Abstract
The efficacy of sonodynamic therapy (SDT), an emerging approach for
tumor treatment, is hindered by the high levels of the antioxidant
glutathione (GSH) in the tumor microenvironment (TME). In this study,
we constructed nanobubbles loaded with the sonosensitizer HMME and the
tumor-targeting peptide RGD (HMME-RGD@C[3]F[8] NBs) for synergistic
SDT/cuproptosis therapy of liver hepatocellular carcinoma (LIHC) in
combination with Elesclomol-Cu as cuproptosis inducers. Endogenous GSH
is consumed by Cu^2+ to modulate the complex TME, thereby amplifying
oxidative stress and further improving SDT performance. Additionally,
intracellular Cu^2+ overload can induce cuproptosis, which is further
amplified by SDT, to initiate irreversible protein toxicity. The
specific mechanism of synergistic SDT/cuproptosis therapy in LIHC was
investigated by RNA sequencing analysis. The synergistic
SDT/cuproptosis therapy reprogrammed the TME to improve the efficacy of
immune checkpoint inhibitor-based immunotherapy. Furthermore, a
risk-scoring model was created and displayed significant promise in the
prognosis of LIHC.
Graphical Abstract
[42]graphic file with name 12951_2024_2995_Figa_HTML.jpg
Supplementary Information
The online version contains supplementary material available at
10.1186/s12951-024-02995-3.
Keywords: Cuproptosis, Sonodynamic therapy, Liver hepatocellular
carcinoma, Synergistic tumor therapy, RNA sequencing
Introduction
Malignant tumors represent a significant global health challenge that
demands a variety of therapeutic strategies including surgery,
chemotherapy, radiation therapy and immunotherapy [[43]1, [44]2].
Recently, emerging therapies including cuproptosis and sonodynamic
therapy (SDT) have received considerable attention owing to their
enhanced antitumor efficacy, noninvasiveness and ability to destroy
deep-seated tumors [[45]3–[46]6]. However, the efficacy of SDT is often
hindered by the presence of abundant intracellular antioxidants, such
as glutathione (GSH), within the tumor microenvironment (TME). These
antioxidants neutralize reactive oxygen species (ROS) to protect cancer
cells from SDT-mediated oxidative stress [[47]7, [48]8]. Therefore,
additional therapeutic strategies are needed to increase oxidative
stress and attenuate the TME to improve the therapeutic outcomes of SDT
[[49]9, [50]10].
A recently discovered cell death pathway driven by copper ions, known
as cuproptosis, has unique characteristics from established mechanisms
such as apoptosis and necroptosis [[51]3, [52]11]. However, the
efficacy of cuproptosis is limited by naturally low levels of
intracellular copper, which drives the development of ionophores
including disulfiram and elesclomol to increase the efficacy of
cuproptosis [[53]12]. Cuproptosis is closely related to mitochondrial
respiration and the lipoic acid pathway, which ultimately induces cell
apoptosis via oxidative stress [[54]13–[55]15]. Concurrently, SDT
induces the production of ROS when exposed to ultrasound, further
exacerbating oxidative stress [[56]16]. Therefore, the synergy of SDT
and cuproptosis represents a promising strategy to combat deep-seated
tumors [[57]17–[58]19].
Targeted drug delivery, an emerging treatment strategy, holds
tremendous potential in liver cancer treatment. Nevertheless, at
present, targeted drug delivery for liver cancer is confronted with
numerous challenges. The existing targeted drug delivery systems
primarily depend on nanotechnology to transport drugs to liver cancer
tissues via nanocarriers [[59]20]. However, nanocarriers in the body
are prone to being recognized and eliminated by the immune system, thus
reducing drug delivery efficiency. Additionally, the targeting
properties of nanocarriers need further enhancement to ensure accurate
delivery of drugs to liver cancer cells [[60]21]. Finally, the safety
of targeted drugs is also a crucial factor. Although targeted drugs can
enhance drug efficacy, they may simultaneously increase drug toxicity
and side effects. Therefore, when designing and developing targeted
drugs, full consideration must be given to drug safety [[61]22].
Nanobubbles have gained significant attention as delivery systems via
ultrasound-targeted microbubble destruction (UTMD), which is an
advanced technology [[62]23]. When ultrasound encounters microbubbles,
the cavitation effect and sonoporation occur, which play significant
biological roles in enhancing the transfection efficiency of genes
[[63]24]. Remarkably, it has the potential to assist nanoparticles in
breaking through the barrier through sonoporation [[64]25]. After the
microbubbles are injected into the body, followed by ultrasonic
irradiation, they continuously expand and contract, exerting an effect
on the blood vessel wall or cell membrane. As a result, the width of
the endothelial cell gap and the permeability of the cell membrane
increase. Simultaneously, the shock wave and high-speed jet generated
instantaneously when the microbubbles were destroyed could also create
large holes in the cell membrane. This helps the nanoparticles break
through the barrier and enter the tumor cells [[65]26]. NBs can
subsequently achieve a high degree of local diffusion and achieve the
goal of targeted therapy in targeted tumor tissue [[66]25].
Perfluoropropane (C[3]F[8])-loaded nanobubbles have shown high
stability with minimum premature leakage and prolonged circulation time
in vivo in previous studies [[67]27, [68]28]. However, the limited
penetration depth in tumors remains a major challenge for nanobubbles
before their clinical translation. Therefore, a variety of strategies
have been employed to increase the permeability of nanobubbles within
tumor tissues [[69]29, [70]30]. For example, the incorporation of
tumor-penetrating peptides with nanobubbles can achieve simultaneous
drug targeting and improved tumor penetration via ligand‒receptor
interactions [[71]31–[72]35].
Here we developed nanobubbles (HMME-RGD@C[3]F[8] NBs) with HMME as a
potent sonosensitizer and Elesclomol-Cu as a cuproptosis inducers for
tumor-targeted SDT and cuproptosis (Fig. [73]1A). The surface-tethered
RGD enables the nanobubbles to actively target tumor cells through
ligand‒receptor interactions, followed by SDT to eradicate tumor
tissues. In addition, ES-Cu was combined with NBs for synergistic
cuproptosis and SDT. The specific mechanism underlying the synergy
between cuproptosis and SDT was investigated by RNA-Seq, further
highlighting the clinical potential of synergistic cuproptosis and SDT
in the treatment of liver hepatocellular carcinoma (LIHC).
Fig. 1.
[74]Fig. 1
[75]Open in a new tab
Mechanism flowchart and characterization of the HMME-RGD@C[3]F[8] NBs.
A Schematic illustration of the underlying mechanism of synergistic
SDT/cuproptosis cancer therapy mediated by HMME-RGD@C[3]F[8] and ES-Cu
under ultrasound irradiation. B Representative TEM images of
HMME-RGD@C[3]F[8] NBs without or with ultrasound irradiation. C UV‒vis
spectra of HMME, C[3]F[8]-NBs and HMME-RGD@C[3]F[8] NBs. D UV‒vis
spectra of ES and ES-Cu. E DLS results showing the size distributions
of the C[3]F[8]-NBs, RGD-C[3]F[8] NBs, and HMME-RGD@C[3]F[8] NBs. F
Size change of HMME-RGD@C[3]F[8] NBs in PBS and FBS solution over
4 days (n = 3). G PDI changes of HMME-RGD@C[3]F[8] NBs in PBS and FBS
solution over 4 days (n = 3). H Release profiles of HMME after
different treatments (HMME-RGD@C[3]F[8] + US/-US)
Results
Preparation and characterization of HMME-RGD@C[3]F[8] NBs
The HMME-RGD@C[3]F[8] NBs were fabricated by self-assembly of the
sonosensitizers HMME, DSPC, and DSPE-PEG-RGD using a solvent exchange
method (Fig. [76]1A). Transmission electron microscopy (TEM) images
revealed that the HMME-RGD@C[3]F[8] NBs were uniform, spherical
nanoparticles with an approximate diameter of 220 nm (Fig. [77]1B and
Supplementary Figure S1A). The HMME-RGD@C[3]F[8] NBs exhibited the same
absorption peak at approximately 400 nm as free HMME did, indicating
that HMME was successfully loaded into the HMME-RGD@C[3]F[8] NBs
(Fig. [78]1C). The encapsulation efficiency and loading content of the
HMME-RGD@C[3]F[8] NBs were calculated to be 48.76% and 4.433%,
respectively (Figure S1C, D). In addition, the characteristic UV
absorption peaks of unbound elesclomol and ES-Cu were significantly
different, which helped to differentiate the bound elesclomol from the
unbound elesclomol (Fig. [79]1D).The size distribution of the
HMME-RGD@C[3]F[8] NBs was further measured as 215.7 ± 5.8 nm by dynamic
light scattering (DLS) (Fig. [80]1E). Monitoring the continuous changes
in size and polydispersity index (PDI) of the NBs in PBS and FBS
revealed high stability of the NBs (Fig. [81]1F, [82]G), likely owing
to their strong negative zeta potential (− 14.02 ± 0.64 mV,
Supplementary Figure S1B). The results of the drug release experiments
revealed that, compared with the group without ultrasound, the
ultrasound group exhibited a burst release phenomenon. The cumulative
drug release rate of HMME reached 70% within 6 h after ultrasound,
while the cumulative release rate of the group without ultrasound was
only 40% within 24 h (Fig. [83]1H). These findings indicate that the
HMME-RGD@C[3]F[8] NBs have the characteristic of US-responsive drug
release. The TEM results of the HMME-RGD@C[3]F[8] NBs after ultrasound
revealed that the NBs were cracked, further verifying the
ultrasound-responsive drug release characteristics of the NBs
(Fig. [84]1B).
In vitro antitumor effects of HMME-RGD@C[3]F[8] NBs combined with ES-Cu
We first examined whether DSPE-PEG-RGD modification could enhance the
cellular internalization of HMME-RGD@C[3]F[8] NBs in liver
hepatocellular carcinoma cells. The fluorescence microscopy results
revealed that the accumulation of HMME in HepG2 and Huh7 cells was
time-dependent. Stronger red fluorescence was observed in the HepG2 and
Huh7 cells incubated with the HMME-RGD@C[3]F[8] NBs than in the cells
treated with the HMME@C[3]F[8] NBs, suggesting enhanced cellular uptake
of the NBs after RGD-conjugation (Fig. [85]2A–D). Next, we conducted a
standard Cell Counting Kit-8 (CCK-8) assay, and the results indicated
that without ultrasound irradiation, the HMME-RGD@C[3]F[8] NBs
intrinsically exhibited remarkable safety. Even at a comparatively high
concentration of HMME, the cell viability was maintained above 75%
(Fig. [86]2E, [87]F). The in vitro cytotoxicity and antitumor effects
of HMME-RGD@C[3]F[8] NBs were also analyzed in HepG2 and Huh7 human
hepatic cancer cells using the CCK-8 assay, with or without the
presence of ES-Cu. Compared with individual SDT treatments, the
combination with ES-Cu markedly improved the inhibition of tumor growth
(Fig. [88]2G, [89]I and Figure S2A–H). We computed the IC50 values for
in vitro antitumor efficacy. The results demonstrated that for HepG2
cells, the IC50 values of HMME in the SDT group and the SDT + ES-Cu
group were 10.81 µg/mL and 3.766 µg/mL, respectively. For Huh7 cells,
the IC50 values of HMME in the SDT group and the SDT + ES-Cu group were
10.72 µg/mL and 4.101 µg/mL, respectively. This finding implies that
the combination of SDT and cuproptosis has favorable anticancer effects
in vitro (Fig. [90]2H, [91]J). This was also confirmed through
calcein-AM (which indicates live cells) and propidium iodide (which
labels dead cells) costaining assays. Compared with that of the cells
treated with PBS, a moderate red fluorescence of PI was observed in the
HMME-RGD@C[3]F[8] + US or ES-Cu groups, likely owing to SDT-mediated
ROS generation and Cu-induced cuproptosis, respectively (Fig. [92]3A,
[93]B). Notably, cells incubated with HMME-RGD@C[3]F[8] + US + ES-Cu
presented the highest rate on tumor cell proliferation inhibition,
indicating enhanced antitumor efficacy via synergistic cuproptosis and
SDT. Similar results were also found in the apoptosis evaluation
(Fig. [94]3C–F), demonstrating the desirable therapeutic effect of the
combination of cuproptosis and SDT on cancer cells.
Fig. 2.
[95]Fig. 2
[96]Open in a new tab
In vitro cellular uptake and cytotoxicity analysis. Intracellular
uptake of HMME-RGD@C[3]F[8] NBs in HepG2 cells (A) and Huh7 cells (B)
after different incubation times. The blue color represents
Hoechst33342-stained cell nuclei, and the red color represents the
fluorescence of HMME. C, D Quantitative analysis of their fluorescence
intensity. E, F Cytotoxicity analysis of HepG2 and Huh7 cells treated
with different concentrations of HMME-RGD@C[3]F[8] NBs without
ultrasound irradiation. G, H In vitro antitumor efficacy of SDT and
SDT + ES-Cu in HepG2 and Huh7 cells. Statistical significance was
calculated via t-test. *P < 0.05, **P < 0.01, ***P < 0.001,
****P < 0.0001
Fig. 3.
[97]Fig. 3
[98]Open in a new tab
In vitro anticancer effects detected by live/dead staining and flow
cytometry. Representative fluorescence images of HepG2 (A) and Huh7 (B)
cells stained with calcein AM (green, live cells) and propidium iodide
(red, dead cells) after different treatments (scale bar = 100 μm). C, D
The percentage of apoptotic cells was analyzed by flow cytometry after
different treatments. The total apoptosis rate was calculated by Q2
(early apoptosis) and Q3 (late apoptosis). E, F Quantification of total
apoptosis rates. The data are expressed as means ± SD (n = 3).
Statistical significance was calculated using one-way analysis of
variance (ANOVA). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001
Increased reactive oxygen species generation and depletion of GSH
As both SDT and cuproptosis induce tumor cell death by oxidative
stress, the ROS levels in HepG2 and Huh7 cancer cells were evaluated
after different treatments (Fig. [99]4A, [100]B). Faint green
fluorescence was detected in cells treated with either
HMME-RGD@C[3]F[8] + US or ES-Cu, indicating that oxidative stress was
induced by individual SDT or ES-Cu-induced cuproptosis, respectively.
In contrast, the highest level of ROS fluorescence was observed in the
cells treated with HMME-RGD@C[3]F[8] + US + ES-Cu, suggesting enhanced
ROS generation by the combination of cuproptosis and SDT. Furthermore,
we detected the ROS levels in HepG2 and Huh7 cells subjected to
different treatments, and the results were consistent with the results
of the ROS fluorescence staining experiments. The level of ROS
generated in the HMME-RGD@C3F8 + US + ES-Cu group was significantly
greater than that in the other groups, and this effect was partially
eliminated by the copper ion chelator ATTM, suggesting that cuproptosis
plays a role in the combined treatment process (Fig. [101]4C–E). JC-1
was used to monitor the depolarization of the mitochondrial membrane
potential induced by ROS. Fluorescence microscopy revealed a weak
signal of JC-1 monomers and a strong signal of JC-1 aggregates in the
control group, indicating the presence of intact mitochondrial
membranes in untreated tumor cells. However, significantly increased
monomer signals and reduced aggregation signals were observed in cells
exposed to HMME-RGD@C[3]F[8] + US + ES-Cu, suggesting a reduction in
the mitochondrial membrane potential caused by abundant ROS
(Fig. [102]5A, [103]B). As the overexpression of the endogenous
antioxidant GSH inevitably compromises the efficacy of ROS-based
antitumor therapies, the intracellular GSH levels in HepG2 and Huh7
cancer cells were measured after different treatments. The
HMME-RGD@C[3]F[8] + ES-Cu group substantially reduced the levels of GSH
in cancer cells subjected to ultrasound irradiation, likely owing to
the oxidization of GSH into GSSG by Cu^2+ (Fig. [104]5C, [105]D). The
MDA level (end products of lipid peroxidation) was also monitored after
different treatments (Fig. [106]5E, [107]F). The highest MDA level was
found in cells exposed to the HMME-RGD@C[3]F[8] + US + ES-Cu,
indicating enhanced oxidative stress induced by the combination of
cuproptosis and SDT. Taken together, the combination of SDT and
cuproptosis induced considerable ROS generation under US irradiation
and eliminated GSH to reinforce intracellular oxidative stress, thus
leading to efficient ablation of deep-seated tumors.
Fig. 4.
[108]Fig. 4
[109]Open in a new tab
Identification of ROS levels. A, B Fluorescence image analyses of
intracellular ROS generation as indicated by DCFH-DA detection after
receiving different treatments as indicated. (Scale bar = 50 μm). C ROS
levels in HepG2 and Huh7 cells detected by flow cytometry after
different treatments as indicated. D, E Quantification of the ROS
levels in HepG2 and Huh7 cells detected by flow cytometry. Statistical
significance was calculated using one-way analysis of variance (ANOVA).
*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001
Fig. 5.
[110]Fig. 5
[111]Open in a new tab
Analysis of the mitochondrial membrane potential (MMP), MDA levels and
GSH levels in synergistic SDT/cuproptosis. A, B Fluorescence images of
JC-1 monomers (green channel) and aggregates (red channel) in the
mitochondria of HepG2 and Huh7 cells after different treatments as
indicated. (scale bar = 50 μm). C, D DTNB assay of GSH levels in HepG2
and Huh7 cells under different treatments. E, F MDA levels in HepG2 and
Huh7 cells subjected to different treatments. Statistical significance
was calculated via one-way analysis of variance (ANOVA). *P < 0.05,
**P < 0.01, ***P < 0.001, ****P < 0.0001
Collaborative effects of cuproptosis and SDT
To confirm the effect of cuproptosis on HepG2 and Huh7 carcinoma cells,
we evaluated the antitumor efficacy of cuproptosis via various
treatments, including PBS (i), HMME-RGD@C[3]F[8] + US (ii), ES-Cu (iii)
and HMME-RGD@C[3]F[8] + US + ES-Cu (iv) (Figure S3A and B). The
expression levels of the Fe-S cluster proteins FDX1 and lipoyl synthase
(LIAS) in the ES-Cu (iii) group were significantly lower than those in
the PBS group (i), indicating the successful induction of cuproptosis
by ES-Cu. In contrast with the ES-Cu (iii) group, the
HMME-RGD@C[3]F[8] + US + ES-Cu (iv) group markedly reduced the
expression levels of FDX1 and LIAS, suggesting that enhanced
cuproptosis was induced by combined cuproptosis and SDT via GSH
depletion and ROS generation.
The antitumor efficacy of HMME-RGD@C[3]F[8] NBs + US + ES-Cu in vivo
Once the tumor volume reached around 100 mm^3, BALB/c nude mice that
had HepG2 cells subcutaneously implanted were randomly divided into
four groups: control group, HMME-RGD@C[3]F[8] + US group, ES-Cu group,
and HMME-RGD@C[3]F[8] + US + ES-Cu group. Initially, HMME-RGD@C[3]F[8]
NBs were administered via tail vein injection. Once they have
accumulated in the tumor for 12 h, ES-Cu was intratumorally injected,
after which ultrasound irradiation was conducted. The therapeutic
process was subsequently monitored for 14 days (Fig. [112]6A). Compared
with the invasive growth of tumors in the control group, both the
HMME-RGD@C[3]F[8] + US and ES-Cu treatment groups showed a certain
degree of tumor suppression, which is mainly attributed to the moderate
SDT mediated by HMME-RGD@C[3]F[8] + US and the cuproptosis induced by
ES-Cu. Notably, the tumor elimination in the
HMME-RGD@C[3]F[8] + US + ES-Cu treatment group was the most obvious,
indicating an excellent antitumor effect (Fig. [113]6B–D). This effect
can be ascribed to the synergistic effect of SDT/cuproptosis induced by
HMME-RGD@C[3]F[8] NBs combined with ES-Cu under US exposure. In
addition, during the whole process, there was no obvious change in the
body weight of tumor-bearing mice in each group, indicating that the
treatment method we constructed has high biological safety
(Fig. [114]6E).
Fig. 6.
[115]Fig. 6
[116]Open in a new tab
Evaluating the antitumor effects of synergistic SDT/cuproptosis therapy
in vivo. A Treatment scheme of subcutaneous HepG2 tumor-bearing mice.
Control (PBS), HMME-RGD@C[3]F[8] + US (with 5 mg/kg of HMME), ES-Cu
(with 6 mg/kg of ES-Cu), and HMME-RGD@C[3]F[8] + US + ES-Cu (with
5 mg/kg of HMME and 6 mg/kg of ES-Cu) were injected every 2 days,
respectively (ultrasound irradiation conditions: 3 W/cm^2, 50% duty
cycle, 10 min). B Representative photos and C weights of the harvested
tumors. D Tumor volume curves of different groups of mice recorded
every 2 days and E the corresponding body weight were measured at
indicated time points. Data are presented as mean ± SD (n = 5).
Statistical analysis was performed via one-way ANOVA. *p < 0.05,
**p < 0.01, ***p < 0.001, ****P < 0.0001
Differential expression of SDCGs after various treatments
In this study, the differential expression of mRNA and lncRNA
transcripts was estimated using expression levels (Count) and the edgeR
R package (p-value < 0.05 and |log2FC| > 1). As a result, 2154
differentially expressed mRNAs were identified between Groups B and A
(Group A: HepG2 cells treated with PBS; Group B: HepG2 cells treated
with HMME-RGD@C[3]F[8] + US + ES-Cu), including 1068 upregulated and
1086 downregulated genes (Fig. [117]7A). Additionally, 5043
differentially expressed mRNAs were identified between Groups D and C
(Group C: Huh7 cells treated with PBS; Group D: Huh7 cells treated with
HMME-RGD@C[3]F[8] + US + ES-Cu), with 2146 upregulated and 2897
downregulated genes (Fig. [118]7B). Similarly, differential expression
was observed for the lncRNA transcripts: 6531 lncRNAs were
differentially expressed between Groups B and A, including 2701
upregulated and 3830 downregulated genes (Fig. [119]7C), while 9095
lncRNAs were differentially expressed between Groups D and C, including
4905 upregulated and 4190 downregulated genes (Fig. [120]7D).
Fig. 7.
[121]Fig. 7
[122]Open in a new tab
Differential expression analysis and enrichment analysis. A Group B vs.
Group A differentially expressed mRNAs; B Group D vs. Group C
differentially expressed mRNAs; C Group B vs. Group A differentially
expressed lncRNAs; D lncRNAs differentially expressed in Group D vs.
Group C; upregulated genes are represented by red dots; downregulated
genes are represented by blue dots; and grey dots are genes whose
expression has not changed significantly. E The first 20 KEGG pathways
enriched with mRNAs differentially expressed between the B and A
groups; F differences between the D and C groups enriched the bubble
chart of the first 20 KEGG pathways enriched with mRNAs; G enriched
bubble chart of the first 20 KEGG pathways enriched with lncRNA target
genes differentially expressed between the B and A groups; H The first
20 KEGG pathways enriched with lncRNA target genes differentially
expressed between the C and D groups
Pathway enrichment analysis was subsequently conducted to provide
insights into the biological functions and processes associated with
synergistic SDT/cuproptosis. The PI3K-Akt signaling pathway, the MAPK
signaling pathway, and cytokine‒cytokine receptor interactions were
significantly enriched after HMME-RGD@C[3]F[8] + US + ES-Cu treatment,
suggesting that synergistic SDT/cuproptosis therapy activated cellular
signaling and the immune response by the synergistic SDT/cuproptosis
therapy (Fig. [123]7E, [124]F). Amino acid metabolism, the estrogen
signaling pathway, and nucleotide excision repair were significantly
enriched after exposure to HMME-RGD@C[3]F[8] + US + ES-Cu for lncRNAs
(Fig. [125]7G, [126]H), suggesting the regulation of lncRNAs in
metabolic processes, hormone signaling, and DNA repair after the
indicated treatments.
Gene Ontology (GO) enrichment analysis was conducted to categorize the
functional roles of the differentially expressed genes. For mRNAs,
significant enrichment was observed in categories related to nucleosome
structure (e.g., GO:0000786 for nucleosomes), indicating that chromatin
changes are potentially relevant to the initiation and progression of
synergistic SDT/cuproptosis therapy. Additionally, terms related to DNA
replication-dependent nucleosome assembly (GO:0006335) were
highlighted, emphasizing changes in DNA packaging that could influence
gene expression during cell cycle progression (Table [127]1). For
lncRNAs, GO terms associated with microtubule structures (e.g.,
GO:0005874) and cellular localization processes (e.g., GO:1903827) were
significantly enriched, reflecting the role of lncRNAs in organizing
the cellular architecture and responding to intracellular signaling.
Furthermore, biosynthetic processes, such as purine biosynthesis
(GO:0009168) and ATP synthesis (GO:0006754), were emphasized,
indicating increased metabolic demand or adaptive metabolic responses
after the corresponding treatment (Table [128]2).
Table 1.
Representative results of GO enrichment analysis of differentially
expressed mRNAs
GO ID Description Category Padj (B vs. A) Padj (D vs. C)
GO:0000786 Nucleosome Cellular component 5.36E−24 0.000499
GO:0044815 DNA packaging complex Cellular component 1.71E−22 0.004403
GO:0032993 Protein‒DNA complex Cellular component 3.93E−16 0.013242
GO:0046982 Protein heterodimerization activity Molecular function
1.26E−11 0.000891
GO:0006335 DNA replication-dependent nucleosome assembly Biological
process 4.86E−09 0.00857
GO:0034723 DNA replication-dependent nucleosome organization Biological
process 4.86E−09 0.00857
GO:0000183 Chromatin silencing at rDNA Biological process 4.86E−09
0.047969
GO:0098761 Cellular response to interleukin-7 Biological process
1.01E−06 0.00102
GO:0098760 Response to interleukin-7 Biological process 1.01E−06
0.00102
GO:0038111 Interleukin-7-mediated signaling pathway Biological process
4.91E−06 0.001992
[129]Open in a new tab
Table 2.
Representative results of GO enrichment analysis of differentially
expressed lncRNAs
GO ID Description Category Pvalue (B vs. A) Pvalue (D vs. C)
GO:0005874 Microtubule Cellular component 0.000149 0.027041
GO:1903827 Regulation of cellular protein localization Biological
process 0.000264 0.026062
GO:0009127 Purine nucleoside monophosphate biosynthetic process
Biological process 0.0003 0.00459
GO:0009168 Purine ribonucleoside monophosphate biosynthetic process
Biological process 0.0003 0.00459
GO:0006754 ATP biosynthetic process Biological process 0.000346
0.000963
GO:0044445 Cytosolic part Cellular component 0.000347 0.000964
GO:0006359 Regulation of transcription by RNA polymerase III Biological
process 0.000543 0.01218
GO:0005759 Mitochondrial matrix Cellular component 0.000547 0.005865
GO:0009124 Nucleoside monophosphate biosynthetic process Biological
process 0.000567 0.006062
GO:0010273 Detoxification of copper ion Biological process 0.000599
0.009877
[130]Open in a new tab
The integration of differential expression data with pathway and GO
enrichment analyses revealed a complex landscape of gene regulatory
networks and cellular pathways induced by synergistic SDT/cuproptosis
therapy. The significant alterations in signaling and immune modulation
pathways, along with changes in metabolic and DNA repair processes,
suggest a multifaceted response of tumor cells to synergistic
SDT/cuproptosis therapy. These molecular insights not only deepen our
understanding of the mechanisms of synergistic SDT/cuproptosis therapy
but also highlight potential biomarkers and targets that could improve
therapeutic outcomes.
Differential immune landscape of synergistic SDT/cuproptosis therapy in LIHC
subtypes
To explore the correlation between synergistic SDT/cuproptosis therapy
and the immune microenvironment, we first utilized the expression
profiles of cuproptosis-related genes from the TCGA-LIHC cohort.
Unsupervised clustering was conducted to categorize LIHC patients into
distinct subtypes according to their responsiveness to synergistic
SDT/cuproptosis therapy. The cumulative distribution function (CDF)
curve was employed to determine the optimal K value by identifying a
notable increase in the area beneath the curve. A K value of 2 was
established, resulting in the most stable clustering with minimal
variation within the CDF curve range (Fig. [131]8A, [132]B). The
consensus heatmap confirmed the two subclusters of LIHC patients: class
1 and class 2 with different expression patterns of cuproptosis genes
(Fig. [133]8C–G).
Fig. 8.
[134]Fig. 8
[135]Open in a new tab
Correlations between the SDT/cuproptosis subgroups and the immune
environment. A Cumulative distribution function (CDF) curve. B CDF area
curve. C–F Consensus clustering matrix for k values from 2 to 5. G
Volcano plot showing the different expression patterns between class 1
and class 2. H Comparison of the expression of immune checkpoint genes
between class 1 and class 2. I–L Comparison of the estimated score (I),
immune score (J), GEP (K) and tumor purity (L) between class 1 and
class 2. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001
To uncover the fundamental biological distinctions contributing to
diverse clinical phenotypes, we next examined the immune cell
composition within the tumor microenvironment (TME) across the two
subtypes of patients. Class 1 was characterized by increased
infiltration of tumor-associated fibroblasts, mast cells, dendritic
cells (DCs), immature dendritic cells (iDCs), and neutrophils, whereas
reduced infiltration of other cell subtypes in the TME was observed in
Class 1 (Figure S4A). In contrast, class 2 patients exhibited increased
infiltration of CD8^+ T cells, CD4^+ T cells, B cells, plasma cells,
and natural killer T (NKT) cells (Figure S4A). Additionally, immune
checkpoints such as PD-1, PD-L1, PD-L2, TIM3, VTCN1, LAG3, and CTLA4
were more highly expressed in class 2 patients than in class 1
patients, suggesting that synergistic SDT/cuproptosis therapy combined
with immune checkpoint inhibitors can potentially improve survival in
LIHC patients (Fig. [136]8H). In addition, class 2 patients had higher
immune scores, estimate scores, and T-cell inflamed GEP scores than
class 1 patients, indicating a more robust antitumor immune response
and potentially greater sensitivity to immunotherapy (Fig. [137]8I–L).
These findings suggest that class 2 patients potentially benefit more
from synergistic SDT/cuproptosis therapy owing to the activation of
antitumor immunity.
Construction and validation of the risk-coring model in TCGA and GEO
We first performed univariate Cox regression on 60 overlapping genes
between SDCRGs and DEGs from the synergistic SDT/cuproptosis therapy
subgroups, identifying 35 candidate genes for further analysis
(Fig. [138]9A). Sequential LASSO regressions were then conducted on
these 35 candidate genes to pinpoint the synergistic SDT/cuproptosis
therapy-related genes that play a critical role in model construction
(Fig. [139]9B, [140]C). We identified 4 key genes (STX3, HILPDA, SMOX,
and ANXA10) for risk score formula construction according to the risk
score analysis,
[MATH:
Risksco
re=0.01719337×
STX3+0.17431249×HILPDA :MATH]
[MATH:
+0.08674283×SMOX+-0.05037749×ANXA10 :MATH]
, which can be used to categorize patients into two distinct risk
groups.
Fig. 9.
[141]Fig. 9
[142]Open in a new tab
Development of the risk-scoring model. A A Venn diagram displaying 60
overlapping DEGs of Class 1 vs. Class 2 and SDCRGs. The overlapping
genes were subsequently analyzed via univariable Cox proportional
hazards regression and the LASSO regression algorithm. B Candidate gene
profiles based on LASSO coefficients. C LASSO coefficient values of the
candidate genes. The vertical dashed lines are the optimal log(λ)
values. D–F Kaplan–Meier curves for overall survival (OS) in the
TCGA-LIHC cohort (D), [143]GSE14520 cohort (E) and [144]GSE76427 cohort
(F) with risk score classes. G–I ROC curve of the risk-scoring model in
the TCGA-LIHC (G), [145]GSE14520 (H) and [146]GSE76427 cohorts (I)
Kaplan‒Meier (KM) curve analysis revealed that patients in the low-risk
group had significantly longer overall survival (OS) in the TCGA-LIHC
cohort (log-rank test, p < 0.0001, Fig. [147]9D), which was further
validated in the [148]GSE14520 cohort (log-rank test, p < 0.0001,
Fig. [149]9E) and the [150]GSE76427 cohort (log-rank test, p = 0.043,
Fig. [151]9F). ROC analysis of OS revealed that the AUCs were 0.72,
0.67, 0.71, and 0.74 at 1, 2, 3 and 4 years, respectively, in the TCGA
cohort (Fig. [152]9G). Similarly, the [153]GSE14520 cohort demonstrated
the reliability of the risk-scoring model with AUC values of 0.64,
0.65, 0.65, and 0.67 at 1, 2, 3, and 4 years, respectively
(Fig. [154]9H). The [155]GSE76427 cohort also presented consistent AUC
values at 1, 2, 3, and 4 years, at 0.63, 0.65, 0.63, and 0.69,
respectively (Fig. [156]9I). These findings suggest that the risk score
could be a significant predictor of prognosis in LIHC patients.
Weighted co-expression network analysis in the TCGA cohort
To identify key gene modules associated with synergistic
SDT/cuproptosis therapy, we employed weighted gene co-expression
network analysis (WGCNA) to construct co-expression networks for LIHC
patients using the TCGA dataset. The expression variances of the
differentially expressed genes (DEGs) were analyzed and the top 25% of
the DEGs with the highest variances were selected for further analysis.
A soft threshold of β = 7 was chosen, achieving a scale-free topology
fit (R^2 = 0.90) to ensure the robustness of the network construction
(Fig. [157]10A). Through hierarchical clustering, we identified three
distinct co-expression modules represented by different colors. A
heatmap of the topological overlap matrix (TOM) illustrated the
interconnectedness within these modules (Fig. [158]10B, [159]C).
Further analysis of the co-expression similarity and associated
clinical features revealed a strong correlation between the turquoise
module (encompassing 421 genes) and class 1 and between the blue module
(containing 169 genes) and class 2 (Fig. [160]10D–F). These findings
highlight the potential of these modules to elucidate the molecular
mechanisms underlying the distinct clinical behaviors of LIHC subtypes
influenced by synergistic SDT/cuproptosis therapy.
Fig. 10.
[161]Fig. 10
[162]Open in a new tab
WGCNA co-expression analysis of Class 1- and Class 2- related genes. A
Scale-free fitting index (left) and average connectivity (right) for
different soft-thresholding powers β. The red line represents a
correlation coefficient of 0.9. B Hierarchical clustering dendrogram of
co-expression modules, with different colors representing different
modules. C Heatmap depicting the correlations between the 3 modules. D
Heatmap showing the correlations between Class 1- and Class 2-related
gene modules and other clinical phenotypes, with each cell containing
the corresponding correlation and p-value. E Scatter plot of the
correlation between the turquoise module and the class 1-related genes.
F Scatter plot of the correlation between the blue module and class
2-related genes
Further examination of the overlap between the 169 genes in the blue
module and the SDCRGs led to the identification of 16 overlapping genes
(TMEM51, DBN1, ARL4C, AGRN, LBH, HTRA3, TM7SF2, COL3A1, MXRA8, IER3,
GEM, COL4A2, PRNP, ITGB2, THBS1, PTGDS). These genes are hypothesized
to influence both the efficacy of synergistic SDT/cuproptosis therapy
and the tumor microenvironment in LIHC. Notably, COL3A1 and ITGB2 are
closely related to extracellular matrix remodeling and immune
modulation, suggesting their pivotal roles in cancer progression and
the response to treatments [[163]36, [164]37]. Further experimental
validation is necessary to elucidate the specific functions of these
genes in the context of cuproptosis and to explore their potential as
therapeutic targets or biomarkers in LIHC.
Discussion
Ultrasound-activated nanobubbles not only enhance ultrasound imaging
but also significantly improve tumor therapy efficacy by acting as drug
delivery vehicles [[165]38–[166]40]. These nanobubbles are more
responsive to ultrasound than traditional nanodelivery systems
[[167]41]. The inclusion of DSPE-PEG-RGD further increases the
targeting and penetration capabilities into cancer cells [[168]42].
Upon reaching the tumor tissues, these nanobubbles undergo contraction,
vibration, expansion, and ultimately rupture under varying intensities
of ultrasound, thereby facilitating ultrasound-responsive drug release.
Simultaneously, ultrasound-induced cavitation and sonoporation create
minute perforations in tumor vascular cell membranes, thus enabling
drug entry into tumor cells for enhanced therapeutic efficiency
[[169]43].
Although SDT has shown promise in inducing cancer cell apoptosis and
inhibiting tumor progression in several preclinical models, its
effectiveness is often compromised by the highly expressed intrinsic
antioxidant GSH, which mitigates the ROS-mediated antitumor efficacy of
SDT. To overcome this limitation, we developed HMME-RGD@C[3]F[8] NBs
with enhanced targeting and penetration capabilities for cancer cells,
which exhibited superior SDT efficacy compared with nonmodified
nanobubbles. Recognizing the critical role of ROS in both cuproptosis
and SDT, we focused on evaluating the ROS production triggered by the
combined therapy. The integration of SDT and cuproptosis offers a
promising approach for cancer treatment, prompting us to investigate
their synergistic effects.
Initially, we successfully synthesized HMME-RGD@C[3]F[8] NBs, owing to
their small size, which function effectively as nuclei for cavitation
during SDT. These NBs demonstrated targeted and permeable
characteristics toward cancer cells, with enhanced antitumor efficacy
compared with that of non-DSPE-PEG-RGD-modified nanoparticles. We
subsequently conducted various assays, including CCK-8, live/dead
staining, and flow cytometry analysis of apoptosis, all of which
confirmed synergistic anticancer effects, with the highest cell death
rate in the synergistic therapy group. Further analysis including
intracellular ROS detection and JC-1 staining, revealed that the
combination of SDT and cuproptosis induced the highest levels of ROS
and the most significant degree of mitochondrial damage. These findings
were supported by measurements of intracellular GSH and MDA levels,
suggesting that ES-Cu depletes intracellular GSH, amplifying oxidative
stress and enhancing the efficiency of SDT. Moreover, the synergistic
SDT/cuproptosis therapy exhibited outstanding anti-tumor capability in
vivo.
Our investigation of the differential expression of mRNAs and lncRNAs
between the synergistic SDT/cuproptosis therapy-treated and control
groups revealed complex molecular mechanisms underlying the distinct
phenotypic variations observed. Pathway enrichment analysis highlighted
the significant involvement of pathways such as PI3K-Akt signaling,
MAPK signaling, and cytokine‒cytokine receptor interactions in the mRNA
profiles, indicating that these pathways play roles in cell
proliferation, survival, and immune modulation in response to
synergistic SDT/cuproptosis therapy. The enrichment of
lncRNA-associated pathways, including amino acid metabolism and
nucleotide excision repair, reflects their potential regulatory roles
in metabolic adaptations and maintaining genomic stability during
cuproptosis.
By stratifying liver hepatocellular carcinoma (LIHC) patients based on
their cuproptosis gene expression profiles, we identified two novel
subtypes, revealing a connection between synergistic SDT/cuproptosis
therapy and changes in the tumor immune microenvironment (TME). This
stratification revealed distinct immune landscapes, with one subtype
characterized by an immune-suppressive environment conducive to tumor
progression, and the other exhibiting an immune-active profile that
could enhance antitumor immunity. These findings are crucial as they
suggest that synergistic SDT/cuproptosis therapy could modulate the
TME, potentially shifting it from a tumor-promoting to a
tumor-suppressing state.
Additionally, the development of a risk-scoring model based on
differentially expressed and cuproptosis-related genes emphasized the
prognostic significance of our findings. This model, validated across
different cohorts, suggests that molecular signatures associated with
synergistic SDT/cuproptosis therapy hold substantial prognostic
potential, suggesting that it is a promising tool for clinical
decision-making in the context of LIHC treatment. Furthermore, our use
of weighted Gene co-expression network analysis (WGCNA) identified key
gene modules associated with each LIHC subtype, providing deeper
insights into the gene networks correlated with the diverse clinical
phenotypes observed. The differential module associations offer a
molecular basis for the phenotypic distinctions, identifying potential
therapeutic targets central to the modulatory effects of synergistic
SDT/cuproptosis therapy on LIHC.
In summary, our findings not only deepen our understanding of the
biological foundations of synergistic SDT/cuproptosis therapy in liver
hepatocellular carcinoma but also pave the way for targeted therapeutic
strategies that exploit the unique molecular and immune profiles
induced by this treatment. However, our research has limitations,
including the lack of validation of therapeutic targets identified
through RNA sequencing to support these findings. Further clinical
research is necessary to validate these mechanisms, with the ultimate
goal of improving LIHC outcomes through precision medicine and
innovative treatment modalities. This study lays the groundwork for
such advancements, underscoring the transformative potential of
integrating molecular biology with clinical oncology to combat LIHC
effectively.
Conclusions
In summary, this study successfully developed HMME-RGD@C[3]F[8]
nanobubbles (NBs) with enhanced targeting and permeability for cancer
cells. In addition, the combination of HMME-RGD@C[3]F[8] NBs with ES-Cu
led to synergistic anticancer effects through cooperative cuproptosis
and SDT. ES-Cu increased intracellular Cu^2⁺ levels and depleted
overexpressed GSH, amplifying oxidative stress and further enhancing
the SDT effect. These results demonstrated that synergistic
SDT/cuproptosis therapy effectively induced ROS generation, oxidative
stress, and anticancer effects in vitro. The results of in vivo
experiments further revealed that the combined therapy centered on
sonodynamic therapy and cuproptosis had achieved outstanding anti-tumor
effects. Additionally, RNA sequencing has indicated that the
combination of SDT and cuproptosis activates antitumor immune
responses, remodels the tumor immune microenvironment, and upregulates
the expression of immune checkpoints such as PD-1, PD-L1, PD-L2, TIM3,
VTCN1, LAG3, and CTLA4. These findings suggest that combining
synergistic SDT/cuproptosis therapy with immune checkpoint inhibitors
may improve the antitumor efficacy and survival rate of cancer
patients. Moreover, the risk-scoring model constructed on the basis of
differentially expressed genes and cuproptosis-related genes has
significant potential for predicting the prognosis of patients with
liver hepatocellular carcinoma (LIHC). Finally, weighted gene
co-expression network analysis (WGCNA) identified 16 key modular genes
associated with synergistic SDT/cuproptosis sensitivity and the tumor
microenvironment in LIHC. Notably, COL3A1 and ITGB2 are closely related
to extracellular matrix remodeling and immunomodulation, respectively,
which could serve as potential new targets or biomarkers for LIHC
therapy.
Supplementary Information
[170]12951_2024_2995_MOESM1_ESM.tif^ (9MB, tif)
Additional file 1: Figure S1. Characterization of the HMME-RGD@C[3]F[8]
NBs. (A) Transmission electron microscopy image showing the
quasispherical morphology of the HMME-RGD@C[3]F[8] NBs with a mean
diameter of approximately 220 nm; scale bar = 500 nm. (B) The zeta
potential of the HMME-RGD@C[3]F[8] NBs was − 14.02 ± 0.64 mV. (C)
Standard curves of HMME at different concentrations detected by UV‒vis
spectroscopy. (D) The formula used to calculate the encapsulation
efficiency and drug loading of the HMME-RGD@C[3]F[8] NBs. (E) The
fluorescence intensity standard curve of HMME at different
concentrations detected by a fluorescence multimode microplate reader.
[171]12951_2024_2995_MOESM2_ESM.tif^ (2.9MB, tif)
Additional file 2: Figure S2. Cytotoxicity analysis of different
treatments at different concentrations. (A–H) Cytotoxicity analysis of
HepG2 and Huh7 cells under different treatments and different
concentrations of HMME-RGD@C[3]F[8] + US or ES-Cu.
[172]12951_2024_2995_MOESM3_ESM.tif^ (4.8MB, tif)
Additional file 3: Figure S3. Western blot results. Western blot
analysis of FDX1 and LIAS expression after various treatments in HepG2
cells (A) and Huh7 cells (B).
[173]12951_2024_2995_MOESM4_ESM.tif^ (27.8MB, tif)
Additional file 4: Figure S4A. Fraction of tumor-infiltrating immune
cells in the TCGA-LIHC cohort.
[174]Supplementary Material 5.^ (37.2KB, docx)
Acknowledgements