Ultrastrong Infrared Emission at 1460 nm from Lanthanide-Doped Nanoparticles with Interfacial Sensitization.

Fluorescent imaging in the second near-infrared window (NIR-II; 1000-1700 nm) has recently received tremendous attention due to its excellent tissue probing depths and high resolution. Under NIR pumping, lanthanide-doped nanoparticles can emit infrared light covering a wide range of 800-3000 nm which has good potential for NIR-II imaging and detection. However, the low efficiency hinders their application. Here, we report intense infrared emission at 1460 nm from lanthanide-doped core/shell nanoparticles with efficient interfacial sensitization. The emitter Tm3+ ion and the sensitizer Yb3+ and Nd3+ ions are spatially separated in core and shells so that the efficient interfacial energy transfer is established between Tm3+ and Yb3+/Nd3+ ions, while thermal vibration spread of high concentration of Yb3+ ions and cross-relaxation among Tm3+, Yb3+, and Nd3+ ions are suppressed. As a result, the ultrastrong NIR-II emission at 1460 nm is achieved, which is more than 100-times that in classic core/shell nanoparticles doped with Tm3+ (NaYF4:20%Yb,0.5%Tm@ NaYF4).

Identification of Biomarkers Associated with Hepatocellular Carcinoma Stem Cell Characteristics Based on Co-Expression Network Analysis of Transcriptome Data and Stemness Index.

Mounting evidence has revealed the key role of cancer stem cells in hepatocellular carcinoma (HCC) metastasis and therapy resistance, yet the genes maintaining HCC stem cell features remain to be explored. This study aimed to identify and validate the key biomarkers associated with HCC stemness. mRNA expression-based stemness index (mRNAsi) was calculating using one-class logistic regression algorithm. RNA-sequencing data and clinical information of HCC samples were downloaded from the cancer genome atlas (TCGA) and merged with the corresponding mRNAsi. We investigated the correlation between mRNAsi and HCC clinical characteristics, including tumor grades, pathologic stages, vascular invasion, and survival outcomes. Significant genes associated HCC stemness features were screened through weighted gene co-expression network analysis (WGCNA) and were functionally annotated using enrichment analysis. Protein-protein interaction network was constructed among significant genes and the key biomarkers were finally identified based on the maximal clique centrality (MCC) method. The expression of key biomarkers and its correlation with HCC clinical outcomes were validated using oncomine and gene expression omnibus (GEO) database. mR-NAsi was significantly higher in HCC tissues and gradually increased according to tumor grades and pathologic stages. Patients with vascular invasion or poor survival exhibited higher mRNAsi. Forty-four highly-correlated significant gens were screened through WGCNA and functionally related to cell cycle, cellular senescence, p53 signaling pathway, DNA replication, and mismatch repair. Four different GEO datasets confirmed that the expression levels of these 44 genes were notably higher in HCC tissues. We finally identified 15 key biomarkers (KIF4A, TTK, CCNB1, CDC20, NCAPG, CCNB2, CDC45, UBE2C, CENPA, AURKB, RRM2, CDCA8, BIRC5, TPX2, and KIF2C) through MCC method. The expression of these biomarkers was up-regulated in multiple types of cancers and showed a gradually increasing trend with HCC tumor grades. Furthermore, high expression levels of these biomarkers were also correlated with HCC metastasis, recurrence, sorafenib resistance, and poor overall survival. We identified 15 key biomarkers associated with HCC stemness features and these genes might serve as promising therapeutic targets for HCC.

Polarization-sensitive switchable display through critical coupling between graphene and a quasi-BIC.

Enhanced light-matter interaction of a local field is of prime importance in optics as it can improve the performance of nanophotonic devices. Such enhancement can be achieved by utilizing the optical bound states in the continuum (BICs). In this study, a dielectric metasurface is proposed that could enhance the light-matter interactions in graphene. A symmetry-protected BIC was observed in such a metasurface, which could transform into a quasi-BIC with a high quality (Q-) factor when the in-plane symmetry is broken. As the graphene monolayer was introduced into the system, its absorption was enhanced by the quasi-BIC resonance. By optimizing the graphene Fermi energy and the asymmetry parameter of the metasurface to satisfy the critical-coupling condition, a tunable absorber could be achieved. The absorbing intensity could be efficiently modulated by varying the polarization direction of the incident light, the maximum difference of which was up to 95.4%. Also, further investigation showed that such a feature indicates potential application in digital switches and image displays, which could be switched by incident polarization only, and therefore without dependence on an additional structural change.

Development and validation of a liquid chromatography-tandem mass spectrometry method for CPUL1, a novel antitumor candidate compound, and its application to pharmacokinetic studies.

An active substance of pyrano[3,2-a]phenazine, also called CPUL1, is a synthesized phenazine derivative and displays broad-spectrum anticancer activities. Quantitative assessment of CPUL1 in biological samples has not been well established, hindering pharmaceutical development and application. According to international guidelines, a sensitive and selective liquid chromatography-tandem mass spectrometry method in negative ion mode was developed and validated for quantification of CPUL1 in human plasma, colorectal cancer cell lines, and rat plasma, whereby linearity and accuracy were demonstrated for the range of 1-1000 ng/ml. The validated liquid chromatography-tandem mass spectrometry method was successfully employed in pharmacokinetic studies of CPUL1 in vitro and in vivo. Notably, the cellular pharmacokinetic behavior of CPUL1 varies in colorectal cancer cell lines. Regarding the pharmacokinetic processes in vivo, oral absorption was less effective than an injection, with a bioavailability of 23.66%. CPUL1 was linearly eliminated after a single administration; however, it could accumulate in tissues (heart, liver, spleen, lung, and kidney) after multiple injections. In summary, this study established a capable bioanalytical method for CPUL1 and provided exploratory pharmacokinetic data, paving the way for use of this promising derivative in disease models.

Prognostic value of myosteatosis in patients with lung cancer: a systematic review and meta-analysis.

The prognostic value of myosteatosis has been widely investigated in lung cancer, yet conclusions remain controversial. The purpose of this meta-analysis was to illuminate this issue. Medline, Embase, Cochrane Library and Web of Science Core Collection online databases were systematically searched from inception to 24 September 2021. Newcastle-Ottawa Scale tool was applied to evaluate the quality of included studies. Pooled hazard ratios (HRs) with 95% confidence intervals (CIs) for overall survival (OS) and progression-free survival (PFS) were used to examine prognostic value of myosteatosis. Subgroup analysis and sensitivity analysis were conducted to assess heterogeneity and stability of results. A total of 484 articles were screened from which 9 eligible studies involving 1667 patients were enrolled in this meta-analysis. Lung cancer patients with myosteatosis had significantly worse OS than patients without myosteatosis (HR 1.10, 95% CI 1.05-1.16, P < 0.001), both in six multivariate analysis (HR 1.46, 95% CI 1.16-1.85, P = 0.001) and in three univariate analysis (HR 1.08, 95% CI 1.03-1.14, P = 0.003). Pooled data from five studies using multivariate survival analysis also showed that patients with myosteatosis had a statistically significant unfavorable PFS (HR = 1.27, 95% CI 1.00-1.62, P = 0.049). Sensitivity analysis showed the result for OS was stable. But for PFS, the result was not robust. Myosteatosis might serve as an independent indicator of unfavorable survival outcomes for OS and PFS in lung cancer patients. Further studies are needed to confirm our results.

Nanoscale Ultrasensitive Temperature Sensing Based on Upconversion Nanoparticles with Lattice Self-Adaptation.

Upconversion nanoparticles have recently received increasing attention due to their outstanding performance in temperature sensing at the nanoscale. Although much effort has been devoted to improve their thermal sensitivity, there is no efficient way for achieving significant enhancement. Here, we show that lattice self-adaptation can unlock a new route for remarkably enhancing the thermal sensitivity of upconversion nanoparticles. The thermally sensitive fluorescence intensity ratio (FIR) of the dopant Er3+ is used for indicating the temperature variation, while a heterojunction of NaGdF4/NaYF4 is prepared as host material to produce a lattice distortion at the interface which is also sensitive to temperature. With the increase of temperature, the FIR of the transitions 2H11/2/4S3/2 → 4I15/2 increases, accompanied by the self-adapted decrease of interface lattice distortion that leads to the additional increase in FIR. Using core/shell upconversion nanoparticles with lattice self-adaptation, we achieve an enhanced thermal sensitivity three times higher than core-only nanoparticles.

Manipulating Energy Transfer in UCNPs@SiO2@Ag Nanoparticles for Efficient Infrared Photocatalysis.

Conventional photocatalysts must be activated by ultraviolet or visible light to meet the energy requirement of populating an initial excited state, while infrared light has a high penetration depth to reaction media but does not have enough photon energy to activate conventional photocatalysts. Here, we report the activation of Ag nanoparticles by upconversion nanoparticles (UCNPs) in UCNPs@SiO2@Ag with manipulated energy transfer for infrared photocatalysis. UCNPs can efficiently convert infrared light to visible and ultraviolet light and are very ideal candidates for bridging the advantage of infrared light and the activation energy requirement of conventional photocatalysts. In the UCNPs@SiO2@Ag nanosystem, we employ the UCNPs to activate conventional Ag nanoparticles under infrared light irradiation. The evanescent field of UCNPs is confined for enhancing the near-field energy-transfer efficiency using a designed core/shell heterostructure, while a SiO2 layer is used for blocking the phonon exchange of thermal vibration between photon upconverters and Ag nanoparticles. Based on the manipulated energy transfer, UCNPs@SiO2@Ag nanoparticles exhibit efficient photocatalytic activity under the irradiation of 980 nm infrared light, while single Ag nanoparticles have negligible catalytic activity under infrared irradiation.

Pre-magnetization for enhancing the iron-catalyzed activation of peroxymonosulfate via accelerating the corrosion of Fe0.

Our findings proved that micron-scale zero-valent iron (mZVI) particles with pre-magnetization combined with peroxymonosulfate (PMS) can markedly enhance the removal of acid orange 7 (AO7). Investigation into the mechanism showed that PMS accelerated the corrosion of ZVI to release Fe2+ under acidic conditions, and the in-situ generated Fe2+ further activated PMS to produce SO4•- and •OH, resulting in AO7 removal. Further, the Lorentz force strengthened the convection in the solution and the field gradient force tended to move Fe2+ from a higher to a lower field gradient at the pre-magnetized ZVI (Pre-ZVI) particle surfaces, thus indicating that pre-magnetization promoted the corrosion of ZVI to release Fe2+, which resulted in the enhancement of PMS activation. Nano-scale ZVI (nZVI) was more effective than mZVI in activating PMS to degrade AO7, but the pre-magnetization effect on mZVI was better than on nZVI. AO7 removal increased with higher ZVI and PMS dosage, lower AO7 concentration, and acidic conditions (pH = 2, 3). This study helps to understand the reactive radicals-based oxidation process with application of pre-magnetized ZVI in activating PMS.

Magnetic tuning of upconversion luminescence in Au/NaGdF4:Yb3+/Er3+ nanocomposite.

Lanthanide-doped upconversion nanoparticles (UCNPs) NaGdF4:Yb3+/Er3+ have received increasing attention due to their unique optical-magnetic bifunctional properties. Here, we show that the luminescent intensity from NaGdF4:Yb3+/Er3+ nanoparticles decreases monotonously with increasing the applied magnetic field from 0 to 37.1 T, while plasmon-enhanced upconversion luminescence in Au/NaGdF4:Yb3+/Er3+ nanocomposite is independent of a magnetic field lower than 6 T. The surface plasmon resonances could compensate for the energetic mismatching between the excitation light and the energy-level gaps induced by magnetic field and enhance the radiative efficiency, which is the main factor for achieving this stable upconversion emission in this nanocomposite under a magnetic field not higher than 6 T. These findings provide a novel route for exploring the magnetic control of upconversion luminescence in lanthanide-doped bifunctional nanoparticles.

Enhanced upconversion luminescence through core/shell structures and its application for detecting organic dyes in opaque fishes.

Here, we report the enhanced upconversion luminescence of NaLuF4:18%Yb(3+),2%Er(3+) through core/shell structures. Among NaYF4, NaGdF4, and NaLuF4 shells, the first one presents the highest efficiency. These upconversion fluorescent nanoprobes with an oleic acid/PEG hybrid ligand can efficiently capture Rhodamine B (RB) and sodium fluorescein (SF) in opaque fishes to present their residues in vivo through luminescence resonant energy transfer (LRET) processes. It can be confirmed based on LRET technology that no RB is absorbed by opaque fishes after incubating in the aqueous solution of 1 µg ml(-1) RB for one day, while SF residue can be obviously detected after incubating in the aqueous solution of 1 µg ml(-1) SF for one day. The merit of this LRET technology with the upconversion nanoparticle (UCNP) donor is ascribed to the deep penetration depth of the infrared pumping laser and high signal to noise ratio.

Energy upconversion in lanthanide-doped core/porous-shell nanoparticles.

Here, we report upconversion nanoparticles with a core/porous-shell structure in which bulk emission and nanoemission are simultaneously observed. The activated porous shell can efficiently tune the bulk emission but has negligible influence on the nanoemission.

Self-Supervised Medical Image Segmentation Using Deep Reinforced Adaptive Masking.

Self-supervised learning aims to learn transferable representations from unlabeled data for downstream tasks. Inspired by masked language modeling in natural language processing, masked image modeling (MIM) has achieved certain success in the field of computer vision, but its effectiveness in medical images remains unsatisfactory. This is mainly due to the high redundancy and small discriminative regions in medical images compared to natural images. Therefore, this paper proposes an adaptive hard masking (AHM) approach based on deep reinforcement learning to expand the application of MIM in medical images. Unlike predefined random masks, AHM uses an asynchronous advantage actor-critic (A3C) model to predict reconstruction loss for each patch, enabling the model to learn where masking is valuable. By optimizing the non-differentiable sampling process using reinforcement learning, AHM enhances the understanding of key regions, thereby improving downstream task performance. Experimental results on two medical image datasets demonstrate that AHM outperforms state-of-the-art methods. Additional experiments under various settings validate the effectiveness of AHM in constructing masked images.

Automatic data augmentation for medical image segmentation using Adaptive Sequence-length based Deep Reinforcement Learning.

Although existing deep reinforcement learning-based approaches have achieved some success in image augmentation tasks, their effectiveness and adequacy for data augmentation in intelligent medical image analysis are still unsatisfactory. Therefore, we propose a novel Adaptive Sequence-length based Deep Reinforcement Learning (ASDRL) model for Automatic Data Augmentation (AutoAug) in intelligent medical image analysis. The improvements of ASDRL-AutoAug are two-fold: (i) To remedy the problem of some augmented images being invalid, we construct a more accurate reward function based on different variations of the augmentation trajectories. This reward function assesses the validity of each augmentation transformation more accurately by introducing different information about the validity of the augmented images. (ii) Then, to alleviate the problem of insufficient augmentation, we further propose a more intelligent automatic stopping mechanism (ASM). ASM feeds a stop signal to the agent automatically by judging the adequacy of image augmentation. This ensures that each transformation before stopping the augmentation can smoothly improve the model performance. Extensive experimental results on three medical image segmentation datasets show that (i) ASDRL-AutoAug greatly outperforms the state-of-the-art data augmentation methods in medical image segmentation tasks, (ii) the proposed improvements are both effective and essential for ASDRL-AutoAug to achieve superior performance, and the new reward evaluates the transformations more accurately than existing reward functions, and (iii) we also demonstrate that ASDRL-AutoAug is adaptive for different images in terms of sequence length, as well as generalizable across different segmentation models.

Long-term high density lipoprotein infusion ameliorates metabolic phenotypes of diabetic db/db mice.

BACKGROUND: Lower quantity and quality of high density lipoprotein (HDL) are important characteristics of type 2 diabetes mellitus. Acute HDL infusion results in a greater fall of plasma glucose in diabetes patients. Here, we aim to investigate the influence of long-term HDL infusion on metabolic phenotypes of diabetic db/db mice. METHODS: High density lipoprotein was introduced to db/db mice twice a week for 4 weeks. The phenotypes of the mice were monitored by analyzing metabolic parameters. Glycogen analysis was performed with amyloglucosidase. The corresponding signaling molecules were detected by western blot. RESULTS: Long-term introduction of HDL decreased plasma glucose levels of db/db mice. Glycogen deposition was enhanced in gastrocnemius muscle, paralleling the elevated glycogen synthase kinase-3 phosphorylation. Meanwhile, increased Akt-Ser473 and adenosine monophosphate-activated protein kinase phosphorylations were detected in the muscle. Moreover, HDL reduced blood glucose and free fatty acids and improved pancreatic islet structure and function with increased C-peptide. Furthermore, decreased interleukin-6, C-reactive protein, monocyte chemoattractant protein-1, resistin, and malondialdehyde, as well as enhanced leptin levels were detected in HDL-treated mice. CONCLUSION: Results of the present study suggest that long-term HDL infusion has positive therapeutic effects on the metabolic disturbances of db/db diabetic mice.

CuO quantum-dot-sensitized mesoporous ZnO for visible-light photocatalysis.

How to extend ultraviolet photocatalysts to the visible-light region is a key challenge for solar-driven photocatalysis. Herein, we show that ultraviolet ZnO photocatalysts can present high visible-light photocatalytic activity when combined with CuO quantum dots (QDs; <3 nm). Theoretical analysis demonstrates that the quantum size effect plays a key role in the photoactivity of the CuO/ZnO composite. For CuO QDs smaller than 3 nm, the separated charges could transfer from CuO QDs to the conduction bands of ZnO due to quantum splitting of the CuO energy level and phonon compensation for the difference in the conduction band minimum of CuO and ZnO; however, this process would not occur with the disappearance of the quantum size effect. Further structural analysis demonstrates that interfacial charge separation and transfer between ZnO and CuO dominate the photocatalytic processes instead of a single CuO or ZnO surface. Compared with ZnO-noble metal structures (e.g., ZnO-Ag or ZnO-Au), these ZnO-CuO QD composites present wider absorption bands, higher visible photocatalytic efficiencies, and lower costs.

Painless and accurate medical image analysis using deep reinforcement learning with task-oriented homogenized automatic pre-processing.

Pre-processing is widely applied in medical image analysis to remove the interference information. However, the existing pre-processing solutions mainly encounter two problems: (i) it is heavily relied on the assistance of clinical experts, making it hard for intelligent CAD systems to deploy quickly; (ii) due to the personnel and information barriers, it is difficult for medical institutions to conduct the same pre-processing operations, making a deep model that performs well on a specific medical institution difficult to achieve similar performances on the same task in other medical institutions. To overcome these problems, we propose a deep-reinforcement-learning-based task-oriented homogenized automatic pre-processing (DRL-HAPre) framework to overcome these two problems. This framework utilizes deep reinforcement learning techniques to learn a policy network to automatically and adaptively select the optimal pre-processing operations for the input medical images according to different analysis tasks, thus helping the intelligent CAD system to achieve a rapid deployment (i.e., painless) and maintain a satisfactory performance (i.e., accurate) among different medical institutes. To verify the effectiveness and advantages of the proposed DRL-HAPre framework, we further develop a homogenized automatic pre-processing model based on the DRL-HAPre framework to realize the automatic pre-processing of key region selection (called HAPre-KRS) in the pneumonia image classification task. Extensive experimental studies are conducted on three pediatric pneumonia classification datasets with different image qualities, and the results show that: (i) There does exist a hard-to-reproduce problem in clinical practices and the fact that having different medical image qualities in different medical institutes is an important reason for the existing of hard-to-reproduce problem, so it is compelling to propose homogenized automatic pre-processing method. (ii) The proposed HAPre-KRS model and DRL-HAPre framework greatly outperform three kinds of state-of-the-art baselines (i.e., pre-processing, attention and pneumonia baseline), and the lower the medical image quality, the greater the improvements of using our HAPre-KRS model and DRL-HAPre framework. (iii) With the help of homogenized pre-processing, HAPre-KRS (and DRL-HAPre framework) can greatly avoid performance degradation in real-world cross-source applications (i.e., thus overcoming the hard-to-reproduce problem).

PAC-Net: Multi-pathway FPN with position attention guided connections and vertex distance IoU for 3D medical image detection.

Automatic medical image detection aims to utilize artificial intelligence techniques to detect lesions in medical images accurately and efficiently, which is one of the most important tasks in computer-aided diagnosis (CAD) systems, and can be embedded into portable imaging devices for intelligent Point of Care (PoC) Diagnostics. The Feature Pyramid Networks (FPN) based models are widely used deep-learning-based solutions for automatic medical image detection. However, FPN-based medical lesion detection models have two shortcomings: the object position offset problem and the degradation problem of IoU-based loss. Therefore, in this work, we propose a novel FPN-based backbone model, i.e., Multi-Pathway Feature Pyramid Networks with Position Attention Guided Connections and Vertex Distance IoU (abbreviated as PAC-Net), to replace vanilla FPN for more accurate lesion detection, where two innovative improvements, a position attention guided connection (PAC) module and Vertex Distance IoU Vertex Distance Intersection over Union loss, are proposed to address the above-mentioned shortcomings of vanilla FPN, respectively. Extensive experiments are conducted on a public medical image detection dataset, i.e., Deeplesion, and the results showed that i) PAC-Net outperforms all state-of-the-art FPN-based depth models in both evaluation metrics of lesion detection on the DeepLesion dataset, ii) the proposed PAC module and VDIoU loss are both effective and important for PAC-Net to achieve a superior performance in automatic medical image detection tasks, and iii) the proposed VDIoU loss converges more quickly than the existing IoU-based losses, making PAC-Net an accurate and also highly efficient 3D medical image detection model.

EFPN: Effective medical image detection using feature pyramid fusion enhancement.

Feature pyramid networks (FPNs) are widely used in the existing deep detection models to help them utilize multi-scale features. However, there exist two multi-scale feature fusion problems for the FPN-based deep detection models in medical image detection tasks: insufficient multi-scale feature fusion and the same importance for multi-scale features. Therefore, in this work, we propose a new enhanced backbone model, EFPNs, to overcome these problems and help the existing FPN-based detection models to achieve much better medical image detection performances. We first introduce an additional top-down pyramid to help the detection networks fuse deeper multi-scale information; then, a scale enhancement module is developed to use different sizes of kernels to generate more diverse multi-scale features. Finally, we propose a feature fusion attention module to estimate and assign different importance weights to features with different depths and scales. Extensive experiments are conducted on two public lesion detection datasets for different medical image modalities (X-ray and MRI). On the mAP and mR evaluation metrics, EFPN-based Faster R-CNNs improved 1.55% and 4.3% on the PenD (X-ray) dataset, and 2.74% and 3.1% on the BraTs (MRI) dataset, respectively. EFPN-based Faster R-CNNs achieve much better performances than the state-of-the-art baselines in medical image detection tasks. The proposed three improvements are all essential and effective for EFPNs to achieve superior performances; and besides Faster R-CNNs, EFPNs can be easily applied to other deep models to significantly enhance their performances in medical image detection tasks.

Berberine induces SOCS1 pathway to reprogram the M1 polarization of macrophages via miR-155-5p in colitis-associated colorectal cancer.

Berberine is approved for the treatment of intestinal infections and diarrhea and has been shown to have anti-inflammatory and anti-tumor effects in pathological intestinal tissues. However, it is unclear whether the anti-inflammatory effect of berberine contributes to its anti-tumor effect on colitis-associated colorectal cancer (CAC). In this study, we found that berberine effectively inhibited tumorigenesis and protected against colon shortening in CAC mouse model. Immunohistochemistry results showed a reduction in the number of macrophage infiltrations in the colon following berberine treatment. Further analysis revealed that most of the infiltrated macrophages were pro-inflammatory M1 type, which berberine effectively limited. However, in another CRC model without chronic colitis, berberine had no significant effect on tumor number or colon length. In vitro studies demonstrated that berberine treatment significantly reduced the percentage of M1 type and levels of Interleukin-1ß (IL-1ß), Interleukin-6 (IL-6) and tumor necrosis factor-α (TNF-α). Additionally, miR-155-5p level was down-regulated, and suppressor of cytokine signaling 1 (SOCS1) expression was up-regulated in berberine-treated cells. Notably, the miR-155-5p inhibitor attenuated the regulatory effects of berberine on SOCS1 signaling and macrophage polarization. Altogether, our findings suggest that the inhibitory effect of berberine on CAC development is dependent on its anti-inflammatory activity. Moreover, miR-155-5p may be involved in the pathogenesis of CAC by regulating M1 macrophage polarization, and berberine could be a promising protective agent against miR-155-5p-mediated CAC. This study provides new insights into pharmacologic mechanisms of berberine and supports the possibility that other anti-miR-155-5p drugs may be beneficial in the treatment of CAC.

Bufotalin induces ferroptosis in non-small cell lung cancer cells by facilitating the ubiquitination and degradation of GPX4.

Ferroptosis is a new form of regulated cell death that is dependent on iron- and lipid reactive oxygen species. Emerging evidence indicate that induction of ferroptosis could inhibit the proliferation of diverse cancer cells, which functions as a potent tumor suppressor in cancer. Here, we firstly reported Bufotalin (BT), a natural small molecule, was a novel glutathione peroxidase 4 (GPX4) inhibitor, which could trigger the ferroptosis in non-small cell lung cancer cells. In vitro, BT significantly inhibited the proliferation of A549 cells and induced the ferroptosis, whereas ferroptosis inhibitor or iron chelator significantly reversed the cytotoxicity of BT on A549 cells. Moreover, BT also increased the intracellular Fe2+. Subsequently, immunoblotting showed that BT could inhibit the protein expression of GPX4. Notably, BT dramatically accelerated the degradation of GPX4 in A549 cells. Immunoprecipitation assay further certified the increased ubiquitination of GPX4 induced by BT. Nevertheless, BT could not further increase the lipid ROS after silencing of GPX4, suggesting the induction of ferroptosis by BT was dependent on GPX4. Furthermore, BT also observably inhibited tumor growth and promoted lipid peroxidation in vivo. In conclusion, our findings indicated that BT could induce ferroptosis and cause lipid peroxidation by accelerating the degradation of GPX4 and raising the intracellular Fe2+, and BT will hopefully serve as a lead compound in developing anti-tumor agents for targeting ferroptosis.