Characterization of hydrogels

As shown in Fig. 1, the hydrogel was composed of two layers. The upper layer was Cu5.4O-loaded thermoresponsive PNIPAM, and the lower layer was a ZnO@SA hydrogel. In our design, upon application of the as-prepared hydrogel, the ZnO@SA layer provides hemostasis and antimicrobial function, and ZnO nanoparticles could produce ROS to improve the antibacterial ability via photocatalytic activity under VL [22]. Subsequently, the upper layer actively contracted wounds by shrinking at approximately 32°C, and the Cu5.4O nanoparticles scavenged ROS at the scheduled time. Figure 2A–B show transmission electron microscopy images of Cu5.4O and ZnO, respectively. Both Cu5.4O and ZnO nanoparticles were uniform, with values of ∼ 5 and 20 nm, respectively. The particle size distribution of Cu5.4O and ZnO nanoparticles was measured using DLS. This analysis showed that the particle sizes were approximately 7 and 25 nm, respectively (Fig. 2C). The optical image of this hydrogel is shown in Fig. 2D, and the inset is the optical microscope image. There was a clear boundary between the upper and lower layers, which demonstrated that the two-layer hydrogel patch was well constructed. Figure 2E shows SEM images of the hydrogel, revealing a good pore structure. Figure 2F shows the hydrogel SEM image after degradation of the lower layer three days postincubation at 37°C; notably, the pore structure of the lower layer no longer existed at that point. This result is consistent with our expectation because we require the upper layer to exert its function after three days. The stress‒strain curve shows that the PTMH possess good mechanical strength (Fig. 2G). The skin of the adult human loss the ability to contract wounds, which has been proved a key role in embryonic wound healing. Therefore, the skin temperature response mechanically active adhesive dressings were designed to actively contract wounds to promoting wound healing. In the temperature-dependent rheological test (Fig. 2H), the elastic modulus (G’) and loss modulus (G’’) changed by approximately 32 °C, which is evidence of the phase change of the PNIPAM-based hydrogel. Figure 2I demonstrates that the hydrogel could shrink at 35 °C, as expected, to actively contract wounds. We conducted an on-off experiment to examine the ROS production ability via photocatalytic activity under VL (Fig. 2J). Following exposure to VL, ZnO exhibited peroxidase-like activity (ultraviolet-visible spectrophotometry spectrophotometer at 652 nm wavelength) [22]. This suggests that the as-prepared ZnO possesses VL activation ability. However, the ability for ROS production gradually decreased with time, and this ability was lost at approximately three days (Fig. S1).

Fig. 2

Preparation and characterization of PTMHs. (A) TEM images of Cu5.4O and (B) ZnO. (Scale bar: 100 nm). (C) The particle size distribution of Cu5.4O and ZnO nanoparticles. (D) Optical image of PTMH. (E) SEM images of PTMH. (Scale bar, left: 100 µm; right: 100 nm) (F) SEM images of ZnO@SA layer view after 3 days of incubation at 35°C. (Scale bar, left: 100 nm; right: 50 nm). (G) The stress‒strain curve of PNIPAM. (H) The elastic modulus (G’) and loss modulus (G’’) of PNIPAM. (I) Images of PNIPAM following incubation at 35 °C. (J) The test of ROS production of ZnO. (K) H2O2, (L) ·OH, and (M) O2·− scavenging capacity of different concentrations of Cu5.4O@PNIPAM. Data in K, L, and M represent the mean ± standard deviation (n = 5). n.s., no significance; one-way ANOVA. TEM, transmission electron microscopy; SA, sodium alginate; PNIPAM, poly(N-isopropylacrylamide); PTMH, programmed time-released multifunctional hydrogel; ANOVA, analysis of variance

H2O2, being a prevalent endogenous ROS variant, was initially employed to examine the ROS elimination capability of Cu5.4O@PNIPAM. The results indicated an obvious reduction of H2O2 following incubation with escalating concentrations of Cu5.4O@PNIPAM, ultimately achieving a rate of 98.3% at 16 ng µL− 1 (Fig. 2K). ·OH radicals are among the most potent and biologically relevant free radicals, exerting a significant influence on inflammation disorders. The scavenging ability of Cu5.4O@PNIPAM toward ·OH radicals was gradually enhanced with the escalating concentration of Cu5.4O@PNIPAM, ultimately achieving a remarkable scavenging rate of 96% at 16 ng µL− 1 (Fig. 2L). Similarly, the O2·− scavenging test showed that Cu5.4O@PNIPAM could scavenge 47.3% of O2·− at a concentration of 16 ng µL− 1 (Fig. 2M). Moreover, to investigate the free radical scavenging ability, 2,2′-azino-bis(3-ethylbenzthiazoline-6-sulfonate) (ABTS) radicals were chosen. The results revealed that the scavenging rate could reach 96% following treatment with Cu5.4O@PNIPAM (Fig. S2). Additionally, Cu5.4O@PNIPAM also showed an efficient scavenging ability towards 2,2-diphenyl-1-(2,4,6-trinitrophenyl) (DPPH) radicals (Fig. S3). To further determine the cytoprotective ability of PTMH, 3T3 cells and HUVECs were pretreated with ZnO@SA, Cu5.4O@PNIPAM, and PTMH. Next, the cells were incubated with 250 µM H2O2. Following Cu5.4O@PNIPAM and PTMH treatment, cell viability remained the higher level compared with that of the ZnO@SA and H2O2 groups, which was attributed to the ROS scavenging effect of Cu5.4O (Fig. S4). The results demonstrated that Cu5.4O@PNIPAM could efficiently scavenge ROS and free radicals and protect cells from excessive oxidative stress damage, which is consistent with those of our previous research.

Furthermore, according to the in vitro cytotoxicity results (Fig. S5), we determined relatively safe concentrations of ZnO@SA, Cu5.4O@PNIPAM, and PTMH for subsequent biological experiments. There was no significant difference in cell viability between the hydrogel treatment group and control group (p > 0.05). These data indicated that the hydrogel is safe for cells.

Antibacterial activity of PTMH

In this study, the antibacterial effect of PTMH was assessed using two representative bacterial species, namely Pseudomonas aeruginosa (P. aeruginosa, ATCC 27,853) and MRSA, ATCC 43,300). The bacterial suspension was coincubated with ZnO@SA, Cu5.4O@PNIPAM, and PTMH with or without VL irradiation. As shown in Fig. 3A–C, ZnO@SA and PTMH groups with VL treatment exhibited a significantly reduced number of bacterial colonies formed by the coincubated bacterial suspension compared to that in the other groups, indicating that ZnO@SA and PTMH groups with VL treatment could efficiently inhibit the growth of P. aeruginosa and MRSA (p < 0.01). Additionally, the absorbance of the bacterial suspension at a wavelength of 600 nm, which serves as a standard metric for monitoring microbial density in liquid cultures, was significantly decreased (p < 0.01) following ZnO@SA and PTMHs treatment with VL (Fig. 3D–E). This trend exhibited a resemblance to the outcomes derived from the bacterial colony formation test. However, as previously shown by our research group [8], Cu5.4O@PNIPAM with or without VL treatment could not inhibit bacterial growth. Notably, compared with the groups without VL treatment, the number of bacterial colonies was significantly decreased in the groups (ZnO@SA and PTMH) with VL treatment (p < 0.01). According to previous research, this observation may be attributed to the photocatalytic effect of ZnO.

To further determine the underlying mechanism of the antibacterial effect of PTMH, we utilized a 2′,7′-dichlorofluorescin (DCFH) diacetate probe to measure the level of oxidative stress in live cells under different treatments. Following VL treatment, the green fluorescence intensity of cells from the ZnO@SA and PTMH groups was significantly increased compared with that of the other groups (Fig. 3F–G). Moreover, the percentage of DCF-positive cells under ZnO@SA and PTMH with VL treatment increased from 8.3 to 57.5% and 54.8% (p < 0.01), respectively (Fig. 3H–I). However, the percentage of DCF-positive cells in the groups without VL treatment remained low (9.14% and 4.93%, respectively). This finding demonstrated that PTMH could produce ROS through VL treatment and inhibit the growth of bacteria.

Fig. 3

Evaluation of the antibacterial activity of PTMH in vitro. (A) Images of bacterial colonies of P. aeruginosa and (B) MRSA following treatment with Cu5.4O@PNIPAM, ZnO@SA, and PTMH with or without VL irradiation. (C) The number of bacterial colonies formed by P. aeruginosa and MRSA under different treatments. (D) OD600 values of P. aeruginosa and (E) MRSA under different treatments. (F) Representative confocal digital images and (G) fluorescence intensity for DCFH-DA (green) staining of 3T3 cells under different treatments. (Scale bar: 50 μm) (H) ROS levels and (I) statistical analysis in 3T3 cells under different treatments. Data in C–E, G, and I represent the mean ± standard deviation (n = 5). *p < 0.05; **p < 0.01; one-way ANOVA. P. aeruginosa, Pseudomonas aeruginosa; MRSA, methicillin-resistant Staphylococcus aureus; VL, visible light; SA, sodium alginate; PNIPAM, poly(N-isopropylacrylamide); PTMH, programmed time-released multifunctional hydrogel; OD, optical density; DCFH-DA, 2′,7′-dichlorofluorescin diacetate; FSC, forward scatter; SSC, side scatter; ROS, reactive oxygen species; ANOVA, analysis of variance

Hemostatic ability assay

Hemorrhage resulting from physical injury, traumatic events, and surgical interventions constitutes a significant concern leading to morbidity and mortality [23]. Therefore, it is necessary to achieve hemostasis to promote tissue regeneration [24]. To examine the hemostatic ability, we conducted a coagulation test, as shown in Fig. S6A. In the absence of intervention, normal blood clots in 5–6 min. However, the as-prepared hydrogel could accelerate hemostasis (∼ 1 min, Fig. S6B). We further checked the clotting using SEM; there was plenty of fibrin in the clotting caused by our hydrogel compared with the control group (Fig. S7A). The size of the fibrin was quantified and is shown in Fig. S7B. The fibrin size in the hydrogel group was approximately 200 nm. However, the fibrin size in the control group was approximately 125 nm. In vivo hemostatic sealing of a rat liver was conducted to examine the hemostatic capability because less bleeding occurred in a cutaneous wound mouse model. The rapid hemostatic sealing of a bleeding rat liver in vivo by the hydrogel is shown in Fig. S8A. Following different treatments, blood loss was reduced in the as-prepared group (Fig. S8B). The time to hemostasis and blood loss until hemostasis is shown in Fig. S8C–D. Based on the above data, the as-prepared hydrogel could significantly reduce the blood clotting time. Bleeding is the first aspect that requires management in the process of wound healing [25]. Therefore, we added Ca2+ to enhance the hemostatic ability. The above results revealed that the prepared PTMHs possess good hemostatic ability.

Promotion of impaired wound healing via PTMH

We further assess the effect of improving delayed wound repairing of PTMH. Diabetic foot ulcers pose a significant challenge in the management of individuals with diabetes due to their inherent vulnerability to bacterial infection, uncontrolled inflammation, and impaired angiogenesis, which impeded spontaneous healing [26, 27]. The high blood glucose and inhospitable environment surrounding the affected wound site frequently facilitates bacterial colonization [28].

The infection models were established one day following the introduction of P. aeruginosa into the wounds that were induced on the dorsal region of streptozotocin-induced diabetic mice (Fig. 4A) [29]. The model mice were randomly divided into four groups: (1) Control group (treated with PBS); (2) ZnO@SA group (with VL irradiation); (3) Cu5.4O@PNIPAM group (with VL irradiation); and (4) PTMH group (ZnO@SA-Cu5.4O@PNIPAM PTMH with VL treatment). The adhesion achieved by amine groups on chitosan are responsive to changes in pH. When pH < 6.5, [NH2] < [NH3+], and the chitosan chains dissolve in water as a polyelectrolyte. When pH > 6.5, [NH2] > [NH3+], and the NH2-OH hydrogen bond promotes the chitosan chains to form a network [30]. The wounds from each group were photographed on days 0, 3, 7, and 14 post-treatment. The wounds treated with PTMH showed obviously accelerated wound closure (p < 0.01) (Fig. 4B–C and Fig. S9). Quantitatively, the PTMH group led to 62.4% and 97.48% wound closure on days 7 and 14 post-treatment, respectively. These rates were markedly higher (p < 0.01) than those of the PBS-treated wounds (11.6% and 22.4%, respectively) (Fig. 4D–E). On day 3 post-treatment, PTMHs led to less inflammatory exudation with less infection status compared with control (p < 0.01), which may be attributed to the antibacterial effect of ZnO. Additionally, five mice from each group were fed continuously until the wounds completely healed. As shown in Fig. 4F, the wounds of mice treated with PTMHs had healed 15.2 days post-treatment and were significantly shorter (p < 0.01) than those of mice treated with PBS, ZnO@SA, and Cu5.4O@PNIPAM (31.6, 25, and 23.6 days, respectively). Notably, Cu5.4O@PNIPAM showed a greater wound-healing promotion effect than ZnO@SA, which was attributed to the antioxidative effect of Cu5.4O nanoparticles and the cutaneous contraction effect of PNIPAM hydrogels.

Fig. 4

Efficiency of PTMHs for the healing of mouse diabetic wounds with P. aeruginosa infection. (A) Schematic of the experimental design for the therapeutic effect evaluation of PTMH in mouse STZ-induced diabetic wounds with P. aeruginosa infection. (B) Digital images of diabetic wounds at various time intervals, with a blue disc measuring 6 millimeters in diameter for scale reference. (C) Schematic images of diabetic wounds under different treatments. (D) The proportions of wound healing achieved through various treatment on days 0, 3, 7, and 14 (n = 5). (E) Area percentages of closed wounds. (F) Ultimate healing time of each group. (G) Representative histological images of H&E staining in the wound site on day 14. Black arrows represent the thickness of the granulation tissue. (Scale bar: 100 μm) (H) Statistical analysis of the thickness of the granulation tissue. (I) The images of Masson’s trichrome staining in diabetic wounds through various treatment. (Scale bar: 100 μm) (J) Statistical analysis of the percentage of collagen volume fraction. (K) Images and (L) quantitative counts of bacterial colonies formed by P. aeruginosa harvested from wound tissues at various time intervals. Data in E, F, H, J, and L represent the mean ± standard deviation (n = 5). *p < 0.05; **p < 0.01; n.s., no significance; one-way ANOVA. SA, sodium alginate; PNIPAM, poly(N-isopropylacrylamide); PTMH, programmed time-released multifunctional hydrogel; P. aeruginosa, Pseudomonas aeruginosa; STZ, streptozotocin; ANOVA, analysis of variance; CFU, colony-forming unit; H&E, hematoxylin and eosin

Histological analysis was employed to investigate the wound healing processes in further detail, utilizing techniques, such as H&E and Masson’s trichrome staining methods [31]. Among all groups, the wounds in the PTMH treatment group exhibited the thickest granulation tissue (p < 0.01) according to H&E section images (Fig. 4G–H). Moreover, we evaluated the collagen content at the diabetic wound sites using Masson’s trichrome staining. As shown in Fig. 4I–J, wounds treated with ZnO@SA, Cu5.4O@PNIPAM, and PTMH showed higher collagen content than the control group (p < 0.01), whereas the PTMH group showed more than the ZnO@SA and Cu5.4O@PNIPAM groups (p < 0.01). It demonstrated that the development of granulation tissue and collagen was significant improved after 14 days of PTMH treatment, which showed an accelerated wound healing.

The antibacterial performance of PTMH was evaluated by collecting and homogenizing the mouse cutaneous tissue in wound areas on days 3, 7, and 14 (Fig. 4K–L). On day 3, a lower number of bacterial colonies was observed in the ZnO@SA and PTMH groups (p < 0.01). However, the number of bacterial colonies from the PTMH groups on days 7 (p < 0.05) and 14 (p < 0.01) was significantly lower than that of the ZnO@SA groups, indicating that ZnO@SA could inhibit bacterial growth in the short term, and that PTMH could release Cu5.4O nanoparticles from day 3 onward, thereby improving the microenvironment of wound tissue in the middle and late stages of wound healing. Notably, PTMHs were associated with the lowest number of bacterial colonies (p < 0.01) during the entire period of the experiment due to their antibacterial effect of ZnO and anti-inflammatory effect of Cu5.4O, indicating long-term effective inhibition of bacteria. Patients with diabetic wound are susceptible to bacterial infection due to the microenvironment of high blood glucose and extensive inflammation, which hinders the normal process of wound healing [32, 33]. Thus, the efficient antibacterial activities of PTMHs provided a sterile microenvironment for subsequent repair stages.

To further determine the effect of PTMH on wound healing promotion by fluorescent staining, wound tissue from day 14 post-treatment was harvested, sectioned, and stained with the corresponding antibodies. PCNA, as a marker of proliferation, was selected for the detection of tissue proliferation under different treatments [34]. The highest number of red fluorescence-positive cells was observed in the PTMH group (p < 0.01), indicating the ability of PTMH to promote tissue repair (Fig. 5A–B). In previous studies [8], we demonstrated that Cu5.4O nanoparticles effectively stimulated the proliferation and migration of vascular endothelial cells, thereby contributing to the promotion of wound healing. Consequently, it is of great interest to assess the in vivo angiogenic impacts of PTMHs. Blood vessels transport essential nutrients to the site of injury, thereby facilitating the process of wound healing. As shown in Fig. 5C–D, the number of blood vessels was significantly increased in the Cu5.4O@PNIPAM-treated wounds (p < 0.01) and PTMH-treated wounds (p < 0.01) compared with that in the PBS-treated wounds, indicating that PTMH exert angiogenic effects in vivo.

Fig. 5

Regulatory effects of PTMH on the microenvironment in vivo. (A) Fluorescent images and (B) statistical analysis of PCNA (red) and DAPI (blue) in mouse diabetic wounds tissue. (Scale bar: 100 μm) (C) Fluorescent images and (D) statistical analysis of CD31 (red) and DAPI (blue) in mouse diabetic wounds tissue. (Scale bar: 100 μm) (E) Fluorescent images and (F) statistical analysis of DHE (red) and DAPI (blue) on days 3 and 14 post-treatment wound tissue. (G) The distribution of cells in wound tissue obtained from different groups on days 3 and 14 post-treatment, stained with rat anti-mouse CD11b-Brilliant Violet 650™, LY6G-FITC, and CD86-Brilliant Violet 421™ (1:200). (H) Statistical analysis of CD86-Brilliant Violet 421TM-positive cells. (I) The concentrations of TNF-α, (J) IL-6, (K) IL-1β, (L) IL-10, (M) VEGF, (N) SOD and (O) MPO in wound tissue were determined using ELISA. Data in B, D, F, H and I–O represent the mean ± standard deviation (n = 5). *p < 0.05; **p < 0.01; one-way ANOVA. SA, sodium alginate; PNIPAM, poly(N-isopropylacrylamide); PTMH, programmed time-released multifunctional hydrogel; PCNA, proliferation cell nuclear antigen; DHE, dihydroethidium; CD, cluster of differentiation; ELISA, enzyme-linked immunosorbent assay; TNF-α, tumor necrosis factor-α; IL, interleukin; VEGF, vascular endothelial growth factor; SOD, superoxide dismutase; MPO, myeloperoxidase; ANOVA, analysis of variance; DAPI, 4’,6-diamidino-2-phenylindole; SSC-A, side scatter area

Furthermore, in order to detect the level of reactive oxygen species (ROS) at the site of the wound, the tissue surrounding the wound was collected on days 3 and 14 after the surgical procedure. Subsequently, the collected tissue was promptly frozen using liquid nitrogen to generate frozen sections, which were then subjected to staining with the fluorescent dye dihydroethidium (DHE) indicating the ROS level of tissues [35]. As shown in Fig. 5E–F, on day 3, the fluorescence intensities were significantly higher for ZnO@SA (p < 0.05) and PTMHs (p < 0.01) compared with those for Cu5.4O@PNIPAM due to the production of oxygen free radicals by ZnO with VL irradiation. However, on day 14, the fluorescence intensities for the PTMHs were apparently decreased compared with those of the other three groups. This may be attributed to the antioxidative effect of Cu5.4O. To further confirm this hypothesis, M1 macrophages in wound tissue were quantified using FACS [13]. M1 macrophages can be activated by a series of inflammatory factors strongly associated with ROS and further produce ROS to enhance inflammation [36]. In contrast, M1 macrophages are involved in the immune system response of mice towards invasive microorganisms, such as bacteria [37]. On day 3, the number of CD86-positive cells (M1 macrophages) was significantly increased in the ZnO@SA (p < 0.05) and PTMH (p < 0.01) groups compared with the control and Cu5.4O@PNIPAM groups (Fig. 5G–H and S10). Conversely, on day 14, the number of CD86-positive cells (M1 macrophages) was significantly decreased in the PTMH group, whereas it was the highest in the control group. This trend was consistent with the results obtained from DHE staining images, confirming that on day 3, M1 macrophages could be activated by ROS produced through ZnO with VL irradiation, which contributed to alleviating the bacterial infection and activating the subsequent signaling pathway for repair promotion. In contrast, on day 14, the release of Cu5.4O inhibited the excessive ROS production and inflammation, promoted the microenvironment, and ultimately enhanced wound healing.

Moreover, proinflammatory factors, as a significant medium, maintained chronic and high-level inflammation in diabetic wounds [38]. Corresponding ELISA kits were used to detect the concentration of proinflammatory factors in wound tissue collected on day 14 post-treatment. As shown in Fig. 5I–K, the concentrations of TNF-α, IL-6, and IL-1β in wound tissue were dramatically decreased following treatment with PTMH (p < 0.01), whereas those of IL-10 (a classic anti-inflammatory factor) were significantly increased compared with those of the other groups (p < 0.01) (Fig. 5I). This demonstrated that PTMH could obviously inhibit inflammation in the wound repairing stage. Vascular endothelial growth factor (VEGF), a cytokine that promotes endothelial cell proliferation, is easily inactivated under excessive oxidative stress and inflammation, thereby inhibiting the process of angiogenesis [39]. Notably, the concentration of VEGF was significantly increased in the PTMH treatment group compared with the other groups (p < 0.01) (Fig. 5M). This trend was consistent with the results obtained from CD31 staining (Fig. 5C).

In addition, superoxide dismutase (SOD) and myeloperoxidase (MPO) were selected as determinants of oxidative stress in wound tissue. SOD, an enzyme with antioxidant properties, is capable of scavenging ROS to uphold a harmonized redox equilibrium within biological systems. SOD demonstrates its efficacy in assessing the degree of oxidative stress and hypoxia by impeding lipid peroxidation within wound tissue and prompting tailored reactions to inflammatory stimuli [40]. It is important to acknowledge that diminished SOD activity can intensify the oxidative harm linked to diabetes [8]. In comparison to the control group, the PTMH treatment group exhibited a sustained elevation in SOD activity (p < 0.01) (Fig. 5N). The glycosylated enzyme MPO is found in the granules of neutrophils and macrophages, playing a vital role in promoting inflammatory injuries caused by oxidative stress and serving as a biomarker for monitoring the advancement of diabetic wounds. Consequently, assessing MPO activity enables the evaluation of immune cell infiltration, including neutrophils and M1 macrophages, in inflamed tissue [14]. Remarkably, the control group demonstrated a noteworthy elevation in MPO activity, which was significantly reduced with PTMH treatment (Fig. 5O). It suggested that the administration of PTMH treatment effectively suppresses oxidative stress within inflamed wound tissue.

Our results of diabetic wound healing showed that there is little difference between ZnO and Cu5.4O groups. On the one hand, ZnO could eliminate bacteria by producing ROS, which could efficiently control the status of bacterial infection in the early stage of wound healing. However, the process of diabetic wound healing was impeded by not only bacterial infection but uncontrolled inflammation. In the stage of inflammation, excessive ROS and proinflammatory factors impaired the function of fibroblasts and vascular endothelial cells. Consequently, the healing process was stalled in the inflammatory phase. On the other hand, Cu5.4O nanoenzyme could scavenge ROS and inhibit the level of inflammation, which could regulate inflammation in wound tissue and improve the effect of angiogenesis. However, moderate ROS could inhibit the growth of bacteria and simply using Cu5.4O nanoenzyme in the early stage of wound healing may aggravate infection. ZnO and Cu5.4O can complement each other to address their respective limitations and capitalize on their advantages. ZnO efficiently eliminated bacteria in the early phase of healing and Cu5.4O scavenged excessive ROS in the stage of inflammation, thereby achieving the improvement of microenvironment of diabetic infected wound.

Therapeutic mechanisms of PTMH in impaired wound healing

In order to investigate the underlying mechanism through which PTMH promote wound healing, transcriptome sequencing was utilized to perform quantitative RNA analysis. According to unguided principal component analysis (PCA) of the data, the transcriptomics profiles of cutaneous wounds in diabetic mice treated with PTMH and wound dressing exhibited significant differences (Fig. 6A). Based on the Venn diagram, 14,259 genes were shared by the PTMH treatment and control groups, 961 and 591 genes were unique to the PTMH and control groups, respectively (Fig. 6B). Furthermore, volcano plots revealed significant differential expression of RNAs (p < 0.05). Among these, 1,370 RNAs were found to be downregulated as determined by a cutoff value of 0.83-fold change, while 1,432 RNAs were upregulated as determined by a cutoff value of 1.2-fold change, following PTMH treatment (Fig. 6C).

Fig. 6

Mechanisms underlying the therapeutic effects of PTMH in impaired wound healing. (A) Principal Component Analysis (PCA) was performed on the differentially expressed genes in the wound tissues of the mice subjected to treatment with PTMH (PTMH group) or PBS (Control group). Each data point within the two groups represents an independent replicate. (B) Venn diagram was constructed based on the transcriptomic data to illustrate the overlap between the two groups. (C) Volcano plots were generated to visualize the genes that were upregulated and downregulated after PTMH treatment. (D) KEGG pathway enrichment analysis of the differentially expressed genes (downregulation). (E) Heatmaps illustrating the noteworthy downregulation of genes implicated in inflammation and oxidative stress subsequent to PTMH treatment. (F) KEGG pathway enrichment analysis of the differentially expressed genes (upregulated). (G) Chord diagram showing the genes involved in inflammation and oxidative stress that were significantly upregulated following PTMH treatment. (H) Relative expression of genes involved in inflammation and wound regeneration. Data in h represent the mean ± standard deviation from three independent replicates (n = 3). *p < 0.05; **p < 0.01; n.s., no significance; t-test. PTMH, programmed time-released multifunctional hydrogel; PCA, principal component analysis; IL, interleukin; KEGG, Kyoto Encyclopedia of Genes and Genomes; ECM, extracellular matrix; Nod1, nucleotide-binding oligomerization domain containing 1

Gene ontology (GO) enrichment analysis was performed to determine the enrichment pathways of differentially expressed genes (DEGs) that were downregulated. Following PTMH treatment, downregulated DEGs were significantly enriched in inflammation-related pathways, such as “positive regulation of IL-1β production” (Fig. S11). As shown in Fig. 6D and S12, the KEGG pathway enrichment analysis showed that the downregulated DEGs in the PTMH treatment group were involved in the nucleotide-binding oligomerization domain (NOD)-like receptor (NLR) signaling pathways associated with inflammation and oxidative stress in cutaneous wounds [41]. Among these downregulated genes, inflammation-associated genes, such as NOD-containing protein 1 (Nod1), signal transducer and activator of transcription 1 (Stat1), Stat2, C-C motif chemokine ligand 2 (Ccl2), Ccl5, and Ccl12, were significantly inhibited following PTMH treatment (Fig. 6E and H). We constructed a protein‒protein interaction (PPI) network comprising key downregulated genes, suggesting that the interaction of these key genes is related to inflammation and oxidative stress (Fig. S13). A subfamily of NLR signaling pathways is primarily involved in the activation of inflammation, which has been implicated in a multitude of disease models and human diseases [41]. Bauernfeind et al. demonstrated that increased ROS levels promote NLR family pyrin domain containing 3 (NLRP3) expression at the transcriptional level [42]. NLRP3 could lead to the maturation of proinflammatory factors [42], such as IL-1β and IL-18, and the decreased concentration of IL-1β following PTMH treatment (Fig. 5K) also verified this trend. The dysregulation and hyperactivation of the NLR signaling pathway are closely related to excessive inflammation and oxidative stress, which may impede the process of diabetic wound healing. Following PTMH treatment, the NLR signaling pathway was significantly downregulated, and this effect may be attributed to two reasons. First, PTMH exerted a strong antibacterial effect, particularly in the first three days postinjury, which controlled the infection and improved the microenvironment of wound healing at the early stage. Second, the NLR signaling pathway may be downregulated via the antioxidative and anti-inflammatory effect of Cu5.4O during the repair stage. By the synergistic effect in time-sequence, the inflammation-associated pathway would be downregulated, and the microenvironment of wound tissue would be improved.

GO and KEGG analyses were performed to further determine the enrichment of upregulated DEGs. The results of GO analysis showed that DEGs were enriched in the “noncanonical Wnt signaling pathway” and “extracellular matrix assembly” (Fig. S14). The results of KEGG analysis demonstrated that the DEGs upregulated following PTMH treatment were involved in “extracellular matrix–receptor interaction” (Fig. 6F–G and S15). Furthermore, the KEGG analysis indicated that the key enrichment pathway is mainly centered on the Wnt signaling pathway (Fig. 6F–G and S16). The heatmaps and PPI networks also showed that regeneration-related genes, such as Wnt 3a and Wnt 5a, were significantly upregulated (Fig. 6H and S17–18). The upregulated Wnt signaling pathway could facilitate the proliferation of vascular endothelial cells and promote angiogenesis [43]. The promotion of accelerated angiogenesis plays a critical role in the process of healing diabetic wounds, as it facilitates the delivery of oxygen and nutrients to the compromised tissue, thereby mitigating extensive inflammation and combating bacterial infection [44]. Nevertheless, the presence of an extensive local inflammatory reaction hinders the progression of angiogenesis in diabetic wounds, including bacterial infection, excessive oxidative stress, and uncontrolled infiltration of inflammatory cells. Excess ROS production may inactivate cytokines, such as VEGF, and cause vascular endothelial cell apoptosis, which may dramatically impede the process of angiogenesis and wound healing. As we previously described [8], Cu5.4O ultrasmall nanoparticles exhibit remarkable antioxidant efficiency both in vitro and in vivo and promote the proliferation of vascular endothelial cells. In the present study (Fig. 5C, D, and M), the number of blood vessels and the concentration of VEGF in the PTMH treatment group were significantly increased, demonstrating the improvement of angiogenesis, which was consistent with the results of the upregulation of Wnt signaling.

In summary, we have developed a PTMH system that can combat bacterial infection, inhibit inflammation, and promote impaired wound healing in a time-dependent manner according to the characteristics of the different stages of the wound healing process. Hence, forthcoming versions of this platform, in conjunction with our PTMHs, have the potential to be programmed for the generation of diverse immunotherapeutic and other substances, facilitating the evaluation of various rational therapeutic combinations. Additionally, we are exploring a real-time monitoring system for the detection of inflammatory cell infiltration and the dynamic assessment of oxidative stress levels, which would enhance the precision and efficiency of this therapeutic system. Finally, identification of better drug-loading options in additional translational animal models is imperative to improve the clinical relevance of this system.

Biocompatibility of PTMH in vivo

The biocompatibility of PTMH was assessed by examining the organ histopathology and blood chemistry of healthy mice. As shown in Fig. S19, there were no signs of necrosis, congestion, or hemorrhage observed in heart, liver, spleen, lung, and kid