Short-range cytokine gradients to mimic paracrine cell interactions in vitro

Abstract

Cell fate decisions in many physiological processes, including embryogenesis, stem cell niche homeostasis and wound healing, are regulated by secretion of small signaling proteins, called cytokines, from source cells to their neighbors or into the environment. Concentration level and steepness of the resulting paracrine gradients elicit different cell responses, including proliferation, differentiation or chemotaxis. For an in-depth analysis of underlying mechanisms, in vitro models are required to mimic in vivo cytokine gradients. We set up a microparticle-based system to establish short-range cytokine gradients in a three-dimensional extracellular matrix context. To provide native binding sites for cytokines, agarose microparticles were functionalized with different glycosaminoglycans (GAG). After protein was loaded onto microparticles, its slow release was quantified by confocal microscopy and fluorescence correlation spectroscopy. Besides the model protein lysozyme, SDF-1 was used as a relevant chemokine for hematopoietic stem and progenitor cell (HSPC) chemotaxis. For both proteins we found gradients ranging up to 50μm from the microparticle surface and concentrations in the order of nM to pM in dependence on loading concentration and affinity modulation by the GAG functionalization. Directed chemotactic migration of cells from a hematopoietic cell line (FDCPmix) and primary murine HSPC (Sca-1(+) CD150(+) CD48(-)) toward the SDF-1-laden microparticles proved functional short-range gradients in a two-dimensional and three-dimensional setting over time periods of many hours. The approach has the potential to be applied to other cytokines mimicking paracrine cell-cell interactions in vitro.

Mimicking Paracrine TGFβ1 Signals during Myofibroblast Differentiation in 3D Collagen Networks

“…msHA showed high affinity for TGFβ1 in previous studies already 11 . As we already discussed in our last paper, available binding sites introduced by GAG functionalization inside the µ-beads exceed binding sites occupied by TGFβ1 even at high cytokine concentrations 10 . µ-beads were laden overnight in TGFβ1 solutions with concentrations of 5, 10, 50 and 100 µg/ml.…”

“…First we proved the TGFβ1 release characteristic of our established µ-bead system 10 . µ-beads were covalently functionalized with a chemically sulfated hyaluronan, thereby creating binding sites for proteins like TGFβ1 (Fig.…”

“…This behavior is characteristic for affinity-based release systems and can be fitted with simplified diffusion models 10, 13, 14 . Thus, TGFβ1 mobility inside the µ-beads can be estimated with a diffusion coefficient of 5 ∙ 10 −4  µm 2 /s.…”

“…Comparable dose-dependent experiments at standard cell culture conditions did not indicate relevant myofibroblast differentiation at such low concentrations of TGFβ1 23 . We attribute this effect to the two important functional properties of our µ-bead release system: i) the sustained release of TGFβ1 over several days in combination with ii) protection of TGFβ1 from proteolytic degradation due to GAG binding inside the µ-beads 10, 24 . The first feature imitates the continuous production of TGFβ1 by neighboring cells, while the second feature mimics functions of in vivo ECM 25 .…”

“…These networks exhibit a fibrillar microstructure and a pore size similar to dermal ECM 9 . To deliver TGFβ1 in a localized, slow and sustained manner, a release system based on glycosaminoglycan (GAG) functionalized agarose microbeads (µ-beads) was used, as recently introduced for the chemokine SDF-1 10 . Differentiation of dermal fibroblasts into myofibroblasts within the 3D biomimetic matrices was determined by staining of TGFβ1 downstream target Smad2/3, which is phosphorylated and afterwards translocated to the nucleus, and αSMA incorporation into the actin cytoskeleton as well-known marker of contractile myofibroblasts (Fig.…”