Targeting the myeloid checkpoint receptor SIRPα potentiates innate and adaptive immune responses to promote anti-tumor activity.

BACKGROUND: Signal regulatory protein α (SIRPα) is a myeloid-lineage inhibitory receptor that restricts innate immunity through engagement of its cell surface ligand CD47. Blockade of the CD47-SIRPα interaction synergizes with tumor-specific antibodies and T-cell checkpoint inhibitors by promoting myeloid-mediated antitumor functions leading to the induction of adaptive immunity. Inhibition of the CD47-SIRPα interaction has focused predominantly on targeting CD47, which is expressed ubiquitously and contributes to the accelerated blood clearance of anti-CD47 therapeutics. Targeting SIRPα, which is myeloid-restricted, may provide a differential pharmacokinetic, safety, and efficacy profile; however, SIRPα polymorphisms and lack of pan-allelic and species cross-reactive agents have limited the clinical translation of antibodies against SIRPα. Here, we report the development of humanized AB21 (hAB21), a pan-allelic anti-SIRPα antibody that binds human, cynomolgus monkey, and mouse SIRPα alleles with high affinity and blocks the interaction with CD47. METHODS: Human macrophages derived from donors with various SIRPα v1 and v2 allelic status were used to assess the ability of hAB21 to enhance phagocytosis. HAB21_IgG subclasses were evaluated for targeted depletion of peripheral blood mononuclear cells, phagocytosis and in vivo efficacy in xenograft models. Combination therapy with anti-PD1/anti-PD-L1 in several syngeneic models was performed. Immunophenotyping of tissues from MC38 tumor-bearing mice treated with AB21 and anti-PD-1 was evaluated. PK, PD and tolerability of hAB21 were evaluated in cynomolgus monkeys. RESULTS: SIRPα blockade with hAB21 promoted macrophage-mediated antibody-dependent phagocytosis of tumor cells in vitro and improved responses to rituximab in the Raji human tumor xenograft mouse model. Combined with PD-1/PD-L1 blockade, AB21 improved response rates by facilitating monocyte activation, dendritic cell activation, and T cell effector functions resulting in long term, durable antitumor immunity. In cynomolgus monkeys, hAB21 has a half-life of 5.3 days at 10 mg/kg and complete target occupancy with no hematological toxicity or adverse findings at doses up to 30 mg/kg. CONCLUSIONS: The in vitro and in vivo antitumor activity of hAB21 broadly recapitulates that of CD47 targeted therapies despite differences in ligand expression, binding partners, and function, validating the CD47-SIRPα axis as a fundamental myeloid checkpoint pathway and its blockade as promising therapeutic intervention for treatment of human malignancies.

Discovery of high affinity, pan-allelic, and pan-mammalian reactive antibodies against the myeloid checkpoint receptor SIRPα.

Targeting the CD47-signal-regulatory protein α (SIRPα) pathway represents a novel therapeutic approach to enhance anti-cancer immunity by promoting both innate and adaptive immune responses. Unlike CD47, which is expressed ubiquitously, SIRPα expression is mainly restricted to myeloid cells and neurons. Therefore, compared to CD47-targeted therapies, targeting SIRPα may result in differential safety and efficacy profiles, potentially enabling lower effective doses and improved pharmacokinetics and pharmacodynamics. The development of effective SIRPα antagonists is restricted by polymorphisms within the CD47-binding domain of SIRPα, necessitating pan-allele reactive anti-SIRPα antibodies for therapeutic intervention in diverse patient populations. We immunized wild-type and human antibody transgenic chickens with a multi-allele and multi-species SIRPα regimen in order to discover pan-allelic and pan-mammalian reactive anti-SIRPα antibodies suitable for clinical translation. A total of 200 antibodies were isolated and screened for SIRPα reactivity from which approximately 70 antibodies with diverse SIRPα binding profiles, sequence families, and epitopes were selected for further characterization. A subset of anti-SIRPα antibodies bound to both human SIRPα v1 and v2 alleles with high affinity ranging from low nanomolar to picomolar, potently antagonized the CD47/SIRPα interaction, and potentiated macrophage-mediated antibody-dependent cellular phagocytosis in vitro. X-ray crystal structures of five anti-SIRPα antigen-binding fragments, each with unique epitopes, in complex with SIRPα (PDB codes 6NMV, 6NMU, 6NMT, 6NMS, and 6NMR) are reported. Furthermore, some of the anti-SIRPα antibodies cross-react with cynomolgus SIRPα and various mouse SIRPα alleles (BALB/c, NOD, BL/6), which can facilitate preclinical to clinical development. These properties provide an attractive rationale to advance the development of these anti-SIRPα antibodies as a novel therapy for advanced malignancies. Abbreviations: ADCC: antibody-dependent cellular cytotoxicity; ADCP: antibody-dependent cellular phagocytosis; CFSE: carboxyfluorescein succinimidyl ester; Fab: fragment antigen binding; Fc: fragment crystallizable; FcγR: Fcγ receptor; Ig: immunoglobulin; IND: investigational new drug; MDM⊘: monocyte-derived macrophage; NOD: non-obese diabetic; scFv: single chain fragment variable; SCID: severe combined immunodeficiency; SIRP: signal-regulatory protein.

ALX148 blocks CD47 and enhances innate and adaptive antitumor immunity with a favorable safety profile.

CD47 is a widely expressed cell surface protein that functions as an immune checkpoint in cancer. When expressed by tumor cells, CD47 can bind SIRPα on myeloid cells, leading to suppression of tumor cell phagocytosis and other innate immune functions. CD47-SIRPα signaling has also been implicated in the suppression of adaptive antitumor responses, but the relevant cellular functions have yet to be elucidated. Therapeutic blockade of the CD47 pathway may stimulate antitumor immunity and improve cancer therapy. To this end, a novel CD47-blocking molecule, ALX148, was generated by fusing a modified SIRPα D1 domain to an inactive human IgG1 Fc. ALX148 binds CD47 from multiple species with high affinity, inhibits wild type SIRPα binding, and enhances phagocytosis of tumor cells by macrophages. ALX148 has no effect on normal human blood cells in vitro or on blood cell parameters in rodent and non-human primate studies. Across several murine tumor xenograft models, ALX148 enhanced the antitumor activity of different targeted antitumor antibodies. Additionally, ALX148 enhanced the antitumor activity of multiple immunotherapeutic antibodies in syngeneic tumor models. These studies revealed that CD47 blockade with ALX148 induces multiple responses that bridge innate and adaptive immunity. ALX148 stimulates antitumor properties of innate immune cells by promoting dendritic cell activation, macrophage phagocytosis, and a shift of tumor-associated macrophages toward an inflammatory phenotype. ALX148 also stimulated the antitumor properties of adaptive immune cells, causing increased T cell effector function, pro-inflammatory cytokine production, and a reduction in the number of suppressive cells within the tumor microenvironment. Taken together, these results show that ALX148 binds and blocks CD47 with high affinity, induces a broad antitumor immune response, and has a favorable safety profile.

Effect of Anti-C5a Therapy in a Murine Model of Early/Intermediate Dry Age-Related Macular Degeneration.

Purpose: A large body of evidence supports a central role for complement activation in the pathobiology of age-related macular degeneration (AMD), including plasma complement component 5a (C5a). Interestingly, C5a is a chemotactic agent for monocytes, a cell type also shown to contribute to AMD. However, the role monocytes play in the pathogenesis of "dry" AMD and the pharmacologic potential of targeting C5a to regulate these cells are unclear. We addressed these questions via C5a blockade in a unique model of early/intermediate dry AMD and large panel flow cytometry to immunophenotype monocytic involvement. Methods: Heterozygous complement factor H (Cfh+/-) mice aged to 90 weeks were fed a high-fat, cholesterol-enriched diet (Cfh+/-∼HFC) for 8 weeks and were given weekly intraperitoneal injections of 30 mg/kg anti-C5a (4C9, Pfizer). Flow cytometry, retinal pigmented epithelium (RPE) flat mounts, and electroretinograms were used to characterize anti-C5a treatment. Results: Aged Cfh+/- mice developed RPE damage, sub-RPE basal laminar deposits, and attenuation of visual function and immune cell recruitment to the choroid that was accompanied by expression of inflammatory and extracellular matrix remodeling genes following 8 weeks of HFC diet. Concomitant systemic administration of an anti-C5a antibody successfully inhibited local recruitment of mononuclear phagocytes to the choroid-RPE interface but did not ameliorate these AMD-like pathologies in this mouse model. Conclusions: These results show that immunotherapy targeting C5a is not sufficient to block the development of the AMD-like pathologies observed in Cfh+/-∼HFC mice and suggest that other complement components or molecules/mechanisms may be driving "early" and "intermediate" AMD pathologies.

Structural basis of C-terminal ß-amyloid peptide binding by the antibody ponezumab for the treatment of Alzheimer's disease.

Alzheimer's disease, the most common cause of dementia in the elderly and characterized by the deposition and accumulation of plaques, is composed in part of ß-amyloid (Aß) peptides, loss of neurons, and the accumulation of neurofibrillary tangles. Here, we describe ponezumab, a humanized monoclonal antibody, and show how it binds specifically to the carboxyl (C)-terminus of Aß40. Ponezumab can label Aß that is deposited in brain parenchyma found in sections from Alzheimer's disease casualties and in transgenic mouse models that overexpress Aß. Importantly, ponezumab does not label full-length, non-cleaved amyloid precursor protein on the cell surface. The C-terminal epitope of the soluble Aß present in the circulation appears to be available for ponezumab binding because systemic administration of ponezumab greatly elevates plasma Aß40 levels in a dose-dependent fashion after administration to a mouse model that overexpress human Aß. Administration of ponezumab to transgenic mice also led to a dose-dependent reduction in hippocampal amyloid load. To further explore the nature of ponezumab binding to Aß40, we determined the X-ray crystal structure of ponezumab in complex with Aß40 and found that the Aß40 carboxyl moiety makes extensive contacts with ponezumab. Furthermore, the structure-function analysis supported this critical requirement for carboxy group of AßV40 in the Aß-ponezumab interaction. These findings provide novel structural insights into the in vivo conformation of the C-terminus of Aß40 and the brain Aß-lowering efficacy that we observed following administration of ponezumab in transgenic mouse models.

Anti-amyloid therapy protects against retinal pigmented epithelium damage and vision loss in a model of age-related macular degeneration.

Age-related macular degeneration (AMD) is a leading cause of visual dysfunction worldwide. Amyloid ß (Aß) peptides, Aß1-40 (Aß40) and Aß1-42 (Aß42), have been implicated previously in the AMD disease process. Consistent with a pathogenic role for Aß, we show here that a mouse model of AMD that invokes multiple factors that are known to modify AMD risk (aged human apolipoprotein E 4 targeted replacement mice on a high-fat, cholesterol-enriched diet) presents with Aß-containing deposits basal to the retinal pigmented epithelium (RPE), histopathologic changes in the RPE, and a deficit in scotopic electroretinographic response, which is reflective of impaired visual function. Strikingly, these electroretinographic deficits are abrogated in a dose-dependent manner by systemic administration of an antibody targeting the C termini of Aß40 and Aß42. Concomitant reduction in the levels of Aß and activated complement components in sub-RPE deposits and structural preservation of the RPE are associated with anti-Aß40/42 antibody immunotherapy and visual protection. These observations are consistent with the reduction in amyloid plaques and improvement of cognitive function in mouse models of Alzheimer's disease treated with anti-Aß antibodies. They also implicate Aß in the pathogenesis of AMD and identify Aß as a viable therapeutic target for its treatment.

Genetic and molecular in vivo analysis of herpes simplex virus assembly in murine visual system neurons.

Herpes simplex virus (HSV) infects both epithelial cells and neuronal cells of the human host. Although HSV assembly has been studied extensively for cultured epithelial and neuronal cells, cultured neurons are biochemically, physiologically, and anatomically significantly different than mature neurons in vivo. Therefore, it is imperative that viral maturation and assembly be studied in vivo. To study viral assembly in vivo, we inoculated wild-type and replication-defective viruses into the posterior chamber of mouse eyes and followed infection in retinal ganglion cell bodies and axons. We used PCR techniques to detect viral DNA and RNA and electron microscopy immunohistochemistry and Western blotting to detect viral proteins in specific portions of the optic tract. This approach has shown that viral DNA replication is necessary for viral DNA movement into axons. Movement of viral DNA along ganglion cell axons occurs within capsid-like structures at the speed of fast axonal transport. These studies show that the combined use of intravitreal injections of replication-defective viruses and molecular probes allows the genetic analysis of essential viral replication and maturation processes in neurons in vivo. The studies also provide novel direct evidence for the axonal transport of viral DNA and support for the subassembly hypothesis of viral maturation in situ.

Temporal expression of herpes simplex virus type 1 mRNA in murine retina.

PURPOSE: During maturation of herpes simplex virus type 1 (HSV) in infected murine retinal F strain ganglion cells, new viral components are axonally transported in two phases. The viral envelope protein (gD) appears 48 hr before the capsid protein (VP5). Our hypothesis was that delayed appearance of VP5 mRNA in the infected eye causes the delayed expression of the VP5 protein in the axon. METHODS: HSV was injected into the ocular posterior chamber. Three to 24 hr later, the mice were euthanized, and the posterior eye was isolated. RNA was extracted, DNAase-treated, and used for amplification by reverse transcription-polymerase chain reaction (RT-PCR) using primers specific to gD, VP5 and a tegument protein VP22. RESULTS: VP22 and gD mRNAs are expressed 6 hr and VP5 mRNA is first detected 9 hr after infection. CONCLUSIONS: The results establish that delayed transcription does not play a significant role in the 48-hr delay in VP5 appearance in the retinal axons.

Axonal transport and sorting of herpes simplex virus components in a mature mouse visual system.

The time course for delivery and transport of two major proteins of herpes simplex virus (HSV) has been determined for mature mouse retinal ganglion cell axons in vivo. Twenty-four hours after intravitreal injection of HSV, valacyclovir was introduced into the drinking water of the mice to inhibit subsequent viral replication. Without treatment, viral spread and replication in periaxonal glial cells confound study of axonal transport. At 2 to 5 days after infection, the animals were sacrificed and contiguous segments of the optic pathway were removed. Immunofluorescence microscopy indicated that the number of infected astrocytes was reduced in the proximal optic nerve and eliminated in the optic tract. Western blots of the retina with antibodies for envelope and capsid components, glycoprotein D (gD) and VP5, respectively, revealed that both components were expressed in retinal homogenates by 2 days. Results of reverse transcription-PCR indicated that there was no gD mRNA present in the treated optic tract 5 days after infection. Therefore, we conclude that gD is transcribed from viral mRNA in the retinal ganglion cell bodies. The gD accumulated in the proximal ganglion cell axon by 2 days and reached the most distal segment after 3 days. The VP5 first appeared in the proximal axons at 4 days, about 48 h after the appearance of gD. Thus, gD entered the axon earlier and independent of VP5. These finding confirm the subassembly model of viral transport in neurons and suggest that there is a 4- to 5-day window for initiation of effective antiviral treatment with valacyclovir.