key: cord-0706428-zq9qra6y
authors: Das, Aurosikha; Behera, Lalita Mohan; Rana, Soumendra
title: Interaction of Human C5a with the Major Peptide Fragments of C5aR1: Direct Evidence in Support of “Two-Site” Binding Paradigm
date: 2021-08-26
journal: ACS Omega
DOI: 10.1021/acsomega.1c03400
sha: 7815a814caa3c8dabee912bb649f0c29dd7f8451
doc_id: 706428
cord_uid: zq9qra6y

[Image: see text] The C5a receptor’s (C5aR1) physiological function in various tissues depends on its high-affinity binding to the cationic proinflammatory glycoprotein C5a, produced during the activation of the complement system. However, an overstimulated complement can quickly alter the C5a–C5aR1 function from physiological to pathological, as has been noted in the case of several chronic inflammation-induced diseases like asthma, lung injury, multiorgan failure, sepsis, and now COVID-19. In the absence of the structural data, the current study provides the confirmatory biophysical validation of the hypothesized “two-site” binding interactions of C5a, involving (i) the N-terminus (NT) peptide (“Site1”) and (ii) the extracellular loop 2 (ECL2) peptide of the extracellular surface (ECS) of the C5aR1 (“Site2”), as illustrated earlier in the reported model structural complex of C5a–C5aR1. The biophysical and computational data elaborated in the study provides an improved understanding of the C5a–C5aR1 interaction at an atomistic resolution, highlighting the energetic importance of the aspartic acids on the NT-peptide of C5aR1 toward binding of C5a. The current study can potentially advance the search and optimization of new-generation alternative “antibodies” as well as “neutraligands” targeting the C5a to modulate its interaction with C5aR1.

The complement system and host defense are complexly intertwined in most of the vertebrates, as it acts as a feedback loop connecting the host's innate and adaptive immune response. On encountering a trigger, complement puts its proteolytic machinery into action, which liberates potent proinflammatory mediators like C3a, C4a, and C5a anaphylatoxins. 1 The termination of the proteolytic signal cascade manifests an inflammatory response 2 by directing the most potent anaphylatoxin C5a to recruit the C5aR1, a G-protein coupled receptor (GPCR) with high binding affinity required for initiating the desired cellular signaling. Under native-like conditions, C5a recruits C5aR1 with picomolar−nanomolar potency, which elicits inflammatory response 3 through the production of reactive oxygen species (ROS) in both myeloid and nonmyeloid tissues expressing the C5aR1, including the activation of platelets. 4 It is noteworthy that C5a−C5aR1 interactions in neutrophils are known to increase the cytosolic pH, affecting the basic functionality of the neutrophils. 5 In addition, overstimulation of the C5a−C5aR1 system triggers the surge of the proinflammatory cytokines, 6 including microvesicle shedding of neutrophils leading to the hyperinflammation-induced neutrophil dysfunction. 7 Given the inflammatory angle, it is logical to regulate the interaction of C5a with C5aR1 so that the risk of collateral damage to the neighboring tissues can be minimal. Dysregulated complement accentuates the C5a−C5aR1 interaction in several tissues leading to several chronic inflammation-induced diseases, 8 such as asthma, lung injury, kidney failure, rheumatoid arthritis, cardiovascular complications, multiorgan failure, and sepsis, including the most recent pandemic COVID-19. 9 Thus, complement and complement-regulated pathways have immense potential and pharmacological value for therapeutic intervention. 10 Therapeutic intervention of the C5a−C5aR1 system 11 usually involves the following strategies: (i) block the ECS of C5aR1, (ii) block the generation of C5a, and (iii) block or neutralize the C5a. 12−15 Large-scale mutagenesis and biomolecular signaling data evidence that the nanomolar to picomolar potency of C5a toward C5aR1 16 is due to the recruitment of two distal sites with large surface areas, respectively, engaging (i) the NT-peptide (Site1) and (ii) ECS of C5aR1 (Site2) with the core and the C-terminus (CT) peptide of C5a through a specific protein−protein interaction. In addition, extensive mutational studies 17−21 on both C5a and C5aR1 had provided the following two key observations: (i) truncation of the NT-peptide region significantly affects the affinity of C5aR1 toward C5a, and the NT-peptide of C5aR1 is very important for activation by C5a, 22, 23 (ii) the NTtruncated C5aR1 could be effectively activated by short peptide analogues based on the CT-region of C5a albeit with relatively weak binding affinity. 20, 24 Therefore, it was hypothesized that the NT-peptide contains the first binding site (Site1), and the poorly defined interhelical crevice in the transmembrane region contains the second binding site (Site2) on C5aR1 for recognizing the C5a to trigger the downstream biomolecular signaling. The high-affinity interaction at the "Site1" plays the anchorage function to arrest the C5a, and the relatively low-affinity interaction at the "Site2" helps in docking the CT-peptide of the C5a, triggering the activation and the cellular signaling of C5aR1. 25 The "two-site" binding paradigm, a hallmark feature consistently illustrated for many peptide/protein binding GPCRs of the secretin family, 26 has also been convincingly illustrated over the years in several chemokines complexed to their cognate receptors belonging to the rhodopsin family. 27, 28 The recently available structural data for the CCL20-CCR6 system 28 clearly illustrates the strong involvement of the ECS (consisting of the three ECLs) of CCR6 as a part of the "Site2" in binding to CCL20. Nevertheless, the full-length structures of the NT-peptides are not entirely resolved in several reported biomolecular complexes of the chemokine receptors, though the importance of the binding interaction involving the synthetic NT-peptides of CXCR4 with the CXCL12 chemokine has been structurally illustrated. 29 While both computational and experimental model structures of C5aR1 both in free and in complex with smallmolecule ligands are available in the literature, 30−33 the structural biology approach has not successfully illustrated the hypothesized biomolecular recognition of the C5a−C5aR1 system at an atomistic resolution so far. Interestingly, most of the structural studies reported so far for C5aR1 demonstrate a truncated structure of the NT-peptide. More importantly, the plausible conformational changes that are likely to be triggered in C5a by binding to the NT-peptide of C5aR1, a vital transition state, perhaps required for the sequential docking of the C5a on the ECS of C5aR1 is poorly understood. Thus, an effort toward the atomistic understanding of the high-affinity intermolecular interactions at "Site1" is essential, as it will pave the way for designing the new age alternative antibody-like molecules for effectively neutralizing the pathophysiological concentration of C5a under disease settings without downregulating the physiological function of the complement system or completely shutting down the C5a-induced lowgrade cellular response of C5aR1.

In this context, a highly refined "two-site" model structural complex describing a plausible activation mechanism of the C5a−C5aR1 system ( Figure 1 ) has been made available from our group earlier, 34 which requires further evaluation to understand the importance of the intermolecular interactions postulated in the model structural complex of C5a−C5aR1. In the current study, major synthetic peptide fragments of C5aR1, such as the NT-peptide and its mutants (codenamed as SR3, SR4, and SR5), including the ECL2 peptide (codenamed as SR1), are subjected to a battery of biophysical studies against the recombinant human C5a to understand their role in anchoring the C5a to C5aR1. The data obtained from the circular dichroism (CD) and fluorescence titration studies find support from the molecular dynamics (MD) and free energy calculation studies, which not only validates the highly refined "two-site" binding model structural complex of C5a−C5aR1 34 but also indicates that the electrostatic interactions between the amino acids with anionic side chains on the NT-peptide of C5aR1 and the amino acids with cationic side chains on the surface of C5a play a crucial role in C5a−C5aR1 interaction over and above the canonical intermolecular interactions involved in arresting the bulk of C5a.

2.1. Design Rationale behind the Synthetic Peptides. The "two-site" model structural complex of C5a−C5aR1 hypothesized in our earlier study 34 suggests that the free NTpeptide of the C5aR1 (Site1), which harbors several amino acids with anionic side-chain structure, has a strong preference to get wrapped around the cationic surface of C5a ( Figure 2 ). Point mutation of the highlighted cationic amino acids on C5a has been shown to affect both the binding and signaling activity of the C5a. 21 Previous mutagenesis studies on NTpeptide of C5aR1 have evidenced the involvement of Asp16, Asp18, Asp21, and Asp27 in the binding and signaling of C5a through C5aR1. 22 Further, in agreement with the mutagenesis studies, the model complex also hints at the involvement of aromatic amino acids like Tyr11 and Tyr14, 35 including the Tyr6 in binding to C5a. In addition to this, the model complex has hypothesized that binding of NT-peptide to the bulk of C5a can trigger conformational changes in C5a, which will lead to the docking of its conformationally altered CT-peptide to the ECS of C5aR1 (Site2), which is composed of ECL2, 36 one of the largest extracellular loop of C5aR1, in addition to the others. Thus, to evaluate the model complex of C5a−C5aR1 further, three NT-peptides (Figures S1−S3) and one ECL2 peptide ( Figure S4 ) of C5aR1 ( Figure 3 ) were synthetically prepared.

Out of the three NT-peptides, SR3 represents the native NT-sequence of C5aR1 (Met1-Lys28), whereas SR4 represents the Asp/Ala (Asp2/Ala, Asp10/Ala, Asp15/Ala, Asp16/ Ala, Asp18/Ala, and Asp27/Ala) and SR5 represents the Tyr/ Ala (Tyr6/Ala, Tyr11/Ala, and Tyr14/Ala) mutant NTpeptide sequences of C5aR1. The ECL2 is the major peptide fragment of the ECS of C5aR1, and its role in the binding and signaling of C5a is well known. The SR1 peptide represents the native ECL2 sequence of C5aR1 (Tyr174-Arg198), except that it is acylated and amidated, respectively, at the N-and Ctermini. Also, it carries the Cys188/Ser mutation to avoid undesired aggregation issues in the solution. The other mutants of the ECL2 peptide were not prepared to avoid disruption of the folded β-hairpin structure of the free ECL2 peptide in solution. In addition, given the short sequence length, the ECL1 and ECL3 peptides were not synthesized, as it is evidenced that both ECL1 and ECL3 may play a more sensitive role in the activation of C5aR1 than the binding of C5a. More importantly, mutations in the ECL1 of C5aR1 have also been shown to have no effect on the binding affinity of C5a. 37 2.2. Conformational Analysis of the Synthetic Peptides. The native and the mutant NT-peptides were completely soluble in 1× PBS (pH ∼ 7.4) and thus subjected to conformational analysis studies by recruiting the CD spectroscopy. As presented in Figure 4 , the NT-peptides demonstrated progressive conformational ordering between 0.05 and 1 μM, demonstrating a signature signal broadly similar to the extended sheet structure of the polypeptides.

Interestingly, the formation of the ordered β-sheet structure was noted earlier over certain regions of the NT-peptide ( Figure 5 ) of C5aR1 complexed to C5a throughout 0.25 μs MD simulation in POPC bilayers. 34 Indeed, ordered β-sheet structure has also been noted over certain sections of the 38mer NT-peptide of CXCR4 complexed to CXCL12 in solution. 29 However, the comparison of the CD signal observed for the SR3 peptide (1 μM) with the mutant peptides (SR4 and SR5) indicates that alanine mutations can potentially affect the overall conformational ordering of the NT-peptide. Further, the characteristic CD signal of the peptides diminished to a greater extent beyond 1 μM, which can be attributed to the formation of soluble aggregates of the peptides under experimental conditions. 38 Thus, most of the further binding studies involving the NT-peptides were performed below 1 μM. On the other hand, up to 100 μM, the ECL2 peptide of C5aR1 (SR1) has been reported to demonstrate a CD signature commonly attributed to twisted short-stranded β-hairpin-like conformation in solution. It is noteworthy that the ECL2 peptide was also predicted to harbor a β-hairpin fold in our earlier studies, 31 which was subsequently confirmed from the observation made in the crystal structure of the thermostabilized C5aR (PDB ID: 5O9H) known as StaR. 33 Interestingly, StaR carried 11 strategic point mutations, truncated by 29 amino acids in the NT-peptide region and 17 amino acids in the CT-peptide region, similar to our truncated model structure of C5aR1 31 that lacked 26 amino acids on the NT-peptide region and 34 amino acids in the CT-peptide region. 2.3. Probing the Intermolecular Interaction between the NT-Peptides and C5a. The intermolecular interactions hypothesized between the NT-peptide of C5aR1 and C5a in the model complex were subjected to scrutiny, respectively, using CD and fluorescence spectroscopy. The concentrations of the native (SR3) and the mutant NT-peptides (SR4 and SR5) were varied between 0 and 1 μM for titration against 0.1 μM recombinant human C5a, and the corresponding conformational changes, as well as the change in fluorescence intensity observed for C5a was monitored for gauging the interaction of the NT-peptides with C5a ( Figure 6 ).

As noted in Figure 6a , the signal intensity of the signature CD spectra demonstrated by the C5a enhanced significantly with the increase in the concentration of the NT-peptides, indicating the strong association of the peptides with the C5a. Interestingly, the rise in the CD signal intensity was also accompanied by the change in the signature CD spectra of C5a, suggesting that binding of the NT-peptides triggers an intrinsically disordered conformational state in C5a. As presented in Figure 6a , the SR4 peptide appears to induce a robust conformational alteration in C5a compared to SR3 and SR5 peptides that demonstrate almost similar interaction patterns with the C5a. It is noteworthy that in comparison to the native NT-peptide SR3, SR5 harbors Tyr6/Ala, Tyr11/Ala, and Tyr14/Ala mutations in its sequence, whereas SR4 carries Asp2/Ala, Asp10/Ala, Asp15/Ala, Asp16/Ala, Asp18/Ala, and Asp27/Ala mutations in its sequence. Further, to maintain some degree of native affinity toward C5a, the Asp21 was not mutated to Ala in the SR4 peptide. In the absence of the anchorage naturally imparted by the transmembrane domain of C5aR1, these synthetic NT-peptides are relatively more labile, which empowers them to establish biologically nonspecific interaction with C5a with the slightest change in their conformational ordering. The SR4 peptide harbors a significantly mutated sequence, which also appears to affect its conformational ordering ( Figure 4) . Thus, it is likely that the SR4 peptide will have a strong potential to drive a biologically nonspecific interaction with C5a compared to the SR5 peptide.

The observation made in the CD titration studies was further subjected to scrutiny by probing the intrinsic tyrosine fluorescence of recombinant C5a both in the presence and absence of increasing concentration of NT-peptides ( Figure  6b ). Among the three NT-peptides, SR5 does not fluoresce at all due to the lack of aromatic amino acids, whereas SR3 and SR4 have very negligible intrinsic tyrosine fluorescence ( Figure  S5 ) at the working concentrations, which does not overlap with the emission maximum of C5a. It is evident from Figure  6b that the intrinsic fluorescence of C5a substantially increases with an increase in the concentration of NT-peptides and eventually gets saturated in the presence of 500 nM peptides, indicating the strong intermolecular interaction between the C5a and the NT-peptides, as observed in the CD studies. Fitting the normalized CD and fluorescence titration data suggests that while SR3 binds to C5a with an estimated K d ∼ 126−193 nM, SR5 binds to C5a with an estimated K d ∼ 105− 123 nM. Interestingly, the binding affinity of the SR4 peptide toward C5a could not be estimated from either CD or fluorescence titration data, as the normalized response was too scattered in response to an increase in the concentration of the peptide.

The observed binding data appears to be in sync with the reported biomolecular signaling studies in cultured cells, where it has been shown that Asp15/Ala, Asp16/Ala, Asp18/Ala, and Asp21/Ala mutations in the NT-region of C5aR1 collectively reduces the binding affinity by ∼42-fold, whereas further addition of Asp10/Ala mutation to NT-region reduces its binding affinity by ∼140-fold toward C5a compared to the native C5aR1. 22 It is also reported that Asp15/Ala and Asp18/ Ala mutations in NT-peptide trigger a tremendous loss in C5aR1 signaling. 39 It is worth mentioning that SR4 carries six Asp/Ala mutations in its sequence compared to SR5, which harbors three Tyr/Ala mutations compared to the native SR3 peptide. The K d values of SR3 and SR5 estimated from the CD data are almost identical. However, SR5 peptide with three Tyr/Ala mutations demonstrated relatively tighter binding to C5a compared to SR3 peptide in the fluorescence titration studies, though the statistical significance of the same remains to be pursued. Thus, broadly, it can be concluded that both SR3 and SR5 peptides appear to have a comparable binding affinity toward C5a. On the other hand, the SR4 peptide that contained all of the tyrosines but lacked the aspartic acids except the Asp21 did not demonstrate a quantifiable binding affinity based on the CD and fluorescence titration studies. This is in contrast to the earlier observations that suggest that tyrosine sulfation on the NT-region of C5aR1 is an important post-translational modification required for the efficient binding of C5a. 35 Overall, the current data indicate that collectively the aspartic acids on the synthetic NT-peptides may be crucial than the tyrosine residues for specific binding of the C5a to C5aR1. However, in the context of the model C5a− C5aR1 complex reported in our earlier studies, the specific contribution of tyrosines toward overall binding affinity cannot be completely ruled out.

2.5. Comparison of the Biomolecular Complexes Formed between the NT-Peptides and C5a. The CD and fluorescence titration studies suggested that mutation of aspartic acids on the NT-peptide region affects the binding affinity toward C5a. However, concerning the mode of interactions of the NT-peptides with the C5a, the data is virtually blind, as the inference is derived from the limited number of variants of NT-peptide, which cannot delineate the specific contribution of each amino acid toward the estimated binding affinity. More importantly, the synthetic NT-peptides are free at both ends, compared to the native conditions, where the C-terminal end will be connected to the transmembrane helix number 1 (TM1) of C5aR1. Thus, the biomolecular complexes involving the C5a and SR4/SR5 peptides were modeled based on the C5a−SR3 complex extracted from the reported C5a−C5aR1 complex. Subsequently, the C5a−SR3, C5a−SR4, and C5a−SR5 complexes were subjected to comparative MD studies over 100 ns each at 300 K.

The data presented in Figure 8 suggests that free NTpeptides of C5aR1 could remain bound to the C5a over the duration of the MD trajectory, irrespective of the number or type of mutations in the peptides, supporting the physical viability of the strong intermolecular interactions between the C5a and the NT-peptides, as noted in the CD and fluorescence titration studies. Interestingly, both the native and mutant NTpeptides were able to experience ∼17 intermolecular hydrogen bonds consistently over the duration of the MD, in addition to the other type of interactions, which could be the reason Figure 7 . Estimation of the binding affinity from the normalized CD and fluorescence titration data points of C5a observed against the variable concentrations of the NT-peptides. Binding affinity could not be obtained reliably for the SR4 peptide due to the incomplete data fitting. 

http://pubs.acs.org/journal/acsodf Article behind the observed stability of the biomolecular complexes involving the mutant NT-peptides that demonstrated altered interactions with the C5a compared to the native SR3 peptide (Figure 9 ). Though the NT-peptides were found to be orientationally drifted to a certain extent compared to the C5a−C5aR1 complex, none of the peptides were dislodged entirely from the surface of C5a over the duration of MD ( Figure 10 ). It is noteworthy that in the case of both SR3 and 

http://pubs.acs.org/journal/acsodf Article SR4 peptides, some tyrosine amino acids on the NT-peptide demonstrated strong interaction with several amino acids on C5a over the duration of the MD. 2.6. Comparison of the Binding Free Energy of the NT-peptide−C5a Complexes. To further understand the residue-specific contributions toward the overall binding affinity observed for the NT-peptides in the experimental studies, the respective MD trajectories of C5a−SR3, C5a−SR4, and C5a−SR5 complexes were subjected to molecular mechanics Poisson−Boltzmann surface area (MM-PBSA) calculation to estimate the binding free energy of the biomolecular complexes. The estimation of the binding free energies involved 500 conformers from each trajectory, randomly selected from the most populated cluster ( Figure  S6 ), evolved over the duration of 100 ns MD. In strong agreement with the experimental data, the binding free energy ( Figure S7 ) estimated for the C5a−SR3 complex (−553.48 ± 8.50 kcal/mol) was found to be similar to that for the C5a− SR5 complex (−568.11 ± 10.57 kcal/mol), which was significantly higher compared to the free energy of binding estimated for the C5a−SR4 complex (−31.99 ± 10.87 kcal/ mol), indicating a tighter binding of the SR3/SR5 peptides to the C5a than the SR4 peptide.

Further, the decomposition of the MM Energy in the context of the amino acids of the NT-peptides clearly evidences ( Figure 11 ) that the aspartates contribute significantly over other amino acids on the NT-region of C5aR1 toward the overall binding free energy. The data indicates that while Asp27 makes the highest contribution, Asp2 makes the lowest contribution. During the MD, it was observed that the Asp2 makes contact with the Arg74 of the C5a in few structures, which could be due to the inherent conformational flexibility of the CT-region of C5a. In addition, Asp10, Asp15/ Asp16/Asp18, and Asp21 also make solid contributions toward the binding free energy.

As presented in Figure 11 , the significant energetic contribution made by the aspartates in the case of SR3 and SR5 peptides is expectedly absent in the case of the SR4 peptide, which harbors Asp/Ala mutations. This could be the reason behind the weakest binding affinity demonstrated by the SR4 peptide toward C5a both in MM-PBSA calculation as well as in the CD and fluorescence titration studies. Interestingly, Asp21, which was not mutated to Ala in SR4 peptide, demonstrated substantial energetic contribution toward binding the C5a. On the other hand, despite harboring the Tyr/Ala mutations, SR5 peptide illustrated comparable binding affinity toward C5a like the SR3 peptide in both MM-PBSA calculations and CD titration studies. Further analysis indicates that though Tyr6/Tyr11/Tyr14 appreciably contributed toward the binding free energy, it was significantly lower than the aspartic acids on the NT-peptides. Nevertheless, the observation is in sync with the earlier studies, which reported that post-translational sulfation of Tyr11/ Tyr14 is essential for the efficient binding of C5a. 35 Overall, the data indicate that selective mutation of any single aspartic acid on the NT of C5aR1 may be able to influence C5a− C5aR1 binding and signaling to an appreciable extent.

2.7. Probing the "Two-Site" Binding Interaction of C5a−C5aR1. Titrations of the NT-peptides against C5a broadly confirmed the hypothesized interactions at "Site1" as illustrated in Figure 1 . The model structural complex of C5a− C5aR1 also suggests that in addition to binding to the NTpeptide, C5a also binds to several amino acids on the ECS of C5aR1 collectively defined as "Site2". The ECL2 is the most prominent polypeptide in the ECS of C5aR1, which has been described to play a significant role in the C5a-induced activation of C5aR1. 40 However, a direct interaction of the C5a with the ECL2 peptide of C5aR1 is not clearly described in the literature. Thus, it was necessary to probe whether activation of C5aR1 is also linked to the interaction of C5a with the ECL2 peptide at "Site2" of C5aR1. To probe the interaction at "Site2", C5a preincubated with saturating concentration of the NT-peptides was subjected to varying concentrations of ECL2 peptide, and the observed response was recorded by, respectively, CD and fluorescence spectroscopy.

The data presented in Figure 12 indicates that binding of NT-peptides does not occlude the interactions of C5a further with the ECL2 peptide, suggesting that C5a perhaps engaged the free NT-peptides at "Site1" in a near-native manner. The addition of 0.5 μM ECL2 peptide (SR1) to C5a preincubated with the native SR3 peptide substantially enhanced the CD signal, which did not change further by the addition of 1 μM SR1 peptide. This suggests that the ECL2 peptide perhaps acted as a part of the distal "Site2", which was able to saturate the secondary binding site of C5a. In agreement, fluorescence data involving the SR1 peptide for the C5a−SR3 system was also found to be consistent with the observation made in the CD. A similar trend was also noticed in both CD and fluorescence for the C5a−SR5 system, in the presence of 0.5− 1 μM SR1 peptide. Further, compared to the free C5a, the C5a preincubated with the native NT-peptide (SR3) of C5aR1 demonstrated a much better response toward the ECL2 peptide, suggesting that binding of the NT-peptide of C5aR1 to C5a may be the first important step necessary for triggering the activation pathway of C5aR1.

The modest change in fluorescence intensity observed for both the systems suggested that the SR1 peptide perhaps does not come in direct contact with the bulk of C5a, further affirming the existence of "Site2". However, compared to the C5a−SR3/C5a−SR5 system, the addition of SR1 peptide demonstrated a relatively strong conformational response in CD for the C5a−SR4 system. On the other hand, the fluorescence signal of the C5a−SR4 system was quenched appreciably on the addition of 0.5−1 μM SR1 peptide. The comparative observations clearly evidence that mutations of important amino acids can trigger an improper mode of interaction between the NT-peptide and the C5a at "Site1", Figure 11 . Comparative summary of the energetic contribution made by each amino acid of the NT-peptides of C5aR1 toward the average free energy of binding, respectively, estimated for the C5a−SR3, C5a−SR4, and C5a−SR5 biomolecular complexes. The specific amino acids that have been subjected to alanine mutation either in SR4 or SR5 peptides are also highlighted within the graph.

which may subsequently alter the further interaction of C5a with the other peptide fragments on the ECS of C5aR1. Further, as evidenced, the NT-peptides do not appear to interact with the ECL2 peptide of C5aR1 in a substantial manner ( Figure S8 ). Thus, it is reasonably clear that C5a interacts with C5aR1 by recruiting "two sites", as illustrated in the model complex of C5a−C5aR1, presented in Figure 1 .

C5aR1 is among the ∼120 GPCRs known in the human genome that recognizes endogenous peptides 41 or proteins as ligands for cellular signaling and physiology. The "two-site" binding paradigm involving the C5a−C5aR1 system is an old concept that was hypothesized three decades ago. 25 The concept had been strongly supported by the deletion and single-point mutation-based biomolecular signaling data obtained from the C5a−C5aR1 system. However, in the absence of any such structural data related to the C5a−C5aR1 system, a highly refined full-length model structure of C5aR1 complexed to C5a 34 was generated in the recent past, affirming the existence of "two-site" contact-based molecular recognition in the C5a−C5aR1 system. The model illustrated that the high-affinity binding at "Site1" is driven by several salt bridge/ hydrogen-bond interactions involving the aspartic acids, as well as the tyrosine amino acids of the NT-peptide of C5aR1. Similarly, the interactions at "Site2" involved several amino acids of the ECL2 and ECL3 peptides of C5aR1. However, the physical viability of the illustrated mode of interaction was untested, which was subject to test in the current study by recruiting the synthetic variants of the NT-peptides and the ECL2 peptide of C5aR1. The titration data of the NT-peptides as well as of the ECL2 peptide presented in Figures 6, 7, 11 , and 12 indicates strong agreement with the intermolecular interaction observed in the model C5a−C5aR1 complex. Moreover, the data also suggests that the aspartic acids on the NT-region of C5aR1 are crucial for the high-affinity binding of C5a, which is in substantial agreement with the independently reported biomolecular signaling studies that indicate both Asp/ Ala or Asp/Asn mutation abrogates binding and signaling of C5aR1. Similarly, mutation of several amino acids with cationic side chains on the bulk of C5a, like Arg37, Arg62, and Arg40, have also been shown to abrogate the binding of C5a to C5aR1, in addition to the mutation of His67, Lys68, and Arg74 in the CT-region of C5a. 21 Further, the binding and signaling activity of C5a toward C5aR1 can also be dampened by altering the biologically active conformer allosterically 42 induced by the mutation of specific amino acids on C5a, which are not necessarily involved in interaction with C5aR1 at "Site1" or "Site2". Thus, the effect of C5a mutation on the binding and signaling of C5aR1 should also be evaluated from the dynamic structural perspective of C5a. Moreover, the binding of C5a to C5aR1 is not purely complementarity of cationic−anionic electrostatic interactions of side chains, as other amino acids on the NT-region of C5aR1 also contribute toward sustained hydrogen bonding and hydrophobic interactions with the C5a. Nevertheless, the current study validates the reported C5a−C5aR1 model complex and provides direct evidence of the involvement of two-step interactions of C5a at two discrete binding sites located on the C5aR1.

A better understanding of the "two-site" binding in the C5a−C5aR1 system can be advantageous from the therapeutic intervention standpoint. Traditionally, C5aR1 has been the preferred target for competitive inhibition of C5a, and currently, several small molecules and peptides are known in the literature that can competitively inhibit the binding of C5a 

http://pubs.acs.org/journal/acsodf Article to C5aR1 by targeting the orthosteric/allosteric "Site2" of C5aR1. 30 On the other hand, direct targeting of C5a for competitively inhibiting the "Site1" interactions of C5aR1 and the "two-site" binding of C5a to C5aR1 is comparatively less exploited. Nevertheless, designer complementary peptides targeting the antisense homology box (AHB) of C5a have shown some exciting results in the earlier cell culture and animal model studies. 43, 44 In addition, recent studies 12 from our group also indicate that prednisone (PDN), a known corticosteroid, can bind to C5a with K d ∼ 0.38 μM, which can potentially modulate the interaction of the C5aR1 with C5a. Preliminary studies, presented in Figure 13 , suggest that C5a preincubated with a near-saturating concentration of PDN (∼0.5 μM) demonstrates a comparatively weaker response toward the near-saturating concentration (∼0.35 μM) of the native NT-peptide (SR3: Met1-Lys28) of C5aR1 than the free C5a, which could be most likely due to the competitive inhibition of the NT-peptide by the binding of PDN to C5a. In this context, it is reasonable to believe that designer peptides targeting C5a will be able to competitively inhibit the binding of the NT-peptide (Site1) and the ECL2 peptide (Site2) of C5aR1 under native conditions. However, subsequent future studies will be necessary for solid validation of the above hypothesis.

The recruitment of C5aR1 by C5a, one of the most proinflammatory anaphylatoxins of the complement system, triggers a plethora of leukocyte responses, and indeed advanced studies over the years suggest that a derailed complement can eventually set the stage for several fatal immunological and inflammatory diseases. The newest addition to the list of fatal diseases is the COVID-19 pandemic, where a direct correlation has been noted between the plasma level of C5a and the severity of COVID-19. 9 Given the complexity of the immune response, an elevated level of C5a can also influence the secretion of other proinflammatory cytokines, leading to a cytokine surge. The favorable outcome observed with anti-C5 (Eculizumab) 45 and anti-C5a (Vilobelimab) 46 antibodies in the case of COVID-19 treatment further highlights the crucial role of C5a in the severe inflammation 47induced pathology of lung injury, followed by the significant damage to the other organ systems. The pleiotropic nature of C5a coupled with its involvement in multiple intertwined signaling pathways labels C5a as a potential target for exploring prospective therapeutics. The anti-C5a antibodies 46−48 neutralize the major functional epitopes on the bulk of C5a, which serve as the binding site for interacting with the NT-peptide of C5aR1. In this context, the current study provides an overtly simplified biophysical overview of the synergistic intermolecular interactions of C5a at the "two sites" of C5aR1, as hypothesized earlier in the model complex of C5a−C5aR1, which can be further exploited to engineer high-affinity designer bidentate peptides for targeted inhibition of C5a, as a secondary alternative to the heavyweight antibodies.

5. MATERIAL AND METHODS 5.1. Synthesis of the Major Peptide Fragments of C5aR1. Total four peptides belonging to the extracellular surface of C5aR1, involving the two major peptide fragments (i) the N-terminus and (ii) the extracellular loop-2 (ECL2), were synthetically prepared using the standard Fmoc chemistry over solid phase by recruiting the services of GenScript (NJ, USA). Out of four, three NT-peptides (SR3, SR4, and SR5) contained 28 amino acids and one ECL2 peptide (SR1) contained 25 amino acids. The SR1 peptide (Tyr174-Arg198) was acylated and amidated, respectively, at the N-and Ctermini and also carried a C188/S mutation for the ease of synthesis and to avoid unwanted aggregation in solution. SR3 peptide has the native NT-sequence of C5aR1 (Met1-Lys28), whereas SR4 and SR5 peptides have mutations at strategic positions based on the model complex of C5a−C5aR1 reported earlier. All of the peptides have ≥ 95% purity as judged from the analytical HPLC profile recorded by recruiting the C18 (4.6 × 250 mm) column at 220 nm, using acetonitrile−water gradient in the presence of 0.05−0.065% trifluoroacetic acid (TFA). The ESI-MS confirmed the integrity of all of the peptides.

5.2. Circular Dichroism (CD) Studies. The CD studies were carried on a Chirascan CD spectrometer system in the far-UV region at 25°C using thoroughly filtered and degassed 1× PBS (pH ∼ 7.4). Each sample was subjected to a minimum of three scans with a time constant of 1s and a step size of 1 nm. The peptides and the recombinant human C5a (R&D systems) were appropriately solubilized only in 1× PBS (pH ∼ 7.4). In all of the titration studies, the concentration of C5a was maintained at 0.1 μM, and the concentration of the NTpeptides (SR3, SR4, and SR5) was varied between 0 and 1 μM. All of the samples were incubated for a minimum 1 h at 4°C prior to the CD studies. The molar ellipticity was converted to mean residue ellipticity [θ MRE ], and the data were normalized and subjected to nonlinear regression in GraphPad Prism for estimating the binding affinity of the NT-peptides toward C5a. In addition, for probing the conformational changes due to the "two-site" binding effect, 0−1 μM ECL2 peptide (SR1) was titrated against 0.1 μM C5a preincubated with 0.5 μM NTpeptides. Each sample of C5a corresponding to the given concentration of the peptides was individually prepared and read separately in the instrument. A similar procedure was followed for the C5a and prednisone system. The CD signals in millidegrees observed for the C5a incubated in the presence of 0−1 μM peptides were subtracted from the CD signals in millidegrees arising from the corresponding concentration of the free peptides in the buffer prior to the final processing of the data.

5.3. Fluorescence Studies. The fluorescence studies were performed in pure 1× PBS (pH ∼ 7.4) at 25°C, using a Cary Eclipse fluorescence spectrophotometer (Agilent Technolo- gies) equipped with the PCB 1500 Water Peltier System. The excitation and emission slit widths were set to 5 nm with excitation wavelength range of 278−280 nm and emission range between 290 and 450 nm. To maintain uniformity with the CD studies, the fluorescence titration studies were also performed in the presence of 0.1 μM C5a by varying the concentration of the NT-peptides between 0 and 1 μM. All of the spectra were recorded with an average of three scans, and the background spectra of the peptides in the corresponding buffer were appropriately subtracted. The fluorescence intensity of the C5a in the presence of the varying concentration of the NT-peptides was normalized at 350 nm, and the data were subjected to nonlinear regression in GraphPad Prism for estimating the binding affinity of the NT-peptides toward C5a. Similar to the CD studies, 0−1 μM ECL2 peptide (SR1) was also titrated against 0.1 μM C5a preincubated with 0.5 μM NT-peptides for probing the conformational changes in C5a due to the "two-site" binding effect. The fluorescence signals observed for the C5a incubated in the presence of 0−1 μM peptides were subtracted from the fluorescence signals arising from the corresponding concentration of the free peptides in the buffer prior to the final processing of the data.

5.4. Molecular Dynamics (MD) Studies. The C5a (PDB ID: 1KJS) complexed to the native (SR3) and mutant (SR4 and SR5) NT-peptides was subjected to MD simulations for 100 ns each at 300 K in the presence of simple point charge (SPC) water molecules, with solvent density set to the value corresponding to 1 atm at 300 K, by recruiting the GROMACS package 49 as described earlier. 34, 42, 50 The C5a−SR3 model complex reported earlier served as the reference for, respectively, generating the C5a−SR4 and C5a−SR5 mutant complexes. All of the systems were neutralized by randomly placing the appropriate number of chloride ions and were equilibrated twice, first under NVT (0.5 ns), followed by NPT (0.5 ns) conditions before the production MD run. Conformational clustering was performed at an interval of 50 ps with a backbone RMSD cutoff ≤ 1.5 Å by recruiting the gromos fitting method, as defined in GROMACS. PyMOL (The PyMOL Molecular Graphics System, Version 1.1r1, Schrodinger, LLC) and Discovery studio (Accelrys) software were utilized for initial processing, visualization, analysis, and presentation of the protein structures. The utility programs available in GROMACS were implemented for the detailed analysis of all of the MD trajectories.

5.5. Binding Energy Calculation Studies. The 100 ns MD trajectories of the C5a complexed to the wild type (SR3) and the mutant (SR4 and SR5) NT-peptides were, respectively, used for calculating the relative binding free energies of the biomolecular complexes by recruiting the molecular mechanics Poisson−Boltzmann surface area (MM-PBSA) method, as described elsewhere. 51 The dielectric constant of the solute and solvent were, respectively, fixed at 20 and 80 for the calculation of polar solvation energy. Variation of solute dielectric in the 2−20 range altered the total binding energy to some extent, but the overall trend remained the same for the peptides. Finally, 500 conformers derived from the first major cluster populated over 100 ns of the MD trajectory for each of the complexes were, respectively, used for calculating the average binding free energy.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.1c03400.

HPLC and ESI-MS profiles of the synthetic peptides; fluorescence signal of the peptides compared to C5a; cluster analysis of the MD trajectories; molecular mechanics energy plot of the C5a complexed to different NT-peptides; and absence of intermolecular crossreactivity between the peptide fragments as judged from CD and fluorescence (PDF) 

Complement and its role in innate and adaptive immune responses

The role of the anaphylatoxins in health and disease

A Novel C5a Receptor-Tissue Factor Cross-Talk in Neutrophils Links Innate Immunity to Coagulation Pathways

Complement C5a Functions as a Master Switch for the pH Balance in Neutrophils Exerting Fundamental Immunometabolic Effects

Complement C5 Inhibition Blocks the Cytokine Storm

Consumptive Coagulopathy By Decreasing Lipopolysaccharide (LPS) Release in E. coli Sepsis. Blood

Complement C5a Induces Pro-inflammatory Microvesicle Shedding in Severely Injured Patients

Therapeutic Targeting of the Complement System: From Rare Diseases to Pandemics

The roles and potential therapeutic implications of C5a in the pathogenesis of COVID-19-associated coagulopathy

New concepts on the therapeutic control of complement anaphylatoxin receptors

The role of human C5a as a non-genomic target in corticosteroid therapy for management of severe COVID19

Binding of raloxifene to human complement fragment 5a ((h)C5a): a perspective on cytokine storm and COVID19

Resveratrol binding to human complement fragment 5a ((h)C5a) may modulate the C5aR signaling axes

A rational search for discovering potential neutraligands of human complement fragment 5a ((h)C5a)

Function, structure and therapeutic potential of complement C5a receptors

Site-specific mutagenesis of residues in the human C5a anaphylatoxin which are involved in possible interaction with the C5a receptor

Residues 21-30 within the extracellular N-terminal region of the C5a receptor represent a binding domain for the C5a anaphylatoxin

Random mutagenesis of the complement factor 5a (C5a) receptor N terminus provides a structural constraint for C5a docking

Structure-function studies in a series of carboxyl-terminal octapeptide analogues of anaphylatoxin C5a

The pharmacophore of the human C5a anaphylatoxin

The amino terminus of the human C5a receptor is required for high affinity C5a binding and for receptor activation by C5a but not C5a analogs

The NH2-terminal region of C5aR but not that of FPR is critical for both protein transport and ligand binding

Determination of structural elements related to the biological activities of a potent decapeptide agonist of human C5a anaphylatoxin

Two-site binding of C5a by its receptor: an alternative binding paradigm for G protein-coupled receptors

Current understanding of the structure and function of family B GPCRs to design novel drugs

Structure of CC Chemokine Receptor 5 with a Potent Chemokine Antagonist Reveals Mechanisms of Chemokine Recognition and Molecular Mimicry by HIV

Structural basis for chemokine receptor CCR6 activation by the endogenous protein ligand CCL20

Structural basis for chemokine recognition by a G protein-coupled receptor and implications for receptor activation

Orthosteric and allosteric action of the C5a receptor antagonists

Model structures of inactive and peptide agonist bound C5aR: Insights into agonist binding, selectivity and activation

Structural complexes of the agonist, inverse agonist and antagonist bound C5a receptor: insights into pharmacology and signaling

Structure of the complement C5a receptor bound to the extrahelical antagonist NDT9513727

The Model Structures of the Complement Component 5a Receptor (C5aR) Bound to the Native and Engineered (h)C5a

Sulfated tyrosines contribute to the formation of the C5a docking site of the human C5a anaphylatoxin receptor

Lifting the lid on GPCRs: the role of extracellular loops

Genetic analysis of the first and third extracellular loops of the C5a receptor reveals an essential WXFG motif in the first loop

Using circular dichroism collected as a function of temperature to determine the thermodynamics of protein unfolding and binding interactions

The role of the N-terminal domain of the complement fragment receptor C5L2 in ligand binding

Essential role for the second extracellular loop in C5a receptor activation

Structure and Function of Peptide-Binding G Protein-Coupled Receptors

Allosterism in human complement component 5a ((h)C5a): a damper of C5a receptor (C5aR) signaling

Inactivation of C5a anaphylatoxin by a peptide that is complementary to a region of C5a

Novel complementary peptides to target molecules

Eculizumab as an emergency treatment for adult patients with severe COVID-19 in the intensive care unit: A proof-of-concept study

Treatment with anti-C5a antibody improves the outcome of H7N9 virus infection in African green monkeys

Structure and characterization of a high affinity C5a monoclonal antibody that blocks binding to C5aR1 and C5aR2 receptors

Algorithms for Highly Efficient, Load-Balanced, and Scalable Molecular Simulation

Computational design of proteins stereochemically optimized in size, stability, and folding speed

Open Source Drug Discovery Consortium. g_mmpbsa-a GROMACS tool for high-throughput MM-PBSA calculations

ACS Omega http://pubs.acs.org/journal/acsodf Article