Distinct Fates of Chemokine and Surrogate Molecule Gradients: Consequences for CCR7-Guided Dendritic Cell Migration

Abstract

Chemokine-guided leukocyte migration is a hallmark of the immune system to cope with invading pathogens. Intruder confronted dendritic cells (DCs) induce the expression of the chemokine receptor CCR7, which enables them to sense and migrate along chemokine gradients to home to draining lymph nodes, where they launch an adaptive immune response. Chemokine-mediated DC migration is recapitulated and intensively studied in 3D matrix migration chambers. A major caveat in the field is that chemokine gradient formation and maintenance in such 3D environments is generally not assessed. Instead, fluorescent probes, mostly labelled dextran, are used as surrogate molecules, thereby neglecting important electrochemical properties of the chemokines. Here, we used site-specifically, fluorescently labelled CCL19 and CCL21 to study the establishment and shape of the chemokine gradients over time in the 3D collagen matrix. We demonstrate that CCL19 and particularly CCL21 establish stable, but short-distance spanning gradients with an exponential decay-like shape. By contrast, dextran with its neutral surface charge forms a nearly linear gradient across the entire matrix. We show that the charged C-terminal tail of CCL21, known to interact with extracellular matrix proteins, is determinant for shaping the chemokine gradient. Importantly, DCs sense differences in the shape of CCL19 and CCL21 gradients, resulting in distinct spatial migratory responses.

Keywords: CCL19; CCL21; CCR7; chemokine gradient formation and maintenance; dendritic cell migration; fluorescent chemokines.

Figure 1

Visualization and characterization of fluorescently labelled dextran, CCL19 and CCL21 gradients in 3D migration chambers. (A) Schematic representation of a 3D migration chamber. Cells are embedded in a defined 3D collagen I matrix of an Ibidi μ-slide suitable for time-lapse fluorescence microscopy. Fluorescently labelled molecules and control medium are filled into the adjacent connected reservoirs, such that a gradient is established over time following a source-and-sink principle. The central chamber corresponds to the observation area, which we subdivided into four quarters (Q1-Q4) for certain analysis. (B-D) Distribution of fluorescently labelled molecules in the central observation chamber was measured by time-lapse fluorescent microscopy. One reservoir was filled with 100 nM dextran^TMR (B), CCL19-S6^Dy549P1 (C) or CCL21-S6^Dy549P1 (D). Mean fluorescence intensity (MFI) of the labelled proteins within the 3D matrix was monitored over time for 5 hours, fitted with a one-phase decay model and color graded from blue (t = 0min) to red (t = 300min). One out of two independent experiments is shown.

Figure 2

CCL19, CCL21 and CCL21trunc establish stable gradients of distinct shapes. (A) Mean fluorescence intensity (MFI) of CCL19-S6^Dy549P1 and CCL21-S6^Dy649P1 in relation to the distance to the reservoir measured after 5 min, 180 min and 300 min of incubation. (B) MFI values of CCL19-S6^Dy549P1, CCL21-S6^Dy649P1 [taken from (A)] and CCL21trunc-S6^Dy549P1 were translated in a pseudocolor representation, and supplemented with an artificial linear gradient for comparison. (C) Applying a one-phase exponential decay function reveals different 50 % initial fluorescence values (d_1/2) for dextran^TMR, CCL19-S6^Dy549P1, CCL21-S6^Dy549P1 and CCL21trunc-S6^Dy549P1 after 180 min. (D) Spatial distribution of d_1/2 as the ‘mean chemokine diffusion’ over time increases for CCL19-S6^Dy549P1, but not for CCL21-S6^Dy549P1.

Figure 3

Distinct shapes of CCL19 and CCL21 gradients translate into different migratory responses of human DCs. (A) Representative cropped microscopic images of the central 3D migration chamber. Bright field (BF) as well as pseudocolored fluorescence images of medium, CCL19-S6^Dy549P1 and CCL21-S6^Dy549P1 at indicated time points derived from one out of three independent experiments are shown (left panel). Velocity and forward migration index in the y-axis (yFMI) of human matured MoDCs migrating along the chemokine gradients in each quarter are shown in the right panel. Mean values ± SD for n = 3, statistical analysis: two-way ANOVA with Sidak post-test. (B) Spider plots of individual tracks of migrating MoDCs in quarters Q1 and Q4.