trajectories, the \u003cem\u003ecis \u003c/em\u003econformation is more frequent, still favoring \u003cem\u003etrans\u003c/em\u003e\u003cem\u003e\u003cstrong\u003e \u003c/strong\u003e\u003c/em\u003eby number (\u003cstrong\u003ee\u003c/strong\u003e). In the QM/MM workflow (\u003cstrong\u003ef\u003c/strong\u003e) the singular best frame for each possible attack path was extracted from all six systems (Supp. Fig. 14) based on SASA, and water distance and angle; these four frames were translated into a QM/MM system with a 3.0 Å distance restraint, subsequently minimized and partial charges extracted via natural bond order (NBO) analysis\u003csup\u003e50,51,87\u003c/sup\u003e. The four minimized topologies (\u003cstrong\u003eg-j\u003c/strong\u003e) show differences in charge separation between the water oxygen and ester carbon, with higher differences having more favorable energetics of the hydrolysis. The maximum difference in the \u003cem\u003etrans\u003c/em\u003e-closed system (\u003cstrong\u003eh\u003c/strong\u003e), which also corresponds to the QM/MM system state of the minimized ester E\u003csub\u003eEH \u003c/sub\u003e(Fig. 3). Notably, \u003cem\u003ecis\u003c/em\u003e-open and \u003cem\u003ecis\u003c/em\u003e-closed systems (\u003cstrong\u003ei-j\u003c/strong\u003e) were extracted from trajectories with H836\u003csup\u003eGPS.-2\u003c/sup\u003e in HSD hetero state with hydrogen at the C\u003csub\u003eδ\u003c/sub\u003e carbon, a different protonation pattern from the HSE systems used for the reactant R.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-7121662/v1/6fe6e54053b3769535815dd5.png"},{"id":86758937,"identity":"97be0bf8-6b18-40fb-9d97-883202a4c221","added_by":"auto","created_at":"2025-07-15 09:53:59","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":184816,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEster hydrolysis proceeds in two possible pathways.\u003c/strong\u003e \u003cstrong\u003ea\u003c/strong\u003e, the flipped ester E\u003csub\u003ef\u003c/sub\u003e from the EH QM/MM system offers an angle of attack for the water molecule with two potential bases. \u003cstrong\u003eb-e\u003c/strong\u003e, the amine route has the amine acting as a base, abstracting the proton from the water attacking the ester carbon (cyan arrow), resulting in T2\u003csub\u003eN\u003c/sub\u003e coordinated by the protonated histidine (\u003cstrong\u003eb\u003c/strong\u003e). Two proton transfers occur to the alkoxide leaving group and from the nascent C-terminus, resulting in the product P\u003csub\u003eNn\u003c/sub\u003e with neutral N-terminus and charged E\u003csup\u003eH6.50\u003c/sup\u003e (\u003cstrong\u003ec\u003c/strong\u003e), from which another proton transfer yields the charged carboxyl and amine in the product P\u003csub\u003eN\u003c/sub\u003e (\u003cstrong\u003ed\u003c/strong\u003e). \u003cstrong\u003ee\u003c/strong\u003e, energies of intermediates relative to the reactant R\u003csub\u003eEH\u003c/sub\u003e, obtained on the B3-LYP/def2-TZVP level; restrained minimizations are shown in grey. \u003cstrong\u003ef-j\u003c/strong\u003e, the acid route has the glutamate E\u003csup\u003eH6.50\u003c/sup\u003e abstract the proton from the attacking water (magenta arrow), resulting in a hydroxide coordinated with the carboxyl in E\u003csub\u003eAp\u003c/sub\u003e (\u003cstrong\u003ef\u003c/strong\u003e). The nucleophilic attack then proceeds via the hydroxide, resulting in the tetrahedral intermediate T2\u003csub\u003eA\u003c/sub\u003e (\u003cstrong\u003eg\u003c/strong\u003e). Two proton transfers take place, first from the R-OH to the neutral amine, and then from the charged amine to the leaving alkoxide, resulting in the product P\u003csub\u003eA\u003c/sub\u003e with uncharged N-terminus and E\u003csup\u003eH6.50\u003c/sup\u003e (\u003cstrong\u003eh\u003c/strong\u003e). Mediating a proton transfer from E\u003csup\u003eH6.50\u003c/sup\u003e to R-NH\u003csub\u003e2\u003c/sub\u003e via water results in energy increase of +8 kcal/mol in P\u003csub\u003eAp\u003c/sub\u003e (\u003cstrong\u003ei\u003c/strong\u003e), indicating P\u003csub\u003eA\u003c/sub\u003e as the lowest-energy state. \u003cstrong\u003ej\u003c/strong\u003e, Energies of intermediates relative to the reactant R\u003csub\u003eEH\u003c/sub\u003e, obtained on the B3-LYP/def2-TZVP level; restrained minimizations are shown in grey.\u003cstrong\u003e k-l\u003c/strong\u003e, semiempirical coordinate scans on the DFTB3 level show that proton transfers from E\u003csub\u003ef\u003c/sub\u003e/E\u003csub\u003eAp\u003c/sub\u003e to the T2\u003csub\u003eN\u003c/sub\u003e/T2\u003csub\u003eA\u003c/sub\u003e intermediate occur stepwise for both routes. \u003cstrong\u003em-n\u003c/strong\u003e, the proton transfer and alkoxide group leaving occur in concert, with lowering energy in the acid route. See also Ext. Data Fig. 6.\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-7121662/v1/721a2b2701585df75ae823b8.png"},{"id":86759388,"identity":"14bb9c4c-af13-4165-b764-72a4b6245688","added_by":"auto","created_at":"2025-07-15 10:02:02","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2411109,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7121662/v1/2b4ca016-6ba0-4f7c-bd4c-8bb0d2c4b68b.pdf"},{"id":86756948,"identity":"d245d4e8-70ad-4f49-ba6f-b0d498f3f984","added_by":"auto","created_at":"2025-07-15 09:37:52","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":9204131,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary File\u003c/p\u003e","description":"","filename":"3SupplementaryFile.docx","url":"https://assets-eu.researchsquare.com/files/rs-7121662/v1/1d276fbac50febb891e44733.docx"},{"id":86757848,"identity":"04d8a075-0b43-4293-8f39-063900292e99","added_by":"auto","created_at":"2025-07-15 09:45:52","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":11663365,"visible":true,"origin":"","legend":"\u003cp\u003eExtended Data Figures\u003c/p\u003e","description":"","filename":"2ExtDataFig.docx","url":"https://assets-eu.researchsquare.com/files/rs-7121662/v1/69e727be7221668b3d602842.docx"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003e\u003cstrong\u003eMechanistic insights into adhesion GPCR autoproteolysis by a multiscale computational approach\u003c/strong\u003e\u003c/p\u003e","fulltext":[{"header":"Introduction","content":"\u003cp\u003eAdhesion G-protein coupled receptors (aGPCR) comprise a family of GPCRs involved in cell adhesion, migration, paracrine signaling and various disease pathologies\u003csup\u003e\u003cspan additionalcitationids=\"CR2\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. aGPCRs are characterized by a heptahelical transmembrane (7TM) domain and their large extracellular regions containing various protein domains. Their hallmark domain is the GPCR autoproteolysis-inducing (GAIN) domain containing the GPCR proteolysis site (GPS)\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. The GAIN domain is composed of two subdomains, a variable helical subdomain A and a more structurally conserved β-sandwich subdomain B containing the GPS with the conserved catalytic triad of H\u003csup\u003eGPS.\u0026minus;2\u003c/sup\u003e, L\u003csup\u003eGPS.\u0026minus;1\u003c/sup\u003e and S/T\u003csup\u003eGPS.+1\u003c/sup\u003e (superscript indicates GAIN domain generic residue numbers\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e). The GPS is flanked by two flexible regions enveloping the catalytic triad termed Flap 1 and 2, mediating triad solvent-accessibility\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. The GPS mediates \u003cem\u003ecis\u003c/em\u003e-autoproteolysis, forming an intact complex between the N-terminal fragment (NTF) composed of all extracellular domains, and the C-terminal fragment (CTF) with the most C-terminal one of 14 β-strands of the GAIN domain followed by the 7TM region\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e,\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eThe cleaved β-strand comprises the N-terminus of the CTF in the NTF-CTF complex. It has been shown to mediate receptor activation as an intramolecular agonist in many aGPCRs like ADGRL1, ADGRD1, ADGRG1/G5/G6, and is termed the \u0026ldquo;\u003cem\u003eStachel\u003c/em\u003e\u0026rdquo; element\u003csup\u003e\u003cspan additionalcitationids=\"CR11 CR12\" citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. The NTF-CTF complex can be disrupted by e.g. mechanostimulation, resulting in NTF shedding off the CTF\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan additionalcitationids=\"CR15 CR16 CR17\" citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. Shed NTF of some receptors moves away from the cell surface\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e,\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e, to stimulate aGPCR signaling in the extracellular space\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e and leave the exposed \u003cem\u003eStachel\u003c/em\u003e acting as an intramolecular receptor agonist on the 7TM domain. Despite recent structural insights into \u003cem\u003eStachel\u003c/em\u003e mediated G protein activation\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e, the cleavability of aGPCRs and the role of cleavage for their physiological functions remains unclear. The population of some receptors exist cleaved, others in an uncleaved state or even partially in both states like ADGRG1\u003csup\u003e9,23\u003c/sup\u003e.⁠ Moreover, the impact of cleavage on receptor function seems to be receptor-specific, as some uncleaved receptors also exhibit activity\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e. It has therefore been postulated that separate pathways may exist for aGPCR signaling depending on their cleavage competence\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. Finally, the GPS turned out to be necessary but not sufficient for cleavage, pointing towards additional structural determinants of cleavage competence\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e,\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. A detailed structural and mechanistic understanding of \u003cem\u003ecis\u003c/em\u003e-autoproteolysis is thus a key issue of current aGPCR research.\u003c/p\u003e\u003cp\u003eIn this work we used a multiscale computational approach combining quantum mechanics (QM), molecular mechanics (MM) and molecular dynamics (MD) simulations to investigate the mechanism of \u003cem\u003ecis\u003c/em\u003e-autoproteolysis. QM allows for simulating chemical reactions by resolving the electronic configurations on a set of atoms. For simple systems, reactions are modeled in vacuum, however for reactions in proteins, the more complex environment of the reaction site is critical for catalysis. For that purpose, hybrid methods such as QM/MM expand the explicit QM system by a flexible, however inert environment treated on the MM level\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e,\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. QM/MM hybrid approaches are established methods to simulate cleavage of peptide bonds by proteases. The local environments of the active sites of these enzymes have evolutionarily been optimized for high catalytic rate, with a k\u003csub\u003ecat\u003c/sub\u003e of e.g. trypsin in the per second range\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. For maximum catalytic effect, the amino acids involved in the reaction and water required for hydrolysis are usually well-coordinated. In this reaction, a base abstracts a proton, creating a nucleophile for acylation of the peptide bond, resulting in an oxyanionic tetrahedral intermediate. In the next step, the N-O acyl shift is completed through protonation of the backbone amide nitrogen and coordination of water for subsequent ester hydrolysis, which finally yields the cleaved product.\u003c/p\u003e\u003cp\u003eIn contrast to proteolytic enzymes, the \u003cem\u003ecis\u003c/em\u003e-autoproteolytic reaction is orders of magnitude slower and has not been evolutionary optimized, likely because it only occurs once in protein lifetime. For such systems, reaction pathways may be less well-defined, with chemical reactions proceeding through multiple conformationally and chemically distinct intermediates. This mechanistic ambiguity limits the applicability of QM/MM approaches that were developed based on studies of classical enzymes. The QM/MM pathway analysis was thus supplemented with MD simulations to sample conformational changes of putative intermediates on time scales exceeding the orders of magnitude faster chemical reactions. Because force field-based MD simulations employ fixed protonation states and cannot account for changes in electronic configurations, the GAIN domain structure was simulated at different protonation states to identify the most thermodynamically favorable protonation configuration. Energetic evaluations of these configurations provided the appropriate starting geometries for subsequent QM/MM reaction pathway analysis.\u003c/p\u003e\u003cp\u003eOur multiscale MD and QM/MM approach suggests an updated sequence of events during GAIN domain autoproteolysis that involves, in addition to the HL|T catalytic triad residues, conserved residues in the direct environment of the GPS: In the first reaction step, the H\u003csup\u003eGPS.\u0026minus;2\u003c/sup\u003e acts as a base, abstracting a proton from the S/T\u003csup\u003eGPS.+1\u003c/sup\u003e hydroxyl group via a pre-formed hydrogen bond (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). The nascent nucleophilic alkoxide initiates a nucleophilic attack on the carbonyl carbon of L\u003csup\u003eGPS.\u0026minus;1\u003c/sup\u003e, forming the five-membered heteroatomic ring intermediate T1, with its oxyanion solely coordinated by water (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb). Subsequently, the intermediate collapses through protonation of the ring amide by a water molecule located behind the catalytic triad in a pocket between T\u003csup\u003eGPS.+1\u003c/sup\u003e and the conserved E\u003csup\u003eH6.50\u003c/sup\u003e, completing the N-O acyl shift, resulting in the ester intermediate E (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec). The nascent hydroxide anion is coordinated by E\u003csup\u003eH6.50\u003c/sup\u003e, which forms part of a larger polar network between the catalytic triad and the C-terminal helix of GAIN subdomain A. Our analysis suggests a one-step mechanism with ester hydrolysis occurring directly after the completed N-O acyl shift and provides explanations why the catalytic triad is necessary but not sufficient for cleavage\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e,\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. Our computational findings are in agreement with previous experimental data\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e and further supplemented by mutational and functional studies on E\u003csup\u003eH6.50\u003c/sup\u003e, a residue that coordinates a water network enveloping the catalytic triad and which protonation state impacts GAIN domain cleavage.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cb\u003ePreparation of the reactant R for the QM/MM analysis workflow\u003c/b\u003e\u003c/p\u003e\u003cp\u003eWe selected the GAIN domain of rat ADGRL1 (L1), a neuronal aGPCR\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e involved in glucose homeostasis and energy regulation\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e,\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e as model to investigate the detailed autoproteolysis mechanism. For this particular system, there is a wealth of experimental data that can inform further interpretation\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e,\u003cspan additionalcitationids=\"CR34 CR35\" citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. To obtain an uncleaved GAIN domain, we re-formed the cleaved peptide bond at the GPS from the X-ray structure of the L1 (Ext. Data Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea-b). After several equilibration runs interposed by steered MD simulations, the backbone of the triad residues shows excellent agreement to the available uncleaved GAIN domain structure of ADGRB3, which lacks the canonical H-L-S/T motif (Ext. Data Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed)\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. The reactant model of L1 selected from the MD simulations (Supp. Figures\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e\u0026ndash;9) was further validated and cross-referenced to the structure of the uncleaved ADGRB2 GAIN domain, with a canonical H-L-S/T motif\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. Finally, the product of ester hydrolysis at the end of the reaction trajectory reveals the configuration of the cleaved peptide bond at the GPS from the X-ray structure of L1, further corroborating the uncleaved reactant model.\u003c/p\u003e\u003cp\u003eTo simulate bond formation and proton transfer reactions we selected the residues from the GPS triad and proximal regions for QM calculations and defined their protonation states (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). The QM selection contains two titratable residues, H\u003csup\u003eGPS.\u0026minus;2\u003c/sup\u003e and E\u003csup\u003eH6.50\u003c/sup\u003e. From the catalytic triad, H\u003csup\u003eGPS.\u0026minus;2\u003c/sup\u003e was protonated at N\u003csub\u003eε\u003c/sub\u003e (HSE). Only in this protonation state, the system adopts the reactant-characterizing hydrogen bond between T\u003csup\u003eGPS.+1\u003c/sup\u003e and H\u003csup\u003eGPS.\u0026minus;2\u003c/sup\u003e in L1 and B2\u003csup\u003e26\u003c/sup\u003e. Our mechanistic approach furthermore reveals that in L1 the probability of this bond to be formed is influenced by the protonation state of E\u003csup\u003eH6.50\u003c/sup\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb). This residue appears to coordinate a water network enveloping the catalytic triad (see below). Empirical pKa estimation by PropKa3 initially indicated E\u003csup\u003eH6.50\u003c/sup\u003e as charged (referred to as E-)\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. Regardless, MD simulations on both E656 protonation states revealed that protonated E\u003csup\u003eH6.50\u003c/sup\u003e (referred to as EH) enhances formation of the hydrogen bond between T\u003csup\u003eGPS.+1\u003c/sup\u003e and H\u003csup\u003eGPS.\u0026minus;2\u003c/sup\u003e (Ext. Data Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, Supp. Table\u0026nbsp;1).\u003c/p\u003e\u003cp\u003eFrom the proximal GPS triad region, the highly conserved F\u003csup\u003eS10.49\u003c/sup\u003e was included in the QM selection. In the reactant, F\u003csup\u003eS10.49\u003c/sup\u003e forms a T-shaped π-π interaction with H\u003csup\u003eGPS.\u0026minus;2\u003c/sup\u003e. Functional analysis of L1 and B2 variants show that F\u003csup\u003eS10.49\u003c/sup\u003e is essential for cleavage competence in Pohl \u003cem\u003eet al.\u003c/em\u003e, 2023\u003csup\u003e26\u003c/sup\u003e. The QM selection, visualized in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, further includes N839\u003csup\u003eS14.45\u003c/sup\u003e (GPS.+2), Q818\u003csup\u003es11s12\u003c/sup\u003e, E656\u003csup\u003eH6.50\u003c/sup\u003e, S770\u003csup\u003eS8.54\u003c/sup\u003e and additional atoms of adjacent residues to allow only carbon-carbon bonds to form the QM system boundary (see Supp. Data 1). The QM/MM system was finally created with the explicit QM atoms surrounded by a MM-treated sphere consisting of all residues with at least one atom within a 5.0 \u0026Aring; sphere of the QM selection (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec).\u003c/p\u003e\u003cp\u003e\u003cb\u003eThe N-O acyl shift is favored by protonated E\u003c/b\u003e\u003csup\u003e\u003cb\u003eH6.50\u003c/b\u003e\u003c/sup\u003e \u003cb\u003eand proceeds on both sides of the peptide plane\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe N-O acyl shift is the first chemical reaction step in autoproteolysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea, b, \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea). Before we studied the reaction mechanism in detail by QM/MM, we performed a potential energy surface (PES) scan of the reaction to obtain a rough estimation of the putative reaction path. We used the self-consistent charge density functional tight binding (SCC-DFTB) employed via DFTB3, a promising approach in treating large biomolecular systems with reasonable accuracy/performance trade-off\u003csup\u003e\u003cspan additionalcitationids=\"CR39\" citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e. To increase the level of resolution of the resulting energy landscapes, intermediate state geometries were generated to be optimized with the DFT workflow. The PES scans comparing both reactant systems EH (protonated E\u003csup\u003eH6.50\u003c/sup\u003e) and E- (deprotonated). (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ek-l) show that the initial reaction proceeds in concert to reach the hydroxy-oxazolidine intermediate T1 (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb). Moreover, our comparison indicates that the N-O acyl shift is energetically favored with protonated E\u003csup\u003eH6.50\u003c/sup\u003e. Finally, it suggests that the ester E (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec) is reached in the next step via two subsequent reactions, amide protonation and removal of the amine leaving group (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003em-n).\u003c/p\u003e\u003cp\u003eTo decipher which of the two protonation states of E\u003csup\u003eH6.50\u003c/sup\u003e provides the energetically more favorable pathway, we started our QM/MM pathway analysis from the two optimized respective reactant conformations R\u003csub\u003eEH\u003c/sub\u003e and R\u003csub\u003eE\u0026minus;\u003c/sub\u003e. In both setups, we observe that H\u003csup\u003eGPS.\u0026minus;2\u003c/sup\u003e abstracts a proton from the hydroxyl moiety of T\u003csup\u003eGPS.+1\u003c/sup\u003e, resulting in an alkoxide, attacking the peptide carbon between L\u003csup\u003eGPS.\u0026minus;1\u003c/sup\u003e and T\u003csup\u003eGPS.+2\u003c/sup\u003e in concert with the proton transfer reaction (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea,b, \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea,f). The resulting tetrahedral intermediate T1\u003csub\u003eEH\u003c/sub\u003e/T1\u003csub\u003eE\u0026minus;\u003c/sub\u003e is characterized by a five-membered heteroatomic ring and an oxyanion (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb,g) \u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e. This hydroxy-oxazolidine is transient due to ring strain and three electronegative binding partners at the former peptide carbon\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e. From both T1\u003csub\u003eEH\u003c/sub\u003e and T1\u003csub\u003eE\u0026minus;\u003c/sub\u003e, the reaction immediately proceeds via protonation of the ring amide\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e. In serine proteases, this step is mediated by the catalytic histidine imidazole group and was suggested to be rate-limiting\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. In L1, the histidine imidazole is sterically hindered by the five-membered oxazolidine ring to form a direct interaction to the amidic nitrogen. A water molecule located on the opposite (\u003cem\u003etrans\u003c/em\u003e) side of the peptide from H\u003csup\u003eGPS.\u0026minus;2\u003c/sup\u003e (\u003cem\u003ecis\u003c/em\u003e side) forms a hydrogen bond to the unprotonated amide nitrogen instead. Strikingly, in both T1\u003csub\u003eEH\u003c/sub\u003e/T1\u003csub\u003eE\u0026minus;\u003c/sub\u003e this water is additionally coordinated by the carboxyl group of E\u003csup\u003eH6.50\u003c/sup\u003e assigning it a key-role for these first steps (Ext. Data Figs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb, \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb).\u003c/p\u003e\u003cp\u003eFrom the two QM/MM systems, the EH system provides the energetically more favorable mechanism, in agreement with the energy landscape from the surface energy scan (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003em-n). A water-mediated proton hop occurs from neutral E\u003csup\u003eH6.50\u003c/sup\u003e to the amide, leading to charged E\u003csup\u003eH6.50\u003c/sup\u003e in TH\u003csub\u003eEH\u003c/sub\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb-c). In the E- system, E\u003csup\u003eH6.50\u003c/sup\u003e is already charged. As a result, the proton is abstracted from the nearby water, forming a hydroxide. The resulting electrostatic repulsion of multiple negatively charged moieties in proximity significantly increases the energy barrier towards TH\u003csub\u003eE\u0026minus;\u003c/sub\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eg-h), making this protonation state/reaction path unlikely.\u003c/p\u003e\u003cp\u003eThe EH system also represents the energetically more favorable configuration during the final step that completes the N-O acyl shift. The ester intermediate E is formed as a result of expulsion of the amidic nitrogen atom in both TH\u003csub\u003eEH\u003c/sub\u003e/TH\u003csub\u003eE\u0026minus;\u003c/sub\u003e, (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed-e). At +\u0026thinsp;37.4 kcal/mol, however, the increase in energy of the ester compared to the reactant is significantly higher in the E- system than in the EH system with +\u0026thinsp;26.0 kcal/mol (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ee,j). E\u003csub\u003eEH\u003c/sub\u003e energy is consistent with ester intermediates observed in related studies on proteases\u003csup\u003e\u003cspan additionalcitationids=\"CR46 CR47\" citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e. In addition, we see proton hopping in E\u003csub\u003eE\u0026minus;\u003c/sub\u003e with the hydroxide anion located at the edge of the QM system (Ext. Data Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ed-e), whereas the water molecule remains coordinated close to the ester bond and the free amine in E\u003csub\u003eEH\u003c/sub\u003e (Ext. Data Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ed).\u003c/p\u003e\u003cp\u003e\u003cb\u003eGAIN domain cleavage is hampered by E656\u003c/b\u003e\u003csup\u003e\u003cb\u003eH6.50\u003c/b\u003e\u003c/sup\u003e \u003cb\u003emutations\u003c/b\u003e\u003c/p\u003e\u003cp\u003eOur analysis attributes a catalytic role to E\u003csup\u003eH6.50\u003c/sup\u003e as a proton donor for the amide protonation step and coordinating water adjacent to the catalytic triad. Interestingly, water accessibility of L\u003csup\u003eGPS.\u0026minus;1\u003c/sup\u003e depends on the protonation state of this residue (Ext. Data Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed-e). Examining the radial density function around the leucine backbone oxygen (forming the oxyanion in the mechanism, Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb), we find significantly more water in one- and two-layer distances in MD simulations of the reactant R with charged E\u003csup\u003eH6.50\u003c/sup\u003e (Ext. Data Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef-g, Supp. Figure\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Charged E\u003csup\u003eH6.50\u003c/sup\u003e reduces solvent-accessibility of the GPS through formation of a salt-bridge with R777\u003csup\u003es8s9\u003c/sup\u003e, a residue located in the flexible Flap 1 (Supp. Figure\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, Supp. Table\u0026nbsp;2)\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. Notably, R777 does not coordinate to the putative oxyanion of L\u003csup\u003eGPS.\u0026minus;1\u003c/sup\u003e, leaving only water to interact with the anionic oxygen in T1 (Supp. Table\u0026nbsp;3, Ext. Data Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Of note, the determined configuration leaves water molecules coordinated in the proteolytic site, still sufficient for the reaction to take place.\u003c/p\u003e\u003cp\u003eTo investigate the influence of E\u003csup\u003eH6.50\u003c/sup\u003e on self-cleavage, we mutated it to Ala, Leu, Gln and Lys, and quantified the effect using two sets of experiments. Firstly, using surface expressed full-length rat L1 receptor (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea-b), and secondly, using secreted and purified HormR-GAIN domains of rat L1 (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec-d). To assess the extent of cleavage, determined 48 hours after transfection, the reaction was slowed down by the additional mutation F803Y\u003csup\u003eS10.49\u003c/sup\u003e in the secreted ectodomains.\u003c/p\u003e\u003cp\u003eThe first mutation E656Q substitutes the carboxyl with an amide group. Replacing the proton donor with an H-bond donor preserves the capability of this residue to coordinate water, but prevents the proton hop during the last step of the N-O acyl shift. This mutation retains the highest cleavage activity of the four variants, with minor reduction compared to native E656 in the full-length constructs, but around 40% reduction of cleavage in the secreted ectodomains. The observation that cleavage is not completely abrogated by the loss of E656 as proton donor suggests that coordination of the water molecule for protonation of the leaving amide nitrogen is the most important role of E\u003csup\u003eH6.50\u003c/sup\u003e for cleavage. To test this hypothesis, we substituted E\u003csup\u003eH6.50\u003c/sup\u003e by Ala and Leu. E656A should abolish the coordination effect of the side chain while still preserving solvent-accessibility of the amide. E656L should have the strongest impact by introducing a hydrophobic bulky sidechain, altering solvation and polarity of the cleavage site. In agreement, the experiments show minor diminishing of cleavage for E656A in the full-length receptor, while cleavage is reduced by \u0026gt;\u0026thinsp;80% for the secreted ectodomain. For E656L, as expected, we see significant cleavage reduction in both experiments, with 40% and more than 90% reduction in the full-length receptors and secreted GAIN domain, respectively. Finally, we replaced E656 with a lysine to alter the water coordination but allow the nascent hydroxide after amide protonation to be coordinated by the protonated lysine. E656K also causes a significant reduction in cleavage activity, with ~\u0026thinsp;30% reduction in the full-length L1 variant and 75% reduction in the full-length ectodomains.\u003c/p\u003e\u003cp\u003e\u003cb\u003eMultiple avenues for ester hydrolysis exist for the ester intermediate\u003c/b\u003e\u003c/p\u003e\u003cp\u003eIn the previously proposed \u003cem\u003ecis\u003c/em\u003e-autoproteolytic mechanism, H\u003csup\u003eGPS.\u0026minus;2\u003c/sup\u003e acts as a base to generate a hydroxide anion for ester hydrolysis\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e. However, H\u003csup\u003eGPS.\u0026minus;2\u003c/sup\u003e is protonated in the obtained ester conformation E (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec), suggesting a different mechanism. The E\u003csub\u003eEH\u003c/sub\u003e path described above suggests that the ester hydrolysis is mediated by a water molecule, coordinated between the ester carbon, the neutral amino group and charged E\u003csup\u003eH6.50\u003c/sup\u003e (Ext. Data Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ee). This water could be deprotonated by either the nearby neutral amine or the carboxylate moiety of E\u003csup\u003eH6.50\u003c/sup\u003e, resulting in a hydroxide for nucleophilic attack on the ester carbon (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ef-g). With the given low efficiency of GAIN domain autoproteolysis, the ester intermediate, however, may have a sufficiently long lifetime to allow conformational re-arrangements and changes in protonation states of H\u003csup\u003eGPS.\u0026minus;2\u003c/sup\u003e. The motility of the ester bond compared to the more rigid planar peptide bond between L\u003csup\u003eGPS.\u0026minus;1\u003c/sup\u003e and T\u003csup\u003eGPS.+1\u003c/sup\u003e, facilitates rotation and changes of dihedral angles. Together with the accessibility of the titratable residues H\u003csup\u003eGPS.\u0026minus;2\u003c/sup\u003e and E\u003csup\u003eH6.50\u003c/sup\u003e to bulk solvent, it seems reasonable to assume that several conformations and avenues for hydrolysis may exist.\u003c/p\u003e\u003cp\u003eWith the goal of identifying the most favorable conformation for hydrolysis, we performed MD simulations of different ester intermediate states. In the various setups, we incorporated a parameterized leucine-threonine ester (LTE) with H\u003csup\u003eGPS.\u0026minus;2\u003c/sup\u003e in HSD, HSE and HSP states as well as charged and neutral E\u003csup\u003eH6.50\u003c/sup\u003e (denoted as E- and EH, respectively; Ext. Data Fig.\u0026nbsp;7), yielding six different setups. In the obtained trajectories, we observe distinct orientations of the ester dihedral relative to the histidine labelled \u003cem\u003ecis\u003c/em\u003e, neutral and \u003cem\u003etrans\u003c/em\u003e (Ext. Data Fig.\u0026nbsp;8, Supp. Figures\u0026nbsp;10\u0026ndash;14). In uncleaved L1 MD simulations, the peptide bond would always correspond to the neutral state. In LTE MD simulations, for HSE and HSP states, the carbonyl oxygen of the ester forms a hydrogen bond with H\u003csup\u003eGPS.\u0026minus;2\u003c/sup\u003e. In this conformation, the ester plane is exposed to water for (\u003cem\u003etrans\u003c/em\u003e-side) hydroxide attack. In the \u003cem\u003etrans\u003c/em\u003e conformation, the carbonyl group of the ester points away from the histidine and is coordinated by water, so that the ester plane faces to H\u003csup\u003eGPS.\u0026minus;2\u003c/sup\u003e for possible (\u003cem\u003ecis\u003c/em\u003e-side) hydroxide attack.\u003c/p\u003e\u003cp\u003eWith the goal of finding suitable geometries for the water attack in the MD trajectories, we examined solvent-accessible surface area (SASA) and water geometries around the electrophilic ester carbon in the six setups. For LTE SASA, we observe a drop to \u0026lt;\u0026thinsp;0.1 nm\u0026sup2; in the first 250 ns in charged E- systems, while this value fluctuates between 0-0.5 nm\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e in the neutral EH systems (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e. a-b, Supp. Figure\u0026nbsp;14). When we analyzed waters within 4.0 \u0026Aring; and within a 15\u0026deg; cone of the ideal B\u0026uuml;rgi-Dunitz angle of 107\u0026deg; relative to the ester plane (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ec-e)\u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e, we find a clear preference for the \u003cem\u003etrans\u003c/em\u003e side in both. From three independent trajectories with 10,000 frames each (1,000 ns with 100 ps trajectory time step), we find 473 frames (1.58%) in \u003cem\u003etrans\u003c/em\u003e to 52 frames in \u003cem\u003ecis\u003c/em\u003e (0.17%) in the EH system and with 353 frames in \u003cem\u003etrans\u003c/em\u003e (1.17%) to 1 frame in \u003cem\u003ecis\u003c/em\u003e (0.003%) in the E- system.\u003c/p\u003e\u003cp\u003eTo examine the best attack geometry, we extracted the four best frames for each recombination of high and low SASA states and \u003cem\u003ecis\u003c/em\u003e / \u003cem\u003etrans\u003c/em\u003e attack geometry and created a separate QM/MM system for each of them. All systems were minimized with the water restrained to 3.00 \u0026Aring; to the ester carbon before we calculated the natural bond orbital (NBO) charges as partial atomic charges to estimate the electrostatic forces for nucleophilic attack\u003csup\u003e\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e,\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e. Stronger electrostatic attraction between the carbon and the water oxygen would indicate a more favorable geometry for hydrolysis.\u003c/p\u003e\u003cp\u003eNotably, the two \u003cem\u003ecis-\u003c/em\u003eframes originate from trajectories with the HSD state, where the histidine would act as a base for generating the hydroxide for hydrolysis as suggested previously and observed for hydrolases\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e,\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e. Both \u003cem\u003etrans\u003c/em\u003e-frames originate from HSP/HSE trajectories, where either E\u003csup\u003eH6.50\u003c/sup\u003e or the R-NH\u003csub\u003e2\u003c/sub\u003e amine could activate the water via base catalysis, corresponding to the E\u003csub\u003eEH\u003c/sub\u003e path. \u003cem\u003eTrans-\u003c/em\u003eattack proceeds differently from the proposed mechanism, with the initial base H\u003csup\u003eGPS.\u0026minus;2\u003c/sup\u003e remaining protonated and coordinating the ester oxygen and the oxyanion after water attack.\u003c/p\u003e\u003cp\u003eTo further assess the reactivity potential of the hydrolytic configurations, we examined charge separation between the reacting water molecule and the electrophilic ester carbon. Charge separation is a key indicator of polarization and potential nucleophilic attack, with a larger difference between partial charges reflecting a more reactive state. To quantify this, we computed Natural Bond Orbital (NBO) charges\u003csup\u003e\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e, which provide chemically intuitive estimates of atomic partial charges based on the electron density distribution. Notably, the two \u003cem\u003ecis\u003c/em\u003e-frames originate from trajectories using the HSD protonation state, where histidine acts as a general base, abstracting a proton from water and generating a hydroxide capable of nucleophilic attack, as previously suggested\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e,\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e. In contrast, \u003cem\u003etrans\u003c/em\u003e-frames stem from HSP/HSE trajectories, where either E\u003csup\u003eH6.50\u003c/sup\u003e or the R-NH₂ amine group can abstract the proton from water, corresponding to the E\u003csub\u003eEH\u003c/sub\u003e pathway.\u003c/p\u003e\u003cp\u003eInterestingly, \u003cem\u003etrans\u003c/em\u003e-frames exhibited slightly higher charge separation: +0.892/+0.904 on the ester carbon and \u0026minus;\u0026thinsp;0.985/-1.003 on the water oxygen in the open and closed GPS states, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eg\u0026ndash;h). In comparison, \u003cem\u003ecis\u003c/em\u003e conformations showed\u0026thinsp;+\u0026thinsp;0.878/+0.871 on the carbon and \u0026minus;\u0026thinsp;0.976/-0.967 on the water oxygen for the open/closed configurations (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ei\u0026ndash;j), respectively. These values suggest a slightly enhanced polarization in the \u003cem\u003etrans\u003c/em\u003e-pathway, which may facilitate nucleophilic attack during hydrolysis. Despite harmonic restraints of 1000 kJ mol\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e nm\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e applied to keep the oxygen-carbon distance at 3.00 \u0026Aring;, we observe strong attraction between both atoms in both trans systems, resulting in 2.85 \u0026Aring; and 2.84 \u0026Aring; for \u003cem\u003etrans\u003c/em\u003e-open and \u003cem\u003etrans\u003c/em\u003e-closed conformations, while in \u003cem\u003ecis\u003c/em\u003e-open and \u003cem\u003ecis\u003c/em\u003e-closed conformations, distances remained at 3.00 \u0026Aring; and 3.02 \u0026Aring;. Thus, the attractive electrostatic force is higher in \u003cem\u003etrans\u003c/em\u003e systems (\u0026ndash;33.4 pN for \u003cem\u003etrans\u003c/em\u003e-closed, -32.1 pN for \u003cem\u003etrans\u003c/em\u003e-open; -27.4 pN for \u003cem\u003ecis\u003c/em\u003e-closed, -27.9 pN for \u003cem\u003ecis\u003c/em\u003e-open). This higher electrostatic attraction indicates that the \u003cem\u003etrans\u003c/em\u003e attack, corresponding to the reaction path taken in the E\u003csub\u003eEH\u003c/sub\u003e QM/MM system, is the preferable path for hydrolysis (Ext. Data Table\u0026nbsp;1).\u003c/p\u003e\u003cp\u003e\u003cb\u003eEster hydrolysis can proceed in two pathways with different product energies\u003c/b\u003e\u003c/p\u003e\u003cp\u003eTo examine the ester hydrolysis on the quantum-mechanical level, we employed the EH QM/MM system, that corresponds to the favored \u003cem\u003etrans\u003c/em\u003e-attack hydrolysis. Since this system is also the natural product of the N-O acyl shift, the entire mechanism can be described as a single trajectory starting from a protonated E\u003csup\u003eH6.50\u003c/sup\u003e reactant geometry. By optimizing the geometry, a hydrogen bond was formed between the ester oxygen and the now cationic H\u003csup\u003eGPS.\u0026minus;2\u003c/sup\u003e, resulting in the flipped ester E\u003csub\u003ef\u003c/sub\u003e with a relative energy of +\u0026thinsp;27.0 kcal/mol to the reactant R, 1.0 kcal/mol higher than the original E\u003csub\u003eEH\u003c/sub\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed). With this configuration, a water molecule is coordinated between the ester carbon, the neutral amino group and charged E\u003csup\u003eH6.50\u003c/sup\u003e (Ext. Data Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ee). During the chemical reaction, the water is first deprotonated before the hydroxide carries out a nucleophilic attack on the ester carbon (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ef-g). Both the amine and the acid moiety of E\u003csup\u003eH6.50\u003c/sup\u003e are potential bases for the hydrolysis step, resulting in potentially two different pathways.\u003c/p\u003e\u003cp\u003eIn the amine pathway (denoted with N), the amine acts as a base and the nucleophilic attack proceeds in concert with the proton transfer from water, resulting in the tetrahedral intermediate T2\u003csub\u003eN\u003c/sub\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eb). Hydrolysis takes place via the alkoxide, which acts as a leaving group and receives a proton from the positively charged amino group, resulting in a product with neutral termini P\u003csub\u003eNn\u003c/sub\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ec,e). This product in turn receives a proton from the nascent C-terminal carboxylic acid group, resulting in a product P\u003csub\u003eN\u003c/sub\u003e with a relative energy of -3.0 kcal/mol, lower than the reactant (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ed-e).\u003c/p\u003e\u003cp\u003eIn the acid (A) pathway, the water transfers a proton (back) to the negatively charged E\u003csup\u003eH6.50\u003c/sup\u003e. The resulting hydroxide attacks the ester carbon, yielding the tetrahedral intermediate T2\u003csub\u003eA\u003c/sub\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ef). Hydrolysis proceeds analogously to the amino route, however now with an uncharged amino group. In restrained minimization, both unsubstituted oxygens of the tetrahedral intermediate T2\u003csub\u003eA\u003c/sub\u003e coordinate a proton through a hydrogen bond; a hydrogen from the original carbonyl-oxygen with H\u003csup\u003eGPS.\u0026minus;2\u003c/sup\u003e and another hydrogen bond from the original hydroxide to the amino group, respectively (Ext. Data Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eg). During the chemical reaction, the neutral amine first completely abstracts the hydrogen bonded proton (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eg), before it transfers a proton to the leaving alkoxide group. With a neutral N-terminal amine and a neutral E\u003csup\u003eH6.50\u003c/sup\u003e, the resulting product P\u003csub\u003eA\u003c/sub\u003e had with \u0026minus;\u0026thinsp;16.3 kcal/mol the lowest energy of all states (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eh,j, Ext. Data Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). Notably, inducing charge separation by simulating water-mediated proton transfers clearly increased the energy to -6 kcal/mol in P\u003csub\u003eAp\u003c/sub\u003e, highlighting the impact of the solvated and polar GPS triad environment (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ei-j). The obtained product conformations were checked for agreement with the conformational space of the cleaved L1 crystal structure (PDB: 4DLQ)\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e employing MD simulations. Principal component analyses of the whole GAIN and the GPS environment showed good agreement of the non-hydrogen coordinates, indicating a continuum between the obtained product states and available structural data (Supp. Figures\u0026nbsp;17\u0026ndash;18).\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003e\u003cem\u003eCis\u003c/em\u003e-autoproteolytic proteins perform a unique self-cleavage reaction only once in their lifetime, distinguishing them from classical enzymes like serine proteases, which are evolutionarily optimized for rapid and repeated catalysis\u003csup\u003e\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e,\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e,\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e\u003c/sup\u003e. This one-time reaction found e. g. in aGPCRs, inteins or glycosylasparaginase\u003csup\u003e\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e,\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e\u003c/sup\u003e (Supp. Table\u0026nbsp;4), combined with presumably high structural flexibility of the cleavage site creates challenges for detailed mechanistic studies. Contrary to enzymatic studies, structures with a bound transition state analogue are not readily available for autoproteolytic proteins. These differences make it challenging to study the mechanism of \u003cem\u003ecis\u003c/em\u003e-autoproteolysis using conventional workflows developed for enzymes with high catalytic rates and optimized reaction geometries.\u003c/p\u003e\u003cp\u003eTo address these challenges, we employed a multiscale computational approach that integrates molecular dynamics (MD) simulations with hybrid quantum mechanics/molecular mechanics (QM/MM) calculations. MD simulations allowed us to explore the flexible structural intermediates of the GAIN domain, the hallmark domain of aGPCRs\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. QM/MM methods provided the detailed electronic and energetic description of the chemical steps involved in the cleavage\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e,\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e,\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e,\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e\u003c/sup\u003e. With this approach, we propose a more detailed perspective on the cleavage mechanism by aGPCRs influenced by conserved residues and water adjacent to the HL|T catalytic triad of the GAIN domain that coordinate solvent accessibility and co-determine the chemical reaction. This can be extended to assess cleavability via evaluating all human GAIN domain models for presence of cleavage determinants (Supp. Table\u0026nbsp;5).\u003c/p\u003e\u003cp\u003eWhile the mechanism of \u003cem\u003ecis\u003c/em\u003e-autoproteolysis has not so far been structurally elucidated in detail, a basic residue has been proposed to initiate the cleavage reaction not only in the GAIN domain, but also in Nucleoporin98, the panthetine hydrolase ThnT and the zona occludens-1 (ZO-1) protein family\u003csup\u003e\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e,\u003cspan additionalcitationids=\"CR60\" citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e–\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e\u003c/sup\u003e (Supp. Table\u0026nbsp;4). By contrast, in glycosylasparaginase or cephalosporin acylase\u003csup\u003e\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e,\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e,\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e\u003c/sup\u003e, an acidic residue serves as the initial base. Generally, mechanistic insights on \u003cem\u003ecis\u003c/em\u003e-autoproteolytic proteins are lacking, with the exception of the SEA domain, where an acid-catalysis of the N-O acyl shift is implied\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e,\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e,\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eOur results reveal that the cleavage reaction in the GAIN domain of L1 starts with a proton transfer from the threonine residue (T\u003csup\u003eGPS.+1\u003c/sup\u003e) at the cleavage site to a nearby histidine residue (H\u003csup\u003eGPS.−2\u003c/sup\u003e). This proton transfer is stabilized by a T-shaped π–π interaction of H\u003csup\u003eGPS.−2\u003c/sup\u003e with a phenylalanine residue (F\u003csup\u003eS10.49\u003c/sup\u003e), which reduces the configurational entropy of the cleavage site and increases proton affinity\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e,\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e\u003c/sup\u003e, thus preparing the system for catalysis. In a concerted mechanism with proton transfer, the nucleophilic threonine alkoxide attacks the adjacent peptide bond, leading to the formation of a five-membered oxazolidine intermediate T1 (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea,b).\u003c/p\u003e\u003cp\u003eThis mechanism differs from that of serine proteases, where the histidine protonates the amide nitrogen after nucleophilic attack by serine. In the GAIN domain, the five-membered ring sterically blocks this proton transfer. Instead, a water molecule on the opposite (\u003cem\u003etrans-\u003c/em\u003e) side of the histidine donates the proton to the amide nitrogen. This water molecule is coordinated by a conserved glutamate residue (E656\u003csup\u003eH6.50\u003c/sup\u003e), which facilitates a proton hop by quickly regaining a proton, maintaining the catalytic cycle (Ext. Data Fig.\u0026nbsp;10). This is only possible with neutral E656\u003csup\u003eH6.50\u003c/sup\u003e, while electrostatic repulsion makes a pathway with charged E656\u003csup\u003eH6.50\u003c/sup\u003e less likely. The shift towards the ester is eventually completed, when the ring opens with the ring nitrogen acting as leaving group.\u003c/p\u003e\u003cp\u003eA similar catalytic mechanism for \u003cem\u003ecis\u003c/em\u003e-autoproteolysis has been proposed in a structural study on the pantetheine hydrolase ThnT\u003csup\u003e\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e\u003c/sup\u003e, where a threonine forms the alkoxide in the first step of the reaction. On the \u003cem\u003etrans-\u003c/em\u003eside opposite the initial nucleophilic attack, two aspartates were suggested to coordinate a water molecule with one in a neutral state facilitating a proton hop for amide protonation. This suggests a conserved chemical reaction theme across different \u003cem\u003ecis\u003c/em\u003e-autoproteolytic proteins involving nucleophilic attack by a hydroxyl group and amide protonation via proton hop mediated by acidic residues and water. Mutation of the acidic residue E656\u003csup\u003eH6.50\u003c/sup\u003e indeed shows that this conserved residue is functionally important for cleavage in aGPCRs (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). The observation that E656Q and E656A do only have a minor effect on cleavage competence compared to E656L, assign a superior role of cleavage competence to water coordination, which is presumably still possible with the former variants. This residue together with the flexible flaps surrounding the cleavage site modulates solvent accessibility, influencing cleavage efficiency. Swapping flap regions between different aGPCRs alters cleavage rates, support the concept of these dynamic protein regions helping to regulate catalysis by controlling water access\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eWater in the \u003cem\u003etrans-\u003c/em\u003epocket, coordinated by E\u003csup\u003eH6.50\u003c/sup\u003e during the N–O acyl shift, also plays a critical role in the subsequent ester hydrolysis. Unlike serine proteases, where the initiating histidine base activates water for nucleophilic attack, the GAIN domain uses E\u003csup\u003eH6.50\u003c/sup\u003e to abstract a proton from water, enabling it to attack the ester bond. This supports a more general catalytic function of E\u003csup\u003eH6.50\u003c/sup\u003e in \u003cem\u003ecis\u003c/em\u003e-autoproteolysis. An alternative mechanism cannot be excluded, where the amine group serves as the proton abstracting base— a mechanism observed in a study on spontaneous hydrolysis of a glycylserine dipeptide, where water is coordinated by two amino groups and activated through simultaneous proton transfer\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e. Remarkably, in both proposed reaction routes, the leaving alkoxide group is protonated by the same amine group, matching the mechanism proposed for GPS cleavage. Notably, the energy profile of glycylserine hydrolysis shows a maximum transition energy state of 29.4 kcal/mol, with water addition to the ester and serine R–OH attack as rate-limiting steps - values similar to the GAIN domain system\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e. Regardless of mechanistic alternative, the negative partial charges surrounding the water promote proton abstraction, generating the hydroxyl for ester hydrolysis.\u003c/p\u003e\u003cp\u003eThe two mechanistic routes for ester hydrolysis can directly proceed from the ester geometry obtained by quantum-mechanical modeling of the N–O acyl shift. This supports the idea that \u003cem\u003ecis\u003c/em\u003e-autoproteolysis may proceed as a single, continuous reaction pathway. However, the conformational flexibility of the ester intermediate could allow for alternate hydrolysis paths (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e) or simply reduce the reaction rate by increasing conformational entropy and obscuring reactive conformations. Analogously, the uncleaved GAIN domain exhibits a similar effect, with the reactive conformation R\u003csub\u003eEH\u003c/sub\u003e/R\u003csub\u003eE−\u003c/sub\u003e representing a small subset of observed conformational space. Thus, assessing reaction speed solely from reaction energies falls short of including the entropic effect of the dynamic GAIN domain on the cleavage rate.\u003c/p\u003e\u003cp\u003eIn conclusion, our study provides detailed mechanistic insights into \u003cem\u003ecis\u003c/em\u003e-autoproteolysis, highlighting how conserved residues and structured water networks enable a chemically precise reaction within an inherently flexible system. These findings deepen our understanding of aGPCR processing and may apply broadly to other autoproteolytic proteins, offering a framework to explore similarly elusive one-time cleavage events in biology.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003eComputational Methods\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCreation and MD of the uncleaved ADGRL1 Model\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eBased on the existing cleaved GAIN\u0026thinsp;+\u0026thinsp;HormR domain structure of rat ADGRL1 (PDB ID: 4DLQ), the geometry of the backbone of the uncleaved ADGRB3 GAIN structure was aimed to be achieved in the uncleaved ADGRL1 model\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e.First, the T838\u003csup\u003eGPS.+1\u003c/sup\u003e residue was rotated along the C-C\u0026alpha; bond to better face the C-terminal carboxyl group of L837\u003csup\u003eGPS.\u0026minus;1\u003c/sup\u003e. The OT2 atom of L\u003csup\u003eGPS.\u0026minus;1\u003c/sup\u003e was deleted and a bond introduced between T838\u003csup\u003eGPS.+1\u003c/sup\u003e:N and L837\u003csup\u003eGPS.‑1\u003c/sup\u003e:C using the Builder utility in PyMol. The OT1 atom was re-named to O and the topology fed into CHARMM-GUI to automatically normalize bond lengths and generate minimization and equilibration inputs using the CHARMM36 forcefield and GROMACS version 2020.2\u003csup\u003e65\u0026ndash;68\u003c/sup\u003e. After minimization using the steepest-descent method for 5,000 steps, a 125,000 step equilibration with 1 fs timestep was performed to yield the equilibrated uncleaved model. After comparing the backbone \u0026phi; and \u0026psi; angles to ADGRB3 and the measuring the distance of H836\u003csup\u003eGPS.\u0026minus;2\u003c/sup\u003e side chain nitrogens to the target T838\u003csup\u003eGPS.+1\u003c/sup\u003e O\u003csub\u003e\u0026gamma;\u003c/sub\u003e of 4.4 \u0026Aring;, in two separate systems for the HSE and HSD state, the respective basic nitrogen atom and the oxygen were pulled together by using biased MD after 5 ns of equilibration with a time step of 1 fs, applying an harmonic umbrella potential on both atoms with a force constant of 1000 kJ/mol*nm\u0026sup2; and a pull rate of -0.001 nm/ns until the run terminated due to a low-distance warning (HSD: 510 ps, HSE: 650 ps). With the resulting configurations, an equilibration cascade with decreasing harmonic potential holding the N\u003csub\u003e\u0026delta;/\u0026epsilon;\u003c/sub\u003e‑O\u003csub\u003e\u0026gamma;\u003c/sub\u003e distance constant for 100 ns with a force constant of 1000, 500, 200 and 100 kJ/mol*nm\u0026sup2;, respectively, in a decreasing order, while simultaneously applying backbone, sidechain and dihedral restraints of 400, 40 and 4 kJ/mol*nm\u0026sup2; or kJ/mol*deg, respectively, to the protein. After equilibration, we performed triplicate unbiased MD simulations for 1,500 ns using the CHARMM36 force field in GROMACS.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eReactant pose selection\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe resulting trajectories were concatenated for HSD and HSE, respectively, and clustered with Principal Component Analysis (PCA) using the scikit-learn package in Python 3.9\u003csup\u003e69\u003c/sup\u003e. The distances in the unbiased MD between the H\u003csup\u003eGPS.\u0026minus;2\u003c/sup\u003e nitrogen atoms and the T\u003csup\u003eGPS.+1\u003c/sup\u003e O\u003csub\u003e\u0026gamma;\u003c/sub\u003e were analyzed to find a potential occupation of the hydrogen-bond distance between the respective unprotonated nitrogen and the oxygen. Based on the distance data, the HSE state was selected and the centroid of the cluster with the lowest N\u003csub\u003e\u0026delta;\u003c/sub\u003e‑O\u003csub\u003e\u0026gamma;\u003c/sub\u003e distance extracted (Ext. Data Fig. 9a-d). The H\u003csup\u003eGPS.\u0026minus;2\u003c/sup\u003e and F\u003csup\u003eS10.49\u003c/sup\u003e \u0026pi;-\u0026pi;-interaction was stable for both HSE runs (Ext. Data Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ej-k), in contrast, three MD systems (HSD-E, HSD-EH, HSP-EH) showed fluctuations in this interaction (Supp. Figures\u0026nbsp;8\u0026ndash;9).\u003c/p\u003e\n\u003cp\u003eWhile we initially hypothesized R777 to coordinate the oxyanion of the tetrahedral intermediate during the mechanism, the simultaneous low-distance between the guanidyl moiety of R777 to the L\u003csup\u003eGPS.\u0026minus;1\u003c/sup\u003e peptide oxygen and H\u003csup\u003eGPS.\u0026minus;2\u003c/sup\u003eN\u003csub\u003e\u0026delta;\u003c/sub\u003e to T\u003csup\u003eGPS.+1\u003c/sup\u003e O\u0026gamma; did not coincide (Supp Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e). We selected the centroid of cluster 2 of the PCA-clustered trajectory as a reasonable reactant with a favorable hydrogen bond configuration between the triad histidine and threonine and ran geometry optimization with the B3-LYP \u003csup\u003e\u003cspan class=\"CitationRef\"\u003e70\u003c/span\u003e\u0026ndash;\u003cspan class=\"CitationRef\"\u003e72\u003c/span\u003e\u003c/sup\u003e functional and the Weigend and Ahlrichs def2-SVP basis set followed by geometry optimization with the def2-TZVP basis set (for detailed QM/MM methods, see below; Ext. Data Fig. 9d-e) \u003csup\u003e\u003cspan class=\"CitationRef\"\u003e73\u003c/span\u003e\u003c/sup\u003e. We duplicated the system, adding an additional proton to the E656\u003csup\u003eH6.50\u003c/sup\u003e residue with identical conformation. From there, for the protonated E656\u003csup\u003eH6.50\u003c/sup\u003e system, we traced the pathway along R-E-P and in reverse P-E-R to generate a reverse reactant for validation of the initial reactant R. The validation pathway was generated via DFTB3 scans and intermediates were optimized using the B3-LYP functional and the def2-SVP basis set. (Ext. Data Fig. 9f). While the QM/MM poses of the reactant and the reversed reactant did not significantly differ (Ext. Data Fig. 9g), we selected the reversed reactant as R for the pathway tracing due to more favorable water coordination.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eQM/MM selection\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe explicit QM selection comprised 146 or 147 (proton at E656) atoms including 8 link atoms with a total of 16 molecules. Link atoms were used for treating the QM/MM boundary, breaking C-C bonds, therefore some backbone atoms of G819, N835 and A769 were included in the QM selection (full list of QM selection in Supp. Data 1). Furthermore, our reactant frame was bulk-solvent accessible, so we included 11 water molecules covering the GPS triad towards Flap 1 and bulk solvent as well as forming a polar network between L\u003csup\u003eGPS.\u0026minus;1\u003c/sup\u003e, T\u003csup\u003eGPS.+1\u003c/sup\u003e E\u003csup\u003eH6.50\u003c/sup\u003e and S\u003csup\u003eS8.54\u003c/sup\u003e (Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ed).\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eReaction coordinate scans with semi-empirical DFTB3\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eReactions were simulated by employing multidimensional coordinate scans with DFTB3 using CHARMM46. We used the previously defined QM/MM system and for each reaction step (R-T1, T1-E, E-T2, T2-P), the changing interatomic distances were scanned to reach reasonable estimated distances for each intermediate, respectively. For each point in the one- to three-dimensional conformational space, a geometry optimization was performed with DFTB3. Upon meeting energy gradient thresholds, the energy and conformation was stored along with the current interatomic distances. For example, the R\u003csub\u003eEH\u003c/sub\u003e-T1\u003csub\u003eEH\u003c/sub\u003e scan was performed on the distances between H\u003csup\u003eGPS.\u0026minus;2\u003c/sup\u003e:N\u003csub\u003e\u0026delta;\u003c/sub\u003e \u0026ndash; T\u003csup\u003eGPS.+1\u003c/sup\u003e:H\u0026gamma; (2.10 \u0026Aring; \u0026minus;\u0026thinsp;1.0 \u0026Aring;) and T\u003csup\u003eGPS.+1\u003c/sup\u003e:O\u0026gamma; \u0026ndash; L\u003csup\u003eGPS.\u0026minus;1\u003c/sup\u003e:C (2.9 \u0026Aring; \u0026minus;\u0026thinsp;1.5 \u0026Aring;) with 0.1 \u0026Aring; spacing and distances \u0026Aring;, respectively. Low-energy intermediates were extracted for geometry optimization with DFT QM/MM.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eDFT QM/MM for Generating Intermediate, Product States and coordinate scans\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eQM/MM DFT was performed with TURBOMOLE 7.4\u003csup\u003e74\u0026ndash;77\u003c/sup\u003e and CHARMM46b2\u003csup\u003e78\u003c/sup\u003e in conjunction with the CHARMM/TURBOMOLE interface\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e79\u003c/span\u003e\u003c/sup\u003e. During setup, initial molecular orbitals were guessed by the Extended H\u0026uuml;ckel method\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e80\u003c/span\u003e\u003c/sup\u003e and the corresponding system charge of 0 (neutral glutamate, EH) or -1 (charged glutamate, E-). Geometry minimizations were performed by first applying the Weigend and Ahlrichs basis set def2-SVP\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e81\u003c/span\u003e\u003c/sup\u003e with the B3-LYP functional\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e70\u003c/span\u003e\u0026ndash;\u003cspan class=\"CitationRef\"\u003e72\u003c/span\u003e\u003c/sup\u003e followed by geometry optimization/minimization with def2-TZVP\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e73\u003c/span\u003e\u003c/sup\u003e and the B3-LYP functional; all employing the subtractive QM/MM scheme and link atoms placed on peptide backbone C-C bonds. Coordinate scans followed a similar system, however the semi-empirical DFTB3 was used for QM treatment\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e. The MM system was treated with the CHARMM36 force field with a flexible sphere of all residues with at least one atom within 5.0 \u0026Aring; of the QM selections and all atoms outside the sphere fixed. Self-consistent field (SCF) threshold was set to 10\u003csup\u003e\u0026minus;\u0026thinsp;7\u003c/sup\u003e and we used the spherical grid of size m3.\u003c/p\u003e\n\u003cp\u003eDue to the ambiguity of the E656\u003csup\u003eH6.50\u003c/sup\u003e protonation state, two separate QM/MM systems were generated based on both states. By following the initial proposed reaction path, one-, two- or multidimensional pathways were generated by incrementally moving the atoms involved in the respective step by intervals of 0.1\u0026ndash;0.2 \u0026Aring; per iteration, applying distance restraints with force constants of 1000 kJ/mol/nm\u0026sup2; and minimizing the structure iteratively until reaching the next reaction state. First, from the reactant conformation R the hydroxide moiety H\u003csub\u003e\u0026gamma;\u003c/sub\u003e of T838\u003csup\u003eGPS.+1\u003c/sup\u003e was moved towards H836\u003csup\u003eGPS.\u0026minus;2\u003c/sup\u003e, while simultaneously the distance between T838\u003csup\u003eGPS.+1\u003c/sup\u003e O\u003csub\u003e\u0026gamma;\u003c/sub\u003e and L837\u003csup\u003eGPS.\u0026minus;1\u003c/sup\u003e peptide carbon was decreased to facilitate the tetrahedral hydroxyl-oxazolidine intermediate T1 following nucleophilic attack on the peptide bond. The next step consisted of proton addition to the T838\u003csup\u003eGPS.+1\u003c/sup\u003e nitrogen from a nearby water molecule within the five-membered ring, followed by distance increase between T838\u003csup\u003eGPS.+1\u003c/sup\u003e nitrogen and L837\u003csup\u003eGPS.\u0026minus;1\u003c/sup\u003e peptide carbon to complete of the N-O acyl shift, resulting in the ester intermediate E. The resulting hydroxyl anion deprotonated the E656 sidechain in the systems containing protonated E656\u003csup\u003eH6.50\u003c/sup\u003e. The water attack on the ester was facilitated by either moving the free hydroxide ion directly towards the L837\u003csup\u003eGPS.\u0026minus;1\u003c/sup\u003e ester carbon or moving the intact water towards the attack, resulting in a second tetrahedral intermediate T2, with a proton transferred to the free neighboring amino moiety if water was attacking. During T2 minimization \u0026ndash; first with restraint of the attacking water oxygen to the carbon atom and subsequently without restraints, various proton transfers occurred and the final products P were generated.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eEster intermediate parameterization\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe leucine-threonine ester (LTE) molecule was extracted and minimized. Link atoms (peptide bond) were simply capped by uncharged hydrogens. Charges and parameters were taken from the respective amino acids in accordance with the CHARMM36 force field parameters where appropriate, leaving the ester bond and its immediate neighbors to parameterize. Point charges, bonds and angles were parameterized using the force field Toolkit (ffTk)\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e82\u003c/span\u003e\u003c/sup\u003e plugin in VMD 1.9.4alpha\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e83\u003c/span\u003e\u003c/sup\u003e coupled with the individual QM runs performed by ORCA 5.0.3\u003csup\u003e84\u003c/sup\u003e. Upon occurring steric clashes, the torsion intervals were reduced for the generation of dihedral parameters for the dihedrals involved in the ester bond. The resulting parameter files were modified for CHARMM to detect the engineered residue as an artificial amino acid with the additional caveat of not being an alpha-amino acid within the backbone.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eEster intermediate MD\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eWith the LTE topology and parameter files, ADGRL1-GAIN-LTE systems were set-up by taking the generated ester intermediate frames minimized via DFT on the B3-LYP/def2-TZVP level and renaming the atoms involved in the new LTE residue according to the topology file via python script. MD systems were then generated using CHARMM-GUI\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e65\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e85\u003c/span\u003e\u003c/sup\u003e with identical parameters to the uncleaved ADGRL1-GAIN model systems. Simulations were run in triplicate for 1,000 ns with H836\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e as HSD, HSE and HSP, with E656 protonated and unprotonated, respectively, using the NAMD2 package on GPU\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e86\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eBiochemical Methods\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003ePlasmids\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eAn N-terminally HA-tagged and C-terminally FLAG-tagged rat LPHN1/ADGRL1 cDNA (gift from Simone Pr\u0026ouml;mel) was subcloned into pcDNA3.1 by sequence-ligation-independent cloning with the following primers: 5\u0026rsquo;-ACTCACTATAGGGAGACCCAAGCTTGCCACCATGGCCCGCTTG and 5\u0026rsquo;-AGTGTGATGGATATCTGCAGAATTCTTACTTATCGTCGTCATCCTTGTAATCACC. GAIN domain cleavage-deficient H838S mutant, and all E656 mutants were generated by Genscript.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eExpression analyses of receptors\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eHEK293T cells were \u0026shy;maintained, transfected and lysed as previously described. Lysates were mixed with Laemmli buffer, and directly subjected to SDS-PAGE and Western blotting. Receptor expressions were detected by first incubating the nitrocellulose membrane overnight at 4 \u003csup\u003eo\u003c/sup\u003eC with rabbit anti-HA primary antibody (1:1000, Cell Signaling, Cat# 3724) or anti-TwinStrepII\u0026reg; primary antibody (1:500, iba Life Sciences, Cat# 2-1507-001), then a 1-hour incubation in room temperature with IRDye\u0026reg; 680RD goat anti-rabbit secondary antibody (1:15000, Licor, Cat# 926-68071). Detection of \u0026beta;-tubulin was done by similar incubation approach, with mouse anti-\u0026beta;-tubulin primary antibody (1:5000, DSHB, Clone E7) and IRDye\u0026reg; 800CW goat anti-mouse secondary antibody (1:15000, Licor, Cat# 926-32210). Visualizations of the immunosignals were done by the Licor Odyssey\u0026reg; XF Dual-Mode Imaging System.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eMolecular cloning and protein expression of ADGRL1 GAIN domain constructs\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eA DNA fragment encoding residues 460\u0026ndash;849 of the \u003cem\u003erat\u003c/em\u003e ADGRL1 ECR was synthesized by Thermo Fisher GeneArt. Point mutation of the \u003cem\u003er\u003c/em\u003eL1 gene was performed on the pMA vector (ThermoFisher Scientific) by \u003cem\u003eQuik Change mutagenesis\u003c/em\u003e. The HormR and GAIN domains of the wild-type and rat L1 point mutants, comprising residues 460\u0026ndash;849, were subcloned into the AgeI\u0026ndash;KpnI site of the pHLsec vector via PCR amplification. A C-terminal Fc tag followed by a hexahistidine sequence was added to increase the mass of the C-terminal fragment generated by GPS cleavage. HEK293T cells were transfected with polyethyleneimine (PEI) at a mass concentration 1.5 times higher than that of the pHLsec expression plasmid. Expression was performed in 10 ml volume for two days at 37\u0026deg;C in a humid atmosphere containing 5% CO\u003csub\u003e2\u003c/sub\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eProtein purification and analysis\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eElution buffer (50 mM Tris-HCl, pH 8; 300 mM NaCl; 500 mM imidazole) was added to the harvested expression medium yielding a final imidazole concentration of 30 mM. Then, 100 \u0026micro;l of washed His-bead slurry (Cat. No. 30410, Qiagen) was added after which the samples were incubated for one hour at 4\u0026deg;C with shaking. The beads were collected by centrifugation at 10000 \u0026times;g. The beads were then washed four times with 1 ml of wash buffer (50 mM Tris-HCl, pH 8; 300 mM NaCl; 30 mM imidazole) and the proteins were eluted by adding 250 \u0026micro;l of elution buffer using an incubation time of two minutes at room temperature. After centrifugation the supernatant was collected. 20 \u0026micro;l of the samples were denatured by adding 5 \u0026micro;l of SDS sample buffer (250 mM Tris-HCl, pH 6.8; 50% (v/v) glycerol; 4% (w/v) SDS; 25% (v/v) \u0026beta;-mercaptoethanol; 0.25% (w/v) bromophenol blue) and boiling at 95\u0026deg;C for 600 seconds. The SDS-PAGE gels were analyzed by semi-dry western blotting on methanol-activated PVDF membranes (Cat. No. 10600023, Cytiva Amersham). After skimmed milk blocking and incubation with anti-His-HRP antibodies (Cat.No. 11965085001, Roche), the proteins were detected using a luminol-peroxidase mixture (Cat.No. 10308449, Cytiva Amersham). Chemiluminescence was measured using a Biostep Celvin\u0026reg; S 420 chemiluminescense imager and intensities were scaled using the program ImageJ to a maximum pixel saturation of 65535. The resulting band intensities were quantified using GelAnalyzer software (version 23.1.1, available at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ewww.gelanalyzer.com\u003c/span\u003e\u003c/span\u003e from Istvan Lazar Jr., PhD and Istvan Lazar Sr., PhD, CSc) using the valley-to-valley analysis method. The plot and the statistical depiction were done by using GraphPad Prism (version 10.4.1 for Windows, GraphPad Software, Boston, Massachusetts USA).\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was funded by the Deutsche Forschungsgemeinschaft (DFG) through CRC 1423, project number 421152132 (project A06 to T.L. and N.S., projects C01 and Z04 to P.W.H.). \u0026nbsp;The authors gratefully acknowledge the scientific support and HPC resources provided by the Erlangen National High Performance Computing Center (NHR@FAU) of the Friedrich-Alexander-Universit\u0026auml;t Erlangen-N\u0026uuml;rnberg (FAU) under the NHR project p101ae. NHR funding is provided by federal and Bavarian state authorities. NHR@FAU hardware is partially funded by the German Research Foundation (DFG) \u0026ndash; 440719683. We thank Paolo Carloni and Emiliano Ippoliti from Forschungszentrum J\u0026uuml;lich GmbH for their support and access to their HPC and QM/MM software infrastructure. We like to acknowledge Ville Kaila and group for their QM/MM tutorial and the interface python script, which we modified to be compatible with Python 3.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eF.S.: Conceptualization; Methodology; Validation; Investigation(computational); Software; Formal Analysis; Visualization; Writing \u0026ndash; original draft; Writing \u0026ndash; review and editing; Data Curation\u003c/p\u003e\n\u003cp\u003eY.K.C.: Investigation(biochemical); Visualization\u003c/p\u003e\n\u003cp\u003eR.S.: Investigation(biochemical); Visualization\u003c/p\u003e\n\u003cp\u003eF.P.: Investigation(biochemical)\u003c/p\u003e\n\u003cp\u003eN.S.: Conceptualization; Writing \u0026ndash; review and editing\u003c/p\u003e\n\u003cp\u003eP.W.H.: Conceptualization; Validation; Writing \u0026ndash; review and editing; Supervision; Project administration; Funding acquisition\u003c/p\u003e\n\u003cp\u003eT.L.: Conceptualization; Writing \u0026ndash; review and editing\u003c/p\u003e\n\u003cp\u003eH.B.: Investigation(computational); Writing \u0026ndash; review and editing;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe generated GAIN domain models generated in this study, QM/MM input and output file, MD models and trajectories, LTE MD parameters and additional data have been deposited in the online repository zenodo under accession code DOI: 10.5281/zenodo.14445825. The PDB entries used in this study are available under the following accession codes: 4DLQ, 4DLO, 1MCT. The original template Python script for TURBOMOLE execution is available under https://villekaila.com/news/. The structural models reviewed are available in the alphafold database under: P15941, Q5T601, Q8IZF2. The entries of proteins reviewed are available in the UniProt knowledge base under: O88917, P15941, Q5T601, Q8IZF2, O60242.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eSeufert, F., Chung, Y. K., Hildebrand, P. W. \u0026amp; Langenhan, T. 7TM domain structures of adhesion GPCRs: what\u0026rsquo;s new and what\u0026rsquo;s missing? \u003cem\u003eTrends in Biochemical Sciences\u003c/em\u003e \u003cstrong\u003e48\u003c/strong\u003e, 726\u0026ndash;739 (2023).\u003c/li\u003e\n\u003cli\u003eRosa, M., Noel, T., Harris, M. \u0026amp; Ladds, G. Emerging roles of adhesion G protein-coupled receptors. \u003cem\u003eBiochem. Soc. Trans.\u003c/em\u003e \u003cstrong\u003e49\u003c/strong\u003e, 1695\u0026ndash;1709 (2021).\u003c/li\u003e\n\u003cli\u003eBondarev, A. D. \u003cem\u003eet al.\u003c/em\u003e Opportunities and challenges for drug discovery in modulating Adhesion G protein-coupled receptor (GPCR) functions. \u003cem\u003eExpert Opin. Drug Discov.\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 1291\u0026ndash;1307 (2020).\u003c/li\u003e\n\u003cli\u003eMorgan, R. K. \u003cem\u003eet al.\u003c/em\u003e The expanding functional roles and signaling mechanisms of adhesion g protein\u0026ndash;coupled receptors. \u003cem\u003eAnnals of the New York Academy of Sciences\u003c/em\u003e \u003cstrong\u003e1456\u003c/strong\u003e, (2019).\u003c/li\u003e\n\u003cli\u003eLangenhan, T., Piao, X. \u0026amp; Monk, K. R. Adhesion G protein-coupled receptors in nervous system development and disease. \u003cem\u003eNature Reviews Neuroscience\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, 550\u0026ndash;561 (2016).\u003c/li\u003e\n\u003cli\u003eSeufert, F. \u003cem\u003eet al.\u003c/em\u003e Generic residue numbering of the GAIN domain of adhesion GPCRs. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 246 (2025).\u003c/li\u003e\n\u003cli\u003eBeliu, G. \u003cem\u003eet al.\u003c/em\u003e Tethered agonist exposure in intact adhesion/class B2 GPCRs through intrinsic structural flexibility of the GAIN domain. \u003cem\u003eMolecular Cell\u003c/em\u003e \u003cstrong\u003e81\u003c/strong\u003e, 905-921.e5 (2021).\u003c/li\u003e\n\u003cli\u003eAra\u0026ccedil;, D. \u003cem\u003eet al.\u003c/em\u003e A novel evolutionarily conserved domain of cell-adhesion GPCRs mediates autoproteolysis. \u003cem\u003eThe EMBO Journal\u003c/em\u003e \u003cstrong\u003e31\u003c/strong\u003e, 1364\u0026ndash;1378 (2012).\u003c/li\u003e\n\u003cli\u003eNieberler, M., Kittel, R. J., Petrenko, A. G., Lin, H.-H. \u0026amp; Langenhan, T. Control of Adhesion GPCR Function Through Proteolytic Processing. \u003cem\u003eHandbook of Experimental Pharmacology\u003c/em\u003e \u003cstrong\u003e234\u003c/strong\u003e, 83\u0026ndash;109 (2016).\u003c/li\u003e\n\u003cli\u003eLiebscher, I. \u003cem\u003eet al.\u003c/em\u003e A Tethered Agonist within the Ectodomain Activates the Adhesion G Protein-Coupled Receptors GPR126 and GPR133. \u003cem\u003eCell Reports\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 2018\u0026ndash;2026 (2014).\u003c/li\u003e\n\u003cli\u003eNazarko, O. \u003cem\u003eet al.\u003c/em\u003e A Comprehensive Mutagenesis Screen of the Adhesion GPCR Latrophilin-1/ADGRL1. \u003cem\u003eiScience\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 264\u0026ndash;278 (2018).\u003c/li\u003e\n\u003cli\u003eStoveken, H. M., Hajduczok, A. G., Xu, L. \u0026amp; Tall, G. G. Adhesion G protein-coupled receptors are activated by exposure of a cryptic tethered agonist. \u003cem\u003eProc. Natl. Acad. Sci.\u003c/em\u003e \u003cstrong\u003e112\u003c/strong\u003e, 6194\u0026ndash;6199 (2015).\u003c/li\u003e\n\u003cli\u003eWilde, C. \u003cem\u003eet al.\u003c/em\u003e The constitutive activity of the adhesion GPCR GPR114/ADGRG5 is mediated by its tethered agonist. \u003cem\u003eFASEB J.\u003c/em\u003e \u003cstrong\u003e30\u003c/strong\u003e, 666\u0026ndash;673 (2016).\u003c/li\u003e\n\u003cli\u003eFrenster, J. D. \u003cem\u003eet al.\u003c/em\u003e Functional impact of intramolecular cleavage and dissociation of adhesion G protein\u0026ndash;coupled receptor GPR133 (ADGRD1) on canonical signaling. \u003cem\u003eJ. Biol. Chem.\u003c/em\u003e \u003cstrong\u003e296\u003c/strong\u003e, 100798 (2021).\u003c/li\u003e\n\u003cli\u003eAra\u0026ccedil;, D. \u003cem\u003eet al.\u003c/em\u003e Dissecting signaling and functions of adhesion G protein-coupled receptors. \u003cem\u003eAnnals of the New York Academy of Sciences\u003c/em\u003e \u003cstrong\u003e1276\u003c/strong\u003e, 1\u0026ndash;25 (2012).\u003c/li\u003e\n\u003cli\u003eScholz, N. \u003cem\u003eet al.\u003c/em\u003e Molecular sensing of mechano- and ligand-dependent adhesion GPCR dissociation. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e615\u003c/strong\u003e, 945\u0026ndash;953 (2023).\u003c/li\u003e\n\u003cli\u003eVizurraga, A., Adhikari, R., Yeung, J., Yu, M. \u0026amp; Tall, G. G. Mechanisms of adhesion G protein\u0026ndash;coupled receptor activation.\u003c/li\u003e\n\u003cli\u003eBarros-\u0026Aacute;lvarez, X. \u003cem\u003eet al.\u003c/em\u003e The tethered peptide activation mechanism of adhesion GPCRs. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e604\u003c/strong\u003e, 757\u0026ndash;762 (2022).\u003c/li\u003e\n\u003cli\u003eKoh, J. T. \u003cem\u003eet al.\u003c/em\u003e Extracellular fragment of brain-specific angiogenesis inhibitor 1 suppresses endothelial cell proliferation by blocking \u0026alpha; v\u0026beta;5 integrin. \u003cem\u003eExperimental Cell Research\u003c/em\u003e \u003cstrong\u003e294\u003c/strong\u003e, 172\u0026ndash;184 (2004).\u003c/li\u003e\n\u003cli\u003eVallon, M. \u0026amp; Essler, M. Proteolytically Processed Soluble Tumor Endothelial Marker (TEM) 5 Mediates Endothelial Cell Survival during Angiogenesis by Linking Integrin \u0026alpha;v\u0026beta;3 to Glycosaminoglycans *. \u003cem\u003eJournal of Biological Chemistry\u003c/em\u003e \u003cstrong\u003e281\u003c/strong\u003e, 34179\u0026ndash;34188 (2006).\u003c/li\u003e\n\u003cli\u003eLehmann, L., Gro\u0026szlig;, V. E., Behlendorf, R. \u0026amp; Pr\u0026ouml;mel, S. The N terminus-only function of adhesion GPCRs: emerging concepts. \u003cem\u003eTrends Pharmacol. Sci.\u003c/em\u003e (2025) doi:10.1016/j.tips.2025.01.004.\u003c/li\u003e\n\u003cli\u003eKleinau, G. \u003cem\u003eet al.\u003c/em\u003e Intramolecular activity regulation of adhesion GPCRs in light of recent structural and evolutionary information. \u003cem\u003ePharmacol. Res.\u003c/em\u003e \u003cstrong\u003e197\u003c/strong\u003e, 106971 (2023).\u003c/li\u003e\n\u003cli\u003eScholz, N. \u003cem\u003eet al.\u003c/em\u003e Mechano-dependent signaling by latrophilin/CIRL quenches cAMP in proprioceptive neurons. \u003cem\u003eeLife\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, (2017).\u003c/li\u003e\n\u003cli\u003eKishore, A. \u0026amp; Hall, R. A. Versatile Signaling Activity of Adhesion GPCRs. \u003cem\u003eHandbook of experimental pharmacology\u003c/em\u003e \u003cstrong\u003e234\u003c/strong\u003e, 127\u0026ndash;146 (2016).\u003c/li\u003e\n\u003cli\u003eLiebscher, I., Sch\u0026ouml;neberg, T. \u0026amp; Thor, D. Stachel-mediated activation of adhesion G protein-coupled receptors: insights from cryo-EM studies. \u003cem\u003eSignal transduction and targeted therapy\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 227 (2022).\u003c/li\u003e\n\u003cli\u003ePohl, F. \u003cem\u003eet al.\u003c/em\u003e Structural basis of GAIN domain autoproteolysis and cleavage-resistance in the adhesion G-protein coupled receptors. \u003cem\u003ebioRxiv\u003c/em\u003e (2023) doi:10.1101/2023.03.12.532270.\u003c/li\u003e\n\u003cli\u003eSenn, H. M. \u0026amp; Thiel, W. QM/MM Methods for Biomolecular Systems. \u003cem\u003eAngew. Chem. Int. Ed.\u003c/em\u003e \u003cstrong\u003e48\u003c/strong\u003e, 1198\u0026ndash;1229 (2009).\u003c/li\u003e\n\u003cli\u003eSenn, H. M. \u0026amp; Thiel, W. QM/MM studies of enzymes. \u003cem\u003eCurr. Opin. Chem. Biol.\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 182\u0026ndash;187 (2007).\u003c/li\u003e\n\u003cli\u003e\u0026Aacute;sgeirsson, B. \u0026amp; Cekan, P. Microscopic rate‐constants for substrate binding and acylation in cold‐adaptation of trypsin I from Atlantic cod. \u003cem\u003eFEBS Lett.\u003c/em\u003e \u003cstrong\u003e580\u003c/strong\u003e, 4639\u0026ndash;4644 (2006).\u003c/li\u003e\n\u003cli\u003eFu, C. \u003cem\u003eet al.\u003c/em\u003e Unveiling Mechanical Activation: GAIN Domain Unfolding and Dissociation in Adhesion GPCRs. \u003cem\u003eNano Lett.\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, 9179\u0026ndash;9186 (2023).\u003c/li\u003e\n\u003cli\u003eKordon, S. P. \u003cem\u003eet al.\u003c/em\u003e Conformational coupling between extracellular and transmembrane domains modulates holo-adhesion GPCR function. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 10545 (2024).\u003c/li\u003e\n\u003cli\u003eMat\u0026uacute;\u0026scaron;, D., Lopez, J. M., Sando, R. C. \u0026amp; S\u0026uuml;dhof, T. C. Essential Role of Latrophilin-1 Adhesion GPCR Nanoclusters in Inhibitory Synapses. \u003cem\u003eJ. Neurosci.\u003c/em\u003e \u003cstrong\u003e44\u003c/strong\u003e, e1978232024 (2024).\u003c/li\u003e\n\u003cli\u003eChhabra, K. H. \u003cem\u003eet al.\u003c/em\u003e ADGRL1 is a glucose receptor involved in mediating energy and glucose homeostasis. \u003cem\u003eDiabetologia\u003c/em\u003e \u003cstrong\u003e67\u003c/strong\u003e, 170\u0026ndash;189 (2024).\u003c/li\u003e\n\u003cli\u003eDietzsch, A. N. \u003cem\u003eet al.\u003c/em\u003e Dysfunction of the adhesion G protein-coupled receptor latrophilin 1 (ADGRL1/LPHN1) increases the risk of obesity. \u003cem\u003eSignal Transduct. Target. Ther.\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 103 (2024).\u003c/li\u003e\n\u003cli\u003eIchtchenko, K. \u003cem\u003eet al.\u003c/em\u003e A novel ubiquitously expressed \u0026alpha;-latrotoxin receptor is a member of the CIRL family of G-protein-coupled receptors. \u003cem\u003eJournal of Biological Chemistry\u003c/em\u003e \u003cstrong\u003e274\u003c/strong\u003e, 5491\u0026ndash;5498 (1999).\u003c/li\u003e\n\u003cli\u003eKrasnoperov, V. \u003cem\u003eet al.\u003c/em\u003e Dissociation of the Subunits of the calcium-independent receptor of \u0026alpha;-Latrotoxin as a result of two-step proteolysis. \u003cem\u003eBiochemistry\u003c/em\u003e \u003cstrong\u003e48\u003c/strong\u003e, 3230\u0026ndash;3238 (2009).\u003c/li\u003e\n\u003cli\u003eOlsson, M. H. M., S\u0026oslash;ndergaard, C. R., Rostkowski, M. \u0026amp; Jensen, J. H. PROPKA3: Consistent Treatment of Internal and Surface Residues in Empirical p \u003cem\u003eK\u003c/em\u003e \u003csub\u003ea\u003c/sub\u003e Predictions. \u003cem\u003eJournal of Chemical Theory and Computation\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 525\u0026ndash;537 (2011).\u003c/li\u003e\n\u003cli\u003eK\u0026ouml;nig, P. H., Hoffmann, M., Frauenheim, Th. \u0026amp; Cui, Q. A Critical Evaluation of Different QM/MM Frontier Treatments with SCC-DFTB as the QM Method. \u003cem\u003eJ. Phys. Chem. B\u003c/em\u003e \u003cstrong\u003e109\u003c/strong\u003e, 9082\u0026ndash;9095 (2005).\u003c/li\u003e\n\u003cli\u003eGaus, M., Cui, Q. \u0026amp; Elstner, M. DFTB3: Extension of the Self-Consistent-Charge Density-Functional Tight-Binding Method (SCC-DFTB). \u003cem\u003eJournal of Chemical Theory and Computation\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 931\u0026ndash;948 (2011).\u003c/li\u003e\n\u003cli\u003eSpiegelman, F. \u003cem\u003eet al.\u003c/em\u003e Density-functional tight-binding: basic concepts and applications to molecules and clusters. \u003cem\u003eAdv. Phys.: X\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, 1710252 (2020).\u003c/li\u003e\n\u003cli\u003eLin, H. H. \u003cem\u003eet al.\u003c/em\u003e Autocatalytic cleavage of the EMR2 receptor occurs at a conserved G protein-coupled receptor proteolytic site motif. \u003cem\u003eJournal of Biological Chemistry\u003c/em\u003e \u003cstrong\u003e279\u003c/strong\u003e, 31823\u0026ndash;31832 (2004).\u003c/li\u003e\n\u003cli\u003eJohansson, D. G. A. \u003cem\u003eet al.\u003c/em\u003e Protein autoproteolysis: Conformational strain linked to the rate of peptide cleavage by the pH dependence of the N \u0026rarr; O acyl shift reaction. \u003cem\u003eJournal of the American Chemical Society\u003c/em\u003e \u003cstrong\u003e131\u003c/strong\u003e, 9475\u0026ndash;9477 (2009).\u003c/li\u003e\n\u003cli\u003eJohansson, D. G. A., Macao, B., Sandberg, A. \u0026amp; H\u0026auml;rd, T. SEA Domain Autoproteolysis Accelerated by Conformational Strain: Mechanistic Aspects. \u003cem\u003eJournal of Molecular Biology\u003c/em\u003e \u003cstrong\u003e377\u003c/strong\u003e, 1130\u0026ndash;1143 (2008).\u003c/li\u003e\n\u003cli\u003eD\u0026iacute;az-Cervantes, E., Robles, J., Sol\u0026agrave;, M. \u0026amp; Swart, M. The peptide bond rupture mechanism in the serine proteases: an in silico study based on sequential scale models. \u003cem\u003ePhys. Chem. Chem. Phys.\u003c/em\u003e \u003cstrong\u003e25\u003c/strong\u003e, 8043\u0026ndash;8049 (2023).\u003c/li\u003e\n\u003cli\u003eMihaylov, T. T., Parac‐Vogt, T. N. \u0026amp; Pierloot, K. A Mechanistic Study of the Spontaneous Hydrolysis of Glycylserine as the Simplest Model for Protein Self‐Cleavage. \u003cem\u003eChem. A Eur. J.\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 456\u0026ndash;466 (2014).\u003c/li\u003e\n\u003cli\u003eEls\u0026auml;sser, B. \u0026amp; Goettig, P. Molecular Sciences Mechanisms of Proteolytic Enzymes and Their Inhibition in QM/MM Studies. (2021) doi:10.3390/ijms22063232.\u003c/li\u003e\n\u003cli\u003eViegas, M. F., Neves, R. P. P., Ramos, M. J. \u0026amp; Fernandes, P. A. QM/MM Study of the Reaction Mechanism of Thermophilic Glucuronoyl Esterase for Biomass Treatment. \u003cem\u003eChemPhysChem\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, e202200269 (2022).\u003c/li\u003e\n\u003cli\u003eDu, S. \u003cem\u003eet al.\u003c/em\u003e Conformational Ensembles Reveal the Origins of Serine Protease Catalysis. \u003cem\u003ebioRxiv\u003c/em\u003e 2024.02.28.582624 (2024) doi:10.1101/2024.02.28.582624.\u003c/li\u003e\n\u003cli\u003eB\u0026uuml;rgi, H. B., Dunitz, J. D., Lehn, J. M. \u0026amp; Wipff, G. Stereochemistry of reaction paths at carbonyl centres. \u003cem\u003eTetrahedron\u003c/em\u003e \u003cstrong\u003e30\u003c/strong\u003e, 1563\u0026ndash;1572 (1974).\u003c/li\u003e\n\u003cli\u003eReed, A. E., Weinstock, R. B. \u0026amp; Weinhold, F. Natural population analysis. \u003cem\u003eJ. Chem. Phys.\u003c/em\u003e \u003cstrong\u003e83\u003c/strong\u003e, 735\u0026ndash;746 (1985).\u003c/li\u003e\n\u003cli\u003eWeinhold, F. \u0026amp; Labdus, C. R. Natural Bond Orbitals and Extensions of Localized Bonding Concepts. \u003cem\u003eChem Educ Res Pr.\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 91\u0026ndash;104 (2001).\u003c/li\u003e\n\u003cli\u003eMerz, T. \u003cem\u003eet al.\u003c/em\u003e Autoproteolytic and Catalytic Mechanisms for the \u0026beta;-Aminopeptidase BapA\u0026mdash;A Member of the Ntn Hydrolase Family. \u003cem\u003eStructure\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 1850\u0026ndash;1860 (2012).\u003c/li\u003e\n\u003cli\u003eElsässer, B. \u003cem\u003eet al.\u003c/em\u003e Distinct Roles of Catalytic Cysteine and Histidine in the Protease and Ligase Mechanisms of Human Legumain As Revealed by DFT-Based QM/MM Simulations. \u003cem\u003eACS Catal.\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 5585\u0026ndash;5593 (2017).\u003c/li\u003e\n\u003cli\u003eRadisky, E. S., Lee, J. M., Lu, C.-J. K. \u0026amp; Koshland, D. E. Insights into the serine protease mechanism from atomic resolution structures of trypsin reaction intermediates. (2006).\u003c/li\u003e\n\u003cli\u003eXu, Q., Buckley, D., Guan, C. \u0026amp; Guo, H. C. Structural insights into the mechanism of intramolecular proteolysis. \u003cem\u003eCell\u003c/em\u003e \u003cstrong\u003e98\u003c/strong\u003e, 651\u0026ndash;661 (1999).\u003c/li\u003e\n\u003cli\u003eRosenblum, J. S. \u0026amp; Blobel, G. Autoproteolysis in nucleoporin biogenesis. \u003cem\u003eProceedings of the National Academy of Sciences of the United States of America\u003c/em\u003e \u003cstrong\u003e96\u003c/strong\u003e, 11370\u0026ndash;5 (1999).\u003c/li\u003e\n\u003cli\u003eKarplus, M. Development of Multiscale Models for Complex Chemical Systems: From H+H2 to Biomolecules (Nobel Lecture). \u003cem\u003eAngew. Chem. Int. Ed.\u003c/em\u003e \u003cstrong\u003e53\u003c/strong\u003e, 9992\u0026ndash;10005 (2014).\u003c/li\u003e\n\u003cli\u003eKubař, T., Elstner, M. \u0026amp; Cui, Q. Hybrid Quantum Mechanical/Molecular Mechanical Methods For Studying Energy Transduction in Biomolecular Machines. \u003cem\u003eAnnu. Rev. Biophys.\u003c/em\u003e \u003cstrong\u003e52\u003c/strong\u003e, 525\u0026ndash;551 (2023).\u003c/li\u003e\n\u003cli\u003eBuller, A. R., Freeman, M. F., Wright, N. T., Schildbach, J. F. \u0026amp; Townsend, C. A. Insights into cis-autoproteolysis reveal a reactive state formed through conformational rearrangement. \u003cem\u003eProc. Natl. Acad. Sci.\u003c/em\u003e \u003cstrong\u003e109\u003c/strong\u003e, 2308\u0026ndash;2313 (2012).\u003c/li\u003e\n\u003cli\u003eD\u0026rsquo;Osualdo, A. \u003cem\u003eet al.\u003c/em\u003e CARD8 and NLRP1 Undergo Autoproteolytic Processing through a ZU5-Like Domain. \u003cem\u003ePLOS ONE\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, e27396 (2011).\u003c/li\u003e\n\u003cli\u003eLiao, Y., Pei, J., Cheng, H. \u0026amp; Grishin, N. V. An ancient autoproteolytic domain found in GAIN, ZU5 and Nucleoporin98. \u003cem\u003eJournal of Molecular Biology\u003c/em\u003e \u003cstrong\u003e426\u003c/strong\u003e, 3935\u0026ndash;3945 (2014).\u003c/li\u003e\n\u003cli\u003eKim, J. K. \u003cem\u003eet al.\u003c/em\u003e Insight into autoproteolytic activation from the structure of cephalosporin acylase: A protein with two proteolytic chemistries. \u003cem\u003eProc. Natl. Acad. Sci.\u003c/em\u003e \u003cstrong\u003e103\u003c/strong\u003e, 1732\u0026ndash;1737 (2006).\u003c/li\u003e\n\u003cli\u003eSandberg, A., Johansson, D. G. A., Macao, B. \u0026amp; H\u0026auml;rd, T. SEA Domain Autoproteolysis Accelerated by Conformational Strain: Energetic Aspects. \u003cem\u003eJournal of Molecular Biology\u003c/em\u003e \u003cstrong\u003e377\u003c/strong\u003e, 1117\u0026ndash;1129 (2008).\u003c/li\u003e\n\u003cli\u003eCalinsky, R. \u0026amp; Levy, Y. Histidine in Proteins: pH-Dependent Interplay between \u0026pi;\u0026ndash;\u0026pi;, Cation\u0026ndash;\u0026pi;, and CH\u0026ndash;\u0026pi; Interactions. \u003cem\u003eJ. Chem. Theory Comput.\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 6930\u0026ndash;6945 (2024).\u003c/li\u003e\n\u003cli\u003eLee, J. \u003cem\u003eet al.\u003c/em\u003e CHARMM-GUI Input Generator for NAMD, GROMACS, AMBER, OpenMM, and CHARMM/OpenMM Simulations Using the CHARMM36 Additive Force Field. \u003cem\u003eJournal of Chemical Theory and Computation\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 405\u0026ndash;413 (2016).\u003c/li\u003e\n\u003cli\u003eSpoel, D. V. D. \u003cem\u003eet al.\u003c/em\u003e GROMACS: Fast, flexible, and free. \u003cem\u003eJournal of Computational Chemistry\u003c/em\u003e \u003cstrong\u003e26\u003c/strong\u003e, 1701\u0026ndash;1718 (2005).\u003c/li\u003e\n\u003cli\u003eAbraham, M. J. \u003cem\u003eet al.\u003c/em\u003e Gromacs: High performance molecular simulations through multi-level parallelism from laptops to supercomputers. \u003cem\u003eSoftwareX\u003c/em\u003e \u003cstrong\u003e1\u0026ndash;2\u003c/strong\u003e, 19\u0026ndash;25 (2015).\u003c/li\u003e\n\u003cli\u003eHuang, J. \u0026amp; MacKerell, A. D. CHARMM36 all-atom additive protein force field: Validation based on comparison to NMR data. \u003cem\u003eJournal of Computational Chemistry\u003c/em\u003e \u003cstrong\u003e34\u003c/strong\u003e, 2135\u0026ndash;2145 (2013).\u003c/li\u003e\n\u003cli\u003ePedregosa, F. \u003cem\u003eet al.\u003c/em\u003e Scikit-Learn: Machine Learning in Python. \u003cem\u003eJ. Mach. Learn. Res.\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 2825\u0026ndash;2830 (2011).\u003c/li\u003e\n\u003cli\u003eBecke, A. D. Density‐functional thermochemistry. III. The role of exact exchange. \u003cem\u003eJ. Chem. Phys.\u003c/em\u003e \u003cstrong\u003e98\u003c/strong\u003e, 5648\u0026ndash;5652 (1993).\u003c/li\u003e\n\u003cli\u003eStephens, P. J., Devlin, F. J., Chabalowski, C. F. \u0026amp; Frisch, M. J. Ab Initio Calculation of Vibrational Absorption and Circular Dichroism Spectra Using Density Functional Force Fields. \u003cem\u003eThe Journal of Physical Chemistry\u003c/em\u003e \u003cstrong\u003e98\u003c/strong\u003e, 11623\u0026ndash;11627 (1994).\u003c/li\u003e\n\u003cli\u003eLee, C., Yang, W. \u0026amp; Parr, R. G. Development of the Colle-Salvetti correlation-energy formula into a functional of the electron density. \u003cem\u003ePhysical Review B\u003c/em\u003e \u003cstrong\u003e37\u003c/strong\u003e, 785\u0026ndash;789 (1988).\u003c/li\u003e\n\u003cli\u003eWeigend, F., H\u0026auml;ser, M., Patzelt, H. \u0026amp; Ahlrichs, R. RI-MP2: optimized auxiliary basis sets and demonstration of efficiency. \u003cem\u003eChem. Phys. Lett.\u003c/em\u003e \u003cstrong\u003e294\u003c/strong\u003e, 143\u0026ndash;152 (1998).\u003c/li\u003e\n\u003cli\u003eGmbH, T., Karlsruhe, U. of \u0026amp; GmbH, F. K. \u003cem\u003eTURBOMOLE V7.4\u003c/em\u003e. (2019).\u003c/li\u003e\n\u003cli\u003eAhlrichs, R., B\u0026auml;r, M., H\u0026auml;ser, M., Horn, H. \u0026amp; K\u0026ouml;lmel, C. Electronic structure calculations on workstation computers: The program system turbomole. \u003cem\u003eChem. Phys. Lett.\u003c/em\u003e \u003cstrong\u003e162\u003c/strong\u003e, 165\u0026ndash;169 (1989).\u003c/li\u003e\n\u003cli\u003eTreutler, O. \u0026amp; Ahlrichs, R. Efficient molecular numerical integration schemes. \u003cem\u003eJ. Chem. Phys.\u003c/em\u003e \u003cstrong\u003e102\u003c/strong\u003e, 346\u0026ndash;354 (1995).\u003c/li\u003e\n\u003cli\u003eH\u0026auml;ser, M. \u0026amp; Ahlrichs, R. Improvements on the direct SCF method. \u003cem\u003eJ. Comput. Chem.\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 104\u0026ndash;111 (1989).\u003c/li\u003e\n\u003cli\u003eBrooks, B. R. \u003cem\u003eet al.\u003c/em\u003e CHARMM: The biomolecular simulation program. \u003cem\u003eJournal of Computational Chemistry\u003c/em\u003e \u003cstrong\u003e30\u003c/strong\u003e, 1545\u0026ndash;1614 (2009).\u003c/li\u003e\n\u003cli\u003eRiahi, S. \u0026amp; Rowley, C. N. The CHARMM\u0026ndash;TURBOMOLE interface for efficient and accurate QM/MM molecular dynamics, free energies, and excited state properties. \u003cem\u003eJ. Comput. Chem.\u003c/em\u003e \u003cstrong\u003e35\u003c/strong\u003e, 2076\u0026ndash;2086 (2014).\u003c/li\u003e\n\u003cli\u003eHoffmann, R. An Extended H\u0026uuml;ckel Theory. I. Hydrocarbons. \u003cem\u003eJ. Chem. Phys.\u003c/em\u003e \u003cstrong\u003e39\u003c/strong\u003e, 1397\u0026ndash;1412 (1963).\u003c/li\u003e\n\u003cli\u003eSch\u0026auml;fer, A., Horn, H. \u0026amp; Ahlrichs, R. Fully optimized contracted Gaussian basis sets for atoms Li to Kr. \u003cem\u003eJ. Chem. Phys.\u003c/em\u003e \u003cstrong\u003e97\u003c/strong\u003e, 2571\u0026ndash;2577 (1992).\u003c/li\u003e\n\u003cli\u003eMayne, C. G., Saam, J., Schulten, K., Tajkhorshid, E. \u0026amp; Gumbart, J. C. Rapid parameterization of small molecules using the force field toolkit. \u003cem\u003eJournal of Computational Chemistry\u003c/em\u003e \u003cstrong\u003e34\u003c/strong\u003e, 2757\u0026ndash;2770 (2013).\u003c/li\u003e\n\u003cli\u003eHumphrey, W., Dalke, A. \u0026amp; Schulten, K. VMD: Visual molecular dynamics. \u003cem\u003eJournal of Molecular Graphics\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 33\u0026ndash;38 (1996).\u003c/li\u003e\n\u003cli\u003eNeese, F. The ORCA program system. \u003cem\u003eWiley Interdisciplinary Reviews: Computational Molecular Science\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 73\u0026ndash;78 (2012).\u003c/li\u003e\n\u003cli\u003eJo, S., Kim, T., Iyer, V. G. \u0026amp; Im, W. CHARMM-GUI: A web-based graphical user interface for CHARMM. \u003cem\u003eJournal of Computational Chemistry\u003c/em\u003e \u003cstrong\u003e29\u003c/strong\u003e, 1859\u0026ndash;1865 (2008).\u003c/li\u003e\n\u003cli\u003ePhillips, J. C. \u003cem\u003eet al.\u003c/em\u003e Scalable molecular dynamics on CPU and GPU architectures with NAMD. \u003cem\u003eJ. Chem. Phys.\u003c/em\u003e \u003cstrong\u003e153\u003c/strong\u003e, 044130 (2020).\u003c/li\u003e\n\u003cli\u003eWeinhold, F. Chemical Bonding as a Superposition Phenomenon. \u003cem\u003eJ. Chem. Educ.\u003c/em\u003e \u003cstrong\u003e76\u003c/strong\u003e, 1141 (1999).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[{"identity":"aa7dc864-0a66-446c-8139-8b1b7ce5ad0b","identifier":"10.13039/501100001659","name":"Deutsche Forschungsgemeinschaft","awardNumber":"421152132","order_by":0}],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"Leipzig University","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"GAIN Domain, adhesion GPCR, autoproteolysis, QM/MM, Molecular Dynamics, ADGRL1, Multiscale Modeling, Stachel peptide","lastPublishedDoi":"10.21203/rs.3.rs-7121662/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7121662/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eIn \u003cem\u003ecis\u003c/em\u003e-autoproteolysis, proteins self-catalyze cleavage of their peptide chains. In the GAIN domain of adhesion G protein-coupled receptors, \u003cem\u003ecis\u003c/em\u003e-autoproteolysis yields a tethered agonist critical for receptor activation. Contrary to current reaction models, the GAIN HL|T catalytic triad is insufficient for cleavage. Here, we investigated the cleavage mechanism in the rat ADGRL1 GAIN domain using multiscale modeling combining molecular dynamics and QM/MM simulations to investigate \u003cem\u003ecis\u003c/em\u003e-autoproteolysis as a once-in-a-lifetime event. We present an updated and unique GAIN domain \u003cem\u003ecis\u003c/em\u003e-autoproteolysis mechanism. The initial N-O acyl shift proceeds via a hydroxy-oxazolidine intermediate, sterically shifting subsequent reaction steps away from the triad base H\u003csup\u003eGPS.-2\u003c/sup\u003e. Water on its opposing side, coordinated by acidic residue E656\u003csup\u003eH6.50\u003c/sup\u003e, is essential for completing the N-O acyl shift facilitating ester hydrolysis. Our study provides detailed mechanistic insights into \u003cem\u003ecis\u003c/em\u003e-autoproteolysis, highlighting how conserved residues and structured water networks adjacent to the triad enable a chemically precise reaction within an inherently flexible protein not evolutionarily optimized for high catalytic rate. These findings deepen our understanding of aGPCR processing and may apply broadly to other autoproteolytic proteins, offering a framework to explore similarly elusive one-time cleavage events in biology.\u003c/p\u003e","manuscriptTitle":"Mechanistic insights into adhesion GPCR autoproteolysis by a multiscale computational approach","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-07-15 09:29:47","doi":"10.21203/rs.3.rs-7121662/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"aa5d6420-ed95-4a22-906e-b8713f4851f5","owner":[],"postedDate":"July 15th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":51508738,"name":"Computational Chemistry"},{"id":51508739,"name":"Molecular Biology"},{"id":51508740,"name":"Biophysics"}],"tags":[],"updatedAt":"2025-07-15T09:29:47+00:00","versionOfRecord":[],"versionCreatedAt":"2025-07-15 09:29:47","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7121662","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7121662","identity":"rs-7121662","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}