Childhood environmental unpredictability and hoarding: Mediating roles of attachment and security.

Life history theory provides a unified perspective for understanding human behaviors as adaptive strategies to specific environmental conditions. Within this theoretical framework, hoarding emerges as a behavior reflecting an evolved strategy in response to unpredictable environmental challenges, serving as a buffer against resource scarcity and enhancing survival prospects. This study aimed to explore the key roles of childhood environmental unpredictability, attachment, and sense of security in the development of hoarding. 662 participants completed scales on childhood environmental unpredictability, Revised Experiences in Close Relationships (ECR-R), sense of insecurity, and Savings Inventory-Revised (SI-R). The results showed that childhood environmental unpredictability was significantly positively correlated with hoarding. Attachment anxiety and sense of security individually mediate the effect of childhood environmental unpredictability on hoarding. Additionally, 'attachment anxiety--sense of security' and 'attachment avoidance--sense of security' serve as chain mediators in this relationship separately. This study offers insights into the cognitive-behavioral model of hoarding, highlighting the importance of life history theory in examining childhood environmental unpredictability's relationship with hoarding. It also integrates insights from the psychosocial acceleration theory into our comprehension of hoarding's development. Future research directions are also discussed.

Suicidal ideation in Chinese patients with advanced breast cancer: a multi-center mediation model study.

PURPOSE: The pathways underpinning suicide ideation (SI) and certain physical and psychological factors in patients with advanced breast cancer remain unclear. This study develops and validates a mediation model that delineates the associations between several multidimensional variables and SI in Chinese patients with advanced breast cancer. METHODS: Patients with advanced breast cancer (n = 509) were recruited as study participants from 10 regional cancer centers across China from August 2019 to December 2020. Participants were required to complete five questionnaires using an electronic patient-reported outcomes (ePRO) system: 9 item- Patient Health Questionnaire (PHQ-9), Hospital Anxiety and Depression Scale (HADS), Insomnia Severity Index (ISI), 5-level EQ-5D (EQ-5D-5L), and MD Anderson Symptom Inventory (MDASI). Risk factors for SI were identified using multivariable logistic regression, and inputted into serial multiple mediation models to elucidate the pathways linking the risk factors to SI. RESULTS: SI prevalence was 22.8% (116/509). After adjusting for covariates, depression (odds ratio [OR] = 1.384), emotional distress (OR = 1.107), upset (OR = 0.842), and forgetfulness (OR = 1.236) were identified as significant independent risk factors (all p < 0.05). The ORs indicate that depression and distress have the strongest associations with SI. Health status has a significant indirect effect (OR=-0.044, p = 0.005) and a strong total effect (OR=-0.485, p < 0.001) on SI, mediated by insomnia severity and emotional distress. CONCLUSIONS: There is a high SI prevalence among Chinese patients with advanced breast cancer. Our analysis revealed predictive pathways from poor health to heightened SI, mediated by emotional distress and insomnia. Regular management of distress and insomnia can decrease suicide risk in this vulnerable population.

Physical symptoms and anxiety and depression in older patients with advanced cancer in China: a network analysis.

BACKGROUND: Little is understood about the association between psychosomatic symptoms and advanced cancer among older Chinese patients. METHODS: This secondary analysis was part of a multicenter cross-sectional study based on an electronic patient-reported outcome platform. Patients with advanced cancer were included between August 2019 and December 2020 in China. Participants (over 60 years) completed the MD Anderson Symptom Inventory (MDASI) and Hospital Anxiety and Depression Scale (HADS) to measure symptom burden. Network analysis was also conducted to investigate the network structure, centrality indices (strength, closeness, and betweenness) and network stability. RESULTS: A total of 1022 patients with a mean age of 66 (60-88) years were included; 727 (71.1%) were males, and 295 (28.9%) were females. A total of 64.9% of older patients with advanced cancer had one or more symptoms, and up to 80% had anxiety and depression. The generated network indicated that the physical symptoms, anxiety and depression symptom communities were well connected with each other. Based on an evaluation of the centrality indices, 'distress/feeling upset' (MDASI 5) appears to be a structurally important node in all three networks, and 'I lost interest in my own appearance' (HADS-D4) had the lowest centrality indices. The network stability was relatively high (> 0.7). CONCLUSION: The symptom burden remains high in older patients with advanced cancer in China. Psychosomatic symptoms are highly interactive and often present as comorbidities. This network can be used to provide targeted interventions to optimize symptom management in older patients with advanced cancer in China. TRIAL REGISTRATION: Chinese Clinical Trial Registry (ChiCTR1900024957), registered on 06/12/2020.

Deciphering the mechanism of Peptostreptococcus anaerobius-induced chemoresistance in colorectal cancer: the important roles of MDSC recruitment and EMT activation.

Peptostreptococcus anaerobius (P. anaerobius, PA) in intestinal flora of patients with colorectal cancer (CRC) are associated with poor prognosis. Studies have shown that P. anaerobius could promote colorectal carcinogenesis and progression, but whether P. anaerobius could induce chemoresistance of colorectal cancer has not been clarified. Here, both in vitro and in vivo experiments showed that P. anaerobius specifically colonized the CRC lesion and enhanced chemoresistance of colorectal cancer to oxaliplatin by recruiting myeloid-derived suppressor cells (MDSCs) into the tumor microenvironment. Furthermore, this study revealed that it was the increased secretion of IL-23 by MDSCs that subsequently facilitated the epithelial-mesenchymal transition (EMT) of tumor cells to induce chemoresistance of CRC by activating the Stat3-EMT pathway. Our results highlight that targeting P. anaerobius might be a novel therapeutic strategy to overcome chemoresistance in the treatment of CRC.

Co-embedding of edges and nodes with deep graph convolutional neural networks.

Graph neural networks (GNNs) have significant advantages in dealing with non-Euclidean data and have been widely used in various fields. However, most of the existing GNN models face two main challenges: (1) Most GNN models built upon the message-passing framework exhibit a shallow structure, which hampers their ability to efficiently transmit information between distant nodes. To address this, we aim to propose a novel message-passing framework, enabling the construction of GNN models with deep architectures akin to convolutional neural networks (CNNs), potentially comprising dozens or even hundreds of layers. (2) Existing models often approach the learning of edge and node features as separate tasks. To overcome this limitation, we aspire to develop a deep graph convolutional neural network learning framework capable of simultaneously acquiring edge embeddings and node embeddings. By utilizing the learned multi-dimensional edge feature matrix, we construct multi-channel filters to more effectively capture accurate node features. To address these challenges, we propose the Co-embedding of Edges and Nodes with Deep Graph Convolutional Neural Networks (CEN-DGCNN). In our approach, we propose a novel message-passing framework that can fully integrate and utilize both node features and multi-dimensional edge features. Based on this framework, we develop a deep graph convolutional neural network model that prevents over-smoothing and obtains node non-local structural features and refined high-order node features by extracting long-distance dependencies between nodes and utilizing multi-dimensional edge features. Moreover, we propose a novel graph convolutional layer that can learn node embeddings and multi-dimensional edge embeddings simultaneously. The layer updates multi-dimensional edge embeddings across layers based on node features and an attention mechanism, which enables efficient utilization and fusion of both node and edge features. Additionally, we propose a multi-dimensional edge feature encoding method based on directed edges, and use the resulting multi-dimensional edge feature matrix to construct a multi-channel filter to filter the node information. Lastly, extensive experiments show that CEN-DGCNN outperforms a large number of graph neural network baseline methods, demonstrating the effectiveness of our proposed method.

MLSFF: Multi-level structural features fusion for multi-modal knowledge graph completion.

With the rise of multi-modal methods, multi-modal knowledge graphs have become a better choice for storing human knowledge. However, knowledge graphs often suffer from the problem of incompleteness due to the infinite and constantly updating nature of knowledge, and thus the task of knowledge graph completion has been proposed. Existing multi-modal knowledge graph completion methods mostly rely on either embedding-based representations or graph neural networks, and there is still room for improvement in terms of interpretability and the ability to handle multi-hop tasks. Therefore, we propose a new method for multi-modal knowledge graph completion. Our method aims to learn multi-level graph structural features to fully explore hidden relationships within the knowledge graph and to improve reasoning accuracy. Specifically, we first use a Transformer architecture to separately learn about data representations for both the image and text modalities. Then, with the help of multimodal gating units, we filter out irrelevant information and perform feature fusion to obtain a unified encoding of knowledge representations. Furthermore, we extract multi-level path features using a width-adjustable sliding window and learn about structural feature information in the knowledge graph using graph convolutional operations. Finally, we use a scoring function to evaluate the probability of the truthfulness of encoded triplets and to complete the prediction task. To demonstrate the effectiveness of the model, we conduct experiments on two publicly available datasets, FB15K-237-IMG and WN18-IMG, and achieve improvements of 1.8 and 0.7%, respectively, in the Hits@1 metric.

MLNet: a multi-level multimodal named entity recognition architecture.

In the field of human-computer interaction, accurate identification of talking objects can help robots to accomplish subsequent tasks such as decision-making or recommendation; therefore, object determination is of great interest as a pre-requisite task. Whether it is named entity recognition (NER) in natural language processing (NLP) work or object detection (OD) task in the computer vision (CV) field, the essence is to achieve object recognition. Currently, multimodal approaches are widely used in basic image recognition and natural language processing tasks. This multimodal architecture can perform entity recognition tasks more accurately, but when faced with short texts and images containing more noise, we find that there is still room for optimization in the image-text-based multimodal named entity recognition (MNER) architecture. In this study, we propose a new multi-level multimodal named entity recognition architecture, which is a network capable of extracting useful visual information for boosting semantic understanding and subsequently improving entity identification efficacy. Specifically, we first performed image and text encoding separately and then built a symmetric neural network architecture based on Transformer for multimodal feature fusion. We utilized a gating mechanism to filter visual information that is significantly related to the textual content, in order to enhance text understanding and achieve semantic disambiguation. Furthermore, we incorporated character-level vector encoding to reduce text noise. Finally, we employed Conditional Random Fields for label classification task. Experiments on the Twitter dataset show that our model works to increase the accuracy of the MNER task.

Multifunctional Role of Ag-Substitution in Enhancing the Photoelectrochemical Properties of LaFeO3 Photocathodes.

Earth-abundant LaFeO3 is a promising p-type semiconductor for photoelectrochemical cells due to its stable photoresponses, high photovoltages and appropriate band alignments, but the photoelectrochemical properties of LaFeO3 , especially the incident-photon-to-current conversion efficiency, need to be further improved. Herein, we propose to partially substitute La3+ of LaFeO3 with Ag+ to enhance the photoelectrochemical performance of LaFeO3 . The combined experimental and computational studies show that Ag-substitution improves surface charge transfer kinetics through introducing active electronic states and increasing electrochemically active surface areas. Furthermore, Ag-substitution decreases grain boundary number and increases majority carrier density, which promotes bulk charge transports. Ag-substitution also reduces the bandgap energy, increasing the flux of carriers involved in photoelectrochemical reactions. As a result, after 8 % Ag-substitution, the photocurrent density of LaFeO3 is enhanced by more than 6 times (-0.64 mA cm-2 at 0.5 V vs RHE) in the presence of oxygen, which is the highest photocurrent gain compared with other cation substitution or doping. The corresponding photocurrent onset potential also demonstrates a positive shift of 30 mV. This work highlights the versatile effects of Ag-substitution on the photoelectrochemical properties of LaFeO3 , which can provide useful insights into the mechanism of enhanced photoelectrochemical performance by doping or substitution.

Bioimaging and detecting endogenous and exogenous cyanide in foods, living cells and mice based on a turn-on mitochondria-targeted fluorescent probe.

A novel fluorescent probe, with advanced features including "turn-on" fluorescence response, high sensitivity, good compatibility, and mitochondria-targeting function, has been synthesized based on structural design for detecting and visualizing cyanide in foods and biological systems. An electron-donating triphenylamine group (TPA) was employed as the fluorescent and an electron-accepting 4-methyl-N-methyl-pyridinium iodide (Py) moiety was used as a mitochondria-targeted localization unit, which formed intramolecular charge transfer (ICT) system. The "turn-on" fluorescence response of the probe (TPA-BTD-Py, TBP) toward cyanide is attributed two reasons, one is the insertion of an electron-deficient benzothiadiazole (BTD) group into the conjugated system between TPA and Py, and the other is the inhibition of ICT induced by the nucleophilic addition of CN-. Two active sites for reacting with CN- were involved in TBP molecule and high response sensitivity were observed in tetrahydrofuran solvent containing 3 % H2O. The response time could be reduced to 150 s, the linear range was 0.25-50 µM, and the limit of detection was 0.046 µM for CN- analysis. The TBP probe was successfully applied to the detection of cyanide in food samples prepared in aqueous solution, including the sprouting potato, bitter almond, cassava, and apple seeds. Furthermore, TBP exhibited low cytotoxicity, clear mitochondria-localizing capability in HeLa cells and excellent fluorescence imaging of exogenous and endogenous CN- in living PC12 cells. Moreover, exogenous CN- with intraperitoneal injection in nude mice could be well monitored visually by the "turn-on" fluorescence. Therefore, the strategy based on structural design provided good prospects for optimizing fluorescent probes.

A dynamic graph Hawkes process based on linear complexity self-attention for dynamic recommender systems.

The dynamic recommender system realizes the real-time recommendation for users by learning the dynamic interest characteristics, which is especially suitable for the scenarios of rapid transfer of user interests, such as e-commerce and social media. The dynamic recommendation model mainly depends on the user-item history interaction sequence with timestamp, which contains historical records that reflect changes in the true interests of users and the popularity of items. Previous methods usually model interaction sequences to learn the dynamic embedding of users and items. However, these methods can not directly capture the excitation effects of different historical information on the evolution process of both sides of the interaction, i.e., the ability of events to influence the occurrence of another event. In this work, we propose a Dynamic Graph Hawkes Process based on Linear complexity Self-Attention (DGHP-LISA) for dynamic recommender systems, which is a new framework for modeling the dynamic relationship between users and items at the same time. Specifically, DGHP-LISA is built on dynamic graph and uses Hawkes process to capture the excitation effects between events. In addition, we propose a new self-attention with linear complexity to model the time correlation of different historical events and the dynamic correlation between different update mechanisms, which drives more accurate modeling of the evolution process of both sides of the interaction. Extensive experiments on three real-world datasets show that our model achieves consistent improvements over state-of-the-art baselines.

Validation of the Chinese Cultural Tightness-Looseness Scale and General Tightness-Looseness Scale.

Introduction: This study aims to revise the Cultural Tightness-Looseness Scale (CTLS) and General Tightness-Looseness Scale (GTLS), and explore the group heterogeneity of tightness-looseness perception in Chinese populations. Methods: Sample 1 (n = 2,388) was used for item analysis and exploratory factor analysis, and sample 2 (n = 2,385) was used for confirmatory factor analysis and latent profile analysis. Sample 3 (n = 512) was used for the reliability test and criterion validity test, among which 162 participants were used for the test-retest reliability examination after a four-week interval. Measurements included the CTLS, GTLS, International Personality Item Pool, Personal Need for Structure Scale, and Campbell Index of Well-Being. Results: The revised CTLS contained four items and retained a single-dimensional structure. The revised GTLS consisted of eight items divided into two dimensions: Compliance with Norms and Social Sanctions. Latent profile analysis extracted two profiles on both CTLS and GTLS scores, indicating that the sample can be divided into two subgroups: high and low perception of tightness. Discussion: The Chinese versions of the CTLS and GTLS can be used as valid and reliable measures of tightness-looseness perception in a Chinese population.

Attention-based message passing and dynamic graph convolution for spatiotemporal data imputation.

Although numerous spatiotemporal approaches have been presented to address the problem of missing spatiotemporal data, there are still limitations in concurrently capturing the underlying spatiotemporal dependence of spatiotemporal graph data. Furthermore, most imputation methods miss the hidden dynamic connection associations that exist between graph nodes over time. To address the aforementioned spatiotemporal data imputation challenge, we present an attention-based message passing and dynamic graph convolution network (ADGCN). Specifically, this paper uses attention mechanisms to unify temporal and spatial continuity and aggregate node neighbor information in multiple directions. Furthermore, a dynamic graph convolution module is designed to capture constantly changing spatial correlations in sensors utilizing a new dynamic graph generation method with gating to transmit node information. Extensive imputation tests in the air quality and traffic flow domains were carried out on four real missing data sets. Experiments show that the ADGCN outperforms the state-of-the-art baseline.

Battery-Type-Behavior-Retention Ni(OH)2-rGO Composite for an Ultrahigh-Specific-Capacity Asymmetric Electrochemical Capacitor Electrode.

Nanosized battery-type materials applied in electrochemical capacitors can effectively reduce a series of problems caused by low conductivity and large volume changes. However, this approach will lead to the charging and discharging process being dominated by capacitive behavior, resulting in a serious decline in the specific capacity of the material. By controlling the material particles to an appropriate size and a suitable number of nanosheet layers, the battery-type behavior can be retained to maintain a large capacity. Here, Ni(OH)2, which is a typical battery-type material, is grown on the surface of reduced graphene oxide to prepare a composite electrode. By controlling the dosage of the nickel source, the composite material with an appropriate Ni(OH)2 nanosheet size and a suitable number of layers was prepared. The high-capacity electrode material was obtained by retaining the battery-type behavior. The prepared electrode had a specific capacity of 397.22 mA h g-1 at 2 A g-1. After the current density was increased to 20 A g-1, the retention rate was as high as 84%. The prepared asymmetric electrochemical capacitor had an energy density of 30.91 W h kg-1 at a power density of 1319.86 W kg-1 and the retention rate could reach 79% after 20,000 cycles. We advocate an optimization strategy that retains the battery-type behavior of electrode materials by increasing the size of nanosheets and the number of layers, which can significantly improve the energy density while combining the advantage of the high rate capability of the electrochemical capacitor.

Strategy of Voltage Match on the Maximum Power Point for a High-Efficiency Photorechargeable Device.

A photorechargeable device can generate power from sunlight and store it in one device, which has a broad application prospect in the future. However, if the working state of the photovoltaic part in the photorechargeable device deviates from the maximum power point, its actual power conversion efficiency will reduce. The strategy of voltage match on the maximum power point is reported to achieve a high overall efficiency (ηoa) of the photorechargeable device assembled by a passivated emitter and rear cell (PERC) solar cell and Ni-based asymmetric capacitors. According to matching the voltage of the maximum power point of the photovoltaic part, the charging characteristics of the energy storage part are adjusted to realize a high actual power conversion efficiency of the photovoltaic part (ηpv). The ηpv of a Ni(OH)2-rGO-based photorechargeable device is 21.53%, and the ηoa is up to 14.55%. This strategy can promote further practical application for the development of photorechargeable devices.

The eco-efficiency evaluation in China's cement industry: A city-level study.

To implement strict environmental targets in China's cement industry into small regions, one should evaluate the city-level eco-efficiency that provides comprehensive instruction. This study establishes a plant-level database with 4000+ production lines located in 341 cities, calculates the energy consumption and CO2, SO2, NOx, and PM emissions, evaluates the eco-efficiency in each city via Slacks-based Measure, and verifies the spatial features of these indicators. Results show that the energy consumption and emissions of the industry are highly concentrated, with ~10 % of the land area contributing to 28.4 %-34.6 % of the total amounts in 2019. The average eco-efficiency value of the clinker calcination and cement grinding processes are 0.761 and 0.714, but the city clusters having low eco-efficiency values are inconsistent with the ones having large energy consumption and emission amounts. The results can contribute to the implementation of the targets such as carbon peaking and pollution cap in China's cement industry.

Pristine GaFeO3 Photoanodes with Surface Charge Transfer Efficiency of Almost Unity at 1.23 V for Photoelectrochemical Water Splitting.

Oxide-based photoelectrodes commonly generate deep trap states associated with various intrinsic defects such as vacancies, antisites, and dislocations, limiting their photoelectrochemical properties. Herein, it is reported that rhombohedral GaFeO3 (GFO) thin-film photoanodes exhibit defect-inactive features, which manifest themselves by negligible trap-states-associated charge recombination losses during photoelectrochemical water splitting. Unlike conventional defect-tolerant semiconductors, the origin of the defect-inactivity in GFO is the strongly preferred antisite formation, suppressing the generation of other defects that act as deep traps. In addition, defect-inactive GFO films possess really appropriate oxygen vacancy concentration for the oxygen evolution reaction (OER). As a result, the as-prepared GFO films achieve the surface charge transfer efficiency (ηsurface ) of 95.1% for photoelectrochemical water splitting at 1.23 V versus RHE without any further modification, which is the highest ηsurface reported of any pristine inorganic photoanodes. The onset potential toward the OER remarkably coincides with the flat band potential of 0.43 V versus RHE. This work not only demonstrates a new benchmark for the surface charge transfer yields of pristine metal oxides for solar water splitting but also enriches the arguments for defect tolerance and highlights the importance of rational tuning of oxygen vacancies.

Suicidal ideation in advanced cancer patients without major depressive disorder.

INTRODUCTION: Major depressive disorder (MDD) is associated with an increased risk of suicide and suicide attempt among cancer patients. However, we do not know how many cancer patients without MDD have suicidal ideation (SI). OBJECTIVES: This study aimed to investigate the prevalence, characteristics and correlated factors of SI among advanced cancer patients without MDD. METHODS: This is a multi-center, cross-sectional study based on an electronic patient-reported outcome systems in patients who were diagnosed with advanced lung, liver, gastric, esophageal, colorectal or breast cancer, the top six prevalent cancers in China. A total of 2930 advanced cancer patients were recruited from 10 regional representative cancer centers across China from August 2019 to December 2020. Patients completed the Patient Health Questionnaire-9 regarding if they had thoughts of being better off dead or of hurting themselves in some way in the previous 2 weeks. Patients also completed the symptom inventory and quality of life assessment. Generalized estimating equation model was performed to explore the correlated factors associated with SI among the patients without MDD. RESULTS: The overall prevalence of SI among advanced cancer patients without MDD was 13.1%. The prevalence was higher in older patients. After adjusted for existing conditions, patients with vomiting symptom (p < 0.001), poorer life quality (p < 0.001), and middle education level (p = 0.031) were correlated factors of SI. CONCLUSIONS: The suicidal ideation is common in advanced cancer patients without MDD. Patients with vomiting, poor quality of life, and middle education level should be screened and monitored for suicidal ideation even without MDD. CLINICAL TRIAL INFORMATION: ChiCTR1900024957.

Linaclotide for treating patients with irritable bowel syndrome with predominant constipation: a multicentre study of real-world data in China.

Introduction: Linaclotide, a guanylate cyclase C agonist that improves the symptoms of irritable bowel syndrome with predominant constipation (IBS-C), has been recently approved for IBS-C treatment. This study aimed to report real-world data on linaclotide treatment in China. Methods: This was a prospective multicentre study of the effectiveness of linaclotide treatment in patients with IBS-C from 10 primary medical institutions. Changes in defecation, abdominal symptoms, the IBS symptom severity scale (IBS-SSS), IBS quality of life questionnaire (IBS-QOL), Zung Self-Rating Anxiety Scale and Self-Rating Depression Scale in patients were evaluated to determine the drug's clinical efficacy and safety. Results: We enrolled 97 patients (mean age: 52.39 ± 13.99 years), 55 of whom were women (56.7%). In terms of efficacy, the number of the patients' defecation per week and Bristol stool form scale scores significantly increased at week 4 and week 12 compared with the values at the baseline. The baseline average IBS-SSS score was 211.01 ± 81.23. Of the patients, 24 had severe IBS-C, and their IBS-SSS scores at week 4 (51.81 ± 54.42) and week 12 (9.3 ± 30.39) significantly decreased and showed a pronounced improvement. The IBS-QOL total scores at week 4 and week 12 gradually decreased compared with that at the baseline and the QOL significantly improved. Treatment satisfaction rate was 79.3% in week 4 and 100% in week 12, showing a gradually increased satisfaction and significant differences. However, 11 cases (11.3%) had diarrhoea. Conclusion: Linaclotide has proved to be a safe and effective drug to improve IBS-C symptoms and severity.

Glutathione-sensitive IPI-549 nanoparticles synergized with photodynamic Chlorin e6 for the treatment of breast cancer.

We combined phosphoinositol-3-kinin inhibitor IPI-549 and photodynamic Chlorin e6 (Ce6) on carboxymethyl chitosan to develop a novel drug delivery nanoparticle (NP) system (Ce6/CMCS-DSP-IPI549) and evaluate its glutathione (GSH) sensitivity and targeting ability for breast cancer treatment. The NPs were spherical with a uniform size of 218.8 nm, a stable structure over 7 days. The maximum encapsulation efficiency was 64.42%, and NPs drug loading was 8.05%. The NPs released drugs within tumor cells due to their high GSH concentration, while they maintained structural integrity in normal cells, which have low GSH concentration. The cumulative release rates of IPI-549 and Ce6 at 108 h were 70.67% and 40.35% (at GSH 10 mM) and 8.11% and 2.71% (at GSH 2µM), respectively. The NPs showed a strong inhibitory effect on 4T1 cells yet did not affect human umbilical vein endothelial cells (HUVECs). After irradiation by a 660 nm infrared laser for 72 h, the survival rate of 4T1 cells was 15.51%. Cellular uptake studies indicated that the NPs could accurately release drugs into tumor cells. In addition, the NPs had a good photodynamic effect and promoted the release of reactive oxygen species to damage tumor cells. Overall, the combination therapy of IPI-549 and Ce6 is safe and effective, and may provide a new avenue for the treatment of breast cancer.