Graphene Oxide-Delivered TIMP‑1 Enhances Skin Regeneration Over 40 Days

Background

Cutaneous injuries arise from trauma, diabetes, burns, or surgery, and traditional autografts or synthetic skin substitutes often face donor shortages, infection risks, and suboptimal integration.

TIMP‑1 binds matrix metalloproteinases (MMPs) to protect extracellular matrix (ECM) integrity and modulates growth factor turnover essential for wound healing. Dysregulation of the TIMP‑MMP axis contributes to impaired healing, keloids, and fibrosis. Therefore, precise spatial and temporal delivery of TIMP‑1 is critical for optimal re‑epithelialization.

Existing cytokine carriers—PLGA, chitosan, PLGA microspheres, and hydrogels—provide limited release profiles. Graphene oxide, a two‑dimensional carbon material with abundant hydrophobic π domains and ionizable edges, offers high loading capacity, aqueous dispersibility, and low intrinsic toxicity, making it a promising platform for protein delivery.

Here, we evaluate recombinant human TIMP‑1 loaded onto GO for controlled release and skin repair efficacy in vitro and in a rat wound model.

Materials and Methods

Cell Culture

Primary rat fibroblasts (Institute of Biochemistry and Cell Biology, CAS) were cultured in DMEM supplemented with 10% FBS at 37 °C, medium changed every 48 h.

GO Characterization

GO flakes (Chengdu Organic Chemicals Co.) were examined by SEM (Hitachi S3000N, 15 kV) and AFM (VEECO MultiMode). Size distribution was measured via dynamic light scattering (Zetasizer 3000 HSA). TGA/DSC (Pyris 1 TGA) assessed thermal stability from 25–1,100 °C at 10 °C/min.

TIMP‑1 Adsorption on GO

GO was labeled with DiI (red, Sigma) and incubated with FITC‑conjugated TIMP‑1 (Thermo Scientific) in PBS for 4 h at 4 °C (GO:TIMP‑1 = 1:1 w/w). Confocal microscopy (Olympus IX81) visualized adsorption; FTIR (Nicolet 5700) confirmed protein binding.

Release Kinetics of TIMP‑1 Protein

GO loaded with 10, 20, or 30 µg/ml TIMP‑1 was suspended in 1.5 ml PBS in 60‑mm dishes and incubated at 37 °C. Supernatants were collected at predetermined intervals, refreshed with PBS, and quantified by TIMP‑1 ELISA (R&D Systems).

GO Biocompatibility Assay

Fibroblasts were exposed to 20 µg/ml TIMP‑1‑GO and monitored over 6–72 h. Viability was measured with CCK‑8 (absorbance at 450 nm, n = 5 per group).

Flow Cytometric Characterization

Cells were stained with Hoechst 33258 and Annexin‑V‑FITC/PI, fixed, and analyzed on a BD FACSCanto II to assess cell‑cycle distribution and apoptosis.

In Vivo Experiment

All procedures adhered to NIH guidelines and Zhejiang University Ethics Committee approval. Four‑week‑old male Sprague–Dawley rats received 10 mm × 10 mm excisional wounds. Fourteen days post‑surgery, rats were randomized into four groups: PBS, GO, TIMP‑1, or TIMP‑1‑GO (1:1 v/w). Treatments were subcutaneously injected (1 ml per site, four sites, once weekly for two weeks). Animals were euthanized 28 days post‑surgery for histological analysis.

Histological and Immunohistochemical Analysis

Tissues were fixed in 4% formalin, decalcified, dehydrated, and embedded in paraffin. Hematoxylin–eosin and Masson’s trichrome stained sections evaluated collagen architecture. CD34 immunostaining (Abcam) quantified neovascularization; sections were processed with DAB visualization and graded by three blinded observers.

Statistical Analysis

Data represent mean ± SD from three independent experiments. One‑way ANOVA with Student–Newman–Keuls post‑hoc test determined significance (p < 0.05 = significant; p < 0.01 = highly significant). Analyses were performed using SPSS 17.0.

Results

GO Characterization

AFM revealed 2‑D sheet‑like GO structures (Fig. 1a). SEM confirmed irregularly shaped flakes (Fig. 1b). Dynamic light scattering indicated a mode size of 1,140 nm and a volumetric peak at 10,674.1 nm³ (Fig. 2a).

GO morphology: (a) AFM, (b) SEM, (c) size distribution.

TIMP‑1 Adsorption on GO

Confocal imaging confirmed 75 ± 1.2% TIMP‑1 loading onto GO after 4 h (Fig. 1c). FTIR spectra showed distinct peaks for TIMP‑1‑GO versus bare GO, confirming protein attachment. TGA curves overlapped, indicating thermal stability was retained (Fig. 2d).

TIMP‑1‑GO characterization: (a) loading efficiency, (b) release profile, (c) FTIR, (d) TGA.

TIMP‑1 Release

Loading 2, 10, or 30 µg/ml TIMP‑1 onto GO produced sustained release. The 2 µg/ml dose reached 50% cumulative release faster than higher doses, yet all formulations maintained detectable TIMP‑1 for ~40 days (Fig. 2c).

Cell Proliferation and Viability on TIMP‑1‑GO

Fibroblast viability, cell‑cycle progression, and apoptosis rates were comparable across control, GO, TIMP‑1, and TIMP‑1‑GO groups (p > 0.05) (Fig. 3a‑c).

Effect of TIMP‑1‑GO on fibroblast function: (a) viability, (b) cell‑cycle, (c) apoptosis.

Efficacy of TIMP‑1‑GO in Excisional Skin Wound Model

After 28 days, TIMP‑1‑GO treatment yielded superior histological scores, increased collagen density, and enhanced hair follicle regeneration compared with PBS, GO, and TIMP‑1 alone (p < 0.05) (Fig. 4b‑c).

In vivo results: (a) experimental scheme, (b) histology/immunohistochemistry, (c) quantitative assessment.

Histologic and Immunohistochemical Analysis

Control wounds displayed fragmented collagen fibers at 4 weeks, whereas TIMP‑1‑GO‑treated wounds showed continuous, dense collagen and abundant CD34⁺ endothelial cells, confirming enhanced angiogenesis (p < 0.05) (Fig. 4c).

Discussion

Our findings establish GO as a safe, high‑capacity platform for the prolonged delivery of TIMP‑1, sustaining therapeutic protein release for up to 40 days. The sustained TIMP‑1 exposure supports balanced MMP activity, promotes ECM deposition, and facilitates neovascularization—key drivers of effective wound closure.

Compared with conventional carriers that release cytokines rapidly, GO’s π‑π stacking and electrostatic interactions afford a controlled, extended release profile. This mitigates the need for repeated dosing and reduces potential local side effects.

While high concentrations of carbon nanomaterials can elicit inflammation, our in vitro and in vivo data show no detectable GO deposition or adverse cellular responses at the tested dose (20 µg/ml). Nevertheless, long‑term safety studies remain essential for clinical translation.

Conclusion

Graphene oxide enables sustained, biocompatible delivery of recombinant TIMP‑1, enhancing collagen formation and angiogenesis in a rat skin wound model. This platform offers a promising strategy for chronic wound management and could inform the design of next‑generation biomaterial‑based therapeutics.