Engineered antibody fusion proteins for targeted disease therapy

Abstract

Since the FDA approval of the first therapeutic antibody 35 years ago, antibody-based products have gained prominence in the pharmaceutical market. Building on the early successes of monoclonal antibodies, more recent efforts have capitalized on the exquisite specificity and/or favorable pharmacokinetic properties of antibodies by developing fusion proteins that enable targeted delivery of therapeutic payloads which are otherwise ineffective when administered systemically. This review focuses on recent engineering and translational advances for therapeutics that genetically fuse antibodies to disease-relevant payloads, including cytokines, toxins, enzymes, neuroprotective agents, and soluble factor traps. With numerous antibody fusion proteins in the clinic and other innovative molecules poised to follow suit, these potent, multifunctional drug candidates promise to be a major player in the therapeutic development landscape for years to come.

Figure 1.. Recent payload immunocytokine formats.

Schematic representation of fusion proteins in which cytokines have been linked to antibodies or recombinant antibody fragments in various formats, including: (A) cytokine flanked C- and N-terminally by single-chain variable fragments (scFvs); (B) cytokine fused C- or N-terminally to a single-chain diabody; (C) cytokines fused C-terminally to a diabody; (D) trimeric cytokine fused C-terminally to scFvs; (E) trimeric dual cytokine fusion with distinct cytokines fused C- and N-terminally to scFvs; (F) cytokine fused C-terminally at the heavy chains of a full IgG; (G) cytokine fused C-terminally to a single-domain antibody (V_HH or nanobody); or (H) cytokine fused C-terminally at the heavy chains of an scFv-Fc. The targets of each payload immunocytokine, as defined by the antibody or antibody fragment specificity, are depicted for: (I) cancer and (J) chronic inflammation and autoimmunity. See Table 1 for the target and mechanisms of action for the fused cytokines. Antibodies and antibody fragments are shown in dark blue and amino acid linkers to the cytokine payload are indicated in gray. Cytokines are shown in lime yellow. Abbreviations: Clec9A, C-type lectin domain containing 9A; EDA(+)Fn, extra domain A of fibronectin; EDB(+)Fn, extra domain B of fibronectin; HER2, human epidermal growth factor receptor 2; PD-1, programmed cell death protein 1.

Figure 2.. Incorporation of cytokines into biased cytokine-antibody fusion proteins.

(A) Schematic of interleukin (IL)-2 signaling through an intermediate affinity receptor complex on effector cells or a high-affinity receptor complex on regulatory T cells (T_Regs). (B) An unmodified cytokine is fused to an anticytokine antibody to modulate cytokine/receptor subunit interactions through occlusion of the subunit/cytokine interface and/or by allosteric enhancement of subunit/cytokine interactions. Fusion proteins can be formed using N-terminal fusions (top) or intra-complementarity-determining region (CDR) fusions (bottom). (C) Mutated cytokines with biased signaling properties [IL-2v, IL-2 variant with abolished IL-2 receptor (IL-2R)α binding; RLI, IL-15 fused to the IL-15Rα minimal binding domain, which enhances the affinity of IL-15 for the receptor β and γ subunits; sumlL-2, IL-2 variant with decreased IL-2Rα binding affinity] can be C-terminally fused to the heavy chain of a full IgG (top) or N-terminally fused to Ffab^-Fc (bottom). (D) Masked cytokines leverage a conditional steric block of the cytokine/receptor interaction. Fc fusion proteins incorporating selectively cleavable receptor fragments have been employed where the fused mask (IL-2Rβ) is conditionally cleaved by tumor-associated matrix metalloproteinases (top). Recent work has generated more elaborate cleavable fusion layouts that target cytokines to a marker of interest (e.g., CD20) and conditionally release them in the presence of tumor-associated proteases (bottom). Antibodies and antibody fragments are shown in dark blue, amino acid linkers to the payload are indicated in gray, and cleavable linkers are shown in red. Wild-type cytokines are shown in lime yellow and mutated cytokines are shown in orange. Abbreviations: CEA, carcinoembryonic antigen; SHEE, serum half-life extension element.

Figure 3.. Recombinant immunotoxin (RIT), antibody-directed enzyme prodrug therapy (ADEPT), targeted enzyme replacement fusion protein, neuroprotective payload fusion protein, and soluble factor trap fusion protein formats.

Schematic representation of fusion proteins in which an antibody or antibody fragment [dsFv, scFv, dscFv, Fab, bisFv, or single-domain antibody (V_HH or nanobody)] is linked to a therapeutic payload, such as: (A) an N-terminal pore-forming toxin (PFT), with or without a caging (DHFR) moiety; (B) an N- or C-terminal ribosome-inactivating protein (RIP); (C) a C-terminal superantigen (SAg); (D) a C-terminal prodrug-activating enzyme, which forms a noncovalent homodimer; (E) a C-terminal lysosomal enzyme; (F) a C-terminal neuroprotective protein; or (G) a C-terminal extracellular domain of a soluble factor receptor. Antibodies and antibody fragments are shown in dark blue and amino acid linkers to the payload are indicated in gray, with the exception of cleavable linkers, which are shown in red. See Table 2 for corresponding fusion protein names and specific protein abbreviations.

Figure 4.. Mechanisms of action for recombinant immunotoxins (RITs), antibody (Ab)-directed enzyme prodrug therapies (ADEPTs), targeted enzyme replacement fusion proteins, neuroprotective payload fusion proteins, and soluble factor trap fusion proteins.

(A) Pore-forming toxin (PFT) (shown in lime yellow) RITs bind their target antigens on a furin-secreting cancer cell. Furin cleavage releases the hydrophilic ‘caging’ moiety for caged PFT RITs. PFTs interact with cancer-specific membrane lipids, resulting in oligomerization and pore formation, which precipitates cell death. (B) The antibody component of ribosome-inactivating protein (RIP) RITs binds the target antigen, triggering receptor-mediated endocytosis. Once in the endosome, furin cleavage releases the catalytic portion of the toxin (shown in lime yellow). Diphtheria toxin-based RIPs are released by furin cleavage alone, whereas Pseudomonas exotoxin A-based RIPs must be trafficked through, and released from, the endoplasmic reticulum (ER). Once in the cytosol, the RIP inactivates the ribosome complex, halting protein translation and triggering cell death. (C) The antibody component of superantigen (SAg) RITs binds the target antigen, while the SAg (shown in lime yellow) engages the T cell receptor (TOR) on cytotoxic T cells. SAg engagement activates the T cell, which then kills the target cell. (D) In ADEPTS, the Ab-enzyme fusion protein is administered first to allow binding to the target tissue. After clearance from healthy tissue, the nontoxic prodrug is administered. The prodrug is therefore converted to the active chemotherapeutic drug only at the targeted site. (E) In addition to facilitating trafficking across the blood–brain barrier, the antibody component of targeted enzyme replacement fusion proteins mediates endocytosis. The endosome containing this protein then fuses with a lysosome that stores accumulated glycosaminoglycans, which are normally degraded by the enzyme that is mutated in the relevant disease. The Ab-enzyme fusion protein replaces the activity of the mutated enzyme, restoring normal cell function. (F) The antibody in targeted soluble factor trap fusion proteins directs depletion of the relevant soluble factor to tissues that express the target antigen, thereby preventing this factor (usually a cytokine or growth factor) from acting on responder cells.