Generic design principles for antibody-based tumour necrosis factor (TNF) receptor 2 (TNFR2) agonists with FcγR-independent agonism

Background: Selective TNFR2 activation can be used to treat immune pathologies by activating and expanding regulatory T-cells (Tregs) but may also restore anti-tumour immunity by co-stimulating CD8+ T-cells. Oligomerized TNFR2-specific TNF mutants or anti-TNFR2 antibodies can activate TNFR2 but suffer either from poor production and pharmacokinetics or in the case of anti-TNFR2 antibodies typically from the need of FcγR binding to elicit maximal agonistic activity. Methods: To identify the major factor(s) determining FcγR-independent agonism of anti-TNFR2 antibodies, we systematically investigated a comprehensive panel of anti-TNFR2 antibodies and antibody-based constructs differing in the characteristics of their TNFR2 binding domains but also in the number and positioning of the latter. Results: We identified the domain architecture of the constructs as the pivotal factor enabling FcγR-independent, thus intrinsic TNFR2-agonism. Anti-TNFR2 antibody formats with either TNFR2 binding sites on opposing sites of the antibody scaffold or six or more TNFR2 binding sites in similar orientation regularly showed strong FcγR-independent agonism. The affinity of the TNFR2 binding domain and the epitope recognized in TNFR2, however, were found to be of only secondary importance for agonistic activity. Conclusion: Generic design principles enable the generation of highly active bona fide TNFR2 agonists from nearly any TNFR2-specific antibody.

Antibody-based soluble and membrane-bound TWEAK mimicking agonists with FcγR-independent activity

Fibroblast growth factor (FGF)-inducible 14 (Fn14) activates the classical and alternative NFκB (nuclear factor 'kappa-light-chain-enhancer' of activated B-cells) signaling pathway but also enhances tumor necrosis factor (TNF)-induced cell death. Fn14 expression is upregulated in non-hematopoietic cells during tissue injury and is also often highly expressed in solid cancers. In view of the latter, there were and are considerable preclinical efforts to target Fn14 for tumor therapy, either by exploiting Fn14 as a target for antibodies with cytotoxic activity (e.g. antibody-dependent cellular cytotoxicity (ADCC)-inducing IgG variants, antibody drug conjugates) or by blocking antibodies with the aim to interfere with protumoral Fn14 activities. Noteworthy, there are yet no attempts to target Fn14 with agonistic Fc effector function silenced antibodies to unleash the proinflammatory and cell death-enhancing activities of this receptor for tumor therapy. This is certainly not at least due to the fact that anti-Fn14 antibodies only act as effective agonists when they are presented bound to Fcγ receptors (FcγR). Thus, there are so far no antibodies that robustly and selectively engage Fn14 signaling without triggering unwanted FcγR-mediated activities. In this study, we investigated a panel of variants of the anti-Fn14 antibody 18D1 of different valencies and domain architectures with respect to their inherent FcγR-independent ability to trigger Fn14-associated signaling pathways. In contrast to conventional 18D1, the majority of 18D1 antibody variants with four or more Fn14 binding sites displayed a strong ability to trigger the alternative NFκB pathway and to enhance TNF-induced cell death and therefore resemble in their activity soluble (TNF)-like weak inducer of apoptosis (TWEAK), one form of the natural occurring ligand of Fn14. Noteworthy, activation of the classical NFκB pathway, which naturally is predominately triggered by membrane-bound TWEAK but not soluble TWEAK, was preferentially observed with a subset of constructs containing Fn14 binding sites at opposing sites of the IgG scaffold, e.g. IgG1-scFv fusion proteins. A superior ability of IgG1-scFv fusion proteins to trigger classical NFκB signaling was also observed with the anti-Fn14 antibody PDL192 suggesting that we identified generic structures for Fn14 antibody variants mimicking soluble and membrane-bound TWEAK.

Tumor Necrosis Factor Receptor 2 (TNFR2): An Emerging Target in Cancer Therapy

Despite the great success of TNF blockers in the treatment of autoimmune diseases and the identification of TNF as a factor that influences the development of tumors in many ways, the role of TNFR2 in tumor biology and its potential suitability as a therapeutic target in cancer therapy have long been underestimated. This has been fundamentally changed with the identification of TNFR2 as a regulatory T-cell (Treg)-stimulating factor and the general clinical breakthrough of immunotherapeutic approaches. However, considering TNFR2 as a sole immunosuppressive factor in the tumor microenvironment does not go far enough. TNFR2 can also co-stimulate CD8⁺ T-cells, sensitize some immune and tumor cells to the cytotoxic effects of TNFR1 and/or acts as an oncogene. In view of the wide range of cancer-associated TNFR2 activities, it is not surprising that both antagonists and agonists of TNFR2 are considered for tumor therapy and have indeed shown overwhelming anti-tumor activity in preclinical studies. Based on a brief summary of TNFR2 signaling and the immunoregulatory functions of TNFR2, we discuss here the main preclinical findings and insights gained with TNFR2 agonists and antagonists. In particular, we address the question of which TNFR2-associated molecular and cellular mechanisms underlie the observed anti-tumoral activities of TNFR2 agonists and antagonists.

CD40- and 41BB-specific antibody fusion proteins with PDL1 blockade-restricted agonism

Background: A strategy to broaden the applicability of checkpoint inhibitors is the combined use with antibodies targeting the immune stimulatory receptors CD40 and 41BB. However, the use of anti-CD40 and anti-41BB antibodies as agonists is problematic in two ways. First, anti-CD40 and anti-41BB antibodies need plasma membrane-associated presentation by FcγR binding to exert robust agonism but this obviously limits their immune stimulatory efficacy by triggering ADCC, CDC or anti-inflammatory FcγRIIb activities. Second, off tumor activation of CD40 and 41BB may cause dose limiting systemic inflammation.

Methods: To overcome the FcγR-dependency of anti-41BB and anti-CD40 antibodies, we genetically fused such antibodies with a PDL1-specific blocking scFv as anchoring domain to enable FcγR-independent plasma membrane-associated presentation of anti-CD40- and anti-41BB antibodies. By help of GpL-tagged variants of the resulting bispecific antibodies, binding to their molecular targets was evaluated by help of cellular binding studies. Membrane PDL1-restricted engagement of CD40 and 41BB but also inhibition of PDL1-induced PD1 activation were evaluated in coculture assays with PDL1-expressing tumor cell lines and 41BB, CD40 and PD1 responsible cell lines or T-cells.

Results: The binding properties of the bispecific antibody fusion proteins remained largely unchanged compared to their parental molecules. Upon anchoring to membrane PDL1, the bispecific antibody fusion proteins activated CD40/41BB signaling as efficient as the parental anti-CD40/anti-41BB antibodies when bound to FcγRs or cells expressing membrane-bound CD40L/41BBL. PD1 inhibition remained intact and the anti-41BB fusion protein thus showed PDL1-restricted costimulation of T-cells activated in vitro with anti-CD3 or a BiTe.

Conclusions: Targeting of anti-CD40 and anti-41BB fusion proteins to membrane PDL1 with a blocking PDL1 scFv links PD1-PDL1 checkpoint blockade intrinsically with engagement of CD40 or 41BB.

A systems-biology model of the tumor necrosis factor (TNF) interactions with TNF receptor 1 and 2

MOTIVATION

Clustering enables TNF receptors to stimulate intracellular signaling. The differential soluble ligand-induced clustering behavior of TNF receptor 1 (TNFR1) and TNFR2 was modeled. A structured, rule-based model implemented ligand-independent pre-ligand binding assembly domain (PLAD)-mediated homotypic low affinity interactions of unliganded and liganded TNF receptors.

RESULTS

Soluble TNF initiates TNFR1 signaling but not TNFR2 signaling despite receptor binding unless it is secondarily oligomerized. We consider high affinity binding of TNF to signaling-incompetent pre-assembled dimeric TNFR1 and TNFR2 molecules and secondary clustering of liganded dimers to signaling competent ligand-receptor clusters. Published receptor numbers, affinities and measured different activities of clustered receptors validated model simulations for a large range of receptor and ligand concentrations. Different PLAD-PLAD affinities and different activities of receptor clusters explain the observed differences in the TNF receptor stimulating activities of soluble TNF.

AVAILABILITY AND IMPLEMENTATION

All scripts and data are in manuscript and supplement at Bioinformatics online.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Membrane lymphotoxin-α2β is a novel tumor necrosis factor (TNF) receptor 2 (TNFR2) agonist

In the early 1990s, it has been described that LTα and LTβ form LTα₂β and LTαβ₂ heterotrimers, which bind to TNFR1 and LTβR, respectively. Afterwards, the LTαβ₂–LTβR system has been intensively studied while the LTα₂β–TNFR1 interaction has been ignored to date, presumably due to the fact that at the time of identification of the LTα₂β–TNFR1 interaction one knew already two ligands for TNFR1, namely TNF and LTα. Here, we show that LTα₂β interacts not only with TNFR1 but also with TNFR2. We furthermore demonstrate that membrane-bound LTα₂β (memLTα₂β), despite its asymmetric structure, stimulates TNFR1 and TNFR2 signaling. Not surprising in view of its ability to interact with TNFR2, LTα₂β is inhibited by Etanercept, which is approved for the treatment of rheumatoid arthritis and also inhibits TNF and LTα.

Receptor Oligomerization and Its Relevance for Signaling by Receptors of the Tumor Necrosis Factor Receptor Superfamily

With the exception of a few signaling incompetent decoy receptors, the receptors of the tumor necrosis factor receptor superfamily (TNFRSF) are signaling competent and engage in signaling pathways resulting in inflammation, proliferation, differentiation, and cell migration and also in cell death induction. TNFRSF receptors (TNFRs) become activated by ligands of the TNF superfamily (TNFSF). TNFSF ligands (TNFLs) occur as trimeric type II transmembrane proteins but often also as soluble ligand trimers released from the membrane-bound form by proteolysis. The signaling competent TNFRs are efficiently activated by the membrane-bound TNFLs. The latter recruit three TNFR molecules, but there is growing evidence that this is not sufficient to trigger all aspects of TNFR signaling; rather, the formed trimeric TNFL–TNFR complexes have to cluster secondarily in the cell-to-cell contact zone for full TNFR activation. With respect to their response to soluble ligand trimers, the signaling competent TNFRs can be subdivided into two groups. TNFRs of one group, designated as category I TNFRs, are robustly activated by soluble ligand trimers. The receptors of a second group (category II TNFRs), however, failed to become properly activated by soluble ligand trimers despite high affinity binding. The limited responsiveness of category II TNFRs to soluble TNFLs can be overcome by physical linkage of two or more soluble ligand trimers or, alternatively, by anchoring the soluble ligand molecules to the cell surface or extracellular matrix. This suggests that category II TNFRs have a limited ability to promote clustering of trimeric TNFL–TNFR complexes outside the context of cell–cell contacts. In this review, we will focus on three aspects on the relevance of receptor oligomerization for TNFR signaling: (i) the structural factors which promote clustering of free and liganded TNFRs, (ii) the signaling pathway specificity of the receptor oligomerization requirement, and (iii) the consequences for the design and development of TNFR agonists.

On-target restoration of a split T cell-engaging antibody for precision immunotherapy

T cell-engaging immunotherapies are changing the landscape of current cancer care. However, suitable target antigens are scarce, restricting these strategies to very few tumor types. Here, we report on a T cell-engaging antibody derivative that comes in two complementary halves and addresses antigen combinations instead of single molecules. Each half, now coined hemibody, contains an antigen-specific single-chain variable fragment (scFv) fused to either the variable light (V_L) or variable heavy (V_H) chain domain of an anti-CD3 antibody. When the two hemibodies simultaneously bind their respective antigens on a single cell, they align and reconstitute the original CD3-binding site to engage T cells. Employing preclinical models for aggressive leukemia and breast cancer, we show that by the combinatorial nature of this approach, T lymphocytes exclusively eliminate dual antigen-positive cells while sparing single positive bystanders. This allows for precision targeting of cancers not amenable to current immunotherapies.

The restriction of appropriate tumour-specific antigens is a current limitation for T cell-engaging immunotherapy. Here, the authors have designed a new system constituted by two halve antibodies, which engage T cells once binding to two different antigens, to specifically eliminate double positive cells in preclinical leukemia and breast cancer mouse models.