Unraveling Protein Complex Dynamics

Case study SPR for protein complex formation studies Dr. Elin Sivertsson1 , Dr. Alexander Fish2 , Dr. Linnea Nygren-Babol1 1 Cytiva, 2 Netherlands Cancer Institute Protein complexes play many important biological roles. Some protein complexes are stable, but others are more transient in nature. This presents a challenge: how can we study a rapidly dissociating protein complex before it falls apart? cytiva.com Solution A Solution C Solution B Solution D Running Solution E buffer Running buffer Biacore™ 1 series introduced a novel surface plasmon resonance (SPR) injection tool, Poly, which allows you to perform up to five sequential injections without any intermediate wash steps. This limits the delay and time for dissociation between injections of different interaction partners. Here, we present two different case studies that illustrate how Poly command can be used to investigate protein complex formation. Introduction Protein complexes Most proteins operate as multimeric complexes, rather than as individual proteins. These protein complexes regulate numerous cellular processes. Advances in proteomic research has led to the identification of thousands of protein complexes, and the number continues to grow. Protein complexes can be dynamic with transient interactions. It can be challenging to study them during their short lifetimes. Moreover, it can be difficult to unravel the binding processes of individual proteins within the complex. Poly command When studying rapidly dissociating protein complexes using SPR, it is crucial to avoid delays between the injections of different proteins. Otherwise, the first protein may dissociate before the next protein has reached the sensor surface, which could potentially lead to reduced binding (or no binding at all) of this protein and subsequent proteins. The Poly command enables the injection of three to five solutions in immediate sequence with no intermediate wash steps (Figure 1). The contact times for the injections are set separately, with a dissociation time for the last injected solution. The same flow rate and flow path are used for the entire command. The absence of wash steps between injections within the Poly command means that the next solution can be injected as soon as the injection of the previous solution has finished and before any significant dissociation has occurred. This gives you new possibilities to study the formation of protein complexes. Fig 1. The Poly command enables studies of complex formation by injection of three to five solutions (solution A to solution E) in sequence with no intermediate wash steps. 2 CY39216-26Oct23-CS MutS binding and ATP-dependent clamp formation. MutS MutL MutS recruits MutL. MutH nicks newly synthetised strand. MutL recruits the helicase UvrD, which separates the strands. The nicked strand is excised by exonuclease. DNA pol III repairs the gap. 1 2 3 4 5 6 MutH ATP ATP ATP ATP UvrD DNA Pol III SSB SSB Exo Case study 1: DNA mismatch repair complex formation DNA mismatch repair (1) is an evolutionarily conserved process for the detection and removal of DNA mismatches introduced during replication. Figure 2 contains a schematic overview of the process and the proteins involved. MutS protein initiates the repair by binding to the DNA mismatch. Upon adenosine triphosphate (ATP) binding, MutS changes conformation to a sliding clamp that can slide on the DNA and recruit MutL protein. Next, MutL activates endonuclease MutH that nicks newly synthetized unmethylated DNA near the hemi-methylated site. Then MutL recruits UvrD helicase to separate the two DNA strands from each other. Exonuclease cleaves the newly synthetized strand. Finally, DNA polymerase III repairs the single strand gap. We used Biacore 1S+ system and the Poly command to study three stages of DNA mismatch repair complex formation (boxed in Figure 2): MutS binding to a mismatch in a DNA ligand, recruitment of MutL, and UvrD binding. Materials and methods Protein and DNA constructs We used MutS (2), MutL (1), and UvrD proteins that were recombinantly expressed and purified. A 100-base-pair DNA ligand was used containing a single nucleotide mismatch (guanine thymine [GT]) and biotinylations on both ends. The DNA strand is coupled on Series S Sensor Chip SA via both biotins, so that the bound MutS cannot slide off the end of the strand. Surface preparation We used 25 mM HEPES pH 7.5, 150 mM KCl, 5 mM MgCl2 , and 0.05% Surfactant P20 as running buffer during surface preparation. First, we conditioned Series S Sensor Chip SA with three consecutive one-minute (min) injections of 1 M NaCl in 50 mM NaOH, using a flow rate of 10 µL/min. We also performed a 1-min injection of 0.5% SDS, which we used as a regeneration solution in our binding assay. We diluted the DNA ligand to 1 µg/mL in the running buffer and injected it into flow cell 4 at a flow rate of 10 µL/min. When the binding level reached 800 RU, we stopped the injection manually using the Injection pause control buttons. Fig 2. Schematic overview of DNA mismatch repair in E. coli. CY39216-26Oct23-CS 3 Solution A MutS binds to mismatch Solution B MutS recruits MutL 12 000 10 000 8000 Response (RU) 6000 4000 2000 0 0 250 500 Time (s) 750 14 000 16 000 Solution C MutL recruits UvrD 1024 nM 512 nM 256 nM 128 nM 64 nM 32 nM 16 nM 8 nM 4 nM 2 nM 1 nM 0 nM Running buffer: 25 mM HEPES pH 7.5, 150 mM KCl, 5 mM MgCl2 , BSA 1 mg/mL, 1 mM ATP, 0.05% P20 Flow rate: 30 μL/min Flow path: 4–3 Contact times: 180 s Dissociation time: 360 s Complex dissociation Binding assay using Poly command We performed the analysis at 25°C. The sample compartment temperature was set to 10°C, which is low enough to preserve sample integrity but not so low that the SDS solution used for regeneration risked precipitating. We used the same running buffer in the binding assay as for DNA-strand coupling, but supplemented it with 1 mg/mL bovine serum albumin (BSA) and 1 mM ATP. To prepare our assay solutions, we diluted the proteins in running buffer to the following final concentrations: • Solution A: 200 nM MutS • Solution B: 200 nM MutS, 200 nM MutL • Solution C: 200 nM MutS, 200 nM MutL, two-fold dilutions of UvrD 0–1024 nM Solution A contained a saturating concentration of MutS. In solution B, we kept the same saturating concentration of MutS and added a saturating concentration of MutL. In solution C, we had saturating concentrations of MutS and MutL, and varying concentrations of UvrD (0, 1, 2, 4, 8, 16, 32, 64, 128, 256, 512, or 1024 nM). We used Poly command with three injections (Injection A, Injection B, and Injection C) to inject the solutions at a flow rate of 30 µL/min. We used flow cell 4 (with the attached DNA ligand) as active flow cell. Flow cell 3, which didn’t have any ligand attached, was used as the reference flow cell. The contact time for each solution was 3 min and the final dissociation time was 6 min. At the end of each cycle, we regenerated the surface with a 1-min injection of 0.5% SDS. First, we performed three startup cycles where we replaced protein solutions with running buffer. Next, we performed 12 analysis cycles, each containing a different concentration of UvrD in solution C. Fig 3. Complex formation of DNA mismatch repair proteins. Series S Sensor Chip SA was coupled with a DNA ligand containing a single nucleotide mismatch and biotinylations on both ends. Using Poly command, MutS was injected to saturation (200 nM MutS, solution A), followed by a mixture of MutS and MutL (200 nM each, solution B), and finally MutS, MutL, and varying concentrations of UvrD (200 nM MutS, 200 nM MutL, and 0 to 1024 nM UvrD, solution C). 4 CY39216-26Oct23-CS Results and discussion The sensorgram in Figure 3 shows the stepwise formation of the protein complex, followed by complex dissociation. By keeping saturating concentrations of a certain protein in the solution, there’s no net dissociation of that protein. In this way, we can study the association of MutL to the complex without interference of simultaneous MutS dissociation. Similarly, we can study the association of UvrD to the complex without interference of simultaneous dissociation of MutS and MutL. Case study 2: Binding dependencies in a SARS-CoV-2 protein complex Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) causes the respiratory disease coronavirus disease 2019 (COVID-19). To gain entry to a host cell, the receptor binding domain (RBD) of the SARS-CoV-2 spike protein binds to angiotensin-converting enzyme 2 (ACE2) on the host cell surface. CR3022 is an antibody that targets a nonoverlapping epitope on RBD (3). We used Biacore 1K+ system and Poly command to investigate the binding dependencies of RBD, CR3022 (termed α-RBD here), ACE2, and an anti-human antibody (termed α-Human here) (Fig 4). Materials and methods Surface preparation We coupled RBD (prepared in-house) on Series S Sensor Chip CM5 (895 RU) using standard amine coupling and HEPES buffered saline containing EDTA and polysorbate (HBS-EP+) as running buffer. We activated the sensor surface with a 50:50 mixture of EDC and NHS for 7 min. We diluted the RBD protein to 5 µg/mL in 10 mM sodium acetate pH 4.5 and injected it over the activated surface for 5 min, which resulted in an attachment level of 895 RU. Finally, we deactivated the surface with a 7-min injection of ethanolamine. We performed the coupling procedure at 25°C in flow cell 4. Binding assay using Poly command We diluted CR3022 (α-RBD) from LifeSpan Biosciences (product code LS-C829135) to 10 nM and ACE2 (prepared in-house) to 750 nM in HBS-EP+. The anti-human IgG (Fc) antibody from Human Antibody Capture Kit (α-Human) was diluted to 10 nM in HBS-EP+. The Poly command was used for three injections (Injection A, Injection B, and Injection C) to inject the solutions over flow cells 3 and 4 at a flow rate of 30 µL/min. We used flow cell 3, which didn’t have any ligand attached, as the reference flow cell. The contact time for each solution was 2 min and the final dissociation time was 3 min. At the end of each cycle, we regenerated the surface with a 1-min injection of 10 mM glycine-HCl, pH 1.5. We used HBSEP+ as running buffer. First, one cycle was performed where we used Poly command to inject all three potential binding partners. Then three cycles were performed where each protein was replaced in turn with running buffer. CY39216-26Oct23-CS 5 Response (RU) 0 600 500 400 300 200 100 0 100 200 300 400 500 Time (s) Response (RU) 0 100 200 300 400 500 600 0 500 Response (RU) 0 10 20 30 40 50 60 0 500 Response (RU) 0 50 100 150 0 500 Time (s) Cycle 1 Cycle 2 Cycle 3 Cycle 4 Running buffer: HBS-EP+ Flow rate: 30 μL/min Flow path: 4–3 Contact times: 120 s Dissociation time: 180 s Regeneration: Glycine 1.5, 30 s Solution A Solution B Solution C Solution A Solution B Solution C Solution A Solution B Solution C Solution A Solution B Solution C Solution A Solution B Solution C Cycle 1 α-RBD ACE2 α-Human Cycle 2 Buffer ACE2 α-Human Cycle 3 α-RBD Buffer α-Human Cycle 4 α-RBD ACE2 Buffer Cycle 1 minus cycle 2 Cycle 1 minus cycle 3 Cycle 1 minus cycle 4 RBD Sensor Chip CM5 ACE2 α-RBD α-Human Data analysis Data analysis was performed using Biacore Insight Evaluation software. To uncover the binding processes of individual components within the formed complex, we made use of the Sensorgram subtraction functionality. We created a Sensorgram item and selected the option Sensorgram subtraction – Subtract cycle to subtract a cycle with buffer from the cycle containing all proteins. We repeated the procedure for all cycles with buffer, to create three subtracted sensorgrams (cycle 1 minus cycle 2, cycle 1 minus cycle 3, and cycle 1 minus cycle 4). Results and discussion You can see the resulting sensorgrams in Figure 4 (right panel), which reveal binding dependencies within the complex. Fig 4. Binding dependencies of SARS-CoV-2 proteins. RBD was coupled on Series S Sensor Chip CM5 (895 RU). Poly command was used to inject α-RBD (10 nM), ACE2 (750 nM), and α-Human (10 nM) (cycle 1). Cycles where each protein in turn was replaced by buffer were also performed (cycles 2, 3, and 4). By subtracting cycles containing buffer segments from the cycle containing all proteins, binding dependencies within the protein complex could be elucidated. 6 CY39216-26Oct23-CS Cycle 1 minus cycle 2 shows that solution A (α-RBD) and solution C (α-Human) bind to each other. That is, the binding of the α-Human antibody depends on previous binding of the α-RBD antibody (as we would expect, since the α-RBD antibody is a human antibody). Note also that the resulting sensorgram mirrors the sensorgram of cycle 3. This shows that the interactions of α-RBD and α-Human aren’t affected by the binding of ACE2. In other words, the presence or absence of ACE2 doesn’t change the antibodies interactions (i.e., noncooperative binding). Cycle 1 minus cycle 3 shows binding of solution B (ACE2) only, as binding of solution A and solution C have been subtracted away. Note that the subtraction has created a sensorgram that mimics the sensorgram from cycle 2. This tells us that ACE2 binds to a different site on RBD compared to the α-RBD antibody (i.e., independent binding sites). This is what we expected. Crystallography studies have shown that the epitope of the α-RBD antibody doesn’t overlap with the ACE2 binding site on RBD (3). This also tells us that ACE2 binds to RBD in the same way whether or not α-RBD is present in the complex (i.e., non-cooperative binding). Cycle 1 minus cycle 4 shows binding of solution C (α-Human) alone. Note that this binding isn’t practically possible without preceding binding of solution A (α-RBD). If the α-RBD antibody wasn’t present, the anti-human antibody would have nothing to bind to. However, by using cycle subtraction we can still separate out the binding curve of the α-Human antibody as a stand-alone sensorgram. Assay considerations You can find general guidelines on surface preparation in Biacore sensor surface handbook (4). Choose your running buffer based on the requirements of the specific model system you are studying. Remember that in some cases, co-factors such as metal ions or ATP affect protein complex formation (see ‘Case study 1’ for an example). Note that there are likely multiple interaction processes occurring simultaneously in the protein complexes (e.g., dissociation of the first component and association of the second component happening at the same time). The case studies we present here handle this in two different ways. In case study 1, we kept the first two proteins at saturating concentrations, so there is no net dissociation of these proteins, while monitoring the association of the third protein. In case study 2, we ran cycles where we replace one solution in turn with running buffer, and subsequently subtracted these cycles to reveal the underlying binding processes of each protein. Conclusions Poly command supports sequential injection of three to five solutions with no intermediate wash steps, thus minimizing the dissociation time between injections. This is beneficial when studying protein complex formation, which may involve transient interactions and complicated binding dependencies. Given the biological importance of protein complexes in various cellular processes, this opens for exciting new studies. Moreover, we envision that Poly command will be a useful tool in the characterization of complexes formed by novel multivalent formats of biotherapeutics, such as bispecific antibodies or PROTACs (proteolysis targeting chimeras) and their binding partners. CY39216-26Oct23-CS 7 cytiva.com Cytiva and the Drop logo are trademarks of Life Sciences IP Holdings Corporation or an affiliate doing business as Cytiva. Biacore is a trademark of Global Life Sciences Solutions USA LLC or an affiliate doing business as Cytiva. Any use of software may be subject to one or more end user license agreements, a copy of, or notice of which, are available on request. © 2023 Cytiva For local office contact information, visit cytiva.com/contact CY39216-26Oct23-CS FAQ about Poly command • Which Biacore systems have Poly command? Biacore 1 series systems all support Poly command. • Can I run Poly command in a single flow cell? Poly command is supported for serial injections through two flow cells (all Biacore 1 series instruments), four or six flow cells (Biacore 1K+ and Biacore 1S+ systems). Poly command isn’t supported for single flow cell use, due to the design of the microfluidics. • How many sequential injections can I perform using Poly command? You can perform three to five sequential injections using Poly command. • What is the required injection volume for Poly command? The injection volume for each segment in Poly command must be in the range 25 to 450 µL, except for the final segment where the injection volume must be in the range 1 to 450 µL. • How can I evaluate data generated with the Poly command? Data is evaluated using Biacore Insight Evaluation software. Results of experiments using Poly command are visualized using Sensorgram, Plot, and Epitope Binning evaluation tools. Poly command isn’t suitable for concentration, affinity, or kinetics measurements. References 1. Fernandez-Leiro R, Bhairosing-Kok D, Kunetsky V, et al. The selection process of licensing a DNA mismatch for repair. Nat Struct Mol Biol. 2021;28(4):373-381. doi:10.1038/s41594-021-00577-7 2. Groothuizen FS, Winkler I, Cristóvão M, et al. MutS/MutL crystal structure reveals that the MutS sliding clamp loads MutL onto DNA. Elife. 2015;4:e06744. Published 2015 Jul 11. doi: 10.7554/eLife.06744 3. Yuan M, Wu NC, Zhu X, et al. A highly conserved cryptic epitope in the receptor binding domains of SARS-CoV-2 and SARS-CoV. Science. 2020;368(6491):630-633. doi:10.1126/science.abb7269 4. Biacore sensor surface handbook, CY28062-27Oct22-HB, https://cdn.cytivalifesciences.com/api/public/ content/digi-16475-pdf Ordering information Product Information Product code Series S Sensor Chip CM5 Pack of 3 BR100530 Series S Sensor Chip SA Pack of 3 BR100531 Amine Coupling Kit Reagents for ligand coupling BR100050 HBS-EP+ Buffer 10x Analysis buffer BR100669 Surfactant P20 Buffer additive BR100054 Human Antibody Capture Kit Capture reagents BR100839 Glycine 1.5 Regeneration solution BR100354 Acetate 4.5 Coupling buffer BR100350 Biacore 1K SPR system for characterization of molecular interactions 29726017 Biacore 1K+ SPR system for characterization of molecular interactions 29726018 Biacore 1S+ SPR system for characterization of molecular interactions 29726019