Nanomechanically Tunable Biomaterials for Directed Musculoskeletal Tissue Differentiation

Introduction: Biomaterials-based tissue engineering approaches have utilized silk fibroin (SF) from the Bombyx mori silkworm because of its biocompatibility and tunable mechanical properties1,2. Despite their excellent versatility and cytocompatibility, a known issue with silk materials is that identical cell types cultured within supposedly identical silk constructs and receiving the same biochemical cues from the scaffolds and the media can behave in disparate and unpredictable ways. Although many studies report similar protocols in synthesizing these silk biomaterials, there are often discrepancies in reported Young’s modulus (YM) values2,3. These discrepancies allude to the fact that cells within SF constructs are experiencing different mechanical signals than those obtained from the traditionally used macroscale testing modalities: compression and tensile testing. Furthermore, our preliminary studies have highlighted significant differences in the YM obtained at the macroscale and nanoscale of various SF biomaterials, using compression testing and atomic force microscopy (AFM) respectively4. This finding suggests the importance of understanding biomaterial mechanical properties at scales more relevant to the cell-length to confidently predict cell behavioral outcomes. To better understand the differences between macro and nanoscale mechanical properties of SF biomaterials, and how they correlate to embedded cellular behavior, 3% SF, 4% SF, and 0.3% tyramine addition 2.7% SF (TA-SF) hydrogels were synthesized and analyzed using compression and AFM, respectively. We hypothesized that the nanomechanical properties of the SF hydrogels would not only differ from the macromechanical properties as previously established, but also that differences in nanoscale properties would more accurately describe changes in cell behavior.

Materials and

Methods: SF for gels was prepared from Bombyx mori cocoons using an established and widely standardized protocol3. TASF was prepared using tyramine hydrochloride as previously described5. Cellular constructs were seeded with 1.5x106 C2C12 mouse myoblasts. All constructs were incubated at 5% CO2, 37°C, in 10% fetal bovine serum (FBS), 1% Antibiotic/Antimycotic (AA) Dulbecco’s Modified Eagle’s Medium (DMEM) for 1, 2, and 3 weeks before testing. AFM imaging was done in ambient conditions using Bruker’s Dimension FastScan Bio SPM with a 64-bit Nanoscope V Controller utilized in combination with PeakForce-HiRs-F-B probes. Macromechanical properties were assessed using unconfined compression testing with a custom loading frame equipped with a 1N load cell where all samples will be preloaded to 0.02N and brought to 30% strain at 1% strain•sec-16. Differences between YM values obtained from AFM and compression testing were compared using an Unpaired Student’s T-test and comparisons of YM between constructs within each modality were made using a Paired Student’s T-test. Confidence intervals were set to 95%. Brightfield images were obtained using a Zeiss Axiovert 5 to assess cellular morphology. C2C12 differential behavior was assessed using WB analysis of myogenic (MyoD), chondrogenic (Sox-9, Col II), tenogenic (Scleraxis), and osteogenic (BMP-2) protein secretion markers at each respective timepoint. Lastly, predictability of the trend in change of YM overtime was assessed in each construct by deriving an R-squared value from a linear regression model of the average YM over the three timepoints for both AFM and compression testing.

Results, Conclusions, and Discussions: At all timepoints, the 3%SF, 4%SF, and TA-SF acellular constructs had significant increases in YM obtained from AFM, compared to compression testing YM. The same was true for cellular constructs except for W2 and W3 3%SF and W3 4%SF comparisons, where no significant differences were found (Figure 1). AFM showed further significant differences in YM between paired timepoints for SF constructs compared to compression testing (Figure 2). Cells in different percentage SF gels had distinct morphologies. WB analysis showed gradual increased production of BMP-2 from W1, to W2, to W3 for 4%SF and TA-SF (Figure 4). Lastly, predictability of YM changes over time was highest in TA-SF constructs, with an average R-squared value of 0.89, compared to 3%SF and 4%SF (R-squared values 0.55 and 0.58, respectively) (Figure 3).
YM obtained from AFM was significantly higher than macroscale YM in most comparisons, suggesting that cells receive differing nanoscale mechanical ques that are not captured by commonly used testing methods, and currently unexploited for directing cell behavior. Observed morphological differences of cells embedded in each SF construct suggest cell behavior responded to the nanoscale properties. Early increases in BMP-2, combined with observed stiffening, suggests an early osteogenic phenotype in the higher YM materials, consistent with prior studies7. TA-SF had the highest predictability in YM change over time, especially in the AFM obtained data, showing that tyramine addition to silk controls both nano- and macromechanical hydrogel properties in a predictable manner for selectively directing mechanotransducive cellular differentiation.
Results of this study highlight differences that exist in the macromechanical and nanomechanical properties of the observed biomaterials. These findings underscore the importance of assessing mechanical properties of biomaterials at the cell-length scale to accurately predict embedded cell behavior. The increased predictability of the TA-SF constructs mechanical changes over time, particularly at the nanoscale level, suggests that tyramine addition is a viable method to control such properties. Understanding how nanomechanical properties can affect cellular behavior, and can be selectively and predictably tuned, would allow for biomaterials with highly tailored mechanical gradients to be synthesized in aims to steer cellular differentiation towards the regeneration of complex tissues.

Acknowledgements (Optional): Acknowledgements: We would like to thank the Surface Science Lab for assistance with obtaining and interpreting the nanomechanical measurements, and the Northwest Tissue Mechanics Lab for the use of their custom load frame. We acknowledge support from the Institutional Development Awards (IDeA) from the National Institute of General Medical Sciences of the National Institutes of Health under Grants #P20GM103408, P20GM109095, and 1C06RR020533. We also acknowledge support from The Biomolecular Research Center at Boise State, BSU-Biomolecular Research Center, RRID:SCR_019174, with funding from the National Science Foundation, Grants #0619793 and #0923535; the M. J. Murdock Charitable Trust; Lori and Duane Stueckle, and the Idaho State Board of Education.

References:
1. Rockwood, D. N. et al. Materials fabrication from Bombyx mori silk fibroin. Nat. Protoc. 6, 1612–1631 (2011).
2. Vepari, C. & Kaplan, D. L. Silk as a biomaterial. Prog. Polym. Sci. 32, 991–1007 (2007).
3. Rockwood, D. N. et al. Materials fabrication from Bombyx mori silk fibroin. Nat Protoc 6, (2011).
4. Jones, C. et al. Defining the Macro- and Nanomechanics of Mechanically Tunable Biomaterials to Improve Control of Cell Behavior. in (Seattle, WA).
5. Ziadlou, R. et al. Optimization of hyaluronic acid-tyramine/silk-fibroin composite hydrogels for cartilage tissue engineering and delivery of anti-inflammatory and anabolic drugs. Mater. Sci. Eng. C 120, 111701 (2021).
6. Semodji, A. M. et al. A novel strain energy-based method to dynamically stimulate three-dimensional cellular constructs. in (Vail, Colorado, 2023).
7. Engler, A. J., Sen, S., Sweeney, H. L. & Discher, D. E. Matrix Elasticity Directs Stem Cell Lineage Specification. Cell 126, 677–689 (2006).