The Kruppel-like factor 5 (KLF5) transcription factor is highly expressed in high-grade and basal-like breast cancers. However, the mechanism by which KLF5 promotes cell migration and invasion is still not completely understood. In this study, we demonstrate that TNFAIP2, a tumor necrosis factor-α (TNFα)-induced gene, is a direct KLF5 target gene. The expression of TNFAIP2 is highly correlated with the expression of KLF5 in breast cancers. The manipulation of KLF5 expression positively alters TNFAIP2 expression levels. KLF5 directly binds to the TNFAIP2 gene promoter and activates its transcription. Functionally, KLF5 promotes cancer cell proliferation, migration and invasion in part through TNFAIP2. TNFAIP2 interacts with the two small GTPases Rac1 and Cdc42, thereby increasing their activities to change actin cytoskeleton and cell morphology. These findings collectively suggest that TNFAIP2 is a direct KLF5 target gene, and both KLF5 and TNFAIP2 promote breast cancer cell proliferation, migration and invasion through Rac1 and Cdc42.

Expression quantitative trait locus (eQTL) analysis, which links variations in gene expression to genotypes, is essential to understanding gene regulation and to interpreting disease-associated loci. Currently identified eQTLs are mainly in samples of blood and other normal tissues. However, no database comprehensively provides eQTLs in large number of cancer samples. Using the genotype and expression data of 9196 tumor samples in 33 cancer types from The Cancer Genome Atlas (TCGA), we identified 5 606 570 eQTL-gene pairs in the cis-eQTL analysis and 231 210 eQTL-gene pairs in the trans-eQTL analysis. We further performed survival analysis and identified 22 212 eQTLs associated with patient overall survival. Furthermore, we linked the eQTLs to genome-wide association studies (GWAS) data and identified 337 131 eQTLs that overlap with existing GWAS loci. We developed PancanQTL, a user-friendly database (http://bioinfo.life.hust.edu.cn/PancanQTL/), to store cis-eQTLs, trans-eQTLs, survival-associated eQTLs and GWAS-related eQTLs to enable searching, browsing and downloading. PancanQTL could help the research community understand the effects of inherited variants in tumorigenesis and development.

A macrophage exosome coated nanoplatform for targeted chemotherapy of triple-negative breast cancer.

Hypoxia exerts a profound impact on diverse aspects of cancer biology. Increasing evidence has revealed novel functions of hypoxia in cancer cell epigenomics, epitranscriptomics, metabolism, and intercellular communication, all hotspots of cancer research. Several drugs have been developed to target intratumoral hypoxia and have entered clinical trials to treat refractory tumors. However, direct targeting of hypoxia signaling still has limitations in the clinic with regard to cancer progression and resistance to therapy. Comprehensive understanding of the molecular mechanisms by which hypoxia reshapes tumors and their microenvironment, as well as how tumor cells adapt to and thrive in hypoxic conditions, will therefore continue to be a focus of cancer research and will provide new directions for hypoxic tumor treatment.

No abstract

a b s t r a c tSince its outbreak in December 2019, a series of clinical trials on Coronavirus Disease 2019 have been registered or carried out. However, the significant heterogeneity and less critical outcomes of such trials may be leading to a waste of research resources. This study aimed to develop a core outcome set (COS) for clinical trials on COVID-19 in order to tackle the outcome issues. The study was conducted according to the Core Outcome Measures in Effectiveness Trials (COMET) handbook (version 1.0), a guideline for COS development. A research group was set up that included experts in respiratory and critical medicine, traditional Chinese medicine, evidence-based medicine, clinical pharmacology, and statistics, in addition to medical journal editors. Clinical trial registry websites (chictr.org.cn and clinicaltrials.gov) were searched to retrieve clinical trial protocols and outcomes in order to form an outcome pool. A total of 78 clinical trial protocols on COVID-19 were included and 259 outcomes were collected. After standardization, 132 outcomes were identified within seven different categories, of which 58 were selected to develop a preliminary outcome list for further consensus. After two rounds of Delphi survey and one consensus meeting, the most important outcomes for the different clinical classifications of COVID-19 were identified and determined to constitute the COS for clinical trials on COVID-19 (COS-COVID). The COS-COVID includes one outcome for the mild type (time to 2019-nCoV reverse transcriptionpolymerase chain reaction (RT-PCR) negativity), four outcomes for the ordinary type (length of hospital

Immunotherapy is recognized as one of the most promising approaches to treat cancers. However, its effect in glioblastoma (GBM) treatment is insufficient, which can in part be attributed to the immunosuppressive tumor microenvironment (TME). Microglia and macrophages are the main immune infiltrating cells in the TME of GBM. Unfortunately, instead of initiating the anti‐tumor response, GBM‐infiltrating microglia and macrophages switch to a tumor‐promoting phenotype (M2), and support tumor growth, angiogenesis, and immunosuppression by the release of cytokines. In this work, a virus‐mimicking membrane‐coated nucleic acid nanogel Vir‐Gel embedded with therapeutic miRNA is developed, which can reprogram microglia and macrophages from a pro‐invasive M2 phenotype to an anti‐tumor M1 phenotype. By mimicking the virus infection process, Vir‐Gel significantly enhances the targetability and cell uptake efficiency of the miR155‐bearing nucleic acid nanogel. In vivo evaluations demonstrate that Vir‐Gel apparently prolongs the circulation lifetime of miR155 and endows it with an active tumor‐targeting capability and excellent tumor inhibition efficacy. Owing to its noninvasive feature and effective delivery capability, the virus‐mimicking nucleic acid nanogel provides a general and convenient platform that can successfully treat a wide range of diseases.

We investigate the following question for machine translation (MT): can we develop a single universal MT model to serve as the common seed and obtain derivative and improved models on arbitrary language pairs? We propose mRASP, an approach to pre-train a universal multilingual neural machine translation model. Our key idea in mRASP is its novel technique of random aligned substitution, which brings words and phrases with similar meanings across multiple languages closer in the representation space. We pre-train a mRASP model on 32 language pairs jointly with only public datasets. The model is then fine-tuned on downstream language pairs to obtain specialized MT models. We carry out extensive experiments on 42 translation directions across a diverse settings, including low, medium, rich resource, and as well as transferring to exotic language pairs. Experimental results demonstrate that mRASP achieves significant performance improvement compared to directly training on those target pairs. It is the first time to verify that multiple lowresource language pairs can be utilized to improve rich resource MT. Surprisingly, mRASP is even able to improve the translation quality on exotic languages that never occur in the pretraining corpus. Code, data, and pre-trained models are available at https://github. com/linzehui/mRASP. * Equal contribution. The work was done when the first author was an intern at ByteDance.