Abstract

   Diabetic wounds require continuous and coordinated modulation of the
   microenvironment concurrent with tissue regeneration, which remains a
   significant challenge. As a proof of concept, we herein propose to use
   dimeric copper peptide (D-CuP) for diabetic wound treatment. The D-CuP
   is synthesized and then incorporated into a reactive oxygen species
   (ROS)-responsive hydrogel matrix to improve therapeutic compliance,
   culminating in the formulation of G/D-CuP. Compared to monomer copper
   peptide (M-CuP), a wound healing agent, D-CuP exhibits multivalency,
   enhanced biological stability against proteases, and broad biological
   activities. Meanwhile, the hydrogel matrix, exhibiting ROS-scavenging
   capabilities, has been engineered to be an intelligent drug reservoir
   for wound-responsive release of D-CuP at the wound site while
   simultaneously attenuating inflammatory responses. Ultimately, the
   G/D-CuP group demonstrates superior therapeutic efficacy, achieving
   97.2% closure of infected wounds.

   Subject terms: Drug delivery, Drug delivery, Medical research
     __________________________________________________________________

   Diabetic wounds are a prevalent type of chronic wound that is
   challenging to treat due to the dynamic and complex wound
   microenvironment. Here, Cong et al. presents a dimeric copper peptide
   incorporated hydrogel for multifunctional treatment of diabetic wounds.

Introduction

   Dimerization is essential in biological systems, playing a key role in
   regulating physiological functions and regulatory mechanisms of living
   organisms^[44]1,[45]2. This process involves the combination of two
   molecules or molecular fragments to form a higher-order structure via
   non-covalent interactions or covalent bonds, enabling the rapid
   construction of complex molecular architectures with diverse
   physiological properties^[46]3,[47]4. In transcriptional regulation
   cascades, dimerization affects gene expression by modulating the
   activity of transcription factors^[48]5,[49]6. It also contributes to
   the modulation of enzyme activity, which in turn impacts metabolic
   pathways and cellular processes^[50]7,[51]8. Additionally, dimerization
   can stabilize proteins, helping them maintain their functional
   conformation and resist denaturation under physiological conditions as
   well as regulate intracellular signaling molecules^[52]9,[53]10.
   Therefore, the formation of dimeric complexes is critical in preserving
   cellular balance and proper functions. Notably, dimerization presents
   opportunities for developing pharmaceuticals, especially those
   necessitating multiple target mechanisms and functions^[54]11,[55]12.
   Dimeric compounds are promising for simultaneously targeting dual
   binding domains, enhancing molecular recognition precision and
   consequently optimizing reaction dynamics and biochemical
   selectivity^[56]13,[57]14. When addressing dimeric protein targets, the
   ability of dimeric therapeutic agents to antagonize paired receptors
   within a single molecular entity potentially offers superior
   therapeutic efficacy and target selectivity^[58]15,[59]16.

   Diabetic wounds are a prevalent type of chronic wound that
   significantly decreases patients’ quality of life and imposes
   substantial clinical and economic burdens^[60]17,[61]18. Studies
   indicate that in the absence of appropriate medical intervention,
   approximately 20% of patients with diabetic wounds will require lower
   limb amputation, with a subsequent one-year mortality rate of 13%
   post-diagnosis^[62]19,[63]20. Key factors hindering the healing of
   diabetic wounds include persistent high level of reactive oxygen
   species (ROS), excessive inflammatory cells and cytokines in the wound
   microenvironment, insufficient new blood vessel formation, and weakened
   cell proliferation^[64]21,[65]22. These elements establish a feedback
   loop that delays tissue repair, making it essential to precisely
   regulate the localized microenvironment to address multiple pathogenic
   mechanisms simultaneously. Current wound management approaches focus on
   scavenging ROS^[66]23,[67]24, mitigating inflammation^[68]25,[69]26,
   modulating immunity^[70]27,[71]28, and enhancing tissue
   regeneration^[72]29,[73]30. However, they are facing challenges in
   coordinating the individual components within the disordered diabetic
   wound microenvironment. Thus, designing an all-rounder with space-time
   ROS scavenging, modulation of macrophage polarization, and promotion of
   angiogenesis and cell proliferation is a prominent subject for
   advancing the clinical treatment of chronic diabetic wounds.

   Many biologically active peptides, such as antimicrobial and
   antioxidant peptides, exhibit physiological functions and have the
   potential for tissue damage repair^[74]31,[75]32. GHK
   (glycyl-l-histidyl-l-lysine) is an endogenous tripeptide isolated from
   plasma, which can coordinate with the organic cation Cu^2+ to form a
   copper peptide (CuP) complex^[76]33,[77]34. CuP demonstrates a wide
   range of biological activities, including antioxidative properties,
   inflammatory modulation, chemotactic cell recruitment, and promotion of
   angiogenesis and cell proliferation^[78]35,[79]36. These properties
   make CuP therapeutically beneficial across all phases of wound healing.
   Nevertheless, CuP is prone to hydrolysis by proteases, resulting in
   treatment failure^[80]34,[81]37. Techniques such as supramolecular
   self-assembly^[82]38,[83]39, macrocyclization^[84]40,[85]41, polymer
   conjugation^[86]42,[87]43 are commonly utilized to enhance the
   stability of bioactive peptides. However, in certain cases, they may
   interfere with the peptide’s ability to interact with its target or
   disrupt bioactive conformation, significantly reducing its
   activity^[88]44,[89]45. Additionally, these modifications may induce
   significant toxicity by triggering immune responses^[90]46,[91]47.
   Consequently, there is an urgent need for innovative chemical
   structures and strategies to bolster the stability of bioactive
   peptides, while the production is convenient and cost-effective.

   In this study, we present a proof-of-concept study to optimize the
   treatment of diabetic wounds (Fig. [92]1). This method integrates a
   dimeric copper peptide incorporated hydrogel (G/D-CuP) to address
   several critical factors: i) The design and synthesis of the dimeric
   copper peptide (D-CuP) feature a broad range of targets and enhanced
   biological activities, such as anti-inflammatory and antioxidative
   properties, while promoting angiogenesis and fibroblast proliferation
   and migration^[93]33,[94]48; ii) Utilizing the steric hindrance
   conferred by dimerization reduces protease access to the active site
   and significantly improving stability against proteases^[95]49,[96]50;
   iii) A ROS-responsive hydrogel matrix (G) has been synthesized,
   exhibiting superior ROS-scavenging capabilities, functioning as an
   intelligent drug reservoir that enables the controlled release of D-CuP
   at the wound interface while simultaneously attenuating inflammatory
   responses in the microenvironment^[97]51; iv) G/D-CuP possesses
   deformability and spreadability, allowing it to adapt to any wound
   shape; v) The synthesis and preparation process of G/D-CuP are
   straightforward and cost-effective, making it highly promising for
   clinical translation. This multifunctional treatment platform is
   anticipated to accommodate the complex and dynamic biological processes
   involved in diabetic wound healing, thereby significantly improving the
   overall efficacy of wound healing treatments.

Fig. 1. Dimeric copper-peptide incorporated hydrogel (G/D-CuP) for enhanced
diabetic wound healing efficacy.

   [98]Fig. 1
   [99]Open in a new tab

   a Illustration of the G/D-CuP production method. b The function of the
   G/D-CuP in the process of diabetic wound healing.

Results

Preparation of the D-CuP

   The dimeric peptide D-P was synthesized through the covalent
   conjugation of two glycyl-l-histidyl-l-lysine (GHK) tripeptide moieties
   utilizing lysine as the bridging molecule (Fig. [100]2a and
   Supplementary Fig. [101]1). The resulting product was characterized by
   ^1H NMR, ^13C NMR and HRMS (Supplementary Figs. [102]2–[103]4),
   confirming successful synthesis. As shown in Fig. [104]2b, upon
   addition of CuSO[4] solution, a blue-purple D-CuP complex was formed in
   a molar ratio of GHK: Cu^2+ = 1: 0.5. The UV-Vis absorption spectrum
   for D-CuP exhibited a distinct absorption peak at 584 nm, consistent
   with that of monomer copper peptide (M-CuP) (Fig. [105]2c). This result
   suggests the successful chelation of Cu^2+ with D-P. When the molar
   ratio of the coordinating functional groups remains constant, the
   absorbance observed at 584 nm for the dimer was larger than that of the
   monomer peptide, indicating that D-P possesses better coordination
   efficacy and greater stability than M-P (Fig. [106]2d). Electron
   paramagnetic resonance (EPR) analysis further confirmed the formation
   of D-CuP through distinctive Cu^2+ signals at ∼3426 mT (Fig. [107]2e).
   The complex pathophysiology of wound microenvironments, characterized
   by fluctuating pH levels and elevated protease activity^[108]52,
   necessitates a thorough assessment of the chemical stability of D-CuP.
   Under normal conditions, the skin exhibits a slightly acidic
   environment (pH 4–6) that helps sustain the epidermal barrier and
   inhibits bacterial intrusion. In contrast, during chronic or infected
   wound states, the pH shifts to alkaline levels (pH 7–9) as a result of
   microbial activity^[109]53,[110]54. We evaluated the pH stability of
   D-CuP across the range of cutaneous wound conditions (pH 4–9). As
   illustrated in Supplementary Fig. [111]5, the characteristic absorption
   peak at 584 nm for both M-CuP and D-CuP exhibited no statistically
   significant variation across the tested pH range. This pH-adaptive
   stability spans both physiological skin acidity (pH 4–6) and the
   alkaline extremes (pH 7–9) typical for chronic wounds, confirming the
   ability of D-CuP to maintain structural integrity under pathological
   conditions. Protease K, a serine protease with broad cleavage activity,
   was employed to evaluate the protease stability of D-CuP in vitro. As
   shown in Fig. [112]2f, M-CuP degraded rapidly, retaining about 58% and
   50% after 0.5 and 4 h, respectively. In contrast, approximately 92% and
   87% of D-CuP remained unchanged after 0.5 and 4 h of protease K
   exposure, highlighting its superior protease stability. This may be
   attributed to the dimeric structure which hinders enzyme access to
   active sites, showcasing significant resistance to enzymatic
   degradation^[113]49,[114]55.

Fig. 2. Preparation and characterization of the D-CuP.

   [115]Fig. 2
   [116]Open in a new tab

   a Structural formula of the D-CuP. b The color of D-CuP solution. c, d
   UV-vis absorption spectra of D-CuP (n = 3 independent replicates).
   Statistical significance was assessed using unpaired t test. e Electron
   paramagnetic resonance spectroscopy of D-CuP. f Stability curve of
   M-CuP and D-CuP against protease K. Data represented as Mean ± SD
   (n = 3 independent replicates). Statistical significance was assessed
   using two-way ANOVA with Šídák’s multiple comparisons test. Source data
   are provided as a Source Data file.

Preparation of the G/D-CuP

   The overproduction of ROS can worsen tissue damage and inflammatory
   reactions, potentially delaying wound healing during the inflammation
   stage. Thus, eliminating surplus ROS is highly advantageous for wound
   healing^[117]56,[118]57. The ROS-responsive linker
   (((ethane-1,2-diylbis(azanediyl))bis(carbonyl))bis(4,1-phenylene))dibor
   onic acid (EBPBA) was synthesized (Supplementary Fig. [119]6), which
   can crosslink polyvinyl alcohol (PVA) to form a hydrogel with
   phenylboronic ester bonds and generate ROS responsiveness. The EBPBA
   structure was confirmed by ^1H NMR and ESI-MS (Supplementary
   Figs. [120]7, [121]8). Then, D-CuP loadedROS- responsive hydrogel was
   prepared and the preparation process was illustrated in Fig. [122]3a.
   G/D-CuP could quickly be formed by simply mixing PVA, EBPBA, and D-CuP
   solution. The hydrogel underwent crosslinking by the interaction of
   phenylboronic acids derived from EBPBA and hydroxyl groups from PVA,
   resulting in the formation of phenylboronic ester bonds. The rheology
   properties further confirmed that the hydrogel was successfully
   prepared. Scanning electron microscope (SEM) images revealed the
   3D-network structure and pore structure of the G/D-CuP (Fig. [123]3b).
   As shown in Fig. [124]3c, under low strain conditions (0.1% to 1200%),
   G/D-CuP showed a storage modulus (G’) consistently surpassing the loss
   modulus (G”). This rheological behavior suggests that the hydrogel
   exhibits a semi-solid consistency with substantial elasticity,
   validating its successful formation. The frequency sweep results, as
   illustrated in Fig. [125]3d, indicated that the G’ of the G/D-CuP
   remained dominant over G” across the tested range (0.1–500 rad·s⁻¹),
   confirming the establishment of a stable hydrogel network. The
   hydrogel, when covering wound surfaces, is susceptible to damage from
   skin movement, which could impede wound healing and increase infection
   risk. However, a self-healing hydrogel is capable of providing
   prolonged protection to wounds^[126]58,[127]59. A continuous step
   strain test was conducted to evaluate the self-healing capacity of
   G/D-CuP. The G’ of G/D-CuP decreased from 2500 Pa (0.1% strain) to
   390 Pa (1200% strain), but upon strain relief, the G’ recovered to
   2500 Pa (Fig. [128]3e), indicating a temporary structure and elasticity
   disruption under high strain conditions with subsequent recovery. The
   hydrogel’s ability to self-repair was also demonstrated in
   Fig. [129]3f, where it healed after being cut into two pieces and
   withstood bending force without cracking. Effective wound filling and
   infection prevention are crucial, especially in cases of irregularly
   shaped wounds^[130]60,[131]61. Figure [132]3g and Supplementary
   Fig. [133]9 illustrated the hydrogel’s ability to effectively spread
   and fill intricate star-shaped wounds. Supplementary Fig. [134]10 also
   evidenced the hydrogel’s strong adhesion capabilities enable close
   contact with wound tissue, which was beneficial to prevent bacterial
   infiltration. Through cleavage of dynamic phenylboronic ester bonds
   under ROS stimulation^[135]62, the hydrogel demonstrated ROS-sensitive
   drug release behavior, as evidenced by the 90% cumulative release of
   D-CuP within 72 h under H[2]O[2] stimulation (Fig. [136]3h). To
   evaluate the stability and retention of G/D-CuP in vivo, it was
   topically applied to full-thickness dorsal wounds in diabetic mice.
   Dynamic imaging was performed at different times to monitor the
   morphological change of hydrogel. The results showed that the hydrogel
   remained at the wound site for over 48 h, indicating favorable
   retention (Supplementary Fig. [137]11a). Weight loss analysis of
   retrieved hydrogels showed 87.44% degradation within 48 h
   (Supplementary Fig. [138]11b). To assess drug release, we applied
   FITC-conjugated D-CuP hydrogel to the wounds and measured the
   fluorescence signal of the residual hydrogel at predetermined time
   intervals in darkness. As shown in Supplementary Fig. [139]11c, the
   results indicated a rapid initial release, with 56.79% released within
   the initial 12 h, followed by gradual and continuous release reaching a
   cumulative payload liberation of 91.12% at 48 h. These findings
   comprehensively address the in vivo stability, retention, and drug
   release properties of the G/D-CuP hydrogel, supporting its potential
   for therapeutic applications. Overall, G/D-CuP demonstrated superior
   mechanical properties like self-healing, strong adhesion, high
   malleability, and ROS responsiveness, making it a promising candidate
   for complex wound healing therapies.

Fig. 3. Characterization of the G/D-CuP.

   [140]Fig. 3
   [141]Open in a new tab

   a The preparation process of G/D-CuP. b Representative SEM image of the
   G/D-CuP (Scale bars = 50 µm). Rheology study of G/D-CuP, c
   Strain-sweep, d Frequency-sweep, e Continuous step strain. f
   Self-healing property of G/D-CuP. g The malleability of G/D-CuP. h
   Cumulative release curves of G/D-CuP under PBS and H[2]O[2] conditions.
   Data represented as Mean ± SD (n = 3 independent replicates).
   Statistical significance was assessed using two-way ANOVA with Šídák’s
   multiple comparisons test. Source data are provided as a Source Data
   file.

In vitro anti-inflammatory effects and antioxidant capacity of the G/D-CuP

   We evaluated the cellular uptake efficiency of D-CuP. As shown in
   Supplementary Fig. [142]12, fluorescence image analysis demonstrated
   that D-CuP achieved enhanced cellular internalization. Quantitative
   flow cytometry analysis indicated that D-CuP exhibited nearly 3.6-fold
   higher intracellular uptake compared to M-CuP, highlighting its
   capability for cellular entry by multivalency. Wound healing is a
   complicated process. High levels of ROS, excessive inflammatory cells
   and cytokines in the wound microenvironment, insufficient new blood
   vessels, and weakened cell proliferation are the key factors that lead
   to the difficult healing of diabetic wounds^[143]24. G/D-CuP was
   engineered to exert therapeutic effects across all phases of wound
   repair while targeting diverse pathological mechanisms, hence in vitro
   effects were validated. Macrophages exhibit a heterogeneous and plastic
   phenotype, capable of polarizing into the pro-inflammatory M1 or the
   reparative and anti-inflammatory M2 phenotypes. However, in diabetic
   wounds, the transition from M1 to M2 is disrupted, leading to the
   accumulation of M1 macrophages, which produce ROS and sustain long-term
   inflammation. As crucial regulators of local immune responses and
   tissue repair, macrophages in diabetic wounds show dysregulated
   switching from M1 to M2, resulting in a poor regenerative environment,
   exacerbated tissue damage, and delayed wound healing^[144]63. Firstly,
   CD206 was employed as the M2 macrophage marker to confirm macrophage
   polarization status using immunofluorescence staining performed under
   inflammatory stimuli (LPS and IFN-γ). The M1 phenotype, stimulated by
   lipopolysaccharide (LPS) and interferon-γ (IFN-γ), was used as a
   positive control, whereas the M2 phenotype, induced by interleukin-4
   (IL-4), functioned as a negative control. As illustrated in
   Fig. [145]4a, the G/D-CuP group exhibited the most intense
   CD206-associated red fluorescence signal among all treatment groups,
   indicating that G/D-CuP promotes macrophage polarization toward the M2
   phenotype. Furthermore, flow cytometry detected the changes of M1 (CD86
   marker) and M2 (CD206 marker) macrophage phenotypes after various
   treatments. Upon treatment of LPS and IFN-γ stimulated M1 macrophages
   with G/D-CuP, the population of M1 macrophages reduced to 36.4%
   (Fig. [146]4b), while the population of M2 macrophages increased to
   12.2%. The further analysis showed that G/D-CuP could significantly
   enhance the transition of macrophage phenotypes from M1 to M2
   (Fig. [147]4c, d), suggesting an improvement in anti-inflammatory
   capacity. D-CuP also showed a better effect on the transition of
   macrophage phenotypes effect than M-CuP. Next, we used the
   pro-inflammatory cytokines tumor necrosis factor-α (TNF-α) and
   interleukin-6 (IL-6) as indicators to evaluate the anti-inflammatory
   response. As illustrated in Fig. [148]4e, f, both G/D-CuP and D-CuP
   exhibited a significant reduction in the expression levels of
   pro-inflammatory cytokines. In chronic diabetic wounds, uncontrolled
   accumulation of reactive oxygen species (ROS) leads to localized
   irreversible oxidative damage and heightened inflammation^[149]64. The
   ability of G/D-CuP to scavenge reactive oxygen species (ROS) was
   evaluated by conducting Fenton-like reactions where hydrogen peroxide
   (H[2]O[2]) generates reactive hydroxyl radicals (•OH) with the help of
   metal catalysts, leading to significant oxidative harm^[150]65. As
   illustrated in Supplementary Fig. [151]13, the ROS scavenging ratio of
   G/D-CuP exceeded 95% at different concentrations, indicating a
   potential beneficial antioxidant effect for treating diabetic wounds.
   To further assess the ROS scavenging ability of G/D-CuP, the 2′,
   7′-dichlorofluorescein diacetate (DCFH-DA) fluorescent probe was used
   to measure intracellular ROS levels. As shown in Fig. [152]4g, the
   negative control group (NC) exposed to H[2]O[2] exhibited strong green
   fluorescence, while the hydrogel-treated group displayed weak green
   fluorescence similar to the blank control group (BC) without H[2]O[2].
   The groups receiving uncross-linked PVA or EBPBA solution showed no
   significant difference relative to the NC group, emphasizing the
   importance of phenylboronic ester bonds for scavenging ROS. M-CuP has
   antioxidant activity by increasing antioxidant enzymes and superoxide
   dismutase (SOD) activity, mainly neutralizing ROS by scavenging harmful
   oxygen-derived free radicals^[153]66. The dimeric copper peptide D-CuP
   showed superior antioxidant properties compared to M-CuP, leading to a
   notable decrease in relative fluorescence intensity. The mean
   fluorescence intensity for ROS-responsive hydrogel with monomer copper
   peptides M-CuP (G/M-CuP) and G/D-CuP groups decreased by 71% and 63%,
   respectively, in comparison to the NC group (Supplementary
   Fig. [154]14). SOD, a critical endogenous antioxidant enzyme, mitigates
   oxidative stress by catalyzing the dismutation of superoxide radicals
   via redox cycling, thus maintaining redox homeostasis^[155]67. The
   G/D-CuP group exhibited a markedly elevated SOD activity compared to
   the G/M-CuP group, indicating improved antioxidant capacity
   (Fig. [156]4h). The multivalency of the dimeric peptide structure
   contributes to the notable ROS-scavenging efficacy of G/D-CuP,
   confirming its great antioxidant capacity. These findings further
   validate the extraordinary antioxidant capacity of G/D-CuP, emphasizing
   its promise for use in wound repair. Hemocompatibility of the G/D-CuP
   was tested using a hemolysis assay. In Fig. [157]4i, the positive
   control (PC) group exhibited an intense red coloration, suggesting
   substantial erythrocyte lysis and hemoglobin release. In sharp
   contrast, neither the negative control (NC) group nor the G/D-CuP group
   exhibited any discoloration, suggesting hemolysis rates below 5% for
   these groups.

Fig. 4. In vitro anti-inflammatory effects and antioxidant capacity of the
G/D-CuP.

   [158]Fig. 4
   [159]Open in a new tab

   a Representative fluorescent image of the macrophage phenotypes after
   different treatments (red: CD206; blue: cell nucleus) (Scale
   bars = 20 µm). b Flow cytometry analysis of CD206 and CD86 expression
   of RAW 264.7 cells after different treatments. Statistical analysis of
   (c) M1(CD86 + , CD206-) or (d) M2(CD206 + , CD86-) phenotype
   macrophages. Data represented as Mean ± SD (n = 3 independent
   replicates). Statistical significance was assessed using one-way ANOVA
   with Tukey’s multiple comparisons test. The concentrations of (e)
   TNF-ɑ, (f) IL-6 were determined by ELISA. Data represented as Mean ± SD
   (n = 3 independent replicates). Statistical significance was assessed
   using one-way ANOVA with Tukey’s multiple comparisons test. g The ROS
   fluorescence images of NIH-3T3 cells after treatment with G/D-CuP.
   (Scale bars = 100 μm) (NC negative control group, BC blank control
   group). h Detection of SOD activities in NIH-3T3 cells after treatment
   with G/D-CuP. Data represented as Mean ± SD (n = 3 independent
   replicates) (NC negative control group, BC blank control group).
   Statistical significance was assessed using one-way ANOVA with Tukey’s
   multiple comparisons test. i Hemolysis ratio of G/D-CuP. Data
   represented as Mean ± SD. (n = 3 independent replicates) (PC positive
   control group, NC negative control group). Source data are provided as
   a Source Data file.

In vitro promotion of cell proliferation, migration and angiogenesis effects
of the G/D-CuP

   The formation of new blood vessels, called angiogenesis, is a crucial
   event during the proliferation stage of wound healing, and poor
   vascularization contributes to the delayed healing of diabetic wounds.
   This process entails endothelial cell proliferation, migration, and the
   formation of capillary-like tubular structures^[160]68.
   Immunofluorescence staining of Ki67, a marker for proliferating cells,
   was used to assess the proliferation of human umbilical vein
   endothelial cells (HUVECs)^[161]63. As shown in Fig. [162]5a, treatment
   with G/D-CuP elevated the number of Ki67-positive cells. Quantitative
   analysis of Ki67-positive cells showed that HUVEC proliferation after
   incubation with G/D-CuP was 69%, significantly higher than that of
   other groups (Fig. [163]5b). Fibroblasts synthesize extracellular
   matrix (ECM) components, glycoproteins, adhesion molecules, and a range
   of cytokines. Their ability to migrate, proliferate, and secrete ECM is
   essential for effective dermal tissue regeneration^[164]69. The
   proliferative activity towards fibroblasts was also evaluated using the
   MTT assay. As depicted in Fig. [165]5c and Supplementary Figs. [166]15,
   [167]16, D-CuP, ROS-responsive hydrogel matrix (G) and G/D-CuP
   demonstrated good biocompatibility. Compared to M-CuP, D-CuP exhibited
   significantly higher proliferative activity at lower concentrations,
   which was more conducive to wound healing. To assess the G/D-CuP’s
   biocompatibility and its ability to support cell growth, we performed
   in vitro three-dimensional (3D) culture experiments using NIH/3T3 cells
   as a model. After 24 h of co-culture with G/D-CuP, NIH/3T3 cells
   displayed well-spread morphology and high viability, as demonstrated by
   both 3D and 2D Calcein-AM fluorescence imaging via CLSM (Supplementary
   Fig. [168]17). These findings confirmed that G/D-CuP provides a
   supportive 3D microenvironment for fibroblast survival and
   proliferation, highlighting its potential for chronic wound healing and
   tissue regeneration.

Fig. 5. In vitro promotion of cell proliferation, migration, and angiogenesis
effects of the G/D-CuP.

   [169]Fig. 5
   [170]Open in a new tab

   a Representative fluorescence images of Ki67 staining of HUVEC cells
   after different treatments (green: Ki67; blue: cell nucleus, Scale
   bars = 20 µm). b Quantitative analysis of Ki67+ cells. Data represented
   as Mean ± SD (n = 3). Statistical significance was assessed using
   one-way ANOVA with Tukey’s multiple comparisons test. c Relative cell
   viability of NIH/3T3 cells co-cultured with D-CuP solution and M-CuP
   solution for 24 h. Data represented as Mean ± SD (n = 4 independent
   replicates, ns not significant). Statistical significance was assessed
   using two-way ANOVA with Tukey’s multiple comparisons test. d
   Representative images of tube formation assay in HUVEC cells stained
   with calcein-AM (green) (Tube: yellow arrow, Scale bars = 400 µm). e
   Quantitative analysis of total tube length. Data represented as
   Mean ± SD (n = 3). Statistical significance was assessed using one-way
   ANOVA with Tukey’s multiple comparisons test. f The concentrations of
   VEGF were determined by ELISA. Data represented as Mean ± SD (n = 3
   independent replicates). Statistical significance was assessed using
   one-way ANOVA with Tukey’s multiple comparisons test. g Representative
   images of transwell migration assay of L929 cells (Scale
   bars = 200 µm). h Quantitative analysis of total tube length. Data
   represented as Mean ± SD (n = 3). Statistical significance was assessed
   using one-way ANOVA with Tukey’s multiple comparisons test. i Images of
   L929 cells scratch experiment after incubating with D-CuP or M-CuP at
   different times (Scale bars = 100 μm). j Cell scratch healing rates at
   different times. Data represented as Mean ± SD (n = 3 independent
   replicates). Statistical significance was assessed using two-way ANOVA
   with Tukey’s multiple comparisons test. Source data are provided as a
   Source Data file.

   An endothelial tube formation was also evaluated. As shown in
   Fig. [171]5d, e and Supplementary Fig. [172]18, a substantial number of
   blood vessel structures were observed in the G/D-CuP group, with both
   tube length and the number of meshes significantly higher than those in
   other treatment groups. Vascular endothelial growth factor (VEGF)
   functions as a powerful mitogen for endothelial cells and a crucial
   regulator of angiogenesis^[173]70. ELISA results of VEGF further
   displayed that G/D-CuP enhanced the ability of angiogenesis
   (Fig. [174]5f). Overall, G/D-CuP demonstrated an ability to promote
   neovascularization, thereby accelerating wound healing.

   Impaired cell migration, as a critical factor, affects diabetic wound
   healing during the proliferative and remodeling phases. To evaluate the
   cell migration, a transwell assay was conducted. Apparently, G/D-CuP
   significantly promoted fibroblast migration (Fig. [175]5g, h). For a
   more detailed understanding of the wound repair efficacy, an in vitro
   scratch assay was carried out using L929 cells. Wound closure was
   simulated and observed for 30 h following treatment of the scratched
   cell monolayer with M-CuP or D-CuP. As shown in Fig. [176]5i, j, cell
   migration was elevated in all treatment groups relative to the control.
   Notably, D-CuP exhibited faster cell migration than M-CuP. The
   increased efficacy is likely due to the enhanced stability and boosted
   bioactivity of D-CuP throughout the scratch wound repair process. The
   above findings indicate that D-CuP exhibits good biocompatibility,
   hemocompatibility, anti-inflammatory, antioxidant capabilities, and
   promotes cell proliferation, migration and angiogenesis.

Proteomics analysis of therapeutic performance of D-CuP

   To elucidate the regulatory mechanisms of D-CuP, we conducted a
   quantitative proteomics analysis on murine L929 fibrosarcoma cells
   before and after treatment under nutrient stress (Fig. [177]6a).
   Initially, the alterations in intracellular protein concentrations
   post-treatment were assessed, and the principal component analysis
   (PCA) outcomes demonstrated a notable divergence in protein expression
   between the D-CuP-treated group and the PBS-treated control group
   (Fig. [178]6b). Compared to the PBS-treated control group, the
   D-CuP-treated group exhibited 74 differentially expressed proteins
   (DEPs), consisting of 29 upregulated and 45 downregulated proteins
   (p < 0.05, fold change ratio >1.2) (Fig. [179]6c). We categorized the
   primary differential proteins into three phases of wound healing:
   inflammatory, proliferative, and remodeling (Fig. [180]6d).
   Specifically, proteins associated with DNA damage, inflammation, and
   oxidative stress, including Cop1, H2-D1, Ndufa6, and Fam76a, showed
   downregulation following D-CuP treatment. Moreover, the levels of
   proteins associated with cell proliferation, migration, and
   angiogenesis (Slc38a2, Psph, Jagn1, and Rorb) were upregulated. A Gene
   Ontology (GO) functional enrichment analysis of differentially
   expressed proteins (DEPs) was also carried out, as depicted in
   Fig. [181]6e and Supplementary Fig. [182]19. The analysis revealed that
   within the molecular function category, molecular function regulator
   and transporter activity were prominent. In the cellular component
   category, cell and membrane were dominant. Moreover, within the
   biological process category, cellular process, regulation of biological
   process, and biological regulation were identified as key processes.
   Furthermore, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway
   enrichment analysis of DEPs mainly included nucleotide excision repair
   and p53 signaling pathway (Fig. [183]6f). Additionally, the mTOR,
   IL-17, NOD-like receptor and HIF-1 signaling pathways were involved
   (Supplementary Fig. [184]20). These findings suggest that D-CuP
   exhibits anti-inflammatory and antioxidant properties, enhances cell
   proliferation and migration, and thereby supports diabetic wound
   healing. To further elucidate the therapeutic mechanisms underlying
   D-CuP’s multivalent interactions, comprehensive proteomic analysis
   revealed two pivotal transcriptional regulators, the tumor suppressor
   protein p53 and histone deacetylase 7 (HDAC7) to be potential
   mechanistic targets. Downregulation of the p53 signaling pathway
   promotes cellular proliferation and modulates inflammatory
   responses^[185]71, while HDAC7 depletion facilitates cell migration,
   angiogenesis, and extracellular matrix remodeling, collectively
   contributing to coordinated tissue repair^[186]72. To evaluate the
   target-binding capacity of D-CuP, molecular docking simulations were
   conducted using AutoDock Vina against the functional domains of p53
   (DNA-binding domain) and HDAC7 (catalytic deacetylase domain).
   Comparative analysis showed that D-CuP exhibited higher binding
   affinities than its monomeric counterpart M-CuP, with Vina scores of
   -5.7 kcal/mol for p53 and -6.4 kcal/mol for HDAC7, compared to
   -4.2 kcal/mol and -4.8 kcal/mol for M-CuP, respectively (Fig. [187]6g).
   To further validate the multivalent interactions of D-CuP, SPR analysis
   was performed to quantify its binding ability to p53 and HDAC7 in
   comparison with M-CuP. As shown in Fig.[188]6h, D-CuP exhibited
   significantly stronger binding affinities, with equilibrium
   dissociation constants (KD) of 3.8 × 10⁻⁶ M for p53 and 1.81 × 10⁻⁶ M
   for HDAC7. In contrast, M-CuP showed negligible binding to either
   target, with maximal responses below 5 response units (RU) across all
   tested concentrations, confirming the essential role of multivalent
   architecture in cooperative target engagement.

Fig. 6. Proteomics assessment of L929 cells before and after D-CuP treatment.

   [189]Fig. 6
   [190]Open in a new tab

   a Diagrammatic sketch of proteomic analysis. b Principal component
   analysis (PCA) was performed based on differentially expressed proteins
   (DEPs) in the L929 cells of the two groups. c Volcano plots showing the
   upregulated and downregulated genes in response to D-CuP treatment.
   Statistical significance was assessed using Student’s t test. d Heat
   map of upregulated and downregulated proteins. (fold change ratio >1.2
   and p <0.05). Statistical significance was assessed using Student’s t
   test. e GO function analysis of DEPs between D-CuP-treated and the
   control cells. f KEGG enrichment for the DEPs between D-CuP-treated and
   the control cells. g Pattern diagram illustrating the molecular docking
   interactions between p53 and M-CuP, p53 and D-CuP, HDAC7 and M-CuP, and
   HDAC7 and D-CuP. h SPR analysis of the binding affinities of M-CuP and
   D-CuP with p53 or HDAC7. Source data are provided as a Source Data
   file.

In vivo effects of the G/D-CuP

   Motivated by the favorable biocompatibility, antioxidant capability,
   and the facilitation of fibroblast cell proliferation and migration, we
   evaluated the efficacy of G/D-CuP hydrogels for diabetic wound healing
   in vivo (Fig. [191]7a). To evaluate the healing efficacy, a
   comprehensive experiment was conducted on diabetic wounds treated with
   various agents, including: i) PBS as negative control (NC group), ii)
   ROS-responsive hydrogel matrix (G group), iii) monomer copper peptides
   M-CuP solution (M-CuP group), iv) dimeric copper peptides D-CuP
   solution (D-CuP group), v) ROS-responsive hydrogel with monomer copper
   peptides M-CuP (G/M-CuP group), and vi) ROS-responsive hydrogel with
   dimeric copper peptides (G/D-CuP group). Healthy mice wounds were
   treated with PBS as the blank control (BC group). Representative images
   of the wounds in various groups from days 0 to 12 were depicted in
   Fig. [192]7b, c. Macroscopically, wound healing was delayed in
   diabetes, and abscesses were observed in the NC group. All groups
   exhibited a significant reduction in wound area after 3 days of
   treatment compared to the NC group. G/D-CuP exhibited a specific
   decrease in wound area from 100% to 17.3% after 7 days (Fig. [193]7d),
   surpassing the reduction observed in the other groups. Notably, on day
   12, significantly large wound areas remained in both the NC and M-CuP
   groups, comprising 34.2% and 17.8%, respectively. Conversely, the D-CuP
   group demonstrated enhanced wound healing, leaving a residual wound
   area of just 9.8%, while the G/D-CuP group exhibited the most
   pronounced wound healing, with only 2.8% of the wound area remaining.
   Histological analysis was further conducted on days 7 and 12 to
   evaluate the quality of the regenerated skin. The H&E staining results
   (Fig. [194]7e and Supplementary Fig. [195]21) showed that the wound
   treated with G/D-CuP exhibited complete recovery, characterized by
   densely arranged granulation tissue, as well as increased the presence
   of skin appendages, including hair follicles (HFs) and sebaceous
   glands. As shown in Fig. [196]7f, the G/D-CuP also enhanced the
   thickness of granulation tissue, thereby enhancing diabetic wound
   healing. G/D-CuP facilitates the proliferation of dermal papilla cells,
   reduces their apoptosis, fosters hair follicle growth and regeneration,
   and enhances wound healing. Moreover, an increased recruitment of
   adipocytes to the wound site further supports simultaneous tissue
   regeneration^[197]37.

Fig. 7. Effects of the G/D-CuP in promoting diabetic wound healing.

   [198]Fig. 7
   [199]Open in a new tab

   a Schematic illustration of the timeline of animal experiments to test
   the therapeutic effect of different groups. b Representative
   photographs of the wounds in different treatment groups (Scale
   bar = 3 mm). c Portions of wounds healed under various treatments on
   days 0, 3, 7, and 12. d Wound healing rate within 12 days. e H&E
   staining of wound tissue with different treatments on the 12th day
   (sebaceous glands: yellow arrow, hair follicles: green arrow. Scale
   bar = 250 µm). f Granulation tissue thickness in each group on the 12th
   day. Data represented as Mean ± SD (n = 3 independent replicates).
   Statistical significance was assessed using one-way ANOVA with Tukey’s
   multiple comparisons test. Source data are provided as a Source Data
   file.

   The effect of the G/D-CuP on the inflammation, proliferation, and
   remodeling stages of tissue repair was also assessed. The initial
   inflammatory phase of wound healing, typically lasting 2–3 days in
   physiological wounds, is extended in chronic wounds^[200]73. This phase
   is consistently characterized by elevated levels of inflammatory
   cytokines. To evaluate levels of inflammatory cytokines, expression
   levels of interleukin-6 (IL-6) and interleukin-10 (IL-10) were measured
   by ELISA after three days of treatment. In diabetic wounds, the G/D-CuP
   group exhibited lower IL-6 expression and higher IL-10 expression
   compared to the G/M-CuP group (Fig. [201]8a, b), suggesting a shorter
   inflammatory phase after G/D-CuP treatment. Histological analysis on
   day 7 revealed a significant decrease in the expression of
   proinflammatory cytokines in the wound, including TNF-α and IL-6 in the
   D-CuP groups, accompanied by an increase in IL-10 expression
   (Fig. [202]8c and Supplementary Fig. [203]21). This result indicated
   that dimeric copper peptides can accelerate the switch from the
   inflammatory to the proliferative phase by regulating the expression of
   inflammatory factors. G/D-CuP could have superior efficacy in
   modulating inflammatory factors compared to G/M-CuP, likely due to the
   multivalency effect of dimeric copper peptides, which enhanced the
   activity.

Fig. 8. G/D-CuP-induced acceleration of the three phases of diabetic wound
repair.

   [204]Fig. 8
   [205]Open in a new tab

   The concentrations of (a) IL-6, (b) IL-10, and d VEGF in the wound
   tissue were determined by ELISA on day 3. Data represented as
   Mean ± SD. (n = 3 independent replicates) Statistical significance was
   assessed using one-way ANOVA with Tukey’s multiple comparisons test. c
   Representative images of immunohistochemistry staining of TNF-α, VEGF,
   CD31, and Ki67 (Scale bars = 100 μm) on day 7 and Masson staining on
   day 12 (Scale bars = 500 μm). Quantitative analysis of
   immunohistochemical staining for (e) CD31 and (f) Ki67 on day 7. Data
   represented as Mean ± SD. (n = 3 independent replicates) Statistical
   significance was assessed using one-way ANOVA with Tukey’s multiple
   comparisons test. g Quantification of the collagen deposition in
   different groups on day 12. Data represented as Mean ± SD (n = 3
   independent replicates). Statistical significance was assessed using
   one-way ANOVA with Tukey’s multiple comparisons test. Source data are
   provided as a Source Data file.

   Immunofluorescence staining was performed to assess macrophage
   polarization, with CD86 serving as a marker for pro-inflammatory M1
   macrophages and CD206 for anti-inflammatory M2 macrophages. As shown in
   Supplementary Fig. [206]22, the G/D-CuP group showed a significantly
   lower proportion of CD86-positive cells compared to the negative
   control (NC) group. Notably, the G/D-CuP group demonstrated the lowest
   CD86/CD206 ratio, further indicating a promoted transition from the M1
   to M2 phenotype (Supplementary Fig. [207]23). Together, these findings
   suggest that G/D-CuP promotes macrophage polarization toward the M2
   phenotype, thereby reducing inflammation, enhancing antioxidant
   capacity, and ultimately accelerating effective wound healing. As shown
   in Supplementary Fig. [208]22, G/D-CuP group exhibited a significantly
   reduced proportion of CD86-positive cells compared to the NC group.
   Notably, the G/D-CuP group showed the lowest ratio of CD86-positive
   macrophages/CD206-positive macrophages, indicating an enhanced M1-to-M2
   transition (Supplementary Fig. [209]23). Immunofluorescence staining
   and semi-quantitative analysis demonstrated that the G/D-CuP group
   significantly upregulated SOD expression compared with the G/M-CuP
   group (Supplementary Figs. [210]24, [211]25). Collectively, these
   findings suggest that G/D-CuP facilitates macrophage phenotype
   switching and enhances antioxidant capacity, thereby mitigating
   inflammation and ultimately accelerating effective wound healing.

   During the proliferation phase, neovascularization is essential for
   delivering nutrients and oxygen to the wound site, thereby promoting
   fibroblast proliferation, collagen synthesis, and
   re-epithelialization^[212]73. Vascular endothelial growth factor
   (VEGF), a typical marker of angiogenesis during wound healing, was
   quantitatively assessed via ELISA on day 3, revealing a significant
   increase in the G/D-CuP group (Fig. [213]8d). Immunohistochemical
   analysis shown in Fig. [214]8c, e, Supplementary Figs. [215]21 and
   [216]26 illustrated a substantial elevation in the expression of both
   VEGF and platelet endothelial cell adhesion molecule-1 (CD31) on day 7
   after treatment with G/D-CuP, compared to the negative control and Gel
   groups. Specifically, quantitative analysis revealed that the relative
   average density (AOD) of CD31 was 1.48-fold higher in the G/D-CuP group
   than in the G/M-CuP group. The initiation of vascularization at an
   earlier stage of wound healing can result in a faster rate of wound
   closure. However, excessive vascularization during the mid-phase of
   wound healing impedes fibroblast growth and wound tissue
   re-epithelialization^[217]74. Furthermore, excessive vascularization
   during wound healing may lead to adverse outcomes, such as wound
   fibrosis and scar formation. On the 12th day, the G/D-CuP group
   exhibited significantly fewer blood vessels, while CD31 expression
   persisted in the M-CuP and G/M-CuP groups. This suggested that wound
   healing was nearly complete in the G/D-CuP group with minimal new blood
   vessel formation on day 12, whereas the M-CuP and G/M-CuP groups, still
   in the healing process, exhibited increased angiogenesis. These
   findings indicate that G/D-CuP may significantly promote angiogenesis,
   as evidenced by increased CD31 expression during the initial stages of
   wound healing. Consequently, G/D-CuP has the potential as a wound
   dressing to expedite the remodeling phase onset and mitigate excessive
   vascularization.

   During the remodeling phase, cell proliferation, proper collagen
   deposition, and remodeling play crucial roles in enhancing tissue
   tensile strength and facilitating improved healing^[218]75. Ki67, a key
   marker of cell proliferation, was examined in wound samples on day 7
   and day 12 to evaluate cell proliferation. Immunostaining results
   showed an increase in Ki67 expression in the G/D-CuP group compared to
   other groups, while the control group exhibited minimal positive cells
   on day 7 (Fig. [219]8c, f, Supplementary Figs. [220]21 and [221]26).
   Interestingly, this pattern reversed on day 14, with delayed
   proliferative fibroblastic cells in the NC group and few detected in
   the G/D-CuP group, indicating that the G/D-CuP treatment facilitates
   wound healing through a time-dependent mechanism. Masson’s trichrome
   staining was utilized to evaluate collagen deposition within the wound
   healing site. Collagen fibers in the G/D-CuP group exhibited increased
   density, thickness, and organization compared to the G/M-CuP group,
   suggesting enhanced collagen deposition (Fig. [222]8g). These results
   highlight the superior wound healing ability of G/D-CuP, with healed
   tissue resembling healthy skin, surpassing outcomes from other
   treatments.

Discussion

   Chronic diabetic wounds represent a huge healthcare challenge,
   necessitating sustained and synchronized regulation of the wound
   microenvironment simultaneously with tissue reconstruction processes.
   However, current wound management approaches face challenges in
   coordinating the individual management approaches within the disordered
   diabetic wound microenvironment. Therefore, developing a comprehensive
   therapeutic platform capable of spatiotemporal ROS scavenging,
   regulation of macrophage polarization, and enhancement of angiogenesis
   and cell proliferation represents a critical research focus for
   improving clinical outcomes in chronic diabetic wound management.

   In this study, a multifunctional dimeric copper peptide incorporated
   hydrogel, G/D-CuP has been developed as an innovative wound dressing to
   promote diabetic wound healing. The hydrogel dressing comprised PVA, a
   ROS-responsive linker EBPBA, and dimeric copper peptide D-CuP. D-CuP
   exhibited multivalency, significantly enhancing biological activity and
   stability against proteases. The dynamic phenylboronic ester hydrogel,
   with ROS responsiveness and scavenging capabilities, helps alleviate
   inflammation and serves as a drug reservoir for the controlled release
   of D-CuP at the wound site. The hydrogel displayed repairability,
   malleability, and ideal mechanical strength, crucial for complex wound
   healing treatments. Notably, G/D-CuP exerted anti-inflammatory and
   antioxidative properties, promoting angiogenesis, fibroblast
   proliferation, and migration to accelerate wound healing without the
   need for additional bioactive substances. In vivo studies demonstrated
   that the G/D-CuP accelerated the three phases of diabetic wound repair
   (inflammation, proliferation, and remodeling). The G/D-CuP was also
   biocompatible, easily synthesized, and showed promise as a dressing
   material for managing diabetic wounds. Given the enhanced therapeutic
   efficacy of D-CuP, it is a potential substitute for M-CuP in wound
   repair, cosmetics, and other applications. Furthermore, the design
   presents a promising approach for leveraging bioactive linear peptides.

Methods

Ethics declarations

   All the animal experiments were performed following the Guide for Care
   and Use of Laboratory Animals, approved by the Animal Experimentation
   Ethics Committee of China Pharmaceutical University (2022-10-017). This
   study involved only non-invasive application of a test hydrogel on the
   dorsal hand skin of a healthy adult volunteer for photographic
   documentation. The study did not involve any clinical procedures or
   personal data collection. According to the policy of the Animal
   Experimentation Ethics Committee of China Pharmaceutical University,
   such minimal-risk procedures are exempt from ethical approval. Written
   informed consent was obtained from the participant for both
   participation and image publication.

Materials

   H-His(Trt)-OMe·HCl, H-Lys(Boc)-OMe·HCl, H-Lys-OMe·2HCl,
   ((1H-Benzo[d][1,2,3]triazol-1-yl)oxy)tri(pyrrolidine-1-yl)phosphonium
   hexafluorophosphate(V) (PYBOP),
   2-(1H-benzotriazole-1-yl)-1,1,3,3-tetramethyluronium
   hexafluorophosphate (HBTU), 1-hydroxy benzotriazole hydrate (HOBT), GHK
   and 4-carboxyphenyl boric acid were obtained from Bide Pharmatech Ltd
   (Shanghai, China). N, N-diisopropylethylamine (DIPEA), Protease K, 30%
   hydrogen peroxide (H[2]O[2]) solution, and Ethylenediamine
   dihydrochloride were purchased from Shanghai Aladdin Biochemical
   Technology Co., Ltd. (Shanghai, China). Fetal bovine serum (FBS) was
   purchased from VivaCell (Shanghai, China). Reactive Oxygen Species
   (ROS) kit, DAPI and Calcein AM were purchased from Beyotime
   Biotechnology Corp., Ltd. (Shanghai, China). DMEM high glucose medium
   and ELISA kits were purchased from Keygen Biotech Corp., Ltd. (Jiangsu,
   China). Streptozotocin (STZ) was purchased from OKA (Beijing, China).
   Matrigel was purchased from Yeasen Biotechnology Co., Ltd. (Shanghai,
   China).

Synthesis of D-P

   H-His(Trt)-OMe·HCl (5.0 g, 11.15 mmol), Boc-Gly-OH (2.15 g,
   12.25 mmol), HATU (8.5 g, 22.35 mmol) and HOAT (3.0 g, 22.35 mmol) were
   dissolved in DMF (80 mL) under nitrogen atmosphere and ice bath
   conditions. DIPEA (7.75 mL, 44.7 mmol) was slowly added to the mixture.
   The reaction mixture was stirred at room temperature for 24 h. After
   completion of the reaction, the solvent was removed under reduced
   pressure, and the residue was dissolved in ethyl acetate. The organic
   phase was washed three times sequentially with saturated NaHCO[3] (aq),
   0.1 M HCl, and saturated NaCl (aq), respectively. The organic layer was
   dried over anhydrous Na₂SO₄, filtered, and concentrated under reduced
   pressure. Compound 1 was isolated by silica gel column chromatography
   using a dichloromethane/methanol (DCM:MeOH = 60:1, V/V) eluent (yield:
   80%).

   ^1H NMR (300 MHz, DMSO) δ 8.26 (d, J = 7.6 Hz, 1H), 7.45 – 7.34 (m,
   9H), 7.25 (d, J = 1.4 Hz, 1H), 7.13 – 7.03 (m, 6H), 7.03 – 6.95 (m,
   1H), 6.70 – 6.61 (m, 1H), 4.49 (q, J = 6.9 Hz, 1H), 3.54 (s, 3H), 3.52
   (d, J = 5.9 Hz, 1H), 2.82 (dd, J = 6.7, 3.4 Hz, 2H), 1.35 (s, 9H). MS
   (ESI) m/z: 568.27, found 569.28 [M + H]^+.

   Compound 1 (6 g, 10.53 mmol) was dissolved in 50 mL MeOH with NaOH
   (2 g, 50 mmol) under ice bath conditions. The reaction mixture was
   stirred at room temperature for 6 h. After completion of the reaction,
   the pH was adjusted to pH = 2-3 with citric acid solution (pH = 2). The
   solvent was removed under reduced pressure. The residue was diluted
   with ethyl acetate and washed with saturated NaCl (aq). The organic
   layer was dried over anhydrous Na₂SO₄, filtered, and concentrated under
   reduced pressure to obtain Compound 2 (yield: 98%).

   ^1H NMR (400 MHz, DMSO) δ 8.74 (s, 1H), 8.22 (d, J = 8.3 Hz, 1H), 7.50
   – 7.39 (m, 9H), 7.22 (s, 1H), 7.18 – 7.09 (m, 6H), 6.91 (t, J = 6.0 Hz,
   1H), 4.57 (td, J = 9.0, 4.8 Hz, 1H), 3.50 (d, J = 10.9 Hz, 2H), 3.14
   (dd, J = 15.0, 4.9 Hz, 1H), 2.94 (dd, J = 15.0, 9.8 Hz, 1H), 1.34 (s,
   9H). MS (ESI) m/z:554.25, found 555.26 [M + H]^+.

   Compound 2 (5.4 g, 9.72 mmol), H-Lys(Boc)-OMe·HCl (3.46 g, 11.52 mmol),
   HATU (7.42 g, 19.44 mmol) and HOAT (2.64 g, 19.44 mmol) were dissolved
   in DMF (80 mL) under nitrogen atmosphere and ice bath conditions. DIPEA
   (6.66 mL, 38.88 mmol) was slowly added into the mixture. The reaction
   mixture was stirred at room temperature for 24 h. After completion of
   the reaction, the solvent was removed under reduced pressure, and the
   residue was dissolved in ethyl acetate. The organic phase was washed
   three times sequentially with saturated NaHCO[3] (aq), 0.1 M HCl, and
   saturated NaCl (aq), respectively. The organic layer was dried over
   anhydrous Na₂SO₄, filtered, and concentrated under reduced pressure.
   Compound 3 was isolated by silica gel column chromatography using a
   dichloromethane/methanol (DCM:MeOH = 40:1, V/V) eluent (yield: 75%).

   ^1H NMR (300 MHz, DMSO) δ 8.23 (d, J = 7.4 Hz, 1H), 8.12 (d,
   J = 8.2 Hz, 1H), 7.41 (d, J = 6.6 Hz, 9H), 7.23 (s, 1H), 7.12 – 7.04
   (m, 6H), 6.78 (d, J = 5.4 Hz, 1H), 6.67 (d, J = 11.0 Hz, 1H), 4.56 (d,
   J = 6.6 Hz, 1H), 4.14 (q, J = 7.4 Hz, 1H), 3.60 (s, 3H), 3.51 (d,
   J = 8.0 Hz, 2H), 2.78 (dq, J = 32.2, 7.5 Hz, 4H), 1.36 (d, J = 2.4 Hz,
   24H). MS (ESI) m/z: 796.42, found 797.43 [M + H]^+.

   Compound 3 (5.0 g, 6.20 mmol) was dissolved in 50 mL MeOH with NaOH
   (2 g, 50 mmol) under ice bath conditions. The reaction mixture was
   stirred at room temperature for 6 h. After completion of the reaction,
   the pH was adjusted to pH = 2-3 with citric acid solution (pH = 2). The
   solvent was removed under reduced pressure. The residue was diluted
   with ethyl acetate and washed with saturated NaCl (aq). The organic
   layer was dried over anhydrous Na₂SO₄, filtered, and concentrated under
   reduced pressure to obtain Compound 4 (yield: 90%).

   ^1H NMR (400 MHz, DMSO) δ 8.72 (s, 1H), 8.25 (s, 1H), 8.02 (d,
   J = 8.3 Hz, 1H), 7.46 (d, J = 6.5 Hz, 9H), 7.14 (dd, J = 7.3, 2.5 Hz,
   6H), 7.08 (s, 1H), 6.94 (t, J = 5.7 Hz, 1H), 6.75 (t, J = 5.8 Hz, 1H),
   4.71 (s, 1H), 4.13 (td, J = 8.2, 5.2 Hz, 1H), 3.52 (d, J = 5.7 Hz, 2H),
   3.07 (d, J = 14.8 Hz, 1H), 2.85 (dd, J = 16.6, 8.5 Hz, 3H), 1.35 (s,
   24H). MS (ESI) m/z: 782.40, found 783.42 [M + H]^+.

   Compound 4 (1.94 g, 2.47 mmol), H-Lys-OMe·2HCl (0.24 g, 1.03 mmol),
   PYBOP (1.62 g, 3.10 mmol) and HOBT (0.42 g, 3.10 mmol) were dissolved
   in DMF (40 mL) under nitrogen atmosphere and ice bath conditions. DIPEA
   (1.08 mL, 6.20 mmol) was slowly added into the mixture. The reaction
   mixture was stirred at room temperature for 24 h. After completion of
   the reaction, the solvent was removed under reduced pressure, and the
   residue was dissolved in ethyl acetate. The organic phase was washed
   three times sequentially with saturated NaHCO[3] (aq), 0.1 M HCl, and
   saturated NaCl (aq), respectively. The organic layer was dried over
   anhydrous Na₂SO₄, filtered, and concentrated under reduced pressure.
   Compound 5 was isolated by silica gel column chromatography using a
   dichloromethane/methanol (DCM:MeOH = 20:1, V/V) eluent (yield: 61%).

   ^1H NMR (500 MHz, DMSO) δ 7.95 (t, J = 38.4 Hz, 5H), 7.41 (p, J = 5.1,
   3.9 Hz, 18H), 7.23 (d, J = 7.4 Hz, 2H), 7.09 (d, J = 7.4 Hz, 12H), 6.96
   (d, J = 19.0 Hz, 2H), 6.70 (q, J = 9.2, 8.1 Hz, 4H), 4.56 (d,
   J = 54.2 Hz, 2H), 4.18 (d, J = 69.9 Hz, 3H), 3.56 (tt, J = 18.3,
   9.3 Hz, 7H), 3.03 – 2.77 (m, 10H), 1.66 – 1.20 (m, 54H). MS (MALDI)
   m/z: 1688.90, found 1712.35 [M + Na]^+.

   Compound 5 (1 g, 0.59 mmol) was dissolved in DCM (2.6 mL) under
   nitrogen atmosphere and ice bath conditions. Trifluoroacetic acid
   (2.6 mL, 35.53 mmol) was slowly added into the mixture. The reaction
   mixture was stirred at room temperature for 6 h. After completion of
   the reaction, the solvent was removed under reduced pressure. The
   organic layer was dried over anhydrous Na₂SO₄, filtered, and
   concentrated under reduced pressure to obtain Compound 6 (yield: 85%).

   ^1H NMR (400 MHz, DMSO) δ 8.96 (s, 2H), 8.83 – 8.63 (m, 2H), 8.44 (dd,
   J = 23.9, 7.0 Hz, 1H), 8.36 – 8.24 (m, 2H), 8.02 (s, 4H), 7.82 (s, 3H),
   7.39 – 7.28 (m, 2H), 4.76 (tt, J = 16.1, 8.0 Hz, 2H), 4.37 – 4.24 (m,
   1H), 4.15 (d, J = 9.7 Hz, 2H), 3.38 (q, J = 7.0 Hz, 4H), 3.13 – 2.87
   (m, 6H), 2.74 (s, 4H), 1.75 – 1.13 (m, 18H), 1.09 (t, J = 7.0 Hz, 5H).
   MS (ESI) m/z:804.47, found 403.25 [M/2 + H]^+.

   Compound 6 (0.80 g, 0.99 mmol) was dissolved in 20 mL MeOH with NaOH
   (0.80 g, 20 mmol) under ice bath conditions. The reaction mixture was
   stirred at room temperature for 6 h. After completion of the reaction,
   the pH was adjusted to pH = 5-6 with hydrochloric acid solution (pH =
   2). The solvent was removed under reduced pressure. The residue was
   diluted with ethyl acetate (100 mL) and deionized water (200 mL). The
   mixture was then sonicated to form a suspension, centrifuged, and the
   supernatant was collected and concentrated to 10 mL. The concentrated
   solution was dialyzed with deionized water for 72 h, and then
   lyophilized to obtain D-P (yield: 90%).

   ^1H NMR (400 MHz, D[2]O) δ 7.78 (s, 2H), 7.01 (s, 2H), 4.71 – 4.60 (m,
   2H), 4.44 – 4.03 (m, 3H), 3.80 – 3.74 (m, 4H), 3.32 – 2.88 (m, 10H),
   1.85 – 1.11 (m, 18H)^13.C NMR (151 MHz, D[2]O) δ 178.88, 172.79,
   172.16, 172.10, 167.38, 167.21, 134.76, 134.56, 129.96, 129.66, 117.45,
   54.20, 54.17, 53.68, 53.46, 53.08, 40.64, 40.59, 39.33, 39.28, 37.38,
   30.79, 30.39, 30.37, 28.05, 27.65, 27.48, 27.40, 22.63, 22.18, 22.04.
   HRMS (m/z): [M]^+ calcd. for C[34]H[58]N[14]O[8], 790.456; found,
   790.4546; analysis (calcd., found for C[34]H[58]N[14]O[8]): C (51.63,
   51.62), H (7.39, 7.40), O (16.18,16.19), N (24.79, 24.80).

Synthesis of
(((ethane-1,2-diylbis(azanediyl))bis(carbonyl))bis(4,1-phenylene))diboronic
acid(EBPBA)

   4-Carboxyphenylboronic acid (3.65 g, 22.00 mmol), ethylenediamine
   dihydrochloride (1.33 g, 10.00 mmol), HATU (11.40 g, 30.00 mmol) and
   HOAT (4.08 g, 30.00 mmol) were dissolved in DMF (40 mL) under nitrogen
   atmosphere and ice bath conditions. DIPEA (10.45 mL, 60.00 mmol) was
   slowly added to the mixture. The reaction mixture was stirred at room
   temperature for 12 h. After completion of the reaction, the solvent was
   removed under reduced pressure. The mixture was diluted with deionized
   water (200 mL) and filtered. The solid product was collected and dried
   under vacuum to obtain EBPBA.

   ^1H NMR (300 MHz, DMSO) δ 8.63 (d, J = 5.2 Hz, 2H), 8.09 (d,
   J = 72.6 Hz, 4H), 7.83 (q, J = 8.2 Hz, 8H), 3.44 (d, J = 4.5 Hz, 4H).
   MS (ESI) m/z:356.14, found 379.13 [M + Na]^+.

Preparation and characterization of D-CuP

   24 mM D-P solution and 24 mM CuSO[4] solution were mixed in equal
   volumes (the D-P: Cu^2+ ratio was 1:0.5) and stirred overnight at room
   temperature to obtain 12 mM D-CuP solution. The absorption spectrum of
   12 mM D-P, D-CuP, CuSO[4], and 24 mM M-CuP aqueous solutions was
   obtained using a UV–Vis spectrophotometer (MAPADA, UV-1800PC). Electron
   paramagnetic resonance (EPR) spectroscopy was conducted using a Bruker
   A300-10/12 to verify the successful coordination of Cu^2+ with D-P.

pH stability of M-CuP and D-CuP

   24 mM D-CuP and 12 mM M-CuP aqueous solutions (1 mL) were separately
   incubated with 1 mL of PBS at varying pH values (pH 4 to 9) at 37 °C
   for 24 h. To simulate extreme acidic conditions, PBS (pH=1) was used as
   the positive control group (PC). The absorbance of D-CuP and M-CuP
   solutions at 584 nm was measured using a UV–Vis spectrophotometer.

In vitro biostability of M-CuP and D-CuP against Proteinase K

   1 mM D-CuP and 2 mM M-CuP aqueous solutions (1 mL) were incubated
   separately with Proteinase K (10 U mL^−1, 100 µL) at 37 °C for 0, 1, 2
   and 4 h. Methanol (400 µL) was then added to stop the reaction. The
   samples were centrifuged at 13000 g for 10 min at 4 °C and the
   supernatant was collected and detected by HPLC to analyze the
   biostability of M-CuP and D-CuP against proteinase K.

Preparation of G/D-CuP

   15% PVA-1788 (w/w) solution (1 mL), 12 mM D-CuP solution (650 µL), and
   EBPBA solution (5 mg mL^−1, 350 µL) were mixed thoroughly and stirred
   rapidly to obtain G/D-CuP.

Rheological properties and characterization of G/D-CuP

   The rheological properties of G/D-CuP were evaluated using a rheometer
   (TA Instruments, DHR-2). The strain sweep was conducted at a constant
   angular frequency of 3.14 rad s^−1 and a strain range of 0.1% to 1200%.
   Frequency sweep measurements were performed at 5% strain and a
   frequency range of 500 to 0.1 rad s^−1. To assess self-healing
   behavior, a continuous step-strain test was conducted by alternating
   between low strain (1%, 65 s) and high strain (1200%, 65 s) for three
   cycles. The morphologies of lyophilized hydrogels were examined by
   scanning electron microscopy (SEM, Hitachi S-3400N) at an accelerating
   voltage of 5 kV. The surface of lyophilized hydrogel was sprayed with
   platinum to increase the electrical conductivity.

D-CuP release from G/D-CuP

   G/D-CuP was loaded into a dialysis bag (MWCO = 3.5 kDa) and immersed in
   PBS (pH 7.4) without or with 1 mM H[2]O[2]. The samples were placed in
   a shaker at 37 °C and shaken at 80 rpm. The release medium was
   collected at predetermined time points and replenished with fresh
   medium. The D-CuP release efficiency was measured by UV-vis
   spectrophotometer at 600 nm.

Cell lines

   Human endothelial cells HUVEC (CC-Y1285) were purchased from
   EK-Bioscience (Shanghai, China). Mouse fibroblast cells NIH/3T3
   (SCSP-515), mouse fibroblast cells L-929 (SCSP-5039) and mouse
   macrophage cells RAW 264.7 (SCSP-5036) were purchased from Chinese
   Academy of Science Cell Bank for Type Culture Collection (Shanghai,
   China). The cells were maintained in an incubator at 37 °C, with above
   95% humidity and 5% CO[2].

The cellular uptake of D-CuP

   NIH/3T3 cells were plated in a 24-well plate (5 × 10⁴ cells per well)
   and cultured for 24 h. The cells were incubated with free FITC-D-CuP or
   FITC-M-CuP (identical in GHK and FITC molar concentrations) for 6 h.
   The nuclei were stained with Hoechst 33342. Fluorescence signals were
   visualized using a fluorescence microscope (Nikon Ts2R) and quantified
   by flow cytometry (NovoCyte 3000, Agilent, USA).

In vitro macrophage phenotype regulation of G/D-CuP

   RAW264.7 cells were seeded on confocal culture dishes (1 × 10^5 cells
   per well) and cultured for 24 h. For M1 polarization, RAW264.7 cells
   were stimulated with LPS (100 ng mL^−1, Sigma-Aldrich, USA) and IFN-γ
   (20 ng mL^−1, ABclonal, China) for 24 h. The culture medium was then
   replaced with 30 μM M-CuP solution, 15 μM D-CuP solution, or hydrogel
   extract at 37 °C for another 24 h. For M2 polarization, RAW264.7 cells
   were stimulated with IL-4 (20 ng mL^−1, ABclonal, China) for 24 h.
   Immunofluorescent staining was performed to analyze the expression of
   M2 gene markers. RAW264.7 cells were fixed with 4% paraformaldehyde,
   permeabilized with ice-cold methanol, and blocked with MinuteBlock
   Blocking Buffer (Affinibody, China). Cells were incubated with rabbit
   polyclonal antibody against CD206 (1:500, Cat No. WL06177, Lot No.
   T05076177, Wanleibio, China), followed by CoraLite594 goat anti-Rabbit
   IgG(H + L) secondary antibody (1:500, Cat No. SA00013-4, Lot No.
   20000842, Proteintech, China). The nuclei of macrophages were stained
   with DAPI. Fluorescence was detected by confocal laser scanning
   microscopy (CLSM) (Zeiss, LSM800, Germany).

   To quantify the proportion of M2 macrophages, cells were stained with
   APC anti-mouse CD206 antibodies (BioLegend, USA) and PE-Cy7 anti-mouse
   CD86 antibodies (BD, Cat No. 560582, Lot No. 0009083, USA) for 30 min
   at 4 °C. Data were obtained by flow cytometry (NovoCyte3000, Agilent,
   USA) and analyzed using FlowJo V10 software.

Elisa

   RAW264.7 cells (3 × 10^5 cells per well) pre-treated with LPS
   (1 μg mL^−1) for 24 h and untreated HUVECs (3 × 10^5 cells per well)
   were incubated with different materials for 24 h. The supernatants were
   collected and the amount of TNF-α, IL-6, and VEGF secreted were
   measured by ELISA kits (KeyGEN, China).

ROS scavenging ability of G/D-CuP

   Fenton reaction was used to determine the efficiency of G/D-CuP to
   scavenge hydroxyl radical. Specifically, 2 mM FeSO[4] solution,
   360 mg mL^−1 safranine solution, 3% (wt%) H[2]O[2] solution, 0.1 M HCl
   (aq) and G/D-CuP were added to each well in a 96-well plate (n = 3).
   The negative control group (NC) did not contain G/D-CuP. The blank
   control group (BC) did not contain H[2]O[2] solution and G/D-CuP. The
   mixture was incubated at 55 °C for 1 h. After the mixture cooled down,
   the absorbance value at 500 nm was measured using a microplate reader.
   H[2]O[2] scavenging ratio (%) = (A – A[n])/(A[b] – A[n]) × 100%, where
   A represents the absorbance value for the sample group, A[b] represents
   the absorbance value for the BC group and A[n] is the absorbance value
   for the NC group. The ability of G/D-CuP to scavenge ROS was evaluated
   using a ROS fluorescent probe (DCFH-DA).

   NIH/3T3 cells were seeded on 6-well plates (2 × 10^4 cells per well)
   and cultured for 24 h. The cell was treated with a medium that included
   500 μM H[2]O[2] solution that was treated with a variety of materials.
   The negative control group (NC) did not contain materials and the blank
   control group (BC) did not contain H[2]O[2] solution and materials.
   After incubation for 1 h, the DCFH-DA probe (10 μM) was added and
   incubated in the dark for 30 min to detect ROS. Hoechst 33342
   (4 μg mL^−1) was then added and incubated at 37 °C for 5 min to stain
   the cell nuclei. The fluorescence was detected by fluorescence
   microscope (Nikon, Ts2R) and flow cytometry (BD, FACSCelesta).

Cellular SOD activity assay

   NIH/3T3 cells were seeded on 6-well plates (2 × 10^5 cells per well)
   and cultured for 24 h. The cells were treated with a medium that
   included 500 μM H[2]O[2] solution that was treated with a variety of
   materials. The negative control group (NC) did not contain materials
   and the blank control group (BC) did not contain H[2]O[2] solution and
   materials. After incubation for 24 h, the cellular SOD activity was
   measured by a Total Superoxide Dismutase Assay Kit with WST-8
   (Beyotime, China).

Hemolysis assay

   The various concentrations of D-CuP solution or hydrogel extract (G or
   G/D-CuP) (500 μL) and 2% (v/v) red blood cells (RBC) suspension
   (500 μL) were added to a 1.5 mL tube as sample groups. Ultrapure water
   was used as the positive control group (PC), and the physiological
   saline was used as the negative control group (NC). After incubating at
   37 °C for 2 h, the samples were centrifuged at 360 g for 10 min, and
   then each sample (200 μL) was transferred to a 96-well plate. The
   absorbance value at 540 nm was measured using a microplate reader.
   Hemolysis ratio (%) = (A – A[n])/(A[p] – A[n]) × 100%, where A
   represents the absorbance value for the sample group, A[p] represents
   the absorbance value for the positive group and A[n] is the absorbance
   value for the negative group.

MTT assay

   NIH/3T3 cells were seeded on 96-well plates (5 × 10^3 cells per well)
   and cultured for 24 h. The culture medium was replaced with various
   concentrations of G extract, D-CuP solution, G/D-CuP extract and blank
   medium (blank control group, BC) (100 µL) (n = 4). After further
   culturing for 24 h, each well was washed, 1 mg mL^−1 MTT solution
   (100 µL) was added, and the cells were re-incubated for 4 h in the
   dark. The MTT solution was removed and DMSO (150 µL) was added to
   dissolve formazan. Finally, the optical density at 490 nm (OD490) was
   measured using a microplate reader (MD, SpectraMax M2e).

Co-culture of cells with G/D-CuP

   G/D-CuP was prepared in a confocal dish. NIH/3T3 cells (5 × 10⁴ cells
   per well) were seeded onto the hydrogel surface and incubated for 24 h.
   NIH/3T3 cells were stained with 0.5 μM Calcein-AM and the fluorescence
   was detected by CLSM.

Ki-67 immunofluorescence staining

   HUVEC cells were seeded on confocal culture dishes (1 × 10^5 cells per
   well) and incubated with different materials for 24 h. HUVEC cells were
   fixed with 4% paraformaldehyde, permeabilized with ice-cold methanol
   and blocked with MinuteBlock Blocking Buffer (Affinibody, China). Cells
   were incubated with rabbit polyclonal antibody against Ki67 (1:500, Cat
   No. 28074-1-AP, Lot No. 00141310, Proteintech, China), followed by
   CoraLite488 goat anti-Rabbit IgG(H + L) secondary antibody (1:500, Cat
   No. SA00013-2, Lot No. 20000839, Proteintech, China). The nuclei of
   macrophages were stained with DAPI. Fluorescence was detected by CLSM.

Tube formation assay

   Angiogenesis was assessed by tube formation assay. High concentration
   Matrigel 10 µL (40187, Yeasen, China) was added into µ-Slide
   Angiogenesis per well and polymerized at 37 °C for 2 h. HUVEC cells
   were seeded onto the Matrigel (1 × 10^4 cells per well) and incubated
   with different materials for 24 h. HUVEC cells were stained with 0.5 μM
   Calcein-AM and the fluorescence was detected by fluorescence
   microscope, and quantitative analysis of each image was performed using
   Image J 1.53t software.

Transwell migration assay

   L929 cells (1 × 10^5 cells per well) were seeded into the upper chamber
   (8 μm pore size). The lower chamber was added with various materials
   supplemented with 10% FBS, while serum-free DMEM was added to the upper
   chamber. After 24 h, migrated cells on the upper surface of the filter
   were removed with a cotton swab, while the cells on the bottom of the
   filter were fixed with 4% paraformaldehyde at room temperature for
   15 min and stained with 3% crystal violet solution for 15 min. The
   migrated cells were detected by a light microscope, and quantitative
   analysis of each image was performed using Image J software.

Cell scratch experiment

   L929 cells were seeded in 6-well plates (8 × 10^5 cells per well) and
   cultured for 24 h to form a cell monolayer. Subsequently, the cell
   monolayer was scratched using a 200 μL pipette and then washed twice
   with PBS. The cells were incubated with the 30 μM M-CuP or 15 μM D-CuP
   solution (2 mL). Cell images were captured at 0, 8, 18, and 30 h by
   fluorescence microscope. Scratch healing rate (%) = (C[0] – C[t])/C[0]
   × 100%, where C[0] represents the scratch area at 0 h and C[t]
   represents the scratch area at specific times. The scratch areas were
   measured using Image J software.

Proteomics analysis

   Label-free LC-MS/MS was used for proteomics analysis of the blank group
   and D-CuP group (n = 3 in each group). After protein extraction,
   Filter-Aided Sample Preparation (FASP) was used for digestion.
   Orbitrap-Fusion mass spectrometer (Thermo Finnigan, San Jose, CA) was
   used for mass spectrometry analysis. Proteome Discoverer 2.4 software
   (Thermo Scientific) with the built-in Sequest algorithm was used for
   database searching and quantitative analysis. P-value < 0.05 and a fold
   change ratio >1.2 were considered as significantly differentially
   expressed proteins (DEPs).

Molecular docking

   The protein structures of p53 (PDB ID: 8E7A) and HDAC7 (PDB ID: 3C0Y)
   were obtained from the Protein Data Bank ([223]https://www.rcsb.org).
   Ligands and receptor proteins were preprocessed using AutoDock Tools
   1.5.6, and molecular docking was performed using AutoDock Vina.
   Structural visualization, image processing, and analysis of docking
   results were performed using PyMOL 2.5.2.

Surface plasmon resonance (SPR) analysis

   SPR measurements were performed using a WeSPR™ OneAuto instrument
   (XLEMENT) maintained at 25 °C. Recombinant human p53 (Proteintech, Cat.
   No. AG0698) or HDAC7 (Proteintech, Cat. No. AG24396) was immobilized
   onto carboxyl sensor chips. Analytes (M-CuP or D-CuP) were diluted in
   PBST buffer (10 mM phosphate-buffered saline, 0.05% Tween-20, pH 7.4)
   and injected at increasing concentrations at a flow rate of 30 μL/min,
   with 120 s association and 180 s dissociation phases. The running
   buffer (PBST, 25 °C) was degassed prior to use. Sensorgrams were
   analyzed using a global fit to the 1:1 Langmuir binding model.

Animal study

   The C57BL/6 mice (male, 5–6 weeks old) were purchased from SPF
   (Beijing) Biotechnology Co., Ltd. All animals were bred in the
   pathogen-free facility with a 12 h light/dark cycle at temperature
   (20–25 °C), relative humidity (40–60%), and had ad libitum access to
   food and water.

In vivo retention of G/D-CuP and release of D-CuP

   To assess the spatiotemporal retention of G/D-CuP, it was topically
   applied to full-thickness dorsal wounds in diabetic mice. Dynamic
   imaging was performed at 0, 2, 12, 16, 20, 32, and 48 h
   post-application to monitor hydrogel distribution. The G/D-CuP denoted
   the unlabeled hydrogel, while the CV-G/D-CuP represented the hydrogel
   labeled with Crystal Violet dye for enhanced visual observation.

   FITC-conjugated D-CuP hydrogel was uniformly applied to the dorsal
   wounds in darkness. Hydrogel residues were collected at specified time
   points, and their weight was recorded to determine the weight loss. To
   analyze drug release, the collected hydrogels were dissolved in
   deionized water, vortexed, and centrifuged. The fluorescence intensity
   of the supernatant was subsequently measured.

In vivo diabetic wound healing of G/D-CuP

   Male C57BL/6 mice were fed with a high-sugar, high-fat diet for four
   weeks, followed by intraperitoneal injection of premade streptozotocin
   (STZ) solution at a dose of 50 mg kg^−1 daily for one week. Mice
   exhibiting fasting blood glucose levels above 16.7 mmol L^−1 were
   considered as successfully established and were subsequently randomly
   assigned into six groups (n = 6): PBS (Blank control group), G, M-CuP
   solution, D-CuP solution, G/M-CuP and G/D-CuP. Mice without injected
   STZ were set as a Negative control group. A full-thickness skin wound
   with a diameter of 6 mm was created on the dorsal region of each mouse.
   On days 3, 7, and 12 post-treatment, wound images were captured using a
   fixed camera setup that included a scale bar within the field of view.
   Wound areas were quantified using ImageJ software, and all measurements
   were normalized to the scale bar.

Expression of cytokines and growth factors

   Wound skin tissue samples were collected from sacrificed mice on the
   3rd day and immersed in normal saline, mechanically broken into tissue
   homogenate, and the supernatants were collected. Protein levels of
   IL-6, IL-10, and VEGF were quantified using ELISA kits and normalized
   to total protein content measured by the BCA Protein Assay Kit.
   Normalized expression (pg mg^−1) = Cytokine concentration (pg
   mL^−1)/Total protein concentration (mg mL^−1).

Histological analysis

   Regenerated skin samples were excised and collected on the 7th and 12th
   day from sacrificed mice. Skin samples were fixed with 4%
   paraformaldehyde, embedded in paraffin, and sectioned for H&E, Masson,
   and immunohistochemical staining. Immunohistochemical staining was
   performed for TNF-α, IL-6, and IL-10 to evaluate the inflammatory
   response in the tissue, VEGF and CD31 to evaluate angiogenesis, and
   Ki67 to assess cell proliferation. Immunofluorescence staining was
   conducted to assess antioxidant capacity via SOD expression and to
   evaluate macrophage polarization using CD86 (M1 marker) and CD206 (M2
   marker). The slide images were captured by an inverted microscope and
   the average optical density (AOD) was measured using Image J 1.53t
   software.

Statistics and reproducibility

   For assessing differences between the two groups, a t-test was
   performed. Differences among more than two groups were analyzed using
   one-way ANOVA or two-way ANOVA with GraphPad Prism 9.0 software
   (GraphPad Software, Inc.; San Diego, CA). Data are expressed as
   Mean ± SD, n = 3 independent replicates. In each group, any p < 0.05
   was considered statistically significant. NS represents no significant
   difference. Results from representative experiments (such as
   micrographs) were obtained from at least three independent fields of
   view with similar results. The enrich plots were drawn using ChiPlot
   ([224]https://www.chiplot.online/).

Supplementary information

   [225]Supplementary Information^ (2.1MB, pdf)
   [226]Transparent Peer Review file^ (4.1MB, pdf)

Source data

   [227]Source Data^ (165.7KB, xlsx)

Acknowledgements