Human IgG Fc-engineering for enhanced plasma half-life, mucosal distribution and killing of cancer cells and bacteria

Abstract

Monoclonal IgG antibodies constitute the fastest growing class of therapeutics. Thus, there is an intense interest to design more potent antibody formats, where long plasma half-life is a commercially competitive differentiator affecting dosing, frequency of administration and thereby potentially patient compliance. Here, we report on an Fc-engineered variant with three amino acid substitutions Q311R/M428E/N434W (REW), that enhances plasma half-life and mucosal distribution, as well as allows for needle-free delivery across respiratory epithelial barriers in human FcRn transgenic mice. In addition, the Fc-engineered variant improves on-target complement-mediated killing of cancer cells as well as both gram-positive and gram-negative bacteria. Hence, this versatile Fc technology should be broadly applicable in antibody design aiming for long-acting prophylactic or therapeutic interventions.

Fig. 1. Fc-engineered IgG1 with improved human FcRn binding and extended plasma half-life.

a The solved co-crystal structure of truncated recombinant human FcRn (green) in complex with IgG1 Fc (blue). The REW amino acid substitutions (Q311R/M428E/N434W) in the Fc are indicated. The Fc residues H435 and H310, required for pH dependent binding, as well as the FcRn residues E115, E116, and L135 are also shown. The N297-linked N-glycan structure attached to the IgG1 Fc is shown in red and the β2-microglobulin subunit of FcRn is shown in yellow. The figure was made in PyMOL using crystallographic data from PDB entry 4NOU. b Illustration showing the human FcRn binding ELISA setup. c, d ELISA showing binding of NIP IgG1 WT, REW, and H310A to human FcRn at pH 6.0 and 7.4, shown as mean±s.d of duplicates. e Illustration showing the human FcRn SPR kinetics assay. f, g SPR sensorgrams showing binding of monomeric human FcRn to immobilized anti-NIP IgG1 WT or REW at pH 6.0. h Illustration outlining the HERA cellular assay. HERA showing (i) uptake, (j) rescue from intracellular degradation (recycling), (k) residual amounts and (l) the derived HERA scores for the antibodies, shown as mean±s.d of triplicates from two independent experiments. m Illustration outlining the in vivo plasma half-life experiments. Plasma clearance of (n) anti-NIP and (o) anti-SARS-CoV-2 mAb4 IgG1-WT and REW (5 mg/kg) in mice pre-loaded with 500 mg/kg IVIg, shown as mean±s.d of percent antibody remaining in plasma over time (n = 5 animals per group). i–l Unpaired two-sided t-test, (n, o) RM Two-way ANOVA with Šídák’s multiple comparison test. Source data are provided as a Source Data file. b, e, h, m were created using BioRender.com.

Fig. 2. REW improves bioavailability, transmucosal delivery and vaccination.

a Illustration outlining the in vivo lung localization experiment in Tg32 mice. b Amount of NIP IgG1 WT and REW in BALF 23 days post i.v. administration of 5 mg/kg of the antibody variants in IVIg (500 mg/kg) pre-loaded mice. Shown as mean ± s.e.m (n = 4 animals (WT) and n = 5 animals (REW). c BALF/plasma ratio of NIP IgG1 WT and REW, 23 days post i.v. administration. Shown as mean ± s.d (n = 4 animals per group). d Illustration outlining i.n. delivery experiment in Tg32 mice. e Plasma concentration of NIP IgG1 WT and REW 24 h post i.n. delivery (2.23 mg/kg) in Tg32 mice. Shown as mean ± s.d. (n = 4 animals (WT) or n = 6 animals (REW)). f Illustration outlining FcRn mediated transcytosis experiments in a transwell system. g Apical to basolateral (A→B) and basolateral to apical (B→A) directed transport of NIP IgG1 WT, REW and IHH across polarized human T84 cells, shown as mean ± s.d (n = 8 independent monolayers from 2 independent experiments) (two data points was excluded from REW A→B and one data point excluded for REW (B→A) due to disrupted cell monolayers). h Apical to basolateral (A→B) and basolateral to apical (B→A) directed transport of NIP IgG1 WT, REW and IHH across MDCK-hFcRn cells, show as mean ± s.d (n = 6 independent cellular monolayers from 2 independent experiments) (one data point were excluded from REW A→B and REW (B→A) due to disrupted cell monolayers). i Illustration outlining mucosal vaccine and challenge experiment in Tg32 mice. j Percent survival of Tg32 mice i.n. vaccinated with HA WT and REW monovalent Fc fusions or NaCl control following challenge with a lethal dose of H1N1 virus. k Illustration showing the ex vivo human placental perfusion model used to measure maternal-to-fetal transport of anti-NIP IgG1 WT and REW. l Ex vivo human placental perfusion model showing maternal/fetal (FM) transport ratio of NIP IgG1 WT and REW, shown as mean ± s.d. (n = 4 placentas per group). b, c, e, g, h Unpaired two-tailed t-test, (l) Two-tailed Wilcoxon t-test. Source data are provided as a Source Data file. (a, d, f, I, k) were created using BioRender.com.

Fig. 3. REW potentiates on-target complement activation and killing of cancer cells.

a Structural model of hexameric IgG with a close-up of the Fc:Fc interface. The REW amino acid substitutions are shown. The model was made in PyMOL using crystallographic data from PDB entry 1HZH^,. b Illustration showing the human C1q binding ELISA setup. c ELISA binding of C1q from NHS to titrated amounts of antigen captured anti-NIP IgG1 WT, REW and PGLALA, shown as mean ± s.d of duplicates. d Illustration showing the C5bC9 deposition ELISA setup. e ELISA showing C5bC9 (TCC) formation from NHS as a function of titrated amounts of antigen captured anti-NIP IgG1 WT, REW and PGLALA, shown as mean±s.d of duplicates. f ELISA binding of C1q from NHS to titrated amounts of randomly immobilized anti-NIP IgG1 WT, REW, RGY and PGLALA, shown as mean ± s.d of duplicates. g Illustration outlining solution phase complement activation assays. h Amount of IgG complexes and (i) C4d in NHS incubated with 100 µg/mL of anti-NIP WT, REW, RGY and PGLALA or NHS only at 37 °C for 1 h, shown as mean ± s.d of duplicates. j Overview of anti-CD20 antibodies and CD20+ cell lines used in CDC assays and (k) Calcein-AM release assay used to measure antibody dependent CDC of cancer cells. l–n CDC activity of anti-CD20 mAb2 (low CDC), mAb1 (intermediate CDC), mAb9 (high CDC) WT and REW against WSU-NHL (low CD20), DOHH-2 (intermediate CD20) and SU-DHL-4 (high CD20) lymphoma target cells, shown as mean±s.d of duplicates. o CDC activity of recombinant forms of ofatumumab (anti-CD20, IgG1) WT, REW and P329A against Raji target cells, shown as mean ± s.d of duplicates. Source data are provided as a Source Data file. a, c, g were created using BioRender.com.

Fig. 4. REW enhances phagocytosis of gram-positive and killing of gram-negative bacteria.

a Illustration showing the ELISA setup used to measure C1q binding to anti-NIP IgG1 WT and REW in presence and absence of a molar excess of SpA. b ELISA showing binding of C1q to antigen captured anti-NIP IgG1 WT and REW in presence and absence of 5-fold molar excess of SpA, shown as mean±s.d of duplicates. c Illustration outlining S. aureus PMN phagocytosis experiments of anti-WTA IgG1 variants in presence of ΔIgG/IgM human serum. d PMN phagocytosis of S. aureus strain Newman-ΔspA/spi and (e) Newman-WT bound by titrated amounts of anti-WTA (clone 4497) IgG1 WT, REW or PGLALA in presence of IgG/IgM depleted human serum, shown as mean±s.d of triplicates. f PMN phagocytosis of S. pneumoniae serotype 6B (CSP6) in presence of ΔIgG/IgM human serum, shown as mean ± s.d of duplicates. g Illustration showing the N. gonorrhoeae CDC assay. h, i CDC activity against N. gonorrhoeae strain 15253 and strain FA1090 bound by titrated amounts of humanized anti-lipo-oligosaccharide IgG1 WT, REW or DAKA (Humab 2C7), shown as mean ± s.d of triplicates. d, e, h, i RM Two-way ANOVA with Šídák’s multiple comparison test. Source data are provided as a Source Data file. b, d, g, j, k were created using BioRender.com.

Fig. 5. Enhanced ADCC by N-glycan-engineering without compromising plasma half-life and IgG subclass specific effect of REW on C1q binding.

a Illustration of ELISA setup and binding of anti-NIP IgG1 WT, REW and low-fucose (2FF) WT and REW to (b) FcγRIIIa-V158 and (c) FcγRIIIa-F158, shown as mean ± s.d of duplicates. d Overview of anti-CD20 IgG1 mAb2 variants and B cell lines used to measure ADCC activity. e The ⁵¹Cr release assay used to measure ADCC activity of MNCs against target cell lines in presence of the antibodies. f–h MNC mediated ADCC of WSU-NHL (low CD20), Carnaval (low CD20) and SU-DHL-4 (high CD20) lymphoma cell lines by the WT mAb2 compared with the Fc-engineered versions; REW, REW-2FF and PGLALA, shown as mean±s.e.m. from 5 replicates performed in parallel. i Experimental outline and (j) plasma clearance of anti-NIP IgG1 WT and REW-2FF in IVIg pre-loaded (500 mg/kg) Tg32 mice, shown as mean±s.d of percent antibody remaining in plasma over time (n = 5 animals per group). k Illustration showing REW-engineered IgG subclasses and a recombinant form of Fc-fusion etanercept. l Cellular FcRn mediated rescue from intracellular degradation of WT and REW engineered IgG subclasses and Fc-fusion etanercept in HERA, shown as mean±s.d of triplicates from three independent experiments. m Plasma clearance of anti-NIP IgG2 WT and REW following i.v. administration (5 mg/kg) in IVIg pre-loaded (500 mg/kg) Tg32 mice, shown as mean ± s.d of percent antibody remaining in plasma over time (n = 4 animals per group). n Plasma clearance of recombinant etanercept WT and REW following i.v. administration (5 mg/kg) in Tg32 mice, shown as mean ± s.d of percent antibody remaining in plasma over time (n = 5 animals per group). o ELISA showing binding of human C1q to titrated amounts of antigen captured anti-NIP IgG2 WT and REW, shown as mean±s.d of duplicates. p CDC activity of mAb2 anti-CD20 IgG2 WT and REW against WSU-NHL (low CD20) using the ⁵¹Cr release assay, shown as mean±s.d of duplicates. q ELISA showing binding of human C1q to titrated amounts of antigen captured anti-NIP IgG4 WT and REW, shown as mean±s.d of duplicates. r, s ELISAs showing relative binding of FcyRIIIa-V158 or FcyRIIIa-F158 to antigen captured anti-NIP IgG2 and IgG4 WT and REW, shown as relative binding compared to anti-NIP IgG1-WT, mean ± s.d of duplicates. f, g, h, j, m, n RM Two-way ANOVA with Šídák’s multiple comparison test. l Unpaired two-tailed t-test. Source data are provided as a Source Data file. a, d, e, I, k were created using BioRender.com.