Microgels as Platforms for Antibody-Mediated Cytokine Scavenging

Abstract

Therapeutic antibodies are the key treatment option for various cytokine-mediated diseases, such as rheumatoid arthritis, psoriasis, and inflammatory bowel disease. However, systemic injection of these antibodies can cause side effects and suppress the immune system. Moreover, clearance of therapeutic antibodies from the blood is limiting their efficacy. Here, water-swollen microgels are produced with a size of 25 µm using droplet-based microfluidics. The microgels are functionalized with TNFα antibodies to locally scavenge the pro-inflammatory cytokine TNFα. Homogeneous distribution of TNFα-antibodies is shown throughout the microgel network and demonstrates specific antibody-antigen binding using confocal microscopy and FLIM-FRET measurements. Due to the large internal accessibility of the microgel network, its capacity to bind TNFα is extremely high. At a TNFα concentration of 2.5 µg mL^-1 , the microgels are able to scavenge 88% of the cytokine. Cell culture experiments reveal the therapeutic potential of these microgels by protecting HT29 colorectal adenocarcinoma cells from TNFα toxicity and resulting in a significant reduction of COX II and IL8 production of the cells. When the microgels are incubated with stimulated human macrophages, to mimic the in vivo situation of inflammatory bowel disease, the microgels scavenge almost all TNFα that is produced by the cells.

Keywords: TNFα scavenging; antibody; autoimmune disease; inflammation; local therapy; microgels.

Figure 1

A) Composition of continuous and aqueous dispersed phases used in microfluidics. Continuous phase consists of Novec HFE 7500 (a fluorinated ether) with 1.5 vol% of Krytox H as a stabilizing agent, while glycidyl methacrylate (GMA) and 2‐Hydroxy‐4′‐(2‐hydroxyethoxy)‐2‐methylpropiophenone (HEMP, also known as Irgacure 2959) is added to the aqueous phase. B) Microfluidic set‐up for the production of spherical microgels with different sPEG‐Ac polymers at a total concentration of 10 wt%. C) Schematic representation of TNFα scavenging microgels:glycidyl‐co‐polymerized PEG microgels are post‐functionalization with TNFα antibodies to scavenge TNFα. D) Brightfield image of collected and purified microgels (10 wt% 4‐arm PEG‐Ac (10 kDa, 10 eq. GMA)) in water. The scale bar represents 50 µm. E) Diffusion coefficients D of fluorescein isothiocyanate (FITC) dextran (20 kDa) in microgels with different polymer compositions and 10 eq. GMA was determined via FRAP technique. Data are shown as mean ± standard deviation, n = 5 microgels. Statistical significance is performed using two‐way ANOVA with Tukey´s test for multiple comparisons: ns = non‐significant. F) Confocal laser‐scanning microscopy (CLSM) images of the diffusion of fluorescein amine dextran (70 kDa) into microgels (10 wt% 4‐arm PEG‐Ac (10 kDa, 10 eq. GMA)) after 96 h of incubation. Scale bars represent 10 µm.

Figure 2

Time‐dependent binding of fluorescein amine dextran to GMA functionalized microgels (10 wt% 4‐arm sPEG‐Ac (10 kDa). A) Flow cytometry‐based mean fluorescence intensity (MFI) of microgels (10 eq. GMA) incubated with fluorescein amine dextran (70 kDa) for 2, 4, 8. and 16 days. B) Measurements of fluorescence in corresponding supernatant via plate reader. C) Flow cytometry plots of microgels containing increasing amounts of GMA ranging from 0.5 to 10 eq. that are incubated with fluorescein amine dextran for 4 days. D) Corresponding plot of the mean fluorescence intensity (MFI) shown in (C), n ≥ 3. E) Confocal images of microgels containing different amounts of GMA (0.5, 1, 4, and 10 eq.) that are incubated with fluorescein amine dextran (70 kDa) for 4 days. Scale bars represent 10 µm. F) Effective Young‘s modulus (E _eff) of microgels containing different eq. of GMA determined via nanoindentation; n ≥ 10 microgels. Data are presented as mean ± standard deviation. Statistical significance is performed using two‐way ANOVA with Tukey´s test for multiple comparisons: ns = non‐significant; ^* p < 0.05, ^** p < 0.01, ^**** p < 0.0001.

Figure 3

Functionalization of microgels (10 wt% 4‐arm sPEG‐Ac (10 kDa, 4 and 10 eq. GMA)) with IgG antibody against human TNFα. The concentration of antibodies in microgels prepared with 4 (A) and 10 (B) eq. of GMA, as well as in the corresponding supernatants (C,D) is measured via a CBQCA protein quantification kit. Data are presented as mean ± standard deviation, n = 3. Statistical significance is performed using two‐way ANOVA with Tukey´s test for multiple comparisons: ns = non‐significant; ^**** p < 0.0001.

Figure 4

TNFα binding to microgels (10 wt% 4‐arm sPEG‐Ac (10 kDa, 10 eq. GMA)) ([microgels] = 200 000 microgels mL⁻¹ = 100 µg mL⁻¹). A) 0.5 µg mL⁻¹ TNFα is incubated with microgels containing different glycidyl‐antibody ratios [IgG:GMA]. Non‐bound TNFα is measured in the supernatant via TNFα ELISA. B) 2.5 µg mL⁻¹ TNFα is incubated with microgels containing different glycidyl‐antibody ratios [IgG:GMA]. Non‐bound TNFα is measured in the supernatant via TNFα ELISA. C) Antibody‐functionalized microgels (IgG:GMA = 0.1) are incubated with increasing concentrations of TNFα. Non‐bound TNFα is measured in the supernatant via TNFα ELISA and shown as a percentage. D) Shows the correlation between the amount of incorporated antibody and scavenged TNFα for a fixed concentration of 2.5 µg mL⁻¹ TNFα. Data are presented as mean ± standard deviation, n = 3. Statistical significance is performed using two‐way ANOVA with Tukey´s test for multiple comparisons: ns = non‐significant, ^*** p < 0.001, ^**** p < 0.0001.

Figure 5

Specific TNFα binding to TNFα antibody‐functionalized microgels ([microgels] = 50 000 microgels mL⁻¹). Confocal images of microgels (10 wt% 4‐arm sPEG‐Ac (10 kDa, 10 eq. GMA)) to show FRET signal. A) Brightfield image, B) TNFα antibody (acceptor, Alexa Fluor 546) functionalized microgels are excited (𝜆 = 561 nm) and detected (𝜆 = 571–700 nm) at acceptor wavelength. C) TNFα (donor, Alexa Fluor 488) incubated with non‐functionalized microgels are excited (𝜆 = 488 nm) and detected (𝜆 = 498–540 nm) at donor wavelength, and D) FRET signal: TNFα antibody (acceptor) functionalized microgels incubated with TNFα (donor) are excited at donor excitation wavelength (𝜆 = 488 nm) and detected (𝜆 = 571–700 nm) at acceptor wavelength. E) TNFα antibody (acceptor) functionalized microgels incubated with TNFα (donor) are excited at donor excitation wavelength (𝜆 = 488 nm) and detected (𝜆 = 498–540 nm) at donor wavelength. Fluorescence lifetime images (FLIM) of F) only donor in solution, G) non‐functionalized microgels incubated with donor, and H) TNFα antibody (acceptor) functionalized microgels incubated with donor (FRET sample). All scale bars represent 10 µm. I) Schematic of only donor in solution, J) Schematic of non‐functionalized microgels incubated with donor, and K) Schematic of TNFα antibody (acceptor) functionalized microgels incubated with donor (FRET sample). L) Lifetime histograms from the FLIM images (F) (red curve), (G) (green curve), and (H) (blue curve). M) The fluorescence decays and respective exponential fittings correspond to the FLIM images (F–H).

Figure 6

Microgels efficiently scavenge TNFα in two different cellular models. A,B) TNFα antibody‐functionalized microgels protect HT29 cells from TNFα induced cyclooxygenase II (COX II) production, as well as interleukin 8 (IL8) release into the cell culture supernatant. Cells are exposed with 5 ng mL⁻¹ TNFα alone, with TNFα together with 100 µg mL⁻¹ of antibody‐functionalized or non‐functionalized microgels, or with microgels only for 24 h. (A) shows a COX‐II western blot of protein lysates of harvested cells. ß‐actin is used as a loading control. B) IL8 concentration in the cell culture supernatant is measured after incubation for 24 h. Data are presented as mean ± standard deviation, n = 3. Statistical significance is performed using two‐way ANOVA with Tukey´s test for multiple comparisons: ns = non‐significant, ^**** p < 0.0001. C) TNFα antibody‐functionalized microgels scavenge TNFα produced by human macrophages upon stimulation with lipopolysaccharide (LPS) (n = 5 donors). Macrophages are stimulated with 100 ng mL⁻¹ LPS, LPS, and either antibody‐functionalized or non‐functionalized microgels, or microgels only for 24 h. As a negative control, the cells are treated with a normal cell culture medium. The graph shows the levels of TNFα in the cell culture supernatant of the different groups. Data are presented as mean ± standard deviation, n = 5. Statistical significance is performed using two‐way ANOVA with Tukey´s test for multiple comparisons: ns = non‐significant, ^** p < 0.01.