Bio-activating ultrafine grain titanium: RNA sequencing reveals enhanced mechano-activation of osteoconduction on nanostructured substrates

Abstract

Titanium is essentially absent from biological systems yet reliably integrates into bone. To achieve osseointegration, titanium must activate biological processes without entering cells, defining it as a bio-activating material. Nanostructuring bulk titanium reduces grain size, increases strength, and improves other quantifiable physical properties, including cytocompatibility. The biological processes activated by increasing grain boundary availability were detected with total RNA-sequencing in mouse pre-osteoblasts grown for 72 hours on nanometrically smooth substrates of either coarse grain or nanostructured ultrafine grain titanium. The average grain boundary length under cells on the conventional coarse grain substrates is 273.0 μm, compared to 70,881.5 μm for cells adhered to the nanostructured ultrafine grain substrates; a 260-fold difference. Cells on both substrates exhibit similar expression profiles for genes whose products are critical for mechanosensation and transduction of cues that trigger osteoconduction. Biological process Gene Ontology term enrichment analysis of differentially expressed genes reveals that cell cycle, chromatin modification, telomere maintenance, and RNA metabolism processes are upregulated on ultrafine grain titanium. Processes related to immune response, including apoptosis, are downregulated. Tumor-suppressor genes are upregulated while tumor-promoting genes are downregulated. Upregulation of genes involved in chromatin remodeling and downregulation of genes under the control of the peripheral circadian clock implicate both processes in the transduction of mechanosensory information. Non-coding RNAs may also play a role in the response. Merging transcriptomics with well-established mechanobiology principles generates a unified model to explain the bio-activating properties of titanium. The modulation of processes is accomplished through chromatin remodeling in which the nucleus responds like a rheostat to grain boundary concentration. This convergence of biological and materials science reveals a pathway toward understanding the biotic-abiotic interface and will inform the development of effective bio-activating and bio-inactivating materials.

Fig 1. Bioinformatic workflow.

Workflow used to generate network analysis. Kits, software and instrumentation used to generate data for each step are shown in parentheses.

Fig 2. Zipf curve.

Dark grey markers are those with adequate coverage to use in comparative analyses, genes shown in light grey were disregarded. Red font indicates genes used as indicators of cell identity. S1 Fig is the same curve annotated with differentially expressed and those of interest in mechanotransduction and osteogenesis.

Fig 3. qPCR results.

Expression levels of the five genes were normalized using Actb levels, which was not differentially expressed in the dataset. The 2^-ΔΔCt method was used to calculate the differences and included three replicates of each sample. Those with asterisks are included in Fig 4.

Fig 4. Biological process gene ontology terms enrichment and network analysis.

Of the 432 differentially expressed genes with p≤0.05, 117 are retained in this network. Diamond-shaped nodes are GO biological process terms, larger diamonds indicate higher levels of significance, p(adj) value for each GO node is shown. Round nodes are genes, blue nodes are downregulated and red are upregulated. Bolded gene abbreviations in the figure are discussed in the text and full gene names are given in Table 4. S3 Table in S1 File provides this information for all 117 genes.

Fig 5. Mechanosensation and transduction model for bio-activating materials.

5A. This model focuses on a single physical characteristic of Ti, the R_GBL/cell. 5B. First contact involves the extracellular matrix (ECM) domain of integrin α (Itgα) with the substrate, which includes a 5 nm layer of TiO₂. Contact causes the association with an integrin ß chain (Itg ß), setting off a complex cascade of protein activation in the cytoplasm. 5C. Mechanosensation transmits the force through focal adhesions (FA) and other cytoplasmic protein complexes. Eventually, the signal reaches the linker of nucleoskeleton and cytoskeleton (LINC) complex embedded in the nuclear envelope (NE). 5D. Chromatin remodeling response involves the differential condensation and opening of chromatin. 5E. RNA transcripts are processed depending upon their function and, if required, transported out of the nucleus through nuclear pores (NP). 5F. Inside-out signal transduction includes regulation by the RNA interference (RNAi), transport to a ribosome, passage through the ER (endoplasmic reticulum) and Golgi, any post-translational modifications, and transport to the appropriate compartment. 5G. Integration of the Ti into bone requires the modification of collagen molecules in the ECM in concert with the maturation of FAs. 5H. The response of the substrate and subsequent cell are topics requiring additional research. Not shown in this model is the role of primary cilia.