(PDF) CD40/anti-CD40 antibody complexes which illustrate...

来自 : 发布时间：2024-05-13

(PDF) CD40/anti-CD40 antibody complexes which illustrate agonist and antagonist structural switchesArticlePDF AvailableCD40/anti-CD40 antibody complexes which illustrate agonist and antagonist structural switches August 2019BMC Molecular and Cell Biology 20(1)DOI:10.1186/s12860-019-0213-4Authors: Maria A. ArgiriadiMaria A. ArgiriadiThis person is not on ResearchGate, or hasn t claimed this research yet. Lorenzo BenatuilLorenzo BenatuilThis person is not on ResearchGate, or hasn t claimed this research yet. Ievgeniia DubrovskaIevgeniia DubrovskaThis person is not on ResearchGate, or hasn t claimed this research yet. David A. EganAbbVieShow all 17 authorsHide Download full-text PDFRead full-textDownload full-text PDFRead full-textDownload citation Copy link Link copied Read full-text Download citation Copy link Link copiedCitations (7)References (31)Figures (10)Abstract and FiguresBackground: CD40 is a 48 kDa type I transmembrane protein that is constitutively expressed on hematopoietic cells such as dendritic cells, macrophages, and B cells. Engagement of CD40 by CD40L expressed on T cells results in the production of proinflammatory cytokines, induces T helper cell function, and promotes macrophage activation. The involvement of CD40 in chronic immune activation has resulted in CD40 being proposed as a therapeutic target for a range of chronic inflammatory diseases. CD40 antagonists are currently being explored for the treatment of autoimmune diseases and several anti-CD40 agonist mAbs have entered clinical development for oncological indications.Results: To better understand the mode of action of anti-CD40 mAbs, we have determined the x-ray crystal structures of the ABBV-323 (anti-CD40 antagonist, ravagalimab) Fab alone, ABBV-323 Fab complexed to human CD40 and FAB516 (anti-CD40 agonist) complexed to human CD40. These three crystals structures 1) identify the conformational CD40 epitope for ABBV-323 recognition 2) illustrate conformational changes which occur in the CDRs of ABBV-323 Fab upon CD40 binding and 3) develop a structural hypothesis for an agonist/antagonist switch in the LCDR1 of this proprietary class of CD40 antibodies.Conclusions: The structure of ABBV-323 Fab demonstrates a unique method for antagonism by stabilizing the proposed functional antiparallel dimer for CD40 receptor via novel contacts to LCDR1, namely residue position R32 which is further supported by a closely related agonist antibody FAB516 which shows only monomeric recognition and no contacts with LCDR1 due to a mutation to L32 on LCDR1. These data provide a structural basis for the full antagonist activity of ABBV-323. a and b: (a) ABBV-323 was incubated with HEK-293 CD40L cells in the presence of CD40L expressing D1.1 cells. The ability of ABBV-323 to inhibit the interaction between CD40 and its receptor CD40L is monitored by the production of SEAP compared to an IgG control. (b) Agonist activity is monitored as above with the following exception: CD40L-expressing D1.1 cells were replaced with assay media. CD40L was titrated alongside ABBV-323 as a positive control (Concentration range = 250 ng/ml -0.026 ng/ml)…  Measuring agonist and antagonist activities of several CD40 antibody variants…  a and b: (a) ABBV-323 strongly inhibits CD40 signaling in B cells (inhibition of CD86 expression) without inducing agonist activity (stimulation of CD86 expression). To measure antagonist activity, CD40L+ jurkat cells were used to stimulate primary human B cells +/− ABBV-323. (b) To measure agonist activity, B cells were incubated with anti-CD40 antibodies…  Electrostatic potential surface calculated in Pymol for Fab ABBV-323 crystal structure. A cleft is formed between HCDR2 and LCDR1. Red patches refer to negative charged regions and blue patches refer to positive charged regions…  +5Complex structure of ABBV-323 Fab (magenta) and human CD40 (shown in grey). The Fab alone structure is superimposed (in cyan) to show how HCDR2 opens to accommodate CD40… Figures - available from: BMC Molecular and Cell BiologyThis content is subject to copyright. Terms and conditions apply. Discover the world s research20+ million members135+ million publications700k+ research projectsJoin for freePublisher Full-text 1Public Full-text 1 Access to this full-text is provided by Springer Nature.Learn moreDownload availableContent available from BMC Molecular and Cell BiologyThis content is subject to copyright. Terms and conditions apply. R E S E A R C H A R T I C L E Open AccessCD40/anti-CD40 antibody complexes whichillustrate agonist and antagonist structuralswitchesMaria A. Argiriadi1*, Lorenzo Benatuil2, Ievgeniia Dubrovska3, David A. Egan3, Lei Gao2, Amy Greischar3,Jennifer Hardman2, John Harlan3, Ramesh B. Iyer3, Russell A. Judge3, Marc Lake3, Denise C. Perron2,Ramkrishna Sadhukhan2, Bernhard Sielaff2, Silvino Sousa2, Rui Wang2and Bradford L. McRae2AbstractBackground: CD40 is a 48 kDa type I transmembrane protein that is constitutively expressed on hematopoieticcells such as dendritic cells, macrophages, and B cells. Engagement of CD40 by CD40L expressed on T cells resultsin the production of proinflammatory cytokines, induces T helper cell function, and promotes macrophage activation.The involvement of CD40 in chronic immune activation has resulted in CD40 being proposed as a therapeutic targetfor a range of chronic inflammatory diseases. CD40 antagonists are currently being explored for the treatment ofautoimmune diseases and several anti-CD40 agonist mAbs have entered clinical development for oncologicalindications.Results: To better understand the mode of action of anti-CD40 mAbs, we have determined the x-ray crystal structuresof the ABBV-323 (anti-CD40 antagonist, ravagalimab) Fab alone, ABBV-323 Fab complexed to human CD40 and FAB516(anti-CD40 agonist) complexed to human CD40. These three crystals structures 1) identify the conformational CD40epitope for ABBV-323 recognition 2) illustrate conformational changes which occur in the CDRs of ABBV-323 Fab uponCD40 binding and 3) develop a structural hypothesis for an agonist/antagonist switch in the LCDR1 of this proprietaryclass of CD40 antibodies.Conclusions: The structure of ABBV-323 Fab demonstrates a unique method for antagonism by stabilizing the proposedfunctional antiparallel dimer for CD40 receptor via novel contacts to LCDR1, namely residue position R32 which is furthersupported by a closely related agonist antibody FAB516 which shows only monomeric recognition and no contacts withLCDR1 due to a mutation to L32 on LCDR1. These data provide a structural basis for the full antagonistactivity of ABBV-323.Keywords: CD40, Crystal structure, Agonist, Antagonist, AntibodyBackgroundCD40 is a 48 kDa type I transmembrane protein that isexpressed on a wide range of hematopoietic (B cells,monocytes/macrophages, dendritic cells) and non-hematopoietic (activated epithelium, endothelium) cells.Its ligand, CD154 or CD40L, has a more restricted ex-pression pattern and is found primarily on activated Tcells, B cells, and platelets. Engagement of CD40 byCD40L results in the recruitment of TNF receptorassociated factors (TRAFs) to the cytoplasmic domain ofCD40 [1]. Phosphorylation of various TRAF proteins re-sults in activation of both canonical and non-canonicalNF-kB pathways. TRAF6-dependent PI3K activation is acritical survival signal while TRAF2/TRAF6 have redun-dant functions in NF-kB activation and upregulation ofCD80 and ICAM-1 expression [2]. Much of our under-standing of CD40/CD40L biology comes from theinteraction between antigen presenting cells [CD40 ex-pression on either dendritic cells (DC) or B cells] andCD40L-expressing T cells. Cell-cell interactions betweenantigen presenting cells and T cells provide bidirectional© The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.* Correspondence: maria.argiriadi@abbvie.com1AbbVie Bioresearch Center, 381 Plantation Street, Worcester, MA 01605, USAFull list of author information is available at the end of the articleBMC Molecular andCell BiologyArgiriadi et al. BMC Molecular and Cell Biology (2019) 20:29 https://doi.org/10.1186/s12860-019-0213-4Content courtesy of Springer Nature, terms of use apply. Rights reserved. signaling that is critical for the activation, maturation,and effector function of both cell types.CD40 expression on epithelium, leukocytes, and vas-cular endothelium is elevated in organ-specific auto-immune diseases such as Crohn’s disease, ulcerativecolitis, and rheumatoid arthritis and in systemic auto-immunity such as systemic lupus erythematosus (SLE).In addition, CD40+ monocytes, dendritic cells, and Bcells are enriched at sites of chronic inflammation. Dis-ruption of this signaling pathway has the potential to re-duce production of proinflammatory cytokines, reduce Thelper cell function, and inhibit macrophage activation,making it a very attractive therapeutic target for patientswith chronic inflammatory disease [3]. X-ray structuralstudies have reported the CD40L-CD40 complex where2 CD40 receptor monomers were bound to the intersu-bunit grooves of the CD40L trimer [4]. This protein-pro-tein association was in part enabled by the receptor’slong and flexible structural fold containing 3 tandemcysteine-rich domains (CRDs). Each CD40 monomer as-sociated within the CD40L intersubunit groove throughvariety of polar and hydrophobic interactions. Chargecomplementarity specifically played a large role in thisassociation. An antibody that would prevent these inter-actions would act as an antagonist.Attempts to disrupt the CD40-CD40L pathway for thetreatment of autoimmunity and transplant rejection havebeen limited due to safety issues linked to the functionalproperties of specific biologic therapies. Two mAb pro-grams (BG9588, Biogen; IDEC-131, IDEC) targetingCD40L entered clinical development for Systemic LupusErythematosis (SLE) and Crohn’s disease [3]. Develop-ment of both BG9588 and IDEC-131 was halted aftermultiple cases of thrombosis were reported. Subsequentstudies suggested that antibodies against CD40Lexpressed on platelets may be cross-linked by FcγR alsoon platelets resulting in platelet degranulation and ag-gregation [5,6]. Antibodies directed against CD40 suchas BMS-224819 have been shown to block CD40-CD40Lbinding but have partial agonist activity resulting insome signaling through CD40 and peripheral B cell de-pletion [7]. In addition to B cell depletion, antibodieswith CD40 agonist activity produce increases in liver en-zymes and cytokine release that present safety concernsin patients with autoimmune disease [8]. Due to thevarious functional properties of anti-CD40 mAbs, it wascritical to understand the molecular interactions of anti-CD40 mAbs with CD40. Therefore, we determined thecrystal structures of ABBV-323 (antagonist) Fab aloneand ABBV-323 and FAB516 (agonist) Fabs complexed tothe 3 extracellular domains of human CD40 along within vitro functional assays to better understand the pro-tein interactions that govern agonist and antagonistactivity of CD40 mAbs. Here we describe the molecularinteractions that produce the full antagonist activity ofABBV-323. ABBV-323 (ravagalimab) has completedphase 1 testing in healthy volunteers and is currently en-tering phase 2 studies for the treatment of ulcerativecolitis.ResultsTo identify CD40 specific antibodies, hybridoma tech-nology was applied by immunizing mice with humanCD40 antigen and adjuvant. Hybridoma antibodies werescreened based on the following criteria: 1) humanCD40 binding and neutralization of CD40 (EC50 10nM), 2) no evidence of agonist activity in reporter assaysor on primary human cells ( 100 nM), 3) retention ofpotent antagonist activity after expressing VH and VL ashuman chimeric antibodies, and 4) cross-reactivity tocynomolgus monkey CD40.One chimeric antibody from the primary screen wasselected for humanization, and affinity and liability en-gineering resulting in the lead mAb. Several variants ofthe humanized lead were produced by introducingsemi-rational mutations in HCDR2. An HCDR2 S56Gmutation variant was chosen to move forward due to a10-fold improvement in affinity. Leucine to alanine mu-tations in the Fc region (234/235) were introduced tominimize FcgR binding and Fc-mediated effector func-tion. Mutations that enhanced FcRn binding (T250Q,M428 L) were introduced to increase antibody half-life.We used a CD40-dependent NF-kB reporter assay toscreen for potent CD40 antagonist. This system usedrecombinant CD40L to stimulate HEK-293 transfectedwith human CD40. NF-kB activation drove SEAP pro-duction which was quantitated using spectrophotom-etry. This was a very sensitive assay of both antagonistactivity (ability to inhibit CD40-mediated signaling) andagonist activity (ability of the anti-CD40 mAbs tostimulate cells through CD40). Our lead mAb, ABBV-323, is a human IgG1/κ(hCg1_z,non-a allotype) anti-body with minimum FcγR binding and Fc-mediatedeffector function due to the introduction of Leucine toalanine mutations in the Fc region (234/235) [9]. It alsocontains mutations that enhance FcRn binding (T250Q,M428 L) to increase antibody half-life [10]. ABBV-323 was apotent CD40 antagonist as measured in this reporter assay(EC50= 3.7 nM) and showed no evidence of agonist activityin the absence of CD40L (data not shown).We developed a primary B cell assay to test the antag-onist and agonist activity of anti-CD40 mAbs in a morephysiologically relevant system. B cells express the high-est levels of CD40 of any cell type in human peripheralblood and also express FcγR which allowed us to evalu-ate potential for residual FcγR binding in promotingCD40 agonist activity. To measure antagonist activity,Jurkat cells were used as a source of native CD40L andArgiriadi et al. BMC Molecular and Cell Biology (2019) 20:29 Page 2 of 13Content courtesy of Springer Nature, terms of use apply. Rights reserved. CD40-dependent B cell activation was measured bycell proliferation and induction of CD86 by flow cy-tometry. ABBV-323 inhibited CD86 upregulation onprimary human B cells with an EC50value of 0.6 nM(Fig 1a). In the agonist assay a known CD40 agonist,CP-870,893 [11], dose-dependently induced CD86 ex-pression on B cells while ABBV-323 showed no evi-dence of agonist activity (Fig. 1b).Receptor crosslinking is known to be critical to initiatedownstream signaling for TNF receptor family members.Although Fc crosslinking was suggested to play an im-portant role under certain conditions, it was shown tobe not essential for agonizing CD40 receptor (Richman,et al. Cancer Immunol Res. 2014) [12–14]. On the con-trary, we have demonstrated that the antibody bindingepitope plays an important role in function and that theFc crosslinking is not required for agonist activity.A crystal structure of ABBV-323 Fab was solved to1.74 Å resolution and the ABBV-323 Fab/CD40 complexstructure was solved to 2.84 Å resolution. Critical obser-vations include the identification of the 3-dimensionalconformational epitope of ABBV-323 and the key struc-tural rearrangement needed for its Fab to accommodatethe CD40 antigen. The structure of ABBV-323 Fab aloneillustrates two loops (HCDR2 and LCDR1) which createa cleft of approximately 25 Å with a diversity of charge(Fig. 2). The structure of ABBV-323 Fab bound to CD40exhibits two complexes in the asymmetric unit. Withinthe asymmetric unit, each complex contains one CD40monomer associated with the Fab on CRD2 (Fig. 3).When overlaying the two complexes within the repeat-ing crystal unit, there is a slight shift in CRD3 due to theinherent flexibility of the cysteine rich domains, howeverthe epitope contacts to CRD2 are constant. When com-paring the complex structure with the Fab alone struc-ture, there is a backbone shift in HCDR2 that allowsopening of the cleft to accommodate CD40 binding(Fig. 3). As described earlier, mutation data supportsthis structural observation: a HCDR2 S56G residuechange results in 20-fold improvement in bindingaffinity due to higher flexibility introduced by a gly-cine at this position. This flexibility most likely aidsFab cleft opening and facilitates a hydrogen bondinteraction of (H)R55 with CD40.The conformational epitope of ABBV-323 showsseveral critical interactions. Specifically, CD40 residueK94 inserts into the central cleft of ABBV-323 andspecifically binds to a negatively charged region out-lined by residues (L)D97 and the backbone carbonylof (H)G101 (Fig. 4a and b). Additional epitope inter-actions are observed for both heavy and light chains:CD40 E64 displays hydrogen bonds to (H)S53/(H)Y50and CD40 T99 displays a hydrogen bond to (L)N31.When overlaying the complex structure with a previ-ously reported structure of CD40/CD40L (PDB code:3QD6 [4]), the CD40 antigens adopt similar confor-mations in the region of CRD1 and CRD2. From thesuperposition, the Fab HCDR2 would minimally clashwith several residues on an external loop of CD40L(amino acid residues S128-K143). Therefore, it is con-ceivable that this loop could move to accommodatesimultaneous binding of CD40L, CD40 monomer andFab in an agonist-like ternary complex.When generating symmetry crystallographic equiva-lents of the CD40/ABBV-323 Fab complex, CD40 is ob-served as a tight antiparallel homodimer (Fig. 5a). Theinterface of the homodimer spans approximately 40 Ålong with extensive hydrogen bond complementarityand VDW contacts, and a buried surface area of 3292.5Å2. Additionally, a PISA analysis predicts this homodi-meric interface to be stable in solution with a calculatedΔGdiss= 24.3 kcal/mol. Not only is ABBV-323 bound tothe primary monomer (\"monomer 1”), a second mono-mer (\"monomer 2”) of the crystallographic dimer associ-ates with ABBV-323 through the N-terminal end ofCD40’s CRD1 which includes the amino acid residuesFig. 1 aand b:(a) ABBV-323 strongly inhibits CD40 signaling in B cells (inhibition of CD86 expression) without inducing agonist activity(stimulation of CD86 expression). To measure antagonist activity, CD40L+ jurkat cells were used to stimulate primary human B cells +/−ABBV-323.(b) To measure agonist activity, B cells were incubated with anti-CD40 antibodiesArgiriadi et al. BMC Molecular and Cell Biology (2019) 20:29 Page 3 of 13Content courtesy of Springer Nature, terms of use apply. Rights reserved. Fig. 2 Electrostatic potential surface calculated in Pymol for Fab ABBV-323 crystal structure. A cleft is formed between HCDR2 and LCDR1. Redpatches refer to negative charged regions and blue patches refer to positive charged regionsFig. 3 Complex structure of ABBV-323 Fab (magenta) and human CD40 (shown in grey). The Fab alone structure is superimposed (in cyan) toshow how HCDR2 opens to accommodate CD40Argiriadi et al. BMC Molecular and Cell Biology (2019) 20:29 Page 4 of 13Content courtesy of Springer Nature, terms of use apply. Rights reserved. T24-C37. Specifically, a position off of the Fab’s LCDR1,(L)R32 bridges two strands of CD40 CRD1 contactingbackbone carbonyls from CD40 (monomer 2) A25/S35and a potential sidechain interaction with Q36 (Fig. 5b).This observation led to a structural hypothesis thatABBV-323 recognizes a biologically relevant CD40 \"in-active”antiparallel homodimer.The concept behind a CD40 functional dimer is alsosupported by a previous report where CD40 1) demon-strated homodimerization in soluble and cell surface-expressed environments, and 2) CD40 dimerization wasdependent on the extracellular domain [15]. Addition-ally, a homology model was created of the homodimerbased on the ligand-free TNFR1 structure and mutationsin the CRD1 domain of CD40 were made based on thedimeric model. When mutating K29 (a key dimerizationresidue in the model) to a non-polar or acidic residue(K29A and K29E), self-assembly decreased [15]. In ourreported crystal structure of ABBV-323 Fab complexedto CD40, K29 participates in the observed antiparallelcrystallographic dimer. Specifically, K29 is proximal tothe dimeric interface and makes a potential interactionwith a nearby residue from the second CD40 monomerE28. It is likely that these two residues engage in a sta-bilizing hydrogen bond interaction for the CD40 homo-dimer. Therefore, mutating this residue position to anon-polar or acidic residue could negatively affect self-assembly which requires charge complementarity fordimerization.Structural insights from previous reports suggest thatunliganded TNF family receptors dimerize in antiparallelorientations between the first two disulfide-rich motifsabFig. 4 aand b:(a) K94 inserts into a negatively charged channel to make interactions with acidic pocket. (b) Interactions between K94 and(H)G101/(L)D97 are illustratedabFig. 5 aand b:(a) Two crystallographic CD40 monomers associate to form a tight antiparallel dimer (grey, green) which bind to ABBV-323 Fab(shown in magenta). (b) LCDR1 R32 inserts in the second crystallographic monomer shown in green. R32 makes interactions with backbonecarbonyls to A25/S35 and sidechain Q36Argiriadi et al. BMC Molecular and Cell Biology (2019) 20:29 Page 5 of 13Content courtesy of Springer Nature, terms of use apply. Rights reserved. of the receptor thereby forming a non-signaling struc-tural state [16,17]. This is potentially a similar scenariofor CD40 receptor which is a member of the TNFαsuperfamily. Antiparallel CD40 homodimers would pre-vent downstream cytoplasmic signaling proteins to bindto CD40 receptor because 1) CD40 antiparallel dimerswould occlude the CD40L binding site which wouldthen 2) prevent CD40 from clustering to recruit TRAFadaptor proteins that activate different signaling path-ways [13]. The co-crystal structure of ABBV-323 Fab/CD40 supports this idea. When comparing the CD40antiparallel dimer with a previously reported TNFR1antiparallel dimer, the antiparallel recognition is differ-ent: CD40 dimerizes with a larger buried surface area viathe CRD1–2 domains while TNFR1 dimerizes primarilythrough the CRD2 domain with a smaller buried dimericsurface area (1475 Å2, Additional file 1: Figure S1) [16].When overlaying the ABBV-323 Fab complex structure(including the crystallographic antiparallel homodimer)with the historical structure of CD40/CD40L (PDB code3QD6 [4]), the superposition demonstrates that CD40Lwould be unable to bind the antiparallel homodimer ofCD40 due to steric clashes spanning approximately 16Å, specifically in the CD40L loop region L195-I205(Fig. 6). We therefore conclude that ABBV-323 func-tions as a CD40 antagonist antibody due to its abilityto capture the non-signaling dimeric state of CD40receptor which sterically masks CD40L recognition. Abetter understanding of the molecular basis for agon-ist and antagonist activity of anti-CD40 antibodies iscritical for the rational design of therapeutics withthe desired efficacy and tolerability profile.To further support this idea of a non-signalingdimeric state, we examined several variants of ABBV-323 with small changes to sequence in LCDR1 in-cluding changing position (L)R32 to proline (FAB516)or leucine (FAB518). These changes resulted in con-version of the antibody from antagonist to agonist ac-tivity as demonstrated in a human CD40 reporterassay (Fig. 7a, b; Table 1). This supports the claimthat the LCDR1 loop in ABBV-323 (specifically aminoacid position R32) may be critical in binding thesecond monomer of CD40 through specific polar in-teractions and thereby stabilizing the inactive antipar-allel homodimer. When changing this residue to anon-polar leucine or proline, this functional dimerwould not form due to lack of polar contacts withLCDR1. We predicted that x-ray complexes of agonistvariants such as FAB516 or FAB518 (which respect-ively have a proline ((L)P32) and leucine ((L)L32) inplace of an arginine on LCDR1) would exhibit differ-ent crystallization packing and oligomerization incontrast to ABBV-323. To confirm this hypothesis,the crystal structure of agonist FAB516 Fab com-plexed to CD40 was solved to 3.1 Å resolution. Thestructure shows very clear electron density for theCD40/Fab interface and crystallographic packing isunambiguous. Comparison of the two antibody com-plexes (CD40/ABBV-323 and CD40/FAB516) was thenneeded to understand oligomerization of the receptor.Fig. 6 Overlay of ABBV-323/CD40 complex (magenta = Fab/CD40 monomer 1, green = CD40 crystallographic monomer 2) with crystal structure ofCD40L/CD40 complex (PDB code: 3QD6 in grey surface) [4]. Red outlined box shows the region of significant steric clash between CD40 dimerand CD40LArgiriadi et al. BMC Molecular and Cell Biology (2019) 20:29 Page 6 of 13Content courtesy of Springer Nature, terms of use apply. Rights reserved. DiscussionWhen overlaying the backbones of FAB516 and ABBV-323 CD40 complex structures, the CD40 monomerrecognition for both Fabs is identical. However as pre-dicted, the crystallographic dimer formation is different.When generating the symmetry equivalents of theFAB516 complex structure, an antiparallel CD40 dimeris again observed however the second CD40 monomeris in a different orientation when compared to theCD40 homodimer in the ABBV-323 complex structure(Fig. 8a and b). A PISA analysis for the FAB516-CD40complex calculates the crystallographic CD40 homodi-mer interface to have a smaller buried surface area(1397.6 Å2) and a smaller predicted ΔGdiss(2.5 kcal/mol) when compared to the values calculated for theABBV-323 complex suggesting that the FAB516 CD40dimer interface is less physiologically relevant. Mostimportantly, the FAB516 Fab binds predominantly toone monomer of CD40. Specifically, LCDR1 (L)L32does not engage in interactions with a second CD40monomer (Fig. 8b) for 4 out of the 5 Fab-antigen com-plexes in the asymmetric unit. Due to these observa-tions, we conclude that FAB516 recognizes CD40monomer and not the functional homodimer as illus-trated with the ABBV-323 complex.This difference in CD40 recognition supports theagonist activity seen for FAB516. When observing asuperposition of the backbone structure of the FAB516complex structure with the CD40L/CD40 complexstructure (PDB code 3QD6 [4]), there are several clashesbetween the FAB516/CD40 monomeric complex andCD40L, specifically in the region of the heavy chainCDRs. This observation is in agreement with blockingresults (Table 1). However, with some structural re-arrangement, an agonist ternary complex (CD40/CD40L/FAB516, Fig. 9a and b) could also be hypothe-sized suggesting that FAB516 could act potentially as anagonist stabilizer. Specifically, when minimizing the pro-posed ternary complex, HCDR2 (residues S52-G56) andCD40L (residues S128-K132) conformationally adjust tocreate the complex. This concept is in agreement withreporter assay results in Table 1(Fig. 7a and b).We conclude that a non-polar residue at positionLCDR1(32) as seen in FAB518 (proline) and FAB516(leucine), allows agonist antibody recognition of mono-meric CD40 and a potential ternary complex withCD40L. In contrast, when this position switches to abasic polar residue (specifically in the case of antagonistABBV-323 (L)R32), the antibody recognizes an antiparal-lel dimeric CD40 which prevents CD40L binding.abFig. 7 aand b:(a) ABBV-323 was incubated with HEK-293 CD40L cells in the presence of CD40L expressing D1.1 cells. The ability of ABBV-323 toinhibit the interaction between CD40 and its receptor CD40L is monitored by the production of SEAP compared to an IgG control. (b) Agonistactivity is monitored as above with the following exception: CD40L-expressing D1.1 cells were replaced with assay media. CD40L was titratedalongside ABBV-323 as a positive control (Concentration range = 250 ng/ml –0.026 ng/ml)Table 1 Measuring agonist and antagonist activities of several CD40 antibody variantsAntibody variants VL LCDR1 Sequence(KSSQSLLNSGNQKNYLT)Blocking of CD40L Agonist: huCD40reporter assay IC50 nMAntagonist: Jurkat/Reporter assay IC50 nMABBV-323 KSSQSLLNRGNQKNYLT Yes No 3.1FAB518 KSSQSLLNPGNQKNYLT Yes 62 NoFAB516 KSSQSLLNLGNQKNYLT Yes 157 NoArgiriadi et al. BMC Molecular and Cell Biology (2019) 20:29 Page 7 of 13Content courtesy of Springer Nature, terms of use apply. Rights reserved. ConclusionIn this study, the structures of ABBV-323 Fab alone andABBV-323/FAB516 Fabs in complex to CD40 weresolved. The structure of ABBV-323 demonstrates aunique method for antagonism by stabilizing the pro-posed functional antiparallel dimer for CD40 receptorvia novel contacts to LCDR1, namely R32. This was fur-ther supported by a closely related agonist antibodyFAB516 which showed only monomeric recognition andno contacts with LCDR1 due to a mutation to L32 onLCDR1. Therefore, a structural agonist/antagonistswitch in this class of antibodies was identified. Anti-bodies for antiparallel dimers have also been recently re-ported for the TNFR2 receptor in the inhibition ofproliferation of ovarian cancer cells [18]. In this report, amodel was created demonstrating how the TNFR2 anti-parallel dimer locked into its quiescent state via theantagonist antibody to prevent TNF binding and signal-ing. It was also hypothesized that these resting stateTNFR2 dimers could also arrange into a higher orderhexagonal lattice on the cell surface. Although we didnot observe a hexagonal arrangement in the crystalpacking of the ABBV-323/CD40 complex structure, itwould be interesting to observe the nature of the CD40resting state antiparallel dimers using orthogonal tech-niques such as SAXS or Cryo-EM with the potential ofobserving higher order CD40 oligomerization. In con-clusion, our structural findings which demonstrate theABBV-323 antibody’s mechanism of action suggest agrowing strategy for designing and identifying antago-nists to not only CD40 receptor but other members ofthe TNF superfamily.MethodsBioactivity measurement of CD40 antibodies –primary Bcell assayPrimary human B cells were purchased from BioSpecialtyand were used to study CD40 signaling. Cells were cul-tured in RPMI with 10% heat-inactivated serum, 2mM L-glutamine, 1mM sodium pyruvate and 100μg/ml PenStrep. To measure agonist activity, the cells were plated at0.1 million per well of a 96-well plate and dilutions ofCD40 antibodies were incubated with the cells for 2days.Cells were then harvested and CD86 upregulation wasabFig.8 aand b(a) ABBV-323 Fab (blue) bound to crystallographic CD40 dimer (second monomer shown in green ribbon) which shows interactionwith LCDR1 R32. (b) FAB516 Fab (orange) bound to crystallographic CD40 dimer (second monomer shown in green ribbon) which shows nointeraction with LCDR1 L32. N and C termini are also labeled to demonstrate the antiparallel dimers for both ABBV-323 and FAB516abFig. 9 aand b:(a) Model of potential ternary complex of agonist FAB516 Fab (cyan), human CD40 (green), human CD40L (purple). Box outlinesregion where minor structural rearrangements may occur to accommodate agonist complex. (b) After ternary complex minimization in MAESTRO,HCDR2 (S52-G56) and CD40L (S128-K132) residues shift to accommodate the interface between CD40L, FAB516 and CD40Argiriadi et al. BMC Molecular and Cell Biology (2019) 20:29 Page 8 of 13Content courtesy of Springer Nature, terms of use apply. Rights reserved. measured by FACS. To assess antagonist activity, B cellswere incubated with CD40L-expressing Jurkat cells(ATCC CRL-10915) at 2:1 ratio per well of 96-well platein the presence of CD40 antibodies. Cells were harvested2 days later and stained for CD19, CD20 (B cell marker)and CD86. Briefly, cells were washed by PBS and incu-bated with antibodies in PBS with 2% serum for 20 minon ice. The cells were then washed by PBS and analyzedby a BD LSRFortessa™cell analyzer. Anti-CD19 and anti-CD20 were purchased from BD bioscience and anti-CD86was from eBiosciences. The ability of the anti-CD40 anti-bodies to inhibit the interaction between CD40 and its re-ceptor CD40L was compared to an IgG control.HEK-293 blue CD40L –antagonist reporter assayHEK-293 Blue CD40L Cells were seeded into 96-well flatbottom TC plates at a density of 5 × 104cells/100 μl/well. The plates were incubated for 1 h at 37 °C, 5%CO2. Stock solutions of CD40 antibodies were preparedin assay medium at a 4x concentration (400 nM), seriallydiluted 1:2.5 and added to HEK-293 cells (50 μl /well).CD40L-expressing D1.1 cells were diluted to a concen-tration of 1 × 106/ml in assay media and added to allwells (50 μl /well) bringing the final Ratio of D1.1 toHEK-293 to 1:1. Plates were incubated for 18 h at 37 °C,5% CO2. Following incubation, 50 μL of supernatantmedia was removed and transferred to a new plate forQUANTI-Blue luminescence detection. IC50 valueswere obtained using logarithm of antibody concentrationvs. luminescence nonlinear regression (4-parameterdose-response curve model:Y¼Bottom þTop‐BottomðÞ=1þ10∧LogIC50‐XðÞHillSlopeÞðÞðÞin GraphPad Prism 5:Three independent experimentswere performed:HEK-293 blue CD40L –agonist reporter assayAssay performed as above with the following excep-tion: CD40L-expressing D1.1 cells were replaced with50 μl assay media. A 4x stock solution of recombinanthuman MegaCD40L (positive control) was prepared at1μg/ml in assay media and serially diluted 1:2.5. Dilu-tions were added alongside CD40 antibodies. Concen-tration range = 250 ng/ml –0.026 ng/ml.Preparation and purification of CD40 antigenA DNA sequence encoding the human CD40 extra-cellular domain (UNIPROT P25942; amino acids 1–193) was synthesized and cloned into our proprietarypHybE (US Patent 8187836 B2) vector followed by anin-frame C-terminal Tev protease cleavage site andhexahistidine tag (sequence ENLYFQGHHHHHH).The pHybE expression vector utilizes an EF-1αpro-moter and an OriP origin of replication derived fromEpstein-Barr virus (EBV). This plasmid was trans-fected into HEK-293-6E cells (NRC Canada) grown inFreestyle 293 medium (Thermo Fischer) at 1 × 106cells/ml using the transfection reagent Polyethyleni-mine (PEI, Polysciences Inc) at a PEI:DNA ratio of 4:1. The transfected cell culture was fed with tryptone-N1 (Organotechnie) (to 0.5%) at 24 h post-transfec-tion. On day 7 post-transfection, the transfected cellculture was cleared by centrifugation followed by filtra-tion through 0.2 μm PES filter (Corning). Clearedmedium was buffer exchanged to PBS, pH 7.4 using aKvick TFF system equipped with 10 kDa membranes(GE Healthcare) and loaded on a 5 ml HisTrap FFcolumn (GE Healthcare) equilibrated with PBS, pH7.4. The column was washed with 25 mM imidazolein PBS, pH 7.4 and bound protein was eluted with250 mM imidazole in PBS, pH 7.4. Eluted protein wasconcentrated using Amicon Ultra-15 centrifugal filterdevices (Millipore) with 10 kDa molecular weight cut-off, and further purified by SEC on a 26/60 Superdex200 column (GE Healthcare) equilibrated and runwith PBS, pH 7.4. Fractions containing CD40 werepooled, concentration measured by absorbance at 280nm, and samples analyzed by SEC, SDS-PAGE, andmass spectrometry. [CD40(h)(21–193)]-Tev-His6 wasstored in aliquots at −80 °C.Preparation and purification of CD40 ABBV-323 FabfragmentFab fragment of CD40 ABBV-323 was prepared bypapain cleavage of the parent mAb ABBV-323 asdetailed below. Papain (Worthington Biochemical,Lakewood, NJ) was activated with 50 mM cysteine inPBS, pH 7.4 buffer. The parent mAb in PBS, pH 7.4buffer was mixed with papain at 1:100 weight ratio ofpapain to mAb and incubated for 1 h at 37 °C. Thereaction was quenched with 5 mM iodoacetamide.The mixture was purified on 10 ml Mab SelectSureresin (GE Healthcare) where the Fab fragment wascollected as flow through. The flow through was con-centrated using an Ultrafree-15 Biomax 10 kDa mo-lecular weight cut-off (MWCO) centrifugal device(Millipore). The concentrated mixture was purified on2.6 cm × 60 cm Sephacryl 200 HiPrep column (GEHealthcare) pre-equilibrated in 50 mM HEPES, 50 mMNaCl, pH 7.5 buffer. Fractions containing Fab frag-ment (monitored by UV absorbance at 280 nm) werepooled and frozen at −80°C.Samplepuritywasassessed by analytical SEC, SDS-PAGE and massspectrometry. Table 3lists protein quality character-izations for the reagents used in this study.Argiriadi et al. BMC Molecular and Cell Biology (2019) 20:29 Page 9 of 13Content courtesy of Springer Nature, terms of use apply. Rights reserved. CD40/CD40 ABBV-323 Fab complex preparationRecombinant human CD40 was expressed and purified asdescribed above. Recombinant human CD40 and CD40ABBV-323 Fab protein were mixed at a 1.05:1 M ratio(1.8 mg/ml final concentration) and incubated for 4 h at4 °C. The complex sample was loaded onto a 2.6 cm × 60cm Sephacryl 200 HiPrep column (GE Healthcare) pre-equilibrated in 50 mM HEPES, 50 mM NaCl, pH 7.5 buf-fer at 1 ml/min. Fractions containing the complex (moni-tored by UV absorbance at 280 nm) were pooled andconcentrated to 18 mg/ml using an Ultrafree-15 Biomax10 kDa molecular weight cut-off (MWCO) centrifugaldevice (Millipore). Sample purity was assessed by analyt-ical SEC and SDS-PAGE. Excess Fab-Complex proteinwas stored frozen at −80 °C.ABBV-323 Fab crystallizationFab alone was supplied at 22.5 mg/ml in 50 mM HEPES,50 mM NaCl, pH 7.5. Crystals grew by vapor diffusion at23 °C. The reservoir contained 25% (w/v) PMME 550,0.1 M MES pH 6.5, 0.01 M zinc sulfate. The drop wasmade by adding equal volumes of protein and reservoirsolution. Crystals grew as thick prisms and were cryo-protected using the reservoir solution with the additionof 10%(v/v) propylene glycol. Crystals were harvested,swished through cryo-solution and cryo-cooled directlyin liquid nitrogen. Diffraction data to 1.74 Å were col-lected under gaseous nitrogen at 100 K at the 17IDbeamline at the Advanced Photon Source at ArgonneNational Laboratories (Argonne IL).ABBV-323 Fab complexed to CD40 antigen crystallizationThe Fab complex was supplied at 18 mg/ml in 50 mMHEPES, 50 mM NaCl, pH 7.5. The antigen constructused was [CD40 (h) (21–193)]-TEV-6His. Crystals grewby vapor diffusion at 23 °C. The reservoir contained 2 Mammonium sulfate, 0.1 M phosphate-citrate pH 4.2. Thedrop was made by adding equal volumes of protein andreservoir solution. Crystals grew as thin rods and werecryo-protected using 2.5 M lithium sulfate. Crystals wereharvested, swished through cryo-solution and cryo-cooled directly in liquid nitrogen. Diffraction data to2.84 Å were collected under gaseous nitrogen at 100 K atthe 17ID beamline at the Advanced Photon Source atArgonne National Laboratories (Argonne IL).Preparation of FAB516 complex by direct co-expressionFermentationThe expression plasmid (pHybE) for heavy chain ofFAB516 was synthesized with secretion leader sequence,variable region, CH1 and hinge region (ending at H224,Eu numbering). The light chain plasmid (pHybE) wassynthesized with secretion leader sequence, variable re-gion and CL region (ending at C214, Eu numbering).The plasmid for antigen expression was constructed asdescribed above. HEK-293-6E cells, which is a suspen-sion adapted human embryonic kidney-293-based cellline stably expressing the Epstein–Barr virus nuclearantigen (EBNA1), were transfected with plasmid DNAencoding the heavy chain (HC) and light chain (LC) ofFAB516 and plasmid for the CD40 antigen ([CD40(h)(1–193)]-TEV-6His). For a 3 L expression, a 5 L flask(Thompson Instrument Company) containing cells at1.2 × 106cells/flask were grown in Freestyle 293 Expres-sion medium at a temperature of 37 °C, with 8% CO2,and shaking at 80 rpm. For transfection, 1.5 mg of DNAin 1:2:3 ratio (HC:LC:Antigen) was mixed with 6 ml of 1mg/ml pH 7.0 PEI solution in a volume of 150 ml Free-style medium. After 10 min of incubation, the mixturewas added to the cells. Tryptone N1 (Organotechnie) inFreestyle medium was added to the flask at 24h posttransfection for 0.5% final concentration. The conditionedmedium was harvested 9–11 days post transfection bycentrifugation at 16 K x G for 10 min, followed by clarifi-cation through a Pall AcroPak 500 0.8/0.45 μmfiltercapsule, and sodium azide was added from a 1 M stock toa concentration of 5 mM. The conditioned media wasstored at 4 °C until purified.CD40/CD40 FAB516 Fab complex preparationThe cell culture supernatant containing Fab/antigencomplex was loaded onto a 22 ml Ni Excel (GE LifeSciences) column at a flow rate of 5 ml/min, and washedwith 5 column volumes of binding buffer (50 mM Tris,350 mM sodium chloride, pH 8.0). The non-specific pro-teins bound to the column were washed with 5 columnvolumes of binding buffer containing 20 mM imidazole.The Fab/antigen complex was eluted by gradient elutionfrom 5 to 100% elution buffer (50 mM Tris, 300 mM so-dium chloride 400 mM imidazole pH 8) in 10 columnvolumes. The Fab/antigen complex was collected in 7 mlfractions. Size-exclusion chromatography was performedas a polishing step using a HiLoad 26/60 Superdex 75column (GE Life Sciences). The Superdex 75 columnwas run at 1 ml/min with 50 mM HEPES, 50 mM NaClpH 7.5. Fractions containing Fab/antigen complex wascharacterized with SEC, SDS-PAGE gel and mass spec,and concentrated to 20.0 mg/ml for crystallography.FAB516 Fab complexed to CD40 antigen crystallizationThe Fab complex crystals grew by vapor diffusion usingsitting drop at 23 °C. The reservoir contained 1.6 MAmmonium Sulfate, 2% w/v PEG1000, 100 mM HEPESSodium Salt pH 7.5 (JBS-II screen from Jena Bioscience,conditions A9). The drop was made by adding equal vol-umes of protein and reservoir solution (0.4 μL + 0.4 μL).Diffracting crystals took 11 months to appear and werecryo-protected using the reservoir solution with 20%Argiriadi et al. BMC Molecular and Cell Biology (2019) 20:29 Page 10 of 13Content courtesy of Springer Nature, terms of use apply. Rights reserved. Ethylene glycol. Crystals were harvested, swishedthrough cryo-solution and cryo-cooled directly in li-quid nitrogen. Diffraction data to 3.13 Å were col-lected under gaseous nitrogen at 100 K at the 17IDbeamline at the Advanced Photon Source at ArgonneNational Laboratories (Argonne IL).Structure determinations of ABBV-323 Fab and CD40antigen complexesDiffraction data for both crystal structures were proc-essed using the program AUTOPROC with isotropicscaling from Global Phasing Ltd. [19] The ABBV-323Fab dataset was processed in the space group C2221withthe following unit cell dimensions: a = 64.65 b = 130.4c = 132.6. A maximum likelihood molecular replacementsolution was determined using the program PHASER[20] using a Fab search model reported previously(Protein Data Bank entry 3QOS [21]). Coordinates for 1Fab molecule in the asymmetric unit were generatedbased on the molecular replacement solution. Prelimin-ary refinement of the resulting solution was conductedusing REFMAC [22,23] and the program BUSTER [24].Iterative protein model building was conducted usingthe program COOT [25] and examination of 2Fo-Fc andFo-Fc electron-density maps. Water molecules wereadded using BUSTER refinement [24]. A final round ofTable 2 Crystallographic statisticsStructure ABBV-323Fab aloneCD40 complexedto ABBV-323CD40 complexedto FAB516PDB code 6PE7 6PE8 6PE9Data CollectionResolution (Å) 132.6–1.74 126.9–2.84 167.5–3.13Space Group C2221P21212C2Unit Cell Lengths (a, b, c; Å) Angles (°) 63.7130.4132.6173.376.0126.1254.8224.0111.4β= 98.0Unique reflections 56956 40335 108254Overall Statistics (Highest Shell)Rmerge(%) 0.041 (0.94) 0.11 (0.974) 0.108 (0.632)I/σI25.5 (2.2) 15.8 (2.3) 11.2 (2.1)Data completeness (%) 100 (100) 100 (100) 99.4 (98.6)Mean multiplicity 6.6 (6.6) 6.6 (6.8) 3.4 (3.2)CC(1/2) 1.00 (0.79) 1.00 (0.72) 0.99 (0.76)RefinementResolution (Å) 24.4–1.74 38.0–2.84 35.3–3.13Reflections used in refinement 56860 40063 107879Rcryst(%) 19.6 20.5 21.6Rfree(%) 23.0 26.5 25.5R.m.s. deviations, bond lengths (Å), bond angles (°) 0.005, 0.865 0.010, 1.13 0.010, 1.13Ramachandran Favored regions (%) 97.7 94.9 93.6Outliers (%) 0.23 0.17 0.21Table 3 Protein quality characterizationProteinnameIsotype Monomer% by SECEndotoxinlevelaMassSpec.IdentitybHeavy chain Light chainFAB518 hCg1_z,non-a Kappa 100 0.1 EU/mg ConsistentFAB516 hCg1_z,non-a Kappa 100 0.1 EU/mg ConsistentCP 870,893 hCg1_z,non-a Kappa 99 0.1 EU/mg ConsistentABBV-323 hCg1_z,non-a Kappa 100 0.4 EU/mg ConsistentaEndotoxin levels assessed by PTS EndoSafe LALbExpected (theoretical) molecular weight (MW) is consistent with measured MW by Mass SpecArgiriadi et al. BMC Molecular and Cell Biology (2019) 20:29 Page 11 of 13Content courtesy of Springer Nature, terms of use apply. Rights reserved. refinement was conducted using PHENIX [26]. Finalrefinement statistics reported Rfree/Rworkvalues of0.230/0.196.The ABBV-323 Fab CD40 complex dataset was proc-essed in the space group P21212 with the following unitcell dimensions: a = 173.3 b = 76.0 c = 126.1. A maximumlikelihood molecular replacement solution was deter-mined using the program PHASER [20] using the previ-ously solved ABBV-323 Fab reported above. Coordinatesfor 2 Fab molecules were found in the asymmetric unitbased on the molecular replacement solution. Prelimin-ary refinement of the resulting solution was conductedusing REFMAC [22,23] and the program BUSTER [24].The model for CD40 was built manually using the pro-gram COOT [25] and examination of 2Fo-Fc and Fo-Fcelectron-density maps. Water molecules were addedusing BUSTER refinement [24]. A final round of refine-ment was conducted using PHENIX [26]. Final refine-ment statistics reported Rfree/Rworkvalues of 0.265/0.205.The FAB516 Fab CD40 complex dataset was processedin the space group C2 with the following unit cell di-mensions: a = 254.8 b = 224.0 c = 111.4 β= 98.0. A max-imum likelihood molecular replacement solution wasdetermined using the program PHASER [20] using thepreviously solved ABBV-323 complex structure reportedabove. Coordinates for 5 Fab/CD40 complexes werefound in the asymmetric unit based on the molecularreplacement solution. Preliminary refinement of theresulting solution was conducted using REFMAC [22,23]and the program BUSTER [24]. A partial model for CD40was built manually using the program COOT [25]andexamination of 2Fo-Fc and Fo-Fc electron-density maps.CRD3 was not seen in electron density for all 5 CD40monomers. Additionally, one Fab molecule (monomers Kand M) only had the variable regions fit due to disordereddensity in the constant region. A final round of refinementwas conducted using PHENIX [26].The refinement statis-tics reported Rfree/Rworkvalues of 0.255/0.216.Ramachandran plots and statistics were calculatedusing MolProbity in the Validation tools from PHENIX[26]. Plots and outliers are included in Additional file 2:Figures: S2a-c.The ternary complex model of FAB516/CD40L/CD40presented in Fig. 9a and b was created using the crystalstructures of FAB516/CD40 and 3QD6 [4]asguides.The complex was minimized in the program MAES-TRO [27] with Prime loop refinement [28]forHCDR2S52-G56.Additional filesAdditional file 1: Figures S1a, b and c: (a) Overlay of TNFR antiparalleldimer as observed in PDB 1NCF (in blue) and CD40-ABBV-323 antiparalleldimer (in green). (b) and (c) show the antiparallel orientations with N andC termini labeled. (PPTX 328 kb)Additional file 2: Figure S2a, b and c: Ramachadran plots and outliersfor all three structures were generated in the program PHENIX(Molprobity) [26]. (PPTX 558 kb)AbbreviationsCD40: Cluster of differentiation 40; CD40L: CD40 ligand;CDR: Complementarity-determining region; CRD: Cysteine-rich domain;FAB: Fragment antigen binding domain; HCDR: Heavy chain CDR; ICAM-1: Intercellular adhesion molecule 1; LCDR: Light chain CDR;mAb: Monoclonal antibody; NF-κB: Nuclear factor κappa light chainenhancer; SEAP: Secreted embryonic alkaline phosphatase; Tev: Tobacco EtchVirus nuclear–inclusion-1 endopeptidase; TNF: Tumor necrosis factor;TRAF: TNF receptor associated factorAcknowledgementsUse of the IMCA-CAT beamline 17-ID (or 17-BM) at the Advanced PhotonSource was supported by the companies of the Industrial MacromolecularCrystallography Association through a contract with Hauptman-WoodwardMedical Research Institute.Use of the Advanced Photon Source was supported by the U.S. Departmentof Energy, Office of Science, Office of Basic Energy Sciences, under ContractNo. DE-AC02-06CH11357.We would like to acknowledge former AbbVie employee, Junjian Liu, forhelpful discussions at the start of this research.Authors’contributionsMAA solved and analyzed all x-ray structures, formed structural hypothesesandwasprimaryauthoronthepaper.LB,ID,DAE,LG,AG,JH1,RBI,RAJ, ML, DCP, RS, BS, SS, and RW performed and interpreted laboratoryexperiments and delivered reagents or data for the study. MAA, ID, RAJ,JH2, and BLM discussed strategy and results of the study. MAA, ID, RJ,JH2, RBI, BS, and BLM helped write the manuscript. All authors are employees of AbbVie and have read and approved the manuscript. The design, study conduct, and financial support for this research were provided by AbbVie. AbbVie participated in the interpretation of data, review, and approval of the publication.FundingNot applicable.Availability of data and materialsTable 2lists data collection and refinement statistics from all structures listedabove. All structural figures were created in the program Pymol(Schrӧdinger) [29]. All structural coordinates have been deposited to theProtein Data Bank (http://www.rcsb.org) with the following PDB codes: 6PE7,6PE8, 6PE9. Table 3lists protein quality characterizations for the reagentsused in this study.Ethics approval and consent to participateNot applicable.Consent for publicationNot applicable.Competing interestsThe authors declare that they have no competing interests.Author details1AbbVie Bioresearch Center, 381 Plantation Street, Worcester, MA 01605, USA.2AbbVie Bioresearch Center, 100 Research Drive, Worcester, MA 01605, USA.3AbbVie Inc., 1 North Waukegan Road, North Chicago, IL 60064, USA.Received: 17 April 2019 Accepted: 17 July 2019References1. Bishop GA, Moore CR, Xie P, Stunz LL, Kraus ZJ. TRAF proteins in CD40signaling. Adv Exp Med Biol. 2007;597:131–51.Argiriadi et al. BMC Molecular and Cell Biology (2019) 20:29 Page 12 of 13Content courtesy of Springer Nature, terms of use apply. Rights reserved. 2. Hostager BS, Haxhinasto SA, Rowland SL, Bishop GA. Tumor necrosis factorreceptor-associated factor 2 (TRAF2)-deficient B lymphocytes reveal novelroles for TRAF2 in CD40 signaling. J Biol Chem. 2007;278:45382–90.3. Law CL, Grewal IS. Therapeutic interventions targeting CD40L (CD154) andCD40: the opportunities and challenges. Adv Exp Med Biol. 2009;647:8–36.4. An HJ, Kim YJ, Song DH, Park BS, Kim HM, Lee JD, Paik SG, Lee JO, Lee H.Crystallographic and mutational analysis of the CD40-CD154 complex andits implications for receptor activation. J Biol Chem. 2011;286:11226–35.5. Koyama I, Kawai T, Andrews D, Boskovic S, Nadazdin O, Wee SL, Sogawa H,Wu DL, Smith RN, Colvin RB, Sachs DH, Cosimi AB. Thrombophiliaassociated with anti-CD154 monoclonal antibody treatment and itsprophylaxis in nonhuman primates. Transplantation. 2004;77:460–2.6. Robles-Carrillo L, Meyer T, Hatfield M, Desai H, Dávila M, Langer F, Amaya M,Garber E, Francis JL, Hsu YM, Amirkhosravi A. Anti-CD40L immunecomplexes potently activate platelets in vitro and cause thrombosis inFCGR2A transgenic mice. J Immunol. 2010;185:1577–83.7. Adams AB, Shirasugi N, Jones TR, Durham MM, Strobert EA, Cowan S, ReesP, Hendrix R, Price K, Kenyon NS, Hagerty D, Townsend R, Hollenbaugh D,Pearson TC, Larsen CP. Development of a chimeric anti-CD40 monoclonalantibody that synergizes with LEA29Y to prolong islet allograft survival. J.Immunol. 2005;174:542–50.8. Hunter TB, Alsarraj M, Gladue RP, Bedian V, Antonia SJ. An agonist antibodyspecific for CD40 induces dendritic cell maturation and promotesautologous anti-tumour T-cell responses in an in vitro mixed autologoustumour cell/lymph node cell model. Scand J Immunol. 2007;65:479–86.9. Wines BD, Powell MS, Parren PW, Barnes N, Hogarth PM. The IgG fc containsdistinct fc receptor (FcR) binding site: the leukocyte receptors fc gamma RIand fc gamma RIIa bind to a region in the fc distinct from that recognizedby neonatal FcR and protein a. J Immunol. 2000;164:5313–8.10. Hinton PR, Xiong JM, Johlfs MG, Tang MT, Keller S, Tsurushita N. Anengineered human IgG1 antibody with longer serum half-life. J Immunol.2006;176:346–56.11. Gladue RP, Paradis T, Cole SH, Donovan C, Nelson R, Alpert R, Gardner J, NatoliE, Elliot E, Shepard R, Bedian V. The CD40 agonist antibody CP-870,893enhances dendritic cell and B-cell activity and promotes anti-tumor efficacy inSCID-hu mice. Cancer Immunol Immunother. 2011;60:1009–17.12. Li F, Ravetch JV. Inhibitory Fcγreceptor engage,ent drives adjuvant and anti-tumor activities of agonistic CD40 antibodies. Science. 2011;333:1030–40.13. White AL, Chan HTC, Roghanian A, French RR, Mockridge CI, Tutt AL, DixonSV, Ajona D, Verbeek JS, Al-Shamkhani A, Cragg MS, Beers SA, Glennie MJ.Interaction with FcγRIIB is critical for the agonist activity of anti-CD40monoclonal antibody. J Immunol. 2011;187:1754–63.14. Richman LP, Vonderheide RH. Role of crosslinking for agonistic CD40monoclonal antibodies as immune therapy of cancer. Cancer ImmunolRes. 2014;2:19–26.15. Smulski CR, Beyrath J, Decossas M, Chekkat N, Wolff P, Estieu-GionnetK, Guichard G, Speiser D, Schneider P, Fournel S. Cysteine-rich domain1 of CD40 mediates receptor self-assembly. J Biol Chem. 2013;288:10914–22.16. Naismith JH, Devine TQ, Brandhuber BJ, Sprang SR. Crystallographicevidence for dimerization of unliganded tumor necrosis factor receptor. JBiol Chem. 1995;270:13303–7.17. Vanamee ES, Faustman DL. Structural principles of tumor necrosis factorsuperfamily signaling. Sci Signal. 2018;11:1–11.18. Torrey H, Butterworth J, Mera T, Okubo Y, Wang L, Baum D, Defusco A,Plager S, Warden S, Huang D, Vanamee E, Foster R, Faustman DL. TargetingTNFR2 with antagonist antibodies inhibits proliferation of ovarian cancercells and tumor-associated Tregs. Sci Signal. 2017;10:1–12.19. Vonrhein C, Flensburg C, Keller P, Sharff A, Smart O, Paciorek W, Womack T,Bricogne G. Data processing and analysis with the autoPROC toolbox. ActaCryst D Biol Crystallogr. 2011;67:293–302.20. McCoy AJ, Grosse-Kunstleve RW, Adam PD, Winn MD, Storoni LC, Read RJ.Phaser crystallographic software. J Appl Crystallogr. 2007;40:658–74.21. Malia TJ, Obmolova G, Almagro JC, Gilliland GL, Teplyakov A. Crystalstructure of human germline antibody 3-23/B3. Mol Immunol. 2011;48:1586–8.22. Winn MD, Ballard CC, Cowtan KD, Dodson EJ, Emsley P, Evans PR, KeeganRM, Krissinel EB, Leslie AGW, McCoy A, McNicholas SJ, Murshudov GN,Pannu NS, Potterton EA, Powell HR, Read RJ, Vagin A, Wilson KS. Overviewof the CCP4 suite and current developments. Acta Cryst. D Biol. Crystallogr.2011;67:235–42.23. Murshodov GN, Vagin AA, Dodson EJ. Refinement of macromolecularstructures by the maximum-likelihood method. Acta Cryst D Biol Crystallogr.1997;53:240–55.24. Bricogne G, Blanc E, Brandl M, Flensburg C, Keller P, Paciorek W, Roversi P,Sharff A, Smart OS, Vonrhein C, Womack TO. BUSTER version 2.9.7.Cambridge, United Kingdom: Global phasing ltd; 2011.25. Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. ActaCryst D Biol Crystallogr. 2004;60:2126–32.26. Adams PD, Afonine PV, Bunkoczi G, Chem VB, Davis IW, Echols N, Headd JJ,Hung LW, Kapral GJ, Grosse-Kunstleve RW, McCoy AJ, Moriarty NW, OeffnerR, Read RJ, Richardson DC, Richardson JS, Terwilliger TC, Zwart PH. PHENIX:a comprehensive python-based system for macromolecular structuresolution. Acta Cryst D Biol Crystalllogr. 2010;66:213–21.27. Schrödinger Release 2018–3. Maestro, Schrödinger. New York, NY: LLC; 2019.28. Jacobson MP, Pincus DL, Rapp CS, Day TJF, Honig B, Shaw DE, FriesnerRA. A hierarchical approach to all-atom protein loop prediction.Proteins. 2004;55:351–67.29. The PyMOL Molecular Graphics System, Version 1.4 Schrödinger, LLC.Publisher’sNoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.Argiriadi et al. BMC Molecular and Cell Biology (2019) 20:29 Page 13 of 13Content courtesy of Springer Nature, terms of use apply. Rights reserved. 1.2.3.4.5.6.Terms and Conditions Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (\"Springer Nature”). Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (\"Users”), for small-scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. Byaccessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (\"Terms”). For thesepurposes, Springer Nature considers academic use (by researchers and students) to be non-commercial. These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personalsubscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used willapply. We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally withinResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will nototherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission asdetailed in the Privacy Policy. While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users maynot: use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent accesscontrol;use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or isotherwise unlawful;falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature inwriting;use bots or other automated methods to access the content or redirect messagesoverride any security feature or exclusionary protocol; orshare the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journalcontent. In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journalcontent cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or anyother, institutional repository. These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information orcontent on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Naturemay revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved. To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or impliedwith respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,including merchantability or fitness for any particular purpose. Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensedfrom third parties. If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner notexpressly permitted by these Terms, please contact Springer Nature at onlineservice@springernature.com