Play‐Doh Inspired a Transformative, Shape‐Adaptive iHEÄLS Enabling Post‐Crosslinking Mechanical‐Biological Modulation and Its Effectiveness in Musculoskeletal Repair

Background

Severe trauma often combines open wounds, comminuted fractures and uncontrolled bleeding, and conventional care separates immediate hemostasis from later reconstruction. Bone wax seals bleeding bone but is non-resorbable and impedes healing, polymethylmethacrylate cements lack bioactivity and polymerize exothermically, and calcium-based fillers are brittle and slow to remodel. Many resorbable alternatives lose moldability after crosslinking, perform poorly in bloody fields and cannot hold small bone fragments that are too fragile for fixation.

Design and approach

iHEÄLS is formed by mixing PVA, gelatin and CNC with borax and microbial transglutaminase (mTG), giving three interacting networks: hydrogen bonding and entanglement among CNC, gelatin and PVA; mTG-catalyzed covalent isopeptide bonds in gelatin that support adhesion and gradual enzymatic degradation; and dynamic borate ester bonds between PVA and borax that provide self-healing and shape adaptability. Because the dough remains moldable after crosslinking, nano-hydroxyapatite (nHA) or bone fragments can be kneaded in afterward. Repeated stretch-fold cycles align the polymer chains and orient the CNC rods, which raises stiffness without additional crosslinkers.

How it was evaluated

Network chemistry and structure were examined by FTIR, SEM and small-angle X-ray scattering, and mechanical behavior by rheology, compression, extension, lap-shear and pulling tests on porcine skin, bone, metal and other substrates, including after 24 h at pH 6.5 and 7.0 and after up to 14 days in PBS at 37 °C. Degradation was measured in PBS and collagenase and after subcutaneous injection in nude mice for 4 and 8 weeks. Hemostasis was tested in a mouse tail amputation model and a rat liver injury model. Human dermal fibroblasts, C2C12 myoblasts and human mesenchymal stem cells (hMSCs) were cultured on different formulations, with and without mechanical training, nHA or the YAP inhibitor verteporfin. Bone repair was assessed in a 6.5 mm critical-sized rat calvarial defect with superior sagittal sinus disruption, comparing sham, untreated defect, nHA/iHEÄLS and nHA/iHEÄLS with autologous bone fragments by µCT and histology at 8 weeks.

Key results

Young's modulus increased from about 5 to about 12 kPa as PVA rose from 6% to 12%. CNC-containing dough stretched 100 times its length, from 5 mm to over 50 cm, whereas the CNC-free dough stretched about 15 times. Five to 20 stretch-fold cycles raised the modulus of CNC-containing dough to about 11 kPa, nearly double, and SAXS showed CNC aggregates breaking up while orientational order persisted. After incubation in PBS at 37 °C, the trained CNC-containing dough decreased from about 13.7 kPa at day 1 to about 4.2 kPa at day 14, while the CNC-free dough fell from about 10.5 to about 0.2 kPa. At pH 6.5, G′ remained about 15.9 kPa compared with about 20.5 kPa at pH 7.0.

Adding 2.5% CNC nearly doubled the adhesive force between porcine skin pieces, and the dough stretched about 150% while staying attached to both sides. In pulling tests, adhesion ranked bone > wood > plastic > porcine skin > glass > metal. Two separately dyed pieces merged into one, and strength recovered over repeated healing cycles; after 24 h at pH 6.5, adhesion on glass and metal showed no substantial loss and self-healing was only modestly reduced.

The dough gained 17% to 40% in weight in PBS over 3 days for all formulations except P10, and CNC slowed degradation; after subcutaneous injection in nude mice, CNC-free dough degraded nearly twice as fast as the 1.5% CNC formulation over 8 weeks, with visible vascular ingrowth at the interface. In the mouse tail amputation model, bleeding stopped within 1 min, while untreated tails lost nearly three times more blood. In the rat liver model, the bleeding area decreased by about 78% within 1 min compared with untreated controls.

On the dough, fibroblasts proliferated and expressed more COL1 and COL3 with CNC, C2C12 cells formed longer myotubes with higher fusion index, and hMSCs showed higher ALP activity, mineralization and ITGα5, ITGβ1, VCL, OPN and OCN expression. Mechanical training further increased ALP, OPN protein and mineralization, and verteporfin lowered YAP1 and PTK2 expression in trained groups; 1% nHA gave higher OCN and OPN expression than 5%. In the rat calvarial defect, nHA/iHEÄLS with bone fragments achieved hemostasis within 50 s, whereas untreated defects bled for over 10 min. At 8 weeks, its BV/TV reached 34.80%, 1.65 to 6.79 times higher than the iHEÄLS-only and defect-only groups (sham 62.58%), with new bone and OPN expression around the implanted fragments.

Open questions

The authors state that a longer-term in vivo tracking study is needed to evaluate biodegradation kinetics, biodistribution, clearance and chronic safety beyond the 8-week implantation period. They also note that collecting uniformly sized bone fragments and distributing them evenly may be difficult in emergency settings, and they suggest future loading of drugs or growth factors through CNC.

How it connects to other research

Tough adhesive hydrogels for wet surfaces separate an adhesive interface from an energy-dissipating matrix, whereas iHEÄLS keeps adhesion, self-healing and moldability in one dough whose stiffness is raised by mechanical training after crosslinking.