Characterisation of novel human CD27-specific monoclonal antibodies for cancer therapy

Abstract

Monoclonal antibodies (mAbs) are employed in the frontline treatment of cancer. These mAbs are known to work through a variety of mechanisms, for example by directly targeting tumour cells to mediate their clearance, or to augment immune responses against cancers. The majority of agonistic mAbs target members of the tumour necrosis factor receptor superfamily (TNFRSF), which provide T-cell costimulatory signalling. Cluster of differentiation (CD) 27 is a member of the TNFRSF and pre-clinical studies investigating mAbs targeting CD27 have demonstrated substantial anti-tumour activity in murine tumour models. In particular, durable cures were observed when anti-CD27 mAb was combined with a direct tumour targeting anti-CD20 mAb. The stimulation of CD27 on T cells promoted myeloid cell activation, leading to enhanced antibody-dependent cellular phagocytosis (ADCP), mediated by the anti-CD20 mAb, which contributed to tumour clearance. Similarly, anti-human CD27 (hCD27) mAb, varlilumab, induced in vitro activation of peripheral blood monocytes, indicated by CD14 downregulation. Despite these promising data, the therapeutic efficacy of anti-hCD27 mAbs in the clinic has been less impressive, potentially due to a lack of understanding of the ideal mAb format. Identifying the key determinants that govern hCD27 agonism might help to achieve better therapeutic outcomes.
Four novel and two clinically relevant anti-hCD27 mAbs were characterised with regards to epitope specificity, and the influence of both epitope and isotype on agonism. A range of affinities and avidities via cell-surface binding and surface plasmon resonance (SPR) were identified. Alanine scanning mutagenesis showed that the mAbs bound across the three cysteine-rich domains (CRDs) of the receptor with membrane-distally binding mAbs being more agonistic than membrane-proximal binding ones in an in vitro reporter cell assay. Further, mAbs binding to the external surface of a hCD27 homodimer were stronger agonists than mAbs binding internally-facing residues. When the agonistic activity of anti-hCD27 mAbs of different isotypes was compared, human immunoglobulin 2 (hIgG2, h2) mAb outperformed hIgG1 (h1) counterparts by inducing stronger transcriptional activation and increased CD8+ T-cell proliferation. Increased agonism was also observed by inhibitory Fcγ receptor (FcγR) IIb-mediated hCD27 cross-linking, through co-culture with FcγRIIb-expressing cells or the use of Fc-engineered h1 mAbs with enhanced affinity to FcγRIIb.
The therapeutic efficacy of hCD27.15 (the best agonist in vitro) in combination with the clinically relevant direct targeting mAb (daratumumab, anti-hCD38) was investigated in vivo. However, no survival benefit was observed in hCD27 transgenic hCD38 A20-tumour-bearing mice. Treatment of Daudi lymphoma bearing, human PBMC-injected mice with hCD27.15 monotherapy resulted in some slowing of tumour growth.
Finally, the bystander effects of anti-hCD27 mAbs in PBMCs were explored. Only stimulation of PBMCs with varlilumab (but no other anti-hCD27 mAb) resulted in indirect effects on monocytes, causing downregulation of CD14 and upregulation of CD16 and CD86, suggesting differentiation of monocytes. The CD14 downregulation might partially be mediated by hCD27- or T-cell receptor signalling. Further work into the underlying mechanism is required.
Together this work highlights the importance of mAb-engineering and the selection of the optimal mAb format in order to generate more clinically efficacious anti-CD27 mAb agonists with better therapeutic outcomes.

More information

Identifiers

ORCID for Franziska Heckel: orcid.org/0000-0002-7564-1932

Catalogue record

Export record