The Effect of the Topmost Layer and the Type of Bone Morphogenetic Protein-2 Immobilization on the Mesenchymal Stem Cell Response

Abstract

Recombinant human bone morphogenetic protein-2 (rhBMP-2) plays a key role in the stem cell response, not only via its influence on osteogenesis, but also on cellular adhesion, migration, and proliferation. However, when applied clinically, its supra-physiological levels cause many adverse effects. Therefore, there is a need to concomitantly retain the biological activity of BMP-2 and reduce its doses. Currently, the most promising strategies involve site-specific and site-directed immobilization of rhBMP-2. This work investigated the covalent and electrostatic binding of rhBMP-2 to ultrathin-multilayers with chondroitin sulfate (CS) or diazoresin (DR) as the topmost layer. Angle-resolved X-ray photoelectron spectroscopy was used to study the exposed chemical groups. The rhBMP-2 binding efficiency and protein state were studied with time-of-flight secondary ion mass spectrometry. Quartz crystal microbalance, atomic force microscopy, and enzyme-linked immunosorbent assay were used to analyze protein-substrate interactions. The effect of the topmost layer was tested on initial cell adhesion and short-term osteogenesis marker expression. The results show the highest expression of selected osteomarkers in cells cultured on the DR-ended layer, while the cellular flattening was rather poor compared to the CS-ended system. rhBMP-2 adhesion was observed only on negatively charged layers. Cell flattening became more prominent in the presence of the protein, even though the osteogenic gene expression decreased.

Keywords: AR-XPS; TOF-SIMS; bone morphogenetic protein 2; cell culture surfaces; chondroitin sulfate; diazoresin.

Figure 1

(A) Graphical representation of the angle-resolved XPS measurements. The outermost and deeper layers were analyzed at angles of 15 and 55 degrees, respectively. Data were collected using a series of sample tilt angles: 15, 25, 35, 45, and 55 degrees. Contribution of varying carbon chemical states for (DR/CS)₆ and (DR/CS)₆DR. XPS-fitted C1s line for photocrosslinked (DR/CS)₆ (B) and photocrosslinked (DR/CS)₆DR (C). The spectra shown were collected at a 15-degree angle.

Figure 2

Changes in the actin cytoskeleton architecture in hUC-MSC grown on both types of DR/CS multilayer films. Cells were cultured in a medium supplemented with 2% FBS for 24 h and were then immunostained for vinculin and counterstained with TRITC-phalloidin (F-actin) and DAPI (nuclei). The scale bar represents 25 μm.

Figure 3

(A) Cell area 24 h after seeding. (B) DNA concentration of hUC-MSC proliferation evaluated after 24, 48, 72 h, and 7 days of cell-culture on a clean substrate, (DR/CS)₆ and (DR/CS)₆DR. (C) Expression of osteogenic markers in hUC-MSCs cultured on a clean substrate, (DR/CS)₆, and (DR/CS)₆DR for 7 days in vitro. Quantitative real-time PCR analysis of selected osteogenic markers (ALPL–alkaline phosphatase; COL10A1–collagen type 10; OCN–osteocalcin) in hUC-MSCs cultured in DMEM/F12, 2% FBS). Gene expression was normalized to the expression of a housekeeping gene–GADPH, shown as the fold-change compared to gene expression in hUC-MSCs cultured on the clean substrate. p values less than 0.05 (p < 0.05) were considered statistically significant in comparison to the control and are labeled with an asterisk (*).

Figure 4

Schematic presentation of the obtained surfaces with preparation steps; E–protein bound electrostatically, C–protein bound covalently. Non-photocrosslinked multilayers are marked as a thicker system, while photocrosslinked layers are thinner.

Figure 5

TOF-SIMS analysis of the rhBMP-2 dimer covalently and electrostatically bonded to (DR/CS)₆ and (DR/CS)₆DR. (A) PC1 is loaded negatively and positively by the mass signals of amino acids and substrates, respectively. (B) PC2 is loaded negatively and positively by the mass signals of amino acids exposed to the surface; PC2 is loaded positively by the mass signals of Arg, Ser, and Gly and negatively by the mass signals of Ala, Pro, and Lys, which relates to the different structural state of rhBMP-2. (C) Scores plot with four groups of data points, centered in different quadrants of the coordinate system PC1 vs. PC2, corresponding to all samples.

Figure 6

(A) rhBMP-2 adsorption/desorption QCM-D measurement using the Au/SiO₂ sensor coated with (DR/CS)₆ before photocrosslinking (dotted curves) and (DR/CS)₆ and (DR/CS)₆DR after photocrosslinking (solid curves), expressed as the mass shift (green and red curves) and dissipation shift (black and grey curves). rhBMP-2 was used as an aqueous solution of 1 μg/mL, pH 6.2, and flow rate of 0.15 mL/min, adhered for 30 min (blue area). (B) Representative topography images of (DR/CS)₆ multilayers in the liquid state before and after photocrosslinking and after the deposition of rhBMP-2 from 1 μg/mL solution: covalent and electrostatic bonding. Measurements were performed in sterile water at room temperature using tapping mode. The root-mean-square roughness (Rq) values corresponding to each image are shown. (C) Release of rhBMP-2 detected by ELISA from surfaces of different types.

Figure 7

(A) Changes in the actin cytoskeleton architecture of hUC-MSCs cultured on selected surfaces with immobilized rhBMP-2. Cells were cultured in a medium supplemented with 2% FBS for 24 h and were then immunostained for vinculin and counterstained with TRITC-phalloidin (F-actin) and DAPI (nuclei). The scale bar represents 25 μm. (B) Cell area after 24 h of seeding measured for 30 cells for each condition, shown as the fold-change vs. appropriate pure polymeric systems (controls). (C) Expression of osteogenic markers in hUC-MSCs cultured on selected surfaces for seven days. Quantitative real-time PCR analysis of selected osteogenic markers (ALPL–alkaline phosphatase; COL10A1–collagen type 10; OCN–osteocalcin) in hUC-MSCs cultured in DMEM/F12, 2% FBS. Gene expression was normalized to the expression of a housekeeping gene–GADPH, shown as the fold-change compared to gene expression in hUC-MSC cultured on the clean substrate–glass coverslips (blue line). p values less than 0.05 (p < 0.05) were considered statistically significant in comparison to the control and are labeled by an asterisk (*).