Search

Understanding “efficiency roll-off” (i.e., the drop in emission efficiency with increasing current) is critical if efficient and bright emissive technologies are to be rationally designed. Emerging light-emitting electrochemical cells (LECs) can be cost- and energy-efficiently fabricated by ambient-air printing by virtue of the in situ formation of a p-n junction doping structure. However, this in situ doping transformation renders a meaningful efficiency analysis challenging. Herein, a method for separation and quantification of major LEC loss factors, notably the outcoupling efficiency and exciton quenching, is presented. Specifically, the position of the emissive p-n junction in common singlet-exciton emitting LECs is measured to shift markedly with increasing current, and the influence of this shift on the outcoupling efficiency is quantified. It is further verified that the LEC-characteristic high electrochemical-doping concentration renders singlet-polaron quenching (SPQ) significant already at low drive current density, but also that SPQ increases super-linearly with increasing current, because of increasing polaron density in the p-n junction region. This results in that SPQ dominates singlet-singlet quenching for relevant current densities, and significantly contributes to the efficiency roll-off. This method for deciphering the LEC efficiency roll-off can contribute to a rational realization of all-printed LEC devices that are efficient at highluminance.

Reinforcement learning from human feedback (RLHF) is the mainstream paradigm used to align large language models (LLMs) with human preferences. Yet existing RLHF heavily relies on accurate and informative reward models, which are vulnerable and sensitive to noise from various sources, e.g. human labeling errors, making the pipeline fragile. In this work, we improve the effectiveness of the reward model by introducing a penalty term on the reward, named as \textit{contrastive rewards}. %Contrastive rewards Our approach involves two steps: (1) an offline sampling step to obtain responses to prompts that serve as baseline calculation and (2) a contrastive reward calculated using the baseline responses and used in the Proximal Policy Optimization (PPO) step. We show that contrastive rewards enable the LLM to penalize reward uncertainty, improve robustness, encourage improvement over baselines, calibrate according to task difficulty, and reduce variance in PPO. We show empirically contrastive rewards can improve RLHF substantially, evaluated by both GPTs and humans, and our method consistently outperforms strong baselines.

We introduce Adversarial Policy Optimization (AdvPO), a novel solution to the pervasive issue of reward over-optimization in Reinforcement Learning from Human Feedback (RLHF) for Large Language Models (LLMs). Over-optimization occurs when a reward model serves as an imperfect proxy for human preference, and RL-driven policy optimization erroneously exploits reward inaccuracies. In this paper, we begin by introducing a lightweight way to quantify uncertainties in rewards, relying solely on the last layer embeddings of the reward model, without the need for computationally expensive reward ensembles. AdvPO then addresses a distributionally robust optimization problem centred around the confidence interval of the reward model's predictions for policy improvement. Through comprehensive experiments on the Anthropic HH and TL;DR summarization datasets, we illustrate the efficacy of AdvPO in mitigating the overoptimization issue, consequently resulting in enhanced performance as evaluated through human-assisted evaluation.

Despite showing increasingly human-like abilities, large language models (LLMs) often struggle with factual inaccuracies, i.e. "hallucinations", even when they hold relevant knowledge. To address these hallucinations, current approaches typically necessitate high-quality human factuality annotations. In this work, we explore Self-Alignment for Factuality, where we leverage the self-evaluation capability of an LLM to provide training signals that steer the model towards factuality. Specifically, we incorporate Self-Eval, a self-evaluation component, to prompt an LLM to validate the factuality of its own generated responses solely based on its internal knowledge. Additionally, we design Self-Knowledge Tuning (SK-Tuning) to augment the LLM's self-evaluation ability by improving the model's confidence estimation and calibration. We then utilize these self-annotated responses to fine-tune the model via Direct Preference Optimization algorithm. We show that the proposed self-alignment approach substantially enhances factual accuracy over Llama family models across three key knowledge-intensive tasks on TruthfulQA and BioGEN., Comment: 19 pages

Aligning large language models (LLMs) with human values is a vital task for LLM practitioners. Current alignment techniques have several limitations: (1) requiring a large amount of annotated data; (2) demanding heavy human involvement; (3) lacking a systematic mechanism to continuously improve. In this work, we study aligning LLMs to a new domain with limited samples (e.g. < 100). We propose an algorithm that can self-align LLMs iteratively without active human involvement. Unlike existing works, our algorithm relies on neither human-crafted instructions nor labeled rewards, significantly reducing human involvement. In addition, our algorithm can self-improve the alignment continuously. The key idea is to first retrieve high-quality samples related to the target domain and use them as In-context Learning examples to generate more samples. Then we use the self-generated samples to finetune the LLM iteratively. We show that our method can unlock the LLMs' self-generalization ability to perform alignment with near-zero human supervision. We test our algorithm on three benchmarks in safety, truthfulness, and instruction-following, and show good performance in alignment, domain adaptability, and scalability.

Recommendation systems aim to predict users' feedback on items not exposed to them yet. Confounding bias arises due to the presence of unmeasured variables (e.g., the socio-economic status of a user) that can affect both a user's exposure and feedback. Existing methods either (1) make untenable assumptions about these unmeasured variables or (2) directly infer latent confounders from users' exposure. However, they cannot guarantee the identification of counterfactual feedback, which can lead to biased predictions. In this work, we propose a novel method, i.e., identifiable deconfounder (iDCF), which leverages a set of proxy variables (e.g., observed user features) to resolve the aforementioned non-identification issue. The proposed iDCF is a general deconfounded recommendation framework that applies proximal causal inference to infer the unmeasured confounders and identify the counterfactual feedback with theoretical guarantees. Extensive experiments on various real-world and synthetic datasets verify the proposed method's effectiveness and robustness. © 2023 ACM.

Food is the basic material basis for human survival. Sudden food safety event risks mainly derive from accidental or natural food safety risks, poor food storage environments, and inefficient government regulation policies. The factor identification of sudden food safety risks is the key to controlling such risks. Therefore, the efficient and scientific identification of risk sources and types will be very important in managing sudden food safety risks. In this study, 16 sudden food safety event risk factors were identified through a literature review, and their interactive relationships were clarified using an interpretive structural model (ISM). Then, the weights of influencing factors were calculated through the analytic hierarchy process (AHP), and the combined weight of indices was determined. Results show that the 16 sudden food safety event risk factors can be divided into four levels. The quality standard for food safety (S5) and food storage (S14) is at the bottom layer of risks of sudden food safety events (the first-layer index weight is 36.899%). The judgment matrices at the four levels passed the consistency check. The influence weight of the factor "whether it contains transgenic raw materials" (S9) ranks second (the total weight is 18.151%). This index system for sudden food safety event risk factors is highly effective, with good operability for managing sudden food safety event risks. The obtained conclusions are important reference values for identifying the factors influencing food safety risk management, determining the emphasis of food safety supervision, realizing food risk prevention and control, and strengthening and guaranteeing the food safety level.

PurposeThe goal of this study was to propose a knowledge-based planning system which could automatically design plans for lung cancer patients treated with intensity-modulated radiotherapy (IMRT).Methods and materialsFrom May 2018 to June 2020, 612 IMRT treatment plans of lung cancer patients were retrospectively selected to construct a planning database. Knowledge-based planning (KBP) architecture named & alpha;Diar was proposed in this study. It consisted of two parts separated by a firewall. One was the in-hospital workstation, and the other was the search engine in the cloud. Based on our previous study, A-Net in the in-hospital workstation was used to generate predicted virtual dose images. A search engine including a three-dimensional convolutional neural network (3D CNN) was constructed to derive the feature vectors of dose images. By comparing the similarity of the features between virtual dose images and the clinical dose images in the database, the most similar feature was found. The optimization parameters (OPs) of the treatment plan corresponding to the most similar feature were assigned to the new plan, and the design of a new treatment plan was automatically completed. After & alpha;Diar was developed, we performed two studies. The first retrospective study was conducted to validate whether this architecture was qualified for clinical practice and involved 96 patients. The second comparative study was performed to investigate whether & alpha;Diar could assist dosimetrists in improving the quality of planning for the patients. Two dosimetrists were involved and designed plans for only one trial with and without & alpha;Diar; 26 patients were involved in this study.ResultsThe first study showed that about 54% (52/96) of the automatically generated plans would achieve the dosimetric constraints of the Radiation Therapy Oncology Group (RTOG) and about 93% (89/96) of the automatically generated plans would achieve the dosimetric constraints of the National

Off-policy learning, referring to the procedure of policy optimization with access only to logged feedback data, has shown importance in various real-world applications, such as search engines, recommender systems, and etc. While the ground-truth logging policy, which generates the logged data, is usually unknown, previous work simply takes its estimated value in off-policy learning, ignoring both high bias and high variance resulted from such an estimator, especially on samples with small and inaccurately estimated logging probabilities. In this work, we explicitly model the uncertainty in the estimated logging policy and propose a Uncertainty-aware Inverse Propensity Score estimator (UIPS) for improved off-policy learning, with a theoretical convergence guarantee. Experiment results on synthetic and three real-world recommendation datasets demonstrate the advantageous sample efficiency of the proposed UIPS estimator against an extensive list of state-of-the-art baselines.

Accuracy and diversity have long been considered to be two conflicting goals for recommendations. We point out, however, that as the diversity is typically measured by certain pre-selected item attributes, e.g., category as the most popularly employed one, improved diversity can be achieved without sacrificing recommendation accuracy, as long as the diversification respects the user's preference about the pre-selected attributes. This calls for a fine-grained understanding of a user's preferences over items, where one needs to recognize the user's choice is driven by the quality of the item itself, or the pre-selected attributes of the item. In this work, we focus on diversity defined on item categories. We propose a general diversification framework agnostic to the choice of recommendation algorithms. Our solution disentangles the learnt user representation in the recommendation module into category-independent and category-dependent components to differentiate a user's preference over items from two orthogonal perspectives. Experimental results on three benchmark datasets and online A/B test demonstrate the effectiveness of our solution in improving both recommendation accuracy and diversity. In-depth analysis suggests that the improvement is due to our improved modeling of users' categorical preferences and refined ranking within item categories., Comment: Accepted to WSDM2023

Building end-to-end task bots and maintaining their integration with new functionalities using minimal human efforts is a long-standing challenge in dialog research. Recently large language models (LLMs) have demonstrated exceptional proficiency in conversational engagement and adherence to instructions across various downstream tasks. In this work, we introduce SGP-TOD, Schema-Guided Prompting for building Task-Oriented Dialog systems effortlessly based on LLMs. Utilizing the symbolic knowledge -- task schema, we instruct fixed LLMs to generate appropriate responses on novel tasks, circumventing the need for training data. Specifically, SGP-TOD comprises three components: a LLM for engaging with users, a DST Prompter to aid the LLM with dialog state tracking, which is then used to retrieve database items, and a Policy Prompter to elicit proper responses adhering to the provided dialog policy. Experimental results on Multiwoz, RADDLE and STAR datasets show that our training-free strategy SGP-TOD, without any task-specific data, yields state-of-the-art (SOTA) zero-shot performance, greatly surpasses the few-shot approaches. In a domain-extension setting, SGP-TOD aptly adapts to new functionalities by merely adding supplementary schema rules. We make our code and data publicly available., Comment: 21 pages, 13 figures

Global cropland expansion and associated greenhouse gas (GHG) emissions are largely driven by the growing demand of animal-sourced food. Thus, options are urgently needed to enhancing feed supply at a low environmental cost. Here we explored a set of scenarios about improving domestic feed supply, using a linear optimization model. Results indicate that the total feed (energy and protein) production may be increased by 18–32% through optimizing the spatial distribution of feed crops across provinces, without additional cropland input. GHG emissions and nitrogen and phosphorus inputs per MJ of produced feed decreased by 18–25%, 18–23%, and 16–21%, respectively. The redistribution strategies provide 10–16% additional animal food, which covers the demand of 195–393 million people. Lastly, we provide suggestions for improving the effectiveness of governmental policies related to spatial planning of crops, so as to alleviate the food-feed competition for cropland and to improve environmental sustainability.

Feeding animals more low-opportunity-cost feed products (LCFs), such as food waste and by-products, may decrease food–feed competition for cropland. Using a feed allocation optimization model that considers the availability of feed sources and animal requirements for protein and energy, we explored the perspectives of feeding more LCFs to animals in China. We found that about one-third of the animal feed consisted of human-edible products, while only 23% of the available LCFs were used as feed during 2009–2013. An increased utilization of LCFs (45–90 Mt) could potentially save 25–32% of feed-producing cropland area without impairing livestock productivity. Parallelly, about one-third of feed-related irrigation water, synthetic fertilizer and greenhouse gas emissions would be saved. Re-allocating the saved cropland could sustain the food energy demand of 30–185 million people. Achieving the potentials of increased LCF use requires improved technology and coordination among stakeholders.

Making moral judgments is an essential step toward developing ethical AI systems. Prevalent approaches are mostly implemented in a bottom-up manner, which uses a large set of annotated data to train models based on crowd-sourced opinions about morality. These approaches have been criticized for potentially overgeneralizing a limited group of annotators' moral stances and lacking explainability. In contrast, top-down approaches make moral judgments grounded in a set of principles. However, it remains conceptual due to the incapability of previous language models and the unsolved debate among moral principles. In this study, we propose a flexible framework to steer Large Language Models (LLMs) to perform moral reasoning with well-established moral theories from interdisciplinary research. The theory-guided top-down framework can incorporate various moral theories. Our experiments demonstrate the effectiveness of the proposed framework on datasets derived from moral theories. Furthermore, we show the alignment between different moral theories and existing morality datasets. Our analysis exhibits the potentials and flaws in existing resources (models and datasets) in developing explainable moral judgment-making systems., Comment: 7 pages

Ensuring alignment, which refers to making models behave in accordance with human intentions [1,2], has become a critical task before deploying large language models (LLMs) in real-world applications. For instance, OpenAI devoted six months to iteratively aligning GPT-4 before its release [3]. However, a major challenge faced by practitioners is the lack of clear guidance on evaluating whether LLM outputs align with social norms, values, and regulations. This obstacle hinders systematic iteration and deployment of LLMs. To address this issue, this paper presents a comprehensive survey of key dimensions that are crucial to consider when assessing LLM trustworthiness. The survey covers seven major categories of LLM trustworthiness: reliability, safety, fairness, resistance to misuse, explainability and reasoning, adherence to social norms, and robustness. Each major category is further divided into several sub-categories, resulting in a total of 29 sub-categories. Additionally, a subset of 8 sub-categories is selected for further investigation, where corresponding measurement studies are designed and conducted on several widely-used LLMs. The measurement results indicate that, in general, more aligned models tend to perform better in terms of overall trustworthiness. However, the effectiveness of alignment varies across the different trustworthiness categories considered. This highlights the importance of conducting more fine-grained analyses, testing, and making continuous improvements on LLM alignment. By shedding light on these key dimensions of LLM trustworthiness, this paper aims to provide valuable insights and guidance to practitioners in the field. Understanding and addressing these concerns will be crucial in achieving reliable and ethically sound deployment of LLMs in various applications., Comment: Fixed several typos

Social network sites (SNSs) allow young people to experiment with and present different aspects of themselves during important periods of self-concept development. Interestingly, whether SNSs have negative or positive effects on self-concept clarity (SCC) is inconclusive. We propose that SNS use may simultaneously produce negative and positive effects on SCC, depending on how people use it and the social connection quality created on-line. Specifically, the suppressing mediation model reveals that the direct effect of SNS use intensity on SCC is negative, whereas the indirect effects via perceived social support and self-esteem are positive, suggesting these variables may suppress the negative effect of SNS use on SCC. Our framework helps to explain how SNS contexts influence identity development in young people.

The position of the emission zone (EZ) in the active material of a light-emitting electrochemical cell (LEC) has a profound influence on its performance because of microcavity effects and doping- and electrode-induced quenching. Previous attempts of EZ control have focused on the two principal constituents in the active material-the organic semiconductor (OSC) and the mobile ions-but this study demonstrates that it is possible to effectively control the EZ position through the inclusion of an appropriate additive into the active material. More specifically, it is shown that a mere modification of the end group on an added neutral compound, which also functions as an ion transporter, results in a shifted EZ from close to the anode to the center of the active material, which translates into a 60% improvement of the power efficiency. This particular finding is rationalized by a lowering of the effective electron mobility of the OSC through specific additive: OSC interactions, but the more important generic conclusion is that it is possible to control the EZ position, and thereby the LEC performance, by the straightforward inclusion of an easily tuned additive in the active material.

[Abstract:] Generating reasonable heterogeneous aquifer structures is essential for understanding the physicochemical processes controlling groundwater flow and solute transport better. The inversion process of aquifer structure identification is usually time-consuming. This study develops an integrated inversion framework, which combines the geological single-sample generative adversarial network (GeoSinGAN), the deep octave convolution dense residual network (DOCRN), and the iterative local updating ensemble smoother (ILUES), named GeoSinGAN-DOCRN-ILUES, for more efficiently generating heterogeneous aquifer structures. The performance of the integrated framework is illustrated by two synthetic contaminant experiments. We show that GeoSinGAN can generate heterogeneous aquifer structures with geostatistical characteristics similar to those of the training sample, while its training time is at least 10 times faster than that of typical approaches (e.g., multi-sample-based GAN). The octave convolution layer and multi-residual connection enable the DOCRN to map the heterogeneity structures to the state variable fields (e.g., hydraulic head, concentration distributions) while reducing the computational cost. The results show that the integrated inversion framework of GeoSinGAN and DOCRN can effectively and reasonably generate the heterogeneous aquifer structures.

Deep learning is applied to many complex tasks in the field of wireless communication, such as modulation recognition of spectrum waveforms, because of its convenience and efficiency. This leads to the problem of a malicious third party using a deep learning model to easily recognize the modulation format of the transmitted waveform. Some existing works address this problem directly using the concept of adversarial examples in the image domain without fully considering the characteristics of the waveform transmission in the physical world. Therefore, we propose a low-intercept waveform~(LIW) generation method that can reduce the probability of the modulation being recognized by a third party without affecting the reliable communication of the friendly party. Our LIW exhibits significant low-interception performance even in the physical hardware experiment, decreasing the accuracy of the state of the art model to approximately $15\%$ with small perturbations., Comment: 4 pages, 4 figures, published in 2021 34th General Assembly and Scientific Symposium of the International Union of Radio Science, URSI GASS 2021

End-to-end task bots are typically learned over a static and usually limited-size corpus. However, when deployed in dynamic, changing, and open environments to interact with users, task bots tend to fail when confronted with data that deviate from the training corpus, i.e., out-of-distribution samples. In this paper, we study the problem of automatically adapting task bots to changing environments by learning from human-bot interactions with minimum or zero human annotations. We propose SL-AGENT, a novel self-learning framework for building end-to-end task bots. SL-AGENT consists of a dialog model and a pre-trained reward model to predict the quality of an agent response. It enables task bots to automatically adapt to changing environments by learning from the unlabeled human-bot dialog logs accumulated after deployment via reinforcement learning with the incorporated reward model. Experimental results on four well-studied dialog tasks show the effectiveness of SL-AGENT to automatically adapt to changing environments, using both automatic and human evaluations. We will release code and data for further research.

Pengfei Liu,1,&ast; Quanhui Chen,1,&ast; Fang Yuan,2 Qingru Zhang,1 Xiaoying Zhang,1 Chan Xue,1 Yuqing Wei,1 Yakun Wang,1 Hanqiao Wang1,2 1Department of Sleep Medicine, The Third Hospital of Hebei Medical University, Shijiazhuang, Hebei, People’s Republic of China; 2Department of Physiology, Hebei Key Laboratory of Neurophysiology, Shijiazhuang, Hebei, People’s Republic of China&ast;These authors contributed equally to this workCorrespondence: Hanqiao Wang, Department of Sleep Medicine, The Third Hospital of Hebei Medical University, Shijiazhuang, 050051, Hebei, People’s Republic of China, Tel +86 18533112218, Email wanghanqiao99@126.comPurpose: Mixed events in obstructive sleep apnea (OSA) patients (mixed-OSA) indicate respiratory regulation instability and are essential for OSA pathogenesis and prognosis. It also shows a decreased compliance with continuous positive airway pressure (CPAP). Using predictors to identify mixed-OSA has significant clinical guidance for OSA precise diagnosis and treatment. This study aimed to establish a simple and accessible method for rapid screening of mixed-OSA, thus promoting OSA precise diagnosis.Patients and Methods: A total of 907 patients with suspected OSA were screened, of which 513 OSA patients, including 344 with pure-OSA and 169 with mixed-OSA, were finally included in the study. The clinical characteristics and polysomnography (PSG) parameters of the two OSA groups were compared. Multivariate logistic regression analysis was used to investigate the factors affecting the morbidity of mixed-OSA. The receiver operating characteristic (ROC) curve was used to explore if some convenient PSG parameters can be used to predict mixed-OSA.Results: About 33% of OSA patients were identified as mixed-OSA. Multivariate logistic regression analysis showed that apnea hypopnea index (AHI) and lowest oxygen saturation (LSO2) were independently associated with mixed-OSA after adjusting for age, sex, body mass index (BMI)

Understanding the dynamic behavior of Lightweight Aggregate Self-Compacting Concrete (LWASCC) is of importance to the safety of concrete structures serving in dynamic loading conditions. In this study, the fundamental dynamic properties of LWASCC with three types of LWA were investigated by the impact resonance method. Results show that the dynamic elastic and shear modulus generally decrease with the increase of LWA volume fraction, whereas three types of LWA exert limited influence on dynamic Poisson’s ratio. The dynamic elastic and shear modulus show good linear dependence upon compressive strength. The inclusion of three types of LWA significantly increases the damping ratio, indicating significantly enhanced damping capacity of LWASCC under dynamic loading conditions. The damping ratio of LWASCC is improved by 2.0%, 4.4%, and 2.9% when adding 1% (by volume) expanded clay, rubber, and expanded polystyrene, respectively. The compressive strength and dynamic performances of LWASCC are highly influenced by the intrinsic properties (elastic modulus, damping capacity, wettability, etc.) and geometrical characteristics (size, surface roughness, etc.) of LWA, as well as the LWA-matrix bonding capacity. Copyright © 2021 Ning Li et al.

The growth location and plant variety may influence the active components and biological activities of plants used in phytomedicine. In this study, nine sets of different Epimedii Folium, from different representative cultivation locations and Epimedium species, were collected for comparison, using HPLC-DAD combined with multivariate analysis. The objective was to investigate the influence of geographical origin and Epimedium species on the quality of Epimedii Folium, and provide applicable guidance for cultivation and quality control of Epimedii Folium. Several Epimedium spp. sets were used to establish the HPLC-DAD fingerprints and 91 peaks (compounds) were selected for the multivariate analysis. Major compounds were analyzed by HPLC-DAD combined with principal component analysis (PCA). HPLC quantitative analysis of known bioactive compounds was performed. Application of PCA to HPLC data showed that Epimedium samples sharing the same geographical origin or species clustered together, indicating that both species and geographical origin have impacts on the quality of Epimedii Folium. The major bioactive flavonoid compounds, epimedin C, icariin and baohuoside I, were identified and quantified. The concentration of bioactive compounds was significantly influenced both by species and geographical origin. E. sagittatum from Sichuan showed the highest content of bioactive compounds. The results showed that both Epimedium species and geographical origin have strong impact into quality of Epimedii Folium. HPLC data combined with multivariate analysis is a suitable approach to inform the selection of cultivation areas and choose Epimedium spp. most suitable for different geographical areas, resulting in improved quality of Epimedii Folium.

Yang Li,1 Long Feng,2 Xiaoying Zhang,1 Lianjun Huang,1 Jie Song,1 Guoqing Chen,1 Yu Zhang,3 Chenggang Zhang,4 Weiguang Li,4,5 Zeguo Feng6 1First Medical Center, Chinese PLA General Hospital, Beijing, People’s Republic of China; 2Department of Anesthesia, Hainan Hostipal of Chinese PLA General Hostipal, Beijing, People’s Republic of China; 3Peking Union Medical College Hospital, Beijing, People’s Republic of China; 4Beijing Institute of Radiation Medicine, State Key Laboratory of Proteomics, Cognitive and Mental Health Research Center, Beijing, People’s Republic of China; 5Faculty of Psychology, Beijing Normal University, Beijing, People’s Republic of China; 6Department of Pain Medicine, First Medical Center, Chinese PLA General Hospital, Beijing, People’s Republic of ChinaCorrespondence: Zeguo FengFirst Medical Center, Chinese PLA General Hospital, Department of Pain Medicine, No. 28 Fuxing Road, Haidian District, Beijing, 100853, People’s Republic of ChinaTel/Fax +86 010-66938317Email beijing_301@sina.comWeiguang LiBeijing Institute of Radiation Medicine, State Key Laboratory of Proteomics, Cognitive and Mental Health Research Center, Beijing Normal University, No. 19 Xinjiekouwai St, Haidian District, Beijing, CN, 100875, People’s Republic of ChinaTel +86 010-66930210Email bnubrain.liwg@mail.bnu.edu.cnPurpose: Propofol is commonly used as an intravenous anesthetic in surgical patients. However, its usage is associated with adverse effects. Auxiliary medication can reduce the dose of intravenous anesthetics. Hence, we investigated whether vitamin C could lower propofol dosage in elderly patients undergoing total knee replacement surgery.Patients and Methods: The trial was carried out in PLA General Hospital in Beijing, China. We enrolled patients aged ≥ 50 years who were undergoing unilateral total knee arthroplasty with total intravenous anesthesia combined with lumbar sciatic nerve block. The patients were randomly assigned to either the vitamin

Dialog systems enriched with external knowledge can handle user queries that are outside the scope of the supporting databases/APIs. In this paper, we follow the baseline provided in DSTC9 Track 1 and propose three subsystems, KDEAK, KnowleDgEFactor, and Ens-GPT, which form the pipeline for a task-oriented dialog system capable of accessing unstructured knowledge. Specifically, KDEAK performs knowledge-seeking turn detection by formulating the problem as natural language inference using knowledge from dialogs, databases and FAQs. KnowleDgEFactor accomplishes the knowledge selection task by formulating a factorized knowledge/document retrieval problem with three modules performing domain, entity and knowledge level analyses. Ens-GPT generates a response by first processing multiple knowledge snippets, followed by an ensemble algorithm that decides if the response should be solely derived from a GPT2-XL model, or regenerated in combination with the top-ranking knowledge snippet. Experimental results demonstrate that the proposed pipeline system outperforms the baseline and generates high-quality responses, achieving at least 58.77% improvement on BLEU-4 score.

Bo Huang,1,* Zhichun Feng,2,3,* Lina Zhu,2 Sheng Zhang,2,3 Jun Duan,4 Chaochao Zhao,1 Xiaoying Zhang1,2 1Public Health, Guilin Medical University, Guilin 541100, People’s Republic of China; 2Affiliated BaYi Children’s Hospital, Seventh Medical Center of PLA General Hospital, Beijing 100700, People’s Republic of China; 3Beijing Key Laboratory of Pediatric Organ Failure, Beijing 100700, People’s Republic of China; 4Department of Pediatrics, The First Affiliated Hospital of Anhui Medical University, Hefei 230022, People’s Republic of China*These authors contributed equally to this workCorrespondence: Xiaoying ZhangPublic Health, Guilin Medical University, 1 Zhiyuan Road, Lingui District, Guilin 541100, People’s Republic of ChinaTel/Fax +86773-3680605Email xiaoyingzhang79@163.comPurpose: Mesenchymal stem cells (MSCs) are largely studied for their potential clinical use. Recently, there has been gained further interest in the relationship between MSCs and tumorigenesis. MSCs are reported to both promote and abrogate tumor growth. The present study was designed to investigate whether miRNAs are involved in the interactions between MSCs and tumor cells in the tumor microenvironment.Materials and Methods: Rat bone marrow-derived MSCs (rMSCs) were cultured with or without tumor-conditioned medium (TCM) to observe the effect upon MSCs by TCM. Microarrays and real-time PCR were performed between the two groups. A series of experiments were used to reveal the functional significance of microRNA-503 (miR-503) in rMSCs. Furthermore, the antitumorigenic effect of silencing of miR-503 in rMSCs (miR-503-i-rMSCs) in vivo was measured.Results: We found that rMSCs in vitro exhibited tumor-promoting properties in TCM, and the microRNA profiles of rMSCs were significantly altered in TCM. However, miR-503-i-rMSCs can decrease the angiogenesis and growth of A549 cells. We also demonstrated in an in vivo tumor model that miR-503-i

Due to the unique combination of properties of high intensity and few optical cycle pulse duration, the ultrashort laser pulses have widespread applications in pump-probe spectroscopy and laser-plasma interaction. Correspondingly, it is critical to precisely measure the electric field in the temporal or spectral domain. Since there are no shorter pulses available to sample them in time, it is not easy to characterize few-cycle pulses. In this thesis, the temporal characterization methods including second harmonic generation (SHG) interferometric autocorrelation, single-shot SHG frequency-resolved optical gating (FROG) and chirp scan are used to measure the laser pulses at different positions in our laser system, the Light Wave Synthesizer-20. The SHG interferometric autocorrelation measured a 25.3 fs pulse duration after the kHz frontend laser with a 24.5 fs Fourier limit (FL). The FROG is tested by measuring seed pulses with a 6.5 fs FL providing 6.7 fs pulse duration with mostly flat phase. Then, this FROG apparatus is used to characterize the amplified pulses with a 4.5 fs FL after the whole laser system. The measured pulse duration is 4.5 fs reaching the FL. Similarly, the chirp scan is also used to measure the amplified pluses with a slightly different spectral phase, which retrieved a longer pulse duration of 5.2 fs. The second part of the thesis is focused on contrast improvement. The temporal intensity contrast is reduced in the amplification process leading to a deteriorated laser-plasma interaction. Contrast improvement based on the nonlinear elliptical polarization rotation (NER) technique in a hollow-core fiber (HCF) is implemented and its optimization is performed by using high extinction ratio polarizer, utilizing Ar gas and testing smaller polarization ellipticity (NER angle). The optimal condition is found to be 7° NER angle and 650 mbar Ar in the HCF. Under this optimal condition, sub-4 fs pulses with a smooth spectrum are generated with a power abo

The intense few-cycle laser pulses play an important role in the investigations of laser-plasma interaction. However, one of the biggest challenges in their generation is the reduction of temporal intensity contrast by introducing undesired pre-pulses and a long pedestal. Two techniques were investigated in this work to improve the contrast. First, the crossed-polarized wave (XPW) generation was optimized to get clean pulse. The conditions for XPW were optimized including crystal thickness and maximal background pressure in the vacuum cell. Second, the method of elliptical polarization rotation (EPR) in a gas-filled hollow-core fiber (HCF) was implemented to produce both broadened and cleaned pulse, since its setup is much simpler. For the tested EPR-based nonlinear filter, the spectral smoothening and broadening were obtained. The contrast of cleaned pulse was characterized providing 2 order of magnitude contrast enhancement, while it had a high average power of 80 mW. The EPR-based nonlinear filter is a promising simplified technique in the development of intense few-cycle lasers.

Temporal filtering and spectral broadening are simultaneously achieved, allowing the compression of 20 fs laser pulses down to sub-4 fs duration through the method of nonlinear elliptical polarization rotation in an argon filled hollow-core fiber. The sub-4 fs source provides ~ 35-µJ energy with an internal efficiency >50%, which is more than from the most commonly used pulse-cleaning methods. Further, the contrast is improved by 3 orders of magnitude when measured after amplifying the pulses to 16 TW using non-collinear optical parametric chirped pulse amplification with a prospect to even further enhancement.

Jian Wang,1,* Yuchao Sheng,2,3,* Jun Ni,4 Jianyu Zhu,5 Zhigang Zhou,5 Tianwei Liu,5 Xiaoying Zhang,6 Qi Zhao2,3 1Department of General Practice, Zhongshan Hospital, Fudan University, Shanghai, People’s Republic of China; 2Department of Epidemiology, School of Public Health, Fudan University, Shanghai, People’s Republic of China; 3The Key Laboratory of Public Health Safety of Ministry of Education, Shanghai, People’s Republic of China; 4Vaccination Clinics, Gangyan Community Health Service Center, Shanghai, People’s Republic of China; 5Vaccination Clinics, Tangqiao Community Health Service Center, Shanghai, People’s Republic of China; 6Vaccination Clinics, Huacao Community Health Service Center, Shanghai, People’s Republic of China*These authors contributed equally to this workCorrespondence: Qi ZhaoDepartment of Epidemiology, School of Public Health, Fudan University, 130 Dongan Road, Xuhui District, Shanghai 200032, People’s Republic of ChinaTel/fax +86 21 5423 7335Email zhaoqi@shmu.edu.cnBackground: Parents play an important role in the management of their children’s illnesses, including in decisions concerning how and whether to use antibiotics. This study aims to assess the knowledge and attitude towards antibiotics usage among young parents from different areas in Shanghai and to identify contributing factors of antibiotics usage that are comprehensible for parents.Methods: A cross-sectional study was conducted at vaccination clinics of three community health centers from April to June 2017. Primary guardians who took children for vaccination under the national expanded immunization program were invited to participate.Results: Of the 1368 young parents interviewed, 78.4% of them have taken their children for medical treatment, and 68.9% of their children have taken antibiotics. Parents with education level of senior high school (OR=1.850, 95% CI: 1.214–2.818), or universit

Contextual bandit algorithms provide principled online learning solutions to balance the exploitation-exploration trade-off in various applications such as recommender systems. However, the learning speed of the traditional contextual bandit algorithms is often slow due to the need for extensive exploration. This poses a critical issue in applications like recommender systems, since users may need to provide feedbacks on a lot of uninterested items. To accelerate the learning speed, we generalize contextual bandit to conversational contextual bandit. Conversational contextual bandit leverages not only behavioral feedbacks on arms (e.g., articles in news recommendation), but also occasional conversational feedbacks on key-terms from the user. Here, a key-term can relate to a subset of arms, for example, a category of articles in news recommendation. We then design the Conversational UCB algorithm (ConUCB) to address two challenges in conversational contextual bandit: (1) which key-terms to select to conduct conversation, (2) how to leverage conversational feedbacks to accelerate the speed of bandit learning. We theoretically prove that ConUCB can achieve a smaller regret upper bound than the traditional contextual bandit algorithm LinUCB, which implies a faster learning speed. Experiments on synthetic data, as well as real datasets from Yelp and Toutiao, demonstrate the efficacy of the ConUCB algorithm., Comment: 11 pages; Accepted by WWW2020

Lijie Zhao,1 Yuhua Zheng,1 Yong Ji,2 Xiaoying Zhang3 1Department of Gynecology, Maternal and Child Health Hospital of Foshan, Foshan 528000, Guangdong, China; 2Department of Surgery, The First People’s Hospital of Foshan, Foshan 528000, Guangdong, China; 3Department of Gynecology, Jiashi County People’s Hospital of Kashi Region, Kashi, Xinjiang 844000, China Background: The oncogenic potential of special AT-rich binding protein 1 (SATB1) has been reported in various types of cancer, but its function in cervical cancer remains not fully investigated. This study aimed to investigate the effect of SATB1 mRNA expression on tumor progression and outcomes in the cervical cancer patients. Methods: A total of 33 cervical cancer patients treated in our hospital from September 2012 to December 2015 were included. The mRNA expression level of STAB1 in cervical cancer tissue was determined by real-time PCR, and the patients were divided into dichotomous groups based on their SATB1 expression level. Clinical characteristics, recurrence, and survival outcomes were compared between groups. Results: Compared with the SATB1-low group, the SATB1-high group had significantly advanced International Federation of Gynecology and Obstetrics (FIGO) stages (P=0.037) and histologic grade (P=0.036). Kaplan–Meier analysis showed that SATB1-high group had a worse overall survival (P=0.078, marginal significant). In the subgroup analysis of pathological types, adenocarcinomas group (n=8) had a significantly higher SATB1 expression level as compared with the squamous cell carcinomas (n=18) and adenosquamous carcinomas (n=7) groups (both P<0.05). Cervical squamous cell carcinomas patients with a high-expression SATB1 (n=8) had more advanced FIGO stages (P=0.015) and histologic grades (P=0.060, marginal significant) as well as a higher (P=0.069, marginal significant) incidence of lymphatic metastasis than those with a low expression of SATB1 (n=10). Conclusion

In this letter, a novel synchronization method for MIMO-OFDM systems is proposed. The new approach has an accurate estimate of both symbol timing and large frequency offest. Simulation results show the excellent robustness of our method in double-selective channel even if the strongest multipath component arrives behind the first path., Comment: 2 pages letter with 4 figures

Yang Liu,* Bin Wang,* Yongxiang Qian, Dongmei Di, Min Wang, Xiaoying Zhang Department of Cardiothoracic Surgery, The Third Affiliated Hospital of Soochow University, Changzhou 213003, China *These authors contributed equally to this work Abstract: Cauda equine syndrome (CES) is a neurological condition caused by compression of the cauda equine. Here, we demonstrate a case of CES as the primary symptom of leptomeningeal metastases from non-small cell lung carcinoma without brain metastases. A 59-year-old male suffered progressive lower extremity motor dysfunction, urinary dysfunction, and lower extremity sensory dysfunction. He was clinically diagnosed with CES. Nuclear magnetic resonance imaging demonstrated several vague nodules in the area of conus medullaris and cauda equine, without lumbar or thoracic herniated discs. The serum carcinoembryonic antigen concentration was 191.20 ng/mL. The conclusion following positron emission tomography–computed tomography was a right upper lung malignant tumor with mediastinal lymph node metastasis and cauda equina metastasis. Pathologic diagnosis was of primary adenocarcinoma of the lung by bronchoscopic biopsy. EML4–ALK fusion and EGFR mutations were absent, and thus the patient received chemotherapy. However, symptoms of intracranial hypertension arose 1 month later, and the patient died 3 months postadmission. Emerging CES may be a sign of metastasis of a malignant tumor, presenting an extremely challenging condition, especially for patients with lung cancer. Positron emission tomography–computed tomography is a fairly effective technique to make the diagnosis. Keywords: cauda equine syndrome, leptomeningeal carcinomatosis, positron emission tomography–computed tomography, lung cancer

Proliferation of ubiquitous mobile devices makes location based services prevalent. Mobile users are able to volunteer as providers of specific services and in the meanwhile to search these services. For example, drivers may be interested in tracking available nearby users who are willing to help with motor repair or are willing to provide travel directions or first aid. With the diffusion of mobile users, it is necessary to provide scalable means of enabling such users to connect with other nearby users so that they can help each other with specific services. Motivated by these observations, we design and implement a general location based system HelPal for mobile users to provide and enjoy instant service, which is called mobile crowd service. In this demo, we introduce a mobile crowd service system featured with several novel techniques. We sketch the system architecture and illustrate scenarios via several cases. Demonstration shows the user-friendly search interface for users to conveniently find skilled and qualified nearby service providers., Comment: 4 pages

Spleen T-lymphocytes, especially CD4(+) T-cells, have been demonstrated to be involved in broad immunomodulation and host-defense activity in vivo. Apolipoprotein M gene (apoM) may have an important role in the regulation of immunoprocess and inflammation, which could be hypothesized to the apoM containing sphingosine-1-phosphate (S1P). In the present study we demonstrate that the splenic CD4(+) T-lymphocytes were obviously decreased in the apoM gene deficient (apoM(-/-)) mice compared to the wild type (apoM(+/+)). Moreover, these mice were treated with lipopolysaccharide (LPS) and it was found that even more pronounced decreasing CD4(+) T-lymphocytes occurred in the spleen compared to the apoM(+/+) mice. The similar phenomena were found in the ratio of CD4(+)/CD8(+) T-lymphocytes. After administration of LPS, the hepatic mRNA levels of tumor necrosis factor-alpha (TNF-alpha) and monocyte chemotactic protein-1 (MCP-1) were markedly increased; however, there were no statistical differences observed between apoM(+/+) mice and apoM(-/-) mice. The present study demonstrated that apoM might facilitate the maintenance of CD4(+) T-lymphocytes or could modify the T-lymphocytes subgroups in murine spleen, which may further explore the importance of apoM in the regulation of the host immunomodulation, although the detailed mechanism needs continuing investigation.

It has been demonstrated that apolipoprotein M (APOM) is a vasculoprotective constituent of high density lipoprotein (HDL), which could be related to the anti-atherosclerotic property of HDL. Investigation of regulation of APOM expression is of important for further exploring its pathophysiological function in vivo. Our previous studies indicated that expression of APOM could be regulated by platelet activating factor (PAF), transforming growth factors (TGF), insulin-like growth factor (IGF), leptin, hyperglycemia and etc., in vivo and/or in vitro. In the present study, we demonstrated that palmitic acid could significantly inhibit APOM gene expression in HepG2 cells. Further study indicated neither PI-3 kinase (PI3K) inhibitor LY294002 nor protein kinase C (PKC) inhibitor GFX could abolish palmitic acid induced down-regulation of APOM expression. In contrast, the peroxisome proliferator-activated receptor beta/delta (PPARβ/δ) antagonist GSK3787 could totally reverse the palmitic acid-induced down-regulation of APOM expression, which clearly demonstrates that down-regulation of APOM expression induced by palmitic acid is mediated via the PPARβ/δ pathway.

Apolipoprotein M (APOM) has been suggested as a vasculoprotective constituent of high density lipoprotein (HDL), which plays a crucial role behind the mechanism of HDL-mediated anti-atherosclerosis. Previous studies demonstrated that insulin resistance could associate with decreased APOM expressions. In agreement with our previous reports, here, we further confirmed that the insulin sensitivity was also reduced in rats treated with high concentrations of glucose; such effect could be reversed by administration of rosiglitazone, a peroxisome proliferator-activated receptor-γ (PPARγ). The present study shows that Apom expression is significantly affected by either rosiglitazone or hyperglycemia alone without cross interaction with each other, which indicates that the pathway of Apom expression regulating by hyperglycemia might be differed from that by rosiglitazone. Further study indicated that hyperglycemia could significantly inhibit mRNA levels of Lxrb (P=0.0002), small heterodimer partner 1 (Shp1) (P<0.0001), liver receptor homologue-1 (Lrh1) (P=0.0012), ATP-binding cassette transporter 1 (Abca1) (P=0.0012) and Pparb/d (P=0.0043). Two-way ANOVA analysis demonstrated that the interactions between rosiglitazone and infusion of 25% glucose solution on Shp1 (P=0.0054) and Abca1 (4E, P=0.0004) mRNA expression was statistically significant. It is concluded that rosiglitazone could increase Apom expression, of which the detailed mechanism needs to be further investigated. The downregulation of Apom by hyperglycemia might be mainly through decreasing expression of Pparg and followed by inhibiting Lxrb in rats.

Apolipoprotein M (ApoM) is a constituent of high-density lipoproteins (HDL). It plays a crucial role in HDL-mediated reverse cholesterol transport. Insulin resistance is associated with decreased ApoM levels.

The present study investigated the correlation among genetic polymorphisms of the proximal promoter region of apolipoprotein M (apoM) gene, the polymorphisms in relation to apoM expressions and the susceptibility to coronary artery diseases (CAD) in a Han Chinese population. Four common polymorphic sites, i.e., T-1628G, C-1065A, T-855C and T-778C, were confirmed, and a new deletion mutation C-724del was found, in 206 CAD patients and 209 non-CAD patients using direct DNA sequencing analyses. Occurrences of alleles T-1628G, T-855C and C-724del were significantly higher in CAD patients compared to non-CAD patients. Moreover we examined all these polymorphisms in relation to apoM expression by applying luciferase reporter assay. It demonstrated that constructs -855C and 724del showed obvious decreased luciferase activities, i.e., (0.93±0.15 vs. 2.11±0.15; P=0.012) and (1.13±0.25 vs. 2.11±0.15; P=0.009) respectively, which indicates these two polymorphisms could confer decreased apoM expressions. Meanwhile the occurrences of these two SNP were also significantly higher in the CAD patients than in non-CAD patients. It is therefore reasonable to speculate that down-regulated apoM expressions in relation to these polymorphisms may affect HDL and cholesterol metabolism in vivo and further influence the susceptibility to CAD, although the underlying mechanisms need further investigation.

Jiahui Zhao, Qingli Cheng, Xiaoying Zhang, Meihua Li, Sheng Liu, Xiaodan WangDepartment of Geriatric Nephrology, Chinese PLA General Hospital, Beijing, ChinaObjectives: Percutaneous transluminal renal angioplasty with stent implantation (PTRAS) has become the treatment of choice for atherosclerotic renal artery stenosis (ARAS). This study evaluates the long-term effects of PTRAS on hypertension and renal function in elderly patients with ARAS.Methods: We conducted a retrospective cohort study of all patients who underwent PTRAS in the geriatric division of a tertiary medical center during the period 2003–2010. The clinical data were extracted from the medical records of each patient. Changes in blood pressure, antihypertensive treatment, and estimated glomerular filtration rate were analyzed before and after PTRAS.Results: Eighty-six stents in 81 elderly patients were placed successfully. The average age of the patients was 76.2 years (65–89 years). Mean follow-up was 31.3 months (range 12 –49 months). There was a significant decrease in both systolic and diastolic blood pressure at the third day after the PTRAS procedure and the reduction in blood pressure was constant throughout the follow-up period until 36 months after PTRAS. However, there was no marked benefit to renal function outcome during the follow-up period. The incidence of contrast-induced nephropathy was 9.9% in this study group. The rate of renal artery restenosis was 14.8%. The survival rate was 96.3% for 4 years after the procedure.Conclusion: It is beneficial to control blood pressure in elderly patients with ARAS up to 36 months after a PTRAS procedure. However, their renal function improvement is limited.Keywords: angioplasty, hypertension, renal function, elderly, renal artery stenosis

Background: There is an inverse secular trend between the incidence of obesity and gastric colonization with Helicobacter pylori, a bacterium that can affect the secretion of gastric hormones that relate to energy homeostasis. H. pylori strains that carry the cag pathogenicity island (PAI) interact more intimately with gastric epithelial cells and trigger more extensive host responses than cag− strains. We hypothesized that gastric colonization with H. pylori strains differing in cag PAI status exert distinct effects on metabolic and inflammatory phenotypes. Methodology/Principal Findings: To test this hypothesis, we examined metabolic and inflammatory markers in db/db mice and mice with diet-induced obesity experimentally infected with isogenic forms of H. pylori strain 26695: the cag PAI wild-type and its cag PAI mutant strain 99–305. H. pylori colonization decreased fasting blood glucose levels, increased levels of leptin, improved glucose tolerance, and suppressed weight gain. A response found in both wild-type and mutant H. pylori strain-infected mice included decreased white adipose tissue macrophages (ATM) and increased adipose tissue regulatory T cells (Treg) cells. Gene expression analyses demonstrated upregulation of gastric PPAR γ-responsive genes (i.e., CD36 and FABP4) in H. pylori-infected mice. The loss of PPAR γ in immune and epithelial cells in mice impaired the ability of H. pylori to favorably modulate glucose homeostasis and ATM infiltration during high fat feeding. Conclusions/Significance: Gastric infection with some commensal strains of H. pylori ameliorates glucose homeostasis in mice through a PPAR γ-dependent mechanism and modulates macrophage and Treg cell infiltration into the abdominal white adipose tissue.

BACKGROUND: G-protein-coupled receptor kinase 2 (GRK2) is a primary regulator of β-adrenergic signaling in the heart. G-protein-coupled receptor kinase 2 ablation impedes heart failure development, but elucidation of the cellular mechanisms has not been achieved, and such elucidation is the aim of this study. METHODS AND RESULTS: Myocyte contractility, Ca(2+) handling and excitation-contraction coupling were studied in isolated cardiomyocytes from wild-type and GRK2 knockout (GRK2KO) mice without (sham) or with myocardial infarction (MI). In cardiac myocytes isolated from unstressed wild-type and GRK2KO hearts, myocyte contractions and Ca(2+) transients were similar, but GRK2KO myocytes had lower sarcoplasmic reticulum (SR) Ca(2+) content because of increased sodium-Ca(2+) exchanger activity and inhibited SR Ca(2+) ATPase by local protein kinase A-mediated activation of phosphodiesterase 4 resulting in hypophosphorylated phospholamban. This Ca(2+) handling phenotype is explained by a higher fractional SR Ca(2+) release induced by increased L-type Ca(2+) channel currents. After β-adrenergic stimulation, GRK2KO myocytes revealed significant increases in contractility and Ca(2+) transients, which were not mediated through cardiac L-type Ca(2+) channels but through an increased SR Ca(2+). Interestingly, post-MI GRK2KO mice showed better cardiac function than post-MI control mice, which is explained by an improved Ca(2+) handling phenotype. The SR Ca(2+) content was better maintained in post-MI GRK2KO myocytes than in post-MI control myocytes because of better-maintained L-type Ca(2+) channel current density and no increase in sodium-Ca(2+) exchanger in GRK2KO myocytes. An L-type Ca(2+) channel blocker, verapamil, reversed some beneficial effects of GRK2KO. CONCLUSIONS: These data argue for novel differential regulation of L-type Ca(2+) channel currents and SR load by GRK2. G-protein-coupled receptor kinase 2 ablation represents a novel beneficial Ca(2+) handling phenoty

We have previously reported that liver X receptor (LXR) agonist, TO901317, could significantly inhibit hepatic apolipoprotein M (apoM) expression. It has been reported that TO901317 could activate the ATP-binding cassette transporter A1 (ABCA1) that mediates cholesterol efflux to the lipid-poor apoA1, which is an essential step for the high-density lipoprotein (HDL) formation. It is unknown if ABCA1 may regulate hepatic apoM expression. In the present study, HepG2 cells were cultured with the synthetic LXR agonists, TO901317 or GW3965 in the presence or absence of ABCA1 antagonist, disodium 4,4'-diisothiocyanatostilbene-2,2'-disulfonate (DIDS). The mRNA levels of ABCA1, apoM and liver receptor homolog-1 (LRH-1) determined by the real-time RT-PCR. It demonstrated that both TO901317 and GW3965 could significantly enhance ABCA1 expression, and simultaneously, inhibit LRH1 expression. However, TO901317 alone could significantly inhibit apoM expression, while GW3965 alone did not influence apoM expression. ABCA1 antagonist, DIDS, have no effects on GW3965 induced upregulation of ABCA1 and downregulation of LRH1. However, apoM mRNA level was significantly decreased when the cells cultured with GW3965 together with DIDS. The present study demonstrated that apoM expression could be elevated by ABCA1 via the RXR/LXR pathway and LRH1 does not involve in the regulation of apoM by the activation of ABCA1, although the direct regulative pathway(s) between ABCA1 and apoM gene is still unknown yet. The detailed mechanism needs further investigation. (C) 2012 Elsevier Inc. All rights reserved.

OBJECTIVE: To investigate effects of lipid lowering drug, simvastatin, on apolipoprotein M expression in the hyperlipidemic mice and in hepatic cell line, HepG2 cells. METHODS: Swiss male mice were randomly divided into the high fat group and control group, and were intragastrically fed with 0.9% saline (control group) or lipid emulsion (high fat group) at the daily dosage of 15 ml/kg body weight, respectively. After 8 weeks feeding, the hyperlipidemic model was successfully induced and these hyperlipidemic mice were then randomly divided into three experimental groups: vehicle control group, high-dose simvastatin-treated group (100 mg/kg body weight), and low-dose simvastatin-treated group (10 mg/kg body weight). Mice were dosed daily for 6 weeks of simvastatin before mice were sacrificed for determining serum lipid profile and apoM protein levels that was determined by using dot blotting analysis. Effects of simvastatin on apoM mRNA expression in the HepG2 cells were determined by real-time RT-PCR. RESULTS: Comparing to high fat model mice without simvastatin treatment, 100 mg/kg simvastatin could significantly increase serum total cholesterol (P < 0.05). Serum apoM levels, in all mice, were significantly lower in the mice at the age of 26 weeks than the mice at 12 weeks old (P < 0.05), which indicated that serum apoM levels were significantly correlated to the mice age. It demonstrated also that treatment of simvastatin did not influence serum apoM levels in these mouse model, although serum apoM levels were increased by about 13% in the 10 mg/kg simvastatin group than in the vehicle control group without simvastatin. In HepG2 cell cultures, simvastatin could significantly decrease apoM mRNA levels with dose- and time-dependent manners. At 10 μM simvastatin treatment, apoM mRNA decreased by 52% compared to the controls. CONCLUSION: The present study suggested that simvastatin, in vivo, had no effect on apoM levels in the hyperlipidemic mouse model. ApoM serum lev

Background: Apolipoprotein M (apoM) may have potential antiatherosclerotic properties. It has been reported that apoM expression could be regulated by many intracellar and extracellar factors. In the present study we further investigated regulation of apoM expression in Caco-2 cells stimulated by a liver X receptor (LXR) agonist, TO901317. Materials and methods: Caco-2 cells were cultured in the presence of either TO901317, farnesoid X receptor (FXR) antagonist guggulsterone or TO901317 together with guggulsterone at different concentrations for 24 hrs. The mRNA levels of ATP-binding cassette transporter A1 (ABCA1), apoA1, apoM, liver receptor homologue-1 (LRH-1) and short heterodimer partner 1 (SHP1) were determined by real-time RT-PCR. Results: When Caco-2 cell cultured with TO901317 alone, the mRNA levels of ABCA1, apoA1, apoM, LRH-1 and SHP1 were significantly increased with dose-dependent manners (p < 0.05), whereas when the cells cultured with guggulsterone alone, the mRNA levels of apoM, SHP1 and LRH-1 (p < 0.05) were strongly inhibited. Moreover, guggulsterone could abolish the TO901317 enhanced mRNA levels of apoA1 apoM, SHP1 and LRH-1. Conclusion: The present study demonstrated that LXR agonist TO901317 induced apoM expression in Caco-2 cells might be mediated via the LXR/FXR pathway.

The present study examined mRNA levels and protein mass of apolipoprotein M (apoM) in human hepatocellular carcinoma (HCC) tissues and in the adjacent tissues. Plasma apoM levels in these HCC patients were also determined and compared to the normal subjects. The mean level of plasma apoM in the HCC patients was 0.61 +/- 0.30 OD mm(-2), which was significantly higher than that in the normal subjects 0.37 +/- 0.07 OD mm(-2) (P < 0.01). However, both apoM mRNA levels and apoM protein mass in the HCC tissues were significantly lower than in the adjacent tissues (P < 0.05). It is concluded that human hepatocellular carcinoma tissues had a reduced capacity to produce apoM than the adjacent non-tumor tissues. However, the plasma apoM levels were higher in the HCC patients than in normal subjects, which suggested that tissues adjacent to the tumors or extra-hepatic apoM production in the HCC patients may contribute to the higher plasma apoM levels in these patients. The clinical significance of apoM in relation to HCC still needs further investigation. (C) 2009 Published by Elsevier GmbH.

Apolipoprotein M (apoM) is present predominantly in high-density lipoprotein (HDL) in human plasma, thus possibly involved in the regulation of HDL metabolism and the process of atherosclerosis. Although estrogen replacement therapy increases serum levels of apoAl and HDL, it does not seem to reduce the cardiovascular risk in postmenopausal women. Therefore, we investigated the effects of estrogen on apoM expression in vitro and in vivo. HepG2 cells were incubated with different concentrations of estrogen with or without the estrogen receptor antagonist, fulvestrant, and apoM expression in the cells was determined. Hepatic apoM expression and serum levels of apoM were also determined in normal and in ovariectomized rats treated with either placebo or estradiol benzoate, using sham operated rats as controls. Estrogen significantly increased mRNA levels of apoM and apoAl in HepG2 cell cultures in a dose- and time-dependent manner; the upregulation of both apolipoproteins was fully abolished by addition of estrogen receptor antagonist In normal rats, estrogen treatment led to an increase in plasma lipid levels including HDL cholesterol, a marked upregulation of apoM mRNA and a significant increase in serum levels of apoM. The same pattern of regulation was found in ovariectomized rats treated with estrogen. Thus, estrogen upregulates apoM expression both in vivo and in vitro by mechanism(s) involving the estrogen receptor. (C) 2011 Elsevier B.V. All rights reserved.