Human immunocompetent choroid-on-chip: a novel tool for studying ocular effects of biological drugs

Abstract

Disorders of the eye leading to visual impairment are a major issue that affects millions of people. On the other side ocular toxicities were described for e.g. molecularly targeted therapies in oncology and may hamper their development. Current ocular model systems feature a number of limitations affecting human-relevance and availability. To find new options for pharmacological treatment and assess mechanisms of toxicity, hence, novel complex model systems that are human-relevant and readily available are urgently required. Here, we report the development of a human immunocompetent Choroid-on-Chip (CoC), a human cell-based in vitro model of the choroid layer of the eye integrating melanocytes and microvascular endothelial cells, covered by a layer of retinal pigmented epithelial cells. Immunocompetence is achieved by perfusion of peripheral immune cells. We demonstrate controlled immune cell recruitment into the stromal compartments through a vascular monolayer and in vivo-like cytokine release profiles. To investigate applicability for both efficacy testing of immunosuppressive compounds as well as safety profiling of immunoactivating antibodies, we exposed the CoCs to cyclosporine and tested CD3 bispecific antibodies.

Fig. 1. Concept of the CoC model.

a Schematic representation of the different cell types within the chip. b Schematic of the layers of the eye. c CoC platform filled with watercolor demonstrating the location of the RPE (green), the endothelial (orange) and melanocyte (blue) channels. d Live/Dead staining performed with Fluorescein diacetate/Propidium Iodide (FDA/PI) to demonstrate the viability of the long-term culture (2 weeks) of the CoC as well an orthogonal view. Scale bar = 500 µm for planar images and 100 µm for the orthogonal view. e Bright field image of the whole chip after 2 weeks in culture and 24 h if immune cell perfusion.

Fig. 2. Endothelial and outer blood-retina barrier (oBRB) characterization after 2 weeks in culture.

a Immunofluorescence staining demonstrating the phenotype and structure of the retinal pigmented epithelium as a confluent monolayer expressing the tight junction marker zonula occludens (ZO-1, in green) at high magnification (left) and throughout the whole cell culture surface on the chip (middle, long image of ZO-1, in green) as well as melanin production, demonstrated by the expression of TYRP1 (red) and pigmentation demonstrated in the bright field imaging (right, gray). b Immunofluorescence staining demonstrating the phenotype and structure of the endothelium formed by microvascular endothelial cells positive for CD31 (red), presenting a stronger staining in the cell junctions when cultured in the absence of RPE and melanocytes (left) than in co-culture (middle, long image of CD31, in red). The DAPI/CD31 staining represents the bottom endothelial layer (left), top endothelial layer (middle) and orthogonal side view of both, with a yellow line corresponding to the bottom membrane. The 3D render (right) confirmed the presence of 2 confluent layers. (c) Quantification of the relative permeability of the endothelial (MVEC)/stromal and oBRB barriers when perfused with low (Carboxyfluorescein, 0.377 kDa) and high molecular weight (Dextran Texas Red, 70 kDa) fluorescence molecules. Relative permeability is presented as a measure of the lag time difference to achieve 50% of the compound concentration (right) extracted from the molecule diffusion kinetic curve (n = 2 for the CoC and n = 3 for the Hydrogel control, bars represent SEM) (left) and a representation of the kinetic profile measured for one chip, normalized to the highest fluorescence intensity in each channel of the CoC and for both molecular weight molecules. d Representative images of the fluorescent measurements, labeled with white circles, performed to quantify relative permeability. The measurements were performed in each of the individual channel, where there is no channel overlap, on the inlet side of the chip. The top circle corresponds to the MVEC channel, the middle to the stromal channel and the bottom to the RPE channel. The images show a fluorophore filled MVEC channel after 552 s of perfusion. Scale bars are 500 µm, except for square images and the orthogonal view in in (a) and (b), where the scale bars represent 50 µm.

Fig. 3. Characterization of the melanocyte compartment after 2 weeks in culture.

Morphology of pigmented melanocytes seeded at low (a, top) and high (b, bottom) densities and embedded in the dextran hydrogel containing Arginylglycylaspartic acid (RGD peptide): Bright field images demonstrate pigmentation; autofluorescence shows melanin, FDA/PI labeling indicates viability (Epifluorescence imaging of the whole chip), and KI-67 staining (Confocal imaging) a low proliferative state. Scale bar: 500 µm (Ki-67 images: 100 µm); (c) 3D representation of the cell distribution in the melanocyte compartment (left). Immunofluorescence microscopy highlighting pigmentation (bright filed; top left), melanin autofluorescence (top right), cell nuclei (DAPI; bottom left) and the presence of TYRP1 (bottom right) (Confocal imaging). d Effect of melanocyte density on the cytokine secretion in the presence of peripheral blood mononuclear cells (PBMCs). The bars represent the average values and the dots the individual values (n = independent biological replicates). Statistical analysis is detailed in Supplementary data 1, including individual data points.

Fig. 4. Validation of PBMC perfusion, viability, recruitment and cytokine release upon activation with anti CD3/CD28 antigens (Act).

a 3D representations of the evaluation of immune cell recruitment on chip containing only hydrogel (left) or supplemented with CCL19 (middle and right) and perfused with Deep Red cell tracker labeled PBMCs, treated with anti CD3/CD28 antigens (middle and left) or non-treated (right) showing that in the absence of a stimuli, there is no immune cell recruitment. The yellow line represents the position of the bottom membrane of the chip. b Evaluation of the cells collected after 24 h of perfusion on its viability (left), concentration (middle) and fraction of recovery relative to the plate control (n = 4–15) measured by flow cytometry. The effluents were collected from the chips with low melanocyte density (black) and from control well plates with (dark gray) or without co-culture with melanocytes (light gray). c Quantification of CD8 positive cells and of T cell activation markers CD25 and CD69 within the CD8 positive population by flow cytometry analysis of the effluents in the chips with low melanocyte density (striped bars) and on the plate controls (n = 4–8). d Evaluation of the immune cell recruitment towards the 3D melanocyte compartment in chips with low melanocyte density and activated (Act) distinguishing PBMCs (red) and T cells (green, labeled with an anti-CD3 antibody) within the melanocytes (*). The figure shows the top of the compartment where the chip bottom membrane is located and its pores are focused (right) and the bottom of the compartment (left), with an orthogonal representation of the whole chip height (bottom). Scale bar is 50 µm. e Quantification of immune cell recruitment by image analysis of the whole chip with low melanocyte density, distinguishing PBMCs (cell tracker labeled prior perfusion) and T cells (CD3 post fixation staining) in the top (dark red and green) and bottom (light red and green) halves of the melanocyte compartment in the chips with low melanocyte density. Statistical analysis represents the comparison of total immune cells between chips perfused with PBMCs and chips perfused with activated cells (*P < 0.1, n = 4 for NT and n = 12 for Act). f Quantification of the cytokines IL-2, IL-10, IL-6, IFN-γ, TNF-α and granzyme B on the effluents of chips after 24 h of perfusion. Statistical analysis represents the comparison between chips with low melanocyte density perfused with PBMCs and chips perfused with activated cells (***P < 0.01, n = 3–5). Bars represent averages ± SEM. Statistical analysis is detailed in Supplementary data 1, including detailed information of the number of biological replicates per group, exact p values and individual data points.

Fig. 5. Effect of the immunosuppressor drug Cyclosporine (CsA) on the Immuno-competent CoC with low melanocyte density after 24 h of perfusion or cell culture (plate controls) respectively, in PBMCs, activated PBMCs (Act), treated and non-treated with high (500 ng/mL) and low (100 ng/mL) CsA concentrations.

a Viability of the immune cells collected in the chip effluents and plate control supernantants (n = 3–10, the dots represent each data point). b Fraction of cells recovered in in the chip effluents and plate control supernantants relative to the number of recovers cells in the PBMC control (n = 3–10). c Flow cytometry quantification of CD8 fraction and the T cell activation markers CD25 and CD69 in the CD8 fraction in activated PBMCs (Act) with and without CsA treatment (n = 4–10). d Cytokine profile in the choroid chip in activated PBMCs (Act) with and without CsA treatment relative to PBMCs (n = 6–12). e Cytokine profile in plate in PBMCs and activated PBMCs (Act) with and without CsA treatment relative to PBMCs (n = 5). f Quantification of immune cell recruitment by image analysis of the whole chip, distinguishing PBMCs (cell tracker labeled prior perfusion) and T cells (CD3 post fixation staining) in the top (dark red and green) and bottom (light red and green) halves of the melanocyte compartment chips with low melanocyte density Statistical analysis represents the comparison of total immune cells between chips perfused with PBMCs and chips perfused with activated cells (*P < 0.1, n = 4–8). g Representative images of the melanocyte bottom compartment containing labeled PBMCs (Cell tracker, red) and T cells (CD3, green) represented as maximum intensity projections for each of the analyzed condition. Scale bar = 500 µm. Bars represent Averages ± SEM. Statistical analysis represents the comparison of the CsA treatment in activated cells with the same condition not exposed to CsA (*P < 0.1, **P < 0.05, ***P < 0.01), unless otherwise labeled. Statistical analysis is detailed in Supplementary data 1, including detailed information of the number of biological replicates per group, exact p values and individual data points.

Fig. 6. Effect of the TCBs on the Immuno-competent Choroid-chip with low melanocyte density after 24 h of perfusion or cell culture (plate controls) respectively, in PBMCs treated and non-treated with TCB A or TCB B (5 µg/mL).

a Viability of the immune cells collected in the chip effluents and plate control supernatants (n = 3–10, the dots represent each data point). b Fraction of cells recovered in the chip effluents and plate control supernatants relative to the number of recovered cells in the PBMC control (n = 3–10). c Cytokine profile in the TCB A and TCB B treated choroid chips relative to PBMCs (n = 6–12). d Quantification of immune cell recruitment by image analysis of the whole chip, distinguishing PBMCs (cell tracker labeled prior perfusion) and T cells (CD3 post fixation staining) in the top (dark red and green) and bottom (light red and green) halves of the melanocyte compartment chips with low melanocyte density (n = 3–9). e Representative images of the melanocyte compartment containing labeled PBMCs (Cell tracker, red) and T cells (CD3, green) represented as maximum intensity projections for each of the analyzed condition. Scale bar = 500 µm. Bars represent Averages ± SEM. Statistical analysis represents the comparison of the TCBs treatment in PBMCs with cells activated with antiCD3/CD28 beads (*P < 0.1, **P < 0.05, ***P < 0.01), unless labeled otherwise, and the comparison of the total and bottom compartments of the TCBs treatment in PBMCs with non-treated PBMCs (^#P < 0.1). Statistical analysis is detailed in Supplementary data 1, including detailed information of the number of biological replicates per group, exact p values and individual data points.