Fcγ receptor binding is required for maximal immunostimulation by CD70-Fc

Abstract

Introduction: T cell expressed CD27 provides costimulation upon binding to inducible membrane expressed trimeric CD70 and is required for protective CD8 T cell responses. CD27 agonists could therefore be used to bolster cellular vaccines and anti-tumour immune responses. To date, clinical development of CD27 agonists has focussed on anti-CD27 antibodies with little attention given to alternative approaches.

Methods: Here, we describe the generation and activity of soluble variants of CD70 that form either trimeric (t) or dimer-of-trimer proteins and conduct side-by-side comparisons with an agonist anti-CD27 antibody. To generate a dimer-of-trimer protein (dt), we fused three extracellular domains of CD70 to the Fc domain of mouse IgG1 in a 'string of beads' configuration (dtCD70-Fc).

Results: Whereas tCD70 failed to costimulate CD8 T cells, both dtCD70-Fc and an agonist anti-CD27 antibody were capable of enhancing T cell proliferation in vitro. Initial studies demonstrated that dtCD70-Fc was less efficacious than anti-CD27 in boosting a CD8 T cell vaccine response in vivo, concomitant with rapid clearance of dtCD70-Fc from the circulation. The accelerated plasma clearance of dtCD70-Fc was not due to the lack of neonatal Fc receptor binding but was dependent on the large population of oligomannose type glycosylation. Enzymatic treatment to reduce the oligomannose-type glycans in dtCD70-Fc improved its half-life and significantly enhanced its T cell stimulatory activity in vivo surpassing that of anti-CD27 antibody. We also show that whereas the ability of the anti-CD27 to boost a vaccine response was abolished in Fc gamma receptor (FcγR)-deficient mice, dtCD70-Fc remained active. By comparing the activity of dtCD70-Fc with a variant (dtCD70-Fc(D265A)) that lacks binding to FcγRs, we unexpectedly found that FcγR binding to dtCD70-Fc was required for maximal boosting of a CD8 T cell response in vivo. Interestingly, both dtCD70-Fc and dtCD70-Fc(D265A) were effective in prolonging the survival of mice harbouring BCL1 B cell lymphoma, demonstrating that a substantial part of the stimulatory activity of dtCD70-Fc in this setting is retained in the absence of FcγR interaction.

Discussion: These data reveal that TNFRSF ligands can be generated with a tunable activity profile and suggest that this class of immune agonists could have broad applications in immunotherapy.

Keywords: CD27; T cells; TNFRSF; cancer; costimulation; immunotherapy; vaccine.

Figure 1

Structure, receptor binding profile, and in vitro T cell costimulatory effects of CD70 fusion proteins. (A) Schematic representation of tCD70 and dtCD70-Fc fusion proteins. (B) Purified tCD70 and dtCD70-Fc proteins (5 µg) were analysed using a 10% SDS-polyacrylamide gel under non-reducing (NR) or reducing (R) conditions. The gel was stained with Coomassie blue. (C) Overlay of SPR sensograms demonstrating binding of CD70 fusion proteins (1.56, 6.25, 25 and 100 nM) to captured recombinant mouse CD27 and their subsequent dissociation. (D) Splenocytes were stimulated for 72 h with various concentrations of soluble anti-CD3 and the indicated proteins (10 μg/ml). Proliferation of T cells as assessed by measurement of [³H]-thymidine incorporation. Data points are the mean of triplicate measurements +/- SE and the data are representative of two independent experiments. Statistical comparisons at the highest anti-CD3 concentration are indicated. **** P < 0.0001, two-way ANOVA with Tukey’s multiple comparison test.

Figure 2

Effects of anti-CD27 mAb and dtCD70-Fc on OT-I T cell expansion in vivo. OT-I TCR transgenic T cells were adoptively transferred into C57BL/6 recipients. Mice were then immunised by i.v. injection of OVA_257-264 in combination with control mouse IgG1 (mIgG1), dtCD70-Fc or anti-CD27. The next day mice received an additional dose of mIgG1, dtCD70-Fc or anti-CD27. Antigen specific CD8⁺ T cells in peripheral blood were enumerated at the indicated time points by staining with anti-CD8α and anti-CD45.1 (A, C, D) or anti-CD8α and OVA_257-264 tetramer (B). (A) Representative dot plots showing the percentage of OVA specific CD8⁺ T cells out of lymphocytes at the peak of the response (day 5). (B–D) Expansion of OVA specific CD8⁺ T cells after adoptive transfer of different numbers of OT-I T cells plotted as percentage out of lymphocytes. Data points represent the mean +/- SE (n = 3 mice/group). * P < 0.05, **** P < 0.0001, two-way ANOVA with Tukey’s multiple comparison test.

Figure 3

Oligomannose-type glycans in dtCD70-Fc contribute to its short half-life in vivo. (A) The concentrations of dtCD70-Fc and anti-CD27 mAb were determined in serum samples by ELISA at the indicated intervals following i.v. injection of proteins (250 μg). (B) Overlay of SPR sensograms demonstrating binding of FcRn at different concentrations (0, 0.8, 4, 20, 100, 500 nM) to dtCD70-Fc or antiCD27 immobilized directly onto a CM5 sensor chip. (C) Site-specific glycan analysis of dtCD70-Fc with and without Endo H treatment and anti-CD27 mAb. Bar graphs represent the average relative abundance of glycans detected across all sites on the molecule. Any composition containing HexNAc(2)Hex(>3) was classified as oligomannose-type, those containing at least one fucose and/or sialic acid were classified as Fucose or NeuAc respectively. Any composition containing Hex(3) was classified as “Hex(3), no galactose”. GlcNAc(1)/GlcNAc(1)Fuc(1) is included as a separate category to highlight the remnant saccharides resulting from Endo H treatment. (D) Analysis of oligomannose digestion by SDS-PAGE. Untreated dtCD70-Fc or an aliquot of the Endo H reaction (~ 5 μg protein) was run on a 10% SDS-polyacrylamide gel under reducing conditions. Proteins were revealed by Coomassie blue staining. (E) The concentrations of untreated or Endo H treated dtCD70-Fc were determined in serum samples by ELISA at the indicated intervals after i.v. injection of proteins (250 μg). Data points represent the mean +/- SE (n = 3 mice/group) and are representative of two independent experiments. *** P < 0.001, unpaired two-tailed t test.

Figure 4

Glycan trimming and FcγR binding potentiate the immunostimulatory activity of dtCD70-Fc in vivo. (A) Purified OT-I CD45.1⁺ congenic CD8⁺ T cells (1 x 10⁴) were adoptively transferred into C57BL/6 recipients. Mice were then immunised with OVA_257-264 in combination with control mIgG1, dtCD70-Fc, Endo H treated dtCD70-Fc or anti-CD27. The next day mice received an additional dose of mIgG1, dtCD70-Fc, Endo H treated dtCD70-Fc or anti-CD27. Antigen specific CD8⁺ T cells in peripheral blood were enumerated at the indicated time points and data are presented as percentage OVA-specific CD8⁺ T cells out of total CD8⁺ T cells. (B) In vivo agonistic activity of Endo H treated dtCD70-Fc in FcγR null mice. Adoptive transfer of OT-I T cells and immunisation was carried out as in (A) except that the recipient mice were FcγR null. (C) Comparison of the agonistic activity of Endo H treated dtCD70-Fc and dtCD70-Fc_(D265A) proteins. Purified OT-I T cells were adoptively transferred into C57BL/6 recipients and mice were immunised as indicated in (A). Data points represent the mean +/- SE (n = 3 mice/group) and are representative of two independent experiments. ** P < 0.01, **** P < 0.0001, two-way ANOVA with Tukey’s multiple comparison test.

Figure 5

Therapeutic activity of dtCD70-Fc against BCL₁ lymphoma. Groups of mice received 5 x 10⁶ BCL₁ cells i.v. on day 0 and then mIgG1 control, dtCD70-Fc, Endo H treated dtCD70-Fc, Endo H treated CD70-Fc_(D265A) or anti-CD27 on days 5, 6, 7 and 8 (200 μg/d). Mice were monitored for tumour development and survival to the humane end point was plotted using the Kaplan-Meier method. *** P < 0.001, **** P < 0.0001, log-rank (Mantle-Cox) test (n = 5 - 10 mice/group).