Preparation and characterization of the Zn-Shik-PEG NPs

Zn-Shik NPs were formed by mixing ZnCl2 aqueous solutions and shikonin (dispersed in ethanol) simply. The phenolic hydroxyl of shikonin allows it to rapidly self-assembled with zinc ions, resulting in a three-dimensional network. Upon adding shikonin dropwise to zinc chloride solution, the solution changed from colorless to blue, which indicating that the metal ions rapidly coordinated with the phenolic hydroxyl groups of shikonin [13, 18]. The as-prepared Zn-Shik nanoparticles were then conjugated with the NH2-mPEG via Schiff-Base reactions, yielding Zn-Shik-PEG nanoparticles. Transmission electron microscopy (TEM) showed that the Zn-Shik-PEG NPs have a uniform size of ≈ 30 nm (Fig. 1A), which was well consistent with the dynamic light scattering (DLS) measurement (Fig. 1B). Additionally, PEGylation did not significantly alter the size or morphology of the Zn-Shik NPs (Figure S1). EDS mapping show that C, O, Zn elements are distributed throughout the nanoparticle (Fig. 1D). Furthermore, there were no obvious size changes of Zn-Shik-PEG nanoparticles in PBS over one week, indicating the good colloidal stability of Zn-Shik-PEG NPs (Figure S2). The UV − vis absorption spectrum shows that characteristic absorption peaks of Zn-Shik-PEG NPs were close to shikonin (Fig. 1C), indicating the coordination between zinc ions and shikonin. Fourier-transform infrared (FTIR) spectra of the Zn-Shik-PEG nanoparticles further indicate that the reaction between phenolic hydroxyl of shikonin and Zn2+, as evidenced by the lowered intensity of HO-C stretching peak around 1275 cm− 1 (Fig. 1E, Table S3) [23]. In addition, the X-ray photoelectron spectroscopy (XPS) survey scan spectra clearly showed the presence of C, O, and Zn elements (Fig. 1F), which consistent with the EDS elemental mapping. The C 1s spectra were deconvoluted into three peaks. the peaks at 284.8, 286.4, and 289.0 eV can be ascribed to C-C, C-O, and C = O bonds, respectively (Fig. 1G) [24]. Figure 1 H show the Zn 2p high resolution XPS spectrum. The two characteristics peaks at 1044.9 and 1021.9 eV can be assigned to Zn 2p1/2 and 2p3/2, respectively [25]. The above results indicated the successful formation of Zn-Shik-PEG nanoparticles.

Fig. 1

Preparation and characterization of Zn-Shik-PEG NPs. (A) TEM image of Zn-Shik-PEG NPs. Scale bar is 100 nm (inset: enlarged image of a single nanoparticle, scale bar: 10 nm). (B) Hydrodynamic size distribution of Zn-Shik-PEG NPs. (C) UV–vis absorption spectra of Shik, NH2-mPEG and Zn-Shik-PEG. (D) Energy dispersion spectroscopy (EDS) mapping images of Zn-Shik-PEG NPs. (E) FTIR spectra of Zn-Shik-PEG NPs. (F − H) Full scan XPS survey spectrum (F) and the high-resolution spectra of Zn-Shik-PEG for C 1s (G) and Zn 2p (H)

ROS scavenging via Nrf2/HO-1 pathway of Zn-Shik-PEG NPs in Vitro

ROS has been implicated as a pivotal factor in the pathophysiology of sepsis. During the progression of sepsis, increased ROS production and/or decreased ROS scavenging capacity significantly contribute to immune cell dysfunction. Accordingly, scavenging excess ROS holds promise for the treatment of sepsis [26, 27]. Since shikonin is a well-known natural polyphenol that has been spotlighted as an ROS scavenger, we tested the ROS scavenging capabilities of Zn-Shik-PEG NPs. We first studied the cytotoxicity of Zn ions, PEG, shikonin and the as-synthesized Zn-Shik-PEG NPs against Raw 264.7 macrophages (Figure S3). We found that shikonin at the concentrations greater than 5 µg/mL effectively suppressed cell growth, while Zn ions, PEG and Zn-Shik-PEG did not exhibit significant cytotoxicity at tested concentrations. The in vitro cytotoxicity results suggested that shikonin cytotoxicity can be reduced by the Zn-Shik-PEG NPs. Furthermore, following treatment with the indicated dose of shikonin or Zn-Shik-PEG NPs, RAW264.7 cells were stimulated with lipopolysaccharide (LPS) to evaluate their anti-inflammatory effects. As shown in Figure S4, LPS significantly decreased the cell viability to 13%, and the cell viability was increased by 20%~48% with different concentrations of shikonin (1, 3, 5 µg/ml) pretreatment, while pretreatment with different concentrations of Zn-Shik-PEG NPs (2, 5, 10 µg/ml) increased the cell viability up to 59%~86%. This result demonstrates that Zn-Shik-PEG NPs can significantly improve the cell viability compared with shikonin. Moreover, when the concentration of Zn-Shik-PEG NPs was 5 µg/mL, the cell viability was significantly increased when compared with control. Therefore, 5 µg/mL of Zn-Shik-PEG NPs was selected as an optimal concentration for subsequent experiments.

Next, the scavenging activities of the Zn-Shik-PEG NPs for intracellular ROS were measured in LPS-induced RAW264.7 cells. As shown in Fig. 2A and Figure S5, an excessive amount of ROS produced after LPS treatment, as evidenced by increased green fluorescence. When the cells were pretreated with shikonin or Zn-Shik-PEG NPs, the mean fluorescence intensity of the cells markedly weakened, suggesting the general intracellular ROS scavenging capability. Meanwhile, we could notice that the Zn-Shik-PEG NPs can more effectively eliminate intracellular ROS compared with the shikonin group, and this is likely due to Zn-Shik-PEG NPs improved water solubility and bioavailability of shikonin. In addition, nuclear factor erythroid 2-related factor (Nrf2), a major regulator of the cellular defense system against oxidative stress, binds with antioxidant response elements (AREs) to initiates transcription of cytoprotective heme oxygenase-1 (HO-1) under oxidative stress conditions [28, 29]. As a result, Nrf2-induced antioxidant kinases can scavenge excess ROS and suppress excessive inflammatory responses [5]. Therefore, here, western blot analysis and immunofluorescence staining were employed to explore the effect of Zn-Shik-PEG NPs on the Nrf2 and HO-1 expression levels in vitro. As shown in Fig. 2B and Figure S6, LPS-mediated reduction of the Nrf2 and HO-1 proteins was significantly suppressed by Zn-Shik-PEG NPs, demonstrating that Zn-Shik-PEG NPs might enhance the endogenous antioxidant defense to counteract LPS-induced oxidative stress. Immunofluorescence staining analysis further supported the results of Western blotting (Fig. 2C, D and Figure S7). After LPS stimulation, the expression of Nrf2 and HO-1 were significantly reduced compared with control condition (P < 0.01). Shikonin treatment could slightly increase the expression of both proteins, whereas the Zn-Shik-PEG NPs group could enhance the expression of NRF2 and HO-1 to a greater degree. Collectively, these results suggest that Zn-Shik-PEG NPs could decrease the LPS-induced ROS via Nrf2/HO-1 pathway.

Fig. 2

ROS scavenging via Nrf2/HO-1 pathway of Zn-Shik-PEG NPs in vitro. (A) Intracellular ROS imaging in RAW264.7 macrophages. (B) Representative western blotting bands of Nrf2, HO-1. (C and D) Representative immunofluorescence images of Nrf2 and HO-1 in RAW264.7 cells after different treatments. (Nucleus: blue; Nrf2 and HO-1: red). Scale bar: 75 μm

Anti-inflammation and anti-apoptosis effect of Zn-Shik-PEG NPs in LPS-Stimulated macrophage cells

Previous studies have found increased proinflammatory cytokines, such as TNF-α and IL-6, in patients with sepsis [30]. A protective strategy for blocking or eliminating these cytokines has been proposed as a valid treatment for sepsis [31]. To evaluate the anti-inflammatory effects of Zn-Shik-PEG NPs, we applied RT-qPCR to detect the mRNA expression levels of pro-inflammatory cytokines in LPS-stimulated RAW264.7 cells. As shown in Fig. 3A, B, LPS stimulation markedly increased the expression levels of TNF-α, IL-6, while pretreatment with Zn-Shik-PEG NPs effectively decreased their expression, which was close to the blank control. Regarding shikonin treatment groups, the increase of related genes also showed somewhat inhibition, whereas Zn ions and PEG does not affect TNF-α or IL-6 expression. These results indicated that the anti-inflammation activity of Zn-Shik-PEG was mainly derived from shikonin.

Fig. 3

Modulation of the AMPK/SIRT1/NLRP3 inflammatory signaling pathway. (A and B) RT-qPCR analysis of the expression of TNF-α and IL-6. (C) Representative western blotting bands of AMPK, p-AMPK, SIRT1, NLRP3. (D and E) Representative immunofluorescence images of SIRT1 and NLRP3 in RAW264.7 cells after different treatments. (Nucleus: blue; SIRT1: red; NLRP3: green). (F) Representative western blotting bands of ASC, cleaved caspase-1, caspase-1, IL-1β, IL-18. (G) Representative TUNEL staining from different treatment of RAW264.7 cells. (Nucleus: blue; TUNEL: green). Scale bar: 20 μm. *p < 0.05, ** p < 0.01, and ***p < 0.001

Adenosine monophosphate-activated protein kinase (AMPK) plays an important regulatory role in the inflammatory response during sepsis [32]. Accumulating evidence indicated that shikonin may exert anti-inflammatory effects by activating AMPK [11, 33]. On the other hand, NLRP3 inflammasome is also a critical signaling mediator of inflammatory damage in sepsis, and SIRT1 plays an important role in the regulation of NLRP3 inflammasome [34]. Furthermore, several studies have suggested a relationship between SIRT1 and the AMPK signaling pathway [35]. Based on the above analysis, we hypothesized that the excellent anti-inflammatory effect of Zn-Shik-PEG NPs might be through the activation of AMPK-SIRT1 signaling pathway to inhibit NLRP3 inflammasome activation in LPS-induced RAW264.7 cells. To verify this hypothesis, we analyzed the total AMPK expression, phosphorylated AMPK (p-AMPK), SIRT1 and NLRP3 by western blotting assay and immunofluorescence assay. As shown in Fig. 3C and Figure S8, LPS stimulation significantly reduced the expression of phosphorylated AMPK (p-AMPK) and SIRT1, while it increased the expression of NLRP3, and which were partly reversed by Zn-Shik-PEG NPs. Similar results were observed in immunofluorescence assay, Zn-Shik-PEG NPs increased the expression level of SIRT1, and reduced NLRP3 expression (Fig. 3D, E and Figure S9). To further confirm whether Zn-Shik-PEG NPs can reduce LPS-stimulated cellular inflammatory damage through AMPK activation, we treated cells with BML-275, an AMPK inhibitor [36]. As shown in Figure S10, Zn-Shik-PEG NPs could indeed reverse the inhibitory effect of BML-275, thus improving p-AMPK and SIRT1 expression, while reducing NLRP3 expression. These results suggest that Zn-Shik-PEG NPs attenuated LPS induced inflammatory damage by activating the AMPK/SIRT1 signaling pathway.

As we know, NLRP3 inflammasomes are multimeric complexes composed of NLRP3, apoptosis-associated spot-like protein (ASC), and cystatin-1 precursor (pro-caspase 1). In response to danger signals, NLRP3 recruits ASC and pro-caspase 1, leading to cleavage of pro-caspase 1 to active caspase 1 (Cas-1), which further promotes maturation and release of pro-inflammatory cytokines (e.g., IL-1β and IL-18), consequently triggering downstream inflammatory responses [37, 38]. Therefore, we next checked the expression levels of ASC, caspase 1 and cleavage caspase 1 (Cle cas-1). As shown in Fig. 3F and Figure S11, ASC, cleaved caspase-1 were upregulated after LPS stimulation, whereas Zn-Shik-PEG NPs attenuated these effects. As a result, Zn-Shik-PEG NPs could reduce the expression of IL-1β, IL-18 in RAW264.7 after LPS stimulation. Moreover, the NLRP3 inflammasome-mediated caspase-1 activation and IL-1β and IL-18 release are closely associated with apoptosis [39, 40]. Therefore, we performed TUNEL staining as further evidence of apoptosis. As shown in Fig. 3G and Figure S12, LPS stimulation significantly increased apoptosis in RAW264.7 cells, but Zn-Shik-PEG NPs can reduce its apoptosis. Taken together, Zn-Shik-PEG NPs could activate AMPK/SIRT1 signaling pathway, decrease NLRP3-related proteins expression, as well as inhibited apoptosis, thus reduced the release of inflammatory factors.

Anti-inflammatory therapeutic activity of Zn-Shik-PEG NPs in Sepsis mice

Next, we used well-established models of LPS injection-induced endotoxemia to explore the in vivo performance of Zn-Shik-PEG NPs. Animals received a single intraperitoneal injection (i.p.) with Zn-Shik-PEG NPs, followed by a single i.p. of LPS (Fig. 4A) (10 mg/kg body weight, 1 h apart). The survival rate, clinical score (clinical score criteria are listed in the Supplementary Table S2), body temperature of mice were recorded. As shown in Fig. 4B, all mice in the LPS group died within 48 h. The shikonin treated group showed a slight improvement at 72 h (17%) while Zn-Shik-PEG NPs delayed mouse death and even rescued more than half of the mice within 72 h after LPS induction (67%), which supporting the protective effect of Zn-Shik-PEG NPs in the acute septic toxicity model. A significantly reduced clinical score was observed in the Zn-Shik-PEG NPs treatment groups (Fig. 4C), demonstrating restoration of physical state after treatment. The changes of body temperature in mice are essential features of sepsis, giving an indication of the gravity of the disease and the impairment of immune response upon infection [24, 41]. Thus, we also monitored body temperature of the mice. As shown in Fig. 4D, E, LPS induced a significant decrease in temperature (from 36.2 to 30.9 °C) within 24 h. while Zn-Shik-PEG NPs showed a strong protective effect on the degree of temperature decrease in mice, demonstrating the beneficial effects of Zn-Shik-PEG NPs in reversing sepsis-associated hypothermia. We next examined the therapeutic efficacy of Zn-Shik-PEG NPs using a cecal ligation and puncture (CLP)-induced sepsis model (Figure S13A). Zn-Shik-PEG NPs recovered body temperature, diminished clinical score, and improved survival in CLP mice (Figure S13B-E). To further evaluate the therapeutic efficacy of the Zn-Shik-PEG NPs, the proinflammatory cytokines TNF-α and IL-6, which are predominantly secreted by activated immune cells, were investigated (Fig. 4F, G and Figure S13F, G). Large quantities of TNF-α and IL-6 were detected in the serum of sepsis model group (either LPS or CLP induced model). When treated with Zn-Shik-PEG NPs, the level of inflammatory cytokines dropped significantly, which is consistent with the results of in vitro experiment.

Fig. 4

Therapeutic efficacy of the Zn-Shik-PEG NPs against LPS-induced sepsis in mice. (A) Experimental procedures for the LPS-induced sepsis mouse model. (B, C) Survival and clinical score analysis of the mice. (D, E) Body temperature change and thermographic images of mice in different treatment groups. (F, G) The level of pro-inflammatory cytokine TNF-α (F) and IL-6 (G) in the serum of septic mice. (H) Distribution of Cy5.5-labelled Zn-Shik-PEG NPs in main organs. Data are presented as mean ± SD (n = 3, 6). *p < 0.05, ** p < 0.01, and ***p < 0.001

Afterwards, we assessed the biodistribution of Zn-Shik-PEG NPs in LPS-treated mice. Mice were injected (i.p.) with Cy5.5-labelled NPs and killed at different time points for ex vivo fluorescence imaging. As shown in Fig. 4H, Cy5.5-labelled Zn-Shik-PEG NPs can reach several organs in mice including liver, kidney, lung, spleen, which are the most vulnerable target organs in sepsis. Zn-Shik-PEG NPs showed long retention in these organs, allowing for long-term protection against multi-organ damage in mice.

To investigate the NPs’ protective effect on multi-organ damage in sepsis mice, we performed biochemical and histopathological analysis. Serum biochemical analysis showed that urea nitrogen (BUN), and creatinine (CRE) (renal function indicators) levels and alanine transferase (ALT) and aspartate aminotransferase (AST) (liver function indicators) levels were significantly increased in LPS-treated group. In contrast, Zn-Shik-PEG NPs could significantly reduce the expression of the above markers (Fig. 5A-D). Furthermore, we also observed severe sepsis-induced lung injury in LPS mice, and the increased severity of lung injury was evidenced by increased lung index as well as myeloperoxidase (MPO) activity levels in the lung of LPS mice, and which can be reduced by Zn-Shik-PEG NPs (Fig. 5E, F). Similar results were obtained in CLP-induced sepsis model (Figure S14A-F). Consistently, hematoxylin-eosin staining (H&E staining) analysis validated the protective effects of the NPs in preventing major organ failure. As shown in Fig. 5G and Figure S14G, the organs of septic mice showed significant pathologic changes, such as inflammatory cell infiltration, alveolar wall thickening, the number of megakaryocytes increased, myocardial fiber breakage, hepatocellular necrosis, tubular epithelial cell edema. In contrast, the degree of organ damage was effectively reversed after Zn-Shik-PEG administration. Semiquantitative assessment using inflammation score also showed similar results, which the score was significantly lowered when septic mice (both LPS and CLP induced model) received Zn-Shik-PEG administration (Figure S15).

Fig. 5

Protective effect of Zn-Shik-PEG NPs on LPS-induced multi-organ damage. (A-D) BUN, CRE, AST, ALT, levels in serum were detected by ELISA kits. (E) Lung index in each treatment group. (F) MPO activity was measured in lungs obtained 24 h after i.p. of LPS. (G) H&E staining of lung, spleen, heart, liver, kidney tissue. Green arrows indicate inflammatory cell infiltration; dark red arrow indicates interstitial thickening; black arrows indicate nuclear pyknosis; blue arrows indicate megakaryocytes; purple arrows indicate myocardial fiber breakage; dark blue arrows indicate hepatocellular necrosis; red arrow indicates renal tubule epithelial cell edema. Scale bar: 50 μm. Data are presented as mean ± SD (n = 3, 6). *p < 0.05, **p < 0.01, and ***p < 0.001

Mechanism of action of Zn-Shik-PEG NPs in vivo

Considering that Zn-Shik-PEG NPs have ROS scavenging and anti-apoptosis effects in vitro, we also investigated the role of Zn-Shik-PEG NPs in ROS and anti-apoptosis in vivo, by performing immunofluorescence staining of ROS-related factors Nrf2, HO-1 and TUNEL staining in lung and spleen tissues. Similar results were clearly found in mice tissues compared with in vitro results. As shown in Fig. 6, LPS decreased the expression of Nrf2 and HO-1, while Zn-Shik-PEG NPs could markedly reverse this trend. In addition, TUNEL staining showed a much lower number of apoptotic cells in the lungs and spleens of LPS/Zn-Shik-PEG-treated mice than in the LPS treated mice. These results suggest that Zn-Shik-PEG NPs possess significant ROS scavenging and anti-apoptosis capacity in both lung and spleen.

Finally, the in vivo biocompatibility of Zn-Shik-PEG NPs was evaluated (Figure S16). There were no abnormalities in blood routine data of healthy mice after Zn-Shik-PEG NPs treatment. In addition, there were no apparent histopathological abnormalities or lesions observed in the heart, liver, spleen, kidney and lung (Figure S16J), indicating that Zn-Shik-PEG NPs are virtually non-toxic to major organs. Collectively, these results demonstrated that the Zn-Shik-PEG NPs are highly biocompatible and their therapeutic effects for sepsis is effective and safe.

Fig. 6

Mechanism of action of Zn-Shik-PEG NPs in vivo (A) Representative fluorescence images of Nrf2 (left), HO-1 (middle) expression and TUNEL staining (right) in lungs and spleen after different treatments. Scale bar: 20 μm. Relative fluorescence intensity of Nrf2, HO-1 and mean fluorescence intensity of Tunel in lung (B) and spleen (C). Data are presented as mean ± SD (n = 3). Statistical analysis: *p < 0.05, ** p < 0.01, and ***p < 0.001