Nanotopography controls cell cycle changes involved with skeletal stem cell self-renewal and multipotency

Abstract

In culture isolated bone marrow mesenchymal stem cells (more precisely termed skeletal stem cells, SSCs) spontaneously differentiate into fibroblasts, preventing the growth of large numbers of multipotent SSCs for use in regenerative medicine. However, the mechanisms that regulate the expansion of SSCs, while maintaining multipotency and preventing fibroblastic differentiation are poorly understood. Major hurdles to understanding how the maintenance of SSCs is regulated are (a) SSCs isolated from bone marrow are heterogeneous populations with different proliferative characteristics and (b) a lack of tools to investigate SSC number expansion and multipotency. Here, a nanotopographical surface is used as a tool that permits SSC proliferation while maintaining multipotency. It is demonstrated that retention of SSC phenotype in culture requires adjustments to the cell cycle that are linked to changes in the activation of the mitogen activated protein kinases. This demonstrates that biomaterials can offer cross-SSC culture tools and that the biological processes that determine whether SSCs retain multipotency or differentiate into fibroblasts are subtle, in terms of biochemical control, but are profound in terms of determining cell fate.

Keywords: Cell cycle; Mesenchymal stem cells; Nanotopography; Skeletal stem cells.

Fig. 1

Determining the optimal seeding density for SQ nanotopography. (a) AFM image representative of our self-renewal promoting nanotopographical pattern; the surface of a SQ substrate produced by injection moulding. Dark regions indicate the location of the pits. The graph to the right of the images shows the depth and spacing of the features. Scale bar: 1 μm. (b) STRO-1 protein levels in SSCs cultured on SQ nanotopographies using a conventional seeding method (suspension method at ∼ 3000 cells/cm²). Retention of STRO-1 was demonstrated with western blot densitometry after 7 days of culture. Results represent relative density of STRO-1 protein normalised to GAPDH, expressed in arbitrary units (a.u.). (c) Schematic of a seeding device for even cell distribution. Each sterile seeder (produced in-house) was placed on top of the polycarbonate substrate. (d) Full substrate images of SSCs stained with Coomassie Blue after 28 days culture. Distinct growth patterns were observed that appeared associated with initial seeding density. At the low and mid-density conditions (500 cells/cm² and 1000 cells/cm² respectively), growth of cells were mostly confined to colonies (clonal growth), whereas at the highest seeding density (2000 cells/cm²), cells spread throughout the surface (fibroblastic growth) reaching full confluence. These observations were not restricted to the SQ nanotopography, and was thus a substrate-independent effect. Representative images shown (n = 3). (e) Percentage of CD271 protein expression at each tested seeding density as a function of the flat control. The data summarises western blot densitometry results (CD721 normalised to GAPDH) following 28 days culture of CD271⁺ SSCs. It can be seen that the highest CD271 stem cell marker levels are found when utilising an initial seeding density of 1000 cells/cm². Each of the other densities tested resulted in a lower percentage of CD271. (f) Levels of CD271 protein in individual cells after 28 day culture were assessed across different seeding densities to provide detailed analysis using CellProfiler image analysis. This involved quantifying integrated intensity levels for each cell across multiple fluorescence images (>40). The data reflects the trends observed with western blotting, in that 1000 cells/cm² gave the highest CD271 readings on SQ relative to flat. At 500 cells/cm², CD271 was similar across the two substrate types whereas at 2000 cells/cm², CD271 was not effectively retained on SQ. *p < 0.05, ANOVA (mean ± SEM, a.u. = arbitrary units). (g) Assessment of growth (mean nuclei count ± SEM) showed that CD271⁺ SSCs increased in number on SQ as time progressed (n = 3). (h) BrdU staining (mean % BrdU positive nuclei ± SEM) indicated a comparable degree of proliferation between flat and SQ (n = 3). (i) Western blot densitometry (STRO-1 normalised to GAPDH) showing that by seeding STRO-1⁺ SSCs at 1000 cells/cm² on SQ, the effect of stem cell marker retention was reproducible. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 2

Linking changes in the metabolome to biochemical networks. MS-metabolomics analysis of CD271⁺ SSCs (7 day culture) highlighted that the top networks on SQ with reference to flat controls had associations with (a) p38 MAPK (b) ERK 1/2 and (c) Akt. Contributions to these signalling hubs came largely from metabolites involved in amino acid biosynthesis, glycolysis and nucleotide biosynthesis. Network maps were generated using IPA (n = 6).

Fig. 3

Cell cycle regulation on nanotopography. (a) AFM image representative of the surface of our osteogenic differentiation nanopattern, NSQ. Dark regions indicate the location of the pits, which have a degree of offset (±50 nm) in comparison to the ordered SQ nanotopography. The graph to the right of the images gives an indication of the pit depth and spacing. Scale bar: 1 μm. (b) Verification of the effects of NSQ on SSCs was tested by culturing CD271⁺ SSCs for 28 days and western blot analysis. Results shown are relative OPN protein levels normalised to alpha tubulin (expressed in a.u.). It can be seen that OPN is higher on NSQ, suggesting that osteogenesis is taking place as expected. (c) Synchronised and subsequently released SSCs were assessed for changes in levels of cell cycle-related proteins. Graphs represent mean fluorescence intensity (a.u.) values ± SEM for the protein of interest, normalised to either total protein (pRb to total Rb, pERK to total ERK) or GAPDH (n ≥ 4). Significant changes were observed in cyclin D1, E2F-1, pRb, pERK and CDK6, suggesting that cell cycle regulation is altered by nanotopography. *p < 0.05, **p < 0.01, by ANOVA. (d) Flow cytometry data illustrating an increased preponderance for SSCs to be in G0/G1 phase on SQ (n = 2).

Fig. 4

Mechanisms underlying SQ induced self-renewal. (a) Graph (inset) shows in cell western results for total ERK 1/2 for SQ (red) and flat (blue) across 0, 1, 4 and 24 h. ERK 1/2 was consistently lower on SQ when compared to the control (significantly at t = 0), and decreased progressively when moving through the cell cycle. ***p < 0.001, t-test. RNAseq analysis generated a network in IPA that links ERK 1/2 and adhesion-based signalling on SQ 24 h following release into the cycle after synchronisation. ERK 1/2 was also indicated to be down-regulated here. (b) Several genes associated with cell cycle progression (CDK1 (cdc2), cyclin B, cyclin D1) and those related to cycle repression (p18INK4C, p19INK4D and p21^cip1) were up-regulated (green) on in cells on SQ (arrows point to all named genes). (c) Canonical pathway analysis underlines differential regulation of cell cycle (denoted as *) and adhesion (denoted as §) in cells on the SQ topography compared to those on control. A general trend of down-regulation can be observed across the majority of signalling groups for cells on on SQ. Note that only pathways with p < 0.05 are shown. A two-fold change cut-off was applied for RNAseq data (n = 3). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 5

Comparison of amino acid networks generated by metabolic analysis. A common amino acid network with associations to ERK 1/2 was identified on SQ and NSQ in STRO-1⁺ SSCs after 7 and 28 days of culture. (a) After 7 days, a high degree of down-regulation (green) was apparent on SQ with (b) some up-regulations (red) being observed after 28 days (c) at 7 days on NSQ more amino acids were up-regulated (red) than SQ. (d) this trend became more obvious after 28 days, where the majority of amino acids linked to ERK 1/2 were up-regulated (n = 3). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)