Skip to main content
NCBI home page
UKPMC Funders Author Manuscripts logoLink to UKPMC Funders Author Manuscripts
. Author manuscript; available in PMC: 2026 Jun 27.
Published in final edited form as: Biomaterials. 2020 Feb 13;240:119854. doi: 10.1016/j.biomaterials.2020.119854

Dimensionality changes actin network through lamin A/C and zyxin

Jip Zonderland 1, Ivan Lorenzo Moldero 1, Shivesh Anand 1, Carlos Mota 1, Lorenzo Moroni 1,*
PMCID: PMC7619213  EMSID: EMS214271  PMID: 32087459

Abstract

Mechanosensing proteins have mainly been investigated in 2D culture platforms, while understanding their regulation in 3D enviroments is critical for tissue engineering. Among mechanosensing proteins, the actin cytoskeleton plays a key role in human mesenchymal stromal cells (hMSCs) activity, but its regulation in 3D tissue engineered scaffolds remains poorly studied. Here, we show that human mesenchymal stromal cells (hMSCs) cultured on 3D electrospun scaffolds made of a stiff material do not form actin stress fibers, contrary to hMSCs on 2D films of the same material. On 3D electrospun and additive manufactured scaffolds, hMSCs also displayed fewer focal adhesions, lower lamin A and C expression and less YAP1 nuclear localization and myosin light chain phosphorylation. Together, this strongly suggests that dimensionality prevents the build-up of cellular tension, even on stiff materials. Knock down of either lamin A and C or zyxin resulted in fewer stress fibers in the cell center. Zyxin knock down reduced lamin A and C expression, but not vice versa, showing that this signal chain starts from the outside of the cell. Lineage commitment was not affected by the lack of these important osteogenic proteins in 3D, as all cells committed to osteogenesis in bi-potential medium. Our study demonstrates that dimensionality changes the actin cytoskeleton through lamin A and C and zyxin, and highlights the difference in the regulation of lineage commitment in 3D enviroments. Together, these results can have important implications for future scaffold design for both stiff- and soft tissue engineering constructs.

Keywords: Dimensionality, Actin, Lamin, Zyxin, Mesenchymal stromal cells

1. Introduction

Understanding cellular responses, such as differentiation and proliferation, to material properties (e.g. stiffness, chemistry, and topography) are critical endeavors in fields like tissue engineering and regenerative medicine. Unraveling the molecular mechanisms underlying these cellular responses can lead to more intelligent design of tissue engineering constructs. Cells adhere to extracellular matrix (ECM) proteins through integrins, forming focal adhesion complexes that connect the actin cytoskeleton to the ECM [1]. The other end of the actin filament, it is attached to either another focal adhesions, or to the nucleus, by binding to the linker of nucleoskeleton and cytoskeleton (LINC) complex, which in turn is linked to lamin A and C [2]. Lamin A and C form a protein meshwork under the nuclear membrane to give structural integrity to the nucleus [3–5]. Lamin B1 and B2 are also part of the lamin protein meshwork. B-type lamins have been shown to influence nuclear integrity, while nuclear stiffness is mainly determined by lamin A and C [4,6]. The different roles and functions of lamin A and C have not been widely investigated; the reader is referred to an excellent review on this topic [7]. Actin filaments join together to form stress fibers, with incorporated non-muscle myosin to create contractile force between its two attachment points [8]. On stiffer materials, focal adhesions and stress fibers have been shown to be bigger and more abundant than on softer materials [9–12], creating a higher cellular tension [12,13]. Indeed, lamin A and C expression, indirectly attached to actin stress fibers, has also been shown to increase on stiffer materials [5]. Besides material stiffness, other factors such as material chemistry and topography have also been shown to influence focal adhesions, actin stress fibers and lamin A and C [14–16]. Yes-associated protein 1 (YAP1) is an important mechanosensitive co-transcription factor that translocates to the nucleus at higher cellular tension to transduce these mechanical changes in the cell to changes in gene expression [17].

On stiffer materials and with more cellular tension human mesenchymal stromal cells (hMSCs) show increased osteogenic differentiation, while softer materials and lower cellular tension enhance differentiation to chondro- and adipogenic lineages [18–22]. These changes in hMSC differentiation have been shown to be orchestrated by actin stress fibers [23], focal adhesions [23], lamin A and C [24] and Yes-associated protein 1 (YAP1) [25,26].

While other material properties are relatively well studied, the dimensionality (2D v. 3D) of a material has not yet been widely studied. All 3D tissue engineering constructs inherently introduce dimensionality, but direct comparison between cells cultured in 2D and 3D, made of the same material, are sparsely reported. It is therefore important to understand the effect of dimensionality on the proteins involved in cellular tension, as they are critical for hMSC differentiation [23–26]. Some studies have investigated the role of the dimensionality on cellular tension in soft hydrogels [27]. While this gives valuable insights for some tissue engineering applications, many 3D tissue engineering constructs are made of stiff materials. Therefore, we investigated the effect of dimensionality in 3D electrospun (ESP) and additive manufactured (AM) scaffolds and compared to flat films made of the same stiff material (300PEOT55PBT45, 100 MPa [28]) (Fig. 1a). Specifically, we focused on four main parts of the cellular tension machinery: focal adhesions (zyxin and paxillin), actin cytoskeleton, nuclear skeleton (lamin A and C) and a mechanosensitive co-transcription factor (YAP1).

Fig. 1. Different actin network organization and lower Lamin A and C in hMSCs cultured on 3D ESP scaffolds.

Fig. 1

Open in a new tab

a, hMSCs cultured on 300PEOT55PBT45 2D films and 3D electrospun (ESP) scaffolds stained for F-actin (green) and nuclei (blue). The bottom panels are magnifications (4 ×) of the respective images above. Scale bars represent 15 μm (top panels) and 3 μm (bottom panels). b, Quantification of the F-actin intensity distribution in hMSCs cultured on 300PEOT55PBT45 2D films or 3D ESP scaffolds. n = 21 for films and n = 23 for ESP from 3 biological replicates. Two-way ANOVA; **p < 0.01, ***p < 0.001, ****p < 0.0001. c, Quantification of F-actin intensity that overlaps with the nucleus in hMSCs cultured on 300PEOT55PBT45 2D films or 3D ESP scaffolds. n = 12 for films and n = 13 for ESP from 3 biological replicates. Student’s t-test, ****p < 0.0001. d, Lamin A and C expression in hMSCs grown on 300PEOT55PBT45 2D films, 3D AM, and ESP scaffolds. TBP shown as loading control in the blots. Graph depicts average expression of lamin A or C/TBP normalized to films, quantified by western blots from 4 independent experiments. One-way ANOVA; *p < 0.05, **p < 0.01, ***p < 0.001 compared with films. e, Lamin A and C expression in hMSCs cultured on polystyrene, as 2D films or 3D AM scaffolds. TBP shown as loading control in the blots. Graph shows the average expression of lamin A or C/TBP normalized to films, quantified by Western blot from 3 independent experiments. Ratio paired t-test; **p < 0.01. b-e, Error bars represent mean ± 95% CI. f, Lamin A and C expression on day 1, 4 and 7 after seeding hMSCs on 300PEOT55PBT45 2D films, 3D additive manufactured (AM) or electrospun (ESP) scaffolds. TBP shown as loading controls. Quantification of western blots from 3 independent experiments shows average expression of lamin A or C/TBP normalized to 2D films at day 1. Error bars represent mean ± SD. One-way ANOVA; *p < 0.05 **p < 0.01 compared with expression on 2D film at day 1. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article).

We demonstrate that dimensionality influenced the actin stress fiber formation, focal adhesion formation, lamin A and C expression and YAP1 nuclear localization in these commonly used 3D tissue engineering scaffolds. The 3D scaffolds changed the actin network of hMSCs through a decrease in zyxin expression, which decreases lamin A and C expression and together change the actin cytoskeleton.

2. Results

2.1. Dimensionality prevents formation of actin stress fibers and reduces lamin A and C

To start investigating how dimensionality influences cellular tension in stiff materials, we cultured hMSCs on 300PEOT55PBT45 3D AM and 3D ESP scaffolds, or 2D films, for 7 days (Figs. S1a–b) and looked at the actin cytoskeleton (Fig. 1a). The electrospun scaffolds had a fiber diameter of 0.99 ± 0.18 μm and a thickness of ~50 μm. Cells infiltrated and grew on top of the scaffolds. Using AM, we created stacks of fibers of ~200 μm, with square pores of ~650 μm, in a 0–90 pattern (Fig. S1). 300PEOT55PBT45 is a stiff material (100 MPa [28]), Young’s moduli of the films, AM and ESP scaffolds were: 98 ± 23 MPa (tensile), 2.0 ± 0.1 MPa (compression) and 1.2 ± 0.2 MPa (tensile), respectively (Figs. S2a–c). As expected, on 2D films large actin stress fibers formed and F-actin fibers were distributed throughout the whole cell (Fig. 1b). When hMSCs were cultured on ESP 3D scaffolds, we observed far fewer actin stress fibers and a change in F-actin distribution. F-actin fibers were mainly located on the cell periphery and very little F-actin was observed in the cell center or overlapping with the nucleus (Fig. 1c). The lack of many stress fibers and a more pronounced peripheral actin network is a sign of lower cellular tension [29], even though cells were cultured on stiff materials. F-actin distribution of hMSCs in AM 3D scaffolds was attempted, but could not be evaluated due to the very high cell density in these 3D constructs. In this high cell density environment, the F-actin distribution of individual cells could not be discriminated from F-actin of surrounding cells.

To understand why few actin fibers were found overlapping with the cell nucleus, we evaluated the expression of lamin A and C, a protein that assists in linking actin filaments to the nucleus [2]. After 7 days of culture, lamin A and C expression were significantly reduced in AM scaffolds, by 30 ± 3.5% (p < 0.001) and 46 ± 14% (p < 0.05), respectively, compared to 2D films (Fig. 1d). Lamin A and C expression was even more reduced in hMSCs cultured on the ESP scaffolds, by 85 ± 6.5% (p < 0.01) and 91 ± 8.0% (p < 0.05), respectively, compared to those cultured on 2D films. Phosphorylation of Lamin A and C at Ser22 induces dislocalization from the nuclear membrane and later degradation [30,31]. The ratio between phosphorylated and total lamin A and C was increased in the 3D ESP scaffolds 2.3 ± 0.5x (p < 0.05) and 5.9 ± 3.5x (p < 0.05), respectively, compared to films (Fig. S3). This shows that the reduction of lamin A and C in 3D ESP scaffolds comes partly from an increase of phosphorylation.

The same trend of lamin A and C expression was observed for polystyrene, another stiff material: cells expressed 44 ± 4.7% (p < 0.01) less lamin A and 71 ± 6.5% (p < 0.01) less lamin C when cultured on AM 3D scaffolds, compared to 2D films (Fig. 1e). Young’s moduli of the polystyrene films and AM 3D scaffolds were 882 ± 112 MPa and 89 ± 33 MPa, respectively. From here on, all 2D film, 3D AM and ESP scaffolds will be 300PEOT55PBT45.

To determine how lamin A and C expression changed over the culture period, we measured their expression levels at days 1, 4, and 7 (Fig. 1f). Lamin A and C expression in hMSCs cultured on the 3D ESP or AM scaffolds remained low at all time points. In comparison, lamin A and C steadily increased over time in hMSCs cultured on 2D films. Lamin A was 3.2 ± 0.43 and 6.2 ± 1.0 times higher after day 4 and day 7, respectively, compared to day 1. Lamin C was 3.7 ± 1.0 and 8.4 ± 3.5 times higher on day 4 and day 7, respectively, compared to day 1. These findings show that due to the dimensionality, lamin A and C expression remains low over the culture period, in contrast to 2D, where lamin A and C expression increases over time.

2.2. Improved migration through small pores of cells cultured in 3D

Lower lamin A and C expression has been shown to decrease nuclear stiffness and improve migration through small pores [5,32]. To test whether the change in lamin A and C expression affected this migration capacity, we observed hMSC migration from 2D films or 3D scaffolds, through a transwell with 3-μm or 8-μm pores. Indeed, cells cultured on 2D films and harboring higher lamin A and C expression migrated less well through 3-μm pores than through 8-μm pores (p < 0.05) (Fig. 2a). Cells migrating from ESP scaffolds migrated equally well through 3-μm or 8-μm pores. The ratio of hMSCs migrating through 3-μm pores to those not migrating was significantly higher from ESP scaffolds than 2D films (p < 0.01). hMSCs on AM scaffolds did not have sufficient physical contact with the transwells to migrate from the AM scaffold to the transwell membrane (data not shown), so their migration ability could not be evaluated. These data indicate that lower lamin A and C expression from hMSCs in 3D ESP resulted in improved migration through small pores compared to hMSCs grown on 2D films, potentially due to lower nuclear stiffness of the cells,.

Fig. 2. More nuclear folds and improved migration in cells from 3D ESP and Lamin A and C influences F-actin organization.

Fig. 2

Open in a new tab

a, Migration of hMSCs from 300PEOT55PBT45 2D films or 3D ESP scaffolds through a transwell with 8-μm or 3-μm pores. 8 μm ESP: n = 18; 3 μm ESP n = 19; 8 μm Film n = 18; 3 μm Film n = 19, from 4 independent experiments.Krusal Wallis test; *p < 0.05, ***p < 0.001. b, Representative images of examples of folded and non-folded nuclei of hMSCs grown on 300PEOT55PBT45 2D films, 3D AM or 3D ESP scaffolds, stained with Lamin A and C (green), scalebars 5 μm. Arrowheads indicate what was considered a nuclear fold. The graph shows the corresponding quantification. Total counted nuclei for films: 491, AM: 159, ESP: 79, in 8, 11 and 10 different images, respectively, from 2 independent experiments. Kruskal Wallis test, **p < 0,01, ****p < 0,0001 compared to films. c, hMSCs transduced with scrambled-(SCR, left) or LMNA-shRNA (right), cultured on TCP, stained for F-actin (green) and nuclei (blue). The bottom panels are 7 × magnifications of the respective images above. Scale bars represent 30 μm (top panels) and 5 μm (bottom panels). d, Quantification of the F-actin intensity distribution in hMSCs transduced with scrambled- or LMNA-shRNA, cultured on 2D TCP. n = 33 for SCR and n = 29 for LMNA-shRNA from 3 biological replicates. Two-way ANOVA; ****p < 0.0001. e, Quantification of F-actin intensity that overlaps with nuclei staining in hMSCs transduced with scrambled- or LMNA-shRNA. n = 25 cells analyzed for SCR and n = 22 for LMNA from 3 biological replicates. Mann-Whitney test, ****p < 0.0001. a-e, Error bars represent mean ± 95% CI. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article).

While a stiff nucleus shows a uniform shape, lower nuclear stiffness results in more heterogeneous shapes and more ‘folds’ [33,34]. Indeed, nuclei of cells cultured in 2D showed mostly uniform nuclei, with folds in 33 ± 11% nuclei. From hMSCs on the 3D scaffolds, 76 ± 12% (p < 0.01) and 96 ± 4% (p < 0.0001) of nuclei displayed folds from AM and ESP scaffolds, respectively (Fig. 2b).

Together, these observations suggest that lower lamin A and C expression, resulting from the dimensionality of the 3D scaffolds compared to 2D films composed of the same stiff material, reduced nuclear stiffness, an observation not previously reported.

2.3. Lamin A and C plays a role in shaping the F-actin network

Actin tension has been shown to influence lamin A and C expression [30], but the role of lamin A and C in shaping the actin network is not known. To investigate this, cells were transduced with LMNA shRNA, knocking down both lamin A and C (Fig. S4), and stained for F-actin (Fig. 2c). A large change in actin organization was observed in the LMNA knock down. Similar to the actin organization of hMSCs on 3D ESP scaffolds, the LMNA knock down showed a clear decrease in F-actin in the cell center and over the nucleus (Fig. 2d and e). Thus, lamin A and C plays an important role in shaping the actin network and could be partly responsible for the difference in actin organization in 2D vs 3D.

2.4. Cell density does not explain differences in lamin A and C expression observed between 2D and 3D

Because cell density inherently differs between 2D and 3D substrates, we wished to determine whether it was the reason for the differences we observed in lamin A and C expression. To determine the influence of cell density on lamin A and C expression, we cultured hMSCs for 7 days with different cell seeding densities on 2D TCP (Fig. 3a). The different cell densities were chosen to reach confluency after 1 day (20k cells/cm2), 2 days (10k cells/cm2), or roughly 40%, 60%, 70%, or 80% confluency after 7 days (1.25k, 2.5k and 5k cells/cm2, respectively), or little cell-cell contact after 7 days (625 cells/cm2). Increasing cell seeding density increased lamin A and C expression after 7 days of culture (Fig. 3a). Lamin A was 2.4 ± 0.5x higher at 10k than at 625 cells/cm2 (p < 0.05), while lamin C was 5.0 ± 0.6x higher at 20k than at 625 cells/cm2 (p < 0.05).

Fig. 3. Cell seeding density influences lamin A and C expression.

Fig. 3

Open in a new tab

a–c, Lamin A and C expression of hMSCs seeded on 2D TCP (a), 300PEOT55PBT45 3D AM (b), or 300PEOT55PBT45 3D ESP (c) in varying densities. Western blots (top) and quantification of western blots (graphs below) from 3 independent experiments. Error bars represent mean ± SD. Values were normalized to 625 cells/cm2 (a), 100k cells/scaffold (b), or 10 K cells/scaffold (c). One-way ANOVA; *p < 0.05 compared with 625 cells/cm2 (a) or 10k cells/scaffold (c);. ***p < 0.001, compared to 10k cells/ scaffold. TBP shown as a loading control. d, Representative Western blot (left) of hMSCs cultured on 300PEOT55PBT45 films at a medium cell density, or on 300PEOT55PBT45 3D ESP scaffolds at a high cell density. Representative images of hMSCs on 2D films seeded at 5k cells/cm2 (left) and on a 3D ESP scaffold seeded with 120k cells (right) show the observed cell density at the time of harvest (day 7), visualized by F-actin staining (green). Scale bars represent 50 μm. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article).

As lamin A and C were influenced by cell density in 2D, we next examined whether this was also true in the 3D scaffolds. For the 3D ESP scaffolds, different cell densities were chosen to have confluency after 1 day (120k cells/scaffold), near confluency after 7 days (30k cells/ scaffold), or little cell contact after 7 days (10k cells/scaffold). For 3D AM scaffolds, confluency was more difficult to determine. With 400k cells/scaffold, most pores were filled after 1–2 days of culture. With 200k and 100k cells/scaffold, full pores were attained after 4–5 and 7 days of culture, respectively. Importantly, little cell proliferation was observed in 3D AM scaffolds (data not shown), thus the increased filling of the scaffold was not related to proliferation, unlike cells on 3D ESP or 2D TCP.

Interestingly, cell density on the AM 3D scaffolds did not influence the lamin A or C expression (Fig. 3b). Similar to 2D TCP, however, increasing cell seeding density significantly increased lamin A expression in 3D ESP scaffolds, with 2.5 ± 0.1 (p < 0.001) and 3.7 ± 1.0 (p < 0.05) times more lamin A from 30k and 120k cells/scaffold, respectively, compared to 10k cells/scaffold (Fig. 3c). Lamin C expression was not significantly changed by cell density on the ESP 3D scaffold, with 2.1 ± 1.0 and 1.9 ± 0.87 times more lamin C expression from 30k and 120k cells/scaffold compared to the 10k cells/scaffold.

At 5k/cm2, cells don’t reach full confluency on 2D films after 7 days, while cells are fully confluent after seeding at 120k/scaffold (Fig. 3d). Lamin A and C expression was still much higher on 2D films than on ESP seeded at this high cell density, regardless of the increase in lamin A and C with increased cell density. This shows that even though lamin A and C increase with increased cell density, it does not explain the differences observed here between 2D and 3D cell culture systems.

2.5. Fewer focal adhesions in 3D environments

To understand why lamin A and C expression and the actin organization change in response to dimensionality, we looked at focal adhesion formation of cells cultured on 3D AM or ESP scaffolds, compared to 2D films. As expected, many large focal adhesions were found on flat films, visualized by zyxin staining, a marker for mature focal adhesions (Fig. 4a). Very few and faint focal adhesions were observed in cells cultured in the 3D AM and ESP scaffolds. hMSCs cultured in 3D AM or ESP scaffolds displayed 76.0 ± 16.3% (p < 0.0001) and 85.8 ± 8.4% (p < 0.0001) fewer zyxin positive focal adhesions per cell area than 2D films (Fig. 4b). The same trend was observed for paxillin positive focal adhesions, an early focal adhesion marker (Fig. S5a). Interestingly, total protein expression of zyxin did not change between 2D films and 3D scaffolds (Fig. 4c). However, expression of paxillin was reduced in both 3D AM and 3D ESP scaffolds, 75% ± 18% (p < 0.05) and 57% ± 11% (p < 0.05), respectively (Fig. S5b).

Fig. 4. Decreased focal adhesions in 3D.

Fig. 4

Open in a new tab

a, hMSCs on 300PEOT55PBT45 2D films, 3D additive manufactured (AM) or electrospun (ESP) scaffolds stained for zyxin (red) and nuclei (blue). The bottom panels are magnifications (4 ×) of the respective images above. Scale bars represent 20 μm (top panels) and 5 μm (bottom panels). b, Quantification of the number of zyxin positive focal adhesions, normalized to cell area, in hMSCs grown on 300PEOT55PBT45 films, AM or ESP 3D scaffolds. One-way ANOVA. ****p < 0,0001. n = 20 cells for each condition. Error bars represent mean ± 95% CI. c, Zyxin expression in hMSCs cultured on 300PEOT55PBT45 2D film, 3D AM and ESP scaffolds. TBP is shown as a loading control. Graph shows Western blot quantifications of zyxin/TBP, normalized to 2D films, from 4 independent experiments. Error bars represent mean ± SD. d, hMSCs seeded on 2D TCP at different cell densities and stained for zyxin (red) and nuclei (blue), with the bottom panels showing magnifications (8 ×) of the respective images above. Scale bars represent 40 μm (top panels) and 5 μm (bottom panels). e, Quantification of the number of zyxin positive focal adhesions, normalized to cell area, in hMSCs grown on 2D TCP at different cell seeding densities. Student’s t-test. ****p < 0,0001. n = 20 cells for each condition. Error bars represent mean ± 95% CI. f, Zyxin expression in hMSCs seeded in different cell densities and cultured on TCP for 7days. TBP is shown as a loading control. Graphs show Western blot quantification as zyxin/TBP, normalized to 625 cells/cm2, from 3 independent. Error bars represent mean ± SD. One-way ANOVA; *p < 0.05 compared to 625 cells/cm2. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article).

Fewer focal adhesions in response to dimensionality correlated to the reduced lamin A and C expression in the 3D scaffolds. In 2D, lamin A and C increased with higher cell seeding densities. To see whether focal adhesion formation also follows the same correlation in 2D, focal adhesion formation and zyxin and paxillin expression were analyzed with different cell seeding densities in 2D TCP, after 7 days of culture. Indeed, at a lower cell seeding density (625 cells/cm2), 65.9 ± 0.1% (p < 0.0001) fewer zyxin positive focal adhesions formed than at higher cell seeding density (20k/cm2) (Fig. 4d–e). Stainings for paxillin revealed a similar trend (Fig. S5c). Total zyxin expression was increased 2.3 ± 0.3, 2.6 ± 0.5 and 2.1 ± 0.2 times (p < 0.05) in 5k, 10k and 20k cells/cm2, compared to 625 cells/cm2, respectively. Paxillin expression also followed this trend, where at 20k cells/cm2 paxillin expression was 2.0 ± 0.43 times higher than at 1.25k/cm2 (p < 0.05) (Fig. S5d).

Together, these data show that dimensionality reduces focal adhesion formation. In addition, a positive correlation between focal adhesions and lamin A and C expression was found, hinting at a possible connection between the two.

2.6. Reduced YAP1 nuclear translocation and pMLC2 in 3D cultures

YAP1 translocates to the nucleus when there is higher cellular tension [35]. To determine the behavior of YAP1 in response to dimensionality, we measured the nuclear translocation of YAP1 in hMSCs cultured on AM or ESP scaffolds, compared to flat films (Fig. 5a). From both AM and ESP scaffolds, the nuclear/cytoplasmic ratio of YAP1 was significantly lower (0.3 ± 0.1 (p < 0.0001) and 0.7 ± 0.2 (p < 0.0001), respectively, compared to that on 2D films 1.9 ± 0.7) (Fig. 5b); these low ratios show that more YAP1 remained in the cytoplasm than translocated to the nucleus.

Fig. 5. Reduced YAP1 nuclear localization and MLC2 phosphorilation in 3D.

Fig. 5

Open in a new tab

a, YAP1 staining (green) of hMSCs grown on 300PEOT55PBT45 2D films, 3D AM and ESP scaffolds. Top panels show YAP1 staining alone, while the bottom panels show YAP1 and nuclei (blue). Scale bars represent 30 μm b, Quantification of the nuclear to cytoplasmic intensity of YAP1 staining in individual cells. Krusal Wallis test; ***p < 0.001, ****p < 0.0001 compared to films. Total counted cells for films: 27, AM: 28 and ESP: 22 from 4 images. c, Phosphorylated myosin light chain 2 (Thr18 and Ser19) (green, top panels) and F-actin staining (grey, middle panels) of hMSCs cultured on 300PEOT55PBT45 2D films, 3D AM or ESP scaffolds. Bottom panels show merge of the respective images above, plus nuclei (blue). Scalebars represent 25 μm d, Quantification of average pMLC2 intensity per cell. One-way ANOVA; ****p < 0.0001 compared to films. Total counted cells for films: 8, AM: 14, ESP: 17. b, d, Error bars represent mean ± 95% CI. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article).

In addition, we tested the phosphorylation of myosin light chain 2 (MLC2) on Thr18 and Ser19 (pMLC2). Phosphorylation of MLC2 induces contraction of the actin-myosin bundles and induces cellular tension [36]. In 2D films, clear pMLC2 staining was observed, overlapping with F-actin fibers (Fig. 5c). hMSCs in both AM and ESP scaffolds, however, displayed very little pMLC2 staining. Average staining intensity per cell was 88,9 ± 5,3% (p < 0,0001) and 67,4 ± 13,1% (p < 0,0001) lower in AM and ESP than in 2D films, respectively. Individual cells are difficult to distinguish in the 3D AM scaffolds. As the quantification was done on maximum-intensity images of Z-stacks, it is possible that the pMLC2 intensity was measured in more than one cell, slightly overestimating the intensity. However, even with this potential overestimation, pMLC2 staining intensity is still much lower in hMSCs cultured in 3D AM than 2D films.

The lack of YAP1 nuclear translocation and phosphorylation of MLC2 strongly indicates that the dimensionality of the 3D cultures causes a lower cellular tension in hMSCs. These findings are in line with the reduced focal adhesions, actin stress fibers and lamin A and C expression in response to dimensionality.

Previous work hypothesized that YAP1 regulates lamin A and C [5,37]. To test if YAP1 plays a role in lamin A and C expression, we knocked down YAP1 in hMSCs cultured on TCP and determined the levels of lamin A and C expression. No difference in lamin A or C expression was found between scrambled control and YAP1 knockdowns (Fig. S6). This shows that lamin A and C are not regulated through YAP1 and indicates that the lower lamin A and C levels in the 3D scaffolds are not due to a lack of YAP1 activity in the 3D scaffolds.

2.7. Zyxin influences lamin A and C expression and actin organization

The dimensionality introduced by the 3D AM and ESP scaffolds caused lower lamin A and C levels and fewer focal adhesions and actin stress fibers. We next determined whether the focal adhesions are involved in controlling the low lamin A and C expression and resulting changed actin network. Zyxin is known to be critical for focal adhesion formation and the connection to the actin cytoskeleton [38,39], and was therefore chosen as target to diminish the focal adhesion formation. We knocked down zyxin in hMSCs using two different ZYX-shRNAs (Fig. S7), cultured the cells on TCP, then measured lamin A and C expression and evaluated the actin network.

Lamin A was reduced 92 ± 5% (p < 0.01) and 71 ± 13% (p < 0.05) and lamin C reduced 86 ± 11% and 76 ± 19% by the two ZYX-shRNAs, respectively, compared to the scrambled control (Fig. 6a). In the ZYX knockdowns, less actin was observed in the cell center and over the nucleus, while more F-actin was found in the cell periphery compared to that of the scrambled control (Fig. 6b–d), a localization pattern similar to the LMNA knockdown and cells cultured on ESP scaffolds (Fig. 1). The reduction of central actin stress fibers after ZYX knockdown is in line with previous reports [38,39].

Fig. 6. Zyxin influences lamin A and C expression, but not vice versa.

Fig. 6

Open in a new tab

a, e, Western blots of zyxin and lamin A and C expression in hMSCs transduced with scrambled or two different ZYX-shRNAs (a) or two different LMNA-shRNAs (e). Graphs show the quantification of lamin A/TBP (a) and zyxin/TBP (e) expression averaged from 4 biological replicates, and normalized to expression from SCR-shRNA. Error bars represent mean ± SD. One-way ANOVA; **p < 0.01 compared to SCR. b, hMSCs transduced with scrambled or ZYX-shRNA, stained for F-actin (green) and nuclei (blue). The bottom panels show 5 × magnifications of the respective images above. Scale bars represent 50 μm (top) and 10 μm (bottom). c, Quantification of the F-actin intensity distribution in hMSCs transduced with scrambled or ZYX-shRNA. n = 33 cells analyzed for SCR and n = 32 for ZYX-shRNA from 3 biological replicates. Two-way ANOVA; ****p < 0.0001. d, Quantification of F-actin intensity that overlaps with nuclei staining in hMSCs transduced with scrambled or ZYX-shRNA cultured on 2D TCP. n = 25 cells analyzed for SCR and n = 26 for ZYX from 3 biological replicates. Mann-Whitney test, ****p < 0.0001. c, d, Error bars represent mean ± 95% CI. f, hMSCs transduced with scrambled or LMNA-shRNAs stained for zyxin (green) and nuclei (blue). Bottom panels show 5 × magnifications of the respective images above. Scale bars represent 50 μm (top panels) and 10 μm (bottom panels). a-f, All cells were cultured on 2D TCP for 7 days. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article).

Zyxin also greatly reduced the number and intensity of paxillin positive focal adhesions, although it could still be visualized (Fig. S8), also in line with previous reports [40]. Paxillin is an important protein for the downstream signaling of focal adhesions [41]. To test if lamin A and C expression are influenced by paxillin, we knocked paxillin down. No difference was observed in lamin A or C expression, demonstrating that the reduction of lamin A and C following zyxin knock down does not occur through paxillin (Fig. S6). Indeed, also the actin distribution in the cells was not affected by paxillin knock down (Fig. S9).

To test whether the influence of zyxin on lamin A and C expression is bidirectional, we looked at zyxin expression in LMNA knockdowns. Zyxin expression and focal adhesion formation were similar in LMNA knockdowns and scrambled controls (Fig. 6e and f), demonstrating that lamin A and C expression does not affect zyxin levels. Furthermore, these data suggest that the faint focal adhesions we observed from hMSCs on 3D scaffolds (Fig. 4a) were not due to the low lamin A and C expression.

Interestingly, in both ZYX and LMNA knockdowns, YAP1 was still located mainly in the nucleus (Fig. S10). This indicates that the lack of YAP1 nuclear localization in 3D is through a mechanism independent of lamin A and C or zyxin.

2.8. Osteogenic preference for hMSCs in 2D and 3D

In 2D, higher expression of lamin A and C, YAP1 nuclear translocation and more focal adhesion, stress fibers and p-MLC2 have all been linked to increased osteogenesis, while a reduction in these have been shown to favor adipogenesis [20,23–26,37,42–45]. To test whether the dimensionality-induced reduction in these proteins has an effect lineage commitment of hMSCs, we cultured hMSCs in 2D films, 3D AM and ESP scaffolds in bipotential medium (mixed osteo- and adipogenic medium) for 21 days. hMSCs differentiated to osteoblasts were visualized with osteocalcin, while PPAR-γ staining was used for adipogenic cells (Fig. S11). As expected, all cells committed to the osteoblast lineage when cultured on 2D films (Fig. 7a). Surprisingly, however, also all cells cultured in 3D AM and ESP scaffolds committed to the osteoblast lineage, and no PPAR-γ positive cells were found in any of the culture platforms.

Fig. 7. Osteogenic preference of hMSCs in 2D and 3D.

Fig. 7

Open in a new tab

a, hMSCs were expanded for 7 days on 300PEOT55PBT45 2D films, 3D AM or 3D ESP scaffolds and then differentiated for 21 days in bipotential medium (osteo and adipogenic differentiation medium mix). Cells were stained for osteocalcin (green, top panels), to identify osteogenic cells, or for PPAR-γ (red, bottom panels) to identify adipogenic cells. Nuclei are stained in blue. Scalebars represent 30 μm b, Model showing differences in actin fiber formation, focal adhesion and lamin A and C in 3D (left) v 2D (right.) In 3D, cellular tension forces (grey arrows) are distributed in multiple directions and focal adhesions (light grey ovals) stay small and few. The reduction of focal adhesions then leads to a reduction in lamin A and C expression (red oval). Together, small and few focal adhesions and low lamin A and C expression then lead to the lack of actin stress fibers (green lines) through the cell center. In 2D, cellular tension forces (grey arrows) are distributed along a single plane, allowing for the build-up tension and large focal adhesions (light grey ovals). This leads to an increase in lamin A and C expression (red oval). Large focal adhesions and high lamin A and C expression then enable the formation of actin stress fibers (green lines) through the cell center.. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

2.9. A model for actin fiber formation mediated by lamin A and C and zyxin in 2D v. 3D

Our data comes together in the model depicted in Fig. 7b. In 3D, dimensionality prevents focal adhesion formation, which leads to a decrease in lamin A and C. The lack of both zyxin and lamin A and C then shapes the actin network and does not allow for the formation of actin stress fibers through the cell center. In 2D, the lack of dimensionality allows for the formation of many large focal adhesions, leading to an increase in lamin A and C. High zyxin and lamin A and C expression then allow for the formation of high cellular tension and large actin stress fibers. This mechanism is independent of both YAP1 and paxillin.

3. Discussion

In this study, we show that hMSCs cultured in 3D ESP or AM scaffolds display a decrease in focal adhesions, lamin A and C expression, YAP nuclear localization, actin stress fibers and MLC2 phosphorylation, compared to 2D films of the same material. We show that zyxin influences lamin A and C expression, but not vise-versa. Both zyxin and lamin A and C play a role in the formation of stress fibers in the cell center and are partly responsible for the different actin organization in 3D.

Fraley et al. also showed a decrease in focal adhesions in response to dimensionality in or on hydrogels [27]. Also, in a 3D in vivo environment, focal adhesions were found to be very different in shape and composition from 2D cultures [46]. How dimensionality initiates a difference in the focal adhesion protein zyxin and focal adhesion formation has not been studied here and remains unclear. Potential mechanisms include a difference in force distribution. On concave surfaces, focal adhesion formation and lamin A and C have been shown to decrease when compared to flat surfaces, while they increase on convex surfaces [47]. This indicates that the force distribution within the cell and the angle at which it is connected to the environment partly determines focal adhesion formation and lamin A and C expression. We hypothesize that because of the dimensionality, forces are not distributed over the nucleus like in 2D, but more throughout the whole cell and along the cell periphery, which could explain the more peripheral actin network in 3D environments. The knock down of lamin A and C could mimic what happens in a 3D environment, where forces can no longer be distributed to the nucleus, which leads to the lack of actin stress fibers near the nucleus. Indeed, stress fibers and focal adhesions can still be found in the LMNA knock downs, but are then located almost exclusively in the cell periphery (Fig. 2b, c, 5f).

We found that YAP1 was excluded from the nucleus on ESP and AM scaffolds, in contrast to 2D films. However, in LMNA and ZYX knockdowns, YAP1 was still located in the nucleus. The opening of nuclear pores due to force on the nucleus has been shown to allow YAP1 to enter the nucleus [35]. A possible explanation is that because of lower nuclear stiffness in both LMNA and ZYX knockdowns (due to lack of lamin A and C), YAP1 can enter the nucleus even if the cell is under less tension [35]. In 3D, the nucleus might be under less tension, as forces are distributed in all 3 dimensions, whereas in 2D forces are concentrated in a single plane. In 2D, a large portion of the forces are going over the nucleus, and pushing down on it [35]. This lack of tension on the nucleus in 3D could potentially result in avoiding YAP1 to enter the nucleus in 3D.

Here, we have presented a link between lamin A and C and the actin cytoskeleton, and lamin A and C and focal adhesions, specifically zyxin. Others have found a similar link between lamin A and C and the actin cytoskeleton, more specifically with actin-cap fibers [33], or total actin [30,48], while our results focus F-actin distribution in the cell. We have found that this process is not mediated by YAP1. Other mechanosensitive (co-)transcription factors could be inhibited to look at the effect on lamin A and C and zyxin, to find the signaling pathway involved in orchestrating the differences in expression in 3D environments.

Decoupling individual variables in 2D vs 3D in stiff materials remains challenging. Local and bulk material properties can change when creating scaffolds, potentially affecting cell behavior. Also, ECM produced by cells can be influenced, directly changing the environment. Using hydrogels, 2D vs 3D can be more strictly controlled, while even there the introduction of the third dimension inevitably introduces changes in nutrient diffusion, cell density, adhesion ligand density, among other factors. This makes the study of dimensionality difficult, but not less valuable. In our research presented here we have decoupled the observed effects of dimensionality from cell density by testing different cell densities on the different scaffolds. Also, we have decoupled material chemistry by using polystyrene as another material. However, there are other scaffold design variables that could affect the results. In the case of 3D AM scaffolds this includes fiber diameter, fiber spacing, layer thickness, pore size and shape, and total porosity. For 3D ESP scaffolds it includes fiber diameter, pore size, total porosity, pore interconnectivity, among others. Rather than decoupling each variable in each system to see if it can explain the observed effects, we have used these two very different 3D systems in a more holistic manner. In both 3D systems we find reduced lamin A and C, cytoplasmic retention of YAP1, fewer focal adhesions, less p-MLC2, compared to 2D. Combining all this data, we show here that dimensionality is responsible for the reduction in important mechanosensing proteins. These results are in line with a study on the effect of dimensionality on focal adhesions in 2D vs 3D hydrogels [27].

In 2D, YAP, focal adhesions, actin-myosin and lamin A and C have all been shown to increase osteogenic commitment, while a reduction in these proteins pushes hMSCs towards adipogenesis [20,23–26,37,42–45]. We have tested the lineage differentiation preference in hMSCs cultured on 300PEOT55PBT45 2D films, 3D AM or 3D ESP scaffolds. Interestingly, even though in 3D these important osteogenic proteins were heavily downregulated, cells still exclusively committed to osteogenesis. This data is contrary to most of the literature on 2D, but is in line with a small but growing body of evidence that suggests a different role for these proteins in 3D. For example, cell morphology, a very important factor in 2D differentiation and greatly influenced by the above mentioned proteins [20,42,49], has been fully decoupled from differentiation in 3D [50–52]. In addition, the lack of both YAP nuclear localization and focal adhesions in 3D hydrogels did not limit osteogenic commitment [50]. The data presented here adds to the small body of research of mechanobiology in 3D, which can have important implications for future tissue engineering scaffold design. Further research is required to decouple the different factors that influence differentiation and lineage commitment in 3D environments.

4. Conclusion

Here, we study the effect of dimensionality on cellular tension of hMSCs in two common 3D tissue engineering constructs: ESP and AM scaffolds. We demonstrate that dimensionality causes fewer stress fibers, fewer focal adhesion, lower lamin A and C expression, less YAP1 nuclear localization. Cell density influences lamin A and C expression and focal adhesion formation in 2D and 3D, but it is not responsible for the observed differences between 2D and 3D, further proving that dimensionality is an important environmental factor. Lamin A and C or zyxin knock down resulted in fewer stress fibers in the cell center and over the nucleus. Zyxin knock down also reduced lamin A and C expression, but not vice versa, showing that this signal chain starts from the outside of the cell. In bipotential differentiation medium, all cells in 2D and 3D committed to osteogenesis. This shows that the reduction in these osteogenic mechanosensing factors did not limit cell osteogenic potential, contrary to literature on 2D substrates. Taken together, our study shows that dimensionality changes the actin cytoskeleton through lamin A and C and zyxin and decreases cellular tension, even on stiff materials. Understanding how these key proteins are influenced and further dictate cell behaviour and differentiation can have important implications for future scaffold design.

5. Experimental section

5.1. Scaffold production

Poly(ethylene oxide terephthalate) and poly(butylene terephthalate) random block co-polymer (PEOT/PBT, 300PEOT55PBT45, PolyActive™) with 300 Da PEO and a PEOT/PBT weight ratio of 55/45 was acquired from PolyVation. Polystyrene (PS, 350 kDa) was acquired from Sigma-Aldrich. PEOT/PBT or PS films were produced by melting the polymer in a circular 23-mm mold between two silicon wafers (Simat, Kaufering, Germany) and two hot plates under slight pressure (~100 kg) to ensure films were fully flat. PEOT/PBT was processed at 180 °C and PS at 210 °C.

AM scaffolds were produced by means of screw-extrusion–based fused deposition modeling (FDM) (Bioscaffolder SYSENG, Germany). The FDM extrusion is controlled by the screw rotation and assisted by N2 (5 bar) gas pressure allowing fine control over deposition of the molten polymer. The fabrication of the 20 × 20 × 4 mm scaffolds was achieved following a layer-by-layer manufacturing with 90° rotation between deposited layers. The 3D scaffold CAD models were uploaded into PrimCAM software (Primus Data, Switzerland) and the deposition patterns were calculated. The fiber spacing, defined as the distance between successive fibers in the same layer was defined as 650 μm, the layer thickness was set to 170 μm, and the fiber diameter obtained was according to the nozzle diameter used, the polymer selected and the processing parameters. The parameters that influence the production of the 3D scaffolds are: temperature, screw rotation, deposition velocity. PEOT/PBT or PS pellets were loaded in the reservoir and molten at a temperature of 195 °C or 220 °C, respectively. The screw rotation for the polymers was 200 rpm. The molten polymer was extruded through a nozzle with G25 (I.D. = 250 μm). The deposition velocity was optimized to 20 and 200 mm min−1 for PS and PEOT/PBT, respectively.

To produce ESP scaffolds, 20% (w v−1) 300PEOT55PBT45 was dissolved in a mixture of 70% chloroform (Sigma-Aldrich) and 30% 1.1,1.3,3.3-Hexafluoro-2-propanol AR (HFIP; Bio-Solve) overnight under agitation at room temperature. Electrospinning was done on a mandrel (diameter: 19 cm) slowly rotating at 100 rpm to produce many scaffolds with randomly oriented fibers at the same time under exactly the same conditions. The following conditions were maintained: 15 cm working distance, 1 ml h−1 flow rate, 23–25 °C and 40% humidity. The mandrel was charged between −2 and −5 kV and the needle between 10 and 15 kV. ESP scaffolds were spun on aluminum foil over a polyester mesh with 12-mm holes. After spinning, ESP scaffolds with a diameter of 15 mm were punched out and scaffolds were removed from the aluminum foil. This method yielded ESP scaffolds of 12 mm with a 1.5-mm supporting ring of polyester mesh around them to improve handleability. Fibers were deposited randomly with a diameter of 0.99 ± 0.18 μm, creating mats of approximately 50-μm thick.

Before cell culture, films, AM and ESP scaffolds were sterilized with 70% ethanol for 15 min and dried until visually dry. The films and scaffolds were then coated by absorption with 1 mg ml−1 rat-tail collagen I solution for 16 h at 37 °C. After coating, they were washed twice with water and air-dried before cell seeding. All experiments on TCP were done without collagen coating.

5.2. Mechanical tests

To test the mechanical properties of 300PEOT55PBT45 films, AM and ESP scaffolds, and polystyrene films and AM scaffolds, a TA Electroforce 3200 mechanical tester was used with a 450 N (for films and AM scaffolds) or a 45 N (for ESP scaffolds) load cell. For films and ESP scaffolds, samples were elongated at 1% strain/s and the force and displacement were recorded. AM scaffolds were compressed at 1% strain/s. The elastic moduli were calculated using the slope of the initial linear region, after the toe region.

5.3. Cell culture

hMSCs were isolated from bone marrow by aspiration from a 24-year old female by Texas A&M Health Science Center [53] after ethical approval from the local and national authorities and written consent from the donor. Briefly, mononuclear cells were isolated by centrifugation and expanded hMSCs were tested for differentiation potential, to be received at passage 1. For expansion, hMSCs were cultured on TCP at 1000 cells/cm2 in αMEM + Glutamax +10% fetal bovine serum (basic medium) (Thermo-Fisher Scientific), until 70–80% confluent. All experiments were performed at passage 5. Films were punched with a diameter of 22 mm and cultured in non-treated, 12-well plates at 5k cells/cm2, unless stated otherwise. ESP scaffolds were 15 mm and cultured in 24-well plates at 30k cells/scaffold, unless stated otherwise, with a rubber O-ring (outer diameter 15 mm, inner diameter 12 mm) to prevent the scaffolds from floating and to cover the polyester ring on which the scaffolds were produced. AM scaffolds were square blocks of 5 × 5 × 3 mm (width, length, height) and cells were seeded in a drop of 50 μl containing 400k cells, unless stated otherwise. Two hours after seeding, the AM scaffold was flipped upside down to increase cell distribution. After a total of 4 h after seeding, scaffolds were transferred to a non-treated, 12-well plate for further culture. Film and scaffold cultures were done in basic medium supplemented with 1 ng ml−1 FGF-2 (Neuromics), 200 μM L-Ascorbic acid 2-phosphate (ASAP) (Sigma-Aldrich) and 100 U ml−1 penicillin-streptomycin (P/S), and were harvested after 7 d of culture, unless stated otherwise. For differentiation experiments, hMSCs were cultured for 7 days after seeding on films, AM or ESP scaffolds, before switching to bipotential medium for 21 days. Bipotential medium was a 1:1 mix of osteo- and adipogenic medium. Osteogenic medium: Basic medium+1% P/S, 200 μM ASAP and 10 nM dexamethasone (Sigma-Aldrich). Adipogenic medium: Basic medium +1% P/S, 0.2 mM indomethacin (Sigma-Aldrich), 1 μM dexamethasone, 0.5 mM 3-Isobutyl-1-methylxanthine (Sigma-Aldrich), 2,5 μg ml−1 Insulin (Sigma-Aldrich).

5.4. Protein isolation and Western blot

Proteins were isolated from cells cultured on films or scaffolds in RIPA buffer (Sigma-Aldrich), supplemented with cOmplete™, Mini, EDTA-free Protease Inhibitor Cocktail (Sigma-Aldrich). To get sufficient proteins, 6–12 films, 15–20 ESP scaffolds or 2–4 AM scaffolds were mixed into 300–400 μl lysis buffer for a single protein isolate. Experiments were repeated 3 or 4 times for replicates. For TCP and films, surfaces were scraped with cell scrapers to ensure cell lysis. AM scaffolds were cut into four smaller pieces and submerged in lysis buffer. ESP scaffolds were removed from the polyester supporting ring and submerged in lysis buffer. Samples were spun down at 10.000g, and the supernatant was used for further processing.

Protein quantification was done using the Pierce BCA protein assay kit (Thermo Fisher Scientific). Twenty micrograms of protein was incubated with laemmli loading buffer (Bio-Rad) and 10% 2-Mercaptoethanol (Sigma-Aldrich) for 5 min at 95 °C and loaded into a 4–15% polyacrylamide gel (Bio-Rad). Proteins were transferred to a 0.45 μm PVDF membrane (Bio-Rad) using the semi-dry transfer method. Membranes were blocked for 1 h with 5% fat free milk powder (Bio-Rad) in TBS + 0.05% tween-20 (Sigma-Aldrich). All primary antibody incubations were performed overnight at 4 °C in blocking buffer. All antibodies (lamin A and C: ab108595; paxillin: ab32084; zyxin: ab58210; YAP1: ab52771; TBP: ab51841) were ordered from Abcam and diluted 1/1000, except for YAP1 which was diluted 1/500. Blots were subsequently incubated with 0.33 μg ml−1 goat-anti-rabbit or mouse horseradish peroxidase (Bio-Rad) in blocking buffer for 1 h at room temperature. Protein bands were then visualized using Clarity Western ECL (Bio-Rad). Quantifications of band intensity were done in Fiji using the gel quantification tool.

5.5. Immunofluorescence and imaging

Prior to staining, hMSCs were fixed in 3.6% (v v−1) paraformaldehyde (Sigma-Aldrich) in PBS for 20 min at room temperature. Cells were permeabilized and blocked in 2% bovine serum albumin (BSA) (VWR) and 0.1% Triton X (VWR) in PBS for 1 h at room temperature. Antibodies mentioned above in the same dilution, or osteocalcin (Abcam, ab93876, 1/250) or PPAR-γ (ThermoFisher Scientific, PA1-824, 1/250) were incubated overnight at 4 °C in 2% BSA and 0.05% tween-20 in PBS (incubation buffer). Secondary antibodies goat-anti-rabbit or mouse Alexa Fluor 488 (Thermo Fisher Scientific) were then incubated overnight at 4 °C in incubation buffer. Instead of antibody staining, F-actin staining was done with phalloidin Alexa Fluor 488 (Thermo Fisher Scientific) at room temperature for 20 min in PBS +0.05% tween-20. Nuclei were stained with DAPI (Sigma-Aldrich). Images were taken on a widefield fluorescence microscope, or confocal microscope. In culture conditions in 3D and where necessary in 2D, Z-stacks were taken to ensure the entire cell was imaged from bottom to top, to ensure equal quantification in the different substrates. To allow for quantitative comparison between images, all images within an experiment were taken with the same settings and on the same day.

Actin quantification was done using a custom-built Fiji macro (paper in press). A line was drawn through the cell perpendicular to the long axis of the cell, and the intensity over that line was measured. The intensity distribution was divided into 10 equal bins to correct for differences in cell size. The total F-actin intensity overlapping with the nucleus was also measured.

For YAP1 quantification, the nucleus and an area in the cytoplasm were selected and the overall pixel intensity in each area was measured using Fiji. The ratio between nuclear and cytoplasmic intensity was then calculated per cell.

5.6. Migration assay

As an indication for nuclear stiffness [32], hMSCs were allowed to migrate directly from films or ESP scaffolds through a Fluoroblok 24-well transwell with 3-μm or 8-μm pores. hMSCs were first cultured on 6-mm films or ESP scaffolds for 7 days in standard scaffold culture conditions (see ‘Cell culture’) and then transferred upside down on the transwell membrane. αMEM without serum was added to the top compartment and basic medium was added to the bottom compartment, to induce cell migration. A screw and nut were placed on top of the film or ESP scaffold as a weight to ensure proper contact between the substrate and transwell. After fixation (see ‘Immunofluorescence and imaging’), the film or ESP scaffold was removed, and the transwells were stained with Syto14 (Thermo Fisher Scientific). The whole transwell was imaged, top and bottom, and cells were quantified using particle analysis in Fiji.

5.7. Lentiviral production for shRNA delivery

To deliver shRNA for gene knock-down, we produced lentiviruses using TRC pLKO.1 constructs from Dharmacon. We used the following clone IDs for LMNA-shRNA 1 and 2: TRCN0000061833 and TRCN0000061836; ZYX-shRNA 1 and 2: TRCN0000074204 and TRCN0000074205; PXN-shRNA 1 and 2: TRCN0000123134 and TRCN0000123136; YAP-shRNA 1 and 21: TRCN0000107265 and TRCN0000107266. To produce the lentiviral particles, human embryonic kidney 293FT (HEK) cells were seeded at 60k cells/cm2 in a TCP dish in DMEM + 10% FBS. The next day, HEK cells were transfected with pMDLg pRRE, pMD2.G, pRSV Rev (Addgene) and one of the pLKO.1 plasmids containing the shRNA, using lipofectamine 2000 (Thermo Fisher Scientific) in ratio of 5:1 (μl:μg of DNA). After overnight incubation, the medium was changed to basic medium for hMSCs. Viral particles were harvested after 24 and 48 h, and filtered through a 0.45-μm filter.

The day before transduction, hMSCs were thawed at 1k cells/cm2 in a 10-cm dish. 3 mL of unconcentrated virus was added to the dish and incubated overnight. The medium was changed for basic medium the following day, and the medium was changed for basic medium +2 μg ml−1 puromycin at 48–72 h later. Cells were treated for 72 h with puromycin. At 9–10 days after initial thawing, cells were passaged at 1k cells/cm2 and cultured for 7 days in basic medium before protein harvest or fixation.

5.8. Statistical analysis

The number of biological replicates and repeated experiments are indicated in the figure captions, as well as the statistical test used. All individual films and scaffolds used in one experiment were randomly assigned to an experimental group. Cells that were imaged for quantification were also randomly picked. All data was tested for normal distribution using the Shapiro-Wilk test. To test the significance of relative expression data with one comparison, a two-tailed ratio t-test was performed, or the Mann-Whitney test as non-parametric equivalent. For relative expression data with multiple comparisons, the log of each value was used for a repeated measures ANOVA, with Tukey’s post hoc to compare individual groups. For non-relative comparisons, a One-way ANOVA with Tukey’s post hoc, or Two-way ANOVA with Sidak’s post-hoc was used, or the Kruskal-Wallis test with Dunn’s post hoc as non-parametric equivalent. Statistical significance was set at p < 0.05. Statistical tests were performed with GraphPad Prism 8.

Supplementary Material

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.biomaterials.2020.119854.

Fig S1-S11

Acknowledgements

We would like to thank Jos Broers for valuable discussions on the lamin A and C data. We are grateful to the European Research Council starting grant “Cell Hybridge” for financial support under the Horizon2020 framework program (Grant #637308). Some of the materials used in this work were provided by the Texas A&M Health Science Center College of Medicine Institute for Regenerative Medicine at Scott & White through a grant from NCRR of the NIH (Grant #P40RR017447). We are grateful to the European Community’s Seventh Framework Program (FP7/2007-2013) (grant agreement No. 305436, STELLAR).

References

  • [1].Wozniak MA, Modzelewska K, Kwong L, Keely PJ. Focal adhesion regulation of cell behavior. Biochim Biophys Acta. 2004;1692(2–3):103–119. doi: 10.1016/j.bbamcr.2004.04.007. [DOI] [PubMed] [Google Scholar]
  • [2].Stroud MJ. Linker of nucleoskeleton and cytoskeleton complex proteins in cardiomyopathy. Biophys Rev. 2018;10(4):1033–1051. doi: 10.1007/s12551-018-0431-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [3].Dahl KN, Kahn SM, Wilson KL, Discher DE. The nuclear envelope lamina network has elasticity and a compressibility limit suggestive of a molecular shock absorber. J Cell Sci. 2004;117(Pt 20):4779–4786. doi: 10.1242/jcs.01357. [DOI] [PubMed] [Google Scholar]
  • [4].Lammerding J, Fong LG, Ji JY, Reue K, Stewart CL, Young SG, Lee RT. Lamins A and C but not lamin B1 regulate nuclear mechanics. J Biol Chem. 2006;281(35):25768–25780. doi: 10.1074/jbc.M513511200. [DOI] [PubMed] [Google Scholar]
  • [5].Swift J, Ivanovska IL, Buxboim A, Harada T, Dingal PC, Pinter J, Pajerowski JD, Spinler KR, Shin JW, Tewari M, Rehfeldt F, et al. Nuclear lamin-A scales with tissue stiffness and enhances matrix-directed differentiation. Science. 2013;341(6149):1240104. doi: 10.1126/science.1240104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [6].Buxboim A, Irianto J, Swift J, Athirasala A, Shin JW, Rehfeldt F, Discher DE. Coordinated increase of nuclear tension and lamin-A with matrix stiffness out-competes lamin-B receptor that favors soft tissue phenotypes. Mol Biol Cell. 2017;28(23):3333–3348. doi: 10.1091/mbc.E17-06-0393. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [7].Al-Saaidi R, Bross P. Do lamin A and lamin C have unique roles? Chromosoma. 2015;124(1):1–12. doi: 10.1007/s00412-014-0484-7. [DOI] [PubMed] [Google Scholar]
  • [8].Weber K, Groeschel-Stewart U. Antibody to myosin: the specific visualization of myosin-containing filaments in nonmuscle cells. Proc Natl Acad Sci U S A. 1974;71(11):4561–4564. doi: 10.1073/pnas.71.11.4561. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [9].Yeh YC, Ling JY, Chen WC, Lin HH, Tang MJ. Mechanotransduction of matrix stiffness in regulation of focal adhesion size and number: reciprocal regulation of caveolin-1 and beta1 integrin. Sci Rep. 2017;7(1):15008. doi: 10.1038/s41598-017-14932-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [10].Pelham RJ, Jr, Wang Y. Cell locomotion and focal adhesions are regulated by substrate flexibility. Proc Natl Acad Sci U S A. 1997;94(25):13661–13665. doi: 10.1073/pnas.94.25.13661. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [11].Discher DE, Janmey P, Wang YL. Tissue cells feel and respond to the stiffness of their substrate. Science. 2005;310(5751):1139–1143. doi: 10.1126/science.1116995. [DOI] [PubMed] [Google Scholar]
  • [12].Prager-Khoutorsky M, Lichtenstein A, Krishnan R, Rajendran K, Mayo A, Kam Z, Geiger B, Bershadsky AD. Fibroblast polarization is a matrix-rigidity-dependent process controlled by focal adhesion mechanosensing. Nat Cell Biol. 2011;13(12):1457–1465. doi: 10.1038/ncb2370. [DOI] [PubMed] [Google Scholar]
  • [13].Tavares S, Vieira AF, Taubenberger AV, Araujo M, Martins NP, Bras-Pereira C, Polonia A, Herbig M, Barreto C, Otto O, Cardoso J, et al. Actin stress fiber organization promotes cell stiffening and proliferation of pre-invasive breast cancer cells. Nat Commun. 2017;8:15237. doi: 10.1038/ncomms15237. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [14].Ankam S, Teo BKK, Pohan G, Ho SWL, Lim CK, Yim EKF. Temporal changes in nucleus morphology, lamin A/C and histone methylation during nanotopography-induced neuronal differentiation of stem cells. Front Bioeng Biotechnol. 2018;6:69. doi: 10.3389/fbioe.2018.00069. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [15].Azatov M, Sun X, Suberi A, Fourkas JT, Upadhyaya A. Topography on a sub-cellular scale modulates cellular adhesions and actin stress fiber dynamics in tumor associated fibroblasts. Phys Biol. 2017;14(6):065003. doi: 10.1088/1478-3975/aa7acc. [DOI] [PubMed] [Google Scholar]
  • [16].Keselowsky BG, Collard DM, Garcia AJ. Surface chemistry modulates focal adhesion composition and signaling through changes in integrin binding. Biomaterials. 2004;25(28):5947–5954. doi: 10.1016/j.biomaterials.2004.01.062. [DOI] [PubMed] [Google Scholar]
  • [17].Dupont S, Morsut L, Aragona M, Enzo E, Giulitti S, Cordenonsi M, Zanconato F, Le Digabel J, Forcato M, Bicciato S, Elvassore N, et al. Role of YAP/TAZ in mechanotransduction. Nature. 2011;474(7350):179–183. doi: 10.1038/nature10137. [DOI] [PubMed] [Google Scholar]
  • [18].Engler AJ, Sen S, Sweeney HL, Discher DE. Matrix elasticity directs stem cell lineage specification. Cell. 2006;126(4):677–689. doi: 10.1016/j.cell.2006.06.044. [DOI] [PubMed] [Google Scholar]
  • [19].Han I, Kwon BS, Park HK, Kim KS. Differentiation potential of mesenchymal stem cells is related to their intrinsic mechanical properties. Int Neurourol J. 2017;21(Suppl 1):S24–S31. doi: 10.5213/inj.1734856.428. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [20].McBeath R, Pirone DM, Nelson CM, Bhadriraju K, Chen CS. Cell shape, cytoskeletal tension, and RhoA regulate stem cell lineage commitment. Dev Cell. 2004;6(4):483–495. doi: 10.1016/s1534-5807(04)00075-9. [DOI] [PubMed] [Google Scholar]
  • [21].Park JS, Chu JS, Tsou AD, Diop R, Tang Z, Wang A, Li S. The effect of matrix stiffness on the differentiation of mesenchymal stem cells in response to TGF-beta. Biomaterials. 2011;32(16):3921–3930. doi: 10.1016/j.biomaterials.2011.02.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [22].Sun M, Chi G, Li P, Lv S, Xu J, Xu Z, Xia Y, Tan Y, Xu J, Li L, Li Y. Effects of matrix stiffness on the morphology, adhesion, proliferation and osteogenic differentiation of mesenchymal stem cells. Int J Med Sci. 2018;15(3):257–268. doi: 10.7150/ijms.21620. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [23].Mathieu PS, Loboa EG. Cytoskeletal and focal adhesion influences on mesenchymal stem cell shape, mechanical properties, and differentiation down osteogenic, adipogenic, and chondrogenic pathways. Tissue Eng B Rev. 2012;18(6):436–444. doi: 10.1089/ten.teb.2012.0014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [24].Zhang B, Yang Y, Keyimu R, Hao J, Zhao Z, Ye R. The role of lamin A/C in mesenchymal stem cell differentiation. J Physiol Biochem. 2019;75(1):11–18. doi: 10.1007/s13105-019-00661-z. [DOI] [PubMed] [Google Scholar]
  • [25].Tang Y, Feinberg T, Keller ET, Li XY, Weiss SJ. Snail/Slug binding interactions with YAP/TAZ control skeletal stem cell self-renewal and differentiation. Nat Cell Biol. 2016;18(9):917–929. doi: 10.1038/ncb3394. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [26].Tang Y, Rowe RG, Botvinick EL, Kurup A, Putnam AJ, Seiki M, Weaver VM, Keller ET, Goldstein S, Dai J, Begun D, et al. MT1-MMP-dependent control of skeletal stem cell commitment via a beta1-integrin/YAP/TAZ signaling axis. Dev Cell. 2013;25(4):402–416. doi: 10.1016/j.devcel.2013.04.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [27].Fraley SI, Feng Y, Krishnamurthy R, Kim DH, Celedon A, Longmore GD, Wirtz D. A distinctive role for focal adhesion proteins in three-dimensional cell motility. Nat Cell Biol. 2010;12(6):598–604. doi: 10.1038/ncb2062. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [28].Moroni L, de Wijn JR, van Blitterswijk CA. 3D fiber-deposited scaffolds for tissue engineering: influence of pores geometry and architecture on dynamic mechanical properties. Biomaterials. 2006;27(7):974–985. doi: 10.1016/j.biomaterials.2005.07.023. [DOI] [PubMed] [Google Scholar]
  • [29].Gupta M, Sarangi BR, Deschamps J, Nematbakhsh Y, Callan-Jones A, Margadant F, Mege RM, Lim CT, Voituriez R, Ladoux B. Adaptive rheology and ordering of cell cytoskeleton govern matrix rigidity sensing. Nat Commun. 2015;6:7525. doi: 10.1038/ncomms8525. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [30].Buxboim A, Swift J, Irianto J, Spinler KR, Dingal PC, Athirasala A, Kao YR, Cho S, Harada T, Shin JW, Discher DE. Matrix elasticity regulates lamin-A,C phosphorylation and turnover with feedback to actomyosin. Curr Biol. 2014;24(16):1909–1917. doi: 10.1016/j.cub.2014.07.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [31].Kochin V, Shimi T, Torvaldson E, Adam SA, Goldman A, Pack CG, Melo-Cardenas J, Imanishi SY, Goldman RD, Eriksson JE. Interphase phosphorylation of lamin A. J Cell Sci. 2014;127(Pt 12):2683–2696. doi: 10.1242/jcs.141820. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [32].Harada T, Swift J, Irianto J, Shin JW, Spinler KR, Athirasala A, Diegmiller R, Dingal PC, Ivanovska IL, Discher DE. Nuclear lamin stiffness is a barrier to 3D migration, but softness can limit survival. J Cell Biol. 2014;204(5):669–682. doi: 10.1083/jcb.201308029. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [33].Kim JK, Louhghalam A, Lee G, Schafer BW, Wirtz D, Kim DH. Nuclear lamin A/C harnesses the perinuclear apical actin cables to protect nuclear morphology. Nat Commun. 2017;8(1):2123. doi: 10.1038/s41467-017-02217-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [34].Kim DH, Li B, Si F, Phillip JM, Wirtz D, Sun SX. Volume regulation and shape bifurcation in the cell nucleus. J Cell Sci. 2015;128(18):3375–3385. doi: 10.1242/jcs.166330. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [35].Elosegui-Artola A, Andreu I, Beedle AEM, Lezamiz A, Uroz M, Kosmalska AJ, Oria R, Kechagia JZ, Rico-Lastres P, Le Roux AL, Shanahan CM, et al. Force triggers YAP nuclear entry by regulating transport across nuclear pores. Cell. 2017;171(6):1397–1410.:e14. doi: 10.1016/j.cell.2017.10.008. [DOI] [PubMed] [Google Scholar]
  • [36].Vicente-Manzanares M, Ma X, Adelstein RS, Horwitz AR. Non-muscle myosin II takes centre stage in cell adhesion and migration. Nat Rev Mol Cell Biol. 2009;10(11):778–790. doi: 10.1038/nrm2786. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [37].Swift J, Discher DE. The nuclear lamina is mechano-responsive to ECM elasticity in mature tissue. J Cell Sci. 2014;127(Pt 14):3005–3015. doi: 10.1242/jcs.149203. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [38].Griffith E, Coutts AS, Black DM. RNAi knockdown of the focal adhesion protein TES reveals its role in actin stress fibre organisation. Cell Motil Cytoskeleton. 2005;60(3):140–152. doi: 10.1002/cm.20052. [DOI] [PubMed] [Google Scholar]
  • [39].Smith MA, Blankman E, Gardel ML, Luettjohann L, Waterman CM, Beckerle MC. A zyxin-mediated mechanism for actin stress fiber maintenance and repair. Dev Cell. 2010;19(3):365–376. doi: 10.1016/j.devcel.2010.08.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [40].Sun Z, Huang S, Li Z, Meininger GA. Zyxin is involved in regulation of mechanotransduction in arteriole smooth muscle cells. Front Physiol. 2012;3:472. doi: 10.3389/fphys.2012.00472. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [41].Deakin NO, Turner CE. Paxillin comes of age. J Cell Sci. 2008;121(Pt 15):2435–2444. doi: 10.1242/jcs.018044. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [42].Fu J, Wang YK, Yang MT, Desai RA, Yu X, Liu Z, Chen CS. Mechanical regulation of cell function with geometrically modulated elastomeric substrates. Nat Methods. 2010;7(9):733–736. doi: 10.1038/nmeth.1487. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [43].Pan JX, Xiong L, Zhao K, Zeng P, Wang B, Tang FL, Sun D, Guo HH, Yang X, Cui S, Xia WF, et al. YAP promotes osteogenesis and suppresses adipogenic differentiation by regulating beta-catenin signaling. Bone Res. 2018;6:18. doi: 10.1038/s41413-018-0018-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [44].Sen B, Uzer G, Samsonraj RM, Xie Z, McGrath C, Styner M, Dudakovic A, van Wijnen AJ, Rubin J. Intranuclear actin structure modulates mesenchymal stem cell differentiation. Stem Cell. 2017;35(6):1624–1635. doi: 10.1002/stem.2617. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [45].Yang C, Tibbitt MW, Basta L, Anseth KS. Mechanical memory and dosing influence stem cell fate. Nat Mater. 2014;13(6):645–652. doi: 10.1038/nmat3889. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [46].Cukierman E, Pankov R, Stevens DR, Yamada KM. Taking cell-matrix adhesions to the third dimension. Science. 2001;294(5547):1708–1712. doi: 10.1126/science.1064829. [DOI] [PubMed] [Google Scholar]
  • [47].Werner M, Blanquer SB, Haimi SP, Korus G, Dunlop JW, Duda GN, Grijpma DW, Petersen A. Surface curvature differentially regulates stem cell migration and differentiation via altered attachment morphology and nuclear deformation. Adv Sci. 2017;4(2):1600347. doi: 10.1002/advs.201600347. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [48].Ho CY, Jaalouk DE, Vartiainen MK, Lammerding J. Lamin A/C and emerin regulate MKL1-SRF activity by modulating actin dynamics. Nature. 2013;497(7450):507–511. doi: 10.1038/nature12105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [49].Yang Y, Wang X, Wang Y, Hu X, Kawazoe N, Yang Y, Chen G. Influence of cell spreading area on the osteogenic commitment and phenotype maintenance of mesenchymal stem cells. Sci Rep. 2019;9(1):6891. doi: 10.1038/s41598-019-43362-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [50].Chaudhuri O, Gu L, Klumpers D, Darnell M, Bencherif SA, Weaver JC, Huebsch N, Lee HP, Lippens E, Duda GN, Mooney DJ. Hydrogels with tunable stress relaxation regulate stem cell fate and activity. Nat Mater. 2016;15(3):326–334. doi: 10.1038/nmat4489. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [51].Khetan S, Guvendiren M, Legant WR, Cohen DM, Chen CS, Burdick JA. Degradation-mediated cellular traction directs stem cell fate in covalently crosslinked three-dimensional hydrogels. Nat Mater. 2013;12(5):458–465. doi: 10.1038/nmat3586. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [52].Huebsch N, Arany PR, Mao AS, Shvartsman D, Ali OA, Bencherif SA, Rivera-Feliciano J, Mooney DJ. Harnessing traction-mediated manipulation of the cell/matrix interface to control stem-cell fate. Nat Mater. 2010;9(6):518–526. doi: 10.1038/nmat2732. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [53].Digirolamo CM, Stokes D, Colter D, Phinney DG, Class R, Prockop DJ. Propagation and senescence of human marrow stromal cells in culture: a simple colony-forming assay identifies samples with the greatest potential to propagate and differentiate. Br J Haematol. 1999;107(2):275–281. doi: 10.1046/j.1365-2141.1999.01715.x. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Fig S1-S11
EMS214271-supplement-Fig_S1_S11.docx (122.6MB, docx)

RESOURCES