Cationic Nanogel-mediated Runx2 and Osterix siRNA Delivery Decreases Mineralization in MC3T3 Cells
- 872 Downloads
Heterotopic ossification (HO) may occur after musculoskeletal trauma, traumatic brain injury, and total joint arthroplasty. As such, HO is a compelling clinical concern in both military and civilian medicine. A possible etiology of HO involves dysregulated signals in the bone morphogenetic protein osteogenic cascade. Contemporary treatment options for HO (ie, nonsteroidal antiinflammatory drugs and radiation therapy) have adverse effects associated with their use and are not biologically engineered to abrogate the molecular mechanisms that govern osteogenic differentiation.
We hypothesized that (1) nanogel-mediated short interfering RNA (siRNA) delivery against Runt-related transcription factor 2 (Runx2) and osterix (Osx) genes will decrease messenger RNA expression; (2) inhibit activity of the osteogenic marker alkaline phosphatase (ALP); and (3) inhibit hydroxyapatite (HA) deposition in osteoblast cell cultures.
Nanogel nanostructured polymers delivered siRNA in 48-hour treatment cycles against master osteogenic regulators, Runx2 and Osx, in murine calvarial preosteoblasts (MC3T3-E1.4) stimulated for osteogenic differentiation by recombinant human bone morphogenetic protein (rhBMP-2). The efficacy of RNA interference (RNAi) therapeutics was determined by quantitation of messenger RNA knockdown (by quantitative reverse transcription-polymerase chain reaction), downstream protein knockdown (determined ALP enzymatic activity assay), and HA deposition (determined by OsteoImage™ assay).
Gene expression assays demonstrated that nanogel-based RNAi treatments at 1:1 and 5:1 nanogel:short interfering RNA weight ratios reduced Runx2 expression by 48.59% ± 19.53% (p < 0.001) and 43.22% ± 18.01% (both p < 0.001). The same 1:1 and 5:1 treatments against both Runx2 and Osx reduced expression of Osx by 51.65% ± 10.85% and 47.65% ± 9.80% (both p < 0.001). Moreover, repeated 48-hour RNAi treatment cycles against Runx2 and Osx rhBMP-2 administration reduced ALP activity after 4 and 7 days. ALP reductions after 4 days in culture by nanogel 5:1 and 10:1 RNAi treatments were 32.4% ± 12.0% and 33.6% ± 13.8% (both p < 0.001). After 7 days in culture, nanogel 1:1 and 5:1 RNAi treatments produced 35.9% ± 14.0% and 47.7% ± 3.2% reductions in ALP activity. Osteoblast mineralization data after 21 days suggested that nanogel 1:1, 5:1, and 10:1 RNAi treatments decreased mineralization (ie, HA deposition) from cultures treated only with rhBMP-2 (p < 0.001). However, despite RNAi attack on Runx2 and Osx, HA deposition levels remained greater than non-rhBMP-2-treated cell cultures.
Although mRNA and protein knockdown were confirmed as a result of RNAi treatments against Runx2 and Osx, complete elimination of mineralization processes was not achieved. RNAi targeting mid- and late-stage osteoblast differentiation markers such as ALP, osteocalcin, osteopontin, and bone sialoprotein) may produce the desired RNAi-nanogel nanostructured polymer HO prophylaxis.
Successful HO prophylaxis should target and silence osteogenic markers critical for heterotopic bone formation processes. The identification of such markers, beyond RUNX2 and OSX, may enhance the effectiveness of RNAi prophylaxes for HO.
KeywordsOsteogenic Differentiation Heterotopic Ossification siRNA Delivery RNAi Treatment MC3T3 Cell
Heterotopic ossification (HO) may occur as a consequence of musculoskeletal trauma from blast and high-energy injuries [2, 21, 40, 52], total joint arthroplasty (TJA) , traumatic brain injury , or spinal cord injury [4, 33, 55, 61]. The incidence rate for HO varies from 15% to 90%, depending on the source of trauma and the grading criteria used [1, 7, 39, 43, 51]. Current treatments for patients with HO emphasize reducing inflammation and/or inactivating tissue-resident stem cells at the trauma site to minimize pathological bone formation. These treatments options include nonsteroidal antiinflammatory drugs (NSAIDs) and radiotherapy (often in combination), although adverse effects have been associated with their use [8, 9, 20, 57]. Surgical excision of HO is another therapeutic option, although recurrence of HO at the surgical site may ensue [10, 13, 65]. Furthermore, contemporary treatment options are not engineered specifically to impede the biological mechanisms responsible for the production of bone in soft tissue. We seek to design a therapeutic that may overcome the limitations of the current generation of treatments and prophylaxes for HO.
In designing a prophylaxis for HO, we emphasized interference with a potent pathway for bone formation and regeneration—the bone morphogenetic protein (BMP)-induced osteogenic signaling cascade. Dysregulation of BMP signaling stimulates osteogenic lineage progression of striated muscle mesenchymal stem cells [14, 56]. Furthermore, a possible HO mechanism may include the inflammatory factor cyclooxygenase-2 (COX-2) and the canonical BMP-based bone-forming pathways [41, 58]. COX-2 expression stimulates prostaglandin E2 (PGE-2) expression in osteoblasts [3, 27]; the PGE-2 and BMP signaling pathways converge at the transcription factor runt-related protein 2 (RUNX2) [37, 38]. RUNX2 and osterix (OSX) are master regulators of osteogenesis and promote expression of alkaline phosphatase (ALP), osteocalcin (OCN), osteopontin (OPN), and osteonectin (ON) [11, 22, 54]. These molecular cues promote the osteogenic cascade and clinical manifestation of HO.
Therefore, we postulated that silencing Runx2 and Osx expression may be a compelling strategy to prevent osteogenic differentiation and potentially, trauma-induced HO. Furthermore, silencing key molecular factors may be accomplished by ribonucleic acid interference (RNAi) mediated by nonviral vectors [19, 24, 34, 42]. Key intermediaries in the RNAi process are short interfering RNAs (siRNA), which facilitate posttranscriptional silencing of target genes. We previously reported that nanostructured polymers (NSPs) are safe and efficient carriers for the delivery of siRNA in mammalian cell cultures and mouse models .
In our current study, we hypothesized that (1) nanogel-mediated siRNA delivery against Runx2 and Osx genes will decrease messenger RNA (mRNA) expression; (2) inhibit activity of the osteogenic marker ALP; and (3) inhibit hydroxyapatite (HA) deposition in osteoblast cell cultures.
Materials and Methods
Cationic nanogels were prepared by activators generated by electron transfer atom transfer radical polymerization [5, 35, 36, 45, 46, 47, 48, 49, 64] in inverse miniemulsion by copolymerizing quaternized dimethyl aminoethylmethacrylate, oligo(ethylene oxide) methacrylate (Mn = 300), and a water-soluble disulfide methacrylate crosslinker with a poly(ethylene glycol 2-bromoisobutyrate) initiator and a copper bromide tris(2-[dimethylamino]ethyl)amine catalyst system dissolved in water. The inverse miniemulsion was prepared by ultrasonication of the aqueous phase in a cyclohexane Span80 solution. After the reaction mixture was degassed, an ascorbic acid solution was injected to generate the active catalyst. Fluorescent nanogels were synthesized by addition of rhodamine methacrylate during synthesis. The nanogels were purified using dialysis and characterized using dynamic light scattering and zeta potential analysis. Further details on nanogel synthesis and characterization can be found in Averick et al. .
Nanogel Polyplex Preparation
Nanogel NSPs were complexed with siRNA by addition of requisite amounts in nuclease-free water (Life Technologies, Thermo Fisher Scientific Inc, Waltham, MA, USA). SiRNA in solution (stored at −20° C at 200 µM) were thawed and added to nanogel NSPs in solution (25 µg/µL) at the desired NSP:siRNA weight ratio and to a dose of 30 pmol siRNA/5 µL treatment. Treatments were prepared in bulk and stored at −20° C between uses.
MC3T3-E1.4 murine calvarial preosteoblasts were acquired from American Type Culture Collection (Manassas, VA, USA). Cells (initially deep frozen) were thawed in a 37° C water bath. Basal \( \alpha \)-minimum essential media (\( \alpha \)MEM) was supplemented with 10% heat-inactivated fetal bovine serum and 1% penicillin/streptomycin to formulate complete \( \alpha \)MEM. The preosteoblasts were cultured in complete \( \alpha \)MEM at 37° C and 5% CO2. Cells were passaged using 0.25% trypsin-ethylenediaminetetraacetic acid (EDTA) once they had achieved 70% to 80% confluency and reseeded at a 1:8 subcultivation ratio.
The siRNA delivery capabilities of nanogel NSPs were validated visually by fluorescently tagging components of our RNAi therapeutics. MC3T3 cells were seeded at a density of 25,000 cells/mL on glass coverslips in 12-well cell-culture plates (each treatment group was performed in triplicate, N = 3). Rhodamine nanogels were prepared and delivered to the MC3T3 cells at concentrations of 200 µg/mL. To visually confirm the delivery of a siRNA payload, cyanine 3 (Cy3)-labeled siRNA were complexed with nonfluorescent nanogels at a 1:1 nanogel:siRNA weight ratio and delivered to cells (siRNA dose: 20 nmol). Cells were rinsed with 1 × phosphate-buffered saline (PBS) after 4 hours and fixed with 4% paraformaldehyde with Triton-X 100. Alexa Fluor 633 Phalloidin and DAPI stains were used to visualize the actin cytoskeleton and nuclei, respectively. Cells mounted in Prolong® Gold Antifade Reagent (Life Technologies, Thermo Fisher Scientific Inc, Waltham, MA, USA) were imaged at × 63 in oil under a Zeiss LSM 700 laser scanning confocal microscope (Carl Zeiss Microscopy GmbH, Jena, Germany).
Runx2 and Osx Gene Expression
The CellsDirect™ One-Step qRT-PCR Kit with ROX (Life Technologies, Thermo Fisher Scientific Inc, Catalog No. 11754) and an ABI Prism 7000 Sequence Detection System (Applied Biosystems, Foster City, CA, USA) were used for the analysis of Runx2 and Osx mRNA expression in MC3T3 cells. MC3T3 cells were plated at 25,000 cells/mL in 96-well plates at 0.2 mL per well. Runx2 (Catalog No. 4390771) and Osx (Catalog No. 4390771) Silencer Select siRNAs were acquired from Life Technologies (Waltham, MA, USA). SiRNA against Runx2 were delivered at 0 hours; siRNA against Osx were delivered at 24 hours along with recombinant human bone morphogenetic protein-2 (rhBMP-2) stimuli at a concentration of 100 ng/mL. This protocol for osteogenic differentiation and knockdown of Runx2 and Osx was previously validated in MC3T3 cultures . Nanogel:siRNA ratios of 1:1 (NG1) and 5:1 (NG5) were used for siRNA delivery. Lipofectamine® RNAiMAX (Life Technologies, Thermo Fisher Scientific Inc, Catalog No. 13778) (abbreviated as LRM) was used as a positive control for siRNA delivery. Target gene expression was evaluated 48 hours after Osx siRNA and rhBMP-2 delivery. Cell lysis, RNA extraction, and complementary DNA (cDNA) preparation were carried out according to the CellsDirect One-Step qRT-PCR Kit protocol. Briefly, each reaction contained SuperScript III RT/Platinum Taq Mix, 2 × Reaction Mix with ROX, Taqman Gene Expression Assay predesigned primers, and 4 μL of processed cell lysates. Expression of Runx2 and Osx were normalized to Actb expression using the comparative CT method. Fold change data were represented as a percent (normalized to rhBMP-2-treated groups) and were reported as mean ± SD with n = 8 (number of culture wells per treatment).
Alkaline Phosphatase Activity and Cell Viability
Silencing Runx2 and Osx through the osteogenic signaling cascade was evaluated by quantifying ALP activity. To extend assay duration beyond 48 hours of rhBMP-2 delivery, experimental groups received RNAi treatments in a repeated 48-hour cycle with Runx2 siRNA delivery occurring at 0 hours and Osx siRNA + rhBMP-2 delivery (100 ng/mL) occurring at 24 hours. Forty-eight-hour treatment cycles were repeated over the course of 4 and 7 days with nanogel NSPs delivering siRNA at 1:1 (NG1), 5:1 (NG5), and 10:1 (NG10) ratios. In 4-day studies, 5:1 and 10:1 NSP:siRNA ratios were evaluated. In 7-day studies, based on the statistically equivalent performance of NG5 and NG10 treatments at 4 days, 1:1 and 5:1 ratios were examined. All groups (except negative controls) received rhBMP-2. Furthermore, scrambled siRNA sequences were delivered as a control to reveal nonspecific gene silencing. Lipofectamine RNAiMAX was used as a positive control for siRNA delivery, defining the positive control as a treatment producing an expected reduction of ALP activity. ALP activity was assessed as previously detailed . Briefly, at experimental endpoints, cells were rinsed with 1 × PBS and lysed with 1 × cell lysis buffer. For cell viability, 100 µL cell lysate were added to a 100 µL solution of Quant-iT™ PicoGreen® dye (1:200 dilution) (Life Technologies, Thermo Fisher Scientific Inc, Waltham, MA, USA). After 2 minutes, DNA content was measured by fluorescence at 490 nm excitation and 535-nm emission wavelengths using a Tecan SpectraFluor (Tecan, Maennedorf, Switzerland). Data were normalized to rhBMP-2-treated groups and expressed as mean ± SD (n = 5). Samples for ALP activity were incubated with p-nitrophenol phosphate (Sigma-Aldrich, St. Louis, MO, USA, Catalog No. N2770), which is dephosphyorylated by ALP to produce a colorimetric change detectable by absorbance at 405 nm. P-nitrophenol standards (Sigma-Aldrich, Catalog No. N7660) were used to produce standard curves. Data were normalized to DNA content (by PicoGreen®) and reported as mean ± SD (n = 5) with two technical repeats.
Mineralization in Osteoblast Cultures
The quantitation of HA deposition was performed at 7, 14, and 21 days in MC3T3 cultures by the OsteoImage™ assay (Lonza, Basel, Switzerland, Catalog No. PA-1503). MC3T3 cells were seeded at 25,000 cells/mL in black, 96-well culture plates. Nanogel:siRNA complexes were prepared with both Runx2 and Osx siRNA at 1:1, 5:1, and 10:1 ratios. Specific groups were treated with 48-hour cycles of RNAi treatments against Runx2 and Osx (previously described). Lipofectamine RNAiMAX was used as a control for siRNA delivery. Osteogenic differentiation was induced by delivery of rhBMP-2 (100 ng/mL), and cultures were carried out for 7, 14, and 21 days. The OsteoImage™ (Lonza) assay was carried out per the manufacturer’s protocol. Briefly, cells were fixed in 70% ethanol for 20 minutes at room temperature, rinsed with 1 × PBS, and rinsed twice with the wash buffer provided. Samples were stained with diluted staining solution and incubated at room temperature for 30 minutes under protection from light. after incubation, cells were rinsed with wash buffer and measurements were taken at 492 nm excitation and 520-nm emission wavelengths using a Tecan SpectraFluor plate reader. Data were reported as relative fluorescence units (scales linearly with fluorescence) normalized to rhBMP-2-treated groups and expressed as mean ± SD. At each temporal period (7, 14, and 21 days), there were six treatment groups, each with four samples (n = 4), and three measurements per sample.
Data are expressed as arithmetic means and reported with SDs. Sample replicates (n) represent the number of wells per treatment and are as indicated. Statistical significance was determined by analysis of variance with Tukey’s test post hoc analysis for multiple comparisons using Graphpad Prism 6 (La Jolla, CA, USA). Statistical significance is reported as indicated.
Current therapies to treat HO (ie, NSAIDs and radiotherapy), if performed early in the HO progression, may successfully modify local microenvironments to be less conducive to heterotopic bone formation. However, these treatments are not biologically engineered to target the molecular mechanisms responsible for bone formation and have notable adverse effects [8, 9, 20, 57]. RNAi, however, may be harnessed to silence master osteogenic factors Runx2 and Osx and consequently prevent the aberrant mineralization of muscle tissue without significant off-target effects (ie, HO). Thus, RNAi technology may produce a compelling alternative prophylaxis for HO for administration at trauma sites. We validated the efficacy of cationic nanogels to deliver Runx2 and Osx siRNA and attenuate osteogenic signaling at mRNA and protein levels. We further identified the capability to inhibit HA deposition in osteoblast cell cultures.
A study of this nature, however, is limited by the innate complexity associated with pathological bone formation mechanisms. The intertwining pathways linking polytrauma to inflammation and HO involve complex signaling and regulatory factors. This interactive biology is difficult to reproduce in vitro. The hallmark of this process is BMP activity; it is known and accepted that BMP-2 is the prototypical molecule for osteogenesis . Terminal osteogenic differentiation is promoted by rhBMP-2; thus, abrogating BMP-induced differentiation has a robust biological basis for HO prophylaxis [32, 42].
The knockdown of transient gene expression by RNAi requires calibration to match the temporal expression profiles of the osteogenic cascade. Runx2 and Osx are key transcriptional molecules of the cascade . Thus, dosing, temporal, and spatial parameters of RNAi must be finely tuned. Temporally, Runx2 and Osx expression have been determined to peak within the first 72 hours after BMP stimulation ; furthermore, Runx2 expression is antecedent to Osx . Thus, our RNAi treatments were engineered to deliver Runx2 siRNA 24 hours before and Osx siRNA simultaneously to rhBMP-2 administration. Data from our work support the hypothesis that silencing specific genes in the osteogenic cascade may be a compelling prophylaxis for HO. Nanogel-to-siRNA ratios from 1:1 to 10:1 have a significant gene-silencing effect in vitro . Our results suggest decreasing Osx gene expression may not affect Runx2 expression (Fig. 2A). This outcome supports the notion that Runx2 regulates Osx expression and not the reverse . However, Runx2 siRNA administration by itself did not suppress Osx expression; both Runx2 and Osx siRNA were required to produce significant Osx gene silencing (Fig. 2B). This suggests Osx expression was upregulated through an alternate pathway , independent of SMAD protein signaling and Runx2 activation .
RNAi treatment outcomes against Runx2 and Osx suggest a significant inhibitory effect on downstream ALP activity (Figs. 3B, 4B). The knockdown was determined to be sequence-specific, as evidenced by negligible changes in ALP activity when negative control siRNA polyplexes were delivered (Fig. 4B). Data suggest LRM-mediated siRNA delivery may produce a nonspecific cytotoxic effect. In contrast, nanogel treatments did not significantly affect cell proliferation (p < 0.05). Thus, a significant reduction in ALP expression may be attained by NG1 and NG5 RNAi treatments against Runx2 and Osx; however, RNAi treatments did not decrease ALP activity to basal levels of MC3T3 cells not treated with rhBMP-2 (p < 0.05). LRM treatments elicited significant reduction in ALP activity, although cytotoxicity issues persisted . Thus, the expression of downstream osteoblast differentiation markers was downregulated as a consequence of RNAi. The reduction of ALP activity may, in turn, prevent late-stage osteoblast maturation, because ALP is known to cleave pyrophosphates that inhibit HA formation and provide the phosphate ions required for HA deposition. To determine if reduction in ALP activity prevented phenotypic cellular change, we analyzed HA deposition in cell cultures.
HA deposition is the hallmark of osteogenic differentiation in MC3T3 cell culture . Silencing Runx2 and Osx over 7, 14, and 21 days produced a significant reduction in HA deposition (Fig. 5). However, reduction of HA deposition to baseline levels was not achieved. Repeated administration of rhBMP-2 (a potent morphogen) over 21 days may have been the cause for this phenomenon. We posit that the repeated delivery of rhBMP-2 challenged the transient efficacy of RNAi therapeutics. To compensate, additional osteogenic signaling factors should be targeted. Mid- and late-stage osteogenic markers including OPN, bone sialoprotein (BSP), and OCN are temporal successors to RUNX2 and OSX expression after rhBMP-2 stimulation . The complex temporal expression profiles of OPN, BSP, and OCN have been elucidated [15, 16, 23, 25, 26, 30, 53, 59, 66]; therefore, a coordinated, temporal RNAi treatment cycle may enhance the duration and intensity of gene silencing and prevent HA deposition in vitro. Consequently, the evolution of RNAi therapeutics must match RNAi targets to their expression profiles. Data suggest that siRNA against Runx2 and Osx should be delivered between Days 0 and 7 and should be succeeded by a permutation of siRNA against Alp, Opn, Bsp, and Ocn between Days 7 and 21. This temporal treatment profile is inspired by the biology and physiology of bone formation and key signaling modulators in the osteogenic cascade.
Ultimately, results suggest that silencing Runx2 and Osx may form the basis of an RNAi prophylaxis for HO in the clinic. However, there are notable challenges we must face. During osteogenic differentiation, both RUNX2 and OSX peak within 72 hours of BMP stimulation . A prophylaxis for HO may require delivery before BMP induction . This treatment schedule, however, may be feasible; the anatomical location of trauma-induced HO is highly predictable [1, 2, 17]. Furthermore, the inflammatory processes preceding bone formation may provide a therapeutic window in which we ultimately deliver HO prophylaxes . These two factors may enable the delivery of a prophylaxis before the maturation of committed osteoblasts at local wound sites.
In conclusion, we report that nanogel-mediated delivery of Runx2 and Osx siRNA inhibits osteoblastic differentiation in murine calvarial preosteoblasts. The nanogel-based siRNA delivery system supports repeated and discrete delivery of therapeutic doses of siRNA without impacting the cellular processes governing viability and proliferation. Knockdown efficacy of nanogels was comparable to LRN treatments without the cytotoxicity of Lipofectamine . By obtaining inspiration from the BMP signaling pathway, we present a biological framework that may prevent HO. However, significant challenges remain: the silencing of mid- and late-stage osteoblast differentiation markers ALP, OPN, BSP, and OCN. Ultimately, the HO prophylactic therapeutic with this biological foundation may eliminate pain and discomfort and improve mobility and quality of life for patients with HO. Further development of this work may produce a compelling clinical contribution to military and civilian medicine.
We acknowledge the assistance of LTC Krinon Moccia, VC USA and the Animal Care Facility staff at Naval Medical Center Portsmouth, for their valuable collaborative contributions. We also acknowledge the Feinberg Laboratory at Carnegie Mellon University for their assistance in confocal microscopy.
- 15.Chen D, Harris MA, Rossini G, Dunstan CR, Dallas SL, Feng JQ, Mundy GR, Harris SE. Bone morphogenetic protein 2 (BMP-2) enhances BMP-3, BMP-4, and bone cell differentiation marker gene expression during the induction of mineralized bone matrix formation in cultures of fetal rat calvarial osteoblasts. Calcif Tissue Int. 1997;60:283–290.CrossRefPubMedGoogle Scholar
- 25.Hassan MQ, Javed A, Morasso MI, Karlin J, Montecino M, van Wijnen AJ, Stein GS, Stein JL, Lian JB. Dlx3 transcriptional regulation of osteoblast differentiation: temporal recruitment of Msx2, Dlx3, and Dlx5 homeodomain proteins to chromatin of the osteocalcin gene. Mol Cell Biol. 2004;24:9248–9261.CrossRefPubMedCentralPubMedGoogle Scholar
- 47.Oh JK, Siegwart DJ, Lee HI, Sherwood G, Peteanu L, Hollinger JO, Kataoka K, Matyjaszewski K. Biodegradable nanogels prepared by atom transfer radical polymerization as potential drug delivery carriers: synthesis, biodegradation, in vitro release, and bioconjugation. J Am Chem Soc. 2007;129:5939–5945.CrossRefPubMedGoogle Scholar
- 63.Shrivats AR, Hollinger JO. The delivery and evaluation of RNAi therapeutics for heterotopic ossification pathologies. In: Turksen K, Vunjak-Novakovic G, eds. Biomimetics and Stem Cells: Methods and Protocols. New York, NY, USA: SpringerLink; 2013.Google Scholar
- 70.Zhu B, Bailey SR, Mauli Agrawal C. Calcification of primary human osteoblast cultures under flow conditions using polycaprolactone scaffolds for intravascular applications. J Tissue Eng Regen Med. 2011 Sep 20 [Epub ahead of print].Google Scholar