Evaluation of in vitro Assays to Assess the Modulation of Dendritic Cells Functions by Therapeutic Antibodies and Aggregates

Abstract

Therapeutic antibodies have the potential to induce immunogenicity leading to the development of anti-drug antibodies (ADA) that consequently may result in reduced serum drug concentrations, a loss of efficacy or potential hypersensitivity reactions. Among other factors, aggregated antibodies have been suggested to promote immunogenicity, thus enhancing ADA production. Dendritic cells (DC) are the most efficient antigen-presenting cell population and are crucial for the initiation of T cell responses and the subsequent generation of an adaptive immune response. This work focuses on the development of predictive in vitro assays that can monitor DC maturation, in order to determine whether drug products have direct DC stimulatory capabilities. To this end, four independent laboratories aligned a common protocol to differentiate human monocyte-derived DC (moDC) that were treated with either native or aggregated preparations of infliximab, natalizumab, adalimumab, or rituximab. These drug products were subjected to different forms of physical stress, heat and shear, resulting in aggregation and the formation of subvisible particles. Each partner developed and optimized assays to monitor diverse end-points of moDC maturation: measuring the upregulation of DC activation markers via flow cytometry, analyzing cytokine, and chemokine production via mRNA and protein quantification and identifying cell signaling pathways via quantification of protein phosphorylation. These study results indicated that infliximab, with the highest propensity to form aggregates when heat-stressed, induced a marked activation of moDC as measured by an increase in CD83 and CD86 surface expression, IL-1β, IL-6, IL-8, IL-12, TNFα, CCL3, and CCL4 transcript upregulation and release of respective proteins, and phosphorylation of the intracellular signaling proteins Syk, ERK1/2, and Akt. In contrast, natalizumab, which does not aggregate under these stress conditions, induced no DC activation in any assay system, whereas adalimumab or rituximab aggregates induced only slight parameter variation. Importantly, the data generated in the different assay systems by each partner site correlated and supported the use of these assays to monitor drug-intrinsic propensities to drive maturation of DC. This moDC assay is also a valuable tool as an in vitro model to assess the intracellular mechanisms that drive DC activation by aggregated therapeutic proteins.

Keywords: aggregates; anti-drug antibodies; dendritic cells; immunogenicity; in vitro assays; intracellular signaling.

Figure 1

Experimental setup. MoDC isolation and differentiation protocols were aligned for four international partner laboratories. Generated moDC cultures were treated with native or aggregated forms of Abs. MoDC maturation was measured by different in vitro assays.

Figure 2

Phenotypic assessment of maturation marker expression on moDC following treatment with native or aggregated infliximab (A) and natalizumab (B). Immature moDC were treated for 48 h with the positive control (MC; maturation cocktail as described in the Methods section), native or stressed (SSL2 or HSL2) therapeutic antibodies at 10 or 100 μg/mL. Cells were collected, washed, and analyzed by flow cytometry for CD80, CD83, and CD86 expression was measured on CD11c+ moDC. Results are expressed as the fold change of marker expression (either as the MFI or as the % of positive cells based on the sensitivity of responses to the positive control for each marker) compared to PBS treated cells (represented by the horizontal dotted line at 1.0). Results from 12–23 independent experiments are shown, individual points representing individual donor responses. The group's mean is represented by the horizontal gray line. *p < 0.05; **p < 0.01.

Figure 3

Expression of proinflammatory cytokine mRNA in moDC following treatment with native or aggregated infliximab (A) and natalizumab (B). Immature moDC were treated with native or stressed (SSL2 or HSL2) antibodies. Transcripts were measured after 6 h (for IL-1β, IL-6, IL-8, and TNFα) or 24 h (for IL-12p40) using real-time RT-PCR. Results are expressed as fold change over PBS control. The results from 12 independent experiments are shown. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Figure 4

Expression of chemokines CCL3 and CCL4 mRNA in moDC treated with native or aggregated infliximab (A) and natalizumab (B). Immature moDC were treated with native or stressed (SSL2 or HSL2) antibodies. Transcript expression was measured after 6 h using real-time RT-PCR. Results are expressed as fold change over PBS control. The results from 12 independent experiments are shown. ***p < 0.001.

Figure 5

Protein levels of proinflammatory cytokines in moDC following treatment with native or aggregated infliximab (A) and natalizumab (B). Immature moDC were treated for 48 h with native or stressed (SSL2 or HSL2) antibodies, or maturation cocktail (MC) or LPS. Cytokine concentrations were measured in culture supernatants using a MSD multiplex assay. Results are expressed as fold chage over PBS control. The results from 11 independent experiments are shown. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Figure 6

Secretion of chemokines CCL3 and CCL4 secretions by moDC treated with native or aggregated infliximab (A), and natalizumab (B). Immature moDC were treated for 48 h with native or stressed (SSL2 or HSL2) antibodies, maturation cocktail (MC) or LPS. Chemokine concentrations were measured in culture supernatants using a MSD multiplex assay. Results from seven independent experiments are shown. *p < 0.05; ***p < 0.001.

Figure 7

Protein phosphorylation in moDC treated with native or aggregated infliximab and natalizumab. (A) Syk phosphorylation, (B) ERK 1/2 phosphorylation, and (C) Akt phosphorylation. MoDC were treated with 100 μg/mL native or aggregated antibodies. Phosphorylation was detected following 15 and 30 min of stimulation. Results are expressed as fold change over PBS control. The results from 12 independent experiments are shown. *p < 0.05; **p < 0.01; ***p < 0.001.