Injectable non-leaching tissue-mimetic bottlebrush elastomers as an advanced platform for reconstructive surgery

The study addresses a key challenge in biomedical implants: conventional gels often mismatch the nonlinear mechanical response of soft tissues and can leach chemicals, causing health risks and complications such as capsular contracture, gel fracture, and inflammatory responses. Biological tissues are characterized by a combination of initial softness (~1 kPa modulus) and pronounced strain-stiffening (firmness), which most gel-based materials fail to replicate due to flexible linear network strands that do not reach finite extensibility. Existing syntheses for tissue-mimetic bottlebrush elastomers typically require solvents and hazardous stimuli (UV, heat), limiting clinical translation. The authors hypothesize that solvent-free, brush-architecture polymer melts with abundant functional chain ends can be injected and cured in vivo to form non-leaching elastomers whose mechanical properties (softness and firmness) can be independently tuned to match surrounding tissues, while also enabling control over gelation time from minutes to hours for diverse surgical needs.

Prior work has shown: (i) gels used in biomedical devices suffer from mechanical mismatch and leachables, contributing to capsular contracture, inflammation, and potential systemic effects; (ii) organogels and silicone gels can bleed, with documented detection of siloxanes and silicone in tissues and breast milk; (iii) bottlebrush architectures enable supersoft, strain-stiffening elastomers matching tissue-like stress–strain behavior without solvents, but typical synthesis involves solvents/UV/heat; (iv) injectable microgels like PAAG have led to severe complications and are now largely banned. Foundational theories of nonlinear elasticity (Dobrynin–Carrillo) and architectural control of rubber elasticity (Sheiko–Dobrynin) support tuning softness and firmness via bottlebrush parameters. These insights motivate solvent-free, injectable, non-leaching bottlebrush networks with clinically compatible curing chemistries.

Materials and synthesis: Random PDMS–PEG bottlebrush polymers (PDMS-r-PEG) were synthesized via ATRP of PDMS-methacrylate (PDMSMA) macromonomers with a controlled, small fraction (~0.02 mol%) of PEGMA macromonomers bearing reactive end-groups (OH or N3). Time-resolved 1H-NMR verified random macromonomer incorporation and low functional fractions; AFM imaged worm-like bottlebrush macromolecules. OH side-chain ends were used directly or post-functionalized (e.g., furan, methacrylate). Azides were reduced to amines to obtain NH2-functionalized brushes. A PDMS diisocyanate macromolecular crosslinker was synthesized from IPDI and PDMS diamine. Formulations and crosslinking chemistries: A dual-syringe system combined (i) bottlebrush melts bearing reactive side-chain ends with (ii) a linear/bifunctional crosslinker. Primary chemistries included isocyanate:hydroxyl (NCO:OH, slow cure; t_gel ~ hours) and isocyanate:amine (NCO:NH2, fast cure; t_gel ~ minutes). Additional schemes explored included reversible Diels–Alder (furan:maleimide) and photocurable methacrylate side chains. Gelation time was tuned by stoichiometry (NCO:OH ratios 1:1 to 1:8), temperature (20–50 °C), and catalyst concentration (e.g., 200–600 ppm Sn(II) 2-ethylhexanoate for NCO:OH), allowing partial decoupling of modulus and cure time. Star-like bottlebrushes further reduced melt viscosity. Rheology and injectability: Gelation was monitored by G′/G″ crossover at 37 °C. Melt viscosity of bottlebrushes versus linear PDMS of similar Mn (~500 kDa) was measured, demonstrating reduced zero-shear complex viscosity with branching and longer side-chains. Injectability was quantified using a bioprinter (pressure vs mass flow; 16G/20G nozzles at 25 and 37 °C). Premixed formulations remained fluid at 0 °C for storage stability. Mechanical testing: Fully cured elastomers underwent uniaxial tensile tests (dogbone 12×2×1 mm; strain rate 0.005 s−1; 20 °C). True stress–stretch curves were fit to σ_true = E(λ^2 − λ^−1) + [1/3(1 − β^3 λ^−3 + 2 λ^−1)]^−1 to extract structural modulus E, firmness β, and E0. Texture Profile Analysis (TPA) assessed springiness, resilience, and cohesiveness under cyclic compression (disks, 8 mm diameter; 20/50/70% strain). Frequency-dependent viscoelasticity was also characterized. Leachability tests: Bulk samples were placed on paper to visualize bleed; aqueous extractions were analyzed by time-resolved 1H-NMR over one month; leached residues were freeze-dried and weighed. In vitro cytocompatibility and proliferation: Cytotoxicity followed ISO 10993-5 using extracts (DMEM+10% FCS+1% Pen/Strep, 3 cm2/mL, 24 h, 37 °C) from injectable elastomers and commercial silicone gels. NIH/3T3 (and HUVECs in supplementary) were exposed; viability measured with PrestoBlue fluorescence. Proliferation on elastomer surfaces was assessed by total DNA (PicoGreen) at 3, 5, 7, 14 days; Sylgard 184 and TCPS served as controls; fluorescence imaging (F-actin, DAPI) monitored cell numbers. In vivo studies: Male Wistar rats received ex vivo cured implants (to match mechanics of silicone gel) subcutaneously and intramuscularly; an additional cohort received a direct in vivo injection of a formulation to assess on-site curing. Explants at 1, 4, 12 weeks were analyzed histologically (H&E; Mallory trichrome). Capsular thickness was quantified by morphometric analysis at 10 locations per sample (n=6 rats per time point). Long-term stability was tested by re-measuring mechanical/TPA properties after implantation and after PBS (pH 7.4) incubation at 70 °C.

- Injectable, solvent-free curing with tunable gelation time: Within NCO:OH chemistry, t_gel spanned over two orders of magnitude by stoichiometry and temperature. For NCO:OH 1:1, t_gel was ~8.5 h at 20 °C, ~1.4 h at 37 °C, and ~0.6 h at 50 °C. Switching to NCO:NH2 reduced t_gel to minutes. Catalyst (Sn(II) 2-ethylhexanoate) concentration at fixed NCO:OH 1:4 (200–600 ppm) altered t_gel while maintaining final modulus, partially decoupling cure time and mechanics. Reversible Diels–Alder crosslinking extended t_gel up to ~10.5–11 h at 37 °C. - High crosslinking efficiency: Gel fraction 91–98% across studied systems, indicating efficient solvent-free curing. - Tissue-mimetic mechanics via architecture: By tuning bottlebrush architectural parameters [n1(backbone between crosslinks), n2(side-chain DP), n3(backbone spacer)] and crosslink density, materials replicated both tissue softness and firmness (strain-stiffening). Representative values (Table 1): • NCO:OH 1:1 (n_side=14): E=18.6 kPa, β=0.24, E0=27.8 kPa, λ_exp=2.1. • NCO:OH 1:4 (n_side=14): E=4.2 kPa, β=0.10, E0=5.1 kPa, λ_exp=3.2. • NCO:OH 1:8 (n_side=14): E=2.1 kPa, β=0.08, E0=2.3 kPa, λ_exp=3.6. • Longer side-chains (n_side=70) at similar E increased firmness (e.g., NCO:OH 1:2, E=1.3 kPa, β=0.26, E0=2.1 kPa). Materials matched weakly strain-stiffening tissues (e.g., chicken gut, dog lung) and a commercial silicone gel’s stress–strain profile, while exhibiting higher elastic deformation before fracture and greater resilience. • Theoretical elongation at break (λ_max,theo = β^−0.5) agreed with experiment, suggesting uniform networks. Structural modulus E followed predicted scaling E ≈ 3kT/l^2 β^(1+3/2). Similar mechanics were achieved across different crosslink chemistries, indicating architectural—not chemical—control of mechanics. - Non-leaching behavior: Commercial silicone gels exhibited visible bleed on paper and detectable leachables by 1H-NMR (peaks at 1.17 and 0.01 ppm) in D2O over one month, whereas injectable elastomers showed no detectable leachables under identical tests. Quantitative extractions confirmed substantially lower leached mass from injectable elastomers than from silicone gels (supplementary data). - In vitro biocompatibility: ISO 10993-5 cytotoxicity assays (NIH/3T3) showed >90% viability after 24 h exposure to extracts from injectable formulations (NCO:OH 1:1–1:8). Extracts from commercial silicone gels reduced viability to ~40–60%. Fibroblast proliferation on elastomer surfaces increased over two weeks with minimal deviation from TCPS controls; fluorescence imaging corroborated cell growth without abnormal stimulation. - In vivo performance: Ex vivo cured injectable elastomers implanted subcutaneously and intramuscularly remained intact, retained shape, and elicited thinner fibrous capsules than silicone gels at 1, 4, and 12 weeks (n=6 rats; capsular thickness measured at 10 locations per sample). Histology showed absence of multinucleated foreign body giant cells and lack of chronic inflammatory infiltrates at later stages. Directly injected samples cured in situ without dispersion and showed only transient inflammation at 7–14 days. Mechanical and TPA properties remained stable after implantation and after PBS incubation at 70 °C, consistent with hydrophobic, hydrolytically stable networks.

The work demonstrates that brush-architecture elastomer networks can reproduce the dual mechanical behavior of soft tissues—initial softness with strong strain-stiffening—without reliance on solvent swelling. The compact, low-viscosity bottlebrush melts enable minimally invasive, solvent-free injection and curing in vivo using clinically relevant, tunable chemistries. This approach addresses key shortcomings of current gel-based implants, namely mechanical mismatch leading to capsular contracture and the risk of chemical leaching. The ability to control both curing kinetics (minutes to hours) and final mechanics by architecture provides a versatile platform for diverse surgical timelines and tissue targets. Non-leachability, together with favorable in vitro cytocompatibility and reduced in vivo fibrotic encapsulation compared to silicone gels, underscores the translational potential for reconstructive surgery and related biomedical devices. The invariance of mechanical behavior across crosslink chemistries further highlights that the core design principle resides in molecular architecture, allowing chemistry to be selected for biocompatibility and processing considerations while maintaining tissue-mimetic performance.

This study introduces a solvent-free, injectable bottlebrush elastomer platform that cures in vivo into non-leaching implants with tunable, tissue-mimetic mechanics and adjustable gelation times. Architectural control over network strand stiffness and extensibility enables independent tuning of softness and firmness, while formulation parameters (stoichiometry, temperature, catalysts, and reversible chemistries) modulate cure time from minutes to hours. The materials exhibit high gel fractions, support cell viability and proliferation in vitro, form thinner capsules and show minimal chronic inflammation in vivo, and retain mechanical integrity under physiological conditions. Future work should expand crosslinking schemes and architectural spaces to more fully decouple curing duration from mechanical properties, adapt the architecture to other polymer chemistries (e.g., polyolefins, polyacrylates, polyesters), integrate functional side-chain chemistries for adhesives and device coatings, and leverage additive manufacturing to fabricate patient-specific, soft medical implants.

- Tuning crosslinker concentration simultaneously affects both curing time and mechanical properties; decoupling requires catalysts or alternative reversible chemistries (e.g., Diels–Alder), which adds formulation complexity. - Slowly curing formulations may risk undesirable in vivo spreading; mitigation includes using fast-curing NCO:NH2 systems or administering small volumes of viscous, hydrophobic melts; high-volume applications may require lumened delivery systems. - Direct injection induced a transient inflammatory response at 7–14 days, indicating short-term tissue reactivity to the curing process. - Long-term degradation was not observed due to hydrophobic, hydrolytically stable networks; while beneficial for stability, the lack of degradability may limit applications requiring resorption. - In vivo data were obtained in a rat model; translation to human physiology and long-term clinical outcomes remains to be established.