Abstract

Objective

   To enhance the therapeutic effect of atezolizumab on osteosarcoma (OS)
   by constructing a pH-responsive hydrogel nanocomplex (Gel@PLGA@FA) as a
   delivery platform for Shikonin.

Methods

   First, Shikonin was initially employed to analyze the [28]GSE14359
   dataset, leading to the identification of 28 differentially expressed
   genes (DEGs). Based on this, a risk score model was constructed and
   molecular dynamics simulations were performed to assess the binding
   ability between Shikonin and cyclin-dependent kinase 1 (CDK1). In
   addition, the in vitro antiproliferative effect of Shikonin on MG63 and
   Saos-2 OS cell lines and its selective toxicity on normal cells were
   assessed. In order to overcome the disadvantages of poor water
   solubility and normal cytotoxicity towards Shikonin, a complex loaded
   with Shikonin by pH-responsive intelligent hydrogel nanomaterials was
   synthesized and its anti-programmed death ligand-1 (PD-L1) therapeutic
   effect on OS cells was evaluated.

Results

   Molecular dynamics simulation showed that Shikonin showed strong
   binding ability to CDK1, showing stable conformation, enhanced
   structural stability and other characteristics. In vitro experiments
   showed that Shikonin had a significant anti-proliferative effect on OS
   cells, while it had selective toxicity on normal liver, kidney and
   osteoblasts. The pH-responsive hydrogel nanomaterial (Gel@PLGA@FA)
   loaded with Shikonin showed good drug release characteristics at
   different pH conditions, especially in the tumor microenvironment to
   achieve controllable drug release. Combined use of Gel@PLGA@Shikonin@FA
   and atezolizumab effectively down-regulated CDK1 and PD-L1 expression,
   inhibited cell proliferation and promoted apoptosis, significantly
   enhancing the anti-PD-L1 therapeutic effect on OS cells. JC-1 staining
   experiments further confirmed that this combination therapy could
   perturb mitochondrial membrane potential and lead to stronger
   apoptosis.

Conclusion

   This study reveals the unique mechanism of action of Shikonin as a
   potential anticancer drug and demonstrates the potential of
   pH-responsive hydrogel nanomaterials as efficient and safe delivery
   systems for targeted cancer therapeutics. The strategy of
   Gel@PLGA@Shikonin@FA combined with atezolizumab provides a new idea and
   experimental basis for OS treatment.

Supplementary Information

   The online version contains supplementary material available at
   10.1007/s12672-025-03742-2.

   Keywords: Osteosarcoma, Shikonin, Nanomaterial, Hydrogels, PD-L1, CDK1

Introduction

   Osteosarcoma (OS), the most common primary malignant bone tumor in
   adolescents, is characterized by high invasiveness and early metastasis
   [[29]1]. Despite advancements in surgical and chemotherapeutic
   techniques, the 5-year survival rate for patients with recurrent or
   metastatic disease remains at approximately 20% [[30]2], highlighting
   an urgent need for more effective treatment strategies.

   The advent of immune checkpoint inhibitors has brought breakthroughs in
   the treatment of solid tumors [[31]3]. However, the single-agent
   response rate of programmed death ligand 1 (PD-L1) inhibitors such as
   atezolizumab in OS remains extremely low. The core reason lies in the
   fact that OS is classified as a “cold tumor”-characterized by sparse
   T-cell infiltration in the tumor microenvironment and a strong
   immunosuppressive state, which makes it difficult to elicit an
   effective immune response [[32]4]. Shikonin, a natural naphthoquinone
   compound extracted from the traditional Chinese medicinal herb
   Lithospermum erythrorhizon [[33]5], possesses multiple biological
   activities including anti-inflammatory [[34]6], antioxidant [[35]7],
   and anti-tumor properties [[36]8]. In particular, it can improve the
   tumor microenvironment by downregulating immunosuppressive factors such
   as TGF-β [[37]9] and regulating the ratio of CD8⁺ T cells to Treg
   cells, thereby showing potential for combination immunotherapy [[38]3,
   [39]10]. However, its clinical application is limited by drawbacks such
   as poor water solubility and certain toxicity to normal cells.

   In recent years, the development of nanotechnology has provided new
   ideas for tumor combination therapy [[40]11]. PH-responsive hydrogel
   nanocomplexes are widely used in targeted drug delivery systems because
   of their unique tumor microenvironment response ability, good
   biocompatibility, and controllable drug release characteristics
   [[41]12]. Such materials are able to trigger structural changes in the
   acidic microenvironment of tumor tissue and achieve site-directed
   release of the drug, thereby improving therapeutic efficiency and
   reducing systemic toxicity [[42]13]. Li et al. [[43]14] used
   pH-sensitive polymeric micelles to load chemotherapeutic drugs to
   achieve effective accumulation and controlled release at the tumor
   site, significantly improving the therapeutic effect and reducing side
   effects.

   Based on this, the present study constructs a pH-responsive hydrogel
   nanocomposite (Gel@PLGA@FA) as a shikonin delivery platform, aiming to
   enhance the immunomodulatory effect of shikonin on OS through targeted
   delivery and controlled release, thereby improving the therapeutic
   efficacy of atezolizumab.

Materials and methods

Differential gene screening and functional analysis of Shikonin targeting
osteosarcoma

   First, we normalized the OS-related gene data (3501 genes in total)
   from GEO database ([44]GSE14359) using GEOquery toolkit and intersected
   it with 84 potential targets of Shikonin predicted by prediction
   databases, and Venn diagrams were drawn to show the intersection. To
   identify differentially expressed genes (DEGs), we used the limma
   algorithm package to analyze [45]GSE14359 transcriptome data (log2 |FC|
   ≥ 1, P < 0.05) and used the ggplot2 package to draw volcano plots to
   visually display the distribution of DEGs. Subsequently, we performed
   multidimensional clustering and heatmap analysis of selected
   differential genes by the ComplexHeatmap package, thereby revealing
   heterogeneous gene expression profiles between different sample groups.
   Further combined with Venn diagram analysis, we identified a common set
   of genes between potential targets of Shikonin and OS-related genes.
   Finally, we performed GO functional enrichment and KEGG pathway
   enrichment analysis of these key target genes using clusterProfiler
   package to systematically annotate the functions of related genes and
   the biological processes involved.

Establishment of a key target prognostic risk model for Shikonin targeting
osteosarcoma

   By combining network pharmacology with transcriptomics, we selected 28
   key targets of Shikonin acting on OS and applied them to a LASSO
   regression model (glmnet package and survival packages) to calculate
   risk coefficients and then construct a risk score model. Subsequently,
   we used the key genes included in this model to establish nomograms
   (rms and survival packages) for predicting the survival prognosis of OS
   patients. To assess the predictive accuracy of the model, we used the
   concordance index (C-index) and the calibration curve for validation.
   At the same time, the clinical utility of the model was further
   evaluated by decision curve analysis (DCA) to ensure its potential
   value in guiding clinical decision-making.

Clinical relevance and prognostic analysis of key targets targeted by
Shikonin in osteosarcoma

   The four key targets finally identified for Shikonin-targetied OS were
   validated for prognostic relevance with the survival analysis toolkit.
   Kaplan-Meier survival analysis was used to plot survival curves to
   reveal a significant association between the target gene and patient
   survival. Further, the regulatory relationships between core genes and
   clinicopathological characteristics (such as gender, metastatic status,
   etc.) were analyzed using multivariate regression models to determine
   their predictive value for disease progression. Finally, based on the
   core algorithm of CIBERSORT (CIBERSORT.R script analysis), using the
   CIBERSORTx website ([46]https://cibersortx.stanford.edu/), 22 immune
   cell markers were used to calculate cyclin-dependent kinase 1 (CDK1)
   immune infiltration in OS.

Molecular dynamics assessment of Shikonin molecular binding to CDK1

   Based on the CHARMM36 force field, we constructed the CDK1 crystal
   structure and the three-dimensional conformation of Shikonin, and
   completed the initialization setup of the complex system by generating
   topology files, using the TIP3P water model, and adding 0.15 M NaCl for
   ion neutralization. In order to eliminate steric hindrance, we first
   implemented the energy optimization of the steepest descent method of
   5000 steps. Then, in order to reach the equilibrium state, we
   sequentially performed the equilibrium treatment of 10 nanoseconds each
   for the NVT ensemble at a temperature of 310 K (temperature controlled
   using the V-rescale method) and the NPT ensemble with a pressure of
   1 bar (pressure controlled using the Parrinello-Rahman method).
   Subsequently, all-atom explicit solvent simulations were performed for
   a total of 70 nanoseconds with a time step of 2 femtoseconds, and
   trajectory frames were acquired every 10 picoseconds. At the analysis
   stage, we cluster trajectories using the gmx cluster tool to identify
   dominant binding conformations. At the same time, using the built-in
   tool in the Gromacs software package, a series of indicators were
   calculated to evaluate the system stability, including root mean square
   deviation (RMSD), root mean square fluctuation (RMSF), radius of
   gyration (Gyrate), accessible surface area (SASA), and hydrogen bond
   formation (HBond). In addition, we calculated the binding free energy
   between Shikonin and CDK1 using MM/GBSA method in order to further
   understand the interaction between the two.

Design and synthesis of pH responsive hydrogel nanomaterial complex

   Shikonin (50 mg, HY-NO882, MCE, China) was co-dissolved with the
   amphiphilic copolymer PLGA-PEG-NH[2] (100 mg) in 5 mL chloroform to
   prepare an oily phase solution. A 1% aqueous solution of polyvinyl
   alcohol (PVA) (50 mL) was also prepared, and the oil phase was evenly
   dispersed in the aqueous phase to form a microemulsion by probe
   sonication (20 kHz, power 300 W for 5 min). The emulsion was
   subsequently transferred to a three-necked flask and organic solvents
   were continuously volatilized for 4 h at 500 rpm under stirring. The
   resulting nanoparticles were purified by centrifuged three times at
   8000 × g for 15 min using 100 kDa ultrafiltration tubes to finally
   obtain a concentrated suspension of Shikonin@PLGA-PEG-NH[2]
   nanoparticles. In the folic acid coupling step, 10 mg of folic acid was
   dispersed in 10 mL MES buffer at pH 5.5, 60 mg EDC and 100 mg NHS were
   added, and the carboxyl group was activated for 30 min at 37 ℃ with
   shaking. The activated mixture was subsequently mixed with the
   nanoparticle suspension described above and supplemented with borate
   buffer (pH 8.5) to a total volume of 20 mL, and the reaction was
   continued for 12 h at room temperature to complete conjugation. At the
   end of the reaction, unbound folic acid molecules were removed by
   centrifugation using 100 kDa ultrafiltration to finally obtain a
   Shikonin@PLGA@PEG@FA targeted drug delivery system. For construction of
   pH-responsive hydrogels, 2 g dextran (DEX) was first dissolved in 15 mL
   deionized water and 1.2 g NaIO[4] was added after stirring until
   complete dissolution and the reaction was performed for 6 h under
   nitrogen atmosphere protection. The reaction was subsequently stopped
   by adding 1.6 g glycerol and the product was transferred to a 10 kDa
   dialysis bag for 24 h dialysis, followed by freeze-drying and
   reconstitution in 20 mL deionized water. During magnetic stirring, 0.5
   mL ethylenediamine was added to continue the reaction for 6 h, and the
   reaction solution was dialyzed again in a 10 kDa dialysis bag for 24 h,
   and finally freeze-dried and reconstituted in 10 mL deionized water to
   obtain hydrogel material with pH response characteristics. In order to
   characterize the structural features of the constructed composite
   system, scanning electron microscopy (SEM) was used to observe its
   micromorphology and confirm that it had a significant three-dimensional
   network structure; transmission electron microscopy (TEM) was also used
   to analyze the morphology of nanoparticles. Effective encapsulation of
   Shikonin and successful implementation of folic acid modification were
   assessed by zeta potential detection. In addition, dynamic light
   scattering (DLS) was used to measure particle size changes under
   different treatment conditions, further verifying drug loading and
   surface modification.

Cell culture

   The human OS cell lines Saos-2 and MG63 were from Wuhan Servicebio
   (Wuhan, China), while the human renal tubular epithelial cell HK-2,
   human liver parenchymal cell L-02, and human osteoblast cell line
   hFOB1.19 were obtained from China Center for Type Culture Collection.
   All cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM,
   Gibco, Carlsbad, CA) containing 10% fetal bovine serum (FBS, PAN,
   Germany) and maintained in a 5% CO[2] incubator at 37℃.

Cell counting kit-8 (CCK-8) proliferation assay

   OS cells were seeded in 96-well plates at a density of 2 × 10³ cells
   per well. After the cells completed adherent growth, the following
   treatments were administered: Shikonin (500 nM), PD-L1 inhibitor
   atezolizumab (10 µg/mL), targeted nanocomplex Gel@PLGA@Shikonin@FA at
   10 µM, and the combination of Gel@PLGA@Shikonin@FA and PD-L1 inhibitor.
   After 6 h of treatment in each group, fresh complete medium was
   replaced and cells were maintained in a 37 ℃ incubator containing 5%
   CO[2] and tested at three time points: 24 h, 48 h, and 72 h. After
   reaching the designated time point, CCK-8 solution prepared with
   complete medium was changed to react for 2 h in the dark. Subsequently,
   absorbance values at 450 nm wavelength were measured using a
   full-wavelength microplate (Synergy H1, Bio Tek, USA) reader to assess
   the viability of cells under different treatment conditions.

EdU proliferation assay

   First, treated OS cells were seeded in 6-well plates and grown to
   approximately 70% confluency under routine culture conditions.
   Subsequently, EdU solution (KTA2031, Abbkine) was added to the culture
   system at a final concentration of 10 µM and incubation was continued
   for 2 h to allow efficient integration of EdU into the DNA strand being
   synthesized. Following completion of labeling, cells were washed twice
   with PBS buffer to remove unincorporated EdU. Cells were then fixed in
   4% paraformaldehyde for 15 min and subsequently permeabilized in 0.5%
   Triton X-100 in PBS for 10 min. According to the instructions for use
   of EdU detection Kit (KTA2031, Abbkine), add Apollo staining reaction
   solution and incubate for 30 min in the dark to make Apollo fluorescent
   dye click chemical reaction with labeled EdU to form a stable
   fluorescent complex. At the end of the staining reaction, cells were
   again washed twice with PBS and subsequently stained for nuclei using
   Hoechst 33,342 staining solution for 10 min. Finally, images were
   collected using fluorescence microscopy (CKX53, Olympus Corporation,
   Japan) for assessment of cell proliferation.

Evaluation of uptake efficiency and cytotoxicity in osteosarcoma cells

   To systematically assess the uptake kinetics and biocompatibility of
   pH-responsive hydrogel nanodelivery systems, especially the uptake
   differences between the targeted modified Gel@PLGA@DiO@FA and the
   non-targeted form Gel@PLGA@DiO in OS cells, we used a DiO fluorescent
   probe labeling strategy. The intracellular uptake efficiency of the two
   nanocarriers was quantitatively analyzed by confocal laser scanning
   microscopy (LSM880, Carl Zessi, Germany) to detect the intracellular
   green fluorescence intensity (excitation wavelength 488 nm), so as to
   evaluate the effect of targeted modification on the cellular uptake
   ability. In addition, to comprehensively understand the biosafety of
   the constructed Gel@PLGA@Shikonin@FA thermosensitive hydrogel
   nanocomplex, we performed CCK-8 cell viability assays in a variety of
   normal human cell lines (including tubular epithelial cell HK-2,
   hepatic parenchymal cell L-02, and osteoblast hFOB1.19) to assess their
   concentration-dependent cytotoxic responses. We systematically assessed
   the biosafety of pH-responsive hydrogel nanosystems by calculating the
   half maximal inhibitory concentration (IC50) and determining the
   maximum non-toxic effect concentration.

Quantitative reverse transcription PCR (qRT-PCR)

   According to the manufacturer ‘s instructions, we used TRIzol solution
   (Acuracy) to extract total RNA. Subsequently, reverse transcription and
   real-time quantitative PCR analysis were performed using PrimeScript RT
   Master Mix (Takara Bio) and SYBR Premix Ex Taq Kit (Takara Bio). All
   results for gene expression were normalized against GAPDH and relative
   quantitation was performed by the 2^−ΔΔCT method.

Apoptosis by flow cytometry

   MG63 cells were seeded at a density of 2 × 10^5 cells per well in
   6-well plates, and after complete attachment of the cells, the
   following treatments were administered: Shikonin (500 nM), PD-L1
   inhibitor atezolizumab (10 µg/mL), Gel@PLGA@Shikonin@FA (10µM), and
   Gel@PLGA@Shikonin@FA combined with PD-1 inhibitor treatment groups,
   while DMSO treatment was used as the control group. All treatments were
   performed in the presence of complete medium for 24 h. At the end of
   treatment, cells from each group were collected and washed with
   pre-chilled PBS to remove residual medium. Subsequently, cells were
   digested with 0.25% trypsin without EDTA for 2 min at 37 °C to prepare
   single cell suspensions. According to the operating instructions of
   Annexin V-FITC/PI Apoptosis Detection Kit, cells were resuspended in
   100 µL binding buffer, 5 µL Annexin V-FITC and 10 µL PI staining
   solution were added sequentially and incubated for 5 min in the dark.
   Following completion of the reaction, 400 µL of buffer was added to
   stop staining. Finally, the samples were detected by flow cytometry
   (Cytoflex, Beckman, USA), and the distribution of subsets at each stage
   of apoptosis was analyzed by FITC and PI dual fluorescence channels to
   assess the effect of different treatments on apoptosis.

JC-1 assay for detecting changes in mitochondrial membrane potential

   First, 1 × 10^6 MG63 cells were seeded in confocal dishes for culture.
   The following day, confocal dishes were supplemented with 0.5 mL of
   JC-1 dye (C2005, Beyotime Bio) and Hoechst 33,342 stain solution
   (C0031, Solaibao) and incubated at 37℃ for 25 min. Afterwards, stained
   cells were observed and images were taken using confocal microscopy
   (LSM880, Carl Zessi, Germany).

Statistical analysis

   All data were presented as mean ± standard deviation (Mean ± SD) from
   at least three independent experiments, each performed in triplicate.
   Statistical differences between groups were compared using unpaired
   Student’s t-test; differences between multiple groups were assessed
   using one-way or two-way analysis of variance (ANOVA), respectively.
   Differences were considered statistically significant when P < 0.05 and
   were marked as “*”; P < 0.05 were marked as “**”; P < 0.001 were marked
   as “***”; and P < 0.0001 were marked as “#”.

Results

Differential gene screening of Shikonin targeting osteosarcoma

   Through analysis of the [47]GSE14359 dataset, we identified a series of
   DEGs. First, DEGs identified in the [48]GSE14359 dataset (|LogFC| >1&
   P < 0.05) clearly revealed a large number of upregulated (n = 944) and
   downregulated (n = 775) genes by volcano plots (Fig. [49]1A). Next, a
   heatmap of the top 50 DEGs (Fig. [50]1B) demonstrates the differences
   in expression levels of these genes between OS and normal samples. To
   further explore the relationship between Shikonin and OS, we showed the
   intersection of DEGs in [51]GSE14359 with Shikonin targeted genes by
   Venn diagram (Fig. [52]1C) and found that a total of 28 genes
   overlapped in these two types of genes. These 28 intersection genes
   were subsequently ranked according to their log2FC and presented as a
   ranked difference plot to highlight the most significant changes in
   these genes across the cross-set (Fig. [53]1D). To show the expression
   changes of these genes more visually, we visualized these 28
   intersection genes to hight differences (15 upregulated and 13
   downregulated genes) via a gradient volcano plot (Fig. [54]1E).
   Finally, the heat map (Fig. [55]1F) compares the expression levels of
   these 28 intersection genes in OS tissues versus normal tissues,
   showing their differential levels between tumor and normal tissues
   (Fig. [56]1F). By performing functional analysis of 28 intersection
   genes, we first constructed protein-protein interaction (PPI) networks
   for these genes (Supplementary Fig. 1A), revealing interaction
   relationships between the proteins they encode. Critical hub proteins
   in the network include CDK1, PIK3R1 and PDGFRB, indicating that these
   proteins may play an important role in the development of OS. Next, we
   performed cellular component (CC) enrichment analysis of these genes
   and showed that they were mainly enrich in cellular components such as
   “transferase complex” and “protein kinase complex”, which implied that
   these genes may be involved in the formation and function of a variety
   of important intracellular complexes (Supplementary Fig. 1B). In
   addition, enrichment analysis of biological processes (BP) showed that
   these genes significantly involved biological processes such as
   “calcium homeostasis” and “ERK1 and ERK2 cascades”, suggesting that
   they may play a key role in regulating cell signaling and ion balance
   (Supplementary Fig. 1C). The enrichment results of molecular function
   (MF) showed that these genes mainly had the function of “protein
   serine/threonine/tyrosine kinase activity”, which showed their
   important roles in cell signal transduction and metabolism
   (Supplementary Fig. 1D).Further, KEGG pathway enrichment analysis
   revealed that these genes were significantly enriched in signaling
   pathways such as “EGFR tyrosine kinase inhibitor resistance” and
   “Rap1”, providing new insights into cancer therapy and drug resistance
   (Supplementary Fig. 1E). Finally, through GO classification and
   circular mapping maps of KEGG pathways (Supplementary Fig. 1F), we can
   visually see the functional annotation and pathway information of these
   genes and thus understand their roles in different biological processes
   and molecular functions more comprehensively.

Fig. 1.

   [57]Fig. 1
   [58]Open in a new tab

   Shikonin targeted differential gene screening for OS. A [59]GSE14359
   dataset volcano plot; B [60]GSE14359 dataset difference ranking top 50
   differential gene heat map; (C) Shikonin targeted OS differential gene
   intersection Venn plot; D 28 intersection gene difference ranking; E 28
   intersection gene gradient volcano plot; F 28 intersection gene
   difference heat map in OS and normal tissues

Construction of Shikonin targeting prognostic model in patients with
osteosarcoma

   In this study, we first screening for prognostic differential genes in
   OS patients treated with Shikonin using a LASSO regression model
   (Supplementary Fig. 2A). The partial likelihood deviance plot showed a
   decreasing trend with increasing log(λ), reaching a minimum at
   approximately log(λ) = − 3, which suggests an optimal λ value of
   approximately 0.001, indicating that the genes selected at this point
   are most relevant for predicting the prognosis of OS patients treated
   with Shikonin (Supplementary Fig. 2B). As λ decreases, more genes are
   selected into the model, and the coefficients of some genes increase or
   decrease dramatically, highlighting their importance for prognostic
   models. Based on the four selected key genes, we constructed a risk
   score model for patients and divided patients into high-risk and
   low-risk groups according to the risk score (Supplementary Fig. 2C).
   Kaplan–Meier survival curves showed a significant difference in
   survival between the two groups, with the high-risk group having a
   shorter median survival time. In addition, we also plotted a heatmap to
   show key target genes (SCD, CCR2, NR3C2, CDK1) and their expression
   patterns in patients with different risk groups, revealing clear
   differences in gene expression levels (Supplementary Fig. 2D). To
   provide a quantitative tool for individualizing prognosis, we developed
   a nomogram prediction model that combined multiple clinical factors
   such as age, sex, metastatic status, and expression levels of key genes
   (Supplementary Fig. 2E). The total number of points can be transformed
   into the probability of survival at 1, 3, and 5 years. The calibration
   curves showed good agreement between the observed partial survival
   probability and the nomogram predicted survival probability, indicating
   that our prediction model had satisfactory predictive accuracy for
   survival outcomes at 1 year (Supplementary Fig. 2F), 3 years
   (Supplementary Fig. 2G), and 5 years (Supplementary Fig. 2H). In
   summary, through this study, we not only identified the key genes of
   Shikonin targeted therapy for OS, but also established a prognostic
   prediction model of Shikonin targeted therapy for OS, providing a
   scientific basis for clinical application. To assess the prognosis of
   key targets in OS, we first assessed the clinical utility of Shikonin
   targeting as a key target for OS (SCD, CCR2, NR3C2, CDK1) by decision
   curve analysis (DCA) at 1, 3 ,5 years (Supplementary Fig. 3A-C). The
   results showed that the net benefit using the risk score model was
   higher than the strategy of “all patients treated” or “none treated”
   over a wide range of threshold probabilities. The curves for each key
   target also showed significant net benefit compared to these two
   extreme strategies, indicating their clinical value in predicting
   short- and long-term survival. In addition, we also plotted receiver
   operating characteristic (ROC) curves for these key targets to assess
   their predictive performance. The results showed areas under the curve
   (AUC) of 0.840 for SCD, 0.852 for CCR2, 0.973 for NR3C2, and 0.977 for
   CDK1. These high AUC values indicate a strong discriminatory power of
   these targets in distinguishing different survival outcomes
   (Supplementary Fig. 3D). Based on selected key targets, we constructed
   an overall survival risk score for patients and compared overall
   survival between the low-risk group and the high-risk group by
   Kaplan-Meier survival curves. The results showed that patients in the
   high-risk group had significantly worse survival outcomes (HR = 2.96,
   95% CI = 1.71–5.10, P < 0.001; Supplementary Fig. 3E). Further, we
   performed survival analysis separately for each key target. High SCD
   expression levels were associated with worse survival (Supplementary
   Fig. 3F). Similarly, high CDK1 expression levels were also associated
   with worse survival (Supplementary Fig. 3G). In contrast, high
   expression of CCR2 (Supplementary Fig. 3H) and NR3C2 (Supplementary
   Fig. 3I) was associated with better survival. In order to better
   investigate the role of targeted genes that are highly expressed and
   have a poor prognosis in OS, SCD and CDK1 were selected for in-depth
   study.

Clinical characteristics of key targets of Shikonin targeted therapy for
osteosarcoma

   To investigate the role of CDK1, a key target of Shikonin targeted
   therapy for OS, and SCD in OS, we first analyzed the pan-cancer
   differential expression of CDK1, a key target, in a variety of cancer
   types. The results showed that CDK1 was significantly overexpressed in
   tumor samples from multiple cancer types (e.g., ACC, BLCA, BRCA, etc.)
   compared to normal tissues. These results suggest that CDK1 may play an
   important role in tumorigenesis in multiple cancer types (Fig. [61]2A).
   Next, we investigated the relationship between the expression levels of
   key targets (SCD and CDK1) and sex in OS patients. The analysis found
   that SCD expression levels were not significantly different between
   male and female patients, while CDK1 expression was higher in female
   patients than in male patients, suggesting that sex may have some
   effect on CDK1 expression (Fig. [62]2B). Further, we investigated the
   relationship between expression levels of key targets (SCD and CDK1)
   and metastatic status. The results showed that the expression level of
   CDK1 was significantly upregulated in patients with metastasis, which
   suggests that it may play a role in promoting OS metastasis
   (Fig. [63]2C). In addition, we analyzed the expression levels of key
   targets (SCD and CDK1) in relation to tumor residual status and found
   that CDK1 expression levels were higher in patients with tumor residual
   (R1 and R2) than in patients with complete resection (R0). This finding
   suggests that these targets may be involved in tumor persistence and
   may serve as markers to predict residual disease after surgery
   (Fig. [64]2D). Therefore, we chose CDK1 for further investigation.
   Finally, we investigated the association between CDK1 expression and
   immune cell infiltration in OS. The heatmap showed that the proportion
   of activated NK cells, M1 macrophages, and activated dendritic cells
   increased, while the proportion of resting immune cells (such as B
   cells and resting NK cells) was relatively low in the high CDK1
   expression group. This suggests that CDK1 may modulate the immune
   microenvironment by promoting activation of certain immune cell types,
   thereby potentially influencing immune responses in OS (Fig. [65]2E).
   In summary, this study not only reveals the clinical features of CDK1,
   a key target of Shikonin targeted therapy for OS, and its association
   with immune cells, but also provides a new perspective for
   understanding the role of CDK1 in tumor biology. These findings lay the
   foundation for further exploration of Shikonin as a novel strategy for
   OS treatment.

Fig. 2.

   [66]Fig. 2
   [67]Open in a new tab

   Clinical features and immunoassay of key targets of Shikonin targeting
   OS. A Pan-cancer differential expression of the key target CDK1; B
   Correlation of the key target with sex in OS; C Correlation of the key
   target with metastasis in OS; D Correlation of the key target with
   residual tumor in OS; E Correlation of the key target CDK1 with immune
   cells. *, P < 0.05; **, P < 0.01; ***, P < 0.001

Shikonin interacts with the CDK1 molecule

   To assess the binding ability between Shikonin and CDK1, we performed
   molecular dynamics simulation analysis. First, the root mean square
   deviation of the Shikonin-CDK1 complex suggests that Shikonin maintains
   a stable conformation in the CDK1 binding pocket throughout the
   simulation, which suggests a strong and sustained interaction between
   Shikonin and CDK1 (Fig. [68]3A). Next, the results of structural
   stability of Shikonin-CDK1 complexes over time showed that both control
   (CDK1 alone) and Shikonin-bound CDK1 were relatively structurally
   stable, but Shikonin-bound CDK1 showed lower RMSD values, indicating
   that ligand binding enhanced its stability (Fig. [69]3B). In addition,
   dynamic stability of certain atomic regions in CDK1 was shown by RMSF
   analysis, and it was found that the motility of some specific residues
   decreased when CDK1 was combined with Shikonin, suggesting that this
   ligand was able to stabilize these regions (Fig. [70]3C). Meanwhile,
   the tightness of the overall structure of the Shikonin-CDK1 complex was
   assessed by radius of Gyrate analysis, which showed that Gyrate values
   remained relatively constant throughout the simulation, indicating that
   the complex maintained a stable and compact structure (Fig. [71]3D).
   Further, the XVG-box comparison provides a thermodynamic comparison of
   the Shikonin-CDK1 complex under different conditions, and the boxplot
   shows that the presence of Shikonin makes the complex more stable, with
   fewer thermal fluctuations and better thermodynamic properties than the
   unbound state (Fig. [72]3E). Meanwhile SASA curves showed that binding
   of Shikonin reduced the exposed surface area of CDK1 and contributed to
   stabilization of the complex (Fig. [73]3F). The number of hydrogen
   bonds between Shikonin and CDK1 varied over time, and hydrogen bond
   analysis showed multiple stable hydrogen bonds formed throughout the
   simulation, supporting strong interactions between ligand and protein
   (Fig. [74]3G). In addition, the evolution of the secondary structure of
   the Shikonin-CDK1 complex over time suggests that binding of Shikonin
   does not significantly alter the secondary structure of CDK1,
   indicating that the ligand achieves binding without causing major
   conformational changes (Fig. [75]3H). The contribution of each residue
   to the total energy identifies that some key residues play an important
   role in stabilizing the complex, highlighting the importance of
   specific interactions for overall binding affinity (Fig. [76]3I). The
   energy landscape of the Gibbs free energy of the Shikonin-CDK1 complex
   as a function of the two principal components shows that the complex is
   in a low-energy state, indicating that it has a stable and favorable
   binding mode (Fig. [77]3J). Finally, 3D free energy topography provides
   a visual representation of the most stable conformation of the complex,
   further confirming a strong and stable interaction between Shikonin and
   CDK1 (Fig. [78]3K). Taken together, the results of molecular dynamics
   simulations show that Shikonin exhibits strong binding ability to CDK1,
   as evidenced by stable conformations, enhanced structural stability,
   reduced atomic fluctuations, compact structures, reduced surface
   exposure, sustained hydrogen bonding, unchanged secondary structures,
   significant residue energy contributions, low-energy states, and
   overall structural stability. These findings provide an important basis
   for an in-depth understanding of the molecular interaction mechanism
   between Shikonin and CDK1 and also support its potential as a
   therapeutic agent targeting CDK1.

Fig. 3.

   [79]Fig. 3
   [80]Open in a new tab

   Molecular dynamics assessment of Shikonin binding ability to CDK1. A
   Frequency profile of Shikonin-CDK1 complexes; B Structural stability of
   Shikonin-CDK1 complexes over time; C Dynamic stability of atoms in some
   regions of CDK1; D Closeness of the overall structure of Shikonin-CDK1
   complexes; E Comparison of XVG-box thermodynamics of Shikonin-CDK1
   complexes under different conditions; F Degree of molecular surface
   exposure of Shikonin-CDK1 complexes; G Number of hydrogen bonds in
   Shikonin-CDK1 complexes over time; H Secondary structure of
   Shikonin-CDK1 complexes over time; I Contribution of each residue to
   total energy in Shikonin-CDK1 complexes; J Gibbs energy of
   Shikonin-CDK1 complexes changes with the two principal components; K 3D
   structure of Shikonin-CDK1 complexes

Inhibitory effect of Shikonin on proliferation and toxicity of osteosarcoma
cells

   To investigate the effect of Shikonin in OS cells in vitro, we first
   determined the half maximal inhibitory concentration (IC50) of Shikonin
   in MG63 cells and showed that its IC50 value was 1.73 µM. Cell
   viability decreased significantly with increasing concentration of
   Shikonin, indicating its strong inhibitory effect on MG63 cells
   (Fig. [81]4A). Similarly, Shikonin showed significant cell growth
   inhibition in Saos-2 cells with IC50 value of 1.38 µM (Fig. [82]4B).
   CCK-8 assay was used to assess the time-dependent effect of Shikonin on
   MG63 and Saos-2 cell proliferation. For MG63 cells, cell viability in
   the Shikonin treated group was significantly reduced compared to the
   DMSO control group at 48 and 72 h, indicating that it effectively
   inhibited cell proliferation (Fig. [83]4C).Similar results were
   observed in Saos-2 cells, showing that Shikonin significantly inhibited
   proliferation of these cells in a time-dependent manner
   (Fig. [84]4D).To assess the potential toxicity of Shikonin on normal
   cells, we examined its effects on L0-2 liver epithelial cells, HK-2
   tubular epithelial cells, as well as hFOB1.19 osteoblasts. The results
   showed that although Shikonin showed some cytotoxicity to these normal
   cells, its IC50 values were 5.69 µM (L0-2 cells, Fig. [85]4E), 10.28 µM
   (HK-2 cells, Fig. [86]4F) and 6.16 µM (hFOB1.19 cells, Fig. [87]4G),
   which were much higher than their IC50 values for OS cells, suggesting
   that Shikonin has some selective toxicity to cancer cells. Finally, the
   inhibitory effect of Shikonin on Saos-2 cell proliferation was visually
   observed and quantified by EdU assay. After 24 h of Shikonin treatment,
   the number of EdU positive cells was significantly reduced, confirming
   its ability to inhibit DNA synthesis and cell proliferation.
   Quantitative analysis showed that the proportion of EdU positive cells
   in the Shikonin-treated group was significantly lower than that in the
   DMSO control group (Fig. [88]4H). In summary, this study not only
   revealed that Shikonin had a significant antiproliferative effect on OS
   cells (MG63 and Saos-2), but also assessed its toxicity to normal
   liver, kidney and osteoblasts cells, finding that it exhibited some
   toxicity to these normal cells.

Fig. 4.

   [89]Fig. 4
   [90]Open in a new tab

   Effect and toxicity assessment of Shikonin in OS cells. A IC50 of
   Shikonin in MG63 cells, n = 5; B IC50 of Shikonin in Saos-2 cells,
   n = 5; C CCK-8 to assess the inhibition of proliferation of MG63 cells
   by Shikonin, n = 5; D CCK-8 to assess the inhibition of proliferation
   of Saos-2 cells by Shikonin, n = 5; E assessment of the toxic effect of
   Shikonin on normal epithelial cells of the liver (L0-2), n = 5; F
   assessment of the toxic effect of Shikonin on normal epithelial cells
   of the renal tubules (HK-2), n = 5; G assessment of the toxic effect of
   Shikonin on osteoblasts (hFOB1.19), n = 5; and H EdU to assess the
   inhibitory effect of Shikonin on the proliferation of OS cell lines
   (Saos-2). ***, P < 0.001

Synthesis of Shikonin complex containing pH-response intelligent hydrogel
nanomaterial

   In order to solve the problem of poor water solubility and normal
   cytotoxicity of Shikonin, we synthesized a complex loaded with Shikonin
   in pH-responsive intelligent hydrogel nanomaterial (Gel@PLGA@FA). To
   verify the successful synthesis, we first analyzed the particle size
   distribution of PLGA nanoparticles by DLS, which showed that their
   average diameter was about 72 nm, and they had uniform particle size
   distribution and good synthesis efficiency (Fig. [91]5A). Further, DLS
   analysis of PLGA@Shikonin nanoparticles revealed a slight increase in
   particle size to about 91.6 nm, indicating that Shikonin was
   successfully encapsulated in PLGA matrix without significantly altering
   the morphology of the nanoparticles (Fig. [92]5B). For PLGA@Shikonin@FA
   nanoparticles, DLS data showed a further increase in particle size to
   approximately 132 nm, which suggests that FA was successfully attached
   to the nanoparticle surface (Fig. [93]5C). To assess the stability of
   these nanoparticles, we measured their zeta potential. PLGA
   nanoparticles showed good colloidal stability at a zeta potential of
   approximately 2.15 mV (Fig. [94]5D). However, the zeta potential of
   PLGA@Shikonin nanoparticles was slightly reduced to about 7.95 mV,
   indicating that the encapsulation of Shikonin did not significantly
   affect the overall charge of nanoparticles and still maintained good
   colloidal stability (Fig. [95]5E). The zeta potential of
   PLGA@Shikonin@FA nanoparticles was about − 44.35 mV, indicating that
   the attachment of folic acid enhances the negative electricity of the
   nanoparticle surface and may help to improve its stability and
   targeting efficiency (Fig. [96]5F). SEM images showed that the
   Gel@PLGA@Shikonin@FA composite had a porous internal structure, in
   which the nanoparticles were evenly distributed and showed a smooth
   spherical morphology, indicating that the nanoparticles were
   successfully embedded in the hydrogel network (Fig. [97]5G). Higher
   magnification TEM images showed the morphology of individual
   nanoparticles more clearly, showed consistent sizes and shapes, and
   confirmed the uniform distribution of nanoparticles in the hydrogel
   (Fig. [98]5H). Subsequently, we characterized PLGA using a fourier
   transform infrared (FTIR) spectrometer. The results revealed a distinct
   absorption peak in the range of 1701 cm^− 1, which corresponds to the
   infrared absorption of the carbonyl (C = O) bond within the ester
   groups (-COO-) in the PLGA molecule. This observation confirms the
   successful synthesis of PLGA. Then, the drug loading capacity and
   encapsulation efficiency of the hydrogel nanocomposite were determined
   using UV-vis spectrophotometry, yielding values of 5.6% and 85.8%,
   respectively. Finally, the pH - responsive drug release profiles
   revealed the drug release behavior of the Gel@PLGA@Shikonin@FA
   composite under different pH conditions. The results showed that
   Shikonin exhibited a continuous release trend within a certain period
   of time. Specifically, the release was faster at pH 5 compared to other
   pH values (6.0, 6.5, 7.0, and 7.4). This pH - responsive characteristic
   indicated that the composite could be used for local tumor drug
   delivery and achieve controlled drug release in the tumor
   microenvironment (low pH) (Fig. [99]5J). In summary, this study not
   only showed that PLGA-based nanocomposites successfully loaded Shikonin
   and achieved folic acid functionalization, but also verified that they
   had ideal physicochemical properties, including appropriate particle
   size, charge, morphology and pH-responsive drug release behavior. These
   findings support the potential application value of this nanocomposite
   as a targeted drug delivery system in cancer therapy.

Fig. 5.

   [100]Fig. 5
   [101]Open in a new tab

   Hydrogel nanomaterial complex synthesis characterization. A DLS plot of
   PLGA, 72 ± 2.31, n = 3; B DLS plot of PLGA@Shikonin, 91.6 ± 2.87,
   n = 3; C DLS plot of PLGA@ Shikonin@FA, 132 ± 4.54, n = 3; D Zeta plot
   of PLGA, 2.15 ± 0.25, n = 3; E Zeta plot of PLGA@Shikonin, 7.95 ± 0.32,
   n = 3; F Zeta plot of PLGA@Shikonin@FA, -44.35 ± 0.87, n = 3; G SEM
   plot of Gel@PLGA@Shikonin@FA; H TEM plot of Gel@PLGA@Shikonin@FA; I
   Infrared spectroscopy chart of PLGA; J PH drug release curve of
   Gel@PLGA@Shikonin@FA

Targeted uptake capacity and cytotoxicity assessment of pH- responsive
hydrogel nanomaterial complexes

   We further assessed the targeted uptake capacity and cytotoxicity of
   pH-responsive hydrogel nanomaterial complexes. First, we observed the
   uptake of Gel@PLGA@DIO in Saos-2 cells using confocal microscopy. The
   results showed that at 0 h, very little green fluorescence was
   observed, indicating low initial uptake (Fig. [102]6A). After 24 h, the
   green fluorescence of Gel@PLGA@DIO was slightly enhanced, indicating
   that only a small amount of nanoparticles were taken up. When Saos-2
   cells were further treated with FA-functionalized nanoparticles
   (Gel@PLGA@DIO@FA), their uptake efficiency was significantly enhanced
   at 24 h (Fig. [103]6B). Saos-2 cells treated with Gel@PLGA@DIO@FA
   showed a significant increase in green fluorescence intensity at 24 h
   compared to unmodified Gel@PLGA@DIO. This suggests that FA modification
   improves the targeting ability of nanoparticles to cancer cells, due to
   folate receptors highly expressed on the surface of Saos-2 cells. Next,
   we evaluated the toxicity of this pH-responsive hydrogel nanomaterial
   complex (Gel@PLGA@Shikonin@FA) on normal cell lines. In L-02 liver
   epithelial cells, cell survival remained above 90% even at the highest
   concentration tested (200 µM/mL), indicating that this material was
   minimally toxic to normal hepatocytes (Fig. [104]6C). Similarly, no
   significant decrease in cell viability was observed in HK-2 tubular
   epithelial cells exposed to Gel@PLGA@Shikonin@FA for up to 48 h at any
   concentration tested (Fig. [105]6D), suggesting that the composite did
   not produce significant toxicity to tubular epithelial cells. Finally,
   we investigated its effect on hFOB1.19 osteoblasts. The results showed
   that cell viability remained stable and approached 100% at all time
   points and concentrations, indicating that the pH-responsive hydrogel
   nanomaterial complex was non-toxic to osteoblasts (Fig. [106]6E). In
   summary, the results of targeted uptake studies and toxicity assessment
   showed that the pH-responsive hydrogel nanomaterial complex
   (Gel@PLGA@Shikonin@FA) showed excellent targeting ability for Saos-2
   cells while maintaining low toxicity in normal hepatocytes, renal
   cells, and osteocytes. These findings highlight their potential as
   efficient and safe delivery systems for targeted cancer therapeutics.

Fig. 6.

   [107]Fig. 6
   [108]Open in a new tab

   Targeted uptake capacity and toxicity assessment of pH responsive
   hydrogel nanomaterial complexes. A Targeted uptake capacity of
   Gel@PLGA@DIO in OS cell line (Saos-2); B Targeted uptake capacity of
   Gel@PLGA@DIO@FA in OS cell line (Saos-2); C Toxicity assessment of pH
   responsive hydrogel nanomaterial complexes to normal liver epithelial
   cells; D Toxicity assessment of pH responsive hydrogel nanomaterial
   complexes to tubular epithelial cells; E Toxicity assessment of pH
   responsive hydrogel nanomaterial complexes to osteoblasts. ns, P > 0.05

pH-responsive hydrogel nanomaterial complex equipped with Shikonin enhances
anti-PD-L1 therapeutic effect in osteosarcoma

   We further evaluated the enhancing effect of Shikonin-loaded
   (Gel@PLGA@Shikonin@FA) pH-responsive hydrogel nanomaterial complexes on
   the anti-PD-L1 effect of OS cells. The results showed that in Saos-2
   and MG63 cells, CDK1 mRNA expression levels were significantly
   decreased in the Gel@PLGA@Shikonin@FA treatment group compared with
   Gel@PLGA alone (Fig. [109]7A). This suggests that Shikonin loaded in
   folate-functionalized nanoparticles is able to effectively downregulate
   CDK1 expression. Similarly, Gel@PLGA@Shikonin@FA significantly reduced
   the mRNA expression level of CD279 in these two cell lines
   (Fig. [110]7B), suggesting that this nanocomplex has the ability to
   regulate immune checkpoint molecules and may enhance the effect of
   immunotherapy. The inhibitory effect on CDK1 expression was further
   enhanced when Gel@PLGA@Shikonin@FA was combined with the anti-PD-L1
   inhibitor atezolizumab (Fig. [111]7C). In Saos-2 and MG63 cells, CDK1
   mRNA expression was significantly reduced in the Gel@PLGA@Shikonin@FA
   plus atezolizumab group compared with the Gel@PLGA plus atezolizumab
   group alone, showing a synergistic effect and highlighting the
   potential of targeted drug delivery systems in combination with
   immunotherapeutic agents. In addition, this combined treatment strategy
   also showed a significant inhibitory effect on the proliferation of OS
   cells. Both MG63 (Fig. [112]7D) and Saos-2 (Fig. [113]7E) cell lines
   showed a significant decrease in cell viability after treatment with
   Gel@PLGA@Shikonin@FA combined with atezolizumab, with OD450 values
   significantly lower than other treatment groups at all time points,
   indicating that this combination therapy potently inhibited the
   proliferation of tumor cells. Finally, flow cytometry analysis showed
   that Gel@PLGA@Shikonin@FA combined with atezolizumab promoted apoptosis
   of Saos-2 cells (Fig. [114]7F). Compared with the DMSO control group
   (apoptosis rate of 2.17%), the apoptosis rate of the combined treatment
   group increased to 33.45% (which was also significantly higher than
   that of the single group), indicating that the combined treatment had a
   significant effect in inducing tumor cell apoptosis. In summary, the
   pH-responsive hydrogel nanomaterial complex (Gel@PLGA@Shikonin@FA),
   especially in combination with atezolizumab, effectively enhances the
   anti-PD-L1 effect on OS cells by mechanisms such as down-regulation of
   CDK1 and PD-1 expression, inhibition of cell proliferation, and
   promotion of apoptosis. These findings provide a strong experimental
   basis and offer new strategies for improving OS treatment by targeted
   drug delivery systems combined with immunotherapy.

Fig. 7.

   [115]Fig. 7
   [116]Open in a new tab

   pH-responsive hydrogel nanomaterial complexes augment anti-PD-L1
   effects in OS cells. A Gel@PLGA@Shikonin@FA decreased the expression
   level of CDK1, n = 3; B Gel@PLGA@Shikonin@FA decreased the expression
   level of CD279, n = 3; C Gel@PLGA@Shikonin@FA combined with
   atezolizumab decreased the expression level of CDK1, n = 3; D
   Gel@PLGA@Shikonin@FA combined with atezolizumab inhibited the
   proliferation of MG63 cells, n = 5; E Gel@PLGA@Shikonin@FA combined
   with atezolizumab inhibited the proliferation of Saos-2 cells, n = 5; F
   Gel@PLGA@Shikonin@FA combined with atezolizumab promoted the apoptosis
   of Saos-2, n = 3. **, P < 0.01; ***, P < 0.001; #, P < 0.0001

   To further investigate the mechanism by which Gel@PLGA@Shikonin@FA
   combined with PD-L1 inhibitor atezolizumab promoted apoptosis in OS
   cells, we assessed changes in mitochondrial membrane potential using
   JC-1 staining. In DMSO-treated controls, most cells showed high
   mitochondrial membrane potential as indicated by red fluorescence
   emitted by JC-1 polymers. This suggests that these cells are in a
   healthy state and have intact mitochondria. However, a significant
   increase in green fluorescence from JC-1 monomers was observed when
   treated with Gel@PLGA@Shikonin@FA alone, indicating a decrease in
   mitochondrial membrane potential and signs of early apoptosis.
   Similarly, treatment with atezolizumab alone also resulted in an
   increase in green fluorescence, suggesting that PD-L1 inhibitors can
   induce some degree of mitochondrial dysfunction and apoptosis. However,
   the effect was more pronounced when Gel@PLGA@Shikonin@FA was combined
   with atezolizumab. In this combined treatment group, green fluorescence
   intensity was significantly higher than in the other groups, while red
   fluorescence was significantly reduced, suggesting substantial loss of
   mitochondrial membrane potential and strong apoptosis induction
   (Fig. [117]8). In summary, the results of JC-1 staining experiments
   showed that the combination of Gel@PLGA@Shikonin@FA and atezolizumab
   was able to effectively perturb mitochondrial membrane potential in OS
   cells, which resulted in more intense apoptosis. This synergistic
   effect highlights the potential of combining targeted drug delivery
   systems with immunotherapeutic agents for improving cancer treatment
   outcomes.

Fig. 8.

   [118]Fig. 8
   [119]Open in a new tab

   JC-1 detection of pH response to hydrogel nanomaterial complex
   (Gel@PLGA@Shikonin@FA) combined with PD-L1 inhibitor (Atezolizumab)
   promotes apoptosis of OS cells

Discussion

   This study represents the first attempt to encapsulate Shikonin into a
   tumor-targeted, pH-responsive nanocarrier, specifically designed to
   enhance the efficacy of PD-L1 inhibitors in osteosarcoma. Previous
   research on Shikonin combined with immunotherapy has mostly relied on
   free drug administration, which is plagued by issues such as poor
   solubility, rapid clearance, and off-target toxicity [[120]9]. In
   contrast, our nanoplatform significantly improves the bioavailability
   and tumor targeting of shikonin. Studies by Wang [[121]15] and Yu et
   al. [[122]16] further supported that the combination of Shikonin with
   PD-L1 inhibitors significantly improved antitumor efficacy and
   prolonged the survival of tumor-bearing mice. In contrast, studies have
   shown limited efficacy of atezolizumab alone in the treatment of liver
   cancer [[123]17], particularly in “cold tumors” such as OS, which
   highlight the importance of combined treatment strategies. Through
   analysis of the [124]GSE14359 dataset, we identified a series of DEGs
   and found that 28 of them overlapped among Shikonin targeted genes.
   Functional analysis of these genes has revealed their critical role in
   regulating cell signaling, ion balance, and so forth, and is
   potentially important in multiple cancer types. Based on this, we
   constructed a patient’s risk score model combining expression levels of
   key genes as well as multiple clinical factors to provide a scientific
   basis for individualized prognosis. However, the bibliometric analysis
   of the role of autophagy in the diagnosis and treatment of OS by Zhu et
   al. [[125]1] emphasizes the necessity of exploring OS treatment
   strategies from multiple perspectives.

   In addition, we investigated the molecular interaction mechanism
   between Shikonin and CDK1. Molecular dynamics simulation results showed
   that Shikonin showed strong binding ability to CDK1, with stable
   conformation, enhanced structural stability, and decreased atomic
   fluctuations. This lays the foundation for further exploration of
   Shikonin as a novel strategy for OS treatment. In contrast, recent
   advances on Shikonin in the treatment of immune-related diseases by Guo
   [[126]5] and others point out that the mechanism of action of Shikonin
   is not limited to direct anticancer effects, but also includes a wide
   range of anti-inflammatory and immunomodulatory functions. In assessing
   the effect of Shikonin in OS in vitro, we found that it had a
   significant antiproliferative effect on MG63 and Saos-2 cell lines.
   However, Shikonin also showed some toxicity to normal liver, kidney,
   and osteoblasts, but its IC50 value was much higher than that for OS
   cells, indicating that Shikonin has some selective toxicity to cancer
   cells. This phenomenon is similar to that observed by Wang et al.
   [[127]15] in studies investigating Shikonin promoting apoptosis of HCC
   cells to enhance anti-PD-L1 efficacy, although they focused on
   different biomarkers and mechanisms of action. In order to overcome the
   disadvantages of poor water solubility and normal cytotoxicity of
   Shikonin, we successfully synthesized a pH-responsive intelligent
   hydrogel nanomaterial (Gel@PLGA@FA) equipped with Shikonin complex. By
   enabling the local release of Shikonin, our nanocomposite enhances
   anti-tumor immune responses and exerts a synergistic effect with
   atezolizumab, achieving superior tumor suppression efficacy and
   prolonged survival in osteosarcoma models. Importantly, compared with
   free Shikonin, this system significantly reduces toxicity. Similarly,
   Zhang et al. [[128]18] used pH-sensitive polymer micelles to load
   chemotherapeutic drugs to achieve effective accumulation and controlled
   release at the tumor site of pancreatic cancer, significantly improving
   the therapeutic effect and reducing side effects.

   Finally, we evaluated the anti-PD-L1 therapeutic efficacy of this
   pH-responsive hydrogel nanomaterial complex on OS cells. The results
   showed that Gel@PLGA@Shikonin@FA combined with atezolizumab could
   effectively downregulate CDK1 and PD-1 expression, inhibit cell
   proliferation, and promote apoptosis, thereby significantly enhancing
   the anti-PD-L1 effect on OS cells. JC-1 staining experiments further
   confirmed that this combination therapy was able to effectively perturb
   mitochondrial membrane potential in OS cells, leading to more intense
   apoptosis. This conclusion echoes the findings of Li et al. [[129]19]
   regarding the simultaneous regulation of ROS generation and copper
   metabolism by nanomedicine for sonodynamic amplification tumor therapy,
   demonstrating the great potential of combined therapeutic strategies.

   In summary, this study not only reveals the unique mechanism of action
   of Shikonin as a potential anticancer drug, but also demonstrates the
   potential of pH-responsive hydrogel nanomaterial complexes as efficient
   and safe delivery systems for targeted cancer therapeutics. These
   findings provide a strong experimental rationale and novel strategies
   for improving OS treatment by targeted drug delivery systems combined
   with immunotherapy.

Conclusion

   Shikonin can significantly down-regulate the expression of CDK1 and
   PD-L1, inhibit OS cell proliferation and promote apoptosis through
   strong and stable binding to the key target, CDK1. In addition, we
   successfully constructed a pH-responsive hydrogel nanocomplex
   (Gel@PLGA@FA) equipped with Shikonin, which showed good drug release
   characteristics at different pH conditions, especially in the tumor
   microenvironment to achieve controllable drug release. The combined use
   of Gel@PLGA@Shikonin@FA and PD-L1 inhibitor (atezolizumab) not only
   effectively enhanced the anti-PD-L1 therapeutic effect on OS cells, but
   also led to stronger apoptosis by disrupting the mitochondrial membrane
   potential. The combination of Shikonin and atezolizumab provides an
   innovative and efficient treatment approach for OS patients. Future
   studies will further optimize the design of this complex and bring new
   hope for OS treatment.

Supplementary Information

   [130]Supplemenatary Material 1.^ (1.2MB, docx)

Acknowledgments