ReviewCancer immunotherapy targeting the CD47/SIRPα axis
Introduction
The field of immuno-oncology has rapidly translated hypothetical concepts into clinical strategies, yielding a new era of cancer investigation and treatment. A large emphasis has been placed on therapies that stimulate the adaptive immune system to attack cancer, in particular T cells. This is rightfully due to the success of immune checkpoint inhibitors targeting CTLA-4 and the PD-1/PD-L1 axis, which disable inhibitory signals to T cells and generate anti-tumour responses. However, both adaptive and innate immune cells are endowed with specialised functions to eliminate pathogens, and many of these functions can be redirected against tumours. Cells of the myeloid lineage are the most abundant immune cells in the body, yet few immunotherapeutic approaches have been aimed at stimulating them to attack cancer. Macrophages in particular have remarkable potential as mediators of anti-cancer therapies based on their robust ability to perform phagocytosis. However, macrophages have a complex relationship with tumours, and in many cases they may promote tumour growth or metastasis. The CD47/signal regulatory protein alpha (SIRPα) axis is a critical molecular interaction that inhibits the activation of macrophages and other myeloid cells against tumours and thereby acts as a myeloid-specific immune checkpoint (Fig. 1). Therapies targeting the CD47/SIRPα axis have demonstrated success in a wide range of preclinical models and are now under investigation in clinical trials for both solid and haematologic malignancies. The CD47/SIRPα axis has emerged as one of the most promising new targets for immuno-oncology.
Section snippets
Discovery of CD47 as a ‘marker of self’
CD47 is a 50 kDa multipass transmembrane protein with an extracellular region composed of a single immunoglobulin superfamily (IgSF) domain. Early studies described CD47 as a molecule expressed by a variety of ovarian cancers [1], a cell-surface protein that interacted with integrins on haematopoietic cells [2] and a glycoprotein on the surface of red blood cells [3]. A unifying theme to these studies and subsequent investigations was widespread expression of CD47 on both normal and malignant
Molecular characterisation of the CD47/SIRPα interaction
Extensive biophysical characterisation of the CD47/SIRPα interaction has been performed, including crystallographic analysis of the extracellular domain of SIRPα alone an in complex with CD47 [10], [11]. These studies have demonstrated that the distal, N-terminal domain of SIRPα is responsible for contacting CD47 [11], [12]. SIRPα was first characterised as a receptor tyrosine kinase that associates with the inhibitory phosphatases SHP-1 and SHP-2 [13]. SIRPα has three extracellular IgSF
The CD47/SIRPα axis in cancer
Expanding on knowledge of CD47 in bone marrow engraftment, the functional role of CD47 in protecting haematopoietic stem cells (HSCs) was further examined. HSCs intermittently enter into a migratory phase during which they exit the bone marrow and circulate systemically through the blood stream [23]. These migratory HSCs upregulate CD47 on their surface to avoid removal by macrophages of the reticuloendothelial system [23]. A similar mechanism is employed by acute myelogenous leukaemia (AML)
Mechanisms of action
Throughout these studies, the predominant mechanism of action investigated has been macrophage activation. In vitro, CD47 blockade stimulates macrophage phagocytosis of cancer cells, a finding that has been extensively validated by microscopy and flow cytometry [24], [25], [28], [31], [33]. In vivo, treatment of mice with liposomal clodronate—which depletes macrophages—abrogates the tumouricidal effects of CD47-blocking therapies, indicating macrophage are required for robust anti-tumour
Approaches to targeting the CD47/SIRPα axis
A variety of therapeutics targeting the CD47/SIRPα axis are under preclinical and clinical investigation, including conventional antibodies, recombinant polypeptides, and bispecific molecules (Table 1). For each of these agents, there exist benefits and tradeoffs for successful translation to the clinic. Special considerations pertain to each agent's efficacy, toxicity, and pharmacokinetic and pharmacodynamic properties.
To date, anti-CD47 antibodies are the best characterised therapies
Combining CD47/SIRPα blockade with tumour-opsonizing antibodies
Although CD47-blocking antibodies are sufficient to induce phagocytosis as single agents, macrophage phagocytosis is dependent on the integration of pro-phagocytic (‘eat me’) and anti-phagocytic (‘don't eat me’) signals. CD47 acts as a predominant inhibitory signal that prevents macrophage phagocytosis of cancer cells. A number of pro-phagocytic signals have also been described, including intrinsic molecules that contribute to immunological cell death such as phosphatidylserine and calreticulin
Considerations for clinical development of CD47/SIRPα-blocking therapies
As described above, one of the greatest distinctions between agents targeting the CD47/SIRPα interaction is whether they act as a ‘monotherapy’ against cancer or an ‘adjuvant’ to tumour-binding antibodies. Any agent that contains an Fc region that engages activating Fc receptors on macrophages has the potential to act as a monotherapy for cancer. These therapies include anti-CD47 antibodies and SIRPα-Fc fusion proteins. As a class, they pose a greater risk for on-target toxicity to normal cells
The CD47/SIRPα axis and the adaptive immune system
Beyond activation of innate immune cells, mounting evidence suggests CD47 blockade can also initiate or augment adaptive immune responses. In a mouse ovalbumin (OVA) tumour model, treatment with anti-CD47 antibodies resulted in greater phagocytosis and antigen presentation [34]. Transfer of macrophages that had engulfed tumour cells into immunocompetent mice was able to protect the mice from a subsequent challenge with B16 cells expressing OVA [34]. These results suggested CD47 blockade could
Additional applications of CD47/SIRPα-blocking therapies
The principles of targeting the CD47/SIRPα axis can also be applied to other non-malignant states of disease. In theory, diseased cells may be more susceptible to macrophage phagocytosis, and any therapeutic antibody that depletes cells by engaging immune effector functions could be enhanced by combination with CD47-blocking agents. Possible combinations include those with antibodies targeting persistently infected cells, such as broadly neutralising HIV antibodies [69], [70]. Therapeutic
Conclusions
Myeloid cells hold much promise as effectors of cancer immunotherapy, and therapeutic targeting of the CD47/SIRPα axis may unlock their potential. Preclinical studies indicate CD47/SIRPα-blocking therapies are effective against a broad range of cancers, and ongoing clinical trials will determine their efficacy as single agents and as combination therapies. The application of these therapeutics extends beyond human cancer, and additional preclinical and clinical studies will reveal their benefit
Conflict of interest statement
K.W. is an inventor of US patent applications pertaining to CD47-blocking assigned to Stanford University. K.W. declares consulting and/or equity ownership in Alexo Therapeutics, Inc. and Forty Seven Inc.
Acknowledgements
The author thanks I.L. Weissman, members of the Weissman laboratory, E. Weiskopf and C. Weiskopf for support and helpful discussions. The author declares no funding sources or writing assistance for this manuscript.
References (76)
- et al.
Platelet homeostasis is regulated by platelet expression of CD47 under normal conditions and in passive immune thrombocytopenia
Blood
(2005) - et al.
Attenuation of phagocytosis of xenogeneic cells by manipulating CD47
Blood
(2007) - et al.
Polymorphic Sirpa is the genetic determinant for NOD-based mouse lines to achieve efficient human cell engraftment
Blood
(2013) - et al.
The structure of the macrophage signal regulatory protein alpha (SIRPalpha) inhibitory receptor reveals a binding face reminiscent of that used by T cell receptors
J Biol Chem
(2007) - et al.
Paired receptor specificity explained by structures of signal regulatory proteins alone and complexed with CD47
Mol Cell
(2008) - et al.
Functional elements on SIRPalpha IgV domain mediate cell surface binding to CD47
J Mol Biol
(2007) - et al.
Integrin-associated protein is a ligand for the P84 neural adhesion molecule
J Biol Chem
(1999) - et al.
Human signal-regulatory protein is expressed on normal, but not on subsets of leukemic myeloid cells and mediates cellular adhesion involving its counterreceptor CD47
Blood
(1999) - et al.
“Velcro” engineering of high affinity CD47 ectodomain as signal regulatory protein alpha (SIRPalpha) antagonists that enhance antibody-dependent cellular phagocytosis
J Biol Chem
(2015) - et al.
Signal regulatory proteins negatively regulate immunoreceptor-dependent cell activation
J Biol Chem
(1999)